Wind Loading_ A Practical Guide to BS 6399-2 ( PDFDrive ).pdf
RodnyThomson1
5 views
131 slides
Jul 18, 2024
Slide 1 of 131
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
About This Presentation
e
Slide Content
Wind loading: A practical guide to’ BS6399-2
. UK wind code, 856399-2, was issued in 1995 and amended in 1997. Ir
laces CP3 Chapter V Part 2, whieh is now obsolescent and will shortly bo
Indtawn. The introduction of a new code is often traumatie, especially so
this case since CP3 was in place for 25 years. Initial reactions to 8553602
ve been mixed: some welcome the additional flexibility and detail, while
es aim its too complex and produces higher loads."Many of the
Salome stem from misuse of CP3 combined with unfamiliarity with
5399.2. its clear that designers would greatly benefit fram this clear ond
thoritotive quidance on the use of the new code.
63992 offers the user a range of choices trom quick ‘back of envelope
imates through 10 lengthy and detailed cakulations, This book guides
user through these choices, demonstrating the benefits and dovwnsides.
each choice. B56389-2 gives 15% lower loads, on average, with the
tonal for larger reductions, depending on the exposure of the site, Ih
ok helps the designer to release this potential, leading to more efficient
d economic designs, Designers! most common questions and problems aro
dressed along with detailed guidance for estimating wind loads to the
Ww code, This book clearly sets out why each change to the code was
cessary and in what context it will increase of dectease loads, illustrated
realistic worked examples.
nd loading: A practica! guide 16 856399.2 isa ‘must read! for all designers
Il and structural engineers and engineering technicians. It has an essen
«e alongside the British Standard B56399-2 in every design office inte
|
sfessor Nicholas Cook worked for many yeors atthe Bullding Resa
fablishment, making oustanding contributions to the understand.
nd loads on buitdings and other structures. He isthe author of:
signers guide to wind loading of building structures from ae» +
6399 2 and other standards derive He has assisted in drafting or «us
number of wind loading standards, incuding CP3-V-2, BSB100. Eure
A the ISO wind code, He isthe expert member of the drafting var“
6393 2 responsible for preparing 110 1995 draft and the 1997
nendiments,
otessor Cook now divides his time between his St Albans bases: wo
à Brstol University where he is Professor of Industrial Ae!ody
ruducting numerous workshops for designers on the use of 8562")
\ developed a unique might into typical misconceptions and prob
+ implementation of 856339-2 and of its predecessor CP3V:2
L! Thomas Telford.
Wind loading
a practical guide to
Wind loads on buildings
. UK wind code, 856399-2, was issued in 1995 and amended in 1997. Ir
laces CP3 Chapter V Part 2, whieh is now obsolescent and will shortly bo
Indtawn. The introduction of a new code is often traumatie, especially so
this case since CP3 was in place for 25 years. Initial reactions to 8553602
ve been mixed: some welcome the additional flexibility and detail, while
es aim its too complex and produces higher loads."Many of the
Salome stem from misuse of CP3 combined with unfamiliarity with
5399.2. its clear that designers would greatly benefit fram this clear ond
thoritotive quidance on the use of the new code.
63992 offers the user a range of choices trom quick ‘back of envelope
imates through 10 lengthy and detailed cakulations, This book guides
user through these choices, demonstrating the benefits and dovwnsides.
each choice. B56389-2 gives 15% lower loads, on average, with the
tonal for larger reductions, depending on the exposure of the site, Ih
ok helps the designer to release this potential, leading to more efficient
d economic designs, Designers! most common questions and problems aro
dressed along with detailed guidance for estimating wind loads to the
Ww code, This book clearly sets out why each change to the code was
cessary and in what context it will increase of dectease loads, illustrated
realistic worked examples.
nd loading: A practica! guide 16 856399.2 isa ‘must read! for all designers
Il and structural engineers and engineering technicians. It has an essen
«e alongside the British Standard B56399-2 in every design office inte
|
sfessor Nicholas Cook worked for many yeors atthe Bullding Resa
fablishment, making oustanding contributions to the understand.
nd loads on buitdings and other structures. He isthe author of:
signers guide to wind loading of building structures from ae» +
6399 2 and other standards derive He has assisted in drafting or «us
number of wind loading standards, incuding CP3-V-2, BSB100. Eure
A the ISO wind code, He isthe expert member of the drafting var“
6393 2 responsible for preparing 110 1995 draft and the 1997
nendiments,
otessor Cook now divides his time between his St Albans bases: wo
à Brstol University where he is Professor of Industrial Ae!ody
ruducting numerous workshops for designers on the use of 8562")
\ developed a unique might into typical misconceptions and prob
+ implementation of 856339-2 and of its predecessor CP3V:2
L! Thomas Telford.
Wind loading
a practical guide to
Wind loads on buildings
Wind loading
Apractical guide to BS 6399-2
Wind loads on buildings
Nicholas J. Cook, FREng
—
SL! Thomas Telford
Apractical guide to BS 6399-2
Wind loads on buildings
Nicholas J. Cook, FREng
—
SL! Thomas Telford
Published by Thomas Telond Publishing. Thoms Telford Limited, | Heron Quay,
Landon RIS AUD.
URL fe do.
Distibotors for Thomas Telford books ste
SAL ASCE Pres, (801 Alexander Well Drive, Reston, VA 20191-4400
“Tas Maruzen Co Lad, Rook Deport, 310 Nihonbast 2<howe, Coes, Tokyo 105
Aura: DA Books and Joual, 688 Whitehore Road, Mitcham 312, Victoria
Fars published 1999
apres amendrents 2002, epimed 2008
A catalogue seen for i hook ls acabe rom the Bri Library
ISBN: 0.7277 2955 9
‘© Nichols 1. Cook, 1999
Alsi ncluding translation, reserved. Except fr fir copying, no par ef this pubition may
Fe reproduced. stored a à real yet or ränsited in amy for or by any ens ler,
mechanical, phalccopyine ar iris, withow the peier wen pension of the Books
Publisher, ihtomas Telford Publishing, Bones Teod Lid, Heron Quay, London EL4 410,
this book ds Quhfihed on the vederstndin that he auth is solely responsible for Ue
eme dt and opinions expressed in le nd tha its publication does ot nes spl
thor auch semer andor opinions are o let te views or opinion ofthe publisher
‘Typeset by MHL Typesting Limite
11.028 N
‘An eunee of practical advice is
4.48 N
> worth a pound of theory”
Landon RIS AUD.
URL fe do.
Distibotors for Thomas Telford books ste
SAL ASCE Pres, (801 Alexander Well Drive, Reston, VA 20191-4400
“Tas Maruzen Co Lad, Rook Deport, 310 Nihonbast 2<howe, Coes, Tokyo 105
Aura: DA Books and Joual, 688 Whitehore Road, Mitcham 312, Victoria
Fars published 1999
apres amendrents 2002, epimed 2008
A catalogue seen for i hook ls acabe rom the Bri Library
ISBN: 0.7277 2955 9
‘© Nichols 1. Cook, 1999
Alsi ncluding translation, reserved. Except fr fir copying, no par ef this pubition may
Fe reproduced. stored a à real yet or ränsited in amy for or by any ens ler,
mechanical, phalccopyine ar iris, withow the peier wen pension of the Books
Publisher, ihtomas Telford Publishing, Bones Teod Lid, Heron Quay, London EL4 410,
this book ds Quhfihed on the vederstndin that he auth is solely responsible for Ue
eme dt and opinions expressed in le nd tha its publication does ot nes spl
thor auch semer andor opinions are o let te views or opinion ofthe publisher
‘Typeset by MHL Typesting Limite
11.028 N
‘An eunee of practical advice is
4.48 N
> worth a pound of theory”
Foreword
British Standard BS 6399: Part 2, has attracted undeserved adverse
criticism, The usual grounds for complaint have been the increased
‘complexity and higher wind loads as compared with the previous code,
namely CP 3-V-2. In part, a number af these complaints stem from a
misunderstanding of both codes. Wind engineering is not generally taught
in undergraduate courses, and many engineers are therefore not properly
equipped to make judgements and to appreciate the reasoning behind
some of the code clauses. There is a tendency to follow a ‘mechanical’
procedure.
This is the background in which Professor Nicholas Cook has written
this practical guide. Professor Cook is an internationally known figure in
the field of wind engincering. He is a member of the Code (BS 6399:
Part 2) Committee, which I chair, He has carried out considerable
research, written numerous papers and produced design aids for BS 6399:
Part 2. This practical guide is the most comprehensive treatment yet of the
subject
sincerely hope that all users of the code will refer to this guide to gain
proper insight into how the code has attempted to quantify the various
aspects of the inherently complex phenomenon of wind loading. The
organisation of the guide should appeal to the user, with commentary of
the code clauses separated from the tips on how to get the best out of the
code. In addition, the worked examples are intended not only to illustrate
the basic procedures of the code but to extend the expertise of the user
beyond these procedures.
“This practical guide will go a long way in promoting proper under-
standing of wind loading. The guide is also a testimony to the author's
‘genuine interest in wanting to assist practising engineers, and it deserves
every success.
Professor R.S. Narayanan, FREng
British Standard BS 6399: Part 2, has attracted undeserved adverse
criticism, The usual grounds for complaint have been the increased
‘complexity and higher wind loads as compared with the previous code,
namely CP 3-V-2. In part, a number af these complaints stem from a
misunderstanding of both codes. Wind engineering is not generally taught
in undergraduate courses, and many engineers are therefore not properly
equipped to make judgements and to appreciate the reasoning behind
some of the code clauses. There is a tendency to follow a ‘mechanical’
procedure.
This is the background in which Professor Nicholas Cook has written
this practical guide. Professor Cook is an internationally known figure in
the field of wind engincering. He is a member of the Code (BS 6399:
Part 2) Committee, which I chair, He has carried out considerable
research, written numerous papers and produced design aids for BS 6399:
Part 2. This practical guide is the most comprehensive treatment yet of the
subject
sincerely hope that all users of the code will refer to this guide to gain
proper insight into how the code has attempted to quantify the various
aspects of the inherently complex phenomenon of wind loading. The
organisation of the guide should appeal to the user, with commentary of
the code clauses separated from the tips on how to get the best out of the
code. In addition, the worked examples are intended not only to illustrate
the basic procedures of the code but to extend the expertise of the user
beyond these procedures.
“This practical guide will go a long way in promoting proper under-
standing of wind loading. The guide is also a testimony to the author's
‘genuine interest in wanting to assist practising engineers, and it deserves
every success.
Professor R.S. Narayanan, FREng
Contents
1. Introduction
LL The role of this Guide
1.2 The need for change
13 The need for continuity
14 Objectives and principles
1.5 The changes to expect
1.5.1. The information chain
1.5.2 Principal rule changes
1.5.3. Calibrations
3.1 Misinterpretation of CP 3-V-2
32. Design dynamic pressure
3 Cladding loads
1.53.4 Structural loads
1.6. Format and content of this Guide
1.6.1. Format
1.6.1.1. Cross-references
16.1.2
1613
1.6.1.4 Tips and examples
1.6.1.5 “Intermediate” and “advanced” guidance
1.6.2. Content
1:63. 2002 Amendments
of the Stand
ds clauses
2. Commentary
21. Scope, definitions and procedure
2.1.1 Scope ($7.1)
21.2 Definitions ($7.3)
2.1.3. Standard and directional methods ($7.8)
2.14 The four steps
2.1.5 Wind direction (82.4.1, §3.1.1)
2.1.5.1 Role
1. Introduction
LL The role of this Guide
1.2 The need for change
13 The need for continuity
14 Objectives and principles
1.5 The changes to expect
1.5.1. The information chain
1.5.2 Principal rule changes
1.5.3. Calibrations
3.1 Misinterpretation of CP 3-V-2
32. Design dynamic pressure
3 Cladding loads
1.53.4 Structural loads
1.6. Format and content of this Guide
1.6.1. Format
1.6.1.1. Cross-references
16.1.2
1613
1.6.1.4 Tips and examples
1.6.1.5 “Intermediate” and “advanced” guidance
1.6.2. Content
1:63. 2002 Amendments
of the Stand
ds clauses
2. Commentary
21. Scope, definitions and procedure
2.1.1 Scope ($7.1)
21.2 Definitions ($7.3)
2.1.3. Standard and directional methods ($7.8)
2.14 The four steps
2.1.5 Wind direction (82.4.1, §3.1.1)
2.1.5.1 Role
viti | Contents Contents | ix
2.1.5.2 Definitions (§2.1.1.2, §2.2.2.3, $21.11) 20 233 Altitude factor (§2.2.2.2) ”
21:53. Options 20 233.1 Basis 7
2153.1 Option 1 —Irrespective of direction, 2332 Application 30
standard method a 2.3.4 Direction factor (§2.2.2.3) 40
2.1.53.2 Option 2—Orthogonal low cases, standard 234.1 Basis a
method at 2.3.4.2 Application 40
21543 Option 3 —Twelve 30°-wide sectors, 2.3.5 Seasonal factor (§2.2.2.4, Annex D) 2
rectional method 2 2.3.5.1 Basis 42
21534 Hybrid options 2 2.3.5.2 Application 42
2.1.6 Topography ($222.22) 2 factor (§2,2.2.5, Annex D) 43
2.1.6.1. Test for significance (82.2.2.2.1) 2 a
2.1.62 Topographic dimensions (§2.2.2.2.4) 25 23.62 Application 44
2163 ‘Effective’ slope dimensions (82.2.2.2.9) 2 23.7 Site wind speed ($2.22, 93.2.2) 4
21.6.4 The location factor, » ($22.2.2.5) a 237.1 Basis 44
2.164.1 Mills and ridges 2 2372 Application 45
211.642 Cliffs and escarpments 2 2.3.7.3 Overseas sites 45
2.1.0.5. Key topography values (§2.2.2.2.3) 2 24 Site exposure 6
2.1.7 The wind load chain 28 24.1 The exposure model 46
2.1.7.1 Dynamic pressure (82.1.2, 83.1.2) 29 24.2 Ground roughness categories and fetch (81.7.2, Annex E) 48
2172. External, intemal and net surface pressures 242.1 Basis 48
($2.1.3, $3.1.3) 2 2422 Application si
2.1.72.1 External surface pressure 29 24.3 Reference height and effective height ($/.7.3, 83.2.3) 52
2.1722 Internal surface pressure 30 243.1 Basis 2
2.1723 Net surface pressure 30 2432 Application 54
2.1.73 Surface friction (§2.1.3.8, §3.1.3.4) 30 243.3 Comments 55
2174 Face loads and overall loads ($2.7.3,5, $2.1.3.6, 2.44 Division-by-parts (§2.2.3.2, §3.2.3.1) 56
$31.32, 83.133) 31 244.1 Basis 56
2.1.7.5. Component loads (82.1.2.5, §3.1:3.2) a 244.2 Application 58
2.2 Dynamic classification 32 245 Tesrain and building factor ($2.2.3.3, $3.2.3.2) 59
22.1. Purpose and method 22 245.1 Basis 5
222 The signpost procedure (87.6) 3 24.52 Application 60
222.1. Building-type factor, Ky (87.6.1) a 2.453 Comments 63
22241 Basis 3 246 Fifective wind speed and dynamic pressure ($2.1.,
22212 Application 3 $223, 88.12, $323) 63
22.2.2 Dynamic augmentation factor, C (81.6.1, $1.6.2) 34 24.6.1 Basis 63
22221 Basis 34 24.62 Application 64
22222 Application 34 247 Commentary on site exposure 64
2.23 Limits to the dynamic classification M 2.5 Building shape factors 66
2.3 Wind climate 35 2.5.1. The shape factor model 66
23.1. The wind climate model 35 25.1.1 Shape factors (§2.3, $3.3) 66
232 Basic wind speed (82.2.1, §3.2.1, Annex B) 35 25.1.2 Extemal pressure coefficient zones. a
232.1 Basis 35 25.13 Directional and standard external pressure
2
2322 Application 36 coefficients
2.1.5.2 Definitions (§2.1.1.2, §2.2.2.3, $21.11) 20 233 Altitude factor (§2.2.2.2) ”
21:53. Options 20 233.1 Basis 7
2153.1 Option 1 —Irrespective of direction, 2332 Application 30
standard method a 2.3.4 Direction factor (§2.2.2.3) 40
2.1.53.2 Option 2—Orthogonal low cases, standard 234.1 Basis a
method at 2.3.4.2 Application 40
21543 Option 3 —Twelve 30°-wide sectors, 2.3.5 Seasonal factor (§2.2.2.4, Annex D) 2
rectional method 2 2.3.5.1 Basis 42
21534 Hybrid options 2 2.3.5.2 Application 42
2.1.6 Topography ($222.22) 2 factor (§2,2.2.5, Annex D) 43
2.1.6.1. Test for significance (82.2.2.2.1) 2 a
2.1.62 Topographic dimensions (§2.2.2.2.4) 25 23.62 Application 44
2163 ‘Effective’ slope dimensions (82.2.2.2.9) 2 23.7 Site wind speed ($2.22, 93.2.2) 4
21.6.4 The location factor, » ($22.2.2.5) a 237.1 Basis 44
2.164.1 Mills and ridges 2 2372 Application 45
211.642 Cliffs and escarpments 2 2.3.7.3 Overseas sites 45
2.1.0.5. Key topography values (§2.2.2.2.3) 2 24 Site exposure 6
2.1.7 The wind load chain 28 24.1 The exposure model 46
2.1.7.1 Dynamic pressure (82.1.2, 83.1.2) 29 24.2 Ground roughness categories and fetch (81.7.2, Annex E) 48
2172. External, intemal and net surface pressures 242.1 Basis 48
($2.1.3, $3.1.3) 2 2422 Application si
2.1.72.1 External surface pressure 29 24.3 Reference height and effective height ($/.7.3, 83.2.3) 52
2.1722 Internal surface pressure 30 243.1 Basis 2
2.1723 Net surface pressure 30 2432 Application 54
2.1.73 Surface friction (§2.1.3.8, §3.1.3.4) 30 243.3 Comments 55
2174 Face loads and overall loads ($2.7.3,5, $2.1.3.6, 2.44 Division-by-parts (§2.2.3.2, §3.2.3.1) 56
$31.32, 83.133) 31 244.1 Basis 56
2.1.7.5. Component loads (82.1.2.5, §3.1:3.2) a 244.2 Application 58
2.2 Dynamic classification 32 245 Tesrain and building factor ($2.2.3.3, $3.2.3.2) 59
22.1. Purpose and method 22 245.1 Basis 5
222 The signpost procedure (87.6) 3 24.52 Application 60
222.1. Building-type factor, Ky (87.6.1) a 2.453 Comments 63
22241 Basis 3 246 Fifective wind speed and dynamic pressure ($2.1.,
22212 Application 3 $223, 88.12, $323) 63
22.2.2 Dynamic augmentation factor, C (81.6.1, $1.6.2) 34 24.6.1 Basis 63
22221 Basis 34 24.62 Application 64
22222 Application 34 247 Commentary on site exposure 64
2.23 Limits to the dynamic classification M 2.5 Building shape factors 66
2.3 Wind climate 35 2.5.1. The shape factor model 66
23.1. The wind climate model 35 25.1.1 Shape factors (§2.3, $3.3) 66
232 Basic wind speed (82.2.1, §3.2.1, Annex B) 35 25.1.2 Extemal pressure coefficient zones. a
232.1 Basis 35 25.13 Directional and standard external pressure
2
2322 Application 36 coefficients
x | Contents
2513.1 Diretional method
2511.32 Standard method
25133 Range
25.134 Lattice structures
25.2. ‘The scaling length and zones
252.1 Basis
2522 Applic
25221 Scaling length (24.1.3, §2.5.1.2, 9.5.22,
$24.12, $332.22, $333.2.)
Re-entrant comers (82.4.3, $3.3.1.5)
Recessed bays (82.4.3, $3.5.1.6)
Irregular Mush faces ($2.4.4.1, §3.3.1.8)
aser storeys (§2.4.4.2, §3.3.1.8)
‘Smaller extensions
Smallest enclosing rectangle ($3.7.3.3.2)
252.3 Comments
2.53. Extemal pressure coefficients for walls ($2.4, §3.3.1)
253.1. Basis
2.53.11 Directional method
253.12 Standard method
25.1.3 Funnelling
2532 Application
25321 Directional method ($3.3)
253.22 Standard method ($24, §3.3.1.5)
253.23 Non-vertical walls
25324. Polygonal-plan buildings (82.4.2)
25325 Cireular-plan buildings ($2.46)
25326. Special considerations
75327 Friction lands (92.45, $2.3.1.9)
2.54 External pressure coefficients for wots ($2.5, §3.3.2-
334)
254.1 Basis
234.11 Directional method
754.12 Standard method
254.13. Range of data
254.14. Funnelling
2.542 Application to all roofs
2542.1 General rules
25422 Roof overhangs (92.58)
25423 Friction loads ($2.5.10, 93.328, 9.3.9)
2.543 Flat roofs
2543.1 Directional method (93.3.2)
25432 Standard method (§2.5.1)
Contents | ai
2543.3 Polygonal-plan buildings
2,5434 Circular-plan buildings
25435 Special considerations
5.44 Pitched roots
2544.1 Reference height ($3.3.3.2.2)
4.2 Alternative values on windward pitch
43. Directional method (§3.3.3)
444 Standard method ($2.52, 92.5.3)
4.5 Special considerations
2545 Barrel-vault roofs
2546 Multi-bay roots
2.54.6.1 Scope of guidance
254.62 Standard and directional methods
2546.3 Onhogonal case 0°
254.64 Orthogonal case 90°
2,547 Comments und examples
25.5 Internal pressure coefficients ($2.6, $3.3.5)
25.5.1 Basis
255.11 Balance of flow
255.12 Response time and equivalent size effect
25.5.2 Enclosed buildings (12.6.1)
255.21 Porosity of enclosed buildings
255.22 Application
2553 Dominant openings (§2.6.2)
255.31 Definitions
25332 Application
25.54. Open-Sided and open-topped buil
255.41 Definitions
25542 Application
25.6 Free-standing canopies (§2.5.9.1
on
25.62 Blockage
22363 Application
.5.7 Free-standing walls, parapets and signboards ($:
2571 Bas zum
25.72 Application (§2.8.1)
2.5.2.1 Free-standing walls, parapets and dense
fences ($2.41. 1)
25122 Signboards ($2.8.2)
25723 Effective wind speed
2.5.8. Pressure coefficients for elements ($2.7)
258.1 Individual sections
258.11 Basis
89
90
90
9
92
92
92
9%
2513.1 Diretional method
2511.32 Standard method
25133 Range
25.134 Lattice structures
25.2. ‘The scaling length and zones
252.1 Basis
2522 Applic
25221 Scaling length (24.1.3, §2.5.1.2, 9.5.22,
$24.12, $332.22, $333.2.)
Re-entrant comers (82.4.3, $3.3.1.5)
Recessed bays (82.4.3, $3.5.1.6)
Irregular Mush faces ($2.4.4.1, §3.3.1.8)
aser storeys (§2.4.4.2, §3.3.1.8)
‘Smaller extensions
Smallest enclosing rectangle ($3.7.3.3.2)
252.3 Comments
2.53. Extemal pressure coefficients for walls ($2.4, §3.3.1)
253.1. Basis
2.53.11 Directional method
253.12 Standard method
25.1.3 Funnelling
2532 Application
25321 Directional method ($3.3)
253.22 Standard method ($24, §3.3.1.5)
253.23 Non-vertical walls
25324. Polygonal-plan buildings (82.4.2)
25325 Cireular-plan buildings ($2.46)
25326. Special considerations
75327 Friction lands (92.45, $2.3.1.9)
2.54 External pressure coefficients for wots ($2.5, §3.3.2-
334)
254.1 Basis
234.11 Directional method
754.12 Standard method
254.13. Range of data
254.14. Funnelling
2.542 Application to all roofs
2542.1 General rules
25422 Roof overhangs (92.58)
25423 Friction loads ($2.5.10, 93.328, 9.3.9)
2.543 Flat roofs
2543.1 Directional method (93.3.2)
25432 Standard method (§2.5.1)
Contents | ai
2543.3 Polygonal-plan buildings
2,5434 Circular-plan buildings
25435 Special considerations
5.44 Pitched roots
2544.1 Reference height ($3.3.3.2.2)
4.2 Alternative values on windward pitch
43. Directional method (§3.3.3)
444 Standard method ($2.52, 92.5.3)
4.5 Special considerations
2545 Barrel-vault roofs
2546 Multi-bay roots
2.54.6.1 Scope of guidance
254.62 Standard and directional methods
2546.3 Onhogonal case 0°
254.64 Orthogonal case 90°
2,547 Comments und examples
25.5 Internal pressure coefficients ($2.6, $3.3.5)
25.5.1 Basis
255.11 Balance of flow
255.12 Response time and equivalent size effect
25.5.2 Enclosed buildings (12.6.1)
255.21 Porosity of enclosed buildings
255.22 Application
2553 Dominant openings (§2.6.2)
255.31 Definitions
25332 Application
25.54. Open-Sided and open-topped buil
255.41 Definitions
25542 Application
25.6 Free-standing canopies (§2.5.9.1
on
25.62 Blockage
22363 Application
.5.7 Free-standing walls, parapets and signboards ($:
2571 Bas zum
25.72 Application (§2.8.1)
2.5.2.1 Free-standing walls, parapets and dense
fences ($2.41. 1)
25122 Signboards ($2.8.2)
25723 Effective wind speed
2.5.8. Pressure coefficients for elements ($2.7)
258.1 Individual sections
258.11 Basis
89
90
90
9
92
92
92
9%
si | Contents
258.12 Application
2.5.8.2 Latice frames
259 Commentary on shape factors
2.6 Cladding, structural and overall loads
2.6.1 Normal pressures, toads and fri
821.3)
2.6.1.1 Normal pressures (62.1.2.1-3, 93.1.3.)
12 Surface loads ($2.1.3,5, §3.1.3.2)
26.13 Overall lnads (§2.1.3.6, 3.1.3.)
26.14 Friction loads (§2.1.3.8, 83.1.3.)
2.62 Diagonal dimension ($2./.34, $3.2.5.3)
26.2.1 Extemal pressures
26.2.2. Intemal pressures
263. Size effeet factor of standard method (§2.1.3.4)
2,64 Asymmetric loads (§2.1.3.7)
3. Making BS 6399-2 work for you
3.1 Taking control
3.1.1 Servant or master
3.1.2. Precision and accuracy
3.13. Reducing conservatism
3.1.3.1 Standard options
3.132 Hybrid options
3.1.3.3 Other opportunities
14 Design aids
Ma Kap suite fou with BS 63992
jon leads ($21.3,
108
no
120
11
121
121
122
12
123
124
124
126
127
127
129
129
129
130
mi
131
131
132
132
132
3.1.42 Ordnance Survey Interactive Alas of Great Britain 132
3143 This Guide
3.14.5 BREWS le
3.1.4.6 BREVe2 i
ven
3.1.7 Wind tunnel tests vA
33.1. Available options
Contents
33.1.1 Option | —Inespective of direction
33.12 Option 2—Orthogonal load cases.
33.1.3 Option 3— Directional method
33.14. Hybrid methods
33 near the coast, estuaries or inland water
3.3.3. Sites in towns or permanent woodland
333.1 Distance in town
3.3.3.2 Coastal towns
Urban shelter
34 Permanent woodland
3.3.3.5. Hybrid options for sites in towns
3.34 Complex topography
335 Temporary buildings
3.3.6. Sorviceability limit
33.6.1 Active safety
6.2. Passive safely management
3.3.6.3 Non-permanent and direct shelter
34 Pressures and loads
3.4.1 Tributary areas across zones
3:42. Diagonal dimension with load sharing
3.43 Internal pressure and dominant openings
343.1. Permeability and porosity
3432 Response time
3.43.2.1 Adiabatic response
34322 Effect of building flexibility
3.4.3.3 Dominant openings
3.4.3.3.1 How small can a dominant opening be?
4.3.3.2 Multiple openings in one face
344 Cladding loads
3.4.4.1. Comer zones
3.44.2 Roof zones for irregular flush faces and inset
storeys
3.45 Loads on intemal partitions
3.46 Structural loads
3.4.6.1 The overall load equation
3.4.6.2 Asymmetry
3,5. Additional tips and tricks
35.1. Minimizing effort
3.5.1.1 Sites on significant topography
35.12 Sites in towns
3.5.1.3. The ‘unfactored load’ method
35.14 Sloping and skewed surfaces
3.5.2 When the exact site details are unknown
146
147
147
147
148
150
150
151
152
153
153
154
156
157
157
157
158
158
158
159
160
160
162
162
163
164
164
165
166
166
166
167
168
168
169
170
170
170
m
m
172
173
258.12 Application
2.5.8.2 Latice frames
259 Commentary on shape factors
2.6 Cladding, structural and overall loads
2.6.1 Normal pressures, toads and fri
821.3)
2.6.1.1 Normal pressures (62.1.2.1-3, 93.1.3.)
12 Surface loads ($2.1.3,5, §3.1.3.2)
26.13 Overall lnads (§2.1.3.6, 3.1.3.)
26.14 Friction loads (§2.1.3.8, 83.1.3.)
2.62 Diagonal dimension ($2./.34, $3.2.5.3)
26.2.1 Extemal pressures
26.2.2. Intemal pressures
263. Size effeet factor of standard method (§2.1.3.4)
2,64 Asymmetric loads (§2.1.3.7)
3. Making BS 6399-2 work for you
3.1 Taking control
3.1.1 Servant or master
3.1.2. Precision and accuracy
3.13. Reducing conservatism
3.1.3.1 Standard options
3.132 Hybrid options
3.1.3.3 Other opportunities
14 Design aids
Ma Kap suite fou with BS 63992
jon leads ($21.3,
108
no
120
11
121
121
122
12
123
124
124
126
127
127
129
129
129
130
mi
131
131
132
132
132
3.1.42 Ordnance Survey Interactive Alas of Great Britain 132
3143 This Guide
3.14.5 BREWS le
3.1.4.6 BREVe2 i
ven
3.1.7 Wind tunnel tests vA
33.1. Available options
Contents
33.1.1 Option | —Inespective of direction
33.12 Option 2—Orthogonal load cases.
33.1.3 Option 3— Directional method
33.14. Hybrid methods
33 near the coast, estuaries or inland water
3.3.3. Sites in towns or permanent woodland
333.1 Distance in town
3.3.3.2 Coastal towns
Urban shelter
34 Permanent woodland
3.3.3.5. Hybrid options for sites in towns
3.34 Complex topography
335 Temporary buildings
3.3.6. Sorviceability limit
33.6.1 Active safety
6.2. Passive safely management
3.3.6.3 Non-permanent and direct shelter
34 Pressures and loads
3.4.1 Tributary areas across zones
3:42. Diagonal dimension with load sharing
3.43 Internal pressure and dominant openings
343.1. Permeability and porosity
3432 Response time
3.43.2.1 Adiabatic response
34322 Effect of building flexibility
3.4.3.3 Dominant openings
3.4.3.3.1 How small can a dominant opening be?
4.3.3.2 Multiple openings in one face
344 Cladding loads
3.4.4.1. Comer zones
3.44.2 Roof zones for irregular flush faces and inset
storeys
3.45 Loads on intemal partitions
3.46 Structural loads
3.4.6.1 The overall load equation
3.4.6.2 Asymmetry
3,5. Additional tips and tricks
35.1. Minimizing effort
3.5.1.1 Sites on significant topography
35.12 Sites in towns
3.5.1.3. The ‘unfactored load’ method
35.14 Sloping and skewed surfaces
3.5.2 When the exact site details are unknown
146
147
147
147
148
150
150
151
152
153
153
154
156
157
157
157
158
158
158
159
160
160
162
162
163
164
164
165
166
166
166
167
168
168
169
170
170
170
m
m
172
173
| Contents
3.53 When the orientation of the building is unknown
54 Automating the balance of flow ealeulation for
internal pressure
35.5. Minimizing wind loads
3.5.5.1 Changing the external shape
35.5.2 Controlling the internal pressure
4, Worked examples
4.1 Introducing the examples
4.1.1. Role of the examples
412 The site at SKA00877
413 The example buildings
ANA Similarities to BRE Digest 436
42 Fffective wind speed and dynamic pressure
4.2.1 Reference parameters
42.2 Option 1 —Irrespeotive of direction
42,3 Option 2— Orthogonal load cases
42.4 Option 3—Directional method
4.2.5 Comparing options
43 Timber-framed house
4.3.1 The house example
432 Scaling parameters
433 Pressure coetficients
434 Racking forces in timber panels
43.5. Overall forces on roof trusses
44 Long-span portal-frame building
44.1 The poral-frame building example
4:42. Pressure coefficients
443 Internal pressures
444 Highest-loaded purlin and rail
4:45. Loads on portal frames, NW case
44.5.1 Loads on individual members
445.2 Basc shear
44.6. Overall horizontal load, NE case
45. Five-storey office building
45.1 ‘The five-storey office example
45.2 Scaling length, b
45.3 Pressure coefficients for the walls
45.4 External pressures
45.5 Internal pressures
45.5.1 Upper storeys and ground floor, ultimate limit
45.5.2 Ground floor, serviceability limit
va
ns
175
175
176
7
177
17
178
179
179
179
179
180
181
183
185
187
187
187
189
191
195
198
198
199
199
200
201
202
203
205
206
206
208
210
210
212
Contents | xv
4.5.6 Loads on cladding panels
‘6 Sie elle acer
4.5.6.2 Ilighest-loaded panels, ultimate limit
4,563. Highest-loaded panels, serviceability limit
46 Tower and podium
4.6.1 The tower and
Dynamie classification
Scaling length D and division-by-pans
External pressure coefficients
Dynamie pressures
Horizontal shear for
of the tower
y al the base of cach storey
Appendix A. Lattice Structures
ALL Lattice frames and trusses
Basis
Dynamic pressure
Net pressure cocffici
Shielding
Application without shielding
A.LÓ Application with shielding
2.2 Ünelad building frames
A21 Basis
A22 Secondary beams
A23 Application
ents
>>>»
Appendix B, Corrected factor tables
References
Index
213
213
213
214
215
215
215
216
217
219
219
21
221
21
21
22
223
25
26
228
228
229
230
233
237
239
3.53 When the orientation of the building is unknown
54 Automating the balance of flow ealeulation for
internal pressure
35.5. Minimizing wind loads
3.5.5.1 Changing the external shape
35.5.2 Controlling the internal pressure
4, Worked examples
4.1 Introducing the examples
4.1.1. Role of the examples
412 The site at SKA00877
413 The example buildings
ANA Similarities to BRE Digest 436
42 Fffective wind speed and dynamic pressure
4.2.1 Reference parameters
42.2 Option 1 —Irrespeotive of direction
42,3 Option 2— Orthogonal load cases
42.4 Option 3—Directional method
4.2.5 Comparing options
43 Timber-framed house
4.3.1 The house example
432 Scaling parameters
433 Pressure coetficients
434 Racking forces in timber panels
43.5. Overall forces on roof trusses
44 Long-span portal-frame building
44.1 The poral-frame building example
4:42. Pressure coefficients
443 Internal pressures
444 Highest-loaded purlin and rail
4:45. Loads on portal frames, NW case
44.5.1 Loads on individual members
445.2 Basc shear
44.6. Overall horizontal load, NE case
45. Five-storey office building
45.1 ‘The five-storey office example
45.2 Scaling length, b
45.3 Pressure coefficients for the walls
45.4 External pressures
45.5 Internal pressures
45.5.1 Upper storeys and ground floor, ultimate limit
45.5.2 Ground floor, serviceability limit
va
ns
175
175
176
7
177
17
178
179
179
179
179
180
181
183
185
187
187
187
189
191
195
198
198
199
199
200
201
202
203
205
206
206
208
210
210
212
Contents | xv
4.5.6 Loads on cladding panels
‘6 Sie elle acer
4.5.6.2 Ilighest-loaded panels, ultimate limit
4,563. Highest-loaded panels, serviceability limit
46 Tower and podium
4.6.1 The tower and
Dynamie classification
Scaling length D and division-by-pans
External pressure coefficients
Dynamie pressures
Horizontal shear for
of the tower
y al the base of cach storey
Appendix A. Lattice Structures
ALL Lattice frames and trusses
Basis
Dynamic pressure
Net pressure cocffici
Shielding
Application without shielding
A.LÓ Application with shielding
2.2 Ünelad building frames
A21 Basis
A22 Secondary beams
A23 Application
ents
>>>»
Appendix B, Corrected factor tables
References
Index
213
213
213
214
215
215
215
216
217
219
219
21
221
21
21
22
223
25
26
228
228
229
230
233
237
239
Introduction
1.1 The role of this Guide
All British Standards are self-contained and complete documents that
need no extra information to implement —at least that is the theoretical
position. The tendency in practice is for Standards to contain only the
rules and data, without supplementary advice on how to apply these rules
and data in the simplest and most effective manner.
The first role of this Guide is to provide a commentary to BS 6399-2
that will assist new designers to become familiar with the basics of the
new Standard and to allow experienced designers Lo convert from the
previous code CP 3-V-2, This is most valuable to the “basic” user.
‘The second role is to help the user to manage the options of BS 6399-2
in the most efficient way, to highlight potential problems and to expose
pitfalls that must be avoided. This is most valuable to the ‘experienced’
user.
The third sole is to extend the limited scope of the Standard with
additional advice and knowledge and to encourage users to exploit the
new features of BS 6399-2, allowing them to make the most efficient use
of design time and to optimize the structure. This is most valuable to the
“advanced” user.
1.2 The need for change
‘The British Standard BS 6399 Loading for buildings, Part 2 (BS 6399-
2) was published in 1995 as a replacement for Basic data for the design of
huildings, Chapter V, Part 2 Wind loads (CP 3-V-2). CP 3-V-2 was first
published in 1970, revised in 1972, and amended in 1986 and 1990.
The change was prompted by a number of Factors:
+ Twenty-five years is a very long time without a major revision.
® Asa member of the International Standards Organisation (ISO), the
British Standards Institution (BSI) is obliged to follow policy
changes agreed internationally and a number of agrecd changes
affect the format of Standards,
1.1 The role of this Guide
All British Standards are self-contained and complete documents that
need no extra information to implement —at least that is the theoretical
position. The tendency in practice is for Standards to contain only the
rules and data, without supplementary advice on how to apply these rules
and data in the simplest and most effective manner.
The first role of this Guide is to provide a commentary to BS 6399-2
that will assist new designers to become familiar with the basics of the
new Standard and to allow experienced designers Lo convert from the
previous code CP 3-V-2, This is most valuable to the “basic” user.
‘The second role is to help the user to manage the options of BS 6399-2
in the most efficient way, to highlight potential problems and to expose
pitfalls that must be avoided. This is most valuable to the ‘experienced’
user.
The third sole is to extend the limited scope of the Standard with
additional advice and knowledge and to encourage users to exploit the
new features of BS 6399-2, allowing them to make the most efficient use
of design time and to optimize the structure. This is most valuable to the
“advanced” user.
1.2 The need for change
‘The British Standard BS 6399 Loading for buildings, Part 2 (BS 6399-
2) was published in 1995 as a replacement for Basic data for the design of
huildings, Chapter V, Part 2 Wind loads (CP 3-V-2). CP 3-V-2 was first
published in 1970, revised in 1972, and amended in 1986 and 1990.
The change was prompted by a number of Factors:
+ Twenty-five years is a very long time without a major revision.
® Asa member of the International Standards Organisation (ISO), the
British Standards Institution (BSI) is obliged to follow policy
changes agreed internationally and a number of agrecd changes
affect the format of Standards,
2 1 Wind loading: a practical guide to BS 6399-2
13
Both CP 3-V-2 and BS 6399-2 are “head codes’, which n
icy: that all
CP 3-V-2 fails to comply with a particular policy
normative information should appear in the body of a Standard and
hot in annexes or appendices. We will realise the wisdom of this
icy later
Rand and building regulation ao hecoming Tess presen
bringing greater choice to the user.
Ve nes of Siete. now have their own Standards
including:
steel chimneys, BS 4076
Tattice towers and masts, BS 8100
hyperbolic cooling towers, BS 4485-6
slating and tiling, BS 5534.
i about 1960 and
CP 3-V-2 was based on knowledge available in
there has been a vast increase in the range, detail and precision of
knowledge since then. o .
A umber of errors in data values und interpretation had bon
Giscovered, the most important being corrected in the 1986 and
1990 amendments to CP 3-V-2.
‘The development and popularity of new forms and shapes of
‘buildings for which CP 3-V-2 did not give data.
inl
The need for continuity sos hat
other codes and Standards refer to them for information, These other
codes and Standards may:
.
refer to the design wind loads directly, as in BS 5950, Structural
use of steelwork in building; |
had and Otto st the parue ct fon.
as in BS 5534, Code of practice for slating and rilings
replace the methods and data entirely, as in BS 8100. Latice
Towers and masts: Part I, Code of Practice for loading: ot
Five prescriptive methods, based on calculations made using the
Wind loading code, as in BS 8103, Structural design of low-rise
buildings, and in Approved Document A to the Building
Regulations.
e more the users become
“The longer a head code remains in use, the more 1 €
famitiar and experienced in its application. Handbooks and other guidance
notes are published. Design aids are developed. Methods from the
‘Standard are incorporated into more Standards. Investment is made in
18. All these aspects are disrupted if large changes are mado in the
methods or format of a head code.
Introduction | 3
Accordingly, RS 6399-2 was designed to build on the foundations of
CP 3-V-2, incorporating the necessary changes, bul preserving the
existing format as closely as possible.
1.4. Objectives and principles
‘The members of the drafting panel for BS 6399-2 were given five main
objectives:
© to make the new Standard independent of material and structural
form
to reflect current understanding as closely as possible
10 maintain the current average level of design risk, but reduce the
range of uncertainty
to retain the CP 3-V-2 engineering model and format
1o minimize the scope for misinterpretation.
In executing these objectives, I
principles.
anel members adopted the following
1, To work in terms of normal stress (pressure) and shear stress
(fiction) on the building envelope.
This ensures independence of material and structural form. We
derive the lowds on structural elements by accumulating the
stresses that act on relevant areas of the building envelope.
To account for all effects that influence: loading by 10% or more.
Many factors affect wind loads but only the more significant are
included, Partial factors on wind loads are typically about
71 — 1-4, so that omitting more than four factors could lead to
erosion of the safety margin if a conservative approach is not
taken,
3. To keep the influence of effects in balance.
Fach factor should be included at a precision that isin balance
with the other factors. ‘There is litle point in determining one
factor to a precision of 2% if ethers are only good to 20%.
4. To make values change smoothly, where possible.
Step changes in value give large step changes in design loads,
so there is a natural tendency for designers to pick values to the
less onerous side ul u step. The step changes in CP 3-V-2 were
so large that the possible difference in loads between nominally
identical buildings could easily excced the margin of safety it
several steps acted together, This is illustrated by Example 1,
which shows that the ratio between the most and the: least
onerous interpretation for a typical building at a typical UK
‘west-coast town site can be as large as 270%,
13
Both CP 3-V-2 and BS 6399-2 are “head codes’, which n
icy: that all
CP 3-V-2 fails to comply with a particular policy
normative information should appear in the body of a Standard and
hot in annexes or appendices. We will realise the wisdom of this
icy later
Rand and building regulation ao hecoming Tess presen
bringing greater choice to the user.
Ve nes of Siete. now have their own Standards
including:
steel chimneys, BS 4076
Tattice towers and masts, BS 8100
hyperbolic cooling towers, BS 4485-6
slating and tiling, BS 5534.
i about 1960 and
CP 3-V-2 was based on knowledge available in
there has been a vast increase in the range, detail and precision of
knowledge since then. o .
A umber of errors in data values und interpretation had bon
Giscovered, the most important being corrected in the 1986 and
1990 amendments to CP 3-V-2.
‘The development and popularity of new forms and shapes of
‘buildings for which CP 3-V-2 did not give data.
inl
The need for continuity sos hat
other codes and Standards refer to them for information, These other
codes and Standards may:
.
refer to the design wind loads directly, as in BS 5950, Structural
use of steelwork in building; |
had and Otto st the parue ct fon.
as in BS 5534, Code of practice for slating and rilings
replace the methods and data entirely, as in BS 8100. Latice
Towers and masts: Part I, Code of Practice for loading: ot
Five prescriptive methods, based on calculations made using the
Wind loading code, as in BS 8103, Structural design of low-rise
buildings, and in Approved Document A to the Building
Regulations.
e more the users become
“The longer a head code remains in use, the more 1 €
famitiar and experienced in its application. Handbooks and other guidance
notes are published. Design aids are developed. Methods from the
‘Standard are incorporated into more Standards. Investment is made in
18. All these aspects are disrupted if large changes are mado in the
methods or format of a head code.
Introduction | 3
Accordingly, RS 6399-2 was designed to build on the foundations of
CP 3-V-2, incorporating the necessary changes, bul preserving the
existing format as closely as possible.
1.4. Objectives and principles
‘The members of the drafting panel for BS 6399-2 were given five main
objectives:
© to make the new Standard independent of material and structural
form
to reflect current understanding as closely as possible
10 maintain the current average level of design risk, but reduce the
range of uncertainty
to retain the CP 3-V-2 engineering model and format
1o minimize the scope for misinterpretation.
In executing these objectives, I
principles.
anel members adopted the following
1, To work in terms of normal stress (pressure) and shear stress
(fiction) on the building envelope.
This ensures independence of material and structural form. We
derive the lowds on structural elements by accumulating the
stresses that act on relevant areas of the building envelope.
To account for all effects that influence: loading by 10% or more.
Many factors affect wind loads but only the more significant are
included, Partial factors on wind loads are typically about
71 — 1-4, so that omitting more than four factors could lead to
erosion of the safety margin if a conservative approach is not
taken,
3. To keep the influence of effects in balance.
Fach factor should be included at a precision that isin balance
with the other factors. ‘There is litle point in determining one
factor to a precision of 2% if ethers are only good to 20%.
4. To make values change smoothly, where possible.
Step changes in value give large step changes in design loads,
so there is a natural tendency for designers to pick values to the
less onerous side ul u step. The step changes in CP 3-V-2 were
so large that the possible difference in loads between nominally
identical buildings could easily excced the margin of safety it
several steps acted together, This is illustrated by Example 1,
which shows that the ratio between the most and the: least
onerous interpretation for a typical building at a typical UK
‘west-coast town site can be as large as 270%,
4 1 Wind loading: a practical guide to BS 6399-2
Example 1. Step changes in wind loads possible using,
cP 3-V-2
Note: The methods and values inthis example relate to CP 3-V-
1986 and 1990 amendes.
2:1972, including the
‘ pret, Coser
sp. gives sep changes ine ah of ge ner ar
aio on de cra Bgl wae Ls (E)
wind speed is 47 ds.
sie
a]
Blackpool, V = 47mis Slope. ]
BEER;
>
sea oe Te
We must ask ourselves several questions about the site
y pie an
se seno inn ua a Category 3A the yp a
ona occ Df the so quiz when te Bing
Decio high lapin wos herve Cto
i ew eee Gon ete if baling toler en he pe
bates)
Te ope sep enue esq ep
testes athe crt of the slope and
opt fac erat de ea,
oeraphy factor to be calculated?
if the slope exceeds 1:20, the
We must also ask several questions about the building
© Je the largest dimension ofthe building atleast 50m? H so, the building is
Chass € for structural loads otherwise itis Class B.
+ Shull we use the pressure coefficients in Table X o the force cosficient in
Table 107 CP 3-V-2 offers a free choice, but the values differ considerably,
depending on building proportions.
Suppose the answers to these questions place the site and building exactly om the
val A cans deer may away ke he
ee (he least onerous choice is also valid. This leads
sr neo Reina cs, a ris
ra large range of possible values as in the following: ‘example calculation of base
shar fore.
Introduetion 1 5
Parameter Most onerous Least onerous Notes
asic wind speed: ms 47 a
Slope 005001 044999 ter side of significance
threshold
Topography factor 106 Lo Site at crest of escarpment
Distance inland: m 499 sol From SAS of CI 3V-2
Roughness category 0 3
Building height: m » 10 Equa to obstruction
ei in towne
Building width: m 49999 50001 Either side of Clase NC
boundary
Chass » c
Factor Sa 095 26
Dynamic press 1373 as
Frontal ares m’ su 500
Loading coetticient table Table 10 TableS Free choice of source,
Table 10 more
Force or net pressure 12 095 anereus for these
coelficient proportions
Base shear force: kN 824 306
5. To ensure that Ihe rules are unambiguous and the values used are
consistent and verifiable.
Consistent and unambiguous rules and data reduce the
possibility of errors and disputes. Verifiable data simplify
quality assurance checks und make it casier to amend the
Standard in response to improved knowledge. CP 3-V-2 was
Frequently ambiguous and the data were neither consistent nor
verifiable.
6. To make as few changes from the CP 3-V-2 format as possible
within the limitations set by the other principles.
‘As well as retaining the existing engineering model used by
CP 3-V-2, it was decided to keep the existing S-factor format
and to make as few changes as possible so that the expertise in
the use of the old code developed over 25 years’ use would
transfer to the new Standard
1.5. The changes to ex
05.1 The information chain
Following the trend to regulations and Standards that are less
prescriptive, BS 6399-2 represents two links in the following consistent
and verifiable chain of information
Example 1. Step changes in wind loads possible using,
cP 3-V-2
Note: The methods and values inthis example relate to CP 3-V-
1986 and 1990 amendes.
2:1972, including the
‘ pret, Coser
sp. gives sep changes ine ah of ge ner ar
aio on de cra Bgl wae Ls (E)
wind speed is 47 ds.
sie
a]
Blackpool, V = 47mis Slope. ]
BEER;
>
sea oe Te
We must ask ourselves several questions about the site
y pie an
se seno inn ua a Category 3A the yp a
ona occ Df the so quiz when te Bing
Decio high lapin wos herve Cto
i ew eee Gon ete if baling toler en he pe
bates)
Te ope sep enue esq ep
testes athe crt of the slope and
opt fac erat de ea,
oeraphy factor to be calculated?
if the slope exceeds 1:20, the
We must also ask several questions about the building
© Je the largest dimension ofthe building atleast 50m? H so, the building is
Chass € for structural loads otherwise itis Class B.
+ Shull we use the pressure coefficients in Table X o the force cosficient in
Table 107 CP 3-V-2 offers a free choice, but the values differ considerably,
depending on building proportions.
Suppose the answers to these questions place the site and building exactly om the
val A cans deer may away ke he
ee (he least onerous choice is also valid. This leads
sr neo Reina cs, a ris
ra large range of possible values as in the following: ‘example calculation of base
shar fore.
Introduetion 1 5
Parameter Most onerous Least onerous Notes
asic wind speed: ms 47 a
Slope 005001 044999 ter side of significance
threshold
Topography factor 106 Lo Site at crest of escarpment
Distance inland: m 499 sol From SAS of CI 3V-2
Roughness category 0 3
Building height: m » 10 Equa to obstruction
ei in towne
Building width: m 49999 50001 Either side of Clase NC
boundary
Chass » c
Factor Sa 095 26
Dynamic press 1373 as
Frontal ares m’ su 500
Loading coetticient table Table 10 TableS Free choice of source,
Table 10 more
Force or net pressure 12 095 anereus for these
coelficient proportions
Base shear force: kN 824 306
5. To ensure that Ihe rules are unambiguous and the values used are
consistent and verifiable.
Consistent and unambiguous rules and data reduce the
possibility of errors and disputes. Verifiable data simplify
quality assurance checks und make it casier to amend the
Standard in response to improved knowledge. CP 3-V-2 was
Frequently ambiguous and the data were neither consistent nor
verifiable.
6. To make as few changes from the CP 3-V-2 format as possible
within the limitations set by the other principles.
‘As well as retaining the existing engineering model used by
CP 3-V-2, it was decided to keep the existing S-factor format
and to make as few changes as possible so that the expertise in
the use of the old code developed over 25 years’ use would
transfer to the new Standard
1.5. The changes to ex
05.1 The information chain
Following the trend to regulations and Standards that are less
prescriptive, BS 6399-2 represents two links in the following consistent
and verifiable chain of information
6 À Wind Inading: a practical guide to BS 6399-2
| Current knowledge: a huge amount of published and proprietary
informati
| Engineering models: design procedures and data derived from
current knowledge.
1. Published design guidance: a consistent set of engineering models
published as the DRE Designer's Guide to Wind Loading of
Building Structures.
| BS 6399 diectionl method hie
understanding as possible wit
| BS 6399 dan! method: which is as close to the CP 3
Format and process as possible without heing tuo conservative,
| Simplified design guidance and design al, including this Guide,
BRE Digest 4362 guidance from various sectors of the industry.
and the design aids discussed later
is as close to current
v2
Should BS 6399-2 be unsuitable for your particular building because of
its dynamic response, unusual shape or complex ste, it should always be
possible 10 move up this chain without invalidating the design proces
Finally, if your building or component can be justified by a very simple
design, should always be possible to move down the chain and reduce
the design efor.
1.5.2 Principal rule changes o
TÍ you are converting from CP 3-V-2, you will notice many apparent
differences. The principal changes are summari
the reason for the change.
+ Scope is now resirited to buildings only, as was CP 3-V-2 after (he
1986 amendments. Most uther forms of structure have their own
Standards.
+ A new structural classification method acts as a ‘signpost’ 10
indicate when it is safe to proceed with BS 6299-2, This extends
the scope to include “mildly dynamic" buildings without the need to,
adopt more complicated dynamic inethous.
+ The number of ground roughness categories are reduced from five
to three, but only two are used since the "sea" category is needed
only to define the distance to sea. Ñ
+ The upwind distance of roughness, the “fetch', becames mors
important than before. This is 10 make best use of improved
Enwledge, 10 remove step changes and prevent misuse of previous
rules.
‘The height of upwind buildings in towns becomes a parameter,
instead of an assumed value. This is to remove step changes.
Introduction | 7
© Size classes A, B and C are replaced by the diagonal dimension, a,
of the loaded arca. The size effect is treated separately in the
standard method to reduce the number of calculations and to
remove step changes.
+ Calculations now start from a basic wind speed that is an hourly-
mean value, but the design wind speed, the effective wind speed,
remains a gust wind speed. This brings BS 6399-2 in line with
European amd international practice. I also improves the accuracy
of the topography method.
* A new alto factor reduces the med vo was the ffs of ill
and cscarpments, 16 remove step changes and miti
ie ray y tigate misuse of
+ The limits to the applicability of the mn-by-parts rulo” are
previous rule.
Pressure zone sizes now vary with the proportions of the building,
This is to make best use of improved knowledge, to remove
‘conservatism and to reduce the number and the complexity of the
pressure coefficient tables.
+ Overall forces are developed from surface pressures, so that
separate force coefficients are no longer required. This removes
ambiguities and improves consistency.
+ High suctions on walls caused by tunnelling between buildings are
predicted. This is a result of improved knowledge and to mitigate
observed damage.
© Asymmetric loads need to be checked when the design is
susceptible to them. This is a new requirement dictated by
changes in structural Standards.
© Effects of hips, parapets, curved eaves and mansan caves are
addressed, all of which reduce uplift on ruofs. This makes best use
of improved knowledge and gives the user an opportunity to
optimize the design.
+ Methods and data are given for non-rectangular plan shapes and for
free-standing walls. This is a result of improved knowledge.
The priniples behind these changes were signaled tothe industry ina
seres of papers presented atthe 1980 Construction Indust Rese an
Information Association seminar, Wind Engineering in the Eighties." It
Lak some in oma the pres me wore proces and
hese were published for the industry to test in 1989 as th |
BRE Digest 346.4 =—
| Current knowledge: a huge amount of published and proprietary
informati
| Engineering models: design procedures and data derived from
current knowledge.
1. Published design guidance: a consistent set of engineering models
published as the DRE Designer's Guide to Wind Loading of
Building Structures.
| BS 6399 diectionl method hie
understanding as possible wit
| BS 6399 dan! method: which is as close to the CP 3
Format and process as possible without heing tuo conservative,
| Simplified design guidance and design al, including this Guide,
BRE Digest 4362 guidance from various sectors of the industry.
and the design aids discussed later
is as close to current
v2
Should BS 6399-2 be unsuitable for your particular building because of
its dynamic response, unusual shape or complex ste, it should always be
possible 10 move up this chain without invalidating the design proces
Finally, if your building or component can be justified by a very simple
design, should always be possible to move down the chain and reduce
the design efor.
1.5.2 Principal rule changes o
TÍ you are converting from CP 3-V-2, you will notice many apparent
differences. The principal changes are summari
the reason for the change.
+ Scope is now resirited to buildings only, as was CP 3-V-2 after (he
1986 amendments. Most uther forms of structure have their own
Standards.
+ A new structural classification method acts as a ‘signpost’ 10
indicate when it is safe to proceed with BS 6299-2, This extends
the scope to include “mildly dynamic" buildings without the need to,
adopt more complicated dynamic inethous.
+ The number of ground roughness categories are reduced from five
to three, but only two are used since the "sea" category is needed
only to define the distance to sea. Ñ
+ The upwind distance of roughness, the “fetch', becames mors
important than before. This is 10 make best use of improved
Enwledge, 10 remove step changes and prevent misuse of previous
rules.
‘The height of upwind buildings in towns becomes a parameter,
instead of an assumed value. This is to remove step changes.
Introduction | 7
© Size classes A, B and C are replaced by the diagonal dimension, a,
of the loaded arca. The size effect is treated separately in the
standard method to reduce the number of calculations and to
remove step changes.
+ Calculations now start from a basic wind speed that is an hourly-
mean value, but the design wind speed, the effective wind speed,
remains a gust wind speed. This brings BS 6399-2 in line with
European amd international practice. I also improves the accuracy
of the topography method.
* A new alto factor reduces the med vo was the ffs of ill
and cscarpments, 16 remove step changes and miti
ie ray y tigate misuse of
+ The limits to the applicability of the mn-by-parts rulo” are
previous rule.
Pressure zone sizes now vary with the proportions of the building,
This is to make best use of improved knowledge, to remove
‘conservatism and to reduce the number and the complexity of the
pressure coefficient tables.
+ Overall forces are developed from surface pressures, so that
separate force coefficients are no longer required. This removes
ambiguities and improves consistency.
+ High suctions on walls caused by tunnelling between buildings are
predicted. This is a result of improved knowledge and to mitigate
observed damage.
© Asymmetric loads need to be checked when the design is
susceptible to them. This is a new requirement dictated by
changes in structural Standards.
© Effects of hips, parapets, curved eaves and mansan caves are
addressed, all of which reduce uplift on ruofs. This makes best use
of improved knowledge and gives the user an opportunity to
optimize the design.
+ Methods and data are given for non-rectangular plan shapes and for
free-standing walls. This is a result of improved knowledge.
The priniples behind these changes were signaled tothe industry ina
seres of papers presented atthe 1980 Construction Indust Rese an
Information Association seminar, Wind Engineering in the Eighties." It
Lak some in oma the pres me wore proces and
hese were published for the industry to test in 1989 as th |
BRE Digest 346.4 =—
8 1 Wind loading; a practical guide to BS 6399-2
1.53. Colibrations
3.1 Misinterpretation of CP 3-V-2
"A number of calibrations of BS 6399-2:1997 against previous practice
wore performed by government and industry agencies, Although
previously suspected, the commun misinterpretations of CP 3-V-2 were
wo confirmed until BS 6399-2 began to be used. Letters began 10 appear
‘he technical press, notably the Verdam column in The Structural
Brgineer, complaining tat design wind loads had significantly increase
Investienion of each individual claim usually revealed misapplication of
CP 3-V-2 in either or both of the following two main areas
1. Failing to assess the topography Factor Sy when topography was
significant.
‘A poll of over 300 designers attending BSL workshops on BS
6399-2 revealed only eight (3%) had assessed $ in the
previous year. Over 20% of all sites in the UK quality for
Regesement under the CP 3-V-2 rule for significant topography
(lope >1:20 within 1 km of site)
2. Using inappropriate roughness categories.
appropriate use af Category 4, city centres’, (ug. Exeter is a
city but, even in the centre, it does not most he building height
and density requirements for Category 4).
Inappropriate use of Category 3. ‘towns’ at sites near to the
town boundary and Caregory 2, “open country’ al sites near the
Sen coast, Appendix A uf CP 3-V-2 clearly stipulates “fetch of
à kilometre or more is necessary to establish a different
roughness category’.
We have already seen in Example 1 that the step changes created in
CP_3.V-2 by topography and ground roughness can casily change loads
hy a factor of 27, which easily eliminates the safety margin given by the
paria factor an Toads yy = 14» 1-6. The aprendices of CP 3.12
Pere rarely used, even though they are an integral and essential part of the
Code, Moving the detailed rules from appendices into the main clauses of
RS 6399-2 reduces the likelihood of these mistakes.
1.5.3.2 Design dynamic pressure
“The simplest and most conservative option of BS 6399-2, the standard
method inespective of direction, gives dynamic pressures about 15% on
{erage less than CP 3-V-2 when used correctly. In Fig. 1, the effective
design wind specds at 10m above open country across the Midlands,
ftjacent to the named towns, are compared. Similarly Fig, 2 compares
Sein wind spoeds in the centre of the named Lowns up the spine of the
UK.
Introduction 1 9
5
pee
en
‘onl gut speed Vt)
A
# +
een
PSV eme 2
8509082 anders
Eng em) ”
Fig. Design wind ape open emp aros the dns
gu) m
E
Noting (km) oe“ ee
ig. 2. Design wind speeds in towns up UK spine
It is important to stress the ‘on average”, e dey
casera rro suas me od ond CP eps on
the expe of th se. The tonal metal rs mame
Br ther 4 on menes Tee maybe tic he rate
cst she de le le, a oie we
ras 99-2 is much better at identifying exposed and
‘ce andthe Gide wil help you explo this to advantage.
1.53. Colibrations
3.1 Misinterpretation of CP 3-V-2
"A number of calibrations of BS 6399-2:1997 against previous practice
wore performed by government and industry agencies, Although
previously suspected, the commun misinterpretations of CP 3-V-2 were
wo confirmed until BS 6399-2 began to be used. Letters began 10 appear
‘he technical press, notably the Verdam column in The Structural
Brgineer, complaining tat design wind loads had significantly increase
Investienion of each individual claim usually revealed misapplication of
CP 3-V-2 in either or both of the following two main areas
1. Failing to assess the topography Factor Sy when topography was
significant.
‘A poll of over 300 designers attending BSL workshops on BS
6399-2 revealed only eight (3%) had assessed $ in the
previous year. Over 20% of all sites in the UK quality for
Regesement under the CP 3-V-2 rule for significant topography
(lope >1:20 within 1 km of site)
2. Using inappropriate roughness categories.
appropriate use af Category 4, city centres’, (ug. Exeter is a
city but, even in the centre, it does not most he building height
and density requirements for Category 4).
Inappropriate use of Category 3. ‘towns’ at sites near to the
town boundary and Caregory 2, “open country’ al sites near the
Sen coast, Appendix A uf CP 3-V-2 clearly stipulates “fetch of
à kilometre or more is necessary to establish a different
roughness category’.
We have already seen in Example 1 that the step changes created in
CP_3.V-2 by topography and ground roughness can casily change loads
hy a factor of 27, which easily eliminates the safety margin given by the
paria factor an Toads yy = 14» 1-6. The aprendices of CP 3.12
Pere rarely used, even though they are an integral and essential part of the
Code, Moving the detailed rules from appendices into the main clauses of
RS 6399-2 reduces the likelihood of these mistakes.
1.5.3.2 Design dynamic pressure
“The simplest and most conservative option of BS 6399-2, the standard
method inespective of direction, gives dynamic pressures about 15% on
{erage less than CP 3-V-2 when used correctly. In Fig. 1, the effective
design wind specds at 10m above open country across the Midlands,
ftjacent to the named towns, are compared. Similarly Fig, 2 compares
Sein wind spoeds in the centre of the named Lowns up the spine of the
UK.
Introduction 1 9
5
pee
en
‘onl gut speed Vt)
A
# +
een
PSV eme 2
8509082 anders
Eng em) ”
Fig. Design wind ape open emp aros the dns
gu) m
E
Noting (km) oe“ ee
ig. 2. Design wind speeds in towns up UK spine
It is important to stress the ‘on average”, e dey
casera rro suas me od ond CP eps on
the expe of th se. The tonal metal rs mame
Br ther 4 on menes Tee maybe tic he rate
cst she de le le, a oie we
ras 99-2 is much better at identifying exposed and
‘ce andthe Gide wil help you explo this to advantage.
10 | Wind loading: a practical guide to BS 6399-2
15.33 Cladding toads |
o rule changes listed above.
“adding loads are affected hy almost all the rule ch :
nacre nose efi y me came
anges to the division by-parts rule. This rule, which is explained |
SINE on 2:44, does not apply to cladding — but has been out ey
applied to cladding by users of CP 3-V-2, despite strong advice 10
SD irary The problem is that CP 3-V-2 did not clealy define what was
SEE an so the rule was routinely taken 0 apply outside its bounds
applicabilit j vy.
9" clon aan eurent pre, mu compar CR 2-2
¿ng the division by-paris rule against BS 6399-2 without using
aa elbration of te highest hlding foals on he
walls uf five-storey buildings for: a
+ Option |, the simplest but most conservative optio in BS 6399-2
2 Option 3, the least conservative option af BS 6399-
il) see in detail later what these options mean. | un,
Wye wah option the comparison is made using the ‘standant inemah
ressure coefficients for eluding of Ga = +0:2 in CP 3.V2, and of hol
Cy = 40:2 and Ca =-03 in BS 63992 BS 6399-2 Eve: ove
«adding loads than CP 3-V-2 on the top storey because here the divisi
ene fect and the other changes decrease loads, But
y-parts rule has no ef ee
An Rennes rule becomes more effective down the building,
N
Bsemacpsva
Fig. 3. Calibration of highest cladding panel louds
Introduction 114
so the loads on the hotiom-storey cladding are substantially higher
using BS 6399-2.
In practice, it is seldom possible to take full advantage of cladding
loads that decrease with height down a building, and panels are usually
designed For the largest loading at the top. In the correct interpretation of
CP 3-V-2, the division-by-parts rule should not have been used for
cladding. In both these cases, the average effect of BS 6399-2 is a
decrease in highest clakling loads by 21%.
The least favourable comparison occurs when the lowest storey is clad
to a different spécification than the remainder of the building and when
the division-hy-parts rule would have been used in CP 3-V-2, By
‘enforcing the limits of applicability of the rule, BS 6399-2 effectively
increases the design load on cladding by an average of about 50% when
the most conservative Option 1 is used and 30% when Option 3 is used
While it is fair to point out that minimizing misapplication was one of the
five main objectives, it is mot much comfort in this particular
circumstance. Nevertheless, BS 6399-2 offers many new opportunities
to reduce conservatism and so mitigate this effect.
15.34 Structural loads
The effect of the division-by-pants rule is diluted in structural loads,
because the structural elements of the lower storeys must curry the wind
forces from all the storeys above. A similar calibration to that above, this
ime for timber-framed buildings, shows that the BS 6399-2 loads on the
panels of the top storey are about 15% lower and on the bottom storey
0-5% lower on average when using BS 6399-2,
‘The improved pressure coefficients in BS 6399-2 make subtle changes
to the distribution of extemal pressures, leading to «l
structural loads on elements. Suctions on the roof edge zones increase in
value but the zones decrease in size, while suctions on the rest of the
roof tend to decrease. This trend favours long-span low-pitched roofs.
Figure 4 illustrates the maximum sheeting rail and purlin loads on a
typical portal frame building with a 7° duopitch roof. Here ‘maximum
loads occur when the intemal pressure coefficient is Cp; = +0:2 and
‘minimum’ loads when Cy — ~0-3. Values from CP 3-V-2 are shown for
both Class À and Class B because this is not the place to debate whether
rail and purlins qualify as ‘cladding’ or ‘structure’ and, in any case, the
difference this makes is small. I is clear that, while BS 6399-2 loads on
the wall sh are similar to CP 3-V-2, loads on the root purlins
are significantly reduced when the roof span is long.
More subtle changes are evident when the roof pitch is increased to 30°.
‘At this pitch angle, CP 3-V-2 predicts zero external pressure on the
windward roof slope, so the loads come from the internal pressure only.
15.33 Cladding toads |
o rule changes listed above.
“adding loads are affected hy almost all the rule ch :
nacre nose efi y me came
anges to the division by-parts rule. This rule, which is explained |
SINE on 2:44, does not apply to cladding — but has been out ey
applied to cladding by users of CP 3-V-2, despite strong advice 10
SD irary The problem is that CP 3-V-2 did not clealy define what was
SEE an so the rule was routinely taken 0 apply outside its bounds
applicabilit j vy.
9" clon aan eurent pre, mu compar CR 2-2
¿ng the division by-paris rule against BS 6399-2 without using
aa elbration of te highest hlding foals on he
walls uf five-storey buildings for: a
+ Option |, the simplest but most conservative optio in BS 6399-2
2 Option 3, the least conservative option af BS 6399-
il) see in detail later what these options mean. | un,
Wye wah option the comparison is made using the ‘standant inemah
ressure coefficients for eluding of Ga = +0:2 in CP 3.V2, and of hol
Cy = 40:2 and Ca =-03 in BS 63992 BS 6399-2 Eve: ove
«adding loads than CP 3-V-2 on the top storey because here the divisi
ene fect and the other changes decrease loads, But
y-parts rule has no ef ee
An Rennes rule becomes more effective down the building,
N
Bsemacpsva
Fig. 3. Calibration of highest cladding panel louds
Introduction 114
so the loads on the hotiom-storey cladding are substantially higher
using BS 6399-2.
In practice, it is seldom possible to take full advantage of cladding
loads that decrease with height down a building, and panels are usually
designed For the largest loading at the top. In the correct interpretation of
CP 3-V-2, the division-by-parts rule should not have been used for
cladding. In both these cases, the average effect of BS 6399-2 is a
decrease in highest clakling loads by 21%.
The least favourable comparison occurs when the lowest storey is clad
to a different spécification than the remainder of the building and when
the division-hy-parts rule would have been used in CP 3-V-2, By
‘enforcing the limits of applicability of the rule, BS 6399-2 effectively
increases the design load on cladding by an average of about 50% when
the most conservative Option 1 is used and 30% when Option 3 is used
While it is fair to point out that minimizing misapplication was one of the
five main objectives, it is mot much comfort in this particular
circumstance. Nevertheless, BS 6399-2 offers many new opportunities
to reduce conservatism and so mitigate this effect.
15.34 Structural loads
The effect of the division-by-pants rule is diluted in structural loads,
because the structural elements of the lower storeys must curry the wind
forces from all the storeys above. A similar calibration to that above, this
ime for timber-framed buildings, shows that the BS 6399-2 loads on the
panels of the top storey are about 15% lower and on the bottom storey
0-5% lower on average when using BS 6399-2,
‘The improved pressure coefficients in BS 6399-2 make subtle changes
to the distribution of extemal pressures, leading to «l
structural loads on elements. Suctions on the roof edge zones increase in
value but the zones decrease in size, while suctions on the rest of the
roof tend to decrease. This trend favours long-span low-pitched roofs.
Figure 4 illustrates the maximum sheeting rail and purlin loads on a
typical portal frame building with a 7° duopitch roof. Here ‘maximum
loads occur when the intemal pressure coefficient is Cp; = +0:2 and
‘minimum’ loads when Cy — ~0-3. Values from CP 3-V-2 are shown for
both Class À and Class B because this is not the place to debate whether
rail and purlins qualify as ‘cladding’ or ‘structure’ and, in any case, the
difference this makes is small. I is clear that, while BS 6399-2 loads on
the wall sh are similar to CP 3-V-2, loads on the root purlins
are significantly reduced when the roof span is long.
More subtle changes are evident when the roof pitch is increased to 30°.
‘At this pitch angle, CP 3-V-2 predicts zero external pressure on the
windward roof slope, so the loads come from the internal pressure only.
12 1 Wind loading: a practical guide to BS 6399-2
1.2 predicts a range of positive and negative extemal pressures
extemal pressure. Figure 5 shows the corresponding. a en
purlin loads. The highest loaded purlin changes from Bein Ih send
putin bob the and he mete the
Fi as mr
roca
Fi, A, Sheets rail an stn los on aporte frame uti with Pop roof
Fig. 5. Sheevng rel and partir leds an a penal frame building with a 30° dunpich roof
Introduction V 13
1.6 Format and content of this Guide
1.6.1 Format
16.1.1 Cross-references
‘This Guide refers to itself, to BS 6399-2 and, occasionally, when
highlighting changes of practico, to CP 3-V-2. The following format has
been adopted tu make these references easier to understand.
+ References to sections, figures and equations in this Guide are
shown in bold text, ex. (see 1.6.1) refers you to this section, and
you have already seen references tothe fist five figures as Fig. 1 to
Fig. 5.
+ References o clauses, figures and equations in BS 6399-2 are
shown in italic text, eg. 93.4.2 refers lo clause 3.4.2 of the
Standard, Figure 6 to the map of basic wind speed, ete.
® References lo. CP 3-V-2, being less frequent, are set in Roman and
the code is explicitly mentioned, eg. ‘in §5.5.2 of CP 3-V-2".
1.6.1.2. Implementation of the Standard's clauses
Specific steps to implement the clauses of the Standard are indicated by
special bulle! points as shown next,
© Main steps are marked hy this box bullet.
O Secondary steps by this bullet,
Tertiary steps by this bullet
‘The idea here is to draw the ronder’s attention 10 the most important
‘guidance.
1613 Notes
Occasionally, where information needs to be inserted that would
otherwise interrupt the flow of the text, notes are inserted after the
relevant paragraph in the form:
Note: There are 10 typographical errors in Figure 7(b) of BS 6399-2:
1997. 5 x slope length if Vu > 0-3" should be ‘5 x slope height if
‘iy > 0.3" and ‘Downwind slope ty > 0-05" should be "Downwind
slope Yo < 0-08".
1.6.14. Tips and examples
Useful tips are inserted at intervals, usually immediately after detail
‘guidance on a subject, in the form:
1.2 predicts a range of positive and negative extemal pressures
extemal pressure. Figure 5 shows the corresponding. a en
purlin loads. The highest loaded purlin changes from Bein Ih send
putin bob the and he mete the
Fi as mr
roca
Fi, A, Sheets rail an stn los on aporte frame uti with Pop roof
Fig. 5. Sheevng rel and partir leds an a penal frame building with a 30° dunpich roof
Introduction V 13
1.6 Format and content of this Guide
1.6.1 Format
16.1.1 Cross-references
‘This Guide refers to itself, to BS 6399-2 and, occasionally, when
highlighting changes of practico, to CP 3-V-2. The following format has
been adopted tu make these references easier to understand.
+ References to sections, figures and equations in this Guide are
shown in bold text, ex. (see 1.6.1) refers you to this section, and
you have already seen references tothe fist five figures as Fig. 1 to
Fig. 5.
+ References o clauses, figures and equations in BS 6399-2 are
shown in italic text, eg. 93.4.2 refers lo clause 3.4.2 of the
Standard, Figure 6 to the map of basic wind speed, ete.
® References lo. CP 3-V-2, being less frequent, are set in Roman and
the code is explicitly mentioned, eg. ‘in §5.5.2 of CP 3-V-2".
1.6.1.2. Implementation of the Standard's clauses
Specific steps to implement the clauses of the Standard are indicated by
special bulle! points as shown next,
© Main steps are marked hy this box bullet.
O Secondary steps by this bullet,
Tertiary steps by this bullet
‘The idea here is to draw the ronder’s attention 10 the most important
‘guidance.
1613 Notes
Occasionally, where information needs to be inserted that would
otherwise interrupt the flow of the text, notes are inserted after the
relevant paragraph in the form:
Note: There are 10 typographical errors in Figure 7(b) of BS 6399-2:
1997. 5 x slope length if Vu > 0-3" should be ‘5 x slope height if
‘iy > 0.3" and ‘Downwind slope ty > 0-05" should be "Downwind
slope Yo < 0-08".
1.6.14. Tips and examples
Useful tips are inserted at intervals, usually immediately after detail
‘guidance on a subject, in the form:
24 | Wind loading: a practical guide to BS 0399-2
Similarly, example calculations are also inserted at intervals afer detailed
juidance on a subject. You have already seen Example 1
6.1.5 ‘Intermediate’ and ‘advanced’ guidance cit
VE ause the Guide is intended to extend the expertise of the us it
unfkely that all the guidance can be absorbed at the ist eng and the
more advanced guidance may only rarely be needed To assist th
ee ace has een med as "basic, “inennediate™ and
‘advances
asias “interme”
A single line down the land, arg i
| sige ponce Ya may wish to sip over this materi 0
the tt fev readings. ;
bres down he and margin indicate ‘aa
Bee Yon may wish offer 1 ths material ony when it
Domos relevant your pre
1.6.2 Content
“This Guide contains four chapters.
ction, You have just read this, so there is mot much point
well, at is definitely overdue for its retirement, BS 6399 sts the
Standard forthe immediate future, uni replaced in about 2005 by
Furocode 1, The rest ofthis Guide is about applying BS 6399-2
cffiviently, CP 3-V-2 will be mentioned only when an importan
Change of practice needs 10 he described. eee
2. Commentary, This gives comments and advice onthe aplication of
BS 6399-2 on a ‘step-by-step’ basis through the require
calculations. This is not the same as a ‘clause-hy-lause) basis
Fecauso the order in which the clauses are presented is not the mo
convenient orde of calculation, Exceptions and special provisions,
Which BS 6999-2 usually presents lst in any cause ae general
described fst to prevent the user from making abortive calculations.
Te directional method will generally be described before Ih
standard method that was derived from jt, but the user may safely
Skip over these diretional method descriptions because the Stand
method descriptions are always complete in themselves. The
Commentary is mostly ‘asi’ and “intermediate? materia UA
3. Making BS 6399-2 work for you. This chapter is a about
applying BS 6399-2 and getting the most from it with the le
elfort. Much of this guidance is rated as “intermediate” or
“advanced.
Introduction | 15
4. Worked examples. A number of example calculations are
provided to demonstrate most of the guidance given in this Guide
and the steps required for various elements of typical structures.
These examples have been chosen to illustrate as many aspects as
possible within the space available, but are not complete. They are
not intended as templates to be copied. The particular choice of
examples does not imply that explicit calculations are necessary
for elements that could be sized using the prescriptive methods
given in Approved Document A to the Building Regulations and in
BS 8103, Structural design of low-rise buildings.
1.6.3. 2002 amendments
In early 2002, shortly after the withdrawal of CP3 Chapter V Part 2,
BS6399-2:1997 was reprinted with a number of minor changes. These
included:
Clowrer definition of the rules for obtaining the effective height.
2. Simplification of the rules for asymmetric luading.
3. Improved consistency between the pressure coefficients for
pitched roofs in the standard and directional methods.
4. Changes to the pressure cnefficients for walls in the standard
method, including introduction of net coefficients for overall
horizontal loads,
5. Introduction of a redu
parapets, depen
n factor for free-standing walls and
1g on the length of the wall.
‘The opportunity was also taken to incorporate other minor editorial
changes to improve clarity.
This guide now includes commentaries on these amendments in the
appropriate sections. However, it is helpful to comment here on’ the
reasons for, and principles, of two of these changes.
“The original clauses on asymmetric ads had been intended to provide
clear advice to enable designers to cater for structures susceptible to
asymmetry without introducing undue conservatism. Objections to these
rules, on the grounds of complexity, were received, particularly from
some designers who applied these rules to multi-bay structures that were
not susceptible, Thc amendments replaced the rules with a general
Statement requiring an allowance for asymmetry to be made, leaving the
implementation to the designer, but adding simplified conser
as options in two notes to the clause.
Coefficients in the standard method had been derived from the
directional method for a range of wind direction +45° either side of the
ortogonal cases and by amalgamating some of the local zones, while
maintaining the required overall risk. This resulted in standard values
rules
Similarly, example calculations are also inserted at intervals afer detailed
juidance on a subject. You have already seen Example 1
6.1.5 ‘Intermediate’ and ‘advanced’ guidance cit
VE ause the Guide is intended to extend the expertise of the us it
unfkely that all the guidance can be absorbed at the ist eng and the
more advanced guidance may only rarely be needed To assist th
ee ace has een med as "basic, “inennediate™ and
‘advances
asias “interme”
A single line down the land, arg i
| sige ponce Ya may wish to sip over this materi 0
the tt fev readings. ;
bres down he and margin indicate ‘aa
Bee Yon may wish offer 1 ths material ony when it
Domos relevant your pre
1.6.2 Content
“This Guide contains four chapters.
ction, You have just read this, so there is mot much point
well, at is definitely overdue for its retirement, BS 6399 sts the
Standard forthe immediate future, uni replaced in about 2005 by
Furocode 1, The rest ofthis Guide is about applying BS 6399-2
cffiviently, CP 3-V-2 will be mentioned only when an importan
Change of practice needs 10 he described. eee
2. Commentary, This gives comments and advice onthe aplication of
BS 6399-2 on a ‘step-by-step’ basis through the require
calculations. This is not the same as a ‘clause-hy-lause) basis
Fecauso the order in which the clauses are presented is not the mo
convenient orde of calculation, Exceptions and special provisions,
Which BS 6999-2 usually presents lst in any cause ae general
described fst to prevent the user from making abortive calculations.
Te directional method will generally be described before Ih
standard method that was derived from jt, but the user may safely
Skip over these diretional method descriptions because the Stand
method descriptions are always complete in themselves. The
Commentary is mostly ‘asi’ and “intermediate? materia UA
3. Making BS 6399-2 work for you. This chapter is a about
applying BS 6399-2 and getting the most from it with the le
elfort. Much of this guidance is rated as “intermediate” or
“advanced.
Introduction | 15
4. Worked examples. A number of example calculations are
provided to demonstrate most of the guidance given in this Guide
and the steps required for various elements of typical structures.
These examples have been chosen to illustrate as many aspects as
possible within the space available, but are not complete. They are
not intended as templates to be copied. The particular choice of
examples does not imply that explicit calculations are necessary
for elements that could be sized using the prescriptive methods
given in Approved Document A to the Building Regulations and in
BS 8103, Structural design of low-rise buildings.
1.6.3. 2002 amendments
In early 2002, shortly after the withdrawal of CP3 Chapter V Part 2,
BS6399-2:1997 was reprinted with a number of minor changes. These
included:
Clowrer definition of the rules for obtaining the effective height.
2. Simplification of the rules for asymmetric luading.
3. Improved consistency between the pressure coefficients for
pitched roofs in the standard and directional methods.
4. Changes to the pressure cnefficients for walls in the standard
method, including introduction of net coefficients for overall
horizontal loads,
5. Introduction of a redu
parapets, depen
n factor for free-standing walls and
1g on the length of the wall.
‘The opportunity was also taken to incorporate other minor editorial
changes to improve clarity.
This guide now includes commentaries on these amendments in the
appropriate sections. However, it is helpful to comment here on’ the
reasons for, and principles, of two of these changes.
“The original clauses on asymmetric ads had been intended to provide
clear advice to enable designers to cater for structures susceptible to
asymmetry without introducing undue conservatism. Objections to these
rules, on the grounds of complexity, were received, particularly from
some designers who applied these rules to multi-bay structures that were
not susceptible, Thc amendments replaced the rules with a general
Statement requiring an allowance for asymmetry to be made, leaving the
implementation to the designer, but adding simplified conser
as options in two notes to the clause.
Coefficients in the standard method had been derived from the
directional method for a range of wind direction +45° either side of the
ortogonal cases and by amalgamating some of the local zones, while
maintaining the required overall risk. This resulted in standard values
rules
16 | Wind loading: a practical guide to BS 6399-2
cose to, but sometime significant Tess than, the maximum dein
wale in cion was made from some industry bodies
value in the range. Representation was mad nd E
the direcional method. On consideration, the drafting panel found (hat be
ing the increased risk of the highest value against
ie one vales cold ma he guaranteed a is tha
Teducad risks of the lower values could nut he gt E
ed withthe industry request. The
‘wore not uniformly exposed, und complied will »
ways the largest found from the
tandard method value is now always Ü a found om
{reetional method in any zone over the range 45° either side of the
Orthogonal case, As a result ofthis change, the increase in conserve
Sas Unacceptable in the net horizontal Yoads obtained by summing
Windward and leeward Faces, requiring the introduction of a table of new
net coefficients.
2. Commentary
2.1 Scope, definitions and procedure
LIA Scope ($1.1}
BS 6399-2 provides gust peak wind loads for use in static methods for the
design of buildings. The scope excludes buildings that are ‘particularly’
dynamic, but now includes buildings that are mildly dynamic, assessed
using the dynamic classification explained in 2.2. The svope excludes
other forms of structure that have their own code.
Turbulence in wind caunes wind loads to fluctuate in space and
time, so the Standard represents the peak load on a surface by the
largest gust that loads the whole of that surface simultaneously. This
is called the “equivalent static gust, The equivalent static gust is
represented in Fig, 6 as having a size a across the wind and 4-54
along the wind, where a is the diagonal dimension (sot 2.6.2) of the
surface. This is swept along by the mean wind speed Vo So the
duration of the gust 7, which is the takes to pass, is given by:
Sa] Vo MED
Its just one of a range of gust sizes that add together to give the
peak gust wind speed Y. We predict this gust wind speed by adding.
the effect of all larger gusts to the mean wind speed. This is called
the “peak factor method? and is shown in Fig. 7, where vis the root-
‘mean-square of the turbulence and gy is the peak factor that depends
‘on the gust duration £, The derivation of the peak factor is described
in Annex F.
In using the equivalent static gust, we assume that the area of a
surface is fully loaded by all gusts that are larger than the diagonal
dimension u and is completely unaffected by gusts that ane smaller
han a. In reality there is the smooth transition, but the effect of
gusts just smaller than a is balanced by the effect of gusts just larger
than a. Clearly this is a very simplified model, but it proves to be
remarkably effective and robust
cose to, but sometime significant Tess than, the maximum dein
wale in cion was made from some industry bodies
value in the range. Representation was mad nd E
the direcional method. On consideration, the drafting panel found (hat be
ing the increased risk of the highest value against
ie one vales cold ma he guaranteed a is tha
Teducad risks of the lower values could nut he gt E
ed withthe industry request. The
‘wore not uniformly exposed, und complied will »
ways the largest found from the
tandard method value is now always Ü a found om
{reetional method in any zone over the range 45° either side of the
Orthogonal case, As a result ofthis change, the increase in conserve
Sas Unacceptable in the net horizontal Yoads obtained by summing
Windward and leeward Faces, requiring the introduction of a table of new
net coefficients.
2. Commentary
2.1 Scope, definitions and procedure
LIA Scope ($1.1}
BS 6399-2 provides gust peak wind loads for use in static methods for the
design of buildings. The scope excludes buildings that are ‘particularly’
dynamic, but now includes buildings that are mildly dynamic, assessed
using the dynamic classification explained in 2.2. The svope excludes
other forms of structure that have their own code.
Turbulence in wind caunes wind loads to fluctuate in space and
time, so the Standard represents the peak load on a surface by the
largest gust that loads the whole of that surface simultaneously. This
is called the “equivalent static gust, The equivalent static gust is
represented in Fig, 6 as having a size a across the wind and 4-54
along the wind, where a is the diagonal dimension (sot 2.6.2) of the
surface. This is swept along by the mean wind speed Vo So the
duration of the gust 7, which is the takes to pass, is given by:
Sa] Vo MED
Its just one of a range of gust sizes that add together to give the
peak gust wind speed Y. We predict this gust wind speed by adding.
the effect of all larger gusts to the mean wind speed. This is called
the “peak factor method? and is shown in Fig. 7, where vis the root-
‘mean-square of the turbulence and gy is the peak factor that depends
‘on the gust duration £, The derivation of the peak factor is described
in Annex F.
In using the equivalent static gust, we assume that the area of a
surface is fully loaded by all gusts that are larger than the diagonal
dimension u and is completely unaffected by gusts that ane smaller
han a. In reality there is the smooth transition, but the effect of
gusts just smaller than a is balanced by the effect of gusts just larger
than a. Clearly this is a very simplified model, but it proves to be
remarkably effective and robust
18 À Wind loading: a practical guide to BS 6399-2
3
Fig, 6, Eguivelen ste gust
Y, gov
EE
Fig. 2. Peak factor method
2.1.2 Definitions 1.3)
Figure 2, which illustrates the difference between ‘fixed dimensions"
(length, width und height that are always constant) and ‘variable
height” is the height above ground — these are often the same value, but
spacing X, but Figure EI in Annex E is moro helpful than $17.33 in
2.1.3. Standard and directional methods ($1.8)
Following the trend towards regulations and Standards that are less
prescriptive, BS 6399-2 offers the user a free choice of two methods. The
Fhoice depends on the level of detail desired and requires the user (0
exercise judgement
dE, andl method is recommended for hand sucios, bt
is between 0% and 30% conservative, depending on the site because it uses
Worst combinations of factors. The directional method is more precise and
"etailed, but is better suited for implementation by computer than by hand
‘also there wre some “hybrid” options that lic between the standard and
Gieetiomal methods thal are particularly useful for eliminating unwanted
Conservatism at sites in towns, The options are there because BS 6399-2
rust cover a wide range of buildings and also act as the “head code” for
specialized structures that have their own Standards.
"Some concerns have been expressed in the technical press that the
range of choice may lead to disputes between the designer and the
building control officer. Always remember that the standard method is a
Commentary | 19
simplification of the directional method. The simpler options are always
conservative and the more complicated options are more precise.
Whichever option you choose to use will be safe for heights above
round less than the 100m limit of the standard method wind speeds.
Its expected that the options will be chosen as follows.
+ For ‘back-of-envelope’ estimates when tendering for jobs: standard
method, irrespective of direction,
For the majority of designs: standard method, orthogonal load cases.
When cos o the ede of own ‘hybrid! option (43:42) onbo-
gonal load case
+ To justify member sizes when close to a li
critical direction only.
+ For complex ar prest
directions.
s build
This Guide explains each of these options.
2.1.4 The four steps
‘The flowchart illustrating the outline procedure, Figure /, breaks the
alculation into ten stages. This Guide groups these into four distinct and
ndependont steps:
1. Dynamic classification — Stages I and 2.
2. Wind climate—Stages 3 and 4.
3. Site exposure — Stages 5 10 8,
4. Building shape factors Stages 9 and 10.
It is valid to switch between options only between these groups. You
may wish to do this to get an optimum ‘hybrid’ solution,
Our choice uf options is affected by the complexity of the site and of
the shape and structure of the building. So to understand these choices, we
must first understand a number of concepts that influence them.
2.1.5. Wind direction ($2.1.1, $3.1.1
215.1 Role rl
“Despite the names “standard method’ and “directional method’, wind
direction plays an important role in both methods. The effect of direction on
the wind climate is given by the direction factor Sy (see 2.3.4). But many of
the other parameters that define the site exposure also depend on wind
direction — distance to sca, distance in town, obstruction height and spacing
all vary with direction, as do the pressure coefficients (see 82.2.2.3 Note).
Wind direction is a parameter we can choose not to consider. To do this
we take the most onerous values irrespective of direction, but we take a
penalty of increased conservatism.
3
Fig, 6, Eguivelen ste gust
Y, gov
EE
Fig. 2. Peak factor method
2.1.2 Definitions 1.3)
Figure 2, which illustrates the difference between ‘fixed dimensions"
(length, width und height that are always constant) and ‘variable
height” is the height above ground — these are often the same value, but
spacing X, but Figure EI in Annex E is moro helpful than $17.33 in
2.1.3. Standard and directional methods ($1.8)
Following the trend towards regulations and Standards that are less
prescriptive, BS 6399-2 offers the user a free choice of two methods. The
Fhoice depends on the level of detail desired and requires the user (0
exercise judgement
dE, andl method is recommended for hand sucios, bt
is between 0% and 30% conservative, depending on the site because it uses
Worst combinations of factors. The directional method is more precise and
"etailed, but is better suited for implementation by computer than by hand
‘also there wre some “hybrid” options that lic between the standard and
Gieetiomal methods thal are particularly useful for eliminating unwanted
Conservatism at sites in towns, The options are there because BS 6399-2
rust cover a wide range of buildings and also act as the “head code” for
specialized structures that have their own Standards.
"Some concerns have been expressed in the technical press that the
range of choice may lead to disputes between the designer and the
building control officer. Always remember that the standard method is a
Commentary | 19
simplification of the directional method. The simpler options are always
conservative and the more complicated options are more precise.
Whichever option you choose to use will be safe for heights above
round less than the 100m limit of the standard method wind speeds.
Its expected that the options will be chosen as follows.
+ For ‘back-of-envelope’ estimates when tendering for jobs: standard
method, irrespective of direction,
For the majority of designs: standard method, orthogonal load cases.
When cos o the ede of own ‘hybrid! option (43:42) onbo-
gonal load case
+ To justify member sizes when close to a li
critical direction only.
+ For complex ar prest
directions.
s build
This Guide explains each of these options.
2.1.4 The four steps
‘The flowchart illustrating the outline procedure, Figure /, breaks the
alculation into ten stages. This Guide groups these into four distinct and
ndependont steps:
1. Dynamic classification — Stages I and 2.
2. Wind climate—Stages 3 and 4.
3. Site exposure — Stages 5 10 8,
4. Building shape factors Stages 9 and 10.
It is valid to switch between options only between these groups. You
may wish to do this to get an optimum ‘hybrid’ solution,
Our choice uf options is affected by the complexity of the site and of
the shape and structure of the building. So to understand these choices, we
must first understand a number of concepts that influence them.
2.1.5. Wind direction ($2.1.1, $3.1.1
215.1 Role rl
“Despite the names “standard method’ and “directional method’, wind
direction plays an important role in both methods. The effect of direction on
the wind climate is given by the direction factor Sy (see 2.3.4). But many of
the other parameters that define the site exposure also depend on wind
direction — distance to sca, distance in town, obstruction height and spacing
all vary with direction, as do the pressure coefficients (see 82.2.2.3 Note).
Wind direction is a parameter we can choose not to consider. To do this
we take the most onerous values irrespective of direction, but we take a
penalty of increased conservatism.
20 À Wind loading: a practical guide to BS 6399-2
ir wind causes the wind direction to fluctuate by
airy de
2.152 Definitions (§2.1.1.2, 522.23, §3.1.1-1)
HS 6399-2 defines wind direction in two ways: u
in degrees east from north, using the symbol p. This is uscd to assess
Me fee climate and se capote, wheter ono the citation of
the building is known, So y = 0° represents winds from the north,
ip 0 from the east, p= 225° from the south-west te,
2. In degrees from normal to a principal axis of the building, using
the symbol 6, This is used to assess the pressure coefficients,
pressures and Toads on the building, So 0 — 0° represents wind
normal to a wall and to the eaves of a roof. while 0 =
Topresents wind parallel to a wall and to the eaves of a roo
venient to convert the wind direction from north 9 into
ing axis 0. as described in the mote to
I is often cor
the angle from normal to the build
83.1.1 and Fig, 8.
2153 Options «sg on
a See nca options for dealing with wind retin
i g one value
1. “Imespeetivo” of direction —taking one val
that isthe most onerous fr al wind direction.
2.10 90*-wide scetors—tiking four values, Increasing
Increasing À Corresponding to the orthogonal load cases conservatism
complexity SF the standard method.
3. In 30*wide sectors —taking twelve values,
Corresponding 10 the directional method
9-0
Fig. 8, Comerting wind direction to angle from normal eaves
Commentary | 21
2.1.5.1. Option | —Irrespective of direction, standard method
Option Lis the least complicated but is the most conservative. For this
we Use the standard method, but we set the direction factor to S4= 1-0
(§2.2.2.3) and use the closest distance to the sea and the shortest distance
in town found over al directions, These distances can be simply measured
as shown in Fig, 9, or averaged over a 30°-wide range.
“This option is useful for initial back-of-cnvelupe calculations and we.
need go no Further if the conservative wind loads this produces are
acceptable. The code states that this option should be used with the
standard method when the orientation of the building is unknown
(§2.2.2.3) but in 3.5.3 we will see how to get around this restriction by
using the hybrid method uf 83.4.2.
Figure 9 illustrates hat the most onerous values of parameters may not
be coincident when taken inespective of direction, giving a conservative
result.
‘Typically, Option 1 will give 15% higher loads that Option 3, but this
will depend on how the parameters combine. Near u west-facing coast
there will be very little conservatism, but near an east-facing coast there
may be up to 30% conservatism. We will see in 3.3.2 how to recover this
conservatism with the least effort.
2.1.5.32 Option 2— Orthogonal load cases, standard method
Option 2 is the method intended for calculating the standard orthogonal
load cases hy hand. Ta this case, we consider the range of wind direction
445° either side of the orthogonal axes of the building, We take the most
onerous value of each parameter within cuch of these ranges, as directed
by $2.1.1.2 and §2.2.2.3 (Note).
Figure 10 shows one of the four orthogonal load eases for the building
in Fig. 9— wind normal to the south-west-facing eaves. The most
Shortest Closest
distance
in tom
Fig. 9. Option 1 — Inespeetse of direction
ir wind causes the wind direction to fluctuate by
airy de
2.152 Definitions (§2.1.1.2, 522.23, §3.1.1-1)
HS 6399-2 defines wind direction in two ways: u
in degrees east from north, using the symbol p. This is uscd to assess
Me fee climate and se capote, wheter ono the citation of
the building is known, So y = 0° represents winds from the north,
ip 0 from the east, p= 225° from the south-west te,
2. In degrees from normal to a principal axis of the building, using
the symbol 6, This is used to assess the pressure coefficients,
pressures and Toads on the building, So 0 — 0° represents wind
normal to a wall and to the eaves of a roof. while 0 =
Topresents wind parallel to a wall and to the eaves of a roo
venient to convert the wind direction from north 9 into
ing axis 0. as described in the mote to
I is often cor
the angle from normal to the build
83.1.1 and Fig, 8.
2153 Options «sg on
a See nca options for dealing with wind retin
i g one value
1. “Imespeetivo” of direction —taking one val
that isthe most onerous fr al wind direction.
2.10 90*-wide scetors—tiking four values, Increasing
Increasing À Corresponding to the orthogonal load cases conservatism
complexity SF the standard method.
3. In 30*wide sectors —taking twelve values,
Corresponding 10 the directional method
9-0
Fig. 8, Comerting wind direction to angle from normal eaves
Commentary | 21
2.1.5.1. Option | —Irrespective of direction, standard method
Option Lis the least complicated but is the most conservative. For this
we Use the standard method, but we set the direction factor to S4= 1-0
(§2.2.2.3) and use the closest distance to the sea and the shortest distance
in town found over al directions, These distances can be simply measured
as shown in Fig, 9, or averaged over a 30°-wide range.
“This option is useful for initial back-of-cnvelupe calculations and we.
need go no Further if the conservative wind loads this produces are
acceptable. The code states that this option should be used with the
standard method when the orientation of the building is unknown
(§2.2.2.3) but in 3.5.3 we will see how to get around this restriction by
using the hybrid method uf 83.4.2.
Figure 9 illustrates hat the most onerous values of parameters may not
be coincident when taken inespective of direction, giving a conservative
result.
‘Typically, Option 1 will give 15% higher loads that Option 3, but this
will depend on how the parameters combine. Near u west-facing coast
there will be very little conservatism, but near an east-facing coast there
may be up to 30% conservatism. We will see in 3.3.2 how to recover this
conservatism with the least effort.
2.1.5.32 Option 2— Orthogonal load cases, standard method
Option 2 is the method intended for calculating the standard orthogonal
load cases hy hand. Ta this case, we consider the range of wind direction
445° either side of the orthogonal axes of the building, We take the most
onerous value of each parameter within cuch of these ranges, as directed
by $2.1.1.2 and §2.2.2.3 (Note).
Figure 10 shows one of the four orthogonal load eases for the building
in Fig. 9— wind normal to the south-west-facing eaves. The most
Shortest Closest
distance
in tom
Fig. 9. Option 1 — Inespeetse of direction
22 1 Wind loading: à practical guide to BS 6399-2
4 10. Option 2 — Orthogonal lod cose
‘onerous values of parameters may still not be coincident within the 90%
Wide range, giving a result that is still conservative, but is less
Conservative than Option 1. Again the shortest distances may be simply
measured as shown, or averaged over a 30F-wide range.
2.1.5.3.3 Option 3—Twelve 30°-wide sectors, directional method
‘Option 3 is the most complex option but it produces the lowest design
Joads by eliminatin all conservatism. In this case we consider twelve 30%
wide sectors of wind direction and we take the average value of each
parameter in each 30*-wide sector, as explained above.
Figure 11 shows this option applied for a mean wind direction of
= 00". Now there is no remaining conservatism, since the values of all
Parameters correspond to the same mean wind direction.
‘Option 3 requires twelve times more ealsulations than Option | and
vee times more than Option 2. Jn addition, the pressure coefficient zones
in the directional method are more complicated than in the standard
Method. For these reasons, is not intended for routine hand calculations,
although we may choose 10 check one or {wo directions that may be
<rical to the design. The repetitive nature ofthese calculations is more
ased spreadsheets or programs such ax BREVe or
suited to computer:
BREWS (see 3.14)
2.5.34 Hybrid options
‘We shall see in 32 that $3.42 allows us to apply Option 3
(directional method) wind speeds withthe Option 2 (standard method)
Commentary | 23
Average
distance
Average tosea
distance
Fig. 11. Option 3 — AT aride sector. centred on p =00
pressure coefficients to eliminate most of the conservatism. Whe
effective wind speeds are obtained automatically by the BREE
program, the risk of accidental error is removed and the optimal
design may be obtained with the minimum of calculation. We shall
also sec how to use Equation 29 instead of Table 4 Tor the Factor S lo
climinate conservatism at sites near the boundary of towns.
| Clause 3.4.4. allows us to use Option 1 or 2 wind speeds with the
directional pressure coefficients of Option 3, There are not many
applications where this would be useful, but one is when you have à
standard building. design that you wish to use in a number of
locations or to align in any direction. This option will not be
described in detail. IF you want to use it, follow the guidance that
applies to the directional method, but use a single value of wind
speed for all directions Ñ
216 | Topography (822.22)
5 6399-2 implements the effect of topography during Step 2— Wind
climate in the standard method and during Step 3 Site exposure the
directional methods. But the tes for significance and the definitions oft
topographic dimensions are common to both methods, whic a
topo dime (0 both methods, which is why they
PS 6399-2 takes the crm “topography? to mean hills, i
I cits andescopments These scene the mean wind spend mer
4 10. Option 2 — Orthogonal lod cose
‘onerous values of parameters may still not be coincident within the 90%
Wide range, giving a result that is still conservative, but is less
Conservative than Option 1. Again the shortest distances may be simply
measured as shown, or averaged over a 30F-wide range.
2.1.5.3.3 Option 3—Twelve 30°-wide sectors, directional method
‘Option 3 is the most complex option but it produces the lowest design
Joads by eliminatin all conservatism. In this case we consider twelve 30%
wide sectors of wind direction and we take the average value of each
parameter in each 30*-wide sector, as explained above.
Figure 11 shows this option applied for a mean wind direction of
= 00". Now there is no remaining conservatism, since the values of all
Parameters correspond to the same mean wind direction.
‘Option 3 requires twelve times more ealsulations than Option | and
vee times more than Option 2. Jn addition, the pressure coefficient zones
in the directional method are more complicated than in the standard
Method. For these reasons, is not intended for routine hand calculations,
although we may choose 10 check one or {wo directions that may be
<rical to the design. The repetitive nature ofthese calculations is more
ased spreadsheets or programs such ax BREVe or
suited to computer:
BREWS (see 3.14)
2.5.34 Hybrid options
‘We shall see in 32 that $3.42 allows us to apply Option 3
(directional method) wind speeds withthe Option 2 (standard method)
Commentary | 23
Average
distance
Average tosea
distance
Fig. 11. Option 3 — AT aride sector. centred on p =00
pressure coefficients to eliminate most of the conservatism. Whe
effective wind speeds are obtained automatically by the BREE
program, the risk of accidental error is removed and the optimal
design may be obtained with the minimum of calculation. We shall
also sec how to use Equation 29 instead of Table 4 Tor the Factor S lo
climinate conservatism at sites near the boundary of towns.
| Clause 3.4.4. allows us to use Option 1 or 2 wind speeds with the
directional pressure coefficients of Option 3, There are not many
applications where this would be useful, but one is when you have à
standard building. design that you wish to use in a number of
locations or to align in any direction. This option will not be
described in detail. IF you want to use it, follow the guidance that
applies to the directional method, but use a single value of wind
speed for all directions Ñ
216 | Topography (822.22)
5 6399-2 implements the effect of topography during Step 2— Wind
climate in the standard method and during Step 3 Site exposure the
directional methods. But the tes for significance and the definitions oft
topographic dimensions are common to both methods, whic a
topo dime (0 both methods, which is why they
PS 6399-2 takes the crm “topography? to mean hills, i
I cits andescopments These scene the mean wind spend mer
24 À Wind loading: a practical guide to BS 6399-2
Commentary 1 25
the summit or crest, because the streamlines arc squeezed together
fas shown in Fig. 12. The effect on the wind gusts of Fig, 7 is that the
Mean component Vo increases, but the gust component gx Y is
unchanged. The increase in speed near Ihe crest can be very large,
{even for small man made embankiments. Close to the ground around
the foot of stecp topography the wind speed can decrease, but BS
6399-2 makes no allowance for this.
2.1.6.1 Test for significance (82.2.2.2.1)
1° "Ihe first step is to decide whether the topography is significant, as
determined by the following two criteria,
1. Upwind slope. Ths must be greater than 1:20to be significant. Draw
os section through the steepest part of the hill and measure the
Slope. Because most natural features curve gently atthe foot and
stas in Fig. 12, the slope should be measured over the middle half
ofthe slope, fe. from between } to ¿ol the height of the bill. I the
Tength ofthe upwind slope is greater than 1200, tren you ean avoid
unnecessary effort by following the advanced guidance in 35.1.1
2. Position of the site, This enterion is defined in Figure 7. For bills
and ridges and the upwind slope of escarpments and cif the sie
must be above halfway up from Foot to erest because most of the
ice in speed occurs over the top half of the feature.
Downwind of escarpments (slope less than 0-3) the site must be
‘within 1,5 slope lengths of the erest. Downwind of cliffs (slope
restos than 0:3) the site must be within five cliff eights of the
crest (sce Note below).
Note: There are two typographical errors in Figure 7(b) of BS 6399.
1997, '5 x slope length ify > 0-3" should be 'S x slope height if
jy 20:3" and * Dowuwind slope Yo >005" should be
*Downwind stope Yo < 005".
1 If both eriteria are met, then topography is significant and we need
to complete the method that has been in use since 1986.
O Ir cither eriterion is nat met, topography is not significant at the
site and no further action is required.
As it is the up
i ras wae
21.6.2 Topographic dimensions ($2,222.24)
When topography is significant you are required to measure
dimensions from cross-sections throug fe
: fons trough he feature defined in gu
for cach wind direction of interest. fined in Hee 8
Determine these parameters:
O the base altitude upwind of the featur
che feature Ay in metres above
O the site altitude As in metres above mean sea level
a E position of the crest or summit
the position of the site X upwind (negative) or downwi
(positive) of the crest PAR ee en
© the height of the feature, taken from the base altitude to the
O the length of the downwind sk
pel Coty wed when “al
° ne ee Ly (only used when ‘hill’ or
Oe iene ors Seema ts gee ales
“The main difficultis in iting the i
The ficulties in Ming the dimensions to real topographic features
Sing the line through th
he upwind slope to obtain the slope length,
sic shuld bo done otero mile. bl af ds sly do
described above, _
2. fixing the position the slope reduces a
y the position of the erest when the slope r gradually
fing U when the slope reduces gradually at
Both these difficulties are illustrated in Example 2, More detailed
guidance on ft
guidance on lng the topographic dimensions to hills of complex shape
Commentary 1 25
the summit or crest, because the streamlines arc squeezed together
fas shown in Fig. 12. The effect on the wind gusts of Fig, 7 is that the
Mean component Vo increases, but the gust component gx Y is
unchanged. The increase in speed near Ihe crest can be very large,
{even for small man made embankiments. Close to the ground around
the foot of stecp topography the wind speed can decrease, but BS
6399-2 makes no allowance for this.
2.1.6.1 Test for significance (82.2.2.2.1)
1° "Ihe first step is to decide whether the topography is significant, as
determined by the following two criteria,
1. Upwind slope. Ths must be greater than 1:20to be significant. Draw
os section through the steepest part of the hill and measure the
Slope. Because most natural features curve gently atthe foot and
stas in Fig. 12, the slope should be measured over the middle half
ofthe slope, fe. from between } to ¿ol the height of the bill. I the
Tength ofthe upwind slope is greater than 1200, tren you ean avoid
unnecessary effort by following the advanced guidance in 35.1.1
2. Position of the site, This enterion is defined in Figure 7. For bills
and ridges and the upwind slope of escarpments and cif the sie
must be above halfway up from Foot to erest because most of the
ice in speed occurs over the top half of the feature.
Downwind of escarpments (slope less than 0-3) the site must be
‘within 1,5 slope lengths of the erest. Downwind of cliffs (slope
restos than 0:3) the site must be within five cliff eights of the
crest (sce Note below).
Note: There are two typographical errors in Figure 7(b) of BS 6399.
1997, '5 x slope length ify > 0-3" should be 'S x slope height if
jy 20:3" and * Dowuwind slope Yo >005" should be
*Downwind stope Yo < 005".
1 If both eriteria are met, then topography is significant and we need
to complete the method that has been in use since 1986.
O Ir cither eriterion is nat met, topography is not significant at the
site and no further action is required.
As it is the up
i ras wae
21.6.2 Topographic dimensions ($2,222.24)
When topography is significant you are required to measure
dimensions from cross-sections throug fe
: fons trough he feature defined in gu
for cach wind direction of interest. fined in Hee 8
Determine these parameters:
O the base altitude upwind of the featur
che feature Ay in metres above
O the site altitude As in metres above mean sea level
a E position of the crest or summit
the position of the site X upwind (negative) or downwi
(positive) of the crest PAR ee en
© the height of the feature, taken from the base altitude to the
O the length of the downwind sk
pel Coty wed when “al
° ne ee Ly (only used when ‘hill’ or
Oe iene ors Seema ts gee ales
“The main difficultis in iting the i
The ficulties in Ming the dimensions to real topographic features
Sing the line through th
he upwind slope to obtain the slope length,
sic shuld bo done otero mile. bl af ds sly do
described above, _
2. fixing the position the slope reduces a
y the position of the erest when the slope r gradually
fing U when the slope reduces gradually at
Both these difficulties are illustrated in Example 2, More detailed
guidance on ft
guidance on lng the topographic dimensions to hills of complex shape
26 | Wind loading: a practical guide to BS 6399-2
Example 2. Topographic dimensions for SX429527
M ass
amp detrnines the toporapic dimensions ora set SX 429527 hs
xa ana unos in you Sound Only wind is = 7
rng erent icon we et sa xi
this
modest
07, 180° and 270 ave show
For the reader.
com HA T
om
Ayaam
(6) Wind region 2707, West
Commentary 1 27
lts often dificult to decido where the rest ies when slope changes gradually. In
case (4) for y =270" the position has been shown at about the upwind limit of the
reasonable range of position
Note: The downwind slop length Ly is measured benseen where the downwind slope
Antersect with the height ofthe crest Z and the terrain base altitude x. Yan
intersection point les of the plotted profe, the value van be determined from
Lye
2.1.6.3 ‘Effective’ slope dimensions (§2.2.2.2.4)
BS 6399-2 expects you to determine the value of the upwind slope vy
and downwind slope vp by dividing the slope height Z by the upwind and
downwind slope length Ly and Lp respectively. You will need to measure
the slope directly when the slope length extends past the end of the cross-
section, as shown in Example 2 above.
The speed-up of wind speed is proportional to the upwind slope vy
only up tu a maximum slope of y =0-3. Slopes in the range 0-05 < vy
<03 are designated ‘shallow’ and slopes steeper than Yy=03 are
designated ‘stexp".
You should determine the effective slope parameters as instructed
by 8222.24:
© effective slope: Ye= vu when shallow and ye =03 when steep
© effective slope length: L.=Ly when shallow and Ly =Z/03
when steep.
2.1.6.4 The location factor, + (§2.2.2.2.5)
The value of the topographie location factor s depends on the position
Of the site relative to the crest and whether the topography is modelled as.
hillfidge (Figure 8fa)) or clifflescarpment (Figure 8(b)).
2.1641. Hillsandridges
The hillridge model applies when the downwind slope sp > 0-05.
You should look up the value of s in Figure 9a) or Nb) corresponding to
the height above ground of your building # and its position from the crest
X. Be very careful, hecause each axis of these figures is expressed as a
différent rato:
© The y-axisis the ratio of height above ground to the effective slope
length: He.
The x-axis for negative values:
Example 2. Topographic dimensions for SX429527
M ass
amp detrnines the toporapic dimensions ora set SX 429527 hs
xa ana unos in you Sound Only wind is = 7
rng erent icon we et sa xi
this
modest
07, 180° and 270 ave show
For the reader.
com HA T
om
Ayaam
(6) Wind region 2707, West
Commentary 1 27
lts often dificult to decido where the rest ies when slope changes gradually. In
case (4) for y =270" the position has been shown at about the upwind limit of the
reasonable range of position
Note: The downwind slop length Ly is measured benseen where the downwind slope
Antersect with the height ofthe crest Z and the terrain base altitude x. Yan
intersection point les of the plotted profe, the value van be determined from
Lye
2.1.6.3 ‘Effective’ slope dimensions (§2.2.2.2.4)
BS 6399-2 expects you to determine the value of the upwind slope vy
and downwind slope vp by dividing the slope height Z by the upwind and
downwind slope length Ly and Lp respectively. You will need to measure
the slope directly when the slope length extends past the end of the cross-
section, as shown in Example 2 above.
The speed-up of wind speed is proportional to the upwind slope vy
only up tu a maximum slope of y =0-3. Slopes in the range 0-05 < vy
<03 are designated ‘shallow’ and slopes steeper than Yy=03 are
designated ‘stexp".
You should determine the effective slope parameters as instructed
by 8222.24:
© effective slope: Ye= vu when shallow and ye =03 when steep
© effective slope length: L.=Ly when shallow and Ly =Z/03
when steep.
2.1.6.4 The location factor, + (§2.2.2.2.5)
The value of the topographie location factor s depends on the position
Of the site relative to the crest and whether the topography is modelled as.
hillfidge (Figure 8fa)) or clifflescarpment (Figure 8(b)).
2.1641. Hillsandridges
The hillridge model applies when the downwind slope sp > 0-05.
You should look up the value of s in Figure 9a) or Nb) corresponding to
the height above ground of your building # and its position from the crest
X. Be very careful, hecause each axis of these figures is expressed as a
différent rato:
© The y-axisis the ratio of height above ground to the effective slope
length: He.
The x-axis for negative values:
28 | Wind loading: a practical guide to BS 6399-2
Commentary | 29
upwind of the rest isthe ratio of position o the upwind slope
Nength: X/La.
© downwind of the exes isthe ratio of position to the downwind
slope length: X.
2164.2. Cliffsand escorpments
‘The clifffescarpment model applies when the downwind slope vin <
6.05. Ie follows that cliffs and escarpments are significant only when the
wind blows up the slope. So for winds down the slope of cliffs and
Escarpments you should ignore the topography and use the site altitud
‘Ds, When the site is on the upwind slope, you should obtain the value of s
in exactly the same way as for hills and ridges. When the site ison the Nat
downwind slope, you should look up the value of s in Figure /Ofa) or
10(b) corresponding to the height above ground of the building H and its
position from the crest X, both expressed as the ratio ofthe effective slope
length Le.
Note: Figures 9 und 10 use height above ground H and not effective
height He.
Jf you prefer tabular data to graphical data you may use Tables
6.1, G2 and G.3, interpolating as requi lace of Figures 9
‘and 10. For computer-based spreadsheet calculations you should use
‘Equations G.1 to G.10, and check that you have implemented them
correctly by comparing the results with the tables.
2.1.6.5 Key topography values (822223)
Al this effort leads to key values of just two parameters:
1. the effective slope ve for cach wind direction, and
2. the location factor s for each wind direction and each height above
ground.
These are applied differently in the standard and the directional methods,
as described in 2.3.3 and 2.4.5, later.
2.1.7. Thewind load chain
ES 6399-2 works in the logical and consistent chain: wind speed
dynamic pressure — surface pressure - surface loads —+ overall loads.
‘This method is independent of structural material and form, but the
structural form controls how the surface loads accumulate into toads in
the structural elements.
2.17.1 Dynamic pressure (§2.1.2, $311.2)
‘The dynamic pressure g represents the kinetic energy of the wind and is
used as a factor on the pressure coefficients C, to scale the surface
pressures p, thus: pq Cp
Dynamic pressure is proportional to the square of 1
Ve when NI Ihe ny ale So
{he constant 613 in Equations 4 and 16 represents p/2. Wis not a
a quay In eset Ma never appa rs ld
oun su thou an src pre coe
The steps you must take to obtain the dynamic pressure are described i
OEA Iynamic pressure are described in
2.172 External, internal and net surf
. nd net surface pressures ($2.13, 63.14
Che ofthe rm srs for oxemal noma! a net reso
that we are interested only in the action of the wind pressure field around
the building on the surface envelope of the building, Pressure is a scalar
ai at any point in space it acts equally in all directions as illustrated in
ig, 13@)-Close toa surface, the pressures at adjacent points, as shown in
Fig. 13(b), act in concert to produce a stress normal (0 the surface, as
shown in Fig. 13(0). Ñ
2.17.21 External surface pressure
‘The effect of building shape on the pressure acting on the external
surfaces is given by the external pressure coefficient, Cpe
A positive pressure coetfici that energy has heen
extracted from the wind. Cpe= +1 implies that 100% of the energy
has been extracted, so that pressure coefficients greater than +1 are
not possible in uniform flow. But in atmospheric wind, speed
increases with height so il is possible to exceed Cye=-+1 if the
NI
=e
Ne 7
YN SI
EL
(a) ©
Fig. 1%. Surface pressure
Commentary | 29
upwind of the rest isthe ratio of position o the upwind slope
Nength: X/La.
© downwind of the exes isthe ratio of position to the downwind
slope length: X.
2164.2. Cliffsand escorpments
‘The clifffescarpment model applies when the downwind slope vin <
6.05. Ie follows that cliffs and escarpments are significant only when the
wind blows up the slope. So for winds down the slope of cliffs and
Escarpments you should ignore the topography and use the site altitud
‘Ds, When the site is on the upwind slope, you should obtain the value of s
in exactly the same way as for hills and ridges. When the site ison the Nat
downwind slope, you should look up the value of s in Figure /Ofa) or
10(b) corresponding to the height above ground of the building H and its
position from the crest X, both expressed as the ratio ofthe effective slope
length Le.
Note: Figures 9 und 10 use height above ground H and not effective
height He.
Jf you prefer tabular data to graphical data you may use Tables
6.1, G2 and G.3, interpolating as requi lace of Figures 9
‘and 10. For computer-based spreadsheet calculations you should use
‘Equations G.1 to G.10, and check that you have implemented them
correctly by comparing the results with the tables.
2.1.6.5 Key topography values (822223)
Al this effort leads to key values of just two parameters:
1. the effective slope ve for cach wind direction, and
2. the location factor s for each wind direction and each height above
ground.
These are applied differently in the standard and the directional methods,
as described in 2.3.3 and 2.4.5, later.
2.1.7. Thewind load chain
ES 6399-2 works in the logical and consistent chain: wind speed
dynamic pressure — surface pressure - surface loads —+ overall loads.
‘This method is independent of structural material and form, but the
structural form controls how the surface loads accumulate into toads in
the structural elements.
2.17.1 Dynamic pressure (§2.1.2, $311.2)
‘The dynamic pressure g represents the kinetic energy of the wind and is
used as a factor on the pressure coefficients C, to scale the surface
pressures p, thus: pq Cp
Dynamic pressure is proportional to the square of 1
Ve when NI Ihe ny ale So
{he constant 613 in Equations 4 and 16 represents p/2. Wis not a
a quay In eset Ma never appa rs ld
oun su thou an src pre coe
The steps you must take to obtain the dynamic pressure are described i
OEA Iynamic pressure are described in
2.172 External, internal and net surf
. nd net surface pressures ($2.13, 63.14
Che ofthe rm srs for oxemal noma! a net reso
that we are interested only in the action of the wind pressure field around
the building on the surface envelope of the building, Pressure is a scalar
ai at any point in space it acts equally in all directions as illustrated in
ig, 13@)-Close toa surface, the pressures at adjacent points, as shown in
Fig. 13(b), act in concert to produce a stress normal (0 the surface, as
shown in Fig. 13(0). Ñ
2.17.21 External surface pressure
‘The effect of building shape on the pressure acting on the external
surfaces is given by the external pressure coefficient, Cpe
A positive pressure coetfici that energy has heen
extracted from the wind. Cpe= +1 implies that 100% of the energy
has been extracted, so that pressure coefficients greater than +1 are
not possible in uniform flow. But in atmospheric wind, speed
increases with height so il is possible to exceed Cye=-+1 if the
NI
=e
Ne 7
YN SI
EL
(a) ©
Fig. 1%. Surface pressure
30 À Wind loading: a practical guide to BS 6399-2
kinetic energy comes from a level above the reference height. As BS
{6309-2 always uses the top of the building as the reference height,
there are no values greater than +1 in any of the tables. À negative
pressure coefficient implies thatthe kinctie energy of the wind has
Merensei. This occurs where the wind accelerates around the
Comers, caves, verges nd ridges of buildings —around the edges.
‘These suctions are theoretically unlimited in value, hut the highest
in BS 6399-2 is Cpe = -2:73 in Table 34 and oceurs inthe upwind
‘comer at the verge of a —15" pitch root with the wind at 30" to the
eaves.
2.1.7.2.2 Internal surface pressure
“The intemal surface pressure of buildings is controlled by the
distribution of external pressures and by the size and positions of
apenings into the building. In principle, i is possible to calcule the
Eternal pressure from the net flow of air in and out of the building, but in
practice it is sufficient to assume standard values of the interna pressure
Coefficient Cy for vesious building forms. The effect of a dominant
opening is an exception to this rulo as we will sce later.
2.17.2.3 Netsurfoce pressure
“The net surface pressure is given by the difference between internal and
external surface pressures and is therefore a stress normal to the surface,
<quivalent to a uniformly distributed load (UBL). When both sides of a
surface are exposed to the wind, e.g. for canopy roofs or boundary walls,
itis often more convenient 10 give the net pressure directly in terms of the
net pressure coefficient, Cp
‘The steps required 10 obtain the surface pressures are described in 2.5
und 2,6, later.
2.73 Surface friction (§2.1.3.8, §3.1.34)
"The flow of wind along a surface creates a shear stress in the direction
‘of the wind by the action of surface friction. This is represented by a
frictional drag coefficient C; that acts on the area of the surface swept by
the wind, as shown in Fig, 14. This shear stress is very much smaller than
the normal pressure stresses, but can accumulate 10 a significant load
when the area of building swept by the wind is very large or when the
Surface is corrugated. While significant friction Toads may occur with
Tong-span low-rise buildings, the contribution from friction is nearly
always small, The steps required to obtain the ficcion Yoads are described
in 2.6.14, later.
Commentary | 31
Hig. 14. Print drag
2.1.7.4. Face loads and overall loads (§2.1.35,
{§3.1.3.2, 63.1.3.3) rn
‘The wind load on any face of the building i
18 is obtained by sum
the noma and ar seis on e ae, with a le for de
ize of the equivalent static gust. This allowance, which is given by the
diagonal dimension a of the face, is a i inthe sida
o . is applied different
nd decora aid: From how ON we shal cal ithe sve
and i his the ‘size
The overall loads are determined by
y summing the face loads tha
Inside te size fe, tut wi an alt 19 ration 1e sun
ur the non-simullaneous action of gusts on the windward and leeward
leew
22. This process is described in 2.6.1, later
21:73 Component loads ($225, 21.32)
„98 63992 provides only a general method fr desemining the loads on
structural components because tis requires knowledge ofthe structural
| onda on clin panc ae found by summing the net muse
ressure over the panel (Equations 6 and 21). Th i
the panel controls the size effect. The diagonal dimension ot
‚od in elements hat support the iting envelopes, hosting ail
and purlins, are determined hy summing the stresses over the area that
atiracts Joads onto thé is the *
ate ie element. We call this the ‘tributary area’ (see
kinetic energy comes from a level above the reference height. As BS
{6309-2 always uses the top of the building as the reference height,
there are no values greater than +1 in any of the tables. À negative
pressure coefficient implies thatthe kinctie energy of the wind has
Merensei. This occurs where the wind accelerates around the
Comers, caves, verges nd ridges of buildings —around the edges.
‘These suctions are theoretically unlimited in value, hut the highest
in BS 6399-2 is Cpe = -2:73 in Table 34 and oceurs inthe upwind
‘comer at the verge of a —15" pitch root with the wind at 30" to the
eaves.
2.1.7.2.2 Internal surface pressure
“The intemal surface pressure of buildings is controlled by the
distribution of external pressures and by the size and positions of
apenings into the building. In principle, i is possible to calcule the
Eternal pressure from the net flow of air in and out of the building, but in
practice it is sufficient to assume standard values of the interna pressure
Coefficient Cy for vesious building forms. The effect of a dominant
opening is an exception to this rulo as we will sce later.
2.17.2.3 Netsurfoce pressure
“The net surface pressure is given by the difference between internal and
external surface pressures and is therefore a stress normal to the surface,
<quivalent to a uniformly distributed load (UBL). When both sides of a
surface are exposed to the wind, e.g. for canopy roofs or boundary walls,
itis often more convenient 10 give the net pressure directly in terms of the
net pressure coefficient, Cp
‘The steps required 10 obtain the surface pressures are described in 2.5
und 2,6, later.
2.73 Surface friction (§2.1.3.8, §3.1.34)
"The flow of wind along a surface creates a shear stress in the direction
‘of the wind by the action of surface friction. This is represented by a
frictional drag coefficient C; that acts on the area of the surface swept by
the wind, as shown in Fig, 14. This shear stress is very much smaller than
the normal pressure stresses, but can accumulate 10 a significant load
when the area of building swept by the wind is very large or when the
Surface is corrugated. While significant friction Toads may occur with
Tong-span low-rise buildings, the contribution from friction is nearly
always small, The steps required to obtain the ficcion Yoads are described
in 2.6.14, later.
Commentary | 31
Hig. 14. Print drag
2.1.7.4. Face loads and overall loads (§2.1.35,
{§3.1.3.2, 63.1.3.3) rn
‘The wind load on any face of the building i
18 is obtained by sum
the noma and ar seis on e ae, with a le for de
ize of the equivalent static gust. This allowance, which is given by the
diagonal dimension a of the face, is a i inthe sida
o . is applied different
nd decora aid: From how ON we shal cal ithe sve
and i his the ‘size
The overall loads are determined by
y summing the face loads tha
Inside te size fe, tut wi an alt 19 ration 1e sun
ur the non-simullaneous action of gusts on the windward and leeward
leew
22. This process is described in 2.6.1, later
21:73 Component loads ($225, 21.32)
„98 63992 provides only a general method fr desemining the loads on
structural components because tis requires knowledge ofthe structural
| onda on clin panc ae found by summing the net muse
ressure over the panel (Equations 6 and 21). Th i
the panel controls the size effect. The diagonal dimension ot
‚od in elements hat support the iting envelopes, hosting ail
and purlins, are determined hy summing the stresses over the area that
atiracts Joads onto thé is the *
ate ie element. We call this the ‘tributary area’ (see
32 | Wind loading: a practical guide to BS 6399-2
Commentary 1 33
In order to maintain independence of material and structural
form, there are no longer any references to “cladding” or ‘structure’,
effect is taken to be solely dependent on the
1 dimension of the loaded or tributary area, following the
fuincipte of the equivalent static gust described in 2.11. This has
"moved a source of frequent disputes between designers and
building contol, eg. whether a shecting rail or purlin qualifies. as
fue”. ft is possible tu use a diagonal dimension
hat is larger than the tributary arca when there is significant load
sharing between structural elements. Detailed guidance an these
aspects is given later in 34.2.
2.2 Dynamic classification
2.2.1. Purpose and method
The main purpose of the dynamic classification is as a ‘signpost
procedure 10 exclude buildings that need to be designed by fully dynamic
Methods, But as a bonus, the procedure extends the scope of BS 6399-2 10
huildings that are mildly dynamic, i.e. buildings where the dynamic
‘component of stress docs nat excead 25% of the static stress
The full classification method, developed especially for BS 6399-2,
turned out to be as complex as the fully dynamic design methods it sought
tu avoid, soit was deemed unsuitable for adoption into BS 6399-2 until it
had heen dramatically simplified. Unfortunately the simplifications
reduce the precision of the method.
Twas simplified in two stages as follows.
Liu, by restricting the dynamic characteristics of the building o
he natural frequency and damping of the first mode uf vibration.
his is sill ton complicated to be applied as an initial ‘signpost’
procedure, so wis confined to Annex C.
2. Secondly hy assuming that:
he characteristics of the building are typical
à the fundamental natural frequency is function of height
(Equation C4) and
‘© the vibration mode shape is fi
ar in the along-wind direction.
“This simplified method provides the model for the dynamics of a
building shawn in Fig. LS, used for the ‘signpost’ procedure in the
body ofthe code ($/.6). Here the building is represented as a sigid
body able 10 rotate about its base, but is restrained by a spring and
Samper system. The spring controls the natural frequency of the
building while the damper controls the structural damping-
i
ATI
Pig. 15. Dynamic clasificasian model
ne ae sie eae oll
wilding, it technically breaches the frst objective in 1.4, but there
way to avoid this " FA e in
2.2.2 The signpost procedure ($1.6)
sa nt ine ens ee rc
Sie hm ae
ee
ciento Ti is very nie, nes the cee fs
ae
22.2.1 Building-type factor,
22211 Basis 1 Ky ($161)
“The building-type factor is a measure of the building's abi
energy. Small values of Ky imply large damping and smaller motion.
Examination of Annex C shows that the value is given by:
Ry = 1/32 @ (C7
where € is the structural damping as a fraction of critical damping.
222.12 Application
5 Selet a value of buildingtype factor from Table J. The values
cover mt ames ii comple seed
hat the range may be extended in the future by valu i
the structural codes. di iii
Commentary 1 33
In order to maintain independence of material and structural
form, there are no longer any references to “cladding” or ‘structure’,
effect is taken to be solely dependent on the
1 dimension of the loaded or tributary area, following the
fuincipte of the equivalent static gust described in 2.11. This has
"moved a source of frequent disputes between designers and
building contol, eg. whether a shecting rail or purlin qualifies. as
fue”. ft is possible tu use a diagonal dimension
hat is larger than the tributary arca when there is significant load
sharing between structural elements. Detailed guidance an these
aspects is given later in 34.2.
2.2 Dynamic classification
2.2.1. Purpose and method
The main purpose of the dynamic classification is as a ‘signpost
procedure 10 exclude buildings that need to be designed by fully dynamic
Methods, But as a bonus, the procedure extends the scope of BS 6399-2 10
huildings that are mildly dynamic, i.e. buildings where the dynamic
‘component of stress docs nat excead 25% of the static stress
The full classification method, developed especially for BS 6399-2,
turned out to be as complex as the fully dynamic design methods it sought
tu avoid, soit was deemed unsuitable for adoption into BS 6399-2 until it
had heen dramatically simplified. Unfortunately the simplifications
reduce the precision of the method.
Twas simplified in two stages as follows.
Liu, by restricting the dynamic characteristics of the building o
he natural frequency and damping of the first mode uf vibration.
his is sill ton complicated to be applied as an initial ‘signpost’
procedure, so wis confined to Annex C.
2. Secondly hy assuming that:
he characteristics of the building are typical
à the fundamental natural frequency is function of height
(Equation C4) and
‘© the vibration mode shape is fi
ar in the along-wind direction.
“This simplified method provides the model for the dynamics of a
building shawn in Fig. LS, used for the ‘signpost’ procedure in the
body ofthe code ($/.6). Here the building is represented as a sigid
body able 10 rotate about its base, but is restrained by a spring and
Samper system. The spring controls the natural frequency of the
building while the damper controls the structural damping-
i
ATI
Pig. 15. Dynamic clasificasian model
ne ae sie eae oll
wilding, it technically breaches the frst objective in 1.4, but there
way to avoid this " FA e in
2.2.2 The signpost procedure ($1.6)
sa nt ine ens ee rc
Sie hm ae
ee
ciento Ti is very nie, nes the cee fs
ae
22.2.1 Building-type factor,
22211 Basis 1 Ky ($161)
“The building-type factor is a measure of the building's abi
energy. Small values of Ky imply large damping and smaller motion.
Examination of Annex C shows that the value is given by:
Ry = 1/32 @ (C7
where € is the structural damping as a fraction of critical damping.
222.12 Application
5 Selet a value of buildingtype factor from Table J. The values
cover mt ames ii comple seed
hat the range may be extended in the future by valu i
the structural codes. di iii
34 À Wind loading: a practical guide 10 BS 6399-2
Commentary | 35
The ıypes of building covered by Table J range trom typical
masonry or timber-framed housing, where there is high structoral
damping (Ky = 0-5, € = 0-063), to welded unelad steel frames,
Were there is low structural damping (Rp = 8, € = 0-004). IF you
‘Camat find a description of your building among the types given in
Sable 1 and do not have an independent estimate of damping 6, you
should compare the expected dynamic behaviour of your building
‘grins the typical buildings listed inthe table. You may interpolate
between the tabulated values if appropriate. For example, if yours is
a ight poral-framed building, but it has more than a few intemal
walls, then Ky = 1-5 is a reasonable value 10 take.
1222 Dynamic augmentation factor, C (6161, 162)
22221 Bosis
“The dynamic augmentation factor C, expresses the addiional dynamic
component of stress in the structural members a5 a fraction of he static
Stress, Equivalent dynamic loads are obtained by multiplying the static
Toads by LC, in Equations 7 and 22.
22222 Application
D Look up the value of dynamic augmentation factor C in Figure 3
fom the value of Ko and the building height H. Interpolate
between the lines if you are using an intermediate value of Ki
Alternatively, you may use Equation C2
D Chock the value of €, against the Timits of applicability of BS
6399-2 shown in Figure 3.
O If C.>0.25 or H>300m the methods of BS 6399-2 are not
applicable and full dynamic design methods should he used
© When C,>0-1, the signpost procedure becomes increasingly
conservative and Anırex € should be used.
“The full classification procedure in Annex € is described in more detail in
32.
2.2.3 Limits to the dynamic classification
Because the classification method is simplified into the model
shown in Fig, 15, it cannot be used to identify all non-typical
buildings that may be dynamic. If the structural form of your
building is so unusual that the fundamental mode of vibration
‘near from its hase, then the classification meihod may not be
applicable. In this case, you must make your own assessment of how
dynamic the building is likely to be.
M is also possible to represent other forms af
cantilever structure by this model, such as the cantilever roof of a
randetnd, In this cue the method of Anner C can bo applied
y changes and the signpost method in $7-6 can he applied
by subsitutng the cantilever span forthe building height A.
23 Wind climate
2. En The wind climate model
wind climate model accounts for five aspects of wind ch
relate specifically to the United Kingdom: " namen
geogrpical variation across he UK
Sie aude
wind rection
ime of year (or temporuy structures)
risk of exceedance. ” >
by depressions generated in the North Atlantic and includes the
effect of frontal thunderstorms and squalls that occur in these
by tormadocs or other thermally generated events, such as lee-waves
‘Tornadoes are excluded because the risk to an individual building is
approximately one twentieth of the design storm event and it is
pate cue sae tales ona
extremely rare. snes
actions’ when the design risk is very much smaller than the standard
pr @ a These ‘accidental actions’, which are
he scope of BS 6399-2, wi
2.3.2. Basicwind speed (§2.2.1, §3.2.1, Annex
EE Mr MR RNA nes
‘The basie wind speed Y, is defined as the mean hourly wind speed with
un anal ih fensa 00.02, cap of wind ration a
m above flat open terrain at sea level which extends at least 100km in
all directions. The annual risk of exceedance Q=002 isthe same as the
previous "once in 50-year wind” (see 2.36.1). Note thatthe basic wind
speed is now a mean value. > Note tha ihe ease wind
Commentary | 35
The ıypes of building covered by Table J range trom typical
masonry or timber-framed housing, where there is high structoral
damping (Ky = 0-5, € = 0-063), to welded unelad steel frames,
Were there is low structural damping (Rp = 8, € = 0-004). IF you
‘Camat find a description of your building among the types given in
Sable 1 and do not have an independent estimate of damping 6, you
should compare the expected dynamic behaviour of your building
‘grins the typical buildings listed inthe table. You may interpolate
between the tabulated values if appropriate. For example, if yours is
a ight poral-framed building, but it has more than a few intemal
walls, then Ky = 1-5 is a reasonable value 10 take.
1222 Dynamic augmentation factor, C (6161, 162)
22221 Bosis
“The dynamic augmentation factor C, expresses the addiional dynamic
component of stress in the structural members a5 a fraction of he static
Stress, Equivalent dynamic loads are obtained by multiplying the static
Toads by LC, in Equations 7 and 22.
22222 Application
D Look up the value of dynamic augmentation factor C in Figure 3
fom the value of Ko and the building height H. Interpolate
between the lines if you are using an intermediate value of Ki
Alternatively, you may use Equation C2
D Chock the value of €, against the Timits of applicability of BS
6399-2 shown in Figure 3.
O If C.>0.25 or H>300m the methods of BS 6399-2 are not
applicable and full dynamic design methods should he used
© When C,>0-1, the signpost procedure becomes increasingly
conservative and Anırex € should be used.
“The full classification procedure in Annex € is described in more detail in
32.
2.2.3 Limits to the dynamic classification
Because the classification method is simplified into the model
shown in Fig, 15, it cannot be used to identify all non-typical
buildings that may be dynamic. If the structural form of your
building is so unusual that the fundamental mode of vibration
‘near from its hase, then the classification meihod may not be
applicable. In this case, you must make your own assessment of how
dynamic the building is likely to be.
M is also possible to represent other forms af
cantilever structure by this model, such as the cantilever roof of a
randetnd, In this cue the method of Anner C can bo applied
y changes and the signpost method in $7-6 can he applied
by subsitutng the cantilever span forthe building height A.
23 Wind climate
2. En The wind climate model
wind climate model accounts for five aspects of wind ch
relate specifically to the United Kingdom: " namen
geogrpical variation across he UK
Sie aude
wind rection
ime of year (or temporuy structures)
risk of exceedance. ” >
by depressions generated in the North Atlantic and includes the
effect of frontal thunderstorms and squalls that occur in these
by tormadocs or other thermally generated events, such as lee-waves
‘Tornadoes are excluded because the risk to an individual building is
approximately one twentieth of the design storm event and it is
pate cue sae tales ona
extremely rare. snes
actions’ when the design risk is very much smaller than the standard
pr @ a These ‘accidental actions’, which are
he scope of BS 6399-2, wi
2.3.2. Basicwind speed (§2.2.1, §3.2.1, Annex
EE Mr MR RNA nes
‘The basie wind speed Y, is defined as the mean hourly wind speed with
un anal ih fensa 00.02, cap of wind ration a
m above flat open terrain at sea level which extends at least 100km in
all directions. The annual risk of exceedance Q=002 isthe same as the
previous "once in 50-year wind” (see 2.36.1). Note thatthe basic wind
speed is now a mean value. > Note tha ihe ease wind
36 | Wind loading: a practical guide to BS 6399-2
Commentary 1 37
This definition provides compatibility with the requirements uf ISO
and the draft Eurocode EC. It also enables other Standards to extract
Mind speed data required for fully dynamic design methods. As a
Consequence, the change from mean to gust values uccurs later, in the
emain and building factor (2.4.5). The geographical variation of the basic
wind speed Va is given by the contour map of Figure 6.
Contours are marked in Figure 6 at intervals of 1 ms, ranging from
2010 in the south Midlands lo 31 m/s north of the Shetlands. This
implies that design wind loads increase by a factor of 24 from south 10
nom across the UK, because loads are proportional to the square of the
wind speed.
The values in Figure 6 come from an analysis of the
Meteorological Office records for all the storms in the period
1970 to 1980 at the SO hest-exposcd anemograph situations.” From
the locations of these anemometess, shown in Fig, 16 (except for
Jersey), we see thut each controlled the value of basic wind speed
ver a radius of about SOkm. The analysis is described in some
detail in Annex A. Although data taken over only 11 years were
used, the storm analysis? is significantly more reliable than previous
“analyses of annual maxima, even for the longest record. However,
there remained the question whether the 11-year period accurately
represents the long-term wind climate. Before the data were
“adopted, the values for twelve of the stations were re anulysed for
the 21-year period 1970-1990 and no significant differences were
Found.
Further guidance on UK basic wind specds is given in 3.1.5.1
2322 Application
M Read the value of basic wind speed Y, at the Jocation of the site
from Figure 6:
© Imerpolate for V, between contours
© Because no contour is shown off the southern coast, you will
‘nee to extrapolate from the 22 m/s contour towards 23 mis for
Sites near the Kent and West Sussex consts. Take Va = 22:5
ms for the Hostings-Folkestone-Dover-Ramsgaic-Margate
coastline,
NA NE
NK
Ei
NL kan 1 nr
wt,
à Mu
OV
SQ | SR] ss
NIET SYT 67
Fig. 16. Location of nemometers contribuin to ma of basic wind speed
2.3.3 Altitude factor (§2.2.2.2)
d spoed map, but the scale of Fi
i ped me, igure 6 is too small to distinguish high
Commentary 1 37
This definition provides compatibility with the requirements uf ISO
and the draft Eurocode EC. It also enables other Standards to extract
Mind speed data required for fully dynamic design methods. As a
Consequence, the change from mean to gust values uccurs later, in the
emain and building factor (2.4.5). The geographical variation of the basic
wind speed Va is given by the contour map of Figure 6.
Contours are marked in Figure 6 at intervals of 1 ms, ranging from
2010 in the south Midlands lo 31 m/s north of the Shetlands. This
implies that design wind loads increase by a factor of 24 from south 10
nom across the UK, because loads are proportional to the square of the
wind speed.
The values in Figure 6 come from an analysis of the
Meteorological Office records for all the storms in the period
1970 to 1980 at the SO hest-exposcd anemograph situations.” From
the locations of these anemometess, shown in Fig, 16 (except for
Jersey), we see thut each controlled the value of basic wind speed
ver a radius of about SOkm. The analysis is described in some
detail in Annex A. Although data taken over only 11 years were
used, the storm analysis? is significantly more reliable than previous
“analyses of annual maxima, even for the longest record. However,
there remained the question whether the 11-year period accurately
represents the long-term wind climate. Before the data were
“adopted, the values for twelve of the stations were re anulysed for
the 21-year period 1970-1990 and no significant differences were
Found.
Further guidance on UK basic wind specds is given in 3.1.5.1
2322 Application
M Read the value of basic wind speed Y, at the Jocation of the site
from Figure 6:
© Imerpolate for V, between contours
© Because no contour is shown off the southern coast, you will
‘nee to extrapolate from the 22 m/s contour towards 23 mis for
Sites near the Kent and West Sussex consts. Take Va = 22:5
ms for the Hostings-Folkestone-Dover-Ramsgaic-Margate
coastline,
NA NE
NK
Ei
NL kan 1 nr
wt,
à Mu
OV
SQ | SR] ss
NIET SYT 67
Fig. 16. Location of nemometers contribuin to ma of basic wind speed
2.3.3 Altitude factor (§2.2.2.2)
d spoed map, but the scale of Fi
i ped me, igure 6 is too small to distinguish high
38 1 Wind loading: a practical guide ta BS 6399-2
Commentary | 39
ds
shown in Example 3, but overesimnted if the site altitude was Tower
Example 3. Effect of altitude at Coalville (SK4214)
Met
vata
J val of in spect a Cosivi aioe 10) ist by u
ee lan, Shaun Wir nd Elm, and hee ae ll
tower atte veces
he ad wn spe mg. e measured site wind pu
GE we de ine ode ator fore site (SF RD co ive he
A yee se ab eens). The sane precdure wes apto Ue ole vo
shes aa nthe ge above ”
: ives Ve=21-5m for Cole, we
She ue wind speed map, Pigne 6, gives Va =21
cn ty Seen te res son vals we get Hy 21a and
tha good match the map. u
Ti ‘altitude factor for Coalville ix 5,=1-150, giving a site wind speed of
yee Shs Dia mean of he ste wid spent othe re
y Wea ham. Tis la undrimat tse Conse abou
507 ih ad
__ ——
The altitude factor was derived from Caton’s analysis? of mean
hourly wind speeds in the UK. Caton included sites on flat ground
with sites on shallow hills and this allows us to substitute the
altitude factor for the more complicated topography factor method
(see 2.1.6) when the hill is shallow or when the site is not close to
the crest, However, the methods for altitude and topography are not
exactly compatible and this leads to small differences in estimates of
wind speeds when a site is assessed by both methods.
2332 Application
‘When the test described in 2.1.6.1 indices that topogeaphy is:
© not significant, the altitude factor $, is calculated from the altitude
of the site Ay b
S,= 1+ 0:001As @ 0.25
9 signilicant, the altitude factor is calculated differently in the
standard and directional methods, as follows.
© Directional method
‘The altitude factor $, is calculated from the altitude ofthe base
of the topography Ar by
S.= 140-0014, @ (20
The effect of topography is addressed later by adding the
topagraphic increment in the equation for terrain and building
factor Sy (Equation 29) (see 2.45).
© Standard method
Sa is calculated from Equation S:
140-0015
5 1
sree of { Fr Set ($) (10, 11)
Application of Euation Sis illutated by Example 4, Using different
site Ag and terrain base Ar altitudes ensures that there is no ‘double
counting” when the topography component is added. I significant
topography is neglected, the larger value of site altitude only parily
compensates fur this omission, The topographie dimensions le ands were
explained earlier in 2.1.6. The weak effect of ground roughness on the
opography component is ignored in the standard method.
Commentary | 39
ds
shown in Example 3, but overesimnted if the site altitude was Tower
Example 3. Effect of altitude at Coalville (SK4214)
Met
vata
J val of in spect a Cosivi aioe 10) ist by u
ee lan, Shaun Wir nd Elm, and hee ae ll
tower atte veces
he ad wn spe mg. e measured site wind pu
GE we de ine ode ator fore site (SF RD co ive he
A yee se ab eens). The sane precdure wes apto Ue ole vo
shes aa nthe ge above ”
: ives Ve=21-5m for Cole, we
She ue wind speed map, Pigne 6, gives Va =21
cn ty Seen te res son vals we get Hy 21a and
tha good match the map. u
Ti ‘altitude factor for Coalville ix 5,=1-150, giving a site wind speed of
yee Shs Dia mean of he ste wid spent othe re
y Wea ham. Tis la undrimat tse Conse abou
507 ih ad
__ ——
The altitude factor was derived from Caton’s analysis? of mean
hourly wind speeds in the UK. Caton included sites on flat ground
with sites on shallow hills and this allows us to substitute the
altitude factor for the more complicated topography factor method
(see 2.1.6) when the hill is shallow or when the site is not close to
the crest, However, the methods for altitude and topography are not
exactly compatible and this leads to small differences in estimates of
wind speeds when a site is assessed by both methods.
2332 Application
‘When the test described in 2.1.6.1 indices that topogeaphy is:
© not significant, the altitude factor $, is calculated from the altitude
of the site Ay b
S,= 1+ 0:001As @ 0.25
9 signilicant, the altitude factor is calculated differently in the
standard and directional methods, as follows.
© Directional method
‘The altitude factor $, is calculated from the altitude ofthe base
of the topography Ar by
S.= 140-0014, @ (20
The effect of topography is addressed later by adding the
topagraphic increment in the equation for terrain and building
factor Sy (Equation 29) (see 2.45).
© Standard method
Sa is calculated from Equation S:
140-0015
5 1
sree of { Fr Set ($) (10, 11)
Application of Euation Sis illutated by Example 4, Using different
site Ag and terrain base Ar altitudes ensures that there is no ‘double
counting” when the topography component is added. I significant
topography is neglected, the larger value of site altitude only parily
compensates fur this omission, The topographie dimensions le ands were
explained earlier in 2.1.6. The weak effect of ground roughness on the
opography component is ignored in the standard method.
40 | Wind loading: a practical guide to BS 6399-2
Commentary | 41
see
Example 4. Standard method for altitude factor at erest of hill or
ridge
Wind
Site on crest, $=1.0,
e
Encciva slope, v= 0.18,
3,=100m
100m.
ea vel
11000145 Oso ats
Sumgreier OF} 140.001 #12 VS 140 OL IDD 1-2%0-18% 1-0 «1-28
2.3.4 Direction factor (82.2.2.3)
234.1 Basis
“The 1983 basic wind speed analysis’"!° gave values of the factor with
an equal tisk in each 30°-wide sector of direction. The drafting panel for
BS 6399-2 took the pragmatic view that use of direetional factors should
not aller the highest desig i
Toads obtained irrespective of direction. So the
Values of directional factor Sa in Table 3 were obtained by scaling up
these factors to give Sq 1-00 in the direction of strongest winds,
p= 240"
2342 Application
“Application ofthe direcion factor depends on the option you chose from
2.1.8 carlier, and whether you have converted the wind direction from north
{into wind angle from normal to the building axis 9 (see Fig. 8).
© Option 1—Irrespective of direction, standard method. IF the
ortentation of the building is unknown or you wish to ignore
direction, as in Fig. 9, take the value Sy-= 1-00. IF your site is near
a coast see the guidance in 3.3.2.
© Option 2—Orthogonal lond cases, standard method. When the
orientation of the building is known, take the largest val
tation of m, ta cat value of
from Table in the range of wind rection a einer side othe
irectional normal to the building fac, as in Fig. 10 and Fig. 17
© Option 3—Twelve MP=wide sectors, directional method. To
work in tems of wind direction from north y, take the value of Sy
in Table 3 for cach 30°-wid sector, as in Fig, 11. To work in terms
of wind angle from normal to the building axis 0, interpolate
between the values in Table 3 as shown in Fig. 18
5-00
5,=085
ig, 17. Direction factors for Option 2, selected for 0 — 07
S=080 S,=076
5-07
$,=089 $,=083
Fig. 18, Direction fotos for Option 3, interpolated In 0 = 4"
Commentary | 41
see
Example 4. Standard method for altitude factor at erest of hill or
ridge
Wind
Site on crest, $=1.0,
e
Encciva slope, v= 0.18,
3,=100m
100m.
ea vel
11000145 Oso ats
Sumgreier OF} 140.001 #12 VS 140 OL IDD 1-2%0-18% 1-0 «1-28
2.3.4 Direction factor (82.2.2.3)
234.1 Basis
“The 1983 basic wind speed analysis’"!° gave values of the factor with
an equal tisk in each 30°-wide sector of direction. The drafting panel for
BS 6399-2 took the pragmatic view that use of direetional factors should
not aller the highest desig i
Toads obtained irrespective of direction. So the
Values of directional factor Sa in Table 3 were obtained by scaling up
these factors to give Sq 1-00 in the direction of strongest winds,
p= 240"
2342 Application
“Application ofthe direcion factor depends on the option you chose from
2.1.8 carlier, and whether you have converted the wind direction from north
{into wind angle from normal to the building axis 9 (see Fig. 8).
© Option 1—Irrespective of direction, standard method. IF the
ortentation of the building is unknown or you wish to ignore
direction, as in Fig. 9, take the value Sy-= 1-00. IF your site is near
a coast see the guidance in 3.3.2.
© Option 2—Orthogonal lond cases, standard method. When the
orientation of the building is known, take the largest val
tation of m, ta cat value of
from Table in the range of wind rection a einer side othe
irectional normal to the building fac, as in Fig. 10 and Fig. 17
© Option 3—Twelve MP=wide sectors, directional method. To
work in tems of wind direction from north y, take the value of Sy
in Table 3 for cach 30°-wid sector, as in Fig, 11. To work in terms
of wind angle from normal to the building axis 0, interpolate
between the values in Table 3 as shown in Fig. 18
5-00
5,=085
ig, 17. Direction factors for Option 2, selected for 0 — 07
S=080 S,=076
5-07
$,=089 $,=083
Fig. 18, Direction fotos for Option 3, interpolated In 0 = 4"
42 1 Wind loading: a practical guide to BS 6399-2
Commentary | 43
2.35. Seasonal factor (§2.2.2.4, Annex D)
235.1 Basis
"The 1983 basic wind speed analysi gave values of the seasonal
factor Se that give the standard annual risk Q=0-02 in cach month
Cotresponding values for any two- or four-month period and for ths six
month summer period April to September were calculated from these
monthly values, assuming no. ‘correlation between months. The risk of
strong wind is much lower in Ihe summer months (han in winter.
The UK Building Regulations specify a minimum annual risk of Q — 0.02
forall permanent buildings, requiring that the season! factor is Ss = 0. For
this repson the relevant clauses and values were placed in Annex D.
2352 Application
“The seasonal factor is only required when the building will be exposed
to the wind for specific sub-annual periods. In practice, this will mean
buildings in a temporary state of construction or renovation and
associated falsework, ar buildings that arc seasonally exposed. such as
portable buildings or temporary grandstands, for shori-pericd events.
© Permanent buildings. Always use the value Ss = 1-0.
(Temporary buildings. Select the value of Ss from Table D.1 that
is appropriate to the stating date and duration of the temporary
period ur exposure and use this in conjunction with an appropriate
value of the probability factor S, (see 2.3.6)
‘The smallest value in Table D.1, Ss=0-62, corresponds to the
month of July and reduces the design wind loads to only 38% of the
Toads for a permanent building. When you apply this value in
design, the corresponding risk during July is the same asthe risk to a
‘permanent structure at any time during the year. Therefore, safety is
ot compromised, provided that the period of exposure does not
extend beyond the assumed period, The value 5s=0-71 applies to
the two-month periods May/June and July/August, but if works
Scheduled for May/June extend over the four-month period May!
‘August the value rises to Sg=0-73 and wind loads increase by 6%
hy virtue of the increased length of exposure alone, If works
scheduled for September/October (Sx=085) extend through
December (S3~0-96), the wind loads increase by 28% due to the
change of season as weil as the length of exposure.
2.3.6 Probability factor ($2.2.
sae ae Peers
The datum value of annual risk, Q=0:02, is set by the definition of
basic wind speed in 2321. Thor a er oF &
interpret this value of risk: PA NS on
‚Annex D)
as a | in 50 chance of exccedance in any one year
as a mean recurrence interval of SO years, or
+ asa 63% chance of exceedance in any 50-year period.
Changing the way that the datum risk is described
Chang tum risk is described. from a high (63%)
— emphasizes that the relevant risk in design is the risk to the public
in every year ia the building is operational. Staring fom this datum
the paria factors on lad uc in be sural Sanda
achieve a risk of damage less than 0:0005 (1 in 2000) and a risk of fal
less than 10°* (1 in 10000) in each year of operation. où
ni probs mall ha lso changed delo improved
nowledge. Figure 19 shows the probability factor 5, in BS
6399-2 compared with previous practice in terms of the mean
recurso real, Bo mods ie $ 1.0 al th aan neral
of SO years, but the new BS 6399-2 model is more onerous for
ol Il fl | TD
PtH] HL |
Be al
al AA |
„AU LAU LU LEI
aan recur naval 10 (me)
Fig, 19. Probability madel compared with previous practice
Commentary | 43
2.35. Seasonal factor (§2.2.2.4, Annex D)
235.1 Basis
"The 1983 basic wind speed analysi gave values of the seasonal
factor Se that give the standard annual risk Q=0-02 in cach month
Cotresponding values for any two- or four-month period and for ths six
month summer period April to September were calculated from these
monthly values, assuming no. ‘correlation between months. The risk of
strong wind is much lower in Ihe summer months (han in winter.
The UK Building Regulations specify a minimum annual risk of Q — 0.02
forall permanent buildings, requiring that the season! factor is Ss = 0. For
this repson the relevant clauses and values were placed in Annex D.
2352 Application
“The seasonal factor is only required when the building will be exposed
to the wind for specific sub-annual periods. In practice, this will mean
buildings in a temporary state of construction or renovation and
associated falsework, ar buildings that arc seasonally exposed. such as
portable buildings or temporary grandstands, for shori-pericd events.
© Permanent buildings. Always use the value Ss = 1-0.
(Temporary buildings. Select the value of Ss from Table D.1 that
is appropriate to the stating date and duration of the temporary
period ur exposure and use this in conjunction with an appropriate
value of the probability factor S, (see 2.3.6)
‘The smallest value in Table D.1, Ss=0-62, corresponds to the
month of July and reduces the design wind loads to only 38% of the
Toads for a permanent building. When you apply this value in
design, the corresponding risk during July is the same asthe risk to a
‘permanent structure at any time during the year. Therefore, safety is
ot compromised, provided that the period of exposure does not
extend beyond the assumed period, The value 5s=0-71 applies to
the two-month periods May/June and July/August, but if works
Scheduled for May/June extend over the four-month period May!
‘August the value rises to Sg=0-73 and wind loads increase by 6%
hy virtue of the increased length of exposure alone, If works
scheduled for September/October (Sx=085) extend through
December (S3~0-96), the wind loads increase by 28% due to the
change of season as weil as the length of exposure.
2.3.6 Probability factor ($2.2.
sae ae Peers
The datum value of annual risk, Q=0:02, is set by the definition of
basic wind speed in 2321. Thor a er oF &
interpret this value of risk: PA NS on
‚Annex D)
as a | in 50 chance of exccedance in any one year
as a mean recurrence interval of SO years, or
+ asa 63% chance of exceedance in any 50-year period.
Changing the way that the datum risk is described
Chang tum risk is described. from a high (63%)
— emphasizes that the relevant risk in design is the risk to the public
in every year ia the building is operational. Staring fom this datum
the paria factors on lad uc in be sural Sanda
achieve a risk of damage less than 0:0005 (1 in 2000) and a risk of fal
less than 10°* (1 in 10000) in each year of operation. où
ni probs mall ha lso changed delo improved
nowledge. Figure 19 shows the probability factor 5, in BS
6399-2 compared with previous practice in terms of the mean
recurso real, Bo mods ie $ 1.0 al th aan neral
of SO years, but the new BS 6399-2 model is more onerous for
ol Il fl | TD
PtH] HL |
Be al
al AA |
„AU LAU LU LEI
aan recur naval 10 (me)
Fig, 19. Probability madel compared with previous practice
44 | Wind loading: a practical guide to BS 6399-2
Commentary 1 45
shorter periods (larger risks) and less onerous for longer periods
(smaller risks). This makes no difference to the normal building
designed to the datum risk Q=0-02, Nuclear installations are
jgned to a much smaller risk, Q= 10" *, ur a mean recurrence
interval of 10000 years, and the probability Factor for this risk falls
from Sp 138 in CP 3-V-2 10 Sp=1:26 in BS 6399-2 with
change of model.
2362 Application
O The concept of “temporary” is not applicable to any structure
exposed to the wind for more than one year. A detailed discussion
of the point is given in 3.35.
For permanent buildings and temporary buildings with general
public access, use the value S= 1.
For temporary buildings with active safety management, choose a
level of annual risk O appropriate for the use of the building.
Determine the value of probability factor from Equation D.1. To
‘avoid errors, you may find it helpful to compare this value with the
values quoted in Annex D.
Note: Ihe values of Sy for various levels of risk Q given in Annex D are
merely examples and should not be taken as design
recommendations.
od 99
ana
We will see in 2.5.5.3, how the probability factor should be used in the
special ease of buildings with dominant openings that can be open or
on
2.37 Sitewind speed (§2.2.2, §3.2.2)
1 Basis
“The site wind speed Y, is an intermediate value of wind speed that was
troduced to separate the factors dependent on UK climate from other
uote, IL represents the mean hourly wind speed 10m above ground level
Me site wit the desired annual risk of exceedance in each desired wind
direction, assuming the site is surrounded by typical UK countryside for at
least 100km in all directions.
© Inthe directional method, the values of site wind speed Y, rel
ai A win relate to
a fat site and any effect of topography is added later in the terra
and building factor S, (see 2.4.5)
+ Tn the standard method, the values of site wind speed V,
À jte wind speed V, include
any effects of topography, but the simplification ignores the weak
effect of ground roughness on the topography component, This is
discussed later in 24,5.
2312 Application
The values of site wind speed V, are obtained from Equati
site A ation 8. Simpl
muhipiy the basic wind speed V y the aude, dicton, scasoral an
probability Factors
O Option 1—Irrespective of direction, standard method. The
jon factor is taken as S¿= 1-00 and you obtain a single value
di
of site wind speed Vs.
M Option 2—Orthogonal loud cases, standard method. You
‘obtain four values of site wind speed V,, one for each of the values
of direction factor S in Fig, 17 for the respective orthogonal cases.
O Option 3—Twelve M'wide sectors, dire
obtain twelve val
sector of wind
normal to the
mal method. You
s of site wind speed V,, one for each 30°-wide
ion from north y or each wind angle from
axis 6.
‘These options are demonstrated later in the worked examples of 4.2.
23.7.3 Overseas sites
‘The Foreword of BS 6399-2 makes i si
for u only at ses in he UR. bo ny coum wi as dee
assessed to the current UK code. In this case you need to replace Ihe UK
site wind speed with the hourly mean wind speed 10m above open level
ground appropriate to the geographical location and altitude of the si
‘This a tandand meteoric! parameter, nd should ths be avaiable
rom the local meteorological authority
oe
done
pes
Commentary 1 45
shorter periods (larger risks) and less onerous for longer periods
(smaller risks). This makes no difference to the normal building
designed to the datum risk Q=0-02, Nuclear installations are
jgned to a much smaller risk, Q= 10" *, ur a mean recurrence
interval of 10000 years, and the probability Factor for this risk falls
from Sp 138 in CP 3-V-2 10 Sp=1:26 in BS 6399-2 with
change of model.
2362 Application
O The concept of “temporary” is not applicable to any structure
exposed to the wind for more than one year. A detailed discussion
of the point is given in 3.35.
For permanent buildings and temporary buildings with general
public access, use the value S= 1.
For temporary buildings with active safety management, choose a
level of annual risk O appropriate for the use of the building.
Determine the value of probability factor from Equation D.1. To
‘avoid errors, you may find it helpful to compare this value with the
values quoted in Annex D.
Note: Ihe values of Sy for various levels of risk Q given in Annex D are
merely examples and should not be taken as design
recommendations.
od 99
ana
We will see in 2.5.5.3, how the probability factor should be used in the
special ease of buildings with dominant openings that can be open or
on
2.37 Sitewind speed (§2.2.2, §3.2.2)
1 Basis
“The site wind speed Y, is an intermediate value of wind speed that was
troduced to separate the factors dependent on UK climate from other
uote, IL represents the mean hourly wind speed 10m above ground level
Me site wit the desired annual risk of exceedance in each desired wind
direction, assuming the site is surrounded by typical UK countryside for at
least 100km in all directions.
© Inthe directional method, the values of site wind speed Y, rel
ai A win relate to
a fat site and any effect of topography is added later in the terra
and building factor S, (see 2.4.5)
+ Tn the standard method, the values of site wind speed V,
À jte wind speed V, include
any effects of topography, but the simplification ignores the weak
effect of ground roughness on the topography component, This is
discussed later in 24,5.
2312 Application
The values of site wind speed V, are obtained from Equati
site A ation 8. Simpl
muhipiy the basic wind speed V y the aude, dicton, scasoral an
probability Factors
O Option 1—Irrespective of direction, standard method. The
jon factor is taken as S¿= 1-00 and you obtain a single value
di
of site wind speed Vs.
M Option 2—Orthogonal loud cases, standard method. You
‘obtain four values of site wind speed V,, one for each of the values
of direction factor S in Fig, 17 for the respective orthogonal cases.
O Option 3—Twelve M'wide sectors, dire
obtain twelve val
sector of wind
normal to the
mal method. You
s of site wind speed V,, one for each 30°-wide
ion from north y or each wind angle from
axis 6.
‘These options are demonstrated later in the worked examples of 4.2.
23.7.3 Overseas sites
‘The Foreword of BS 6399-2 makes i si
for u only at ses in he UR. bo ny coum wi as dee
assessed to the current UK code. In this case you need to replace Ihe UK
site wind speed with the hourly mean wind speed 10m above open level
ground appropriate to the geographical location and altitude of the si
‘This a tandand meteoric! parameter, nd should ths be avaiable
rom the local meteorological authority
oe
done
pes
46 | Wind loading: a practical guide ty BS 6399-2
Commentary À 47
Example 5. Converting a basic gust
site
Method
Take the gust speed 10 be the same as Ihe effective wind speed Vo or
100 in country tera, in the standard method.
© Take the corresponding reference value of temain and building factor S from
Table forthe reference height Hy» 10m, couriry terrain and he distance to
sea of the site
© Solve Equation 12 for Va by dividing the gust speed by Se.
he effet of aude and distance to sea is assumed o be included in Ih reference
hat wind speed forthe ste. So you shoul! use the same distance 1 sea I sep >
Bone that you use fore design values terrain and building Factor Sp forthe bold
Aescrbod in 245.
Example
‘Determine the equivalent site wind speed at an overseas si
inland, given a asc gust speed of 44 ms,
1:62 in Table 4 for He= 101m in country, more than
more than 100K
O Reference value ofS
100 km from the sea
Equivalent she wind speed, V.-=4471-62
amv
2.4 Site exposure
2.4.1. The exposure model
“The wind climate factors described in the previous section all relate to a
datum height of 10m above typical UK country terrain. This standard
datum is required because the character of che wind varies with height
above ground and changes as it moves over terrain of differing roughness.
‘The zone of wind that is affected by the rough ground surface is called the
atmospheric boundary layer (ABL) and this controls the site exposure, The
Variation with height of any parameter through the ABL is called a ‘profile’.
Figure 20 shows the equilibriurh profiles of mean wind speed,
which increase with increasing height above ground and decrease
With increasing ground roughness. Figure 21 shows the corre
ponding equilibrium profiles of turbulence, which decrease with
Frereasing height above ground and increase with increasing ground
roughness. Because the mean wind speed and the turbulence react in
opposite ways to the ground roughness, they reduce the effect of
ground roughness changes when combined by the peak, factor
Prethod (2.1.1) into the gust wind speed, as shown in Fig. 22.
Me in se ate
Fig. 20. Equilibrium proßles of mean wind speed
Tune nay cor
Fig. 21. Equilibrium profiles of nubulence
Here, the term “equilibrium® means that the ground 1
extends Far cnn upwind forte ADL To at completly. The
upwind extent of each kind of ground roughness is called the ‘fetch’
Whas only reconily been established that u Fech of over 100km
required to achieve complete equilibrium. so that the ABL is in a
constant state of adjustment to changes in ground roughness. The BS
6399-2 exposure model is a simplification of the model in the BRE
Commentary À 47
Example 5. Converting a basic gust
site
Method
Take the gust speed 10 be the same as Ihe effective wind speed Vo or
100 in country tera, in the standard method.
© Take the corresponding reference value of temain and building factor S from
Table forthe reference height Hy» 10m, couriry terrain and he distance to
sea of the site
© Solve Equation 12 for Va by dividing the gust speed by Se.
he effet of aude and distance to sea is assumed o be included in Ih reference
hat wind speed forthe ste. So you shoul! use the same distance 1 sea I sep >
Bone that you use fore design values terrain and building Factor Sp forthe bold
Aescrbod in 245.
Example
‘Determine the equivalent site wind speed at an overseas si
inland, given a asc gust speed of 44 ms,
1:62 in Table 4 for He= 101m in country, more than
more than 100K
O Reference value ofS
100 km from the sea
Equivalent she wind speed, V.-=4471-62
amv
2.4 Site exposure
2.4.1. The exposure model
“The wind climate factors described in the previous section all relate to a
datum height of 10m above typical UK country terrain. This standard
datum is required because the character of che wind varies with height
above ground and changes as it moves over terrain of differing roughness.
‘The zone of wind that is affected by the rough ground surface is called the
atmospheric boundary layer (ABL) and this controls the site exposure, The
Variation with height of any parameter through the ABL is called a ‘profile’.
Figure 20 shows the equilibriurh profiles of mean wind speed,
which increase with increasing height above ground and decrease
With increasing ground roughness. Figure 21 shows the corre
ponding equilibrium profiles of turbulence, which decrease with
Frereasing height above ground and increase with increasing ground
roughness. Because the mean wind speed and the turbulence react in
opposite ways to the ground roughness, they reduce the effect of
ground roughness changes when combined by the peak, factor
Prethod (2.1.1) into the gust wind speed, as shown in Fig. 22.
Me in se ate
Fig. 20. Equilibrium proßles of mean wind speed
Tune nay cor
Fig. 21. Equilibrium profiles of nubulence
Here, the term “equilibrium® means that the ground 1
extends Far cnn upwind forte ADL To at completly. The
upwind extent of each kind of ground roughness is called the ‘fetch’
Whas only reconily been established that u Fech of over 100km
required to achieve complete equilibrium. so that the ABL is in a
constant state of adjustment to changes in ground roughness. The BS
6399-2 exposure model is a simplification of the model in the BRE
48 | Wind loading: a practical guide to BS 6399-2
Commentary | 49
0 gu oca
Fig. 22. Esquliuium profiles of Ys gust speed
i tures! whi
Designer's Guide to Wind Loading of Building Struct
self is u simplification of a more sophisticated model of the ABL.
be how the equilibrium profiles of gust
‘Two main components describe how the eg
speed respond to changes in the roughness of the ground surface
1. Permanent obstacles displace the whole profile upwards and this
happens very quickly, within 100m in towns.
2. ‘The profile shape adjusts gradually from the ground upwards and
this happens very slowly.
i i oundary layer over a town
Figure 23 illusteates how the new boundary layı
inc wuly replacing
rows slowly upwards as the town fetch increases, slowly
Ae kan country boundary Taye. Note that a 30m high lng
that is 5 km inside a town will stick up through the town boundary
layor into wind that is still unaffected by the town.
ed in BS 6399-2 is fundamentally
“The way that these effects are addressed in BS 63 dan
different from CP 3-V-2, but they are expressed in the same ‘terrain and
building factor’. We will see ater that typical site exposures give similar
values in both Standards.
2.4.2. Ground roughness categories and fetch ($1.7.2, Annex E)
€ ground roughness
ying the principle of balanced effects o de.
Re Previa four rouliness categories could be reduced 10
Figure 24 shows how the transi
that describes how the wind speeds in the new town boundary layer
| decrease with fetch. About 50% of the effect eecurs in the first À km
sl | _—
Fok lon (on)
Fig. 23. Growth of imernal wn boundary layer
just two land categories, ‘country’ and ‘town’, plus the category ‘sea’
which is only used to determine the land fetch:
1. Sea-—This includes the sea and inland areas of water that are
larger than I km in extent and closer than 1 km to the site. These
d blue on OS 1:50000 Landranger series maps.
area of land that dues not qualify as town.
3. Town— This is any area of land built up to a plan density of 8%
and with an average level of roof tops at least Sm above ground
level that extends at least 100m upwind of the site. Urban arcas
axe coloured pink on OS 1:50000 Landranger series maps.
‘The previous four categories of CP 3-V-2 were a constant cause of
dispute and the minimum fetch of “at feast Lkm specified in SAS of
CP 3-V-2 was often ignored, leading to the kind of disparity demonstrated
in Example 1, eavlier. At first sight it might seem that reducing. the
number of categories would increase the steps between them, but the
model counters this by applying x gradual transition that depends on the
value of fetch after the change of ground roughness.
is achieved by a fetch factor
and 75% of the effect occurs by 10 km. The original model predicts
a slow asymplotic convergence to unity, but the simplified model is
forced to converge between Fetches of 50km and 100km. The sea=
country and country-town curves in Fig, 24 are the relevant two
transitions from a family predicted by the model. Downwind of a
town, the wind speed gradually increases hack to the country values
but, for simplicity, BS 6399-2 takes no account of this (but the full
Designer's Guide method in the design aid BREVe does (see 3.1.4)).
When the equilibrium gust speed profile for town terrain is
lt
ied by the fetch factor for a country-to-town change, the new
Commentary | 49
0 gu oca
Fig. 22. Esquliuium profiles of Ys gust speed
i tures! whi
Designer's Guide to Wind Loading of Building Struct
self is u simplification of a more sophisticated model of the ABL.
be how the equilibrium profiles of gust
‘Two main components describe how the eg
speed respond to changes in the roughness of the ground surface
1. Permanent obstacles displace the whole profile upwards and this
happens very quickly, within 100m in towns.
2. ‘The profile shape adjusts gradually from the ground upwards and
this happens very slowly.
i i oundary layer over a town
Figure 23 illusteates how the new boundary layı
inc wuly replacing
rows slowly upwards as the town fetch increases, slowly
Ae kan country boundary Taye. Note that a 30m high lng
that is 5 km inside a town will stick up through the town boundary
layor into wind that is still unaffected by the town.
ed in BS 6399-2 is fundamentally
“The way that these effects are addressed in BS 63 dan
different from CP 3-V-2, but they are expressed in the same ‘terrain and
building factor’. We will see ater that typical site exposures give similar
values in both Standards.
2.4.2. Ground roughness categories and fetch ($1.7.2, Annex E)
€ ground roughness
ying the principle of balanced effects o de.
Re Previa four rouliness categories could be reduced 10
Figure 24 shows how the transi
that describes how the wind speeds in the new town boundary layer
| decrease with fetch. About 50% of the effect eecurs in the first À km
sl | _—
Fok lon (on)
Fig. 23. Growth of imernal wn boundary layer
just two land categories, ‘country’ and ‘town’, plus the category ‘sea’
which is only used to determine the land fetch:
1. Sea-—This includes the sea and inland areas of water that are
larger than I km in extent and closer than 1 km to the site. These
d blue on OS 1:50000 Landranger series maps.
area of land that dues not qualify as town.
3. Town— This is any area of land built up to a plan density of 8%
and with an average level of roof tops at least Sm above ground
level that extends at least 100m upwind of the site. Urban arcas
axe coloured pink on OS 1:50000 Landranger series maps.
‘The previous four categories of CP 3-V-2 were a constant cause of
dispute and the minimum fetch of “at feast Lkm specified in SAS of
CP 3-V-2 was often ignored, leading to the kind of disparity demonstrated
in Example 1, eavlier. At first sight it might seem that reducing. the
number of categories would increase the steps between them, but the
model counters this by applying x gradual transition that depends on the
value of fetch after the change of ground roughness.
is achieved by a fetch factor
and 75% of the effect occurs by 10 km. The original model predicts
a slow asymplotic convergence to unity, but the simplified model is
forced to converge between Fetches of 50km and 100km. The sea=
country and country-town curves in Fig, 24 are the relevant two
transitions from a family predicted by the model. Downwind of a
town, the wind speed gradually increases hack to the country values
but, for simplicity, BS 6399-2 takes no account of this (but the full
Designer's Guide method in the design aid BREVe does (see 3.1.4)).
When the equilibrium gust speed profile for town terrain is
lt
ied by the fetch factor for a country-to-town change, the new
50 | Wind loading: a practical guide to BS 63992
Commentary 151
“—Á y PAY
> Cou TH
= | ll
Fig. 24, Ferch fatara
“fiective non-equilibrium profile is given by this new profile. Above
the intersection point the effective profile is unchanged from the
previous land profile. The height of the intersection point, 24m in
Fig. 25 corresponds to the height of the new town houndary layer at
ilar transition also occurs downwind of the
Skim in Fig. 23. A si
profile intersects the country profile as illustrated by the dashed
Curve in Fig. 25 for a 5 km fetch, Below the intersection point, the
coast
The datum calibration point is between CP 3-V-2 Ground
roughmess 2 and BS 6399-2 country terrain at 20km from the sea,
“able 1 shows that, provided we account for the factor of 178
difference between the definitions of basic gust wind speed in CP 3-
‘V-2and basic mean wind speed in BS 6399-2, then the site exposure
methods of the two Standards match well at 10m above ground for
the first three categories, BS 6399-2 is slightly more onerous at 30 m
above ground because it predicts thal this height is above the
interface of the new boundary layer. BS 6399-2 is also oncrous at
10m in “city centre” sites for the reason explained in the next
section,
BS 6399-2 shields the user from the complications of this roughness
change model by providing tables of factors for various values of fetch
and effective height.
: = /
“Town x Fetch factor
ee
at in so aor =
ig, 25. Effective profile af Ls gue speed for Shi fetch fou
Table 1. CP 3-V2 standard ground roughnese categories ercas In tern of BS 399.2
on Hn Ham Si
Caegory Tersin description
1 Finger og, Zimio e y OO
De Br
2 Mia fran, 20km sa won
cons a
32km ine tw, 1m ie aly na
um thee o er)
4 20m nem. 20m Woh bit na
Dim me a tag tat
EDS 6199 des not ascrme 27m high permanent obstructions in “Country” Aral
2422 Application
in order to select the appropriate ground roughness upwind of the
for each of the wind directions or orthogonal cases we need to distinguish
between the coastline (oF shoreline of lakes) and the boundaries of tows
The contin i obvious on any map. The urban houndary is the edge of
the pink zones on Ordnance Survey 1:50,000 Landranger series maps and
is sufficient to determine the distance to the nearest 100m, which is all
that is required, The groups of individual buildings, shown as black
rectangles on these: maps, are too small to qualify as town terrain, so
vidual buildings in country terrain.
Commentary 151
“—Á y PAY
> Cou TH
= | ll
Fig. 24, Ferch fatara
“fiective non-equilibrium profile is given by this new profile. Above
the intersection point the effective profile is unchanged from the
previous land profile. The height of the intersection point, 24m in
Fig. 25 corresponds to the height of the new town houndary layer at
ilar transition also occurs downwind of the
Skim in Fig. 23. A si
profile intersects the country profile as illustrated by the dashed
Curve in Fig. 25 for a 5 km fetch, Below the intersection point, the
coast
The datum calibration point is between CP 3-V-2 Ground
roughmess 2 and BS 6399-2 country terrain at 20km from the sea,
“able 1 shows that, provided we account for the factor of 178
difference between the definitions of basic gust wind speed in CP 3-
‘V-2and basic mean wind speed in BS 6399-2, then the site exposure
methods of the two Standards match well at 10m above ground for
the first three categories, BS 6399-2 is slightly more onerous at 30 m
above ground because it predicts thal this height is above the
interface of the new boundary layer. BS 6399-2 is also oncrous at
10m in “city centre” sites for the reason explained in the next
section,
BS 6399-2 shields the user from the complications of this roughness
change model by providing tables of factors for various values of fetch
and effective height.
: = /
“Town x Fetch factor
ee
at in so aor =
ig, 25. Effective profile af Ls gue speed for Shi fetch fou
Table 1. CP 3-V2 standard ground roughnese categories ercas In tern of BS 399.2
on Hn Ham Si
Caegory Tersin description
1 Finger og, Zimio e y OO
De Br
2 Mia fran, 20km sa won
cons a
32km ine tw, 1m ie aly na
um thee o er)
4 20m nem. 20m Woh bit na
Dim me a tag tat
EDS 6199 des not ascrme 27m high permanent obstructions in “Country” Aral
2422 Application
in order to select the appropriate ground roughness upwind of the
for each of the wind directions or orthogonal cases we need to distinguish
between the coastline (oF shoreline of lakes) and the boundaries of tows
The contin i obvious on any map. The urban houndary is the edge of
the pink zones on Ordnance Survey 1:50,000 Landranger series maps and
is sufficient to determine the distance to the nearest 100m, which is all
that is required, The groups of individual buildings, shown as black
rectangles on these: maps, are too small to qualify as town terrain, so
vidual buildings in country terrain.
52 | Wind loading: a practical guide to BS 6399-2
Commentary 1 53
© Option 1—Errespective of direction, standard method.
© You need the shortest distance to the sea and the shortest
distance in town in any direction, as shown in Fig. 9 earlier.
© These can be averaged over a 30"-wide sector.
© Option 2—Orthogonal lond cases, standard method.
© You need the shortest distance to the sea and the shortest
‘ance in town in the range of direction +45" either side of
each orthogonal case. This was shown for one case in Fig, 10
earlier,
© These can also be averaged over a 30'=wide sector.
A Option 3— directional method.
© You need the average distance to the sea und distance in town
in each of twelve Wide sectors af wind direction. This was
shown for one sector in Fig. 11 earlier.
+ For sites further than 20 km to the sea, simply count 10 km
grid square
+ For sites between 2km and 20k« to the sea estimate by
tenths of grid squares.
+ For sites closer than 2km, use a larger scale map or site
2.4.3 Reference height ond effective height (§1.7.3, §3.2.3)
243.1 Basis
We require design values of wind speed at reference heights A, above
ground that depend on the form of the building. The definition of
reference height is given in the key figure Lo each table of pressure
coefficients in BS 6399-2.
‘Within town terrain, the profiles are displaced upwards by the presence
of the upwi s, almost up to the average height of the roof tops
‘when the buildings are dense or typically spaced, but less when there are
large spaces between them. This displacement occurs very quickly, within
100m of the town boundary as illustrated in Fig, 26. While the profile
shape may not have changed significantly, the wind speed at any given
height above ground decreases because ofthis displacement. But as height
Wind
<A 1000
Fig. 26. Displacement height in towns
above ground increases and the displacement height becomes a smaller
proportion, the shelter provided by the displacement reduces.
The displacement height Hg, previously defined in Annex E, was
moved into the main clauses in 2002. ‘The displacement height Hu is
useful in breaking down the effective height rules of $7.7.3.3 into two
simpler stages, involving much Tess work when there is more than one
reference height to consider.
‘The displacement is set by the average height and spacing of upwind
buildings in terms of: j nn
L ‘The obstruction height, Hu.
2. The obstruction spacing, Xy
Note: Ihe 2002 amendment introduced X, as the symbol for obstruction
spacing to distinguish it from X, the symbol for fetch.
‘The obstruction height M, is not a now parameter. CP 3-V-2
assumed standard values of H, in Appendix A. The problem with a
standard value, such as H,= 10m for Ground roughness 3 in
CP 3-V-2, is the following.
+ If the building is 10m high and surrounded by two-storey
housing (4,=6m), it sticks up above roottop height and is
mare exposed than assumed,
IF the building is 10m high and surrounded by four-storey
buildings (21, — 12m), itis more sheltered than assumed.
1 bsrctonapacing XI a new pamınte. CP 3-V-2 aroma
that buildings in towns and cities were always close enough together
to give the full effect. But if there was an open space upwind of the
building, CP 3-V-2 required you to use the next lower category (ie.
Category 32 and Category 43).
_ BS 6399-2 addresses this problem by determining the
displacement from the obstruction height Ho and spacing Xy, then
Commentary 1 53
© Option 1—Errespective of direction, standard method.
© You need the shortest distance to the sea and the shortest
distance in town in any direction, as shown in Fig. 9 earlier.
© These can be averaged over a 30"-wide sector.
© Option 2—Orthogonal lond cases, standard method.
© You need the shortest distance to the sea and the shortest
‘ance in town in the range of direction +45" either side of
each orthogonal case. This was shown for one case in Fig, 10
earlier,
© These can also be averaged over a 30'=wide sector.
A Option 3— directional method.
© You need the average distance to the sea und distance in town
in each of twelve Wide sectors af wind direction. This was
shown for one sector in Fig. 11 earlier.
+ For sites further than 20 km to the sea, simply count 10 km
grid square
+ For sites between 2km and 20k« to the sea estimate by
tenths of grid squares.
+ For sites closer than 2km, use a larger scale map or site
2.4.3 Reference height ond effective height (§1.7.3, §3.2.3)
243.1 Basis
We require design values of wind speed at reference heights A, above
ground that depend on the form of the building. The definition of
reference height is given in the key figure Lo each table of pressure
coefficients in BS 6399-2.
‘Within town terrain, the profiles are displaced upwards by the presence
of the upwi s, almost up to the average height of the roof tops
‘when the buildings are dense or typically spaced, but less when there are
large spaces between them. This displacement occurs very quickly, within
100m of the town boundary as illustrated in Fig, 26. While the profile
shape may not have changed significantly, the wind speed at any given
height above ground decreases because ofthis displacement. But as height
Wind
<A 1000
Fig. 26. Displacement height in towns
above ground increases and the displacement height becomes a smaller
proportion, the shelter provided by the displacement reduces.
The displacement height Hg, previously defined in Annex E, was
moved into the main clauses in 2002. ‘The displacement height Hu is
useful in breaking down the effective height rules of $7.7.3.3 into two
simpler stages, involving much Tess work when there is more than one
reference height to consider.
‘The displacement is set by the average height and spacing of upwind
buildings in terms of: j nn
L ‘The obstruction height, Hu.
2. The obstruction spacing, Xy
Note: Ihe 2002 amendment introduced X, as the symbol for obstruction
spacing to distinguish it from X, the symbol for fetch.
‘The obstruction height M, is not a now parameter. CP 3-V-2
assumed standard values of H, in Appendix A. The problem with a
standard value, such as H,= 10m for Ground roughness 3 in
CP 3-V-2, is the following.
+ If the building is 10m high and surrounded by two-storey
housing (4,=6m), it sticks up above roottop height and is
mare exposed than assumed,
IF the building is 10m high and surrounded by four-storey
buildings (21, — 12m), itis more sheltered than assumed.
1 bsrctonapacing XI a new pamınte. CP 3-V-2 aroma
that buildings in towns and cities were always close enough together
to give the full effect. But if there was an open space upwind of the
building, CP 3-V-2 required you to use the next lower category (ie.
Category 32 and Category 43).
_ BS 6399-2 addresses this problem by determining the
displacement from the obstruction height Ho and spacing Xy, then
sat
Wind loading: a practical guide to BS 6399-2
Commentary 1 55
defining the effective height He as the reference height less the
displacement. Within the building layer BS 6399-2 sets a minimum
value to the effective wind specd by setting a lower limit to the
effective height of He= 04H.
“This is u change from CP 3-V-2, which allowed the wind speed
to decrease linearly through the building layer 10 a minimum
height of 2m. This change was necessary to account for the higher
wind speeds brought down to ground level by the flow around the
buildings. However, for buildings of a similar height to the
‘upwind neighbours, the most common situation, both Standards
give about the same 30% reduction in loads compared with an
‘open exposure.
2 Application
:ept for the simplest building, there are usually several reference
heights to consider. The corresponding effective heights A. are defined in
1.7.3.3 which is easicr to apply since the 2002 amendments adopted the
advice given previously in this guide.
o
o
Determine the reference heights, HL. The general rule is that the
reference height is taken to the highest point of the wall or roof:
© for walls, H, is the height above ground of the exves or the top
af the gable or the parapet
O for roofs, M, is the height above ground of the ridge for
duopitch roots or the high eaves for monopitch or troughed
roof.
Determine the obstruction height and separation, Ho and X, for
each orthogonal case or wind direction. Refer to Figure 2(c) for
the definitions of these parameters.
© In finding the average height of upwind obstructions, you
should concentrate unly on the rooftops of the principal
buildings and ignore any ground level cluter, such as garages,
sheds, outbuildings or boundary walls and fences.
© Take the average height of buildings upwind of the site over a
distance of about 100m upwind. In the standard method, a
sector 445° cher side of the normal to the building face
should he considered. In the directional method, each 30°
sector should be considered.
© Ifyou cannot find the exact value of Hq, estimate it from the
typical storey height of 3m given in the Nore to $J.7.2,.
O The breton separation, Ki tedster rom the upwind
face of the building to the nearest principal buildings.
© there are loo few bulking upwind lo Gnd a exsonable
average height (8% plan density, or 12 houses per hi
the buildings extend less than 100m upwind, then tl
for town terrain is not met and the site should be treated as
Country terrain with H~0. In this case, the site should have
been treated as Ground roughness 2 under CP 3-V-2.
© When the exact site details are unknown, refer to the guidance
in 38.2 later.
Determine the displacement height, Ha for each orthogonal case or
wind direction
© From Figure 2(c) the displacement height is given by:
08 Xo 2H
Hy = 4 1-21 02X) Ï 2e < Xo < Oto ©
o if X, 2 6H
© This is most easily obtained as follows,
1. Make a first estimate of displacement height from:
Ha= 12 He=02 Xo Ñ
2. Check that this value of Hy does not exceed 0-8 Ha;
if Ha > O-8 Hy then Hy—O-8 Ho
3. Check that this value of Hg is not less than 0:
if My <0 then Hy=0
Finally, determine the effective height H, for each reference
height.
© From $1.7.3,3, the effective height is given by:
He He
04H,
© This is most easily obtained as follows.
1. Estimate the effective height, He, by subtracting Ha from
the reference height Hy: He — Hy--H
2. Check that this value is not less than 0-4 2: Me <0-4 H,
then H¿=04 H,
He reno
243.3 Comments
‘The parameters far town terrain which give the standard displacement
Ha= 10m and were assumed by CP 3-V-2 for Category 3 are shown in
Wind loading: a practical guide to BS 6399-2
Commentary 1 55
defining the effective height He as the reference height less the
displacement. Within the building layer BS 6399-2 sets a minimum
value to the effective wind specd by setting a lower limit to the
effective height of He= 04H.
“This is u change from CP 3-V-2, which allowed the wind speed
to decrease linearly through the building layer 10 a minimum
height of 2m. This change was necessary to account for the higher
wind speeds brought down to ground level by the flow around the
buildings. However, for buildings of a similar height to the
‘upwind neighbours, the most common situation, both Standards
give about the same 30% reduction in loads compared with an
‘open exposure.
2 Application
:ept for the simplest building, there are usually several reference
heights to consider. The corresponding effective heights A. are defined in
1.7.3.3 which is easicr to apply since the 2002 amendments adopted the
advice given previously in this guide.
o
o
Determine the reference heights, HL. The general rule is that the
reference height is taken to the highest point of the wall or roof:
© for walls, H, is the height above ground of the exves or the top
af the gable or the parapet
O for roofs, M, is the height above ground of the ridge for
duopitch roots or the high eaves for monopitch or troughed
roof.
Determine the obstruction height and separation, Ho and X, for
each orthogonal case or wind direction. Refer to Figure 2(c) for
the definitions of these parameters.
© In finding the average height of upwind obstructions, you
should concentrate unly on the rooftops of the principal
buildings and ignore any ground level cluter, such as garages,
sheds, outbuildings or boundary walls and fences.
© Take the average height of buildings upwind of the site over a
distance of about 100m upwind. In the standard method, a
sector 445° cher side of the normal to the building face
should he considered. In the directional method, each 30°
sector should be considered.
© Ifyou cannot find the exact value of Hq, estimate it from the
typical storey height of 3m given in the Nore to $J.7.2,.
O The breton separation, Ki tedster rom the upwind
face of the building to the nearest principal buildings.
© there are loo few bulking upwind lo Gnd a exsonable
average height (8% plan density, or 12 houses per hi
the buildings extend less than 100m upwind, then tl
for town terrain is not met and the site should be treated as
Country terrain with H~0. In this case, the site should have
been treated as Ground roughness 2 under CP 3-V-2.
© When the exact site details are unknown, refer to the guidance
in 38.2 later.
Determine the displacement height, Ha for each orthogonal case or
wind direction
© From Figure 2(c) the displacement height is given by:
08 Xo 2H
Hy = 4 1-21 02X) Ï 2e < Xo < Oto ©
o if X, 2 6H
© This is most easily obtained as follows,
1. Make a first estimate of displacement height from:
Ha= 12 He=02 Xo Ñ
2. Check that this value of Hy does not exceed 0-8 Ha;
if Ha > O-8 Hy then Hy—O-8 Ho
3. Check that this value of Hg is not less than 0:
if My <0 then Hy=0
Finally, determine the effective height H, for each reference
height.
© From $1.7.3,3, the effective height is given by:
He He
04H,
© This is most easily obtained as follows.
1. Estimate the effective height, He, by subtracting Ha from
the reference height Hy: He — Hy--H
2. Check that this value is not less than 0-4 2: Me <0-4 H,
then H¿=04 H,
He reno
243.3 Comments
‘The parameters far town terrain which give the standard displacement
Ha= 10m and were assumed by CP 3-V-2 for Category 3 are shown in
56 | Wind loading: a practical guide (0 BS 6399-2
Commentary 1 57
Wind
Fig. 27. Equivalence between BS 63992 and CP 3 V-2 Category 3
Fig. 27. This implies (ha the effective heights for all buildings less than
16m tall will be set by the Tower limit 77,=0:4/1, in this terrain. In reas
of typical two-storey pitched-100f housing, we would expect #75 m
(3m storey height, counting the roof as halla storey) in which case the
lower limit He=0-4H, will apply to all buildings less than 10m tall
Typically, this shelter reduces structural loads to about 70% of the loads
for an isolated building in a town.
244 Division-by-parts (§2.2.3.2, §3.2.3.1)
2441 Basis
Previous practice in CP 3-V-2 allowed the walls of buildings to be
divided into ‘convenient parts’ and the Toads on each part determined
using the wind speed at the top of each part. This allowed wind loads to be
reduced down the height of the building. Unfortunately, lax definitions in
CP 3-V-2 of the meaning of ‘part allowed the rule to be used beyond the
Jimits of its applicability. Although only intended for overall structural
Toads, the rule was commonly applied to the design of cladding, despite
‘clear guidance® to the contrary dating frum 1974.
“The new rule sets a minimum height to the lowest part that depends on
the slendermess of the building. Application is restricted to ‘lateral loads’.
‘This is intended to exclude surface pressures, but the Standard does not
slate this sufficiently clear.
“The aerodynamics that lead tothe division-by-parts rule apply only
for the postive pressures on the middle region of windward walls Of
slender buildings, as defined in Figure 17. There is copious evidence
fiom full-scale and model-scale measurement thatthe suctions which
ceur on side and rear faces vary with horizontal distance from the
corner, but are almost uniform wih height and so depend solely on
the dynamic pressure atthe op ofthe building. To ilustrate this, Fig,
28 shows recent measurements" of the maximum suctions that occur
on the wall of « 6:1 tower when al possible wind directions are taken
Ha
nerves
ig. 28. Marimum suction om face of tall ter
into account. The small variation with height that dues occur acts in
the opos see ooh rule, wi higher ution towards the base of
the wal
The influence of the previous rule on overall loads and cladding
pressures is shown in Fig, 29, which shows the ratio of the loads
wen i ten at ds vo ofeach tre I he loads wi taken
at the eaves. By treating a single-storey building as one part the rule
some part the role
has no effect. The rule produces a maximum reduction of overall
loads to 71%, which occurs for four-storey buildings, But the effect
Commentary 1 57
Wind
Fig. 27. Equivalence between BS 63992 and CP 3 V-2 Category 3
Fig. 27. This implies (ha the effective heights for all buildings less than
16m tall will be set by the Tower limit 77,=0:4/1, in this terrain. In reas
of typical two-storey pitched-100f housing, we would expect #75 m
(3m storey height, counting the roof as halla storey) in which case the
lower limit He=0-4H, will apply to all buildings less than 10m tall
Typically, this shelter reduces structural loads to about 70% of the loads
for an isolated building in a town.
244 Division-by-parts (§2.2.3.2, §3.2.3.1)
2441 Basis
Previous practice in CP 3-V-2 allowed the walls of buildings to be
divided into ‘convenient parts’ and the Toads on each part determined
using the wind speed at the top of each part. This allowed wind loads to be
reduced down the height of the building. Unfortunately, lax definitions in
CP 3-V-2 of the meaning of ‘part allowed the rule to be used beyond the
Jimits of its applicability. Although only intended for overall structural
Toads, the rule was commonly applied to the design of cladding, despite
‘clear guidance® to the contrary dating frum 1974.
“The new rule sets a minimum height to the lowest part that depends on
the slendermess of the building. Application is restricted to ‘lateral loads’.
‘This is intended to exclude surface pressures, but the Standard does not
slate this sufficiently clear.
“The aerodynamics that lead tothe division-by-parts rule apply only
for the postive pressures on the middle region of windward walls Of
slender buildings, as defined in Figure 17. There is copious evidence
fiom full-scale and model-scale measurement thatthe suctions which
ceur on side and rear faces vary with horizontal distance from the
corner, but are almost uniform wih height and so depend solely on
the dynamic pressure atthe op ofthe building. To ilustrate this, Fig,
28 shows recent measurements" of the maximum suctions that occur
on the wall of « 6:1 tower when al possible wind directions are taken
Ha
nerves
ig. 28. Marimum suction om face of tall ter
into account. The small variation with height that dues occur acts in
the opos see ooh rule, wi higher ution towards the base of
the wal
The influence of the previous rule on overall loads and cladding
pressures is shown in Fig, 29, which shows the ratio of the loads
wen i ten at ds vo ofeach tre I he loads wi taken
at the eaves. By treating a single-storey building as one part the rule
some part the role
has no effect. The rule produces a maximum reduction of overall
loads to 71%, which occurs for four-storey buildings, But the effect
58 | Wind loading: a practical guide to BS 6399-2
Commentary | 59
E
Number free
ig. 29. Effect of previous division-by parts ale on lands
of the previous rule is unlimited if allowed to apply to surface
pressures, Le. lu cladding loads.
Because the codes for most other European countries do not
include a division-by-parts rule, it was deleted from the initial draft
of BS 6399-2 to assist the transition towards Eurocode 1. The rule
was reinstated with the limits of applicability clearly defined in
response to public comments. In this new Form, it was proposed by
the UK for inclusion in Eurocode 1 and accepted.
2442 Application
© The procedure fur dividing the walls of the building into parts
given in §2.2.3.2 should be suictly observed.
© Afthe height His less than or equal to the crosswind breath B,
the wall must be trated as one part and the reference height
Halt
© ihe height Fis greater than the crosswind breadth B, but less
than or equal to 28, the wall may be treated as two parts: the
lower part with HL, =8 and upper part with H,=H,
Af the height H is greater than 2B, the wall may be divided
1. One part of height B at the base, with H,
2. A second part of height at the top, with JJ, H and
3. Any number of horizontal strips between the top and base
parts, with 4,22, the top of each stip.
© When H> 2B, you are advised tu alga the horizontal strips 10
the storeys.
© In general, expect to obtain different parts at wind angle 8 —0*
(B=1) from the parts al 9=90° (B=W).
These parts may be used only for lateral loads, ie. storey loads,
shear and moment at any level and foundation Toads.
© The diagonal dimension a (see 2.6.2) for the size effect factor C,
should always be the diagonal of the whole loaded area being
considered.
O Surface pressures should always bo deter
‘of the walls, including any parapet.
ed u
ng A, o the top
See the worked example in 4.6 Tower and podium for an example of
division-by-parts.
2.4.5. Terrain and building factor (§2.2.3.3, $3.2.3.2)
245.1 Basls
In the directional method all the effects of site exposure are expressed
by the terrain and building factor Sp, But in the standard method, $,
represents the smallest size of gust and he size effeet is applied separately
by the factor Ca. The terrain and building factor therefore encapsulates
1. the profile of mean wind speed
2. the profile of turbulence a
3. the effect of ground rough
Given the number of variables, the general case could be very complex,
as illustrated in Fig. 30. Even though the site may be some distance into a
town
+ high enough above ground, at point ‘a’, the gust profile is
unchanged from the profile above the se
+ Tower down, at point *h’, the gust profile is controlled by the fetch
of open country
® only close to the rooftops, at point ‘c’, is the gust profile controlled
by the fetch of town.
Wind
aoe ‘County Town
Fig. 30. General case for ste exposure
Commentary | 59
E
Number free
ig. 29. Effect of previous division-by parts ale on lands
of the previous rule is unlimited if allowed to apply to surface
pressures, Le. lu cladding loads.
Because the codes for most other European countries do not
include a division-by-parts rule, it was deleted from the initial draft
of BS 6399-2 to assist the transition towards Eurocode 1. The rule
was reinstated with the limits of applicability clearly defined in
response to public comments. In this new Form, it was proposed by
the UK for inclusion in Eurocode 1 and accepted.
2442 Application
© The procedure fur dividing the walls of the building into parts
given in §2.2.3.2 should be suictly observed.
© Afthe height His less than or equal to the crosswind breath B,
the wall must be trated as one part and the reference height
Halt
© ihe height Fis greater than the crosswind breadth B, but less
than or equal to 28, the wall may be treated as two parts: the
lower part with HL, =8 and upper part with H,=H,
Af the height H is greater than 2B, the wall may be divided
1. One part of height B at the base, with H,
2. A second part of height at the top, with JJ, H and
3. Any number of horizontal strips between the top and base
parts, with 4,22, the top of each stip.
© When H> 2B, you are advised tu alga the horizontal strips 10
the storeys.
© In general, expect to obtain different parts at wind angle 8 —0*
(B=1) from the parts al 9=90° (B=W).
These parts may be used only for lateral loads, ie. storey loads,
shear and moment at any level and foundation Toads.
© The diagonal dimension a (see 2.6.2) for the size effect factor C,
should always be the diagonal of the whole loaded area being
considered.
O Surface pressures should always bo deter
‘of the walls, including any parapet.
ed u
ng A, o the top
See the worked example in 4.6 Tower and podium for an example of
division-by-parts.
2.4.5. Terrain and building factor (§2.2.3.3, $3.2.3.2)
245.1 Basls
In the directional method all the effects of site exposure are expressed
by the terrain and building factor Sp, But in the standard method, $,
represents the smallest size of gust and he size effeet is applied separately
by the factor Ca. The terrain and building factor therefore encapsulates
1. the profile of mean wind speed
2. the profile of turbulence a
3. the effect of ground rough
Given the number of variables, the general case could be very complex,
as illustrated in Fig. 30. Even though the site may be some distance into a
town
+ high enough above ground, at point ‘a’, the gust profile is
unchanged from the profile above the se
+ Tower down, at point *h’, the gust profile is controlled by the fetch
of open country
® only close to the rooftops, at point ‘c’, is the gust profile controlled
by the fetch of town.
Wind
aoe ‘County Town
Fig. 30. General case for ste exposure
60 | Wind loading: a practical guide to BS 6399-2
Commentary | 61
To simplify application of the standard method, values of terrain ancl
building factor So were pre-computed for standard values of height, etch
and peak factor, and presented in Table 4.
"The peak factor method defines the wind speed for a gust of
duration ras:
Y Voll + gl) 6
where Vis the mean wind speed, gis the peak factor depending on
the gust duration » and J is the turbulence intensity. This was
represented in Rig. 7, where Y = x Vo is the rms value of
turbulence.
"The gust duration rs given from the dimension a by Equation 1
and the peak factor is given by the empirical equation:!
+42 In( 3600/0) 0)
‘The mean wind speed Vo and turbulence intensity vary with height
above ground, surface roughness, fetch and which internal boundary
layer has control, as indicated hy ‘a’, °b’ and “e” in Fig, 30.
To simplify application of the dircetional method, Equation 8
was reformulated using a number of sub-S-factors
Sy = Sel + (BST + Si] (10) (29)
where Se is the fetch factor for country terrain, T is the fetch
adjustment factor for town terrain, 3, is the turbulence factor for
&
country terrain, 7, is the turbulence adjustment factor for town
terrain and Sy is the topographie increment.
In country terrain the adjustment factors for town lorrain
1 and Equation 10 simplifies to Equation 28. The form
of Equation 8 is recognizable in Equation 10 if we note that $.7. =
Vo] Ya and ST, = over flat terrain.
‘There is the additional compleaity that the directional method
includes the effect of topography in S, while the standard method
includes topography in the altitude factor Sy
TT
2452 Application
O Interpolation. Interpolation should be made logarithmically
hecause this is the way that the rows of effective height and
«columns of distance to sea are spaced. In interpolating hetween Ori
and 2 km, you should treat 0 kim as being 0-1 km, the minimum feich
used in Fables 23 and 24, Logarithmic interpolation is demonstrated
in Example 6. The accuracy of linear and logarithmic interpotation
iscussed later in 3.12,
Example 6. Logarithmic interpolation in a table
‘To interpolate Iogarhmically within Table 4 for
effective height 4
© Distance to sea X—20km
© Sie in conniry terrain,
1, Hind the values of Closest distance to sea ether side of the fetch (IOKm and
100 kan).
2. Find de vales of ite ig soe and blow he ove fight Sm
und 10m).
3. Caleulate the logarithuns of each ul the values of distance to sea and effective
height
4. Look up the values of Sh in Table 4 for these datum values
5. Lay these values out on à cxlculaion sheet as shown below
6. Either
(a) For each datum effective height (Sm and 10m), interpolate for the
corresponding value of $, between the values of log X.
(b) Then inerpolate for the value of Sy between the values of log
or
€) For each datum fetch (10km and 100km), imerpolate for the
corresponding value at $ between the values of log HL.
(4) Then interpolate for the value of S between the values of log X.
Both these options give the value Sh = 1:644, as shown below.
ET
xa ET
AA,
T i
I
SEI Sada
io iS
Ben
Dan
a
PDT
Commentary | 61
To simplify application of the standard method, values of terrain ancl
building factor So were pre-computed for standard values of height, etch
and peak factor, and presented in Table 4.
"The peak factor method defines the wind speed for a gust of
duration ras:
Y Voll + gl) 6
where Vis the mean wind speed, gis the peak factor depending on
the gust duration » and J is the turbulence intensity. This was
represented in Rig. 7, where Y = x Vo is the rms value of
turbulence.
"The gust duration rs given from the dimension a by Equation 1
and the peak factor is given by the empirical equation:!
+42 In( 3600/0) 0)
‘The mean wind speed Vo and turbulence intensity vary with height
above ground, surface roughness, fetch and which internal boundary
layer has control, as indicated hy ‘a’, °b’ and “e” in Fig, 30.
To simplify application of the dircetional method, Equation 8
was reformulated using a number of sub-S-factors
Sy = Sel + (BST + Si] (10) (29)
where Se is the fetch factor for country terrain, T is the fetch
adjustment factor for town terrain, 3, is the turbulence factor for
&
country terrain, 7, is the turbulence adjustment factor for town
terrain and Sy is the topographie increment.
In country terrain the adjustment factors for town lorrain
1 and Equation 10 simplifies to Equation 28. The form
of Equation 8 is recognizable in Equation 10 if we note that $.7. =
Vo] Ya and ST, = over flat terrain.
‘There is the additional compleaity that the directional method
includes the effect of topography in S, while the standard method
includes topography in the altitude factor Sy
TT
2452 Application
O Interpolation. Interpolation should be made logarithmically
hecause this is the way that the rows of effective height and
«columns of distance to sea are spaced. In interpolating hetween Ori
and 2 km, you should treat 0 kim as being 0-1 km, the minimum feich
used in Fables 23 and 24, Logarithmic interpolation is demonstrated
in Example 6. The accuracy of linear and logarithmic interpotation
iscussed later in 3.12,
Example 6. Logarithmic interpolation in a table
‘To interpolate Iogarhmically within Table 4 for
effective height 4
© Distance to sea X—20km
© Sie in conniry terrain,
1, Hind the values of Closest distance to sea ether side of the fetch (IOKm and
100 kan).
2. Find de vales of ite ig soe and blow he ove fight Sm
und 10m).
3. Caleulate the logarithuns of each ul the values of distance to sea and effective
height
4. Look up the values of Sh in Table 4 for these datum values
5. Lay these values out on à cxlculaion sheet as shown below
6. Either
(a) For each datum effective height (Sm and 10m), interpolate for the
corresponding value of $, between the values of log X.
(b) Then inerpolate for the value of Sy between the values of log
or
€) For each datum fetch (10km and 100km), imerpolate for the
corresponding value at $ between the values of log HL.
(4) Then interpolate for the value of S between the values of log X.
Both these options give the value Sh = 1:644, as shown below.
ET
xa ET
AA,
T i
I
SEI Sada
io iS
Ben
Dan
a
PDT
62 | Wind loading: a practical guide to BS 6399-2
Commentary 1 63
O Directinnal method, The terrain and building factor is given by
Equation 10.
1. Read the values of Se and $, from Table 22. As the rows of
effective height and columns of distance to sea are given in
logarithmic steps, logarithmic interpolation should be used.
2. If the site is in town terrain, read the values of 7. and 7, from
“able 23, intcrpolating logarithmically. IF the site is in country
terrain, take both these factors as unity
3. Read the value of gust peak factor g, from Table 24, interpolating
logarithmically, ur from Figure FJ. To do this, you need to know
the relevant values of diagonal dimension a (see 2.6.2, below).
4. [topography is significant, the topographic increment Sy is given
by:
an
effective slope Y and location
in 21.6.2
$204
The topographic dimension:
factor s, were described carl
The constant 2-0 in Equation 11 is different fram the constant 1-2
in Equation 5, since the former is applied to the mean wind speed
while the later is applied to the gust wind speed.
Note:
M Standard method. Look up the value of terrain and building
Factor in Table 4.
© In country terrain or less than 2km inside town terrain,
For each value of effective height, look up the corresponding.
valuc of Sy from the lefi-hand set of values labelled “Site in
count”, interpolating between the values of effective height
and distance to sea.
© More than 2km inside town terrain. For each value of
effective height, look up the corresponding value of 5, from
the right-hand set of values labelled ‘Site in town.
extending > 2 km upwind from the sit”, interpolating between
the values of effective height and distance to sea
2.45.3 Comments
‘The terrain and building factor accounts for topography and size effects
in the directional method, whereas the standard method accounts for
topography in the altitude factor S, and excludes size effect. We shall see
later, in 3.23.5, that replacing the standard method value of Sp by the
directional method value for the datum diagonal dimension a = 5m over
flat ground is a valid hybrid option permitted hy §3.4.2, This is useful for
removing, must of the unnecessary conservatism in the standard method
wind speeds, particularly for sites in towns.
Logarithmic interpolation can be avoided by using the Wind Loading
Ready-Reckoner for BS 6399-2, which expands Table 4 so that
interpolation is not necessary, or by using the computer program BREVe
(Gee 3.1.4 Design aids). These design aids give a higher precision to the
value of $, (the precise value in Example 6 is Sy — 1-659, compared with
the value Sa — 1-644 obtained by interpolation) but the effect on the wind
loads is minimal (32%).
The effect af topugraphy is to increase the mean wind speed
over hills and escarpments without making any significant change
to the rms turbulence Y, as described in 2.16 earlier. The
directional method applies this increase in mean wind speed
through the topographic increment $, in Kquation 10, so that the
overall effect of topography on design gust speeds depends
properly on the ground roughness, ‘This increase in mean wind
speed dilutes the turbulence intensity /, which is now given by
PAS TAO + Sn).
‘The standard method applies a topographic factor to both mean
and turbulence components as part of the altitude factor Sa, so
excludes the dependence on ground roughness. As this dependence
on ground roughness is very weak, the difference in the two
methods is not si
24.6 Effective wind speed and dynamic pressure (§2.1.2, §2.2.3,
$31.2, §3.2.3)
24.6.1 Basis
‘The effective wind speed is the design gust wind speed for the
building or component heing considered, including all the effects of
wind climate and site exposure, The dynamic pressure is a measure of
the kinetic energy of the effective wind speed, as described earlier in
2171.
Commentary 1 63
O Directinnal method, The terrain and building factor is given by
Equation 10.
1. Read the values of Se and $, from Table 22. As the rows of
effective height and columns of distance to sea are given in
logarithmic steps, logarithmic interpolation should be used.
2. If the site is in town terrain, read the values of 7. and 7, from
“able 23, intcrpolating logarithmically. IF the site is in country
terrain, take both these factors as unity
3. Read the value of gust peak factor g, from Table 24, interpolating
logarithmically, ur from Figure FJ. To do this, you need to know
the relevant values of diagonal dimension a (see 2.6.2, below).
4. [topography is significant, the topographic increment Sy is given
by:
an
effective slope Y and location
in 21.6.2
$204
The topographic dimension:
factor s, were described carl
The constant 2-0 in Equation 11 is different fram the constant 1-2
in Equation 5, since the former is applied to the mean wind speed
while the later is applied to the gust wind speed.
Note:
M Standard method. Look up the value of terrain and building
Factor in Table 4.
© In country terrain or less than 2km inside town terrain,
For each value of effective height, look up the corresponding.
valuc of Sy from the lefi-hand set of values labelled “Site in
count”, interpolating between the values of effective height
and distance to sea.
© More than 2km inside town terrain. For each value of
effective height, look up the corresponding value of 5, from
the right-hand set of values labelled ‘Site in town.
extending > 2 km upwind from the sit”, interpolating between
the values of effective height and distance to sea
2.45.3 Comments
‘The terrain and building factor accounts for topography and size effects
in the directional method, whereas the standard method accounts for
topography in the altitude factor S, and excludes size effect. We shall see
later, in 3.23.5, that replacing the standard method value of Sp by the
directional method value for the datum diagonal dimension a = 5m over
flat ground is a valid hybrid option permitted hy §3.4.2, This is useful for
removing, must of the unnecessary conservatism in the standard method
wind speeds, particularly for sites in towns.
Logarithmic interpolation can be avoided by using the Wind Loading
Ready-Reckoner for BS 6399-2, which expands Table 4 so that
interpolation is not necessary, or by using the computer program BREVe
(Gee 3.1.4 Design aids). These design aids give a higher precision to the
value of $, (the precise value in Example 6 is Sy — 1-659, compared with
the value Sa — 1-644 obtained by interpolation) but the effect on the wind
loads is minimal (32%).
The effect af topugraphy is to increase the mean wind speed
over hills and escarpments without making any significant change
to the rms turbulence Y, as described in 2.16 earlier. The
directional method applies this increase in mean wind speed
through the topographic increment $, in Kquation 10, so that the
overall effect of topography on design gust speeds depends
properly on the ground roughness, ‘This increase in mean wind
speed dilutes the turbulence intensity /, which is now given by
PAS TAO + Sn).
‘The standard method applies a topographic factor to both mean
and turbulence components as part of the altitude factor Sa, so
excludes the dependence on ground roughness. As this dependence
on ground roughness is very weak, the difference in the two
methods is not si
24.6 Effective wind speed and dynamic pressure (§2.1.2, §2.2.3,
$31.2, §3.2.3)
24.6.1 Basis
‘The effective wind speed is the design gust wind speed for the
building or component heing considered, including all the effects of
wind climate and site exposure, The dynamic pressure is a measure of
the kinetic energy of the effective wind speed, as described earlier in
2171.
64 | Wind loading: a practical guide to BS 6399-2
Commentary 65
+ In the directional method, the effective wind speed and dynamic
pressure include the size effect, so are equivalent to the previous
Class A, B or C design wind speeds from CP 3-V-2.
+ Inthe standard method, the effective wind speed and dynamic
pressure represent the datum 15 duration gust (a= 5m), so are
equivalent to the Class A design wind speed and dynamic pressure
from CP 3-V-2. The size effect is treated separately by the size
effect factor Cy
In general the dynamic pressure is denoted by the symbol 9. In the
nal method it is denoted by the symbol ge for deviving external
pressures and by q, for derivi
ving internal pressures. Inthe standard method,
the dynamic pressure is denoted by the symbol q, so that, when size effect
is included:
474€ 02
2462 Application
O The effective wind speed Ve is obtained from Equation 12 or 27:
¥e= Vs Se
© The dynamic pressure q is obtained from Equation 1 or 16
q=0613 V.? or from Table 2.
2.4.7 Commentary on site exposure
‘The main strength of the treatment of site exposure in BS 6399-2 is thal
it is much better able to distinguish between sites that are exposed and
sites that are sheltered! than previously. This section has described a
number of changes from previous practice:
‘The number of roughness categories is reduced.
Wind speed profiles develop gradually after a change of roughness
instead of in steps.
+ The minimum fetch to establish a change of roughness is replaced
by the actual distance from sea and in town.
+ The fixed displacement height in towns now varies, depending on
the height and spacing of permanent upwind obstacles.
+ Appl Ision-by-parts rule is restricted to structural
loads on slender buildings.
One of the principal aims was to make values change smoothly so as lu
eliminate large step changes in design loads. Two steps still remain:
1. In both directional and standard methods, step changes occur on
either side of the minimum 0.05 (1:20) slope for significant
topography. However the size of this potential step is reduced by
th corresponding change Between site and erin base aldi
the factor.
2. The 2km fetch for town terrain in Table 4 of the standard method
induces a step change in wind Ions of shou 58, We will see
ww to eliminate the conservatism introduced by this step later in
En y this step Inter
We can judge the success of this aim in Example 7, by using Option
1 —Inrespective of direction, standard method (see 2.1.5.3), the most
conservative option of BS 6399-2, to recalculate Example 1. The
range of possible values is now only 7% und this is due t the
remaining step around the topography threshold. In this particular
example, a further 3% reduction can be obtained by applying §3.4.2
(3335) to eliminate conservatism from the ~500m fetch of town
terrain
Example 7. Example 1 recalculated using BS 6399-2
Note: The methods and values inthis exemple relate to US 6399.2:1997.
We reconsider the site shown below, just inland from the const at Blackpool, where
basic wind speed is now Ve =22:5 mis,
site
x" |
Blackpool, V,=29.5m/s Siope.
a tete
ei + “Town >
We repeat the calculi in Example 1 wing Option 1 — respective of det,
Mandard ced. Ihe ithe man conrewive open, hat the depne al
comen small mar a west-faing com. BS 6399-2 rela de ap
1:20 topographie sipicance atts À
offset bythe td base alte nthe lod factor $y For
the eyramie clsienlen. we shall tae the balling as Ding el mon
coca
Commentary 65
+ In the directional method, the effective wind speed and dynamic
pressure include the size effect, so are equivalent to the previous
Class A, B or C design wind speeds from CP 3-V-2.
+ Inthe standard method, the effective wind speed and dynamic
pressure represent the datum 15 duration gust (a= 5m), so are
equivalent to the Class A design wind speed and dynamic pressure
from CP 3-V-2. The size effect is treated separately by the size
effect factor Cy
In general the dynamic pressure is denoted by the symbol 9. In the
nal method it is denoted by the symbol ge for deviving external
pressures and by q, for derivi
ving internal pressures. Inthe standard method,
the dynamic pressure is denoted by the symbol q, so that, when size effect
is included:
474€ 02
2462 Application
O The effective wind speed Ve is obtained from Equation 12 or 27:
¥e= Vs Se
© The dynamic pressure q is obtained from Equation 1 or 16
q=0613 V.? or from Table 2.
2.4.7 Commentary on site exposure
‘The main strength of the treatment of site exposure in BS 6399-2 is thal
it is much better able to distinguish between sites that are exposed and
sites that are sheltered! than previously. This section has described a
number of changes from previous practice:
‘The number of roughness categories is reduced.
Wind speed profiles develop gradually after a change of roughness
instead of in steps.
+ The minimum fetch to establish a change of roughness is replaced
by the actual distance from sea and in town.
+ The fixed displacement height in towns now varies, depending on
the height and spacing of permanent upwind obstacles.
+ Appl Ision-by-parts rule is restricted to structural
loads on slender buildings.
One of the principal aims was to make values change smoothly so as lu
eliminate large step changes in design loads. Two steps still remain:
1. In both directional and standard methods, step changes occur on
either side of the minimum 0.05 (1:20) slope for significant
topography. However the size of this potential step is reduced by
th corresponding change Between site and erin base aldi
the factor.
2. The 2km fetch for town terrain in Table 4 of the standard method
induces a step change in wind Ions of shou 58, We will see
ww to eliminate the conservatism introduced by this step later in
En y this step Inter
We can judge the success of this aim in Example 7, by using Option
1 —Inrespective of direction, standard method (see 2.1.5.3), the most
conservative option of BS 6399-2, to recalculate Example 1. The
range of possible values is now only 7% und this is due t the
remaining step around the topography threshold. In this particular
example, a further 3% reduction can be obtained by applying §3.4.2
(3335) to eliminate conservatism from the ~500m fetch of town
terrain
Example 7. Example 1 recalculated using BS 6399-2
Note: The methods and values inthis exemple relate to US 6399.2:1997.
We reconsider the site shown below, just inland from the const at Blackpool, where
basic wind speed is now Ve =22:5 mis,
site
x" |
Blackpool, V,=29.5m/s Siope.
a tete
ei + “Town >
We repeat the calculi in Example 1 wing Option 1 — respective of det,
Mandard ced. Ihe ithe man conrewive open, hat the depne al
comen small mar a west-faing com. BS 6399-2 rela de ap
1:20 topographie sipicance atts À
offset bythe td base alte nthe lod factor $y For
the eyramie clsienlen. we shall tae the balling as Ding el mon
coca
66 1 Wind loading a practical guide to BS 6399-2
Commentary | 67
Parameter Most onerous Least onerous. Notes
Basie wind speed: ads 235 233
Slope 005001 004999 Eller side of significance
threshold
Distance inland: nı 499 soi Distance to sea and in
tom
Altitude m o 2s To base and 10 site,
respectively
Effective slope xtoeation 005 o sol ut crest
fa
Altitude factor 106 1025 Site at cres of escarpment
Roughness category Tom Town
Reference height: m 10 10 Equal to obstruction
height
Effective high: m 4 4 He=0-4H, when H,= Ho
Factor Se 158 138 (054m from sea, < Zn in
Dynamic pressure: Pa 950 mar
Fromal area: m 500 500
Net pressure coefficient 1-1 11 Table 5, span ratio =
Size effect factor 0853 0853 Figure 4, line À
a=50m
Dynamic angmentaion 001 oor Figure 3, Ky 05
factor
Bso shear force: KN 383 357 Equation 7, includes 085
CP 3.V-2 values frum 824 306
ample 1: KN
25 Bullding shape factors
2.5.1 The shape factor model
25.1.1 Shape factors (§2.3, 53.3) o
ape factor” is Ihe term used by codes of practico for all the factors
that describe the effect of building shape, CP 3-V-2 used pressure
coefficients and force coefficients. BS 6399-2 uses only the pressure
coefficients:
© Cye—the external pressure coefficient, for external pressures
© Ch — the internal pressure coefficient, For internal pressures
© Cy —the net pressure coefficient, for net pressure difference,
p= Copia
all mean values Ce obtained by dividing the mean pressures by the
mean dynanie pressu
| ‘The shape cocfficients for external pressures in CP 3-V-2 were
Tye=Fe/ (where the ~ symbol indicates a mean value) (13)
‘The peak gust pressure was obtained from this by assuming that
the equivalent static gust acts like the mean wind:
Pe= Te à (where the * symbol indicates a peak value) (14)
This is called the ‘quasi-steady assumption’ and assumes that all
: (he fluctuations of wind pressure are caused by approaching gusts.
But we know that each building generates its own turbulence that
increases the high suctions around the edges of walls and roofs,
External pressure coefficients in the latest generation of codes are
obtained directly from measured peak pressures divided by the peak
dynamic pressure:
Cre là (where the” symbol indicates these new values) (15)
In practice, these new values are close to the uld mean values but
are more précise because they include the effect of building
generated turbulence. All the data in BS 6399-2 except Table?, Cpe
for walls of cireular-plan buildings, are this new type of value. (For
a detailed explanation of this process see Part 2 of reference 1.)
25.1.2 External pressure coefficient zones
‘The values of C,« vary smoothly aver the surface of a building but, like
most other codes, BS 6399-2 simplifies this to a series of zones of uniform
pressure, as shown in Fig, 31. On the one hand, the value for each zone
needs to be the average value for each zone to allow the load to be
summed over a number of zones without adding conservatism. On the
other hand, the value for each zone needs to he the highest value found in
each zone lo give the maximum surface pressure for cladding design.
Both criteria cannot be met simultaneously and some compromise value
has to be used. As Fig. 31 shows, the smallest edge zones are set 10 the
highest value to cover the requirements for cladding fixings while the
‘other zones take values close to the mean.
To meet the 10% objective in 1.4 requires the zones to be smaller
around the edges of buildings where pressure changes value quickly.
Even so, this creates step changes in loading that are not really
wanted, This is one occasion where the steps cannot be completely
removed without adding undue complexity. In the directional
method, the zones are laid out asymmetrically from the upwind
Commentary | 67
Parameter Most onerous Least onerous. Notes
Basie wind speed: ads 235 233
Slope 005001 004999 Eller side of significance
threshold
Distance inland: nı 499 soi Distance to sea and in
tom
Altitude m o 2s To base and 10 site,
respectively
Effective slope xtoeation 005 o sol ut crest
fa
Altitude factor 106 1025 Site at cres of escarpment
Roughness category Tom Town
Reference height: m 10 10 Equal to obstruction
height
Effective high: m 4 4 He=0-4H, when H,= Ho
Factor Se 158 138 (054m from sea, < Zn in
Dynamic pressure: Pa 950 mar
Fromal area: m 500 500
Net pressure coefficient 1-1 11 Table 5, span ratio =
Size effect factor 0853 0853 Figure 4, line À
a=50m
Dynamic angmentaion 001 oor Figure 3, Ky 05
factor
Bso shear force: KN 383 357 Equation 7, includes 085
CP 3.V-2 values frum 824 306
ample 1: KN
25 Bullding shape factors
2.5.1 The shape factor model
25.1.1 Shape factors (§2.3, 53.3) o
ape factor” is Ihe term used by codes of practico for all the factors
that describe the effect of building shape, CP 3-V-2 used pressure
coefficients and force coefficients. BS 6399-2 uses only the pressure
coefficients:
© Cye—the external pressure coefficient, for external pressures
© Ch — the internal pressure coefficient, For internal pressures
© Cy —the net pressure coefficient, for net pressure difference,
p= Copia
all mean values Ce obtained by dividing the mean pressures by the
mean dynanie pressu
| ‘The shape cocfficients for external pressures in CP 3-V-2 were
Tye=Fe/ (where the ~ symbol indicates a mean value) (13)
‘The peak gust pressure was obtained from this by assuming that
the equivalent static gust acts like the mean wind:
Pe= Te à (where the * symbol indicates a peak value) (14)
This is called the ‘quasi-steady assumption’ and assumes that all
: (he fluctuations of wind pressure are caused by approaching gusts.
But we know that each building generates its own turbulence that
increases the high suctions around the edges of walls and roofs,
External pressure coefficients in the latest generation of codes are
obtained directly from measured peak pressures divided by the peak
dynamic pressure:
Cre là (where the” symbol indicates these new values) (15)
In practice, these new values are close to the uld mean values but
are more précise because they include the effect of building
generated turbulence. All the data in BS 6399-2 except Table?, Cpe
for walls of cireular-plan buildings, are this new type of value. (For
a detailed explanation of this process see Part 2 of reference 1.)
25.1.2 External pressure coefficient zones
‘The values of C,« vary smoothly aver the surface of a building but, like
most other codes, BS 6399-2 simplifies this to a series of zones of uniform
pressure, as shown in Fig, 31. On the one hand, the value for each zone
needs to be the average value for each zone to allow the load to be
summed over a number of zones without adding conservatism. On the
other hand, the value for each zone needs to he the highest value found in
each zone lo give the maximum surface pressure for cladding design.
Both criteria cannot be met simultaneously and some compromise value
has to be used. As Fig. 31 shows, the smallest edge zones are set 10 the
highest value to cover the requirements for cladding fixings while the
‘other zones take values close to the mean.
To meet the 10% objective in 1.4 requires the zones to be smaller
around the edges of buildings where pressure changes value quickly.
Even so, this creates step changes in loading that are not really
wanted, This is one occasion where the steps cannot be completely
removed without adding undue complexity. In the directional
method, the zones are laid out asymmetrically from the upwind
681
Wind loading: a practical guide to BS 6399-2
Commentary 1 69
zer
Bis. 31, Pressure coefficient zones
Note:
comer, so that the method always predicts an asymmetric loading,
even when the building is symmetrical and the wind is normal to a
face. The standard method has fewer zones that are lid out
symmetrically, with the smaller h don zones along the
upwind edge. This requires asymmetric loads to be considered
separately (in §2./.3.7, see 2.644 Inter)
‘A major change has been made to the way that the pressure
coefficient zones are sized and Fig. 32 illustrates why this was
necessary. The zones of highest suction un the roof and side walls of
a building are caused by bubbles of separated flow. Provided that
the separation bubble is smaller than the depth of the roof D, its size:
does not change as D increases further, as shown in Fig. 32(0). The
sizes of the zones of suction corresponding to this bubble do not
increase in the fixed proportions to the building plan, as assumed by
CP 3-V-2 and shown in Fig, 32(b). Instead, the zone sizes in BS
6399-2 are controlled hy the scaling length b that is sot by the
proportions of the front elev nee the front elevation is the
‘same fur each case in Fig. 32, the zones remain constant in size as.
shown in Fi basis and application of the scaling
parameter b are described in the next section
Unlike in CP 3-V-2 (Fig, 32(b)) BS 6399-2 does not predict hig
suction zones along the side caves of the ruef in the standard
method. fc). High-suction zones occur along these eaves when they
are at the windward edge, ie. in another orthogonal direction.
The practical consequence of this change is that the uplift on
long-span roofs on low-rise buildings is greatly reduced, as
illustrated for a flat roof in Fig, 33. On the other hand, the up
on roots of tall slab-like buildings increases because the higher
suction zones may uccupy the whole roof. IL also requires fewer
combinations of values to be tabulated,
—_
(a) Flow over oot
(©) Zones h CP 3.V.2, constant proportion ol plan
CI | |
(e) Zones in BS 63992, constant properon of font levain
Fig. 32. lifer of mind depth on foe aver a fla roof
“ES == === ==
ARR EEE
AAA
Fig. 33. oval uplift on Pat reef H= 3mm, Gy= 402
25.13 Directional and standard external pressure coefficients
25.1.3.) Directional method " cire
In the directional method, external pressure coefficients are given for
teach zone in 15° or 30° increments of wind angle @ from normal to the
building. On flat roofs, the pressure coefficients on each wall and hehind
each of the upwind eaves depend only on the angle of the wind normal to
‚each wall, as illustrated in Fig. 34. This now allows the data to be applied
10 any non-rectangular-plan building, shown by the definition Figure 35.
However, the data for pitched rmofs are still limited to rectangular plan
shapes. Similarly, the external coefficients for the upwind fave of à
duopitch roof are taken to be independent of the coefficients on the
Wind loading: a practical guide to BS 6399-2
Commentary 1 69
zer
Bis. 31, Pressure coefficient zones
Note:
comer, so that the method always predicts an asymmetric loading,
even when the building is symmetrical and the wind is normal to a
face. The standard method has fewer zones that are lid out
symmetrically, with the smaller h don zones along the
upwind edge. This requires asymmetric loads to be considered
separately (in §2./.3.7, see 2.644 Inter)
‘A major change has been made to the way that the pressure
coefficient zones are sized and Fig. 32 illustrates why this was
necessary. The zones of highest suction un the roof and side walls of
a building are caused by bubbles of separated flow. Provided that
the separation bubble is smaller than the depth of the roof D, its size:
does not change as D increases further, as shown in Fig. 32(0). The
sizes of the zones of suction corresponding to this bubble do not
increase in the fixed proportions to the building plan, as assumed by
CP 3-V-2 and shown in Fig, 32(b). Instead, the zone sizes in BS
6399-2 are controlled hy the scaling length b that is sot by the
proportions of the front elev nee the front elevation is the
‘same fur each case in Fig. 32, the zones remain constant in size as.
shown in Fi basis and application of the scaling
parameter b are described in the next section
Unlike in CP 3-V-2 (Fig, 32(b)) BS 6399-2 does not predict hig
suction zones along the side caves of the ruef in the standard
method. fc). High-suction zones occur along these eaves when they
are at the windward edge, ie. in another orthogonal direction.
The practical consequence of this change is that the uplift on
long-span roofs on low-rise buildings is greatly reduced, as
illustrated for a flat roof in Fig, 33. On the other hand, the up
on roots of tall slab-like buildings increases because the higher
suction zones may uccupy the whole roof. IL also requires fewer
combinations of values to be tabulated,
—_
(a) Flow over oot
(©) Zones h CP 3.V.2, constant proportion ol plan
CI | |
(e) Zones in BS 63992, constant properon of font levain
Fig. 32. lifer of mind depth on foe aver a fla roof
“ES == === ==
ARR EEE
AAA
Fig. 33. oval uplift on Pat reef H= 3mm, Gy= 402
25.13 Directional and standard external pressure coefficients
25.1.3.) Directional method " cire
In the directional method, external pressure coefficients are given for
teach zone in 15° or 30° increments of wind angle @ from normal to the
building. On flat roofs, the pressure coefficients on each wall and hehind
each of the upwind eaves depend only on the angle of the wind normal to
‚each wall, as illustrated in Fig. 34. This now allows the data to be applied
10 any non-rectangular-plan building, shown by the definition Figure 35.
However, the data for pitched rmofs are still limited to rectangular plan
shapes. Similarly, the external coefficients for the upwind fave of à
duopitch roof are taken to be independent of the coefficients on the
70 1 Wind loading: a practical guide to BS 6399-2
Commentary | 71
I &
Fig. 34. Rosie of pressure coeficiente for flat rs
downwind face, allowing the same coefficients to be used for monopiteh
roofs. This allows more building shapes to be covered using fewer tables
uf data and allows complex shapes to be assembled from simple
components.
2.5.1.3.2 Standard method
"The external pressure coefficient tables for the standard method have
been compiled from the directional coefficients for the range of wind
angle 45° from the orthogonal case and by amalgamating some of the
Zones. Essentially, the orthogonal case 0° is derived from wind angles O°
and 30°, while the orthogonal ease 90” is derived from wind angles 60
and 90", By forcing symmetry and confining the high-suction zones to the
windward edge, the standard method greatly simplifies the zones, but
introduces some minor inconsistencies on which guidance is given later.
251.33 Ronge
‘The range of pressure coefficient data has been extended to cover
common building forms previously excluded or made popular in
recent years. Newly included arc troughed and hipped roofs and the
effects of parapets, curved eaves und mansard eaves on Mat roof.
‘Some rare shapes, for which reliable data exist, are not included, c.g.
‘domes, barrel vault, hypesboloid and skew-hipped roofs. Com-
patible data for barrel vaulls are given in 2.545. Compatible
pressure coefficients for other forms may be found in Part 2 of
reference 1.
2.5.1.34 Lottice structures
Cocticients for single and multiple lattice frames, previously
included in CP 3-V-2, are missing in BS 6399-2. The justification
for this was two fold. Firstly that lattice structures are excluded
from the scope and secondly that they are covered by other
specialized standards. However, lattices in the form of exposed
trusses are a common component of many buildings, and most
framed buildings form lattices during construction before the
cladding is applied. This exclusion may need to be reviewed but, in
the meantime, guidance is provided here in 2.5.8.2 and Appendix
A, derived from Part 2 of reference 1. The more detailed treatment
given in an earlier BRE Report Design Guide for Wind Loads on
Unclad Building Frames During Construction” is still valid and
useful, provided the references to the CP 3-V-2 design wind speed
are replaced hy the BS 6399-2 eff
25.2. Thescaling length and zones
252.1 Basis
“The sizes ofthe pressure coefficient zones are set by the scaling length
b. The basis of this parameter is the ‘laziness’ of the wind which always
finds the path of least resistance, as illustrated in Fig. 35. When the front
elevation of the building is squat, B > 2H, the wind finds it easier to pass
over the building than around it. In doing this, the wind has to diverge
from a straight path hy a maximum distance of H. When the front
elevation of the building is slender, B< 211, the wind finds it easier to
pass around the building than over it. In doing this, the wind diverges hy à
maximum distance of H or B/2, whichever is the smaller. This
“divergence distance” is a helpful concept when dealing with complicated
shapes.
pat
Bram Ena iogo OVER tan ARQUND 62H Casio o AROUND wan OVER
Fig. 35. Base of eating tngth b
Commentary | 71
I &
Fig. 34. Rosie of pressure coeficiente for flat rs
downwind face, allowing the same coefficients to be used for monopiteh
roofs. This allows more building shapes to be covered using fewer tables
uf data and allows complex shapes to be assembled from simple
components.
2.5.1.3.2 Standard method
"The external pressure coefficient tables for the standard method have
been compiled from the directional coefficients for the range of wind
angle 45° from the orthogonal case and by amalgamating some of the
Zones. Essentially, the orthogonal case 0° is derived from wind angles O°
and 30°, while the orthogonal ease 90” is derived from wind angles 60
and 90", By forcing symmetry and confining the high-suction zones to the
windward edge, the standard method greatly simplifies the zones, but
introduces some minor inconsistencies on which guidance is given later.
251.33 Ronge
‘The range of pressure coefficient data has been extended to cover
common building forms previously excluded or made popular in
recent years. Newly included arc troughed and hipped roofs and the
effects of parapets, curved eaves und mansard eaves on Mat roof.
‘Some rare shapes, for which reliable data exist, are not included, c.g.
‘domes, barrel vault, hypesboloid and skew-hipped roofs. Com-
patible data for barrel vaulls are given in 2.545. Compatible
pressure coefficients for other forms may be found in Part 2 of
reference 1.
2.5.1.34 Lottice structures
Cocticients for single and multiple lattice frames, previously
included in CP 3-V-2, are missing in BS 6399-2. The justification
for this was two fold. Firstly that lattice structures are excluded
from the scope and secondly that they are covered by other
specialized standards. However, lattices in the form of exposed
trusses are a common component of many buildings, and most
framed buildings form lattices during construction before the
cladding is applied. This exclusion may need to be reviewed but, in
the meantime, guidance is provided here in 2.5.8.2 and Appendix
A, derived from Part 2 of reference 1. The more detailed treatment
given in an earlier BRE Report Design Guide for Wind Loads on
Unclad Building Frames During Construction” is still valid and
useful, provided the references to the CP 3-V-2 design wind speed
are replaced hy the BS 6399-2 eff
25.2. Thescaling length and zones
252.1 Basis
“The sizes ofthe pressure coefficient zones are set by the scaling length
b. The basis of this parameter is the ‘laziness’ of the wind which always
finds the path of least resistance, as illustrated in Fig. 35. When the front
elevation of the building is squat, B > 2H, the wind finds it easier to pass
over the building than around it. In doing this, the wind has to diverge
from a straight path hy a maximum distance of H. When the front
elevation of the building is slender, B< 211, the wind finds it easier to
pass around the building than over it. In doing this, the wind diverges hy à
maximum distance of H or B/2, whichever is the smaller. This
“divergence distance” is a helpful concept when dealing with complicated
shapes.
pat
Bram Ena iogo OVER tan ARQUND 62H Casio o AROUND wan OVER
Fig. 35. Base of eating tngth b
72.1 Wind loading: a practical guide to BS 6399-2
Commentary | 73
2522 Application
2522.1 Scaling length §2.4.1.3, §2.5.1.2,§2.5.2.2,§9.3-11.2,§3.3.22.2,
§3.3.3.2.1)
BS 6399-2 usos the scaling length b to describe
is effect, where:
B
b= smaller off oH «ao
so that b represents twice the divergence distance in Fig. 35.
‘The rule of Equation £6 is used consistently throughout BS 6399-2 for
walls (824.14, $3.3.1.1.2) and for roots (62.5.1.2, 825.22, 93.222,
$3.3.3.2.1). In some clauses the length I. or width W is used in place of D.
but this is always consistent with the definition of B in Figure 2.
“The rule is easy to apply when the shape of the building is simple,
‘ew. the house shape shown in Fig. 36(a) or even the tower and
podium in (b). However, selection of the appropriate valuc of B 10
compare with 2H becomes much more difficult as the shape of the
building becomes more complex. Simply adding a single-storey
extension to the house, as in Fig. 37, gives three values of b for
different areas of the building
A. for the walls of the upper storey and the roof of the house:
B=8m, H—75m, so b=8m as before
B. for the walls of the lower storey of the hou
H=15m, so b= 12m
C. forthe walls and roof of the extension: B= 12m, H=3m, so
b=6m.
‘The scaling length b is also used to dimension the effects of re-
entrant corners, inset storeys and other departures from a regular
‘cuboidal shape, for which no advice was previously given. Some
advice on the value of A to use in these circumstances is given in
82.4.3 and Figure 13 in the section for wind pressures on walls
When there is a wing of the building that protrudes upwind from the
"uílding. you are advised to use the smaller crosswind breadth of the
wing for sizing the zones on the walls of that wing. You should take
this advice to apply 10 the root zones also. Unfortunately, the advice
is unclear on how long this wing necds to be to qualify as a wing.
2.5.2.2.2 Resantrant corners (§2.4.3,§3.3.1 5)
Figure 38 shows a case that BS 6399-2 calls a re-cnirant comer. By
implication, all walls are the same or similar in height. (Guidance for
smaller extensions to buildings is given below.) The upwind wing is
shown shaded in Fig. 38(a) when its depth is greater than
2
So
sana {8
4 „20m
= omater ot (2,0,
atom
pr
em re
(2) House (0) Tower on podium
Pi 36, lg racer or sin buis stes
é B= an i
E sont» Ken
“Li ai!
Dame (¿059, ocean (RUE poemas (852
de His Hs
ig. 37. Seating Tee for hone and extension
sos] y
3
Kol Te
hepa
(a) Upwind wing (©) Not an upwind wing
Fig. 38. Upwind wing caused by a re-entront comer
_ The criterion is whether the wind will divide to flow around
either side of the wing hefore reaching the main building or will
prefer to flow around the building as a whole. If the ‘wing’ is too
short to qualify, as in Fig, 38(b), then the full breadth of the building
must be used in Equation 16, The resulting value of b sets the sizes
‘of the high-suction zones on the side walls and roof, so when in any
doubt he largest value corresponding to the whole hung should
36 use
Commentary | 73
2522 Application
2522.1 Scaling length §2.4.1.3, §2.5.1.2,§2.5.2.2,§9.3-11.2,§3.3.22.2,
§3.3.3.2.1)
BS 6399-2 usos the scaling length b to describe
is effect, where:
B
b= smaller off oH «ao
so that b represents twice the divergence distance in Fig. 35.
‘The rule of Equation £6 is used consistently throughout BS 6399-2 for
walls (824.14, $3.3.1.1.2) and for roots (62.5.1.2, 825.22, 93.222,
$3.3.3.2.1). In some clauses the length I. or width W is used in place of D.
but this is always consistent with the definition of B in Figure 2.
“The rule is easy to apply when the shape of the building is simple,
‘ew. the house shape shown in Fig. 36(a) or even the tower and
podium in (b). However, selection of the appropriate valuc of B 10
compare with 2H becomes much more difficult as the shape of the
building becomes more complex. Simply adding a single-storey
extension to the house, as in Fig. 37, gives three values of b for
different areas of the building
A. for the walls of the upper storey and the roof of the house:
B=8m, H—75m, so b=8m as before
B. for the walls of the lower storey of the hou
H=15m, so b= 12m
C. forthe walls and roof of the extension: B= 12m, H=3m, so
b=6m.
‘The scaling length b is also used to dimension the effects of re-
entrant corners, inset storeys and other departures from a regular
‘cuboidal shape, for which no advice was previously given. Some
advice on the value of A to use in these circumstances is given in
82.4.3 and Figure 13 in the section for wind pressures on walls
When there is a wing of the building that protrudes upwind from the
"uílding. you are advised to use the smaller crosswind breadth of the
wing for sizing the zones on the walls of that wing. You should take
this advice to apply 10 the root zones also. Unfortunately, the advice
is unclear on how long this wing necds to be to qualify as a wing.
2.5.2.2.2 Resantrant corners (§2.4.3,§3.3.1 5)
Figure 38 shows a case that BS 6399-2 calls a re-cnirant comer. By
implication, all walls are the same or similar in height. (Guidance for
smaller extensions to buildings is given below.) The upwind wing is
shown shaded in Fig. 38(a) when its depth is greater than
2
So
sana {8
4 „20m
= omater ot (2,0,
atom
pr
em re
(2) House (0) Tower on podium
Pi 36, lg racer or sin buis stes
é B= an i
E sont» Ken
“Li ai!
Dame (¿059, ocean (RUE poemas (852
de His Hs
ig. 37. Seating Tee for hone and extension
sos] y
3
Kol Te
hepa
(a) Upwind wing (©) Not an upwind wing
Fig. 38. Upwind wing caused by a re-entront comer
_ The criterion is whether the wind will divide to flow around
either side of the wing hefore reaching the main building or will
prefer to flow around the building as a whole. If the ‘wing’ is too
short to qualify, as in Fig, 38(b), then the full breadth of the building
must be used in Equation 16, The resulting value of b sets the sizes
‘of the high-suction zones on the side walls and roof, so when in any
doubt he largest value corresponding to the whole hung should
36 use
74 | Wind loading: a practical guide to BS 6399-2
Commentary | 75
25.2.2.3 Recessed bays (§2.4.3, 533.1.6)
Figure 39 shows a case that BS 6399-2 calls a ‘recessed bay
an extension of the previous casc that gives two potential upwi
again of similar height to the main building. The same rule applies, except
that the wind will flow between the wings only if it prefers lo flow over
the main building, as in Fig. 39(a), and und it, as in (b). So an
additional criterion here is that Ihe breadth of the main building must be at
Teast twice its height.
2.52.24 Irregular flush faces (§2.4.4.1, §3.3.1.8)
‘BS 6999-2 calls the case in Fig. 37 an ‘irregular flush face" because the
‘windward wall is a single flat or ‘lush’ face, but is not a simple rectangle.
In the standard method, BS 6399-2 assumes that the peak of the gable wall
under the house roof has no significant effect, although you can define
‘able zones using the directional method ($3.3./.3).
When wind blows from the right, the wall of the house above the
extension roof creates what BS 6399-2 calls an ‘insct face’. The relevant
height # for sizing the zones on the side wall and roof of the house is
taken from the level of the extension roof, exactly as the tower on the
podium in Fig. 36(b). The resulting wall zones arc defined in Figure 14,
‘which is clarified in Fig. 40. The reference height for the © zone on a wall
is the local height of the wall, so u step change of pressure occurs at the
change of height, as shown in Fig, 40(a). The reference height for the A
and B zones on walls, which represent the separation bubble, is always the
height of the front comer of the wall that creates the bubble, as shown in
(b). Put more simply, you may split the C zone into “high” and “low walls
as shown in (a), but the A and B zones should not be split, as in (b)
au
(a) Uomind wings (0) Not upwind wings
Fig. 39. Upwind wings caused by a recessed bay
(9) Lower pat arg enough or © zone
Wind msnm
(6) Lower part 100 small or G zone
Fig. 40, Claricacion of Figure 14
25.225 Inset storeys (§2.4.4.2, §3.3.1.8)
BS 6399-2 calls the tower on the podium in Fig. 36(b) an ‘inset storey"
because the faces are sot back from the podium. In this case, the scaling
length b is controlled by the dimensions of the tower, assuming that His
‘measured from the roof of the podium. The eviterion for an inset storey is
defined by Figure 15 us an inset greater than 0.25. Pressures on the walls
‘of an inset storey affect the pressures on adjacent areas of the roof at its
base, creating the extra zones defined in Figure 18.
‘The side wall of the house above the mof of the extensior
covered by Figure 15(b) that defines an extra E zone with C,
the upwind corner formed by the extension roof. Additional
where high-suction zones appear is given in 344, later
dance on
25.2.2.6 Smaller extensions
When the extension to the house in Fig. 37 is set huck from the
Front face, we get the case shown in Fig. 41. This is not addressed
Siret by BS 6399-2 and we must improvise from the rules we
already have,
1. Our first question is: ‘how does the wind prefer to flow
around the house?” It is not obvious that we have the case
in Fig. 38 because the extension is lower than the house:
the house is the ‘iain’ building and the extension is the
‘wing’. Nor do we obviously have the inset storey case in
Fig. 36(b), but the extension is set back more than 0-2».
Commentary | 75
25.2.2.3 Recessed bays (§2.4.3, 533.1.6)
Figure 39 shows a case that BS 6399-2 calls a ‘recessed bay
an extension of the previous casc that gives two potential upwi
again of similar height to the main building. The same rule applies, except
that the wind will flow between the wings only if it prefers lo flow over
the main building, as in Fig. 39(a), and und it, as in (b). So an
additional criterion here is that Ihe breadth of the main building must be at
Teast twice its height.
2.52.24 Irregular flush faces (§2.4.4.1, §3.3.1.8)
‘BS 6999-2 calls the case in Fig. 37 an ‘irregular flush face" because the
‘windward wall is a single flat or ‘lush’ face, but is not a simple rectangle.
In the standard method, BS 6399-2 assumes that the peak of the gable wall
under the house roof has no significant effect, although you can define
‘able zones using the directional method ($3.3./.3).
When wind blows from the right, the wall of the house above the
extension roof creates what BS 6399-2 calls an ‘insct face’. The relevant
height # for sizing the zones on the side wall and roof of the house is
taken from the level of the extension roof, exactly as the tower on the
podium in Fig. 36(b). The resulting wall zones arc defined in Figure 14,
‘which is clarified in Fig. 40. The reference height for the © zone on a wall
is the local height of the wall, so u step change of pressure occurs at the
change of height, as shown in Fig, 40(a). The reference height for the A
and B zones on walls, which represent the separation bubble, is always the
height of the front comer of the wall that creates the bubble, as shown in
(b). Put more simply, you may split the C zone into “high” and “low walls
as shown in (a), but the A and B zones should not be split, as in (b)
au
(a) Uomind wings (0) Not upwind wings
Fig. 39. Upwind wings caused by a recessed bay
(9) Lower pat arg enough or © zone
Wind msnm
(6) Lower part 100 small or G zone
Fig. 40, Claricacion of Figure 14
25.225 Inset storeys (§2.4.4.2, §3.3.1.8)
BS 6399-2 calls the tower on the podium in Fig. 36(b) an ‘inset storey"
because the faces are sot back from the podium. In this case, the scaling
length b is controlled by the dimensions of the tower, assuming that His
‘measured from the roof of the podium. The eviterion for an inset storey is
defined by Figure 15 us an inset greater than 0.25. Pressures on the walls
‘of an inset storey affect the pressures on adjacent areas of the roof at its
base, creating the extra zones defined in Figure 18.
‘The side wall of the house above the mof of the extensior
covered by Figure 15(b) that defines an extra E zone with C,
the upwind corner formed by the extension roof. Additional
where high-suction zones appear is given in 344, later
dance on
25.2.2.6 Smaller extensions
When the extension to the house in Fig. 37 is set huck from the
Front face, we get the case shown in Fig. 41. This is not addressed
Siret by BS 6399-2 and we must improvise from the rules we
already have,
1. Our first question is: ‘how does the wind prefer to flow
around the house?” It is not obvious that we have the case
in Fig. 38 because the extension is lower than the house:
the house is the ‘iain’ building and the extension is the
‘wing’. Nor do we obviously have the inset storey case in
Fig. 36(b), but the extension is set back more than 0-2».
76 | Wind loading: a practical guide to BS 6399-2
Commentary | 77
Fig. 41. Principle of ‘reflection
We make te judgement that the size and the set-back of
the extension mens that it does not affect the flow around
the front of the house. So, for the whole house; A— Rn,
H=75m, giving b=8m and the flow prefers to flow
around, rather than over the house. This value of 6=8m
sets the zone sizes for the walls and roof of the house.
2. Our next question is “how docs the wind prefer to flow
around the extension?” It can either flow aver or around one
side. It is prevented from flowing around the other side by
the wall ofthe house. That wall acts like a mirror, making the
effective breadth of the extension twice the actual breadth.
We will call this the principle of ‘reflection’. In this case,
B=8m, H—3m, giving b=6m, and the flow prefers lo
flow over rather than around the’ extension. This value of
b=6m sels the zone sizes for the walls and roof of the
extension
“The principle of “reflection” should be used whenever there is a
small extension on a farger wall. When an extension links two taller
buildings, Now between the buildings must rise over the extension,
so the scaling dimension for the link is always = 2H. For elevated
bridges linking two buildings, where the wind can pass below as
well as above the link, the scaling dimension =f, where 4 is the
vertical distance between bottom (soffit) and top (roof) surfaces of
the link,
25.227 Smollest enclosing rectangle ($3.1.3.3.2)
For a building or arbitrary plan shape, Figure 30 defines the length L
and width W in terms of the smallest enclosing rectangle. Since B=L or
B=W, depending on wind direction, this sets a limit to the maximum
possible value of the scaling length b when used with H in Equation 16,
2523 Comments
There is no doubt that the rules for selecting an appropriate value of h
can be complicated for all but the simplest building shape. But the
guidance given in this section by rules and by example should be
sufficient to avoid gross error. I is impossible to give examples that cover
all possibilities. The key is to think of # as twice the ‘divergence
distance’, First decide whether the wind prefers to flow around or over the
a whole, then decide whether this flow prefers to flow around
2.5.3 External pressure coefficients for walls (§2.4, §3.3.1)
253.1 Basis
253.11 Directional method
‘The directional extemal pressure coefficients for walls are derived from
extensive measurements taken at 15° increments of wind direction. Zones
are defined as vertical strips of increasing width from the upwind comer
of the wall. When the wind is normal to a wall, @—0° or 180°, this
arrangement leads to a notional asymmetry to the loads as shown in Fig.
42, where à small change in wind direction will cause the ‘upwind corner"
to “move” to the opposite end of the wall. This is intended to provide an
automatic allowance for asymmetry in the overall loads, but requires you
Lo assess both the anti-symmettic cases shown in Fig. 42 to ensure (hat the
at À
Pig. #2. Notional osyownety for wind normal o face of building
Commentary | 77
Fig. 41. Principle of ‘reflection
We make te judgement that the size and the set-back of
the extension mens that it does not affect the flow around
the front of the house. So, for the whole house; A— Rn,
H=75m, giving b=8m and the flow prefers to flow
around, rather than over the house. This value of 6=8m
sets the zone sizes for the walls and roof of the house.
2. Our next question is “how docs the wind prefer to flow
around the extension?” It can either flow aver or around one
side. It is prevented from flowing around the other side by
the wall ofthe house. That wall acts like a mirror, making the
effective breadth of the extension twice the actual breadth.
We will call this the principle of ‘reflection’. In this case,
B=8m, H—3m, giving b=6m, and the flow prefers lo
flow over rather than around the’ extension. This value of
b=6m sels the zone sizes for the walls and roof of the
extension
“The principle of “reflection” should be used whenever there is a
small extension on a farger wall. When an extension links two taller
buildings, Now between the buildings must rise over the extension,
so the scaling dimension for the link is always = 2H. For elevated
bridges linking two buildings, where the wind can pass below as
well as above the link, the scaling dimension =f, where 4 is the
vertical distance between bottom (soffit) and top (roof) surfaces of
the link,
25.227 Smollest enclosing rectangle ($3.1.3.3.2)
For a building or arbitrary plan shape, Figure 30 defines the length L
and width W in terms of the smallest enclosing rectangle. Since B=L or
B=W, depending on wind direction, this sets a limit to the maximum
possible value of the scaling length b when used with H in Equation 16,
2523 Comments
There is no doubt that the rules for selecting an appropriate value of h
can be complicated for all but the simplest building shape. But the
guidance given in this section by rules and by example should be
sufficient to avoid gross error. I is impossible to give examples that cover
all possibilities. The key is to think of # as twice the ‘divergence
distance’, First decide whether the wind prefers to flow around or over the
a whole, then decide whether this flow prefers to flow around
2.5.3 External pressure coefficients for walls (§2.4, §3.3.1)
253.1 Basis
253.11 Directional method
‘The directional extemal pressure coefficients for walls are derived from
extensive measurements taken at 15° increments of wind direction. Zones
are defined as vertical strips of increasing width from the upwind comer
of the wall. When the wind is normal to a wall, @—0° or 180°, this
arrangement leads to a notional asymmetry to the loads as shown in Fig.
42, where à small change in wind direction will cause the ‘upwind corner"
to “move” to the opposite end of the wall. This is intended to provide an
automatic allowance for asymmetry in the overall loads, but requires you
Lo assess both the anti-symmettic cases shown in Fig. 42 to ensure (hat the
at À
Pig. #2. Notional osyownety for wind normal o face of building
78 | Wind loading: a practical guide to BS 6399-2
Commentary 479
highest loads are found in every structural member us well as the highest
cladding pressures.
“The less common cases of non-rectangular (polygonal) plan buildings.
and non-vertical walls ae only covered in the directional method. Walls
with triangular gables, resulting from monopitch or duopitch roots, are not
treated differently in the standard method, but are given special zanes in
253.12 Standard method
Standard method coefficients are simplified to a single zone on wind.
ward (front) and leeward (rear) walls, with zones on side walls defined
from the upwind comer, forcing symmetry as noted earlier. This requires
the effect of asymmetric loads to be considered separately (in $27.37.
see 26.4 later). The windward and leeward values, introduced in the 2002
amendment, eortespond to the must onerous cases at wind directions I
and 45°, respectively. These would give over-conservative net loads, so
net pressure coefficients are given separately in Table Sa).
25.3.1.3 Funnelling
Because other changes have led to an average reduction in loads, the
effects uf funnelling of the wind between buildings can become cri
and are included for Ihe first time in any code. It was previously subsumed
into the average conservatism of CP 3-V-2, but the level of observed
damage lo cladding and to gable ends of houses gave rise 10 concerns,
prompting a parametric study of the effects on which the rules are based.
‘The 2002 amendment added sub-clauses (§2.1.3.7e and $3.3.1.1.3d) which
exclude funnelling when completely sheltered by upwind buildings.
Note: Funnelting is one of two conditions that may act ta increase wind
loads on an existing building when a new building is constructed
nearby. The other condition is when a tal building causes higher wind
speeds to act un low-rise buildings around its base (see $17.34)
2532 Application
2.5.3.2.1 Directional method (83.3.1)
For each wind direction in the directional method the following
procedure should be carried out.
© Determine the scaling length b applicable tw the walls.
Determine the span ratio of the building DVI.
I dhere is a neighbouring building facing a side wall and the effec-
tive height of the lower building is greater than 0-4 H,, determine the.
idth of the gap, otherwise funnelling does not oces
le the side walls into the zones A, B and € from the upwind cor-
ee re
a
Look up the values of extn pressure coefficen fr each zone
tom:
Q Table 26 for vertical walls, interpolating between the two sets
of values in the range 1 < D/H <4
Table 28 for triangular gables and vertical gables adjacent to
non-venical walls (A-frame buildings), interpolating between
the two sets of values in the range 1 < DH < 4, and
© Table 29 for windward-facing non-vertical walls. In this case
DH has no influence.
If the gap between buildings is in the range 0:25b-< gap width <b
funnelling will occur on the region of wall within the gap, eas in
the hatched zone shown in Fig. 43.
© Funnelling occurs for wind directions 445" either side of the
axis of the gap, but itis reasonable to assume that this angle
would be limited by any sheltering of the gap by the building
or its neighbour as indicated in the figure
© The rules of §3.3.1.1.3 are simplified in Fig. 44 as factors to be
applied to zone A and zone B coefficients, depending on the
gap width. You must also replace zone C coefficients by the
factored zone B values,
For polygonal-plan buildings, first check that the length of the
auljacent upwind wall is greater than /5, then look up and apply the
reduction factor given in Table 27 to the values of suctions in zone
AA. You should not apply these factors to positive values. If the
ona
regent
cee
‘comeing eg
Fig. 43. Range of wind direction for fimnellng
Commentary 479
highest loads are found in every structural member us well as the highest
cladding pressures.
“The less common cases of non-rectangular (polygonal) plan buildings.
and non-vertical walls ae only covered in the directional method. Walls
with triangular gables, resulting from monopitch or duopitch roots, are not
treated differently in the standard method, but are given special zanes in
253.12 Standard method
Standard method coefficients are simplified to a single zone on wind.
ward (front) and leeward (rear) walls, with zones on side walls defined
from the upwind comer, forcing symmetry as noted earlier. This requires
the effect of asymmetric loads to be considered separately (in $27.37.
see 26.4 later). The windward and leeward values, introduced in the 2002
amendment, eortespond to the must onerous cases at wind directions I
and 45°, respectively. These would give over-conservative net loads, so
net pressure coefficients are given separately in Table Sa).
25.3.1.3 Funnelling
Because other changes have led to an average reduction in loads, the
effects uf funnelling of the wind between buildings can become cri
and are included for Ihe first time in any code. It was previously subsumed
into the average conservatism of CP 3-V-2, but the level of observed
damage lo cladding and to gable ends of houses gave rise 10 concerns,
prompting a parametric study of the effects on which the rules are based.
‘The 2002 amendment added sub-clauses (§2.1.3.7e and $3.3.1.1.3d) which
exclude funnelling when completely sheltered by upwind buildings.
Note: Funnelting is one of two conditions that may act ta increase wind
loads on an existing building when a new building is constructed
nearby. The other condition is when a tal building causes higher wind
speeds to act un low-rise buildings around its base (see $17.34)
2532 Application
2.5.3.2.1 Directional method (83.3.1)
For each wind direction in the directional method the following
procedure should be carried out.
© Determine the scaling length b applicable tw the walls.
Determine the span ratio of the building DVI.
I dhere is a neighbouring building facing a side wall and the effec-
tive height of the lower building is greater than 0-4 H,, determine the.
idth of the gap, otherwise funnelling does not oces
le the side walls into the zones A, B and € from the upwind cor-
ee re
a
Look up the values of extn pressure coefficen fr each zone
tom:
Q Table 26 for vertical walls, interpolating between the two sets
of values in the range 1 < D/H <4
Table 28 for triangular gables and vertical gables adjacent to
non-venical walls (A-frame buildings), interpolating between
the two sets of values in the range 1 < DH < 4, and
© Table 29 for windward-facing non-vertical walls. In this case
DH has no influence.
If the gap between buildings is in the range 0:25b-< gap width <b
funnelling will occur on the region of wall within the gap, eas in
the hatched zone shown in Fig. 43.
© Funnelling occurs for wind directions 445" either side of the
axis of the gap, but itis reasonable to assume that this angle
would be limited by any sheltering of the gap by the building
or its neighbour as indicated in the figure
© The rules of §3.3.1.1.3 are simplified in Fig. 44 as factors to be
applied to zone A and zone B coefficients, depending on the
gap width. You must also replace zone C coefficients by the
factored zone B values,
For polygonal-plan buildings, first check that the length of the
auljacent upwind wall is greater than /5, then look up and apply the
reduction factor given in Table 27 to the values of suctions in zone
AA. You should not apply these factors to positive values. If the
ona
regent
cee
‘comeing eg
Fig. 43. Range of wind direction for fimnellng
80 | Wind loading: a practical guide to BS 6399-2
Commentary 1 81
i
PE
Ea]
on
Fig. 44. Interpolation of fnmeling factor in directional method
reduced A zone suetions are less than (less negative) the suctio
the B or C zones, you should replace the B or C zone values wit
these less onerous values. Example 8 shows these rules applied to a
hoxagonal-plan building.
Example 8. Pressure coefficients for walls of a regular
hexagonal-plan building
We consider a 20m high building with a plan shape that is à regular hexagon with
sides of 10m in length and determine she external pressure coefficients on the walls
for wind angles of # —4P and O = 307 Figure (a) below shows thatthe dimensions.
4 ~ 20m ul B =20m give b =20m, so that the width ofthe A zone on each wall is
an and zones C and D do not exis,
‘
al B ABRE
HAE EN
: el E Fl4 el E
oh A mr
|
stile | ale [Sa
SAS dla alas [a] = [3/5
Weal EP ARAL
Soie ee
The span ratio PYH= 1, so that values of Cp are taken from the Ind si
able 26. The length ofthe tace is greater than AS and he corner angle is = L
that the reduction factor For suetions in zone A is 06 from Table 27.
Figures (b) and (0) show all the walis ofthe building opened out Mat
© Values of Co for wind angle
(nr, 120" and 180° in Table 27.
© Values of Cpe for wind angle 8" are shown in (e) and correspond 10
= 30", 90" and 150° in Table 27.
® Te reduced zone A value is lower and replaces the zone B value in every
instance.
ate shown in (b) and correspond to 0 0,
Note: Depending on the building proportions, zone C of the standard
method, and zones C or D of the directional method may not exit.
2.5.3.2.2 Standard method (§2.4, 63.3.1.5)
For each orthogonal load case in the standard method the follow
procedure should be followed.
a
de the side walls into zones A, B and C in vertical strips
starting from the upwind comer, using the key Figures 12, 14 and
15 as reference,
9 Look up the values of external pressure coefficient for each zone
from Table 5, and the net coefficients frum Table Sa)
O Windward (front) and leeward (rear) cocflicients should be
lincarly interpolated for span ratio in the range 1 < D/H < 4.
To avoid interpolation, you may safely use the values for
D/H — 1 in this range.
© Zone A, Band C coefficients on the side walls should be taken
from the ‘Isolated’ column when funnelling does not occur,
The criteria for funnelling are shown in Fig, 43. The rules of
§2.4.1.4 are simplified in Kig. 45 as coeffi
that depend on the gap width,
“Treat these as being vertical walls of the same height for a
pe Non-vertical walls
It.
‘conservative re
Commentary 1 81
i
PE
Ea]
on
Fig. 44. Interpolation of fnmeling factor in directional method
reduced A zone suetions are less than (less negative) the suctio
the B or C zones, you should replace the B or C zone values wit
these less onerous values. Example 8 shows these rules applied to a
hoxagonal-plan building.
Example 8. Pressure coefficients for walls of a regular
hexagonal-plan building
We consider a 20m high building with a plan shape that is à regular hexagon with
sides of 10m in length and determine she external pressure coefficients on the walls
for wind angles of # —4P and O = 307 Figure (a) below shows thatthe dimensions.
4 ~ 20m ul B =20m give b =20m, so that the width ofthe A zone on each wall is
an and zones C and D do not exis,
‘
al B ABRE
HAE EN
: el E Fl4 el E
oh A mr
|
stile | ale [Sa
SAS dla alas [a] = [3/5
Weal EP ARAL
Soie ee
The span ratio PYH= 1, so that values of Cp are taken from the Ind si
able 26. The length ofthe tace is greater than AS and he corner angle is = L
that the reduction factor For suetions in zone A is 06 from Table 27.
Figures (b) and (0) show all the walis ofthe building opened out Mat
© Values of Co for wind angle
(nr, 120" and 180° in Table 27.
© Values of Cpe for wind angle 8" are shown in (e) and correspond 10
= 30", 90" and 150° in Table 27.
® Te reduced zone A value is lower and replaces the zone B value in every
instance.
ate shown in (b) and correspond to 0 0,
Note: Depending on the building proportions, zone C of the standard
method, and zones C or D of the directional method may not exit.
2.5.3.2.2 Standard method (§2.4, 63.3.1.5)
For each orthogonal load case in the standard method the follow
procedure should be followed.
a
de the side walls into zones A, B and C in vertical strips
starting from the upwind comer, using the key Figures 12, 14 and
15 as reference,
9 Look up the values of external pressure coefficient for each zone
from Table 5, and the net coefficients frum Table Sa)
O Windward (front) and leeward (rear) cocflicients should be
lincarly interpolated for span ratio in the range 1 < D/H < 4.
To avoid interpolation, you may safely use the values for
D/H — 1 in this range.
© Zone A, Band C coefficients on the side walls should be taken
from the ‘Isolated’ column when funnelling does not occur,
The criteria for funnelling are shown in Fig, 43. The rules of
§2.4.1.4 are simplified in Kig. 45 as coeffi
that depend on the gap width,
“Treat these as being vertical walls of the same height for a
pe Non-vertical walls
It.
‘conservative re
82 | Wind loading: a practical guide to BS 6399-2
Commentary 1-83
Fig, 45. Interpolarion of Cy. for felling un wal in standard method
25324. Polygonol-plan buildings (62.4.2)
Clause 2.4.2 requires the coefficients for polygonal plan
buildings to be obtained using the directional method. However
the concept of the ‘smallest enclosing rectangle’ in 2.5,2.2.7 allows
vs to get around this restriction, as Follows
O External pressures. (As Table 27 shows that the high local
suctions in the A zone of walls are greatest for a $ =90°
corner, it follows that the standurd external pressure
coefficients in Table 5 are always adequate for cladding
design,
© Overall loads. (Because the area of the front and rear faces of
the ‘smallest enclosing rectangle’ are always equal to the
sum of the face arcas resolved into the wind direction (see
35.14) and the corresponding cocfficienls in Table 5 are
always more onerous than those in Table 29, it follows that
the overall loads on the equivalent rectangular building are
conservative.
25.325. Circular-plan buildings (82.4.6)
Coefficients for che walls of eircular-plan buildings are given in
Table 7 and are the same as used previously in CP 3-V-2.
© You may use the values in Table 7 for ares of curved walls,
provided that you include zones A and B of flat walls at any
sharp comer. You may also use these values for regular
polygonal buildings with ten or more faces, or ares made of
many flat faces.
You may estimate the overall drag force on a cylindrical
building by taking a net pressure coefficient of C,=06
if on the arca in elevation. (This value is compatible
with the factor of 0-85 in Equation 7 —see 2.6.1.3.)
25.3.2.6 Special considerations
Clauses 2.4.3 and 2.4.4 in the standard method and §3.3.1.5 (0
$3.2.1.8 in the directional method define the special considerations
needed for buildings with re-entran! corners, recessed bay’, internal
wells and imegular or inset faces. They cover the cases of upwind
wings and zones on inset faces, explained above. The following
additional considerations apply to walls:
© Zones for re-entrant corners. Figure 46 shows the special
considerations for this ease.
(4) Standard method, When the re-entrant comer faces down-
wind, the side wall is treated as the leeward (rear) face, but
when it faces upwind the side wall has the normal zones.
(b) Directional method. When the re-entrant corner faces
downwind, the side wall is ueated as Zine C of the leeward
Gear) face. When it faces upwind, the side wall has the
normal zones, but the zones on the windward facing wall
depend on whether the outside or inside corner ix upwind.
When the outside corer faces upwind, the A and B zones.
are defined from the corner and you also need to add the
additional wedge-shaped zone of positive pressure defined
in Figure 33. Zones A and B should never be defined from
aan inside corner- see 3.4.4.1.)
Fig. 46. Walls of reentrant comer
Commentary 1-83
Fig, 45. Interpolarion of Cy. for felling un wal in standard method
25324. Polygonol-plan buildings (62.4.2)
Clause 2.4.2 requires the coefficients for polygonal plan
buildings to be obtained using the directional method. However
the concept of the ‘smallest enclosing rectangle’ in 2.5,2.2.7 allows
vs to get around this restriction, as Follows
O External pressures. (As Table 27 shows that the high local
suctions in the A zone of walls are greatest for a $ =90°
corner, it follows that the standurd external pressure
coefficients in Table 5 are always adequate for cladding
design,
© Overall loads. (Because the area of the front and rear faces of
the ‘smallest enclosing rectangle’ are always equal to the
sum of the face arcas resolved into the wind direction (see
35.14) and the corresponding cocfficienls in Table 5 are
always more onerous than those in Table 29, it follows that
the overall loads on the equivalent rectangular building are
conservative.
25.325. Circular-plan buildings (82.4.6)
Coefficients for che walls of eircular-plan buildings are given in
Table 7 and are the same as used previously in CP 3-V-2.
© You may use the values in Table 7 for ares of curved walls,
provided that you include zones A and B of flat walls at any
sharp comer. You may also use these values for regular
polygonal buildings with ten or more faces, or ares made of
many flat faces.
You may estimate the overall drag force on a cylindrical
building by taking a net pressure coefficient of C,=06
if on the arca in elevation. (This value is compatible
with the factor of 0-85 in Equation 7 —see 2.6.1.3.)
25.3.2.6 Special considerations
Clauses 2.4.3 and 2.4.4 in the standard method and §3.3.1.5 (0
$3.2.1.8 in the directional method define the special considerations
needed for buildings with re-entran! corners, recessed bay’, internal
wells and imegular or inset faces. They cover the cases of upwind
wings and zones on inset faces, explained above. The following
additional considerations apply to walls:
© Zones for re-entrant corners. Figure 46 shows the special
considerations for this ease.
(4) Standard method, When the re-entrant comer faces down-
wind, the side wall is treated as the leeward (rear) face, but
when it faces upwind the side wall has the normal zones.
(b) Directional method. When the re-entrant corner faces
downwind, the side wall is ueated as Zine C of the leeward
Gear) face. When it faces upwind, the side wall has the
normal zones, but the zones on the windward facing wall
depend on whether the outside or inside corner ix upwind.
When the outside corer faces upwind, the A and B zones.
are defined from the corner and you also need to add the
additional wedge-shaped zone of positive pressure defined
in Figure 33. Zones A and B should never be defined from
aan inside corner- see 3.4.4.1.)
Fig. 46. Walls of reentrant comer
84 | Wind Inading: a practical guide to BS 6399-2
Commentary 1 85
(o) ifgap<b/2 (b) Define zones as solid wall (0) Add extra A zones
Fig, 47, Wall of recessed bay
Note: This ‘wedge’ is important for tall buildings, where b ~ B,
because it will fll the corner in skew winds (9 = 45°) and is a
significant component af the overall horizontal load.
Remember that in the directional method the wind angle à is
taken from normal to the wall, so the value is different for each wall.
Recessed bays. If the gap across the bay is les than D/2 then
follow the rules in §2-4.3.2 or 83.3.1.6 and Figure 34a)
Figure 47 simplifies the rules into the three steps shown,
‘Assess the extemal wall zones as if the bay was not there.
Apply the average pressure across the gap 10 the walls
inside the bay.
‘Add the extra A zones adjacent to each upwind corner,
If the gap is greater than b/2, treat the bay as two re
entrant comers.
oo 00
Note: The threshold gap size is given as b in the standard method
§2.4.3, which is not consistent with b/2 used in the
directional method §3.3.1.6. You are advised to adopt the
more onerous criterion of the directional method.
Note: Remember ro add the ‘wedge’ of positive pressure in skew
winds as defined in Figure 34(b).
© Internal wells. IF the gap across the well is les than b/2 then
follow the rules in $2.4.3.2 or $3.3.1.7. The steps are the
same as for re-entrant corners, except thatthe pressure zones
ae assessed for the roof instead of the external wal, IF the
gap is greater than b/2, teat the bay as four verentrant
comers.
25.3.2.7 Friction loads ($24.5, §3.3.1.9)
wind only on the C zones and D
shod and the C zones in the standard method.
zones in the directional n
O Determine the load using Equation 7a, where the swept arca A
the area of the C and D zones on both side faces and the friction
coefficients are given in Table 6.
Take this load to act parallel to the walls at half their height.
(Remember that there will be an additional component from the
mol)
Buildings need to be very long before friction loads become significant.
‘The horizontal force on a long terrace when the wind is parallel to the
Jong axis is a typical example. (See also 2.6.14 and 3.4.6.1.)
2.54 External pressure coefficients for roofs (§2.5, $3.3,2-$3.3.4)
2541 Basis
254.1.1 Directional method
‘The directional external pressure coefficients for ruofs are derived from
extensive measurements taken at 15° increments of wind direction for flat
roofs and 30° for pitched roofs. In the directional method, the wind angle
is taken from normal tothe eaves, and zones are defined as parallel strips
of various widths from the upwind eaves and the upwind verge. When the
wind is normal to the caves ur verge, 8=0°, 90°, 180° or 270°, this
arrangement Jeads to a notional asymmetry to the loads as shown in Fig.
48, where a small change in wind direction will cause the ‘upwind corner
10 move 10 the opposite verge. This is intended to provide an automatic
allowance for asymmetry in the overall loads, but requires you to assess
both the anti-symmetric cases shown in Fig. 48 to ensure that the highest
loads are found in every structural member as well as the highest cladding
pressure.
2.5.4.1.2 Standard method
In che standard method, coefficients are simplified to give zones behind
the upwind caves and ridge for the orthogonal case of wind normal 10
Eu
Eco pind
4“
Fig. 48. Novional asymunciy for wind normal o eaves
Commentary 1 85
(o) ifgap<b/2 (b) Define zones as solid wall (0) Add extra A zones
Fig, 47, Wall of recessed bay
Note: This ‘wedge’ is important for tall buildings, where b ~ B,
because it will fll the corner in skew winds (9 = 45°) and is a
significant component af the overall horizontal load.
Remember that in the directional method the wind angle à is
taken from normal to the wall, so the value is different for each wall.
Recessed bays. If the gap across the bay is les than D/2 then
follow the rules in §2-4.3.2 or 83.3.1.6 and Figure 34a)
Figure 47 simplifies the rules into the three steps shown,
‘Assess the extemal wall zones as if the bay was not there.
Apply the average pressure across the gap 10 the walls
inside the bay.
‘Add the extra A zones adjacent to each upwind corner,
If the gap is greater than b/2, treat the bay as two re
entrant comers.
oo 00
Note: The threshold gap size is given as b in the standard method
§2.4.3, which is not consistent with b/2 used in the
directional method §3.3.1.6. You are advised to adopt the
more onerous criterion of the directional method.
Note: Remember ro add the ‘wedge’ of positive pressure in skew
winds as defined in Figure 34(b).
© Internal wells. IF the gap across the well is les than b/2 then
follow the rules in $2.4.3.2 or $3.3.1.7. The steps are the
same as for re-entrant corners, except thatthe pressure zones
ae assessed for the roof instead of the external wal, IF the
gap is greater than b/2, teat the bay as four verentrant
comers.
25.3.2.7 Friction loads ($24.5, §3.3.1.9)
wind only on the C zones and D
shod and the C zones in the standard method.
zones in the directional n
O Determine the load using Equation 7a, where the swept arca A
the area of the C and D zones on both side faces and the friction
coefficients are given in Table 6.
Take this load to act parallel to the walls at half their height.
(Remember that there will be an additional component from the
mol)
Buildings need to be very long before friction loads become significant.
‘The horizontal force on a long terrace when the wind is parallel to the
Jong axis is a typical example. (See also 2.6.14 and 3.4.6.1.)
2.54 External pressure coefficients for roofs (§2.5, $3.3,2-$3.3.4)
2541 Basis
254.1.1 Directional method
‘The directional external pressure coefficients for ruofs are derived from
extensive measurements taken at 15° increments of wind direction for flat
roofs and 30° for pitched roofs. In the directional method, the wind angle
is taken from normal tothe eaves, and zones are defined as parallel strips
of various widths from the upwind eaves and the upwind verge. When the
wind is normal to the caves ur verge, 8=0°, 90°, 180° or 270°, this
arrangement Jeads to a notional asymmetry to the loads as shown in Fig.
48, where a small change in wind direction will cause the ‘upwind corner
10 move 10 the opposite verge. This is intended to provide an automatic
allowance for asymmetry in the overall loads, but requires you to assess
both the anti-symmetric cases shown in Fig. 48 to ensure that the highest
loads are found in every structural member as well as the highest cladding
pressure.
2.5.4.1.2 Standard method
In che standard method, coefficients are simplified to give zones behind
the upwind caves and ridge for the orthogonal case of wind normal 10
Eu
Eco pind
4“
Fig. 48. Novional asymunciy for wind normal o eaves
86 | Wind loading: a practical guide to BS 6309-2
Commentary | 87
eaves (0 =0" and 180%) and behind the upwind verges for wind parallel to
the eaves (0—90" and 270), forcing symmetry as noted earlier. This
requires the sensitivity of the Structure to possible asymmetric loads 10 be
considered separately (in $2..3.7, see 26.4 later).
2.54.13 Range ofdata
Previous information given for Mat roofs has been improved and
extended to apply 10 polygonal buildings and to include the effects of
parapets, curved and mansard eaves. Information on pitched mofs now
includes specific data for hipped roofs. Data for some other unusual root
forms not included in BS 6399-2 (hyperboloid, skew-hipped) may be
found in Part 2 of reference 1. Some guidance is given for the effect of
parapet and eave details on pitched roofs, but this guidance is speculative
since itis based on flat roof data, The data for free-standing canopy roofs,
introduced in CP 3-V-2 in the 1986 amendments, have been retained.
‘Allowance for the reduced loads on downwind bays af multi-hay or ‘sas
tooth’ roofs has been retained by replacing the previous data tables wit
reduction factors that are applied to the pressures for the upwind bay.
254.14 Funnelling
imnelling of the wind between buildings is assumed not to have u
significant effect on roofs. However, roofs are susceptible to increased
wind speeds at roof level caused by a neighbouring high-rise building. A
warning to this effect is given in $7.7.3.4, but no useful data are included
10 support it.
2.5.42 Application to all roofs
Note: The guidance in this section applies to all roof forms, When the
steps in this section have been completed, follow the remaining
{guidance for the specific roof form.
2542.1 General rules
O All roots with a pitch angle fess than «x = 5° should be considered
(0 be Mat. Any ridges or troughs that result from the Fall of the roof
are ignored and zones are defined only from the upwind eaves.
7 For each wind direction in the directional method or orthogonal case
in the standard method the following guidance should be used.
© Determine the scaling length b applicable to the roof. Bear in
mind that wich duopitch, mult-pitch and multi bay buildings
the crosswind breadth Y applies to the whole building, unless
there is an upwind wing or a step change in height.
© Determine the pitch angle a of each face of the roof. ‘This is
defined in the key figure for each roof type. The pitch angle
for troughed roofs is negative.
© If there is a parapet, determine the parapet height ratio Iyb,
using Figure 17 as reference.
© If the eaves are curved, determine the radius ratio 7/0, using
Figure 17 as reference.
© For mansard eaves, first check that the mansard extends at
Teast 4/10 behind the wall, then determine the mansard pitch
angle 0.
Note: In Figure 17(c) the horizontal dimension of she mansard show
read ‘> b/10" dl ind
254.22. Roof overhangs ($2.5.8)
D When a roof overhangs the supporting walls at eaves or verge hy
less than 8/10, apply the pressure on the adjacent wall to the under-
surface (soffit) of the overhang.
© When a roof overhangs the supporting walls at eaves or verge by
more than B/10, the pressure on the adjacent wall to the under-
surface (soffit) of the overhang should be the internal pressure for
an open-sided building, as described later in 2.5.5.4
25423 Frictionloads (§2.5.10, 63.3.2.8,63.3.3.9)
Friction loads are generated by the wind on long roofs. Determine the
load using Zquarion 7a.
The area swept by the wind As is for:
© Mat roofs, the arca of the D zone of the standard method
(82.5.10) and the G zone of the directional method (§3.3.2.8)
© pitched roofs, when the wind is parallel to the ridge, the area
of the D zone of the standard method ($2,5.10) and zones F
and P of the directional method (3.3.2.8)
© barrel-vault roofs, when the wind is parallel to the axis of the
vault, the zone F defined in Fig, 54 later.
© Friction coefficients are given in Table 6. Take this load to act
parallel to the roof in the direction of the wind. (Remember that
there will be an additional component from the side wall.) (See
also 2.6.1.4 and 3.4.6.1.)
25.43 Flat roofs
2.543.) Directional method (§3.3.2)
For each wind angle of interest
Commentary | 87
eaves (0 =0" and 180%) and behind the upwind verges for wind parallel to
the eaves (0—90" and 270), forcing symmetry as noted earlier. This
requires the sensitivity of the Structure to possible asymmetric loads 10 be
considered separately (in $2..3.7, see 26.4 later).
2.54.13 Range ofdata
Previous information given for Mat roofs has been improved and
extended to apply 10 polygonal buildings and to include the effects of
parapets, curved and mansard eaves. Information on pitched mofs now
includes specific data for hipped roofs. Data for some other unusual root
forms not included in BS 6399-2 (hyperboloid, skew-hipped) may be
found in Part 2 of reference 1. Some guidance is given for the effect of
parapet and eave details on pitched roofs, but this guidance is speculative
since itis based on flat roof data, The data for free-standing canopy roofs,
introduced in CP 3-V-2 in the 1986 amendments, have been retained.
‘Allowance for the reduced loads on downwind bays af multi-hay or ‘sas
tooth’ roofs has been retained by replacing the previous data tables wit
reduction factors that are applied to the pressures for the upwind bay.
254.14 Funnelling
imnelling of the wind between buildings is assumed not to have u
significant effect on roofs. However, roofs are susceptible to increased
wind speeds at roof level caused by a neighbouring high-rise building. A
warning to this effect is given in $7.7.3.4, but no useful data are included
10 support it.
2.5.42 Application to all roofs
Note: The guidance in this section applies to all roof forms, When the
steps in this section have been completed, follow the remaining
{guidance for the specific roof form.
2542.1 General rules
O All roots with a pitch angle fess than «x = 5° should be considered
(0 be Mat. Any ridges or troughs that result from the Fall of the roof
are ignored and zones are defined only from the upwind eaves.
7 For each wind direction in the directional method or orthogonal case
in the standard method the following guidance should be used.
© Determine the scaling length b applicable to the roof. Bear in
mind that wich duopitch, mult-pitch and multi bay buildings
the crosswind breadth Y applies to the whole building, unless
there is an upwind wing or a step change in height.
© Determine the pitch angle a of each face of the roof. ‘This is
defined in the key figure for each roof type. The pitch angle
for troughed roofs is negative.
© If there is a parapet, determine the parapet height ratio Iyb,
using Figure 17 as reference.
© If the eaves are curved, determine the radius ratio 7/0, using
Figure 17 as reference.
© For mansard eaves, first check that the mansard extends at
Teast 4/10 behind the wall, then determine the mansard pitch
angle 0.
Note: In Figure 17(c) the horizontal dimension of she mansard show
read ‘> b/10" dl ind
254.22. Roof overhangs ($2.5.8)
D When a roof overhangs the supporting walls at eaves or verge hy
less than 8/10, apply the pressure on the adjacent wall to the under-
surface (soffit) of the overhang.
© When a roof overhangs the supporting walls at eaves or verge by
more than B/10, the pressure on the adjacent wall to the under-
surface (soffit) of the overhang should be the internal pressure for
an open-sided building, as described later in 2.5.5.4
25423 Frictionloads (§2.5.10, 63.3.2.8,63.3.3.9)
Friction loads are generated by the wind on long roofs. Determine the
load using Zquarion 7a.
The area swept by the wind As is for:
© Mat roofs, the arca of the D zone of the standard method
(82.5.10) and the G zone of the directional method (§3.3.2.8)
© pitched roofs, when the wind is parallel to the ridge, the area
of the D zone of the standard method ($2,5.10) and zones F
and P of the directional method (3.3.2.8)
© barrel-vault roofs, when the wind is parallel to the axis of the
vault, the zone F defined in Fig, 54 later.
© Friction coefficients are given in Table 6. Take this load to act
parallel to the roof in the direction of the wind. (Remember that
there will be an additional component from the side wall.) (See
also 2.6.1.4 and 3.4.6.1.)
25.43 Flat roofs
2.543.) Directional method (§3.3.2)
For each wind angle of interest
38 1
Wind loading: a practical guide ro BS 6399-2
Commentary | 89
Determine the zones using the following steps, as defined in
$33.22, using Figure 35 as the key and Figure 36 as an example.
Draw lines parallel (othe wind direction from each upwind corner
of the wo. |
From each extemal comer, mark out he edge zones A, B, C and
D Depending on the roof proportions, zone D may not exis.
When the eaves are curved or mansard, the zones start from the
upwind edge of the fat roof as shown in Figure 17.
From any intemal corner, mark out the edge zone as zone D only.
Behind the edge zones, mark out zones E and F. Zone F may not
‘Any remaining area of ruof behind these zones is taken as Zone G.
+ Depending on the building proportions, zones D, F or G of the
directional method may not exist
Look up the values of external pressure coefficient for exch zone
as follows.
© Sharp eaves (§3.3.2.3). From Table 30.
© Parapets ($3.3.2.4). Pressure coefficients from Table 30 and
the corresponding reduction factor from Table 31. You may
interpolate in Tuhle 3/ for the parapet height rat
duads on the parapet are obtained by eating
standing wall.
© Curved eaves (§3.3.2.5). From Table 32. You may interpolato.
for the radius 1/6, taking Table 30 to apply for #b—0. You
obtain the pressures around the arc of the eaves by
interpolating between the pressure on the adjacent wall and
roof zones (92.52.32).
O Mansard eaves (§3.3.2.6). From Table 33. You may
interpolate for the mansard pitch angle ay, taking Table 30 to
apply for © =90*, If ax < 30° ignore the mansard eaves and
treat the roof as sharp-edged. You obtain the pressures an the
mansard by treating it as a pitched roof.
as a free
oe
254.3.2 Standard method (§2.5.1)
This is suitable for ruofs of rectangular plan only, but may include cut
‘outs from recessed bays or wells, providing these are also rectangular, For
each orthogonal case consider the follow
7 When the caves are curved or mansard, the zones start from the
‘upwind edge of the flat roof as shown in Figure 17.
1 Divide the ruaf into strips parallel to the upwind eaves: the edge
stip is 8/10 deep and the next stip is /2 deep, measured from the
ind edge of the Mat roof. Use Figure 16 as the reference.
ide the first strip into zones A and B, where zone A extends a
stance of bid from either corner.
M The second strip becomes zone C and any area of the roof that
exists downwind is zone D.
© Look up the values of external pressure cucffici
from Table 8.
© Sharp eaves ($2.5.1.3). Read the values directly from the top
row of Table 8.
O Parapets (§2.5.1.4). You may interpolate for the parapet
height ratio Ab, taking the sharp eaves values to apply for
‘h/o—0. The load on the parapet is obtained by treating it as
a free-standing wall
© Curved eaves (§2.5.1.5). You may interpolate for the radius
2%, taking the sharp eaves values to apply for »/9=0, You
obtain the pressures around the are of the caves by
interpolating between the pressure on the adjacent wall and
roof zones ($2.5.2.5.2).
© Mansard eaves (2.5.1.6). You may interpolate for the
mansard pitch angle a, taking the sharp eaves values to apply
for a =90", If à < 30’ ignore the mansard eaves and treat the
roof as sharp-eilgeil. You obtain the pressures on the mansard
by treating it as a pitched roof.
O Coefficient for polygonal-plan buildings must be obtained using
the directional method.
nt for each zone
Note: Depending on the building proportions, zone D of the standard
method may not exist.
2.5433 Polygonol-plan buildings
It is expected that you will wish to use the directional method for
fatroofed polygonal-plan buildings, e.g. for the hexagonal-plan
building in Example 8. However, the standard external pressure
coefficients in Table 8 are conservative for polygonal-plan
Wind loading: a practical guide ro BS 6399-2
Commentary | 89
Determine the zones using the following steps, as defined in
$33.22, using Figure 35 as the key and Figure 36 as an example.
Draw lines parallel (othe wind direction from each upwind corner
of the wo. |
From each extemal comer, mark out he edge zones A, B, C and
D Depending on the roof proportions, zone D may not exis.
When the eaves are curved or mansard, the zones start from the
upwind edge of the fat roof as shown in Figure 17.
From any intemal corner, mark out the edge zone as zone D only.
Behind the edge zones, mark out zones E and F. Zone F may not
‘Any remaining area of ruof behind these zones is taken as Zone G.
+ Depending on the building proportions, zones D, F or G of the
directional method may not exist
Look up the values of external pressure coefficient for exch zone
as follows.
© Sharp eaves (§3.3.2.3). From Table 30.
© Parapets ($3.3.2.4). Pressure coefficients from Table 30 and
the corresponding reduction factor from Table 31. You may
interpolate in Tuhle 3/ for the parapet height rat
duads on the parapet are obtained by eating
standing wall.
© Curved eaves (§3.3.2.5). From Table 32. You may interpolato.
for the radius 1/6, taking Table 30 to apply for #b—0. You
obtain the pressures around the arc of the eaves by
interpolating between the pressure on the adjacent wall and
roof zones (92.52.32).
O Mansard eaves (§3.3.2.6). From Table 33. You may
interpolate for the mansard pitch angle ay, taking Table 30 to
apply for © =90*, If ax < 30° ignore the mansard eaves and
treat the roof as sharp-edged. You obtain the pressures an the
mansard by treating it as a pitched roof.
as a free
oe
254.3.2 Standard method (§2.5.1)
This is suitable for ruofs of rectangular plan only, but may include cut
‘outs from recessed bays or wells, providing these are also rectangular, For
each orthogonal case consider the follow
7 When the caves are curved or mansard, the zones start from the
‘upwind edge of the flat roof as shown in Figure 17.
1 Divide the ruaf into strips parallel to the upwind eaves: the edge
stip is 8/10 deep and the next stip is /2 deep, measured from the
ind edge of the Mat roof. Use Figure 16 as the reference.
ide the first strip into zones A and B, where zone A extends a
stance of bid from either corner.
M The second strip becomes zone C and any area of the roof that
exists downwind is zone D.
© Look up the values of external pressure cucffici
from Table 8.
© Sharp eaves ($2.5.1.3). Read the values directly from the top
row of Table 8.
O Parapets (§2.5.1.4). You may interpolate for the parapet
height ratio Ab, taking the sharp eaves values to apply for
‘h/o—0. The load on the parapet is obtained by treating it as
a free-standing wall
© Curved eaves (§2.5.1.5). You may interpolate for the radius
2%, taking the sharp eaves values to apply for »/9=0, You
obtain the pressures around the are of the caves by
interpolating between the pressure on the adjacent wall and
roof zones ($2.5.2.5.2).
© Mansard eaves (2.5.1.6). You may interpolate for the
mansard pitch angle a, taking the sharp eaves values to apply
for a =90", If à < 30’ ignore the mansard eaves and treat the
roof as sharp-eilgeil. You obtain the pressures on the mansard
by treating it as a pitched roof.
O Coefficient for polygonal-plan buildings must be obtained using
the directional method.
nt for each zone
Note: Depending on the building proportions, zone D of the standard
method may not exist.
2.5433 Polygonol-plan buildings
It is expected that you will wish to use the directional method for
fatroofed polygonal-plan buildings, e.g. for the hexagonal-plan
building in Example 8. However, the standard external pressure
coefficients in Table 8 are conservative for polygonal-plan
90 1 Wind loading: practical guide to BS 6399-2
Commentary | 91
buildings, provided that the A and B zones aro defined from every
corner along the upwind eaves.
254.34 Circular-plon buildings
Coefficients for flat roofs of cireular-plan buildings arc not
specifically given in BS 6399-2, so we should use the procedure
given in Part 2 of reference 1, as follows.
O Mark out the standard method zones and corresponding
coefficients on the flat roof of a square-plan building with
breadth and depth equal to the diameter d of the building.
© Superimposc the cylindrical plan as shown in Fig, 49.
Zone A pressures do not occur because there are no €
so substitute zone B for any arcas of zone A.
3,
254.3.5 Special considerations
© Inset storeys. Clause 2.5.1.7 in the standard method and
82.327 in the directional method define the special
‘considerations needed for a flat roof at the base of an inset
storey.
© Mark out the normal roof zones, as described above, except
where the roof is replaced by the inset storey.
© Determine the sealing length 5 for the inset storey. This will
usually be smaller than the value for the roof.
(© Mark out the extra zones on the ruo around the base of the
inset storey, using Figure 18 or Figure 37 as referencx
Note: These figures are inconsistent. Figure 18 shows the extra zone to
extend b/2 all around, whereas Figure 37 shows the upwind zones
X extending b/2 and the downwind zones Y extending for b. Figure
Fig. 49. Zones on the e roof of a circular plan bring
37 is the correct figure, and you would be justified in extending the
extra zone in Figure 18(a} a distance b from the rear face of the
inset storey.
1] O Apply the pressure on the adjacent wall of the inset
storey to the extra roof zone.
Note: Clause 2.5.1.7 refers to ‘pressure coefficient’, whereas §3.3.2.7
correctly refers to “pressure”. The dynamic pressure applicable to
the wall pressure coefficients is at he reference height ofthe inset
Morey walls, not of the lower roof,
17 Upwind wings, re-entrant comers, recossed bays and internal
wells. Specific provisions for re-entrant corners, recessed
bays and internal wells are not given as they are covered by
the general case of Figure 35. But some further explanation
may he helpful.
(a) Upwind wings. The rules for upwind wings in Figure 13
apply to the roof as well as the walls, as described
2.5.2 and Fig. 38. This may result in smaller zones on
the roof of the upwind wing.
(b) Cut out areas. ‘The standard method is still applicable to
fat roofs with cut out areas ifthe building plan and the
areas both have a rectangular envelope,
© Re-entrant comers and recessed hays facing upwind. Define
zones behind all upwind eaves.
O Intemal wells and recessed bays facing crosswind. If the
gap across a cut out is less than »/2 (the rule for walls in
$2.3.1.6) apply only zones A and B behind the up
eaves in the cut out, otherwise define zones behind all
upwind eaves.
© Zones A in the standard method and zones B and C in the
directional method occur only at external comers.
‘These rules are demonstrated in Fig. 50.
Commentary | 91
buildings, provided that the A and B zones aro defined from every
corner along the upwind eaves.
254.34 Circular-plon buildings
Coefficients for flat roofs of cireular-plan buildings arc not
specifically given in BS 6399-2, so we should use the procedure
given in Part 2 of reference 1, as follows.
O Mark out the standard method zones and corresponding
coefficients on the flat roof of a square-plan building with
breadth and depth equal to the diameter d of the building.
© Superimposc the cylindrical plan as shown in Fig, 49.
Zone A pressures do not occur because there are no €
so substitute zone B for any arcas of zone A.
3,
254.3.5 Special considerations
© Inset storeys. Clause 2.5.1.7 in the standard method and
82.327 in the directional method define the special
‘considerations needed for a flat roof at the base of an inset
storey.
© Mark out the normal roof zones, as described above, except
where the roof is replaced by the inset storey.
© Determine the sealing length 5 for the inset storey. This will
usually be smaller than the value for the roof.
(© Mark out the extra zones on the ruo around the base of the
inset storey, using Figure 18 or Figure 37 as referencx
Note: These figures are inconsistent. Figure 18 shows the extra zone to
extend b/2 all around, whereas Figure 37 shows the upwind zones
X extending b/2 and the downwind zones Y extending for b. Figure
Fig. 49. Zones on the e roof of a circular plan bring
37 is the correct figure, and you would be justified in extending the
extra zone in Figure 18(a} a distance b from the rear face of the
inset storey.
1] O Apply the pressure on the adjacent wall of the inset
storey to the extra roof zone.
Note: Clause 2.5.1.7 refers to ‘pressure coefficient’, whereas §3.3.2.7
correctly refers to “pressure”. The dynamic pressure applicable to
the wall pressure coefficients is at he reference height ofthe inset
Morey walls, not of the lower roof,
17 Upwind wings, re-entrant comers, recossed bays and internal
wells. Specific provisions for re-entrant corners, recessed
bays and internal wells are not given as they are covered by
the general case of Figure 35. But some further explanation
may he helpful.
(a) Upwind wings. The rules for upwind wings in Figure 13
apply to the roof as well as the walls, as described
2.5.2 and Fig. 38. This may result in smaller zones on
the roof of the upwind wing.
(b) Cut out areas. ‘The standard method is still applicable to
fat roofs with cut out areas ifthe building plan and the
areas both have a rectangular envelope,
© Re-entrant comers and recessed hays facing upwind. Define
zones behind all upwind eaves.
O Intemal wells and recessed bays facing crosswind. If the
gap across a cut out is less than »/2 (the rule for walls in
$2.3.1.6) apply only zones A and B behind the up
eaves in the cut out, otherwise define zones behind all
upwind eaves.
© Zones A in the standard method and zones B and C in the
directional method occur only at external comers.
‘These rules are demonstrated in Fig. 50.
92 | Wind loading: a practical guide to HS 6399-2
Commentary | 93
(a) gap <br2 (0) gap > 2
Fig. 50. Internal wel in a fla ro, standard meted
25.44 Pitched roofs
25.44. Reference height (§3.3.3.2.2) | \
“The reference height for pitched roofs is always the height above
ground of the highest part of the roof.
2544.2. Alternetive values on windward pitch o
For some zones on the windward pitch, at pitch angles in the range
15°< a < 45%, the tables of external pressure coefficient give (wo values:
‘one positive and one negative (suction). This occurs because the mean
pressure is near zero, bul the gust value varies from positive to negative
¿ue to wind turbulence. You should use the value that is mare onerous in
the design
Note: It was previously possible using CP 3-¥-2 10 abtain zero values of
external pressure pe and net pressure p at pitch angles around
a 30", This never occurs using BS 6399-2,
2.54.43. Directional method (§3.3.3)
For each wind angle of interest adopt the following procedure.
© Monopitch rots ($3.3.3.3)
© The wind angle 0 is taken from normal to the eaves.
© Determine the zones using Figure 38 as the key and Figure
391a) as an example.
© Pitch angle c is positive with the low caves upwind and
negative with the low caves downwind. In the range
=5"< a < +5" you should (reat the roof as being Mat.
© Look up pressure coefficients for each zone from Table 34,
interpolating for piteh ungke.
© Duopiteh roofs (3.3.3.4)
© First check for a couple of special cases:
1. pitch angle or < 10° and the building is narrow W < B.
then §3.3.3.4.3 applies. It is simpler to revert to the
standard method with the corresponding clause §2.5.2.4.2
than to apply §3.3.3.4.3.
2. If there is a difference of more than 5° between the angles
‘of each pitch, itis easier to use the corresponding standard
method procedure described below.
© The wind angle is taken from normal 10 the eaves.
© Determine the zones using Figure 40 as the key and Figure
391b) as an example.
© Pitch angle a is positive when the roof is ridged and negative
when it is troughed. In the range 5°< cr < 45° you should
(reat the roof as flat.
Look up pressure evetficients for cach zone from Table 34 for
the upwind pitch und Table 35 for the downwind pitch,
interpolating for pitch angle.
Note: Depending on the building proportions, zones D, P, N or P may
‚nor exist.
© Hipped roofs ($3.3.3.5)
Unless the building plan is square, you will need to distinguish
betsween the longer main roof, which has a section of ridge, and
the hip end. The key Figure 41 uses the subscript “1° for the
Commentary | 93
(a) gap <br2 (0) gap > 2
Fig. 50. Internal wel in a fla ro, standard meted
25.44 Pitched roofs
25.44. Reference height (§3.3.3.2.2) | \
“The reference height for pitched roofs is always the height above
ground of the highest part of the roof.
2544.2. Alternetive values on windward pitch o
For some zones on the windward pitch, at pitch angles in the range
15°< a < 45%, the tables of external pressure coefficient give (wo values:
‘one positive and one negative (suction). This occurs because the mean
pressure is near zero, bul the gust value varies from positive to negative
¿ue to wind turbulence. You should use the value that is mare onerous in
the design
Note: It was previously possible using CP 3-¥-2 10 abtain zero values of
external pressure pe and net pressure p at pitch angles around
a 30", This never occurs using BS 6399-2,
2.54.43. Directional method (§3.3.3)
For each wind angle of interest adopt the following procedure.
© Monopitch rots ($3.3.3.3)
© The wind angle 0 is taken from normal to the eaves.
© Determine the zones using Figure 38 as the key and Figure
391a) as an example.
© Pitch angle c is positive with the low caves upwind and
negative with the low caves downwind. In the range
=5"< a < +5" you should (reat the roof as being Mat.
© Look up pressure coefficients for each zone from Table 34,
interpolating for piteh ungke.
© Duopiteh roofs (3.3.3.4)
© First check for a couple of special cases:
1. pitch angle or < 10° and the building is narrow W < B.
then §3.3.3.4.3 applies. It is simpler to revert to the
standard method with the corresponding clause §2.5.2.4.2
than to apply §3.3.3.4.3.
2. If there is a difference of more than 5° between the angles
‘of each pitch, itis easier to use the corresponding standard
method procedure described below.
© The wind angle is taken from normal 10 the eaves.
© Determine the zones using Figure 40 as the key and Figure
391b) as an example.
© Pitch angle a is positive when the roof is ridged and negative
when it is troughed. In the range 5°< cr < 45° you should
(reat the roof as flat.
Look up pressure evetficients for cach zone from Table 34 for
the upwind pitch und Table 35 for the downwind pitch,
interpolating for pitch angle.
Note: Depending on the building proportions, zones D, P, N or P may
‚nor exist.
© Hipped roofs ($3.3.3.5)
Unless the building plan is square, you will need to distinguish
betsween the longer main roof, which has a section of ridge, and
the hip end. The key Figure 41 uses the subscript “1° for the
94 | Wind loading: a practical guide to BS 6399-2
| Commentary | 95
main roof and “2° for the hip end to define wind angles 0, and
0%, pitch angles a and az and the zones Ay, Az, Bi, Ba, ete
© The wind angle is taken from normal to the upwind eaves of
the main roof 9, or of the hipped end Or.
O Determine the zones using Figure 47 as the key.
© Pitch angles ay and em are positive when the roof is ridged and
negative when it is roughed. In the range ~5°< cr < +3" you
should treat the roof as flat.
© Look up pressure coefficients for each zone from Table 34 for
the upwind face, Table 35 for the downwind face and Table 36
for the extra zones behind the hip ridges, interpolating for the
respective pitch angle ay or az
Note: Depending on the building proportions, zones D, F or P may not
exist. The extra zones T-Y do not exist when the pitch angle is
negative (troughed)
25444. Stondard method (§2.5.2,§2.5.3)
This is intended for ronfs of rectangular plan only, but is suitable for
roofs that include cut outs from recessed bays or wells, providing these
are also rectangular.
O Scaling length (§2.5.2.2)
As B=L for 0=0° and 180° and B= W for 0=90* and 270°, the
corresponding scaling lengths b from Equation 16 are denoted by
a, and buy in the key figures, Figures 19-21. All zone dimensions
for the orthogonal cases with wind normal to the eaves are defined
using bu and all zone dimensions for the orthogonal cases with
wind normal to the verges are defined using bw.
O Monopiteh mots (§2.5.2.3)
© Wind angle @ is taken from normal to the low eaves,
Symmeiry requires a minimum of three orthogonal directions,
9=00 (low eaves upwind), 90° and 180° (high eaves upwind),
© Determine the zones using Figure 19 and luck up pressure.
coefficients for euch zone from Table 9.
© Duopiteh roots (2.5.2.4)
© Wind angle 0 is taken from normal to the caves, Symmetry
requires a minimum of two orthogonal directions, &
(normal to eaves) and 90° (normal to verge). Positive
angles correspond to ridged roofs, and negative values to
(roughed roofs,
© Check for a couple of special cases:
AR
amo]
EA i
ne
Hg. SÍ. Special cose, riled duoptch roof with < 7 and W <b,
1, Ifpiteh angle a < 7° and the building is narrow W < by, then
$2.5.242 applies. In this ease, extend zone C over E pl
line 10 a distance by/2 from the eaves, making the zone the
sume size as the flat roof case und replacing the ridge zones
Band E This special case is illustrated in Fig. SL.
Note: This special case is quite rare, since the roof is treated as flat for
& <S it can only occur over a 2° range of maf pitch.
2. IP there is a difference of more than 5° between the angles of
each pitch, $2.5.2.4.1 refers you to the original source data in
Part 2 of reference 1. Sale rules for positive-pitch (ridged)
duopitch roofs are as follows.
G Take ay as the angle of the upwind pitch and ap as the
angle of the downwind pitch,
€) For ridged roofs when a> ay (downwind pitch is
stecper) and all troughed roofs, use ay for the zones of the
upwind pitch and cap forthe zones of the downwind pitch
This is conserva
(©) For ridged roofs when ap < cx) (upwind pitch is steeper),
use ay for the zones of the upwind pitch, but treat the
deremsind ich a5 a lat rt wih mansrd eves with
& = au, This is conservative, but is only applicable wi
ay> 30", 7e wien
© Determine the zones using Figure 20 as the key and look up
pressure coefficients for each zone from Table 10. The values
in Table JO assume that the pitch angle is the same for both
pitches, but $2.5.2.4.. notes that they can be upplied as long as
the pitch angles are similar to within 5°,
Note: For 0 =0", zone B will not exist when b= L because each A zon
is b/2 wide, Ñ dd
| Commentary | 95
main roof and “2° for the hip end to define wind angles 0, and
0%, pitch angles a and az and the zones Ay, Az, Bi, Ba, ete
© The wind angle is taken from normal to the upwind eaves of
the main roof 9, or of the hipped end Or.
O Determine the zones using Figure 47 as the key.
© Pitch angles ay and em are positive when the roof is ridged and
negative when it is roughed. In the range ~5°< cr < +3" you
should treat the roof as flat.
© Look up pressure coefficients for each zone from Table 34 for
the upwind face, Table 35 for the downwind face and Table 36
for the extra zones behind the hip ridges, interpolating for the
respective pitch angle ay or az
Note: Depending on the building proportions, zones D, F or P may not
exist. The extra zones T-Y do not exist when the pitch angle is
negative (troughed)
25444. Stondard method (§2.5.2,§2.5.3)
This is intended for ronfs of rectangular plan only, but is suitable for
roofs that include cut outs from recessed bays or wells, providing these
are also rectangular.
O Scaling length (§2.5.2.2)
As B=L for 0=0° and 180° and B= W for 0=90* and 270°, the
corresponding scaling lengths b from Equation 16 are denoted by
a, and buy in the key figures, Figures 19-21. All zone dimensions
for the orthogonal cases with wind normal to the eaves are defined
using bu and all zone dimensions for the orthogonal cases with
wind normal to the verges are defined using bw.
O Monopiteh mots (§2.5.2.3)
© Wind angle @ is taken from normal to the low eaves,
Symmeiry requires a minimum of three orthogonal directions,
9=00 (low eaves upwind), 90° and 180° (high eaves upwind),
© Determine the zones using Figure 19 and luck up pressure.
coefficients for euch zone from Table 9.
© Duopiteh roots (2.5.2.4)
© Wind angle 0 is taken from normal to the caves, Symmetry
requires a minimum of two orthogonal directions, &
(normal to eaves) and 90° (normal to verge). Positive
angles correspond to ridged roofs, and negative values to
(roughed roofs,
© Check for a couple of special cases:
AR
amo]
EA i
ne
Hg. SÍ. Special cose, riled duoptch roof with < 7 and W <b,
1, Ifpiteh angle a < 7° and the building is narrow W < by, then
$2.5.242 applies. In this ease, extend zone C over E pl
line 10 a distance by/2 from the eaves, making the zone the
sume size as the flat roof case und replacing the ridge zones
Band E This special case is illustrated in Fig. SL.
Note: This special case is quite rare, since the roof is treated as flat for
& <S it can only occur over a 2° range of maf pitch.
2. IP there is a difference of more than 5° between the angles of
each pitch, $2.5.2.4.1 refers you to the original source data in
Part 2 of reference 1. Sale rules for positive-pitch (ridged)
duopitch roofs are as follows.
G Take ay as the angle of the upwind pitch and ap as the
angle of the downwind pitch,
€) For ridged roofs when a> ay (downwind pitch is
stecper) and all troughed roofs, use ay for the zones of the
upwind pitch and cap forthe zones of the downwind pitch
This is conserva
(©) For ridged roofs when ap < cx) (upwind pitch is steeper),
use ay for the zones of the upwind pitch, but treat the
deremsind ich a5 a lat rt wih mansrd eves with
& = au, This is conservative, but is only applicable wi
ay> 30", 7e wien
© Determine the zones using Figure 20 as the key and look up
pressure coefficients for each zone from Table 10. The values
in Table JO assume that the pitch angle is the same for both
pitches, but $2.5.2.4.. notes that they can be upplied as long as
the pitch angles are similar to within 5°,
Note: For 0 =0", zone B will not exist when b= L because each A zon
is b/2 wide, Ñ dd
96 | Wind loading: a practical guide to RS 6399-2
Commentary | 97
Note: Depending on the building proportions, zones B and F may nor
exist on monopitch or duopitch roofs.
(1 Hipped roots (82.5.3)
© Determine the zones using Figure 2/ as the key.
© The pitch angle ofthe hip ends may be different from the pitch
angle of the main roof.
‘© Look up pressure coefficients for each zone from Table 11 for
the corresponding pitch angle.
2.54.45 Special considerations
J) Mixed gable and hip ($3.3.3.6) o
Sometimes a duopitch roof has a gable at one end and a hip
fat the other. Treat each face according to the form that
‘applies at the upwind comer, as shown in Fig. 52, using
either directional, standard or hybrid methods.
Pitched roofs on polygonal buildings
BS 6399-2 does not give specific advice, but Fig. 53 shows
that the directional method can be adapted to apply. Case (a)
often occurs because of site boundary constraints — mark the
verge zones and take the wind angle for these zones from
parallel o the skewed verge, as shown. Many-sided buildings
generally have hipped roots as shown in (b) —treat each facet
as a hip and the wind angle from normal to the eaves. The
standard external pressure coefficients are conservative for
[7
opten
à hip) |
pc Hp
A A
Fig. 32. Re for mixed gable and hip rufe
(6) Duopiten (0) Hipped
Fig. 5%. Pitched ro om polygonal buildings
polygonal-plan buildings, provided that the À and B zones are
defined from each corner along the windward eaves
Pitched roofs on circular-plan buildings
Coefficients for pitched roofs of circular-plan buildings are
not specifically given in BS 6399-2. We should use the
procedure given in Part 2 of reference 1, by applying a
similar procedure as for flat rovfs in 25.4.3.
© Mark out the standard method zones and corresponding
coefficients for a square-plan building of the same roof
form and pitch.
© Superimpose the cylindrical plan as shown in Fig. 49,
only in this case the zones must be defined for the form
af the pitched roof,
© Zone A pressures do not occur because there are no
comers, so substitute zone B for any areas of zone A.
© Inset storeys, upwind wings, re
hays and internal wells ($2.56)
Follow the advice for flat roofs given in the previous section,
but use the zones and cocflicients for the form and pitch of
the roof.
© Mansard or multi-piteh roots (82.54, $3.3.4.1)
‘These are roofs that have one or more changes of pitch angle a
on any face of the roof. A roof with two pitch angles where the
lower piteh is steeper than the upper pitch, as defined in Figure
22{a), is usually called a mansard roof. The del
22 is largely self-explanatory. The general rule
Pitch is treated as if it were the face of a duopiteh or
roof ofthe same pitch angle, depending on whether the venges
are gabled or hipped, except that the edge zones are applied
‘only along the upwind eaves, the main ridge and behind any
local ridge on the windward face created by a change in pitch
angle.
© Parapets (62.57, $3.3.3.7)
© Directional method. Clause 3.3.3.7 gives detailed and
complex rules for parapets on monopitch, duopiteh and
ipped ronfs, While these give more accurate and detailed
estimates than the standard method, the additional benefit
is probably not worth the effort
mL corners, recessed
Commentary | 97
Note: Depending on the building proportions, zones B and F may nor
exist on monopitch or duopitch roofs.
(1 Hipped roots (82.5.3)
© Determine the zones using Figure 2/ as the key.
© The pitch angle ofthe hip ends may be different from the pitch
angle of the main roof.
‘© Look up pressure coefficients for each zone from Table 11 for
the corresponding pitch angle.
2.54.45 Special considerations
J) Mixed gable and hip ($3.3.3.6) o
Sometimes a duopitch roof has a gable at one end and a hip
fat the other. Treat each face according to the form that
‘applies at the upwind comer, as shown in Fig. 52, using
either directional, standard or hybrid methods.
Pitched roofs on polygonal buildings
BS 6399-2 does not give specific advice, but Fig. 53 shows
that the directional method can be adapted to apply. Case (a)
often occurs because of site boundary constraints — mark the
verge zones and take the wind angle for these zones from
parallel o the skewed verge, as shown. Many-sided buildings
generally have hipped roots as shown in (b) —treat each facet
as a hip and the wind angle from normal to the eaves. The
standard external pressure coefficients are conservative for
[7
opten
à hip) |
pc Hp
A A
Fig. 32. Re for mixed gable and hip rufe
(6) Duopiten (0) Hipped
Fig. 5%. Pitched ro om polygonal buildings
polygonal-plan buildings, provided that the À and B zones are
defined from each corner along the windward eaves
Pitched roofs on circular-plan buildings
Coefficients for pitched roofs of circular-plan buildings are
not specifically given in BS 6399-2. We should use the
procedure given in Part 2 of reference 1, by applying a
similar procedure as for flat rovfs in 25.4.3.
© Mark out the standard method zones and corresponding
coefficients for a square-plan building of the same roof
form and pitch.
© Superimpose the cylindrical plan as shown in Fig. 49,
only in this case the zones must be defined for the form
af the pitched roof,
© Zone A pressures do not occur because there are no
comers, so substitute zone B for any areas of zone A.
© Inset storeys, upwind wings, re
hays and internal wells ($2.56)
Follow the advice for flat roofs given in the previous section,
but use the zones and cocflicients for the form and pitch of
the roof.
© Mansard or multi-piteh roots (82.54, $3.3.4.1)
‘These are roofs that have one or more changes of pitch angle a
on any face of the roof. A roof with two pitch angles where the
lower piteh is steeper than the upper pitch, as defined in Figure
22{a), is usually called a mansard roof. The del
22 is largely self-explanatory. The general rule
Pitch is treated as if it were the face of a duopiteh or
roof ofthe same pitch angle, depending on whether the venges
are gabled or hipped, except that the edge zones are applied
‘only along the upwind eaves, the main ridge and behind any
local ridge on the windward face created by a change in pitch
angle.
© Parapets (62.57, $3.3.3.7)
© Directional method. Clause 3.3.3.7 gives detailed and
complex rules for parapets on monopitch, duopiteh and
ipped ronfs, While these give more accurate and detailed
estimates than the standard method, the additional benefit
is probably not worth the effort
mL corners, recessed
98 1 Wind loading: a practical guide to BS 6399-2
| Commentary | 99
© Standard method. Clause 2.5.7 recon
parapet for a conservative result for ruof pitches less
than 30°.
© Hybrid method. Guidance is not specifically given in BS
6399-2, but is implied by $3:4.1. We may simplity the
rules in §3.3.3.7 for use with the standard method
coefficients.
+ For all zones of the roof above the level of the
purapet, you should ignore the parapet.
+ All zones on upwind-facing pitches of the roof that
Tie below the parapct should he treated as zones on a
flat roof with a parapet of the same height,
imespective of the actual roof pitch
> On downwind-facing pitches, apply the reduction
factor from Table 37:10 the corresponding pressure
coefficients for all edge zones (A-D, H-I and Q-S)
that lie below the level of the parapet, You muy
interpolate in Table 31 forthe parapet height ratio Ib.
& The loads on the parapet are obtained by tre
as a free-standing wall.
© Curved or mansard eaves.
BS 6399-2 docs not give any guidance on the effect of
curved or mansard caves on pitched roofs because the
‘corresponding coefficient values have never been measured.
However, we can be sure that they will be beneficial,
reducing the high suctions immediately behind 1
Similarly, curved ridges are also expected to be beneficial.
Unfortunately, no data exist on which to base guidance and
the values of Ce given in BS 6399-2 for pitched roofs should
be used for a conservative result,
5 Barrel-vault roofs
“There is nothing in BS 6399-2 for this roof furm hecause reliable
data are very sparse. Nevertheless, this form is often built and
design values are needed. Part 2 of reference I recommends values
‘obtained from studies'* 5 in Brazil.
2
71 Determine the following dimensions defined in the key Fig.
54,
+ The length along the harrel-vault L.
+ The span of the barrel-vault $.
—— à — —s
Pig. 54. Key for bamekwaultruofs (width of A and B zones I 410)
<> "The height of the eaves of the wall H. For arched
buildings springing directly from the ground, À =.
© The rise of the barrel-vault r.
The reference height, should be taken at the highest point
Of the roof, Hr = H + r.
Divide the barrel-vault roof into the zones defined in Fig. $4,
‘The dimensions of these zones are all based upon the span of
the roof S. (The BS 6399-2 scaling length D is not needed)
Zones a to f in Fig. 54 correspond tu the orthogonal case
00" while zones A to F correspond to 9 = 90",
© Look up values of the external pressure coefficient from
Table 2, depending on the roof rise ratio »/$ and wall height
ratio H/S. You may interpolate within the ranges in the table,
01< 1/8 <03 and 0 < HIS <0:5, Outside these ranges use
the follow
Whe rise ratio 7/8 <0-1, trat the roof as being flat.
Whe rise ratio r/8>03 proceed as follows.
+ For the orthogonal case 9 =0" use the pressure
‘coefficients for circular walls of buildings in Table 7
in place of zones a to f.
+ For the orthogonal case 9 =90" use the pressure
coefficients for zones A 10 F in Table 2 für
1/S=03.
Note: Adina da for oe, less comun roof forms are given la
Part 2 of reference 1. af à
| Commentary | 99
© Standard method. Clause 2.5.7 recon
parapet for a conservative result for ruof pitches less
than 30°.
© Hybrid method. Guidance is not specifically given in BS
6399-2, but is implied by $3:4.1. We may simplity the
rules in §3.3.3.7 for use with the standard method
coefficients.
+ For all zones of the roof above the level of the
purapet, you should ignore the parapet.
+ All zones on upwind-facing pitches of the roof that
Tie below the parapct should he treated as zones on a
flat roof with a parapet of the same height,
imespective of the actual roof pitch
> On downwind-facing pitches, apply the reduction
factor from Table 37:10 the corresponding pressure
coefficients for all edge zones (A-D, H-I and Q-S)
that lie below the level of the parapet, You muy
interpolate in Table 31 forthe parapet height ratio Ib.
& The loads on the parapet are obtained by tre
as a free-standing wall.
© Curved or mansard eaves.
BS 6399-2 docs not give any guidance on the effect of
curved or mansard caves on pitched roofs because the
‘corresponding coefficient values have never been measured.
However, we can be sure that they will be beneficial,
reducing the high suctions immediately behind 1
Similarly, curved ridges are also expected to be beneficial.
Unfortunately, no data exist on which to base guidance and
the values of Ce given in BS 6399-2 for pitched roofs should
be used for a conservative result,
5 Barrel-vault roofs
“There is nothing in BS 6399-2 for this roof furm hecause reliable
data are very sparse. Nevertheless, this form is often built and
design values are needed. Part 2 of reference I recommends values
‘obtained from studies'* 5 in Brazil.
2
71 Determine the following dimensions defined in the key Fig.
54,
+ The length along the harrel-vault L.
+ The span of the barrel-vault $.
—— à — —s
Pig. 54. Key for bamekwaultruofs (width of A and B zones I 410)
<> "The height of the eaves of the wall H. For arched
buildings springing directly from the ground, À =.
© The rise of the barrel-vault r.
The reference height, should be taken at the highest point
Of the roof, Hr = H + r.
Divide the barrel-vault roof into the zones defined in Fig. $4,
‘The dimensions of these zones are all based upon the span of
the roof S. (The BS 6399-2 scaling length D is not needed)
Zones a to f in Fig. 54 correspond tu the orthogonal case
00" while zones A to F correspond to 9 = 90",
© Look up values of the external pressure coefficient from
Table 2, depending on the roof rise ratio »/$ and wall height
ratio H/S. You may interpolate within the ranges in the table,
01< 1/8 <03 and 0 < HIS <0:5, Outside these ranges use
the follow
Whe rise ratio 7/8 <0-1, trat the roof as being flat.
Whe rise ratio r/8>03 proceed as follows.
+ For the orthogonal case 9 =0" use the pressure
‘coefficients for circular walls of buildings in Table 7
in place of zones a to f.
+ For the orthogonal case 9 =90" use the pressure
coefficients for zones A 10 F in Table 2 für
1/S=03.
Note: Adina da for oe, less comun roof forms are given la
Part 2 of reference 1. af à
100 | Wind loading: a practical guide to BS 6399-2
Commentary À 101
Table 2, External pressure evefcents for Barcel vet anf
Zone
oe 6 AA
Fe ra,
na -02 0-4 -04 -02 02 -10 “LS 07 402 302 402
MiS 2028 08-045 07 07-055 04-14 -16 075 -02 402 =02
TJS 205-127-088 085-015 -06 045 -L:3 —15 098-025 -02 +02
Rise ral, 75 =02
SO DS 01 06-04 -02 02-11 -13 07 402 402 202
055 056 -12 17 08 -02 202 402
M/S =025 -02 -07 095 0
SOS 04 -08 095 08-035 O55 1-4 -20 —1:0-0:35 -02 402
Rise ratio, 115=09,
ad 406 402-076 05-02 02-11 -11 07402402 202
MUS =025 402-058 085 -07 US -04 13 —13 -07 -07 202 302
Ms mos 102 -03 095-0.75-055-095S -14 -20 -04 -04 02 402
‘Nowe The opograplical errors ia Table 2025 of reference have been correcte in Table 2.
2546 Multi-bay roofs
7.546.) Scope of guidance
This guidance applies unly to mulli-bay roofs of enclosed
buildings or open-sided buildings with permanent walls, Mulli-bay
canopy roofs are addressed in 2.8.6, later.
{BS 6399-2 states (82.5.5) that extemal pressure coefficients on
downwind bays of multi-bay roofs may conservatively be taken to
be the same as for a single-bay roof. This is true for pitch angles:
“a. < 30°. For higher pitch angles the upwind pitch ofthe frst upwind
bay will experience positive pressures, while the upwind pitches of
the other bays will experience negative pressures (suctions) that may
be higher in value. You are therefore advised to follow the
procedure in §2.5.5 or $3.3.4.2 as described below.
254.82. Standard and directional methods
“The procedure is essentially the same in both methods, except for
the way that the directional method defines the upwind bay A,
sccond bay B, ete. for skewed wind angles in Figure 42. However,
the zoning of the directional method is considerably more complex.
and gives little extra advantage. Accordingly, the following
guidance applics to the orthogonal load cases of the standard
method only.
254.63 Orthogonal case 0”
© Monopitch (sawtooth). See key Figure 23(a).
© Divide each mof pitch into zones, as if they were
individual monopitch roofs.
© Obtain the pressure coefficients for each zone fr
Table 9. mien
© Replace any positive value you find on the second bay
and any further downwind bays with the value
Cr -04,
© Apply the reduction factors in Table 12 to the
coefficients for each respective bay
© Duopitch, See key Figure 23(c) and (d). Provided the pitch
angles on either side of each ridge ave the same to within 5°,
© Treat the first pitch behind the upwind eaves as an
upwind pitch containing zones A, B and C
© Treat each pitch behind a ridge as a downwind pitch with
positive pitch angle (ridged) conta
© Treat each pitch behind a trough as a downwind pitch
‘with negative pitch angle (troughed) containing zones E,
Fond G.
O Uneguat-pitch duopitch. See key Figure 23(b). When the
Pic anges on ihr sido ofeach ridge ae nor the sum to
within 5°
© From the first ridge, divide the roof into troughed bays as
shown in Figure 23(b)
© Assess the pressure coefficients for exch troughed bay
using the pitch angle corresponding to the upwind and
downwind pitch. Both these pitch angles will bo
negativo, indicating a troughed roof
© Treat the first pitch (between the upwind eaves and the
first ridge) and the last pitch (between the last ridge and
the downwind eaves) as if they were isolated monopiteh
roofs, Both these Jes will he positive, @ — 0° for
the upwind pitch and @ =90° for the downwind pitch.
‘This procedure is safe, but is usually very conservative.
25464 Orthogool cose 90°
here mo ration in loading For this unse. Trea enh bay the
same as for sngl-bay rot ofthe same form and pitch engi,
Commentary À 101
Table 2, External pressure evefcents for Barcel vet anf
Zone
oe 6 AA
Fe ra,
na -02 0-4 -04 -02 02 -10 “LS 07 402 302 402
MiS 2028 08-045 07 07-055 04-14 -16 075 -02 402 =02
TJS 205-127-088 085-015 -06 045 -L:3 —15 098-025 -02 +02
Rise ral, 75 =02
SO DS 01 06-04 -02 02-11 -13 07 402 402 202
055 056 -12 17 08 -02 202 402
M/S =025 -02 -07 095 0
SOS 04 -08 095 08-035 O55 1-4 -20 —1:0-0:35 -02 402
Rise ratio, 115=09,
ad 406 402-076 05-02 02-11 -11 07402402 202
MUS =025 402-058 085 -07 US -04 13 —13 -07 -07 202 302
Ms mos 102 -03 095-0.75-055-095S -14 -20 -04 -04 02 402
‘Nowe The opograplical errors ia Table 2025 of reference have been correcte in Table 2.
2546 Multi-bay roofs
7.546.) Scope of guidance
This guidance applies unly to mulli-bay roofs of enclosed
buildings or open-sided buildings with permanent walls, Mulli-bay
canopy roofs are addressed in 2.8.6, later.
{BS 6399-2 states (82.5.5) that extemal pressure coefficients on
downwind bays of multi-bay roofs may conservatively be taken to
be the same as for a single-bay roof. This is true for pitch angles:
“a. < 30°. For higher pitch angles the upwind pitch ofthe frst upwind
bay will experience positive pressures, while the upwind pitches of
the other bays will experience negative pressures (suctions) that may
be higher in value. You are therefore advised to follow the
procedure in §2.5.5 or $3.3.4.2 as described below.
254.82. Standard and directional methods
“The procedure is essentially the same in both methods, except for
the way that the directional method defines the upwind bay A,
sccond bay B, ete. for skewed wind angles in Figure 42. However,
the zoning of the directional method is considerably more complex.
and gives little extra advantage. Accordingly, the following
guidance applics to the orthogonal load cases of the standard
method only.
254.63 Orthogonal case 0”
© Monopitch (sawtooth). See key Figure 23(a).
© Divide each mof pitch into zones, as if they were
individual monopitch roofs.
© Obtain the pressure coefficients for each zone fr
Table 9. mien
© Replace any positive value you find on the second bay
and any further downwind bays with the value
Cr -04,
© Apply the reduction factors in Table 12 to the
coefficients for each respective bay
© Duopitch, See key Figure 23(c) and (d). Provided the pitch
angles on either side of each ridge ave the same to within 5°,
© Treat the first pitch behind the upwind eaves as an
upwind pitch containing zones A, B and C
© Treat each pitch behind a ridge as a downwind pitch with
positive pitch angle (ridged) conta
© Treat each pitch behind a trough as a downwind pitch
‘with negative pitch angle (troughed) containing zones E,
Fond G.
O Uneguat-pitch duopitch. See key Figure 23(b). When the
Pic anges on ihr sido ofeach ridge ae nor the sum to
within 5°
© From the first ridge, divide the roof into troughed bays as
shown in Figure 23(b)
© Assess the pressure coefficients for exch troughed bay
using the pitch angle corresponding to the upwind and
downwind pitch. Both these pitch angles will bo
negativo, indicating a troughed roof
© Treat the first pitch (between the upwind eaves and the
first ridge) and the last pitch (between the last ridge and
the downwind eaves) as if they were isolated monopiteh
roofs, Both these Jes will he positive, @ — 0° for
the upwind pitch and @ =90° for the downwind pitch.
‘This procedure is safe, but is usually very conservative.
25464 Orthogool cose 90°
here mo ration in loading For this unse. Trea enh bay the
same as for sngl-bay rot ofthe same form and pitch engi,
102 | Wind loading: a practical guide to BS 6399-2
Commentary | 103
254.7 Comments and examples
“The principal change from previous practice is in the way that zones
are sized depending an the proportions of the front elevation, instead of
being fixed proportions of the plan. This will take a litle time to get used
10, but the reasons have been explained in 2.5.1 and the potential rewards,
lustrated in Fig. 33, are well worth having.
The voning of the standard method does not introduce much
conservatism, so we must ask ourselves whether the complexity of
the directional method is justified, This depends on your
application but, as al! structures can be safely designed using the standard
method, the answer must be ‘no’ for typical hand calculations.
“The derivation of scaling length and zoning for x fat-roofed building
by the directional and standard methods is shown in Example 9, This is
extended to duopitch and hipped roofs on the same building in Example
10, using the standard method only.
—— __ -—— o —
Example 9. Scaling length and pressure cocfficient zones on a
flat-roofed building
Dimensions of the example building
“The example building isa two-storey Matroofed office. with H—6m, in the shape of
aT. The main office dimensions are L=25m and WS. ‘There is an entrance/
Staiewell atthe centre of one long face that is Sm wide and protrudes by Tim. his is
‘at long enough to qualify as a separate wing using our criterion in Fig, 38, so we
shall teat i a a “small extension’
Directional method, wind angle 9 = 60"
For this wind angle: A =Gm, 8 = 5 x sin 60° +25 x cos 60° = 16-8m, giving
12m. The key dimensions tor the zones in Figure 35 ae: b/10~ 1-2m, b/4
and 5/2 = 6m
Directional method, wind unge @ = 90"
For the main building at this wind angle: H = Gm, B = 5m, giving b = Sm. Using
the principle of reflection defined in Fig. 41, tor he extension: 11 — 6m, & =6m,
giving b = 6m, AU is wind angle, dhe sealing length forthe small extension i larger
than forthe main building but this is not unusual for Long buildings.
Standard method, orthogonat case 8 =!
“Moving ta the orthogonal case of the stand method hat comesponds tothe same
wind angle. we have the same sealing ll for the main building and the extension,
but the distribution of zones ik simple.
Standard method, orthogonal case 6
Por the unthogenal case ofthe standard method vith the extension upuind: H = 6m,
5m, giving b = 12m for the moin building, We eunsider the entrancestarwell
extension 100 small to be a wing, so that the wind flows over, rather than round, und
‘b= 12m also. Note that this means there is no space for zone B along the eaves ofthe
extension, If we consider the extension Jung enough to be a wing,
zones on the extension would be smaller. We huve chosen the safer opio.
Commentary | 103
254.7 Comments and examples
“The principal change from previous practice is in the way that zones
are sized depending an the proportions of the front elevation, instead of
being fixed proportions of the plan. This will take a litle time to get used
10, but the reasons have been explained in 2.5.1 and the potential rewards,
lustrated in Fig. 33, are well worth having.
The voning of the standard method does not introduce much
conservatism, so we must ask ourselves whether the complexity of
the directional method is justified, This depends on your
application but, as al! structures can be safely designed using the standard
method, the answer must be ‘no’ for typical hand calculations.
“The derivation of scaling length and zoning for x fat-roofed building
by the directional and standard methods is shown in Example 9, This is
extended to duopitch and hipped roofs on the same building in Example
10, using the standard method only.
—— __ -—— o —
Example 9. Scaling length and pressure cocfficient zones on a
flat-roofed building
Dimensions of the example building
“The example building isa two-storey Matroofed office. with H—6m, in the shape of
aT. The main office dimensions are L=25m and WS. ‘There is an entrance/
Staiewell atthe centre of one long face that is Sm wide and protrudes by Tim. his is
‘at long enough to qualify as a separate wing using our criterion in Fig, 38, so we
shall teat i a a “small extension’
Directional method, wind angle 9 = 60"
For this wind angle: A =Gm, 8 = 5 x sin 60° +25 x cos 60° = 16-8m, giving
12m. The key dimensions tor the zones in Figure 35 ae: b/10~ 1-2m, b/4
and 5/2 = 6m
Directional method, wind unge @ = 90"
For the main building at this wind angle: H = Gm, B = 5m, giving b = Sm. Using
the principle of reflection defined in Fig. 41, tor he extension: 11 — 6m, & =6m,
giving b = 6m, AU is wind angle, dhe sealing length forthe small extension i larger
than forthe main building but this is not unusual for Long buildings.
Standard method, orthogonat case 8 =!
“Moving ta the orthogonal case of the stand method hat comesponds tothe same
wind angle. we have the same sealing ll for the main building and the extension,
but the distribution of zones ik simple.
Standard method, orthogonal case 6
Por the unthogenal case ofthe standard method vith the extension upuind: H = 6m,
5m, giving b = 12m for the moin building, We eunsider the entrancestarwell
extension 100 small to be a wing, so that the wind flows over, rather than round, und
‘b= 12m also. Note that this means there is no space for zone B along the eaves ofthe
extension, If we consider the extension Jung enough to be a wing,
zones on the extension would be smaller. We huve chosen the safer opio.
104 | Wind loading: a practical guide to BS 6399-2
Commentary | 105
Standard method, orthogonal case 0 = 180" and 270°
“ortogonal case 9 = 27" ie symmetrical (09 = 907, o select the larger design wind
speed to cover these cases, Orthogonal case 0 — 180° has the extension downwind, 0
the B zone continues across the upwind eaves and the downwind 2m of the extension
roof is zone D.
Comments i
“The complexity ofthe zoning in the directional method makes it unsuitable for Rand
‘calculations and the detail it provides is rarely necessary. The standard method
fvalgersaes rones A and B and zones 13 and Fino single zones, adopting the larger
vale of Cy. This leeds 10 a very small degree of conservatism.
THis generally sate to ignore zone D and replace it wilh highe-teaded zone C,
simplifies the standard method to comer zone A, eaves zone D and main zone C.
—_—————————————
Example 10, Scaling length and pressure coefficient zones on a
pitched roof building
Dimensions of the example building
‘The example building is the same T shaped two-storey ofice used in Example 9, bot
wit 30" cuopiteh or hipped roof added. The pitch of the oof is rt relevant 0 the
position and size of the zones when viewed in plan, but increases the height 10
117-25 mt the ridge, We take the wind angle @ from normal o the ridge and eaves
‘of the main office. We mus! note thatthe extension is aligned at 90° tothe main ice.
‘When the zones on the oof ofthe main building are given by Figure 20tb). he zones
‘onthe roof of the extension will be given by Figure 20(c) and vice versa. To avoid
confusion, the zones for 0=0> in Figure 2010) are given the subscript "0° and the
ones für 9 =90° in Figure 20fe) are given the subscript "90" inthe diagrams below.
Standard method, duopitch and hipped, orthogonal case 0 — 7
For tie main building at this wind angle: 17 =7-25m, B=WW=5m, giving
dy = 5m. Using the principle of reflection defined in Fig. 41, for the extension:
N=125m, BaL=6m, giving hy Grn. AL this wind angle, the scaling length for
{he small extension is larger than forthe rain office, but this is not unusual far lon
ings. "Ihe main office is shown with gables and with hips
oem
Standard method, duoplich, orthogonal case
For the orthogonal case with the extension upwind: A 5m,
ivine b—14Sim for the main building. We consider the entrance/stairwell
extension too small o be u wing, so tha the wind Rows over, rather than around,
and so we take dy = 145 m also. If we consider the extension long enough to be 4
‘wing, then by = Sm and the zones on he extension would be smaller. We have chosen
Ue safer option,
E
$ am patas
The duopitch roofed building in Example 10 has been developed to
‘cover one of the three possible cases of multi-bay roofs, the most common
‘equal-pitched case. This is illusvated in Example 11.
Example 11. Scaling length and pressure coefficient zones on a
multi-bay building
Dimensions of the example building
Far this example, additional bays have been added to the T shaped two-storey office
used in Example 10, ult with the same 30° duopiteh roots. Only the wind angle
is Hlustated for the case of equal-pich duopiteh roots (Figure 24).
‘The zones A, Band C on the upwindpitch ofthe fist bay are given the subscripts
"0° and "90° as explained in Example 10. The zones on all remaining pitches wre the
zones E, Fund G for downwind pitches, hat are alternately for o = +30" ‘ridged? and
Commentary | 105
Standard method, orthogonal case 0 = 180" and 270°
“ortogonal case 9 = 27" ie symmetrical (09 = 907, o select the larger design wind
speed to cover these cases, Orthogonal case 0 — 180° has the extension downwind, 0
the B zone continues across the upwind eaves and the downwind 2m of the extension
roof is zone D.
Comments i
“The complexity ofthe zoning in the directional method makes it unsuitable for Rand
‘calculations and the detail it provides is rarely necessary. The standard method
fvalgersaes rones A and B and zones 13 and Fino single zones, adopting the larger
vale of Cy. This leeds 10 a very small degree of conservatism.
THis generally sate to ignore zone D and replace it wilh highe-teaded zone C,
simplifies the standard method to comer zone A, eaves zone D and main zone C.
—_—————————————
Example 10, Scaling length and pressure coefficient zones on a
pitched roof building
Dimensions of the example building
‘The example building is the same T shaped two-storey ofice used in Example 9, bot
wit 30" cuopiteh or hipped roof added. The pitch of the oof is rt relevant 0 the
position and size of the zones when viewed in plan, but increases the height 10
117-25 mt the ridge, We take the wind angle @ from normal o the ridge and eaves
‘of the main office. We mus! note thatthe extension is aligned at 90° tothe main ice.
‘When the zones on the oof ofthe main building are given by Figure 20tb). he zones
‘onthe roof of the extension will be given by Figure 20(c) and vice versa. To avoid
confusion, the zones for 0=0> in Figure 2010) are given the subscript "0° and the
ones für 9 =90° in Figure 20fe) are given the subscript "90" inthe diagrams below.
Standard method, duopitch and hipped, orthogonal case 0 — 7
For tie main building at this wind angle: 17 =7-25m, B=WW=5m, giving
dy = 5m. Using the principle of reflection defined in Fig. 41, for the extension:
N=125m, BaL=6m, giving hy Grn. AL this wind angle, the scaling length for
{he small extension is larger than forthe rain office, but this is not unusual far lon
ings. "Ihe main office is shown with gables and with hips
oem
Standard method, duoplich, orthogonal case
For the orthogonal case with the extension upwind: A 5m,
ivine b—14Sim for the main building. We consider the entrance/stairwell
extension too small o be u wing, so tha the wind Rows over, rather than around,
and so we take dy = 145 m also. If we consider the extension long enough to be 4
‘wing, then by = Sm and the zones on he extension would be smaller. We have chosen
Ue safer option,
E
$ am patas
The duopitch roofed building in Example 10 has been developed to
‘cover one of the three possible cases of multi-bay roofs, the most common
‘equal-pitched case. This is illusvated in Example 11.
Example 11. Scaling length and pressure coefficient zones on a
multi-bay building
Dimensions of the example building
Far this example, additional bays have been added to the T shaped two-storey office
used in Example 10, ult with the same 30° duopiteh roots. Only the wind angle
is Hlustated for the case of equal-pich duopiteh roots (Figure 24).
‘The zones A, Band C on the upwindpitch ofthe fist bay are given the subscripts
"0° and "90° as explained in Example 10. The zones on all remaining pitches wre the
zones E, Fund G for downwind pitches, hat are alternately for o = +30" ‘ridged? and
106 | Wind loading: a practical guide 10 BS 6399-2
Commentary | 107
UP oughed® pitch angles. “Ridged' zones are given the subscript 1 and
coughed? zones the subscript "1", followed by the number ofthe bay from the fist
‘upwind bay. Por example, Gra isthe G zone of a troughed ruot on the second bay.
‘Standard method, duoplteh, orihugonal case 6 = 0”
From Example 10, b, =145a1 for the main building and by = 145m for the
‘entrance extension. I follows hat dy = 14:5 forthe additional downwind hays. The
ones on the extension and the fist bay are exactly the same as for Example 10 but
for example, zone Fy becomes Fay in the new notation,
‘We take the values of pressure coeficent for each zone from Tables 10 and 12 xs
follows.
® Al remeining bays: En 054; Mrs = Gr
22042 Bee 07:06 04x06
sm
am
Internal pressure coefficients (§2.6, §3.3.5)
Basis
5.1.1 Balenceof flow
Fe may rane Her inside a enclosed building a hing pr tected
from the wind by the building envelope. The internal pressure is the
pressure inside the building and acts on the internal surfaces. The value is
set by the balance flow of air in and out of the building through openings,
in the envelope, driven by the external pressures. The openings will be of
many sizes, from completo faces (e.g. grandstands), through doors,
windows, ventilation grilles, down to the small pores through the
‘envelope of ‘sealed” air-conditioned offices. A need to supply fresh air to
‘the occupants means that there is no such thing as a ‘Tully seated’
building. If there were, changes in atmospheric pressure would generate
pressure differences ten times greater than possible by the wind.
When storeys and ruums divide a building into compartments, the
internal pressure will vary between compartments, generating loads on
(ernal partitions. You may als need to consider the internal pressure.
within components, such as between the layers of à built-up insulated
raof, to determine the fixing loads between layers
‘The net flow of air Q through any apertur
Qo Ap" an
where A is the area of the aperture and p is the pressure difference across
it. The power » depends on the form of the opening. Kor a simple orifice
n=05, but for à libyrinthine opening #1. Measurements of real
buildings show m in the range 0:5 < n < 04, In practice itis sufficient 10
take n 05, ie. a square rout.
‘The total flow in and out of a compartment must balance:
iow cow
Yen Dar DARRO an
Note the different order of the external and internal pressures in the
inflow and outflow terms of Equation 18. This is necessary to ensure that
Equation 18 can be solved:
Pron = Pi _ (Ane)?
Ps Pre (ax) a)
where the subse
Because the pr
th
1 “front” and ‘rear’ refer to the front and rear faces,
sure drop across either wall is proportional to square of
arca ratio, increasing one of the areas quickly makes it dom
PAP AD Pr
LI
Fig. 55, Internal prete with two openings
Commentary | 107
UP oughed® pitch angles. “Ridged' zones are given the subscript 1 and
coughed? zones the subscript "1", followed by the number ofthe bay from the fist
‘upwind bay. Por example, Gra isthe G zone of a troughed ruot on the second bay.
‘Standard method, duoplteh, orihugonal case 6 = 0”
From Example 10, b, =145a1 for the main building and by = 145m for the
‘entrance extension. I follows hat dy = 14:5 forthe additional downwind hays. The
ones on the extension and the fist bay are exactly the same as for Example 10 but
for example, zone Fy becomes Fay in the new notation,
‘We take the values of pressure coeficent for each zone from Tables 10 and 12 xs
follows.
® Al remeining bays: En 054; Mrs = Gr
22042 Bee 07:06 04x06
sm
am
Internal pressure coefficients (§2.6, §3.3.5)
Basis
5.1.1 Balenceof flow
Fe may rane Her inside a enclosed building a hing pr tected
from the wind by the building envelope. The internal pressure is the
pressure inside the building and acts on the internal surfaces. The value is
set by the balance flow of air in and out of the building through openings,
in the envelope, driven by the external pressures. The openings will be of
many sizes, from completo faces (e.g. grandstands), through doors,
windows, ventilation grilles, down to the small pores through the
‘envelope of ‘sealed” air-conditioned offices. A need to supply fresh air to
‘the occupants means that there is no such thing as a ‘Tully seated’
building. If there were, changes in atmospheric pressure would generate
pressure differences ten times greater than possible by the wind.
When storeys and ruums divide a building into compartments, the
internal pressure will vary between compartments, generating loads on
(ernal partitions. You may als need to consider the internal pressure.
within components, such as between the layers of à built-up insulated
raof, to determine the fixing loads between layers
‘The net flow of air Q through any apertur
Qo Ap" an
where A is the area of the aperture and p is the pressure difference across
it. The power » depends on the form of the opening. Kor a simple orifice
n=05, but for à libyrinthine opening #1. Measurements of real
buildings show m in the range 0:5 < n < 04, In practice itis sufficient 10
take n 05, ie. a square rout.
‘The total flow in and out of a compartment must balance:
iow cow
Yen Dar DARRO an
Note the different order of the external and internal pressures in the
inflow and outflow terms of Equation 18. This is necessary to ensure that
Equation 18 can be solved:
Pron = Pi _ (Ane)?
Ps Pre (ax) a)
where the subse
Because the pr
th
1 “front” and ‘rear’ refer to the front and rear faces,
sure drop across either wall is proportional to square of
arca ratio, increasing one of the areas quickly makes it dom
PAP AD Pr
LI
Fig. 55, Internal prete with two openings
108 | Wind loading: a pructicul guide 10 BS 6399-2
Commentary | 169
Unfortunately Equation 18 cannot be solved directly for more than two
openings, but can be solved by repeated iteration from assumed initial
values of internal pressure until the flow balance is achieved. We can also
work in terms of the pressure difference coefficient Cp= Cpe-Cp instead
‘of the net pressure when the reference dynamic pressure isthe same for ll
‘openings. This is illustrated for a single compartment by Example 12 and
for a two-room building by Example 13.
A A - _AAXXA<<<€€úéááf<>,
Example 12. Internal pressure with multiple openings
Note: We may use pressure cooficents Cy. and Cy, instead of pressures when the
same value of dynamic pressure applies to each wall (ie. when euch wall has
the same effective height
03
|
AZAR <5 — 04
Ca
08—2> 08
1
1
+08
For this example, we shall assume six upenings of equal area in the building walls
as shown above, As the areas are equal, we may take A- and flow through each
‘opening to be Qe Gq — Tu for intlow and Qe y/C = Ce for cul, trom
Equation 18, We reed lo irate the value of Cy, untl the net inflow is zero.
1. Start with an initial guess For Cj, = — 0:32, the mean value ofthe six values of
Gre
2. Calculate Q for each opening and sum this to give IQ ~ -0-782.
This negative value indicates a net outflow, so ie ternal pressure must he
reduced to reach the flow balan.
‘Make frst iteration by decrensing internal pressure to Cy= 04.
“This now gives a net nilo, bal the value is smaller than the previous oullow.
‘Make second iteraion by increasing internal pressure to Cu => 039.
This gives a net outflow. Try a smaller increment
The third eration to Ci= 0396 balances the flow to two significant
CR
U a a Fe
il guess 0:32 1057 —049S —0695 -0289 - 0289 0-129 -0782 Ouiflow
Fis iteration — 040 1095 0632 0432 0000 0000 0316 0.147 Inflow
Second iteration 0391091 0640 0640 —0 100 -0:100 0300 0030 Outfow
Td iteraion 0306 LUN ~0.636 —D636 MAC 0063 0310 0006 Inflow
Now if we assume A=2 for opening 1 in the windward face and for opening 6 in
the leeward face, Le. twice the area A= 1 of the vier openings, we make each face
‘equally permeahle Starting to iterate from the previous soluion we get
L 0% % % % &% Te
Initial guess -040 2.191 0632 0652 0000 DOM 0632 1558 Inflow
Fit ir =030 2098 0207-0707 -0316-0316 0.000. 0051 Inflow
Second iteration -029 2065 _ 0714 -0714 —0332 0332-0200 -0204 Oulow
“Th iertion - 0299 2097 008 ~0708 —0318 DIR —0.063 -0018 Outfow
Note: This isthe source ofthe internal pressure coefficient value Cp -0-3 or ‘four
walls equally permeuble, rouf impermeable” in Table (6,
Example 13. Internal pressure for bt
03
ding with two rooms
Et
+08
To this example the building of Example 12 is assumed to have apacition installed,
dividing it Into two rooms A and B, with internal pressure eelficints Cua and Cy.
is example differs from Example 12 in that we need to iterate both Ca nd Crus
ti the net flow into both rooms is zero. Fit we shall assume that the aren of the
‘opening in the paritlon I Ihe same as the other openings.
1. Stat with initial values oF Cp = Cn = 0-4.
2. Calcolate the flow dough each opening and sum the net inflow into each
Commentary | 169
Unfortunately Equation 18 cannot be solved directly for more than two
openings, but can be solved by repeated iteration from assumed initial
values of internal pressure until the flow balance is achieved. We can also
work in terms of the pressure difference coefficient Cp= Cpe-Cp instead
‘of the net pressure when the reference dynamic pressure isthe same for ll
‘openings. This is illustrated for a single compartment by Example 12 and
for a two-room building by Example 13.
A A - _AAXXA<<<€€úéááf<>,
Example 12. Internal pressure with multiple openings
Note: We may use pressure cooficents Cy. and Cy, instead of pressures when the
same value of dynamic pressure applies to each wall (ie. when euch wall has
the same effective height
03
|
AZAR <5 — 04
Ca
08—2> 08
1
1
+08
For this example, we shall assume six upenings of equal area in the building walls
as shown above, As the areas are equal, we may take A- and flow through each
‘opening to be Qe Gq — Tu for intlow and Qe y/C = Ce for cul, trom
Equation 18, We reed lo irate the value of Cy, untl the net inflow is zero.
1. Start with an initial guess For Cj, = — 0:32, the mean value ofthe six values of
Gre
2. Calculate Q for each opening and sum this to give IQ ~ -0-782.
This negative value indicates a net outflow, so ie ternal pressure must he
reduced to reach the flow balan.
‘Make frst iteration by decrensing internal pressure to Cy= 04.
“This now gives a net nilo, bal the value is smaller than the previous oullow.
‘Make second iteraion by increasing internal pressure to Cu => 039.
This gives a net outflow. Try a smaller increment
The third eration to Ci= 0396 balances the flow to two significant
CR
U a a Fe
il guess 0:32 1057 —049S —0695 -0289 - 0289 0-129 -0782 Ouiflow
Fis iteration — 040 1095 0632 0432 0000 0000 0316 0.147 Inflow
Second iteration 0391091 0640 0640 —0 100 -0:100 0300 0030 Outfow
Td iteraion 0306 LUN ~0.636 —D636 MAC 0063 0310 0006 Inflow
Now if we assume A=2 for opening 1 in the windward face and for opening 6 in
the leeward face, Le. twice the area A= 1 of the vier openings, we make each face
‘equally permeahle Starting to iterate from the previous soluion we get
L 0% % % % &% Te
Initial guess -040 2.191 0632 0652 0000 DOM 0632 1558 Inflow
Fit ir =030 2098 0207-0707 -0316-0316 0.000. 0051 Inflow
Second iteration -029 2065 _ 0714 -0714 —0332 0332-0200 -0204 Oulow
“Th iertion - 0299 2097 008 ~0708 —0318 DIR —0.063 -0018 Outfow
Note: This isthe source ofthe internal pressure coefficient value Cp -0-3 or ‘four
walls equally permeuble, rouf impermeable” in Table (6,
Example 13. Internal pressure for bt
03
ding with two rooms
Et
+08
To this example the building of Example 12 is assumed to have apacition installed,
dividing it Into two rooms A and B, with internal pressure eelficints Cua and Cy.
is example differs from Example 12 in that we need to iterate both Ca nd Crus
ti the net flow into both rooms is zero. Fit we shall assume that the aren of the
‘opening in the paritlon I Ihe same as the other openings.
1. Stat with initial values oF Cp = Cn = 0-4.
2. Calcolate the flow dough each opening and sum the net inflow into each
110 À Wind landing: a practical guide to BS 6399-2
Commentary | 111
3. Note when summing the et inflow thatthe flow Q; is defined as ongfow from
room À and flow to room #
44, rate the intemal pressure hat has the largest nel inflow or outlaw.
Ga Gem HD EL EM
tidal pes 080 040 LIM 9692 -0652 00m DOS O00 O16 0316
Vale 204 OO 1099-0612 OR =D -D1 ON 10D 005) 000
Son At 2039 110 1624 9628 -010 DM 000 C141 -D008 OH
Mr ec oat 0398 1100-0628 0408 ‚at OCA OMS 0.130 —0019 0007
Tooth caldo ZO SIS. 0393. 101 0872 ~D622 008: 0084 08 0:41 -0001 0008
it OS. O 33% 140-0622 -0622 „0002 ~D0R2 0008 ~O-40 DU 0001
Su won ZO AI LD AS 1102 „062? „0622 DE? —D08 0208 Win DINO 000
We see iha both interval presenres are negative, and they have similar values
‘But if we close the openings in th side wall, we get a different balance as shown
below.
“RU 0% Du
Ga Ga o
interral pressure and the difference
Now the windward room has a strongly pos
across the intemal partition is Cy= 0:365. seo eet Ñ
However. internal patios ure expected to heal Least thee times more porows Lan
the external Façade. Inerensing the area of opening 7 10 thre times the other arcas
gives:
ba Ga AL A AAA A A EW
Sixth men 0279 OI 0772 000 DUO 0000 Dam 0722 0722 000 mul
Now the intemal pressure is similar between the (wo rooms and the difference
cross the Internal partition Is only Cy ~ 0-058.
Note: This is the source of the internal pressure coefictent value Cy = +02 for
“mv opposite walls equally permeable, other faces impermeable, wind normal
10 permeable face" in Table 16.
_ > — _ _2z—--—JHHIÁKA
2.5.5.1.2 Response time and equivalent size effect |
The balance of flows hus previously been assumed to occur instantly,
but we know that it will take some time to establish, The response time
depends on a number of factors, principally the size of the openings and
ly shorter than the duration of the
equivalent static gust that Joads the building. This was discussed at some
length in the 1972 Wind Loading Handbook.”
Following the ‘10% influence” principle in 1.4, BS 6399-2 sets
standard values to all the parameters that control response time, leaving
only the volume of the building O as a variable parameter, converting this
to an equivalent diagonal dimension a. This allows the size effect method
for the extemal pressures to be used for the response time of internal
pressures.
There is litle advantage to be gained in departing from the simplified
approach in BS 6399-2 unless the building is very large and/or the
cladding envelope is very flexible, in which case some conserva
be eliminated by a more sophisticated approach. More detailed guidance
is given Inter in 34.3.
25.5.2 Enclosed buildings (826.1)
25.5.2.1 Porosity of enclosed buildings
In the summation of Equation 18, a number of openings in any
pressure coefficient zone can be treated as a single opening of the total
area. This allows us o treat walls having a uniform porosity $, where:
atea of openings
area of wall
as having an equivalent single opening of area PA. In order to use
Equation 20 we need to know the porosity of typical glazing, cladding
and walls. Unfortunately, there is very litle information on this subject,
but the available information that is reliable is given in Table 3.
Fortunately, for most enclosed buildings we only need to know the
relative porosity of the building faces within a factor of about three in
order to apply Equation 18. When the porosity is the same on each face,
we need only the area of each wall,
For typical buildings, the solution of Equation 18 gives only a small
range of internal pressure coefficients Ga which are given in Table 16 of
BS 6399-2. Experience with multi-storey buildings shows that:
(20)
Table 3. Typical porosity of construction
Fax of constuction Porosity
‘Open areaoral area —3.5% 107%
Open areata area = 7310
“Typical housing in UK Open arcalotal area = 105 107%
ergy-efficient housing Open areauta ue = 4310
Single teat door Calcule sing gap width — 1- mm when closed
Commentary | 111
3. Note when summing the et inflow thatthe flow Q; is defined as ongfow from
room À and flow to room #
44, rate the intemal pressure hat has the largest nel inflow or outlaw.
Ga Gem HD EL EM
tidal pes 080 040 LIM 9692 -0652 00m DOS O00 O16 0316
Vale 204 OO 1099-0612 OR =D -D1 ON 10D 005) 000
Son At 2039 110 1624 9628 -010 DM 000 C141 -D008 OH
Mr ec oat 0398 1100-0628 0408 ‚at OCA OMS 0.130 —0019 0007
Tooth caldo ZO SIS. 0393. 101 0872 ~D622 008: 0084 08 0:41 -0001 0008
it OS. O 33% 140-0622 -0622 „0002 ~D0R2 0008 ~O-40 DU 0001
Su won ZO AI LD AS 1102 „062? „0622 DE? —D08 0208 Win DINO 000
We see iha both interval presenres are negative, and they have similar values
‘But if we close the openings in th side wall, we get a different balance as shown
below.
“RU 0% Du
Ga Ga o
interral pressure and the difference
Now the windward room has a strongly pos
across the intemal partition is Cy= 0:365. seo eet Ñ
However. internal patios ure expected to heal Least thee times more porows Lan
the external Façade. Inerensing the area of opening 7 10 thre times the other arcas
gives:
ba Ga AL A AAA A A EW
Sixth men 0279 OI 0772 000 DUO 0000 Dam 0722 0722 000 mul
Now the intemal pressure is similar between the (wo rooms and the difference
cross the Internal partition Is only Cy ~ 0-058.
Note: This is the source of the internal pressure coefictent value Cy = +02 for
“mv opposite walls equally permeable, other faces impermeable, wind normal
10 permeable face" in Table 16.
_ > — _ _2z—--—JHHIÁKA
2.5.5.1.2 Response time and equivalent size effect |
The balance of flows hus previously been assumed to occur instantly,
but we know that it will take some time to establish, The response time
depends on a number of factors, principally the size of the openings and
ly shorter than the duration of the
equivalent static gust that Joads the building. This was discussed at some
length in the 1972 Wind Loading Handbook.”
Following the ‘10% influence” principle in 1.4, BS 6399-2 sets
standard values to all the parameters that control response time, leaving
only the volume of the building O as a variable parameter, converting this
to an equivalent diagonal dimension a. This allows the size effect method
for the extemal pressures to be used for the response time of internal
pressures.
There is litle advantage to be gained in departing from the simplified
approach in BS 6399-2 unless the building is very large and/or the
cladding envelope is very flexible, in which case some conserva
be eliminated by a more sophisticated approach. More detailed guidance
is given Inter in 34.3.
25.5.2 Enclosed buildings (826.1)
25.5.2.1 Porosity of enclosed buildings
In the summation of Equation 18, a number of openings in any
pressure coefficient zone can be treated as a single opening of the total
area. This allows us o treat walls having a uniform porosity $, where:
atea of openings
area of wall
as having an equivalent single opening of area PA. In order to use
Equation 20 we need to know the porosity of typical glazing, cladding
and walls. Unfortunately, there is very litle information on this subject,
but the available information that is reliable is given in Table 3.
Fortunately, for most enclosed buildings we only need to know the
relative porosity of the building faces within a factor of about three in
order to apply Equation 18. When the porosity is the same on each face,
we need only the area of each wall,
For typical buildings, the solution of Equation 18 gives only a small
range of internal pressure coefficients Ga which are given in Table 16 of
BS 6399-2. Experience with multi-storey buildings shows that:
(20)
Table 3. Typical porosity of construction
Fax of constuction Porosity
‘Open areaoral area —3.5% 107%
Open areata area = 7310
“Typical housing in UK Open arcalotal area = 105 107%
ergy-efficient housing Open areauta ue = 4310
Single teat door Calcule sing gap width — 1- mm when closed
112 | Wind loading: a practical guide to BS 6399-2
Commentary | 113
+ the internal pressure for each storey can be assumed to be
independent of the other storeys, and
© the porosity of internal walls and partitions is generally three times
‘greater than that of the external walls,
2.5.5.2.2 Application
O For typical enclosed buildings, use §2.6.1.1.
© Take the internal pressure coefficients from Table 16. These
values are the same as previously given by CP 3-V-2.
© Calculate the effective diagonal dimension a from the internal
volume of each storey using Equation 13.
© The net pressure coefficient across internal walls and
partitions is nominally zero, but should be taken as C,=05
for the serviceability limit to account for apen windows.
17 For buildings where the external walls are very permeable, eg
diherately ventilated, or when intemal walls are impermeable, c.
party walls between houses or industial units, $2.6 1.2 applies.
© For windward walls, where Coe is positive, take Cy) = 03.
O For side and leeward walls, where Ce is negative, take
Cu= +02.
O For roofs where the loft space is ventilated around all eaves
lake Cy=-03, otherwise take the more onerous of
Cy= +02 oF -03.
Note: When $2:6.12 applies there will be unticeable resistance to
opening internal doors in windy weather.
2.5.5.3 Dominant openings (§2.6.2)
2.5.5.3.1 Definitions
A single opening is taken as dominant when its area isa least twice
that of all other openings added together, including the arcas
calculated from distributed porosity using Equativn 20, even when
this opening is relatively small, We shall see Inter in 3.4.3.3 thatthe
equivalent diagonal dimension given by Equation 15 is valid for an
opening that is only 1% uf the face area of a cubic building.
® More than one opening in a single face should be treated as a
single dominant opening of the total arca, provided this arca is at
east twice that of all other openings, as defined above.
1 When there are openings in more than one face, but none is large
enough to be dominant, use the following guidelines.
© Ithe storey is not divided into rooms, then $2.6./ applies instead.
© Whe storey is divided into rooms, each opening is a dominant
opening into the respective ruom.
2.5532. Application
(Take the intemal pressure coefficient Cy from Table 17. I unsure
about the ratio of the opening area to the other openings, take the
most onerous value, Ci =09 Cpe.
9 When the storey is not divided into rooms, calculate the effective
diagonal dimension a using Equation 15, from the internal volume
of the storey
When the storey is sub-divided into rooms apply the following.
© For louds on external walls, we must assume that internal
doors can be open. Use the volume of the storey in Equati
17 and apply the resulting Gy lo the whole storey.
© For loads on the internal walls or partitions of the room
containing the opening, we must assume that the dows tu the
room can be closed. Use the volume of the room in Equation
17 and apply the resulting Cy; to the room.
O If there is more than one opening in the same face, but the
‘openings lie in different zones, take the value of Coe averaged over
the area of the zones.
‘Elective’ dominant openings, ie. openings that are normally closed,
should be treated as a serviccahility condition with a suitable value
of probability factor $ Further advice is given in 3346, later.
2.5.5.4 Open-sided and open-topped buildings
2.5.54.1 Definitions
Open-sided buildings are buildings with at least one wall completely
open, ¢.g. grandstands, or where at least 80% of the wall is open.
ings with roof but with no permanent walls are treated as Fre
standing canopies in $2.5,9./. (See 2.5.6, later.)
2.5.5.4.2 Application
O Internal pressure coefficients for open-sided buildings are given in
Table 18. Figure 56 gives a key to the column headings of Table
18 and indivates how the positive and negative values apply to the
“two adjacent open faces’ case at 0 = 9°
O Open-topped buildings, such as a building without à roof during
renovation, should he treated as free-standing walls using §2.8
(See 25,7, below.)
Commentary | 113
+ the internal pressure for each storey can be assumed to be
independent of the other storeys, and
© the porosity of internal walls and partitions is generally three times
‘greater than that of the external walls,
2.5.5.2.2 Application
O For typical enclosed buildings, use §2.6.1.1.
© Take the internal pressure coefficients from Table 16. These
values are the same as previously given by CP 3-V-2.
© Calculate the effective diagonal dimension a from the internal
volume of each storey using Equation 13.
© The net pressure coefficient across internal walls and
partitions is nominally zero, but should be taken as C,=05
for the serviceability limit to account for apen windows.
17 For buildings where the external walls are very permeable, eg
diherately ventilated, or when intemal walls are impermeable, c.
party walls between houses or industial units, $2.6 1.2 applies.
© For windward walls, where Coe is positive, take Cy) = 03.
O For side and leeward walls, where Ce is negative, take
Cu= +02.
O For roofs where the loft space is ventilated around all eaves
lake Cy=-03, otherwise take the more onerous of
Cy= +02 oF -03.
Note: When $2:6.12 applies there will be unticeable resistance to
opening internal doors in windy weather.
2.5.5.3 Dominant openings (§2.6.2)
2.5.5.3.1 Definitions
A single opening is taken as dominant when its area isa least twice
that of all other openings added together, including the arcas
calculated from distributed porosity using Equativn 20, even when
this opening is relatively small, We shall see Inter in 3.4.3.3 thatthe
equivalent diagonal dimension given by Equation 15 is valid for an
opening that is only 1% uf the face area of a cubic building.
® More than one opening in a single face should be treated as a
single dominant opening of the total arca, provided this arca is at
east twice that of all other openings, as defined above.
1 When there are openings in more than one face, but none is large
enough to be dominant, use the following guidelines.
© Ithe storey is not divided into rooms, then $2.6./ applies instead.
© Whe storey is divided into rooms, each opening is a dominant
opening into the respective ruom.
2.5532. Application
(Take the intemal pressure coefficient Cy from Table 17. I unsure
about the ratio of the opening area to the other openings, take the
most onerous value, Ci =09 Cpe.
9 When the storey is not divided into rooms, calculate the effective
diagonal dimension a using Equation 15, from the internal volume
of the storey
When the storey is sub-divided into rooms apply the following.
© For louds on external walls, we must assume that internal
doors can be open. Use the volume of the storey in Equati
17 and apply the resulting Gy lo the whole storey.
© For loads on the internal walls or partitions of the room
containing the opening, we must assume that the dows tu the
room can be closed. Use the volume of the room in Equation
17 and apply the resulting Cy; to the room.
O If there is more than one opening in the same face, but the
‘openings lie in different zones, take the value of Coe averaged over
the area of the zones.
‘Elective’ dominant openings, ie. openings that are normally closed,
should be treated as a serviccahility condition with a suitable value
of probability factor $ Further advice is given in 3346, later.
2.5.5.4 Open-sided and open-topped buildings
2.5.54.1 Definitions
Open-sided buildings are buildings with at least one wall completely
open, ¢.g. grandstands, or where at least 80% of the wall is open.
ings with roof but with no permanent walls are treated as Fre
standing canopies in $2.5,9./. (See 2.5.6, later.)
2.5.5.4.2 Application
O Internal pressure coefficients for open-sided buildings are given in
Table 18. Figure 56 gives a key to the column headings of Table
18 and indivates how the positive and negative values apply to the
“two adjacent open faces’ case at 0 = 9°
O Open-topped buildings, such as a building without à roof during
renovation, should he treated as free-standing walls using §2.8
(See 25,7, below.)
114 | Wind loading: a practical guide to BS 6399-2
Commentary 1 115
(a) One open face, shorter (b) One open face, longer
90 — ae cen ar
Era eg
of o
(6) Two adjacent open faces
of
domos cagon
o 56. Fat persed ings
Internal pressure coefficients for open-topped vertical cylinders are.
given in Table 19.
2.5.6 Free-standing canopies ($2.5.9.1)
256.1 Basis
“The provisions for free-standing canopies that had been introduced in
the 1986 amendments to CP 3-V-2 were transferred into BS 6399-2
without change. This is the only case where the previous ‘overall’ und
“local? coefficient format of CP 3-V-2 still applics. Values of net pressure
coefficient arc given in Tables 13 and 14. Positive valucs are detined as
downward loads and negative values as uplift. As no directional load ease
is ascribed to the values of net pressure coefficient C, they should be used
with the most onerous wind speed found for the site aver all wind
directions.
2
2 Blockage
he net pressure coefficients depend on the degree by which the wind
is blocked from flowing freely under the canopy. The blockage ratio ¢ is
defined as the height of obstructions divided by the height to caves in
Figure 24b, The blockage ratio affects the value of minimum (most
negative or uplift) coefficient, which increases as the blockage ratio
increases. Blockaye affects only the part of the canopy that is upwind of
the position of maximum blockage, so is most onemus when the blockage
is to the downwind eaves.
Values of C, for ¢=0 (no blockage) were obtained from measure-
ments on frce-standing canopies, but values for ¢ = 1 (fully blocked) were
derived from the external pressure coefficients in Tables 9 and 10 and the
internal pressure coefficients For open-sided buildings in Table 18, Values
for intermediate blockage ratios are obtained by interpolating between
2563 Application
© For loads on structural members, including shecting pu
the ‘Overall cuefficients
For loads on cladding and its fixings, use the “Local coefficients’
For downward loads, take the values of Cy for “Maximum, all €.
n
ins, use
irrespective of blockage ratio,
For upward loads, refer to Figure 25(b) and Fig. 57 to determine
the blockage ratio ¢ as follows.
© Interpolate values of Cy for blockage ratio between “Minimum
=U" and ‘Minimum ¢=1" and apply to all parts of the
canopy that lic upwind of the position of maximum blockage
(cross-hatched zone in Fig. 57
© Apply the values of C for "Minimum € =0" (o all parts of the
canopy that fie downwind of the position of maximum
blockage.
Use these values of Cy with the most onerous effective wind speed
Ve found for all wind directions. This i
© the value obtained using Option 1 — Irrespective of direction,
standard method, o
© the highest value found using the other options. (See 2.1.5.)
© For multi-bay free-standing canopy roofs the reduction factors
given in Table 15 may be applied to the net pressure coefficients,
Commentary 1 115
(a) One open face, shorter (b) One open face, longer
90 — ae cen ar
Era eg
of o
(6) Two adjacent open faces
of
domos cagon
o 56. Fat persed ings
Internal pressure coefficients for open-topped vertical cylinders are.
given in Table 19.
2.5.6 Free-standing canopies ($2.5.9.1)
256.1 Basis
“The provisions for free-standing canopies that had been introduced in
the 1986 amendments to CP 3-V-2 were transferred into BS 6399-2
without change. This is the only case where the previous ‘overall’ und
“local? coefficient format of CP 3-V-2 still applics. Values of net pressure
coefficient arc given in Tables 13 and 14. Positive valucs are detined as
downward loads and negative values as uplift. As no directional load ease
is ascribed to the values of net pressure coefficient C, they should be used
with the most onerous wind speed found for the site aver all wind
directions.
2
2 Blockage
he net pressure coefficients depend on the degree by which the wind
is blocked from flowing freely under the canopy. The blockage ratio ¢ is
defined as the height of obstructions divided by the height to caves in
Figure 24b, The blockage ratio affects the value of minimum (most
negative or uplift) coefficient, which increases as the blockage ratio
increases. Blockaye affects only the part of the canopy that is upwind of
the position of maximum blockage, so is most onemus when the blockage
is to the downwind eaves.
Values of C, for ¢=0 (no blockage) were obtained from measure-
ments on frce-standing canopies, but values for ¢ = 1 (fully blocked) were
derived from the external pressure coefficients in Tables 9 and 10 and the
internal pressure coefficients For open-sided buildings in Table 18, Values
for intermediate blockage ratios are obtained by interpolating between
2563 Application
© For loads on structural members, including shecting pu
the ‘Overall cuefficients
For loads on cladding and its fixings, use the “Local coefficients’
For downward loads, take the values of Cy for “Maximum, all €.
n
ins, use
irrespective of blockage ratio,
For upward loads, refer to Figure 25(b) and Fig. 57 to determine
the blockage ratio ¢ as follows.
© Interpolate values of Cy for blockage ratio between “Minimum
=U" and ‘Minimum ¢=1" and apply to all parts of the
canopy that lic upwind of the position of maximum blockage
(cross-hatched zone in Fig. 57
© Apply the values of C for "Minimum € =0" (o all parts of the
canopy that fie downwind of the position of maximum
blockage.
Use these values of Cy with the most onerous effective wind speed
Ve found for all wind directions. This i
© the value obtained using Option 1 — Irrespective of direction,
standard method, o
© the highest value found using the other options. (See 2.1.5.)
© For multi-bay free-standing canopy roofs the reduction factors
given in Table 15 may be applied to the net pressure coefficients,
116 | Wind loading: a practical guide to BS 6399-2
Commentary | 117
Postion of
ig. 37. Position of maximum blockage
2.5.7 Free-standing walls, parapets and signboards (§2.8)
257.1 Basis
Free-standing, walls were not included in CP 3-V-2, but are commonly
found as boundary walls. The 1972 Wind Loading Handbook? included
guidance predicting the largest luads on long boundary walls. Recent
research in full and model scale shows much of this guidance 10 have
been quite wrong—grossly conservative in the middle of long, walls,
but unsafe near any free ends. The highest loads occur at the free end of
à wall when the wind is incident at an angle of 0 =45", Loads increase
towards the comers of walls with return corners, but not as much as for a
free end.
Parapets on rectangular-plan buildings and the walls of buildings
without roofs, e.g. during renovation, should be treated xs houndary walls
with return comers. A moderate amount of porosity (20%) in the wall, ex.
a denso fence, suppresses these high loads.
Loads on more porous fences (less than 80% solid) are less than on à
solid wall of the same height. The boundary wall coefficients can be used
as safe upper-bound values for fences when applied 10 the overall
envelope arca (height x width) of the fence. Very porous fences should be
treated as latice frames (see 2.5.8.2)
‘A boundary wall or fence provides shelter downwind that extends for
‘many wall heights. Boundary walls offen uceur in parallel rows, each
sheltering the next wall downwind. This is the only case where BS 6399-2
permits the direct action of shelter to be considered for the ulti it
state, At very close spacing, the luad on the downwind wall can reduce
through zero to become negative (acting upwind). BS 6399-2 does not
allow a reduction factor smaller than 30%.
Signboards are also commonly associated
and BS 6399-2 includes design values in 32.8.2.
th building construction
25.7.2 Application (§2.8.1)
2.5.7.2.1 Free-standing walls, parapets and dense fences ($2.8.1.1)
O Look up the net pressure coefficient C in Table 21 using Figure
26 as the key, The loaded zones are dimensioned in terms of wall
or parapet height.
O Interpolate for solidity between the values for € — 1 (solid) and
0:8 (80% solid). You may use the values for & = 08 as safe
upper-bound values for more porous fenees, bul these may be
better assessed as lattice frames (see 2.8.8.2).
9 When the wall, parapet or fence is sheltered by another wall,
parapet or fenco of atleast the same height, apply the shelter factor
for fences in Figure 27 to the nel pressure coefficient Cy
257.22 Signboards (§2.8.2)
9 Take a value of net pre 4 to apply for
signboards suspended above ground, leaving a gap of at least half
their height, Take the normal force to act at half the height of the
board and anywhere within the range 25% of the width uf the
board either side of the centre, as shown in the key Figure 28.
2.57.23 Effectivewind speed
D Use these values of Cy withthe most onerous tre wind speed
Ve found for all wind directions. This is: =
© the value obtained using Option 1 Irrespective of direction,
standard method, or
© the highest value found using (he other options. (see §2.1.5.)
2.5.8 Pressure coefficients for elements (§2.7)
25.8.1 Individual sections
2581.1 Basis
Previously, CP 3-V-2 gave force cocfficients for a large number of
structural sections, measured in smooth uniform flow. IL is now
recognised that the apparent precision this gave when summed for a
large number of different sections is not real. In BS 6399-2 the force
coefficients become net pressure coefficients based on the area of the
Commentary | 117
Postion of
ig. 37. Position of maximum blockage
2.5.7 Free-standing walls, parapets and signboards (§2.8)
257.1 Basis
Free-standing, walls were not included in CP 3-V-2, but are commonly
found as boundary walls. The 1972 Wind Loading Handbook? included
guidance predicting the largest luads on long boundary walls. Recent
research in full and model scale shows much of this guidance 10 have
been quite wrong—grossly conservative in the middle of long, walls,
but unsafe near any free ends. The highest loads occur at the free end of
à wall when the wind is incident at an angle of 0 =45", Loads increase
towards the comers of walls with return corners, but not as much as for a
free end.
Parapets on rectangular-plan buildings and the walls of buildings
without roofs, e.g. during renovation, should be treated xs houndary walls
with return comers. A moderate amount of porosity (20%) in the wall, ex.
a denso fence, suppresses these high loads.
Loads on more porous fences (less than 80% solid) are less than on à
solid wall of the same height. The boundary wall coefficients can be used
as safe upper-bound values for fences when applied 10 the overall
envelope arca (height x width) of the fence. Very porous fences should be
treated as latice frames (see 2.5.8.2)
‘A boundary wall or fence provides shelter downwind that extends for
‘many wall heights. Boundary walls offen uceur in parallel rows, each
sheltering the next wall downwind. This is the only case where BS 6399-2
permits the direct action of shelter to be considered for the ulti it
state, At very close spacing, the luad on the downwind wall can reduce
through zero to become negative (acting upwind). BS 6399-2 does not
allow a reduction factor smaller than 30%.
Signboards are also commonly associated
and BS 6399-2 includes design values in 32.8.2.
th building construction
25.7.2 Application (§2.8.1)
2.5.7.2.1 Free-standing walls, parapets and dense fences ($2.8.1.1)
O Look up the net pressure coefficient C in Table 21 using Figure
26 as the key, The loaded zones are dimensioned in terms of wall
or parapet height.
O Interpolate for solidity between the values for € — 1 (solid) and
0:8 (80% solid). You may use the values for & = 08 as safe
upper-bound values for more porous fenees, bul these may be
better assessed as lattice frames (see 2.8.8.2).
9 When the wall, parapet or fence is sheltered by another wall,
parapet or fenco of atleast the same height, apply the shelter factor
for fences in Figure 27 to the nel pressure coefficient Cy
257.22 Signboards (§2.8.2)
9 Take a value of net pre 4 to apply for
signboards suspended above ground, leaving a gap of at least half
their height, Take the normal force to act at half the height of the
board and anywhere within the range 25% of the width uf the
board either side of the centre, as shown in the key Figure 28.
2.57.23 Effectivewind speed
D Use these values of Cy withthe most onerous tre wind speed
Ve found for all wind directions. This is: =
© the value obtained using Option 1 Irrespective of direction,
standard method, or
© the highest value found using (he other options. (see §2.1.5.)
2.5.8 Pressure coefficients for elements (§2.7)
25.8.1 Individual sections
2581.1 Basis
Previously, CP 3-V-2 gave force cocfficients for a large number of
structural sections, measured in smooth uniform flow. IL is now
recognised that the apparent precision this gave when summed for a
large number of different sections is not real. In BS 6399-2 the force
coefficients become net pressure coefficients based on the area of the
118 À Wind loading: a practical guide to BS 6399-2
Commentary | 119
clement normal to the wind. Simple standard values of nel pressure
coefficient, Cy=-2-0 for Mat-faced sections and C,— 1-2 for circular
(ubcritical) sections give almost as good a result. The largest values that
vceur for long soctions reduce with the length L of the section, so a
reduction factor for length x is given in Table 25.
“The value Cp =1-2 for circular sections assumes that the flow around
the clement will be ‘suberitical, which will be the case if B x Ve < 6 ms,
where the breadth 2 is the diameter of the section, For sections of large
diameter, the flow will become ‘superesitical’ at the effective wind spoud
Vz and the net pressure coefficient falls to C,- 0-5. However the
reduction factor for length does not apply to supercritical flow, so that
can halve in value between ‘subcritical’ and
supercritical” Now, itis possible for the largest load to occur at a wind
speed below the effective wind speed. The single value used by BS 6399
2 is therefore a safe approximation, but is conservative for long large-
Note: The overall load equation, Equations 7 and the implied factor of
0.85 do nut apply to individual elements. (See also 2.532.)
As structural sections are always slender, the division-by-parts rule
always applies (see 2.44). The element may be divided up into parts 28 in
length, or longer, and the effective wind speed taken at the top of each
pa
25.8.1.2 Application
Take the value of net pressure coefficient af Cy = 220 for flut-faced
sections and = 1-2 for circular sections.
19 Look up the reduction factor for length x in Table 25 and apply
this to reduce the value of C,. Note that a structural section or pipe
that spans between two walls is taken to be infinitely long and
K= 10.
CO Determine the effective wind speed und dynamic pressure along
the section at intervals not smaller than twee the section breadth
‘Alternatively take the effective wind speed at the highest point of
the section
For large diameter circular sections, e.g. pipe ducts and booms,
taking Cy = is equivalent to previous practice, hut the reduction
factor for length xs must be taken as unity.
The shielding factor for latice frames and trusses, explained in the
ext section, can be applicd to individual structural sections that
are shielded by an upwind lattice frame or wuss.
O Always use Equation 5 (not Equation 7) to obtain wind forces,
258.2 Lattice frames
BS 6399-2 recommends that loads on fences, lattice frames and
E a trusses
should be obtained by summing the loads on individual sections. ‘This
works well for very open lattices, Le. solidity ratio ¢ <0-1, but becomes
increasingly conservative with increasing, sulidity as more of the wind is
flow around the lattice instead of through it. BS 6399-2
: : For this by using the length of each element as the distance
hetween nodes in the lattice to calculate the reduction factor for length ».
AA the lattice becomes denser, the length between nodes becomes smaller,
so Figure 25 gives a smaller value of reduction factor. This is a pragmatic
‘compromise in the same way that altitude factor is used to compensate for
topography (ee 2.6 alien) AL very high wlll o, ¢ > O, a dense
atico becomes a porous wall (see 2.5.7, earlier). The
deinonstrated in Example 14. ii
Note: The solidity ratio € is defined as the
lefined as the projected area of elements
normal to the wind divided by the area of the envelope of the
lattice. It is related to porosity $ by &
Example 14. Net pressure coefficient for individual elements of
a lattice
&
‘The lali is composed of 400 mm wide Mat elements, and is suspenad clear of the
ground. For wind normal (o the plane ofthe Lance (ie. blowing into payo):
Commentary | 119
clement normal to the wind. Simple standard values of nel pressure
coefficient, Cy=-2-0 for Mat-faced sections and C,— 1-2 for circular
(ubcritical) sections give almost as good a result. The largest values that
vceur for long soctions reduce with the length L of the section, so a
reduction factor for length x is given in Table 25.
“The value Cp =1-2 for circular sections assumes that the flow around
the clement will be ‘suberitical, which will be the case if B x Ve < 6 ms,
where the breadth 2 is the diameter of the section, For sections of large
diameter, the flow will become ‘superesitical’ at the effective wind spoud
Vz and the net pressure coefficient falls to C,- 0-5. However the
reduction factor for length does not apply to supercritical flow, so that
can halve in value between ‘subcritical’ and
supercritical” Now, itis possible for the largest load to occur at a wind
speed below the effective wind speed. The single value used by BS 6399
2 is therefore a safe approximation, but is conservative for long large-
Note: The overall load equation, Equations 7 and the implied factor of
0.85 do nut apply to individual elements. (See also 2.532.)
As structural sections are always slender, the division-by-parts rule
always applies (see 2.44). The element may be divided up into parts 28 in
length, or longer, and the effective wind speed taken at the top of each
pa
25.8.1.2 Application
Take the value of net pressure coefficient af Cy = 220 for flut-faced
sections and = 1-2 for circular sections.
19 Look up the reduction factor for length x in Table 25 and apply
this to reduce the value of C,. Note that a structural section or pipe
that spans between two walls is taken to be infinitely long and
K= 10.
CO Determine the effective wind speed und dynamic pressure along
the section at intervals not smaller than twee the section breadth
‘Alternatively take the effective wind speed at the highest point of
the section
For large diameter circular sections, e.g. pipe ducts and booms,
taking Cy = is equivalent to previous practice, hut the reduction
factor for length xs must be taken as unity.
The shielding factor for latice frames and trusses, explained in the
ext section, can be applicd to individual structural sections that
are shielded by an upwind lattice frame or wuss.
O Always use Equation 5 (not Equation 7) to obtain wind forces,
258.2 Lattice frames
BS 6399-2 recommends that loads on fences, lattice frames and
E a trusses
should be obtained by summing the loads on individual sections. ‘This
works well for very open lattices, Le. solidity ratio ¢ <0-1, but becomes
increasingly conservative with increasing, sulidity as more of the wind is
flow around the lattice instead of through it. BS 6399-2
: : For this by using the length of each element as the distance
hetween nodes in the lattice to calculate the reduction factor for length ».
AA the lattice becomes denser, the length between nodes becomes smaller,
so Figure 25 gives a smaller value of reduction factor. This is a pragmatic
‘compromise in the same way that altitude factor is used to compensate for
topography (ee 2.6 alien) AL very high wlll o, ¢ > O, a dense
atico becomes a porous wall (see 2.5.7, earlier). The
deinonstrated in Example 14. ii
Note: The solidity ratio € is defined as the
lefined as the projected area of elements
normal to the wind divided by the area of the envelope of the
lattice. It is related to porosity $ by &
Example 14. Net pressure coefficient for individual elements of
a lattice
&
‘The lali is composed of 400 mm wide Mat elements, and is suspenad clear of the
ground. For wind normal (o the plane ofthe Lance (ie. blowing into payo):
120 | Wind loading: a practical guide to BS 6399-2
Commentary | 121
“The net pressure coefficient for Hat-taced sections is Cp=24.
2. The breadth of the elemenis Is ¿04m and the length between nodes is
1.250, giving LN 625.
2. From Figure 25, the reduction factor for length ix « 0-82
4, The effective nel pressure coefficient for each element is RG
0.82 2.0 164,
5. The effective net pressure cucficie for the whe lative is alo Cy
because it is hased on the area ofthe elements and not the overall envelope
6, "he projected area of elements normal tothe wind is À =envelope area-area
of holes = 10416 2.1 = 37-6"
7. The tal horizontal load Is Pa gs AR Cy = gs x 37-6 1:64
Ss
Better estimates are obtained by considering the lallice frame as a
whole, in terms of its solidity. This is how BS 8100 assesses lattice
towers, and which gives net pressure coefficients for plane lattice frames
and for three- and four-boom towers that are based on the solid area of the
‘windward face. For plane lattice frames the BS 8100 values are very
similar to the previous CP 3-V-2 values, but are still conserv
Reference 6 of BS 6399-2, the BRE Designer's Guide to Wind Loading of
Building Structures, Part 2, gives a more accurate methodology that is
summarized in Appendix À. You are recommended to use the method
and data in Appendix A for all latice frames and trusses, including
unclad building frames during construction.
2.5.9 Commentary on shape factors
“This scetion has described a number of changes from previous practice.
© All the shape evefticients are pressure coefficients:
— external pressure coefficients, Cpe
—internal pressure coefficients, Ca or
—net pressure coefficients, Cp = Ce Ce
» The pressure cucfficients were derived from measured peak
pressures divided by peak dynamic pressure, instead of mean
pressures divided by mean dynamic pressure, wherever possible.
‘These new values are close to the old values, but are more precise
and include building-generated turbulence.
+ The sizcs of the pressure coefficient zones are controlled by the
proportions of the front elevation, instead of by the plan
dimensions, using the scaling length b. This reflects the
distribution of pressure more accurately and requires fewer tables.
+ The zones of high suction around the periphery of buildings tend to
be smaller than previously, but the value of the cocfficient is often
higher. These zones are no longer ‘separately shown’, but act
together with the other zones to indicate the distribution of pressure.
across wall and roof surfaces.
+ External pressure coefficients have been provided for a wider range
of roof types and eaves details
+ The provisions fur internal pressure for enclosed buildings are
similar to before, but guidance is now given on sizes uf dominant
openings. New internal pressure coefficients are given for open-
Sided tidings.
+ Guelicients are now given for tee-stnding walls, parapets and
signboards. , —
Provisions for exposed structural elements and latices are
sili ta Comente, Addons gine Tor us
ames, ses and unelad building frames is given in Append
of this Guide. z Pates
2.6 Cladding, structural and overall loads
2.6.1 Normal pressures, loads and friction loads (§2.1.3, $3.1.3)
26.1.1 Normal pressures (52.1.3.1-3, §3.1.3.1)
‘The normal pressure coefficients are applied with the dynamic pressure.
at the reference height lo give the surface pressures. The force on the
building caused by these pressures acts normal to the surface.
In the directional method, the size effect is included in the equivalent
dynamic pressures ge and qj, so th
© External surface pressures are given by: p
lo Cre um
9 Internal surface pressures are given by: p=aG 18
Net surface pressures are given hy: me F4
O p=pe-p fur enclused buildings (19
O p=90C) for canopies, free-standing walls and
structural elements eo
where the values of the cquivalent dynamic pressures ge and q, depend on
the size ofthe loaded area and the internal volume respectively.
Inthe standard method, the size effect is separated wut into the size
effect factor C,, which depends on the sizeof the loaded area for external
pressures and the internal volume for intemal pressures, leaving the
standard dynamic pressure gs independent of size. So that:
© External surface pressures are given by: pe= gs CeCe (2)
© Internal surface pressures are given by: Pi=g eC, GQ)
O Not surface pressures are given by:
Commentary | 121
“The net pressure coefficient for Hat-taced sections is Cp=24.
2. The breadth of the elemenis Is ¿04m and the length between nodes is
1.250, giving LN 625.
2. From Figure 25, the reduction factor for length ix « 0-82
4, The effective nel pressure coefficient for each element is RG
0.82 2.0 164,
5. The effective net pressure cucficie for the whe lative is alo Cy
because it is hased on the area ofthe elements and not the overall envelope
6, "he projected area of elements normal tothe wind is À =envelope area-area
of holes = 10416 2.1 = 37-6"
7. The tal horizontal load Is Pa gs AR Cy = gs x 37-6 1:64
Ss
Better estimates are obtained by considering the lallice frame as a
whole, in terms of its solidity. This is how BS 8100 assesses lattice
towers, and which gives net pressure coefficients for plane lattice frames
and for three- and four-boom towers that are based on the solid area of the
‘windward face. For plane lattice frames the BS 8100 values are very
similar to the previous CP 3-V-2 values, but are still conserv
Reference 6 of BS 6399-2, the BRE Designer's Guide to Wind Loading of
Building Structures, Part 2, gives a more accurate methodology that is
summarized in Appendix À. You are recommended to use the method
and data in Appendix A for all latice frames and trusses, including
unclad building frames during construction.
2.5.9 Commentary on shape factors
“This scetion has described a number of changes from previous practice.
© All the shape evefticients are pressure coefficients:
— external pressure coefficients, Cpe
—internal pressure coefficients, Ca or
—net pressure coefficients, Cp = Ce Ce
» The pressure cucfficients were derived from measured peak
pressures divided by peak dynamic pressure, instead of mean
pressures divided by mean dynamic pressure, wherever possible.
‘These new values are close to the old values, but are more precise
and include building-generated turbulence.
+ The sizcs of the pressure coefficient zones are controlled by the
proportions of the front elevation, instead of by the plan
dimensions, using the scaling length b. This reflects the
distribution of pressure more accurately and requires fewer tables.
+ The zones of high suction around the periphery of buildings tend to
be smaller than previously, but the value of the cocfficient is often
higher. These zones are no longer ‘separately shown’, but act
together with the other zones to indicate the distribution of pressure.
across wall and roof surfaces.
+ External pressure coefficients have been provided for a wider range
of roof types and eaves details
+ The provisions fur internal pressure for enclosed buildings are
similar to before, but guidance is now given on sizes uf dominant
openings. New internal pressure coefficients are given for open-
Sided tidings.
+ Guelicients are now given for tee-stnding walls, parapets and
signboards. , —
Provisions for exposed structural elements and latices are
sili ta Comente, Addons gine Tor us
ames, ses and unelad building frames is given in Append
of this Guide. z Pates
2.6 Cladding, structural and overall loads
2.6.1 Normal pressures, loads and friction loads (§2.1.3, $3.1.3)
26.1.1 Normal pressures (52.1.3.1-3, §3.1.3.1)
‘The normal pressure coefficients are applied with the dynamic pressure.
at the reference height lo give the surface pressures. The force on the
building caused by these pressures acts normal to the surface.
In the directional method, the size effect is included in the equivalent
dynamic pressures ge and qj, so th
© External surface pressures are given by: p
lo Cre um
9 Internal surface pressures are given by: p=aG 18
Net surface pressures are given hy: me F4
O p=pe-p fur enclused buildings (19
O p=90C) for canopies, free-standing walls and
structural elements eo
where the values of the cquivalent dynamic pressures ge and q, depend on
the size ofthe loaded area and the internal volume respectively.
Inthe standard method, the size effect is separated wut into the size
effect factor C,, which depends on the sizeof the loaded area for external
pressures and the internal volume for intemal pressures, leaving the
standard dynamic pressure gs independent of size. So that:
© External surface pressures are given by: pe= gs CeCe (2)
© Internal surface pressures are given by: Pi=g eC, GQ)
O Not surface pressures are given by:
122 1 Wind loading: a practical guide 10 BS 6399-2
Commentary | 123
pe=m… for enclosed huildings @
1s Cp Ca. for eanopies, frec-standing walls and
structural elements. o
‘The size effect factor C, is described in 2.6.3, below.
“The major advantage of the separate size effect factor is that it allows
us ta calculate a single set of pressures und loads with C,—1, which we
‘will call ‘unfactored® loads. We may then apply the relevant sive effect
factor to these loads with the size effect factor applicable to each part or
‘element of the building at the end of the assessment, This procedure is
described later in 3.5.1.3.
2.6.12 Surface loads ($2.1.35, 99.132)
le loads normal Lo the building surface are determined by summing,
he pressures over the area of the surface, giving:
P=pA ©,
where A is the “loaded area’. The choice of the appropriate loaded area
epends on the form of the structure. This choice is usually obvious when
dealing with overall loads or loads on a whole component, e.g. uplift on a
roof. The choice is less obvious when dealing with an individual element
Supporting only part of a component, 6. the load on a single rafter, In
this ease, the area is the arca over which load is attracted to the structural
element, sometimes called the ‘wibutary area’. Detailed guidance on
‘options for wibutary areas is given in 3.4.1, later.
26.13 Overall toads ($2.1.26, 83.1.3.3)
“The overall Toads are the sum of the surface loads on the building, The
‘overall vertical and horizontal loads on a roof are therefore determined from
the distribution of extemal and internal pressures acting on the area of the
100F in plan and elevation respectively. The overall horizontal loads on the
‘whole building are determined from the loads on the roof and the walls. The
Size effect factor deals with the non-simoltancous action of gusts across each
of the faces of the building, but (here isa further reduction in overall load
caused by the nan-simultaneous action of gusts on windward and leeward
faves, BS 6399-2 deals with this by applying the factor 0-85 in Equations 7,
22 and 23 The value 0-85 was determined from quasi-steady theory (see
25.1.1) and confirmed by measurements in Full and model scale.
For the directional method, the component loads from all windward-
Facing walls and roof faces are included in Y) Pow and the component
loads from all leeward-facing walls and roof faves are included in > Pro
These are resolved into the wind direction using the wind angle normal to
each face 0 as shown in Fig. 58;
{7
x) ho
P< 088 COOP RE ang"
ig. 58. Overall load in wind direction by dirotional method (exeluding friction)
P=0:85 [ge cos 4) ~ EP cos D) (1+ Cr) (D (22)
This assumes that any cross-wind component of overall force is small
enough to be insignificant and that the overall force along this axis is
greater in the corresponding orthogonal wind directions.
For the exthogonal cases of the standard method:
P= 085( Pio — E Pun) (4a) am
When Equation 22 is used, $2.1.3.6 requires you to take ‘the inwind
dep of the building D as the smaller of width W or length L This
reduces the ratio D/H, and so makes it ess likely that the lower values of
pressure coefficient Co given in Table 4 for windward and leeward walls
can be used. This is now known to he unnecessarily oncrous.
‘When wind at skew angles is important, for stresses in comer columns
at 6 =45" for example, $2.1.3.6 recommends taking 80% of the sum of
the overall toads from both orthogonal casos. An altemativo to this is
given by Equation 23, which estimates the overall load in the wind
direction from the standard method pressure coefficients. This alternative
is explained by Note 3 of §3..3.3.2.
Note that Equation 21 and Equation 22 include the dynamic
augmentation factor C, which was obtained during the dynamic
classification, described in 2.2 earlier. These are the unly times C, is
needed since it applies only to the main structural members that resist the
overall dynamic response of the building,
2.6.1.4 Friction loads (§2.1.3.8, 83.134)
‘The flow of wind past the building surface generates a friction
rare othe Sac the direction of the ow: sings
Commentary | 123
pe=m… for enclosed huildings @
1s Cp Ca. for eanopies, frec-standing walls and
structural elements. o
‘The size effect factor C, is described in 2.6.3, below.
“The major advantage of the separate size effect factor is that it allows
us ta calculate a single set of pressures und loads with C,—1, which we
‘will call ‘unfactored® loads. We may then apply the relevant sive effect
factor to these loads with the size effect factor applicable to each part or
‘element of the building at the end of the assessment, This procedure is
described later in 3.5.1.3.
2.6.12 Surface loads ($2.1.35, 99.132)
le loads normal Lo the building surface are determined by summing,
he pressures over the area of the surface, giving:
P=pA ©,
where A is the “loaded area’. The choice of the appropriate loaded area
epends on the form of the structure. This choice is usually obvious when
dealing with overall loads or loads on a whole component, e.g. uplift on a
roof. The choice is less obvious when dealing with an individual element
Supporting only part of a component, 6. the load on a single rafter, In
this ease, the area is the arca over which load is attracted to the structural
element, sometimes called the ‘wibutary area’. Detailed guidance on
‘options for wibutary areas is given in 3.4.1, later.
26.13 Overall toads ($2.1.26, 83.1.3.3)
“The overall Toads are the sum of the surface loads on the building, The
‘overall vertical and horizontal loads on a roof are therefore determined from
the distribution of extemal and internal pressures acting on the area of the
100F in plan and elevation respectively. The overall horizontal loads on the
‘whole building are determined from the loads on the roof and the walls. The
Size effect factor deals with the non-simoltancous action of gusts across each
of the faces of the building, but (here isa further reduction in overall load
caused by the nan-simultaneous action of gusts on windward and leeward
faves, BS 6399-2 deals with this by applying the factor 0-85 in Equations 7,
22 and 23 The value 0-85 was determined from quasi-steady theory (see
25.1.1) and confirmed by measurements in Full and model scale.
For the directional method, the component loads from all windward-
Facing walls and roof faces are included in Y) Pow and the component
loads from all leeward-facing walls and roof faves are included in > Pro
These are resolved into the wind direction using the wind angle normal to
each face 0 as shown in Fig. 58;
{7
x) ho
P< 088 COOP RE ang"
ig. 58. Overall load in wind direction by dirotional method (exeluding friction)
P=0:85 [ge cos 4) ~ EP cos D) (1+ Cr) (D (22)
This assumes that any cross-wind component of overall force is small
enough to be insignificant and that the overall force along this axis is
greater in the corresponding orthogonal wind directions.
For the exthogonal cases of the standard method:
P= 085( Pio — E Pun) (4a) am
When Equation 22 is used, $2.1.3.6 requires you to take ‘the inwind
dep of the building D as the smaller of width W or length L This
reduces the ratio D/H, and so makes it ess likely that the lower values of
pressure coefficient Co given in Table 4 for windward and leeward walls
can be used. This is now known to he unnecessarily oncrous.
‘When wind at skew angles is important, for stresses in comer columns
at 6 =45" for example, $2.1.3.6 recommends taking 80% of the sum of
the overall toads from both orthogonal casos. An altemativo to this is
given by Equation 23, which estimates the overall load in the wind
direction from the standard method pressure coefficients. This alternative
is explained by Note 3 of §3..3.3.2.
Note that Equation 21 and Equation 22 include the dynamic
augmentation factor C, which was obtained during the dynamic
classification, described in 2.2 earlier. These are the unly times C, is
needed since it applies only to the main structural members that resist the
overall dynamic response of the building,
2.6.1.4 Friction loads (§2.1.3.8, 83.134)
‘The flow of wind past the building surface generates a friction
rare othe Sac the direction of the ow: sings
124 1 Wind loading: a practical guide to RS 6399-2
Commentary 1 125
Pr=gs CiAs Ca (a
where Gris the friction coefficient and A, isthe “arca swept by the wind’.
The size effect factor C, is relevant here because the smaller gusts cannot
‘sweep’ the whole of the arca A, simultaneously. The corresponding.
diagonal dimensiun « is the diagonal of the swept area.
Clause 2.3.1.8 states that the frictional forces fram Equation 7a are
additional 10 the overall forces calculated from Equation 7 or 22, ie. the
total load is P + Py, sa the 0-85 factor and €, do not apply to the frictional
component.
“The friction coefficient Ci, whi ven in Table 6, is very much
smaller in value than the normal pressure coefficient. This means that a
very much greater area is required before the friction forces are significant,
bout this may happen when the building is very long in the wind direction.
The “arca swept by the wind’ A, is defined in §2.4.5, $2.5.10 for the
standard method as Zone C on side walls and Zone Don flat roofs in all
wind directions and on pitched roots when the wind is parallel to the
ridge. For the disectional method A, is defined in §3.3.1.9, §3.32.8,
§3.3.3.9 and $3.3.4.3 as Zones C and D on side walls, Zone G on flat roofs
in all wind directions, and Zones F and P on pitched roofs when the wind
is parallel to the ridge.
2.6.2. Diagonal dimension ($2.1.3.4, $3.2.3.3)
2.62.1 External pressures
"The diagonal dimension is used to describe the size effect. [replaces
the ‘greatest horizontal dimension’ and ‘greatest vertical dimension’ in
85.5.2 of CP 3-V-2 and it eliminates the large steps between the previous
Classes A, B and €. The diagonal dimension is defined us ‘the largest
diagonal over which load sharing takes place’, so it is a function of the
structural form.
‘When there is no load sharing outside the component being considered,
the diagonal dimension is simply the largest dimension across the loaded
area. Figure 5 of BS 6399-2 gives a number of examples for whole
‘components, such as roof and wall faces, for shear at à given storey on a
tower and for a cladding panel (Figure 5(c)).
Figure 59 shows the example of one pitch of a roof, where panels are
supported between rafters. For the design of a panel, the loaded
tributary area and the corresponding diagonal dimension a are shown in
Fig. $9{a). It is likely that a zone boundary will tie within the wibutary
ara of the end panel when the wind is normal to the gable end, and
‘uidsnce on this is given later in 3.4.1. If the panels are simply supported
fon the rafters, then the rafters will attract half the load from the panel on
either side, leading to the tributary areas and diagonal dimensions shown
{a} root panels raters
(0) beams
lorend tame for ypleal ame
Fig. 60. Diagonal dimensions for portal frames
in Fig, $9(b). Note that the tributary area for the end rafter is half that of
‘the central rafters and the diagonal dimension is correspondingly smaller.
Larger, fabricated structural clements, such as portal frames, will
attract load fram more than one face of the building. Figure 60
demonstrates the case for horizontal racking loads on the frames of a
portal-framed building. This w to the rafters in Figure S9(b),
except that the diagonal dimension is taken for the projected area
spanning from halfway between frames and from ground to ridge,
assuming no load sharing between adjacent frames.
When there is significant load sharing, the area that defines the
diagonal dimension, a, will be larger than the luaded area. This means that
some of the localized gust loads on the surface are shed onto adjacent
elements that are not simultaneously loaded by the same gusts. This larger
Commentary 1 125
Pr=gs CiAs Ca (a
where Gris the friction coefficient and A, isthe “arca swept by the wind’.
The size effect factor C, is relevant here because the smaller gusts cannot
‘sweep’ the whole of the arca A, simultaneously. The corresponding.
diagonal dimensiun « is the diagonal of the swept area.
Clause 2.3.1.8 states that the frictional forces fram Equation 7a are
additional 10 the overall forces calculated from Equation 7 or 22, ie. the
total load is P + Py, sa the 0-85 factor and €, do not apply to the frictional
component.
“The friction coefficient Ci, whi ven in Table 6, is very much
smaller in value than the normal pressure coefficient. This means that a
very much greater area is required before the friction forces are significant,
bout this may happen when the building is very long in the wind direction.
The “arca swept by the wind’ A, is defined in §2.4.5, $2.5.10 for the
standard method as Zone C on side walls and Zone Don flat roofs in all
wind directions and on pitched roots when the wind is parallel to the
ridge. For the disectional method A, is defined in §3.3.1.9, §3.32.8,
§3.3.3.9 and $3.3.4.3 as Zones C and D on side walls, Zone G on flat roofs
in all wind directions, and Zones F and P on pitched roofs when the wind
is parallel to the ridge.
2.6.2. Diagonal dimension ($2.1.3.4, $3.2.3.3)
2.62.1 External pressures
"The diagonal dimension is used to describe the size effect. [replaces
the ‘greatest horizontal dimension’ and ‘greatest vertical dimension’ in
85.5.2 of CP 3-V-2 and it eliminates the large steps between the previous
Classes A, B and €. The diagonal dimension is defined us ‘the largest
diagonal over which load sharing takes place’, so it is a function of the
structural form.
‘When there is no load sharing outside the component being considered,
the diagonal dimension is simply the largest dimension across the loaded
area. Figure 5 of BS 6399-2 gives a number of examples for whole
‘components, such as roof and wall faces, for shear at à given storey on a
tower and for a cladding panel (Figure 5(c)).
Figure 59 shows the example of one pitch of a roof, where panels are
supported between rafters. For the design of a panel, the loaded
tributary area and the corresponding diagonal dimension a are shown in
Fig. $9{a). It is likely that a zone boundary will tie within the wibutary
ara of the end panel when the wind is normal to the gable end, and
‘uidsnce on this is given later in 3.4.1. If the panels are simply supported
fon the rafters, then the rafters will attract half the load from the panel on
either side, leading to the tributary areas and diagonal dimensions shown
{a} root panels raters
(0) beams
lorend tame for ypleal ame
Fig. 60. Diagonal dimensions for portal frames
in Fig, $9(b). Note that the tributary area for the end rafter is half that of
‘the central rafters and the diagonal dimension is correspondingly smaller.
Larger, fabricated structural clements, such as portal frames, will
attract load fram more than one face of the building. Figure 60
demonstrates the case for horizontal racking loads on the frames of a
portal-framed building. This w to the rafters in Figure S9(b),
except that the diagonal dimension is taken for the projected area
spanning from halfway between frames and from ground to ridge,
assuming no load sharing between adjacent frames.
When there is significant load sharing, the area that defines the
diagonal dimension, a, will be larger than the luaded area. This means that
some of the localized gust loads on the surface are shed onto adjacent
elements that are not simultaneously loaded by the same gusts. This larger
126 | Wind loading: a practical guide to BS 6399-2
arca must be defined from consideration of the structural form and the
relative stiffness of components and joints. The size effect factor is
reduced by only about 5% when the diagonal dimension is doubled.
UI
22 Internal pressures
2 ternal pressure. he siz effect of wind gui as 1 be replaced
by an equivalent value that represents the response time of the internal
volume, The basis of this procedure was described in 2,5.5.1.2 and will be
covered in more detail in 34.3.2 later.
For enclosed buildings without dominant openings (see 2.5.5.2), BS
6399-2 docs this by calculating an equivalent diagonal dimension from
the internal volume of the building or room:
10v0 (23) 19,04)
where O is:
© the internal volume of the storey for the typical case, where internal
doors are at least three times mone porous than extemal doors and
§2.6.1.1 applies, or
© the internal volume of the room, where internal doors are less than
three times more porous than external doors and $2.6.1.2 applies.
For enclosed buildings with a dominant opening (or more than one
opening in the same face) (sce 2.5.5.3), BS 6399-2 compares the diagonal
dimension of the opening with the equivalent from the internal volume
and uses the larger value:
@ of opening 7
a eae o 0.2Y0 of internal volume ©
The constants in Equations 23 and 24 are different.
Commentary | 127
2.6.3 Size effect factor of standard method (82.1.3.4)
The size effect factor ofthe standard method €, describes the decrease
in load intensity as the diagonal dimension increases, as represented by
the equivalent static gust (2.1.1 and Fig. 6). I isthe ratio between the
dynamic pressure for the diagonal dimension a to the dynamic pressure
for the datum a = 5 m.
Values of €, are given in Figure 4 as three lines, labelled “A”, ‘B* and
°C”, which correspond to the ‘sea’, “country” and ‘town’ terrain categories
respectively. The three lines are necessary hecause the size and intensity
of turbulent eddies differs over each terrain. The appropriate line is
selected using the key table under Figure 4, In towns, the change From °C’
lo Ro “A? occurs at the height of the sea-country and country-town
interface described in 24.1 and Fig. 23.
Note: The reason that Figure 4 gives Ca for diagonal dimensions up to
a= 1000m is not for buildings Tkm in size, but to accommodate
the equivalent diagonal dimension of large internal volumes.
(Note: Uquation 25 is incorrect in the first edition of this guide.)
2.64 Asymmetric loads (§2.1.3.7)
‘The standard method of BS 6399-2 and the previous CP 3-V-2 predict
symmetrical loadings for the orthogonal load cases that give no net
torque. Clause 2.1.3.7 is principally intended to cover the case illustrated
in Fig, 61. Without this clause, it would be possible to provide shear walls
— > > >
= 3
zZ | Ë —= 3
= 5 — E
= Ss =
|: 3 3 ES
3 3 = E
= 3 $ E
(0 Smet addons nctenvie (8 Anymneiieiairogine
any torsional resistance
Fig. 61. Asymmetric Inde
‘adequate torna resistance.
arca must be defined from consideration of the structural form and the
relative stiffness of components and joints. The size effect factor is
reduced by only about 5% when the diagonal dimension is doubled.
UI
22 Internal pressures
2 ternal pressure. he siz effect of wind gui as 1 be replaced
by an equivalent value that represents the response time of the internal
volume, The basis of this procedure was described in 2,5.5.1.2 and will be
covered in more detail in 34.3.2 later.
For enclosed buildings without dominant openings (see 2.5.5.2), BS
6399-2 docs this by calculating an equivalent diagonal dimension from
the internal volume of the building or room:
10v0 (23) 19,04)
where O is:
© the internal volume of the storey for the typical case, where internal
doors are at least three times mone porous than extemal doors and
§2.6.1.1 applies, or
© the internal volume of the room, where internal doors are less than
three times more porous than external doors and $2.6.1.2 applies.
For enclosed buildings with a dominant opening (or more than one
opening in the same face) (sce 2.5.5.3), BS 6399-2 compares the diagonal
dimension of the opening with the equivalent from the internal volume
and uses the larger value:
@ of opening 7
a eae o 0.2Y0 of internal volume ©
The constants in Equations 23 and 24 are different.
Commentary | 127
2.6.3 Size effect factor of standard method (82.1.3.4)
The size effect factor ofthe standard method €, describes the decrease
in load intensity as the diagonal dimension increases, as represented by
the equivalent static gust (2.1.1 and Fig. 6). I isthe ratio between the
dynamic pressure for the diagonal dimension a to the dynamic pressure
for the datum a = 5 m.
Values of €, are given in Figure 4 as three lines, labelled “A”, ‘B* and
°C”, which correspond to the ‘sea’, “country” and ‘town’ terrain categories
respectively. The three lines are necessary hecause the size and intensity
of turbulent eddies differs over each terrain. The appropriate line is
selected using the key table under Figure 4, In towns, the change From °C’
lo Ro “A? occurs at the height of the sea-country and country-town
interface described in 24.1 and Fig. 23.
Note: The reason that Figure 4 gives Ca for diagonal dimensions up to
a= 1000m is not for buildings Tkm in size, but to accommodate
the equivalent diagonal dimension of large internal volumes.
(Note: Uquation 25 is incorrect in the first edition of this guide.)
2.64 Asymmetric loads (§2.1.3.7)
‘The standard method of BS 6399-2 and the previous CP 3-V-2 predict
symmetrical loadings for the orthogonal load cases that give no net
torque. Clause 2.1.3.7 is principally intended to cover the case illustrated
in Fig, 61. Without this clause, it would be possible to provide shear walls
— > > >
= 3
zZ | Ë —= 3
= 5 — E
= Ss =
|: 3 3 ES
3 3 = E
= 3 $ E
(0 Smet addons nctenvie (8 Anymneiieiairogine
any torsional resistance
Fig. 61. Asymmetric Inde
‘adequate torna resistance.
128 1 Wind loading: a practical guide to BS 6399-2
that gave no torsional resistance as in Fig, 61(a), but, clearly, a torsion
core is needed as in (b). A
The original version of $2./.3.7 was implemented as a ‘catch-ull” to
over any case where increasing the wind load benefits the design, but
was found by many to be too complicated, ‘The 2002 amendment simply
directs the user to make ‘an allowance for asymmetry" in the standard
method where specific guidance is not given for a particular building
form, Le. most buildings.
Note 1 to 82.1.3.7 now advises the user to reduce all beneficial londs by
40%, ie. to 60% of their full value, Previously, anly the largest beneficial
loud contributing to the load effect was reduced.
Note 2 to $2.1.3.7 now advises the user to account for torsional effects
by displacing the loads on each face by 10% of the face width. Although
mot stated, the direction of the displacement on each face should always
act o increase the torque applied to the building.
3. Making BS 6399-2 work for you
3.1 Taking control
3.1.1 Servant or master
‘The very first point that needs to be made is that the role of BS 6399-2 is to
facilitate the design, not to constrain it. The Standard has the flexibility to
«ope with all but the most unusual of building forms. If you use the
information in BS 6399-2 to modify the building shape so that wind loads
are reduced and the design is more efficient (see 3.5.5, later), then you are
‘working smartly and the Standard is your servant. Bul if you change the
form of a building to match the simplified provisions of the Standard, then
you are working dumbly and allowing the Standard to become your master.
The true master of the design is the UK Building Regulations, which
stipulate à minimum level of performance. Approved Document A quotes
BS 6399-2 as ono way of meeting these requirements. You are free (0 use
any other method you choose, as indicated in $/.7 Scope—‘Other
methods may be used in place of the two methods given in this Standard,
provided that they can be shown to be equivalent.”
Equivalent ‘other methods” will include the following.
+ The guidance and additional data given in the appendives to this
Guide.
© The detailed guidance and data in References 6 and 8 of the
Standard," from which the Standard was built. This gives pressure
coefficients for shapes not included in the Standard.
+ The design aids recommended in 3.1.4, late.
© Fully dynamic design methods for buildings shown by the dynamic
classification to be outside the scope of BS 6399-2, for example
Annex B of the draft Burocode 1."
Note: I is unlikely that a typical building will fall outside the scope of BS
6399-2 unless it is very tall.
+ Wind tunnel tests (see 3.1.7, later).
‘The only thorn in this rosy picture is the role of Building Control. By
that gave no torsional resistance as in Fig, 61(a), but, clearly, a torsion
core is needed as in (b). A
The original version of $2./.3.7 was implemented as a ‘catch-ull” to
over any case where increasing the wind load benefits the design, but
was found by many to be too complicated, ‘The 2002 amendment simply
directs the user to make ‘an allowance for asymmetry" in the standard
method where specific guidance is not given for a particular building
form, Le. most buildings.
Note 1 to 82.1.3.7 now advises the user to reduce all beneficial londs by
40%, ie. to 60% of their full value, Previously, anly the largest beneficial
loud contributing to the load effect was reduced.
Note 2 to $2.1.3.7 now advises the user to account for torsional effects
by displacing the loads on each face by 10% of the face width. Although
mot stated, the direction of the displacement on each face should always
act o increase the torque applied to the building.
3. Making BS 6399-2 work for you
3.1 Taking control
3.1.1 Servant or master
‘The very first point that needs to be made is that the role of BS 6399-2 is to
facilitate the design, not to constrain it. The Standard has the flexibility to
«ope with all but the most unusual of building forms. If you use the
information in BS 6399-2 to modify the building shape so that wind loads
are reduced and the design is more efficient (see 3.5.5, later), then you are
‘working smartly and the Standard is your servant. Bul if you change the
form of a building to match the simplified provisions of the Standard, then
you are working dumbly and allowing the Standard to become your master.
The true master of the design is the UK Building Regulations, which
stipulate à minimum level of performance. Approved Document A quotes
BS 6399-2 as ono way of meeting these requirements. You are free (0 use
any other method you choose, as indicated in $/.7 Scope—‘Other
methods may be used in place of the two methods given in this Standard,
provided that they can be shown to be equivalent.”
Equivalent ‘other methods” will include the following.
+ The guidance and additional data given in the appendives to this
Guide.
© The detailed guidance and data in References 6 and 8 of the
Standard," from which the Standard was built. This gives pressure
coefficients for shapes not included in the Standard.
+ The design aids recommended in 3.1.4, late.
© Fully dynamic design methods for buildings shown by the dynamic
classification to be outside the scope of BS 6399-2, for example
Annex B of the draft Burocode 1."
Note: I is unlikely that a typical building will fall outside the scope of BS
6399-2 unless it is very tall.
+ Wind tunnel tests (see 3.1.7, later).
‘The only thorn in this rosy picture is the role of Building Control. By
130 | Wind loading: a practical guide to BS 6399-2
convincing your Building Control Officer (BCO) of the adequacy of your
design. This is because BS 6399-2 provides an area af common ground
for the designer and BCO in which both can have confidence through its
status and familiarity ofits use,
3.1.2. Precision and accuracy :
Values in BS 6399-2 tables are generally given to three significant
figures, and this is the minimum level of precision to which hand
calculations need to be performed. With the number of factors and
‘coelTicients required in the calculations, this level of precision results in a
{ypical accuracy of about 5% on wind speed and about 10% on loath. In
order to maintain this level of precision, you were advised in 2.4.5.2 10
use logarithmic interpolation in the S-factor tables, because columns of
distance and rows of effective height are given in logarithmic steps.
lication of Ingarithmic interpolation was demonstrated in Example 6.
Linear interpolation will tend to overestimate for distance to sea or in
(own, but underestimate for effective height. Nevertheless, linear
interpolation is generally adequate, except when the effective height is
less than about 10m.
‘There is a second source of error that affects the final accuracy. The S-
factor tables in BS 6399-2 were compiled by hun calculations to an
accuracy of about 2%. Most of the values in Tables 4, 22 and 25 we either
exact or differ by one in the last figure. However, the values of the
adjustment factor for sites in town terrain 7, in Table 23 are inexplicably
tou small by a consistent factor of 1-016, This is shown in Fig, 62 between
Y km and 3 km inside Ihe town for an effective height of 10 m. Also shown
EES
Datacom)
sai els ea atc ERENEENUN.3=:2223:
Making BS 6399-2 work for you | 131
are the interpolations using the values given in Table 23. This model error
was identified before the 1997 amendments, hut was deemed too small to
justify replacing Table 23 at that time. Nevertheless, you may expect the
tables to be corrected when there is a major revision.
Overall, the model errors in the 1997 revision of BS 6399-2 lead to
variations in effective wind speed of about 2%, which is far too small to
be of any concern in comparison with the uncertainties in defining the site
parameters. However, corrected versions of Tables 4, 22, 23 and 24 are
given in Appendix B.
3.1.3. Reducing conservatism
.1. Standard options
‘The major oppontunitics to reduce conservatism have already been
described. We have seen that the standard method for wind speed is
between 0% and 30%, but typically 14%, more conservative in towns than
the directional method. On the other hand, the standard method pressure
coefficients are not much more conservative than the directional method
values. Accordingly, this Guide and the BRE Digest 436? both
recommend you use Option 2—Onthogonal load cases, standard
method, for hand calculations, but you should use the terrain and
building factor from Equation 29 instead of the value from Table 4 for
sites in towns, Ihe BRE Digest 436° adopts this option for its example
calculations of pressures and loads on typical building forms,
2 Hybrid options
However, Option 2 may still be significantly conservative at sites
where the exposure varies greatly by direction, as demonstrated by 1
wind speed calculations of Chapter 4. These demonstrate the hybri
i ied by 83.42. The hybrid method that releases the most
+ 10 calculate Option 3, directional method dynamic pressures for
twelve 30%wide sectors (see Example 21 in 4.2.4) and then
+ to select the mont onerous values for each of the four orthogonal
cases (sec Example 22 in 4.2.4) to use with standard method
pressure coefficients,
‚This hybrid option can be significantly less conservative than Option 2
(Example 20 in 4.2.3). In hand calculations this would involve twelve
assessments of site exposure, distance In sen and distance in town.
However, use of the design aid BREVe eliminates all effort in assessing
the site exposure as well as the potential of errors in calculation, As the
aim of this Guide is to demonstrate optimum use of the Standard, the
example calculations of pressure and load in Chapter 4 use this hybri
pr a
convincing your Building Control Officer (BCO) of the adequacy of your
design. This is because BS 6399-2 provides an area af common ground
for the designer and BCO in which both can have confidence through its
status and familiarity ofits use,
3.1.2. Precision and accuracy :
Values in BS 6399-2 tables are generally given to three significant
figures, and this is the minimum level of precision to which hand
calculations need to be performed. With the number of factors and
‘coelTicients required in the calculations, this level of precision results in a
{ypical accuracy of about 5% on wind speed and about 10% on loath. In
order to maintain this level of precision, you were advised in 2.4.5.2 10
use logarithmic interpolation in the S-factor tables, because columns of
distance and rows of effective height are given in logarithmic steps.
lication of Ingarithmic interpolation was demonstrated in Example 6.
Linear interpolation will tend to overestimate for distance to sea or in
(own, but underestimate for effective height. Nevertheless, linear
interpolation is generally adequate, except when the effective height is
less than about 10m.
‘There is a second source of error that affects the final accuracy. The S-
factor tables in BS 6399-2 were compiled by hun calculations to an
accuracy of about 2%. Most of the values in Tables 4, 22 and 25 we either
exact or differ by one in the last figure. However, the values of the
adjustment factor for sites in town terrain 7, in Table 23 are inexplicably
tou small by a consistent factor of 1-016, This is shown in Fig, 62 between
Y km and 3 km inside Ihe town for an effective height of 10 m. Also shown
EES
Datacom)
sai els ea atc ERENEENUN.3=:2223:
Making BS 6399-2 work for you | 131
are the interpolations using the values given in Table 23. This model error
was identified before the 1997 amendments, hut was deemed too small to
justify replacing Table 23 at that time. Nevertheless, you may expect the
tables to be corrected when there is a major revision.
Overall, the model errors in the 1997 revision of BS 6399-2 lead to
variations in effective wind speed of about 2%, which is far too small to
be of any concern in comparison with the uncertainties in defining the site
parameters. However, corrected versions of Tables 4, 22, 23 and 24 are
given in Appendix B.
3.1.3. Reducing conservatism
.1. Standard options
‘The major oppontunitics to reduce conservatism have already been
described. We have seen that the standard method for wind speed is
between 0% and 30%, but typically 14%, more conservative in towns than
the directional method. On the other hand, the standard method pressure
coefficients are not much more conservative than the directional method
values. Accordingly, this Guide and the BRE Digest 436? both
recommend you use Option 2—Onthogonal load cases, standard
method, for hand calculations, but you should use the terrain and
building factor from Equation 29 instead of the value from Table 4 for
sites in towns, Ihe BRE Digest 436° adopts this option for its example
calculations of pressures and loads on typical building forms,
2 Hybrid options
However, Option 2 may still be significantly conservative at sites
where the exposure varies greatly by direction, as demonstrated by 1
wind speed calculations of Chapter 4. These demonstrate the hybri
i ied by 83.42. The hybrid method that releases the most
+ 10 calculate Option 3, directional method dynamic pressures for
twelve 30%wide sectors (see Example 21 in 4.2.4) and then
+ to select the mont onerous values for each of the four orthogonal
cases (sec Example 22 in 4.2.4) to use with standard method
pressure coefficients,
‚This hybrid option can be significantly less conservative than Option 2
(Example 20 in 4.2.3). In hand calculations this would involve twelve
assessments of site exposure, distance In sen and distance in town.
However, use of the design aid BREVe eliminates all effort in assessing
the site exposure as well as the potential of errors in calculation, As the
aim of this Guide is to demonstrate optimum use of the Standard, the
example calculations of pressure and load in Chapter 4 use this hybri
pr a
132 | Wind loading: a practical guide to BS 6399-2
Making BS 6399-2 work for you 1 133
3.1.33 Other opportunities
‘There are other minor reductions in conservatism within the
pressure and load clauses of the Standard, These include parapets
around flat and pitched roofs and the downwind bays of multi-bay
roofs, where the relevant clause in BS 6399-2 often includes the
phrase “may conservatively be taken’. Here the choice is often
whether itis beter to accept the conservatism in order to simplify
the caleulation or to follow the more complex rules.
‘Another opportunity is to allow for the effect of load sharing on
(he size effect factor (see 3.4.2).
The most significant opportunity to reduce conservatism that is
acknowledged, but not implemented, by BS 6399-2 is in loads on
lattice frames and trusses. Appendix À to this Guide gives the
method und data required (see 2.5.8.2 Lattice frames).
3.1.4 Design aids
3.1.4.1. Maps sultable for use with BS 6399-2
The most useful aid for obtaining the distances to sea and in town
anywhere inthe UK is a book of road maps that shows the 10km National
Grid Squares. Two books produced by the Automobile Association, The
AA Big Road Atlas and The AA Truckers Atlas, are very suitable and are
available at most petrol stations. These have the added advantage of a
large gazetteer-index giving the National Grid Reference of most named
locations. Strangely, most road atlases produced by the Ordnance Survey
lexed by an arbitrary grid and so gie unsuitable.
M For sites further than 20 km to the sea, simply count
10km cach.
1 Forsites between 2km and 20km from the sea or in town, estimate
by considering one tenths of grid squares as Ikm.
© For sites eloser than 2km to the sea or in town, use a larger scale
map.
id squares as
A large-scale site plan is required to determine the site altitude and the
position of surrounding huildings. Failing this, the latest Ordnance Survey
1/25000 scale Explorer series maps show both altitude contours and
vidual buildings. Both the Ordnance Survey 1/25 000 Explorer and
1/50000 Landranger series are suitable for determining the topographic
dimensions when topography is significant at the site.
3.142 Ordnance Survey Interactive Atlas of Great Britain
This is a computer-based atlas of Great Britain, requiring Windows
95™ or later running on a multimedia PC with a CD drive. It covers the
Aa di
‘The main strengths of this interactive atlas are the ability to search for
the National Grid reference of over 45000 place names and the facility to
measure distances. This makes the interactive atlas a useful “site finder’
for the BREVe program described below.
‘The main weakness is that the topography contours are shown at 200m
intervals. This indicates hilly areas, but is not sufficient 10 determine the
topographic dimensions or site altitude, for which a larger-scale site map
is needed.
The Ordnance Survey Interactive Atlas of Great Britain costs less than
£30 and is stocked by major computer stores.
13 This Guide
As well as noting some minor comections tu the current 1997 revision
of BS 6399-2, this Guide gives some additional information, including:
© external pressure coefficients for barrel-vault roofs in ‘Table 2
+ net pressure coefficients and shielding factors for lative frames and
trusses in Appendix A
© corrected factor tables in Appendix B.
‘This additional information complies with the requir
Scope, and so can be used as an aid to desi
ments of $1.7
3.1.44 Wind Loading Ready-Reckoner
‘The Wind Loading Ready-Reckoner for BS 6399 Part 2 1997 is
published by the BSI" to complement the Standard. It provides a quick
route through the standard method that eliminates most, but not all, the
wm. IL comprises six sets of tables.
1. Table A gives the displacement height Ha from the obstruction
height £4, and spacing X, (817.3, 24.3, Equation 6).
2. Table B gives the effective height 4, from the displacement height
Ha and reference height H, ($1.7.3, 24.3, Equation 7).
3. Tables C.1 10 C.13 give the terrain and building factor $, from the
effective height A, the distance from site to sea and the distance
in town (§2.2.3.3, 2.4.5).
4, Table D gives the dynamic pressure q, from the effective wind
speed Ve, replacing the Table 2 increments of Lm/s with the finer
increments of 0-1 nvs.
5. Table E gives the size effect factor C, from the diagonal dimension
a, replacing Figure 4 with a table ($2.1.3.4, 2.6.3, Equation 25).
6, Table F gives the dynamic augmentation facior Cy from building
height and building-type factor Ky, replacing Figure 3 (81.6.1).
‘The main value of the ready-reckoner tables lies in Tables C.1 to C.13.
The standard method table of terrain and building factor $,. Table 4,
Making BS 6399-2 work for you 1 133
3.1.33 Other opportunities
‘There are other minor reductions in conservatism within the
pressure and load clauses of the Standard, These include parapets
around flat and pitched roofs and the downwind bays of multi-bay
roofs, where the relevant clause in BS 6399-2 often includes the
phrase “may conservatively be taken’. Here the choice is often
whether itis beter to accept the conservatism in order to simplify
the caleulation or to follow the more complex rules.
‘Another opportunity is to allow for the effect of load sharing on
(he size effect factor (see 3.4.2).
The most significant opportunity to reduce conservatism that is
acknowledged, but not implemented, by BS 6399-2 is in loads on
lattice frames and trusses. Appendix À to this Guide gives the
method und data required (see 2.5.8.2 Lattice frames).
3.1.4 Design aids
3.1.4.1. Maps sultable for use with BS 6399-2
The most useful aid for obtaining the distances to sea and in town
anywhere inthe UK is a book of road maps that shows the 10km National
Grid Squares. Two books produced by the Automobile Association, The
AA Big Road Atlas and The AA Truckers Atlas, are very suitable and are
available at most petrol stations. These have the added advantage of a
large gazetteer-index giving the National Grid Reference of most named
locations. Strangely, most road atlases produced by the Ordnance Survey
lexed by an arbitrary grid and so gie unsuitable.
M For sites further than 20 km to the sea, simply count
10km cach.
1 Forsites between 2km and 20km from the sea or in town, estimate
by considering one tenths of grid squares as Ikm.
© For sites eloser than 2km to the sea or in town, use a larger scale
map.
id squares as
A large-scale site plan is required to determine the site altitude and the
position of surrounding huildings. Failing this, the latest Ordnance Survey
1/25000 scale Explorer series maps show both altitude contours and
vidual buildings. Both the Ordnance Survey 1/25 000 Explorer and
1/50000 Landranger series are suitable for determining the topographic
dimensions when topography is significant at the site.
3.142 Ordnance Survey Interactive Atlas of Great Britain
This is a computer-based atlas of Great Britain, requiring Windows
95™ or later running on a multimedia PC with a CD drive. It covers the
Aa di
‘The main strengths of this interactive atlas are the ability to search for
the National Grid reference of over 45000 place names and the facility to
measure distances. This makes the interactive atlas a useful “site finder’
for the BREVe program described below.
‘The main weakness is that the topography contours are shown at 200m
intervals. This indicates hilly areas, but is not sufficient 10 determine the
topographic dimensions or site altitude, for which a larger-scale site map
is needed.
The Ordnance Survey Interactive Atlas of Great Britain costs less than
£30 and is stocked by major computer stores.
13 This Guide
As well as noting some minor comections tu the current 1997 revision
of BS 6399-2, this Guide gives some additional information, including:
© external pressure coefficients for barrel-vault roofs in ‘Table 2
+ net pressure coefficients and shielding factors for lative frames and
trusses in Appendix A
© corrected factor tables in Appendix B.
‘This additional information complies with the requir
Scope, and so can be used as an aid to desi
ments of $1.7
3.1.44 Wind Loading Ready-Reckoner
‘The Wind Loading Ready-Reckoner for BS 6399 Part 2 1997 is
published by the BSI" to complement the Standard. It provides a quick
route through the standard method that eliminates most, but not all, the
wm. IL comprises six sets of tables.
1. Table A gives the displacement height Ha from the obstruction
height £4, and spacing X, (817.3, 24.3, Equation 6).
2. Table B gives the effective height 4, from the displacement height
Ha and reference height H, ($1.7.3, 24.3, Equation 7).
3. Tables C.1 10 C.13 give the terrain and building factor $, from the
effective height A, the distance from site to sea and the distance
in town (§2.2.3.3, 2.4.5).
4, Table D gives the dynamic pressure q, from the effective wind
speed Ve, replacing the Table 2 increments of Lm/s with the finer
increments of 0-1 nvs.
5. Table E gives the size effect factor C, from the diagonal dimension
a, replacing Figure 4 with a table ($2.1.3.4, 2.6.3, Equation 25).
6, Table F gives the dynamic augmentation facior Cy from building
height and building-type factor Ky, replacing Figure 3 (81.6.1).
‘The main value of the ready-reckoner tables lies in Tables C.1 to C.13.
The standard method table of terrain and building factor $,. Table 4,
134 | Wind loading: a practical guide 10 BS 6399-2
Making BS 6399-2 work for you | 135
derives from the directional method Equation 29 (Equation 10), but
Jestricts distance in town lo zero (open country) or greater than 2km.
Each of the Tables C.1 to C.13 gives an expanded version of Table 4 for a
different distance in town: Okm, 0-1 km, 02km, 0-5km, 07km, | km,
2km, Skm, 7km, 10km, 15 km, 20km and 30km. This allows the user 10
implement the hybrid method of 63.42, described in 3.3.3.5 later, as
simply as the standard method.
“The steps of value inthe index rows and columns of the tables ae fine
enough to avoid the need for interpolation, but additional conservatism is
removed if interpolation is used, particularly for small effective heights
"The teady-reckoner tables!” are available directly from BSI Sales.
2.145 BREWS
BREWS was a computer-based implementation of BS 6399-2 which
ran only under the Windows™ 95, 98 or NT operating systems. St had the
very ambitious scope of covering all the clauses ofthe Standard, but only
implemented the more common building shapes. For example, for
buildings with a T plan shape it covered the walls and flat roofs, but not
pitched roots. Except for the basic wind speed for some major towns,
BREWS did not include any information about UK sites. All the exposure
parameters required manual input
The main weaknesses of BREWS were that it has very limited
Atexiility, so is more ‘master’ than ‘servant? and a tendeney to ‘rash’ the
PC on which it was run. In view of the more successful implementations
of BREVe2 and BRECp, BREWS has been withdrawn from sale,
3.146 BREVeZ
BREVe2 is an automated implementation of the effective wind speed
dynamic pressure methods of BS 6399-2 that runs under all 32-bit
Windows™ operating systems since Windows95. BREVe2 automates the
wind specd parts of the standard and directional methods of BS 6399-2,
the "hybrid" method of BRE Digest 436 and the BRE Designer's Guide 10
Wind Leading of Building Structures‘ method from which the Standard
was derived.
The main strengths of BREVe2 ate
1. which allows the user to focus on
1. the high degree of automat
the deta of his site, and
2. its implementation as an ActiveX™ component which allows the
user to embed i into many different applications, including spread:
sheets.
These attributes have led to its incorporation into the design packages of
several major manufacturers.
BREVe2 obtains the input parameters for sites in the UK (presently
excluding Northern Ireland) automatically from the grid reference using
BRE and Ordnance Survey databases of ground roughness and topography.
“The ground roughness model of the Standard accounts for three roughness
categories - Sea, Land and Town — while the BRE Designer's Guide uses
additional intermediate categories. BREVe2 uses the full BRE Designer's
Guide catcgorics, thon simplifies these for use with the Standard. Fig, 63
e2 analysis of ground roughness forthe frst SOkm around
a site at 1km resolution. Roughness changes aro determined to 200km
from the site in each of the twelve 30"-wide sectors of direction from the
BRE database, Topographic parameters are determined from the Ordnance
Survey database. This enables BREVe2 to ealeulate the design wind specils
appropriate for the most=cxposed location in every 1 km square of the UK
bout need for user input.
To optimize a particular site, a ‘New Site Wizard’ offers a series of
standard choices to obtain only such additional information as is needed
about the site, including heights and spacing of surrounding obstructions
‘when the site is in a town, These choices depend on the location of the site
Fig. 6. BREVE ground roughness dirty for SOK aroun Path, Maut Baron,
SK492S27 # sie
Making BS 6399-2 work for you | 135
derives from the directional method Equation 29 (Equation 10), but
Jestricts distance in town lo zero (open country) or greater than 2km.
Each of the Tables C.1 to C.13 gives an expanded version of Table 4 for a
different distance in town: Okm, 0-1 km, 02km, 0-5km, 07km, | km,
2km, Skm, 7km, 10km, 15 km, 20km and 30km. This allows the user 10
implement the hybrid method of 63.42, described in 3.3.3.5 later, as
simply as the standard method.
“The steps of value inthe index rows and columns of the tables ae fine
enough to avoid the need for interpolation, but additional conservatism is
removed if interpolation is used, particularly for small effective heights
"The teady-reckoner tables!” are available directly from BSI Sales.
2.145 BREWS
BREWS was a computer-based implementation of BS 6399-2 which
ran only under the Windows™ 95, 98 or NT operating systems. St had the
very ambitious scope of covering all the clauses ofthe Standard, but only
implemented the more common building shapes. For example, for
buildings with a T plan shape it covered the walls and flat roofs, but not
pitched roots. Except for the basic wind speed for some major towns,
BREWS did not include any information about UK sites. All the exposure
parameters required manual input
The main weaknesses of BREWS were that it has very limited
Atexiility, so is more ‘master’ than ‘servant? and a tendeney to ‘rash’ the
PC on which it was run. In view of the more successful implementations
of BREVe2 and BRECp, BREWS has been withdrawn from sale,
3.146 BREVeZ
BREVe2 is an automated implementation of the effective wind speed
dynamic pressure methods of BS 6399-2 that runs under all 32-bit
Windows™ operating systems since Windows95. BREVe2 automates the
wind specd parts of the standard and directional methods of BS 6399-2,
the "hybrid" method of BRE Digest 436 and the BRE Designer's Guide 10
Wind Leading of Building Structures‘ method from which the Standard
was derived.
The main strengths of BREVe2 ate
1. which allows the user to focus on
1. the high degree of automat
the deta of his site, and
2. its implementation as an ActiveX™ component which allows the
user to embed i into many different applications, including spread:
sheets.
These attributes have led to its incorporation into the design packages of
several major manufacturers.
BREVe2 obtains the input parameters for sites in the UK (presently
excluding Northern Ireland) automatically from the grid reference using
BRE and Ordnance Survey databases of ground roughness and topography.
“The ground roughness model of the Standard accounts for three roughness
categories - Sea, Land and Town — while the BRE Designer's Guide uses
additional intermediate categories. BREVe2 uses the full BRE Designer's
Guide catcgorics, thon simplifies these for use with the Standard. Fig, 63
e2 analysis of ground roughness forthe frst SOkm around
a site at 1km resolution. Roughness changes aro determined to 200km
from the site in each of the twelve 30"-wide sectors of direction from the
BRE database, Topographic parameters are determined from the Ordnance
Survey database. This enables BREVe2 to ealeulate the design wind specils
appropriate for the most=cxposed location in every 1 km square of the UK
bout need for user input.
To optimize a particular site, a ‘New Site Wizard’ offers a series of
standard choices to obtain only such additional information as is needed
about the site, including heights and spacing of surrounding obstructions
‘when the site is in a town, These choices depend on the location of the site
Fig. 6. BREVE ground roughness dirty for SOK aroun Path, Maut Baron,
SK492S27 # sie
136 | Wind loading: a practical guide to BS 6399-2
Making RS 6399-2 work for you | 137
Fig. 64. Example of BREVE2 New Ste Wizard
and the safest choice is always pre-selected; a typical example is shown in
Fig. 64, Ail the site parameters can be adjusted and sites outside the UK
can be input by hand.
BREVe2 reports the effective gust speeds and dynamic pressures for static
design of buildings and structures and also the mean wind speed and
turbulence intensity for use in dynamic design methods. Tt includes the
option to report the values of all the intermediate BS 6399-2 and BRE
Designer's Guide parameters and factors for quality assurance purposes.
Figure 65(a) shows an example of a typical report of design dynamic press-
ture corresponding to Option 3 — twelve 30?-wide scctors, directional method.
The main weakness of BREVe2 is that the resolution of the full
automation is 1 km At sites close to the coast or the edge of a town, the
‘New Site Wizard’ asks for the necessary inform ie the site.
For sites in complex topography, the resolution of the Ordnance Survey
database used by BREVe2 is not fine enough to define the hill shape, so
that topography is always represented as a conical hill, Le. the topographic
increment isthe same in all wind directions. However, this weakness may
be overcome by using the “Wind over hills’ service, described below.
BREVe2 may be purchased for use on a single PC from:
+ Anemos Associates Ltd, 14 The Chestnuts, Hemel Hempstead,
HP3 ODZ. Tel: 01442 212292. Fax: 01442 256155.
www anemos.co.uk
+ British Standards Institution, Sales Department, 389 Chiswick High
Road, London Wa 4A. Tel: 0181 996 7000. Fax: 0181 996 7001.
+ BRE Bookshop, Building Research Establishment, Watford, WD25
9XX. Tel: 10923 664262, Fax: 01923 664605.
Fig. 65), Example of DREVEZ report
Development and distribution licences for embedding BREVe2 into
‘commercial applications or far use over networks may be obtained from
Anemos Associates,
3147 BRECp
BRECp is an ActiveX component that is complementary to
BREVe2. It provides validated values of pressure coefficient C, and
scaling parameter b from the parameters specifying the building form
and zane, interpolating for wind angle, roof pitch, ete, wherever
required, It automatically complies with all the relevant notes and
‘caveats in the Standard, reporting the choices it has made to the user.
For example, given parameters that specify an eaves zone on a
pitched roof with a parapet, it determines the pressure coefficient for
‘that zone on a pitched roof with a sharp eave and compares this with
the value for a Mat root with parapet, using $3.3.7./ to determine
PAN RA AA
Making RS 6399-2 work for you | 137
Fig. 64. Example of BREVE2 New Ste Wizard
and the safest choice is always pre-selected; a typical example is shown in
Fig. 64, Ail the site parameters can be adjusted and sites outside the UK
can be input by hand.
BREVe2 reports the effective gust speeds and dynamic pressures for static
design of buildings and structures and also the mean wind speed and
turbulence intensity for use in dynamic design methods. Tt includes the
option to report the values of all the intermediate BS 6399-2 and BRE
Designer's Guide parameters and factors for quality assurance purposes.
Figure 65(a) shows an example of a typical report of design dynamic press-
ture corresponding to Option 3 — twelve 30?-wide scctors, directional method.
The main weakness of BREVe2 is that the resolution of the full
automation is 1 km At sites close to the coast or the edge of a town, the
‘New Site Wizard’ asks for the necessary inform ie the site.
For sites in complex topography, the resolution of the Ordnance Survey
database used by BREVe2 is not fine enough to define the hill shape, so
that topography is always represented as a conical hill, Le. the topographic
increment isthe same in all wind directions. However, this weakness may
be overcome by using the “Wind over hills’ service, described below.
BREVe2 may be purchased for use on a single PC from:
+ Anemos Associates Ltd, 14 The Chestnuts, Hemel Hempstead,
HP3 ODZ. Tel: 01442 212292. Fax: 01442 256155.
www anemos.co.uk
+ British Standards Institution, Sales Department, 389 Chiswick High
Road, London Wa 4A. Tel: 0181 996 7000. Fax: 0181 996 7001.
+ BRE Bookshop, Building Research Establishment, Watford, WD25
9XX. Tel: 10923 664262, Fax: 01923 664605.
Fig. 65), Example of DREVEZ report
Development and distribution licences for embedding BREVe2 into
‘commercial applications or far use over networks may be obtained from
Anemos Associates,
3147 BRECp
BRECp is an ActiveX component that is complementary to
BREVe2. It provides validated values of pressure coefficient C, and
scaling parameter b from the parameters specifying the building form
and zane, interpolating for wind angle, roof pitch, ete, wherever
required, It automatically complies with all the relevant notes and
‘caveats in the Standard, reporting the choices it has made to the user.
For example, given parameters that specify an eaves zone on a
pitched roof with a parapet, it determines the pressure coefficient for
‘that zone on a pitched roof with a sharp eave and compares this with
the value for a Mat root with parapet, using $3.3.7./ to determine
PAN RA AA
138 | Wind loading: a practical guide to BS 6399-2
Making BS 6399-2 work for you | 139
Cees
ig 65), Example of BREC test application
directional method for this ease, BRECp calculates the directional
method values 15° increments over the required range of wind
direction and selects the most onerous value for the standard
method. This is illustrated in Fig. 65(b), where the choices used are
listed on the right-hand side
BRECp is embedded in a number of commercial applications. IL
comes as un ActiveX component, with examples of its use in
Microsoft Excel spreadshccts, Visual Basic and Pascal-based Delphi.
‘The Delphi example shown in Fig. 65(b) s a useful applica
own right, but lacks the sophistication and reporting features of
BREVe2. Single- ices and licences for embedding BRECP
into commercial applications or fur use over networks may be
‚obtained from Anemos Associates at Ihe address given above.
3.14.8 Wind speeds over complex terrain design service
the topography model of BS 6399-2 cupes well with simple
isolated hills, ridges, cliffs and escarpments, it may not be adequate
when the topography is complex, such as multiple hills, or hills that
are not a simple shape, where a more sophisticated model should be
used instead. The wind speeds over complex terrain design service
provides detailed information on changes in wind speed and direction
‘over complex hill terrain, compatible with BS 6399-2.
The scrvico combines the well established numerical model
MSMicro for flow over threc-dimensional topography, developed for
nd ar HG AS ia]
digital elevation data licensed by the Ordnance Survey from their
LandForm Panorama model of the United Kingdom (excluding North-
em Ireland). For sites overseas the dala are derived by digitally
scanning topographic maps. The MSMicro model is valid for “moder-
at slopes, meaning slopes up to a gradient of 1 in 3. In practice
will include the majority of sites where construction is possible. Where
thre are gradients greater than about 1 in, flow separation is kely to
occur, giving regions of shelter that will not be predicted by MSMicro.
‘The model is therefore conservative and ‘fail-safe”.
‘The results ofthe service may be used us a direct replacement for
the topographic inererment, Sy, They are supplied in a form that may
be incorporated into reports or imported directly into the BRE
design wind speed program BREVe or into spreadsheets.
“The wind speeds over complex terrain design service is currently
‘operated by:
+ Anemos Associates Ltd, 14 The Chestnuts, Beechwood Park,
‘Hemel Hempstead, HP3 0DZ. Tel: 01442 212292, Fax: 01442
256155. Web: www.anemos.co.uk
‘© Building Research Establishment, Wind Loading Section, Garston,
‘Watford WD2 JR. Tel: 01923 664533. Fax: 01923 664096.
3.1.5. Alternative source of design wind data
3.1.5.1 For the United Kingdom
The basic wind speed map Figure 6 was compiled in 1984 from
the best available data available up to 1981. ‘The values from
anemogreph stations in Fig. 16 that do not conform to the definition
of Va have been corrected, e.g. at sites ncar the coast or at altitude
‘These exposure corrections match, as far as possible, the exposure
model used in BS 6399-2 (see 2.4.1).
Since this map was compiled wind data for a further 18 years
have been recorded and the anemometer exposures re-appraised!®
using exactly the BS 6399-2 exposure model. This means that the
BS 6399-2 wind map is likely to be revised some time soon. In the
‘meantime, the basic wind speed map, Figure 6 in BS 6399-2,
represents the best available estimates of Vp, Accordingly, you are
strongly discouraged from adopting values from other sources.
15.2 For overseas sites
‘The procedure to make data from overseas sites compatible with
BS 6399-2 was described in 2.3.7.3 and the conversion of gust data
was ilusirated by Example 5. You should make every effort to obi
A RN NN ae gee
Making BS 6399-2 work for you | 139
Cees
ig 65), Example of BREC test application
directional method for this ease, BRECp calculates the directional
method values 15° increments over the required range of wind
direction and selects the most onerous value for the standard
method. This is illustrated in Fig. 65(b), where the choices used are
listed on the right-hand side
BRECp is embedded in a number of commercial applications. IL
comes as un ActiveX component, with examples of its use in
Microsoft Excel spreadshccts, Visual Basic and Pascal-based Delphi.
‘The Delphi example shown in Fig. 65(b) s a useful applica
own right, but lacks the sophistication and reporting features of
BREVe2. Single- ices and licences for embedding BRECP
into commercial applications or fur use over networks may be
‚obtained from Anemos Associates at Ihe address given above.
3.14.8 Wind speeds over complex terrain design service
the topography model of BS 6399-2 cupes well with simple
isolated hills, ridges, cliffs and escarpments, it may not be adequate
when the topography is complex, such as multiple hills, or hills that
are not a simple shape, where a more sophisticated model should be
used instead. The wind speeds over complex terrain design service
provides detailed information on changes in wind speed and direction
‘over complex hill terrain, compatible with BS 6399-2.
The scrvico combines the well established numerical model
MSMicro for flow over threc-dimensional topography, developed for
nd ar HG AS ia]
digital elevation data licensed by the Ordnance Survey from their
LandForm Panorama model of the United Kingdom (excluding North-
em Ireland). For sites overseas the dala are derived by digitally
scanning topographic maps. The MSMicro model is valid for “moder-
at slopes, meaning slopes up to a gradient of 1 in 3. In practice
will include the majority of sites where construction is possible. Where
thre are gradients greater than about 1 in, flow separation is kely to
occur, giving regions of shelter that will not be predicted by MSMicro.
‘The model is therefore conservative and ‘fail-safe”.
‘The results ofthe service may be used us a direct replacement for
the topographic inererment, Sy, They are supplied in a form that may
be incorporated into reports or imported directly into the BRE
design wind speed program BREVe or into spreadsheets.
“The wind speeds over complex terrain design service is currently
‘operated by:
+ Anemos Associates Ltd, 14 The Chestnuts, Beechwood Park,
‘Hemel Hempstead, HP3 0DZ. Tel: 01442 212292, Fax: 01442
256155. Web: www.anemos.co.uk
‘© Building Research Establishment, Wind Loading Section, Garston,
‘Watford WD2 JR. Tel: 01923 664533. Fax: 01923 664096.
3.1.5. Alternative source of design wind data
3.1.5.1 For the United Kingdom
The basic wind speed map Figure 6 was compiled in 1984 from
the best available data available up to 1981. ‘The values from
anemogreph stations in Fig. 16 that do not conform to the definition
of Va have been corrected, e.g. at sites ncar the coast or at altitude
‘These exposure corrections match, as far as possible, the exposure
model used in BS 6399-2 (see 2.4.1).
Since this map was compiled wind data for a further 18 years
have been recorded and the anemometer exposures re-appraised!®
using exactly the BS 6399-2 exposure model. This means that the
BS 6399-2 wind map is likely to be revised some time soon. In the
‘meantime, the basic wind speed map, Figure 6 in BS 6399-2,
represents the best available estimates of Vp, Accordingly, you are
strongly discouraged from adopting values from other sources.
15.2 For overseas sites
‘The procedure to make data from overseas sites compatible with
BS 6399-2 was described in 2.3.7.3 and the conversion of gust data
was ilusirated by Example 5. You should make every effort to obi
A RN NN ae gee
140 |. Wind loading: a practical guide to BS 6399-2
Making BS 6399-2 work for you | 141
À standard Gumbel analysis of annual maximum wind speeds is
demonstrated in Example 15. This is the form of analysis uscd to
compile CP 3-V-2. There are other methods that are more
aceurate.! The method used 10 compile Figure 6 requires access
to the full hourly record of wind speeds and is not suitable for
manual analysis.
the relevant meteorological authority. Only in the last resort should
‘you attempt your own analysis of wind speed records.
Example 15. Gumbel analysis of annual maximum wind speeds
For this example we use the 21-year record of annual maximum gust wind spocd at
Jersey Airport, 1958-1978.
"The recorded values ae as follows.
Year Year Peas Yea Dim
PRET 1965385 wn 3905
1959 37 1966 405 285
1960 1 196739 1368
1961 196839 1975 5
CCR 169 34 1976
1963 3 ma 1977
ES CRE 197
‘here are W = 21 onal values of wind speed
1. Son the values of wind speed ino ascending order of value and assign each
value a rank mn from the smallest m = 1 10 the largest nı = N = 21.
2. Caleulare the
w+).
3. Square the values of wind speed, Y
‘his procedure is summarized in the following table.
jumbel “reduced variate’ for each rank -In{-Injm/
Hank A loja 1 D) Dm Ranke Inn à D) PsP
' 112851 Fran 0.500651 132
2 “087439 su B 0642217 16
3 0.68936 961 1 0794106 naa
4 sa 96 15 0939741 1434
5 00a m 1164278 1482
& = 076181 mm 1355458 1182
7 uns 1056 IS 1.60609 1321
8 os ns 19 1920024 1321
3 0412253 usé 20 2350619 1600
10 027677 no a 3057873, 1630
m 0.366513 1296
4. Now plot the values of In{—lnfn/(W + 1) agains the values ot 92 as
shown in she graph elow anf he bos straight line though the data points
5 Read off the value Y? where —in{inin/(¥ +1) — 3:9, which
corresponds to Q = 002 In this example V2 — 2012m1%
6 Take the square oo! to ive the value with a annua ik of © = 0:02 otis
example Y = 44-85 ms,
7. Time wind sped is gust speed,
procedure given in Example 5
8. Inthisexamplo,V,=44:89/1-78 - 25.2 ms. The denominatoris 1:78 because
the anermometer is Kan from the sea
wert this to the site wind speed using the
Note: The calculated wind speed is higher than the value in Figure 6 because the
quality of te annual maximum wind speed data in this analysis és nut as good
5 the Jul record used in BS 6399-2.
COM
‘This example demonstates chat real data will not always fall nel onto the
‘expected siraight line in the Gumbel pot, as shown abone. In this case, the st
lime was fitted by the method of least squares by using the "add linear trendline” and
‘show equation on chart” options of the Microsoft Excel spreadsheet,
‘This analysis assumes that the anemumeler is exposed at the meteorological
standard. Ifthe anemomeler isnot a Ihe standard height of 10m above ground or not
in the standard open country exposure you should seek expert advice,
Clause 1.1 Scope of BS 6399-2 provides a route to use extemal
pressure coefficients from other sources, provided that they can he
shown to be equivalent, This will only be necessary for unusual
shapes that are not covered by the Standard. There are two main
en As
| | 3.1.6 Alternative sources of data for shapes not in the Standard
Making BS 6399-2 work for you | 141
À standard Gumbel analysis of annual maximum wind speeds is
demonstrated in Example 15. This is the form of analysis uscd to
compile CP 3-V-2. There are other methods that are more
aceurate.! The method used 10 compile Figure 6 requires access
to the full hourly record of wind speeds and is not suitable for
manual analysis.
the relevant meteorological authority. Only in the last resort should
‘you attempt your own analysis of wind speed records.
Example 15. Gumbel analysis of annual maximum wind speeds
For this example we use the 21-year record of annual maximum gust wind spocd at
Jersey Airport, 1958-1978.
"The recorded values ae as follows.
Year Year Peas Yea Dim
PRET 1965385 wn 3905
1959 37 1966 405 285
1960 1 196739 1368
1961 196839 1975 5
CCR 169 34 1976
1963 3 ma 1977
ES CRE 197
‘here are W = 21 onal values of wind speed
1. Son the values of wind speed ino ascending order of value and assign each
value a rank mn from the smallest m = 1 10 the largest nı = N = 21.
2. Caleulare the
w+).
3. Square the values of wind speed, Y
‘his procedure is summarized in the following table.
jumbel “reduced variate’ for each rank -In{-Injm/
Hank A loja 1 D) Dm Ranke Inn à D) PsP
' 112851 Fran 0.500651 132
2 “087439 su B 0642217 16
3 0.68936 961 1 0794106 naa
4 sa 96 15 0939741 1434
5 00a m 1164278 1482
& = 076181 mm 1355458 1182
7 uns 1056 IS 1.60609 1321
8 os ns 19 1920024 1321
3 0412253 usé 20 2350619 1600
10 027677 no a 3057873, 1630
m 0.366513 1296
4. Now plot the values of In{—lnfn/(W + 1) agains the values ot 92 as
shown in she graph elow anf he bos straight line though the data points
5 Read off the value Y? where —in{inin/(¥ +1) — 3:9, which
corresponds to Q = 002 In this example V2 — 2012m1%
6 Take the square oo! to ive the value with a annua ik of © = 0:02 otis
example Y = 44-85 ms,
7. Time wind sped is gust speed,
procedure given in Example 5
8. Inthisexamplo,V,=44:89/1-78 - 25.2 ms. The denominatoris 1:78 because
the anermometer is Kan from the sea
wert this to the site wind speed using the
Note: The calculated wind speed is higher than the value in Figure 6 because the
quality of te annual maximum wind speed data in this analysis és nut as good
5 the Jul record used in BS 6399-2.
COM
‘This example demonstates chat real data will not always fall nel onto the
‘expected siraight line in the Gumbel pot, as shown abone. In this case, the st
lime was fitted by the method of least squares by using the "add linear trendline” and
‘show equation on chart” options of the Microsoft Excel spreadsheet,
‘This analysis assumes that the anemumeler is exposed at the meteorological
standard. Ifthe anemomeler isnot a Ihe standard height of 10m above ground or not
in the standard open country exposure you should seek expert advice,
Clause 1.1 Scope of BS 6399-2 provides a route to use extemal
pressure coefficients from other sources, provided that they can he
shown to be equivalent, This will only be necessary for unusual
shapes that are not covered by the Standard. There are two main
en As
| | 3.1.6 Alternative sources of data for shapes not in the Standard
142 Y Wind loading: a practical guide to BS 6399-2
Making BS 6399-2 work for you | 143
1. Published design guidance from authoritative sources such as:
@
(b)
©
other British, European and 150 Standards
Building Research Establishment Digests and other
publications
ESDU International Data Skeets (ESDU International,
27 Corsham Street, London N} GUA. Tel: 0171 490
5151).
2. Research papers published in pecr-roviewed journals,
You may expect that data from the sources listed under point 1
above will automatically comply with the provisions of he
Standard, but you should always check to make sure, BS 6399-2
cites the BRE Designer's Guide" as a source and thi
includes data
for domes, hyperboloid roofs and skew-hipped roofs deemed tou
rare for inclusion, While most other British Standards on wind loads
are more likely to take data from BS 6399.
, some will include
additional information, such as the lattice frame and truss
information in BS 8100.
Research data in peer-reviewed journals wi
always need to be
assessed for compatibility with BS 6399-2 and for compliance with
the provisions for wind tunnel tests given in Annex A (see below). It
would be safest to seek expert guidance.
3.1.7 Wind tunnel tests
‘The note to Clause 1.1 Scope recommends wind tunnel tests when:
© the form of the
fing is not covered by the data in the
Standard
‘© the form of the building can be changed in response to the
test
* londing data are required
sults to give un optimized design or
more detail than the Standard can
provide.
Annex A. Necessary provisions for wind tunel testing lists the
principles that should be followed for wind tunnel tests on static and
dynamic buildings. Note that Annex A is marked as being normative,
o that these provisions must be met for such tests to be vali.
The provisions of Aner A are given as a set of general principles
‘chat will he understood hy the wind tunnel laboratory, but the
designer has no simple way of knowing whether they have been
met, In selecting a laboratory, the first thing you should do is to ask
how they intend to meet these requirements and ask to be shown the
facilities that will be used in your test. The following four main
o cl he seat lé
rer
omen a
}
30% toa
—Wind—> 4
eos rh
<n afte ET
IT 0 D 0 043350 0
ooo» O
Bm)
3/0
Pig 66, Typical urban wind anne! simulation
1
The wind tunnel. Wind tunnels capable of meeting the pro-
s of Annex A are specially developed for representing
buildings in their environment and are called ‘houndary-
layer wind tunnels". They have a working section of constant
cross-section that is at least six times longer than it is his
order to develop the proper flow characteristics from a
representation of the ground roughness upwind of the site.
Short-section wind tunnels developed for aircraft testing will
not be able to meet the provisions.
The simulation method. This is the physical method of
representing the building in its environment at mudel scale.
The elements you should expect for an urban simulation are
shown in Fig. 66. Check that the following points apply.
@
0)
©
@)
The height of the model of your building is less than
50% of the height ofthe tunnel. Flow simulation quality
is best near the ground and may not be sufficiently good
in the upper half of the wind tunnel.
“the neighbouring huildings are represented on the same
scale as your building. The number of neighbouring
buildings will depend on the size of the turntable and the
scale of the model, but you will require all ofthe imme
ate neighbours and all of their immediate neighbours: that
is, at least two blocks around your building. Any tall
building that may cast a wake on to your building may
nee (0 be separately represented off the turntable,
‘The general uiban roughness is represented for at least
five wind tunnel heights upwind and one wind tunnel
ight downwind of your model.
That the initial turbulence intensity and ground-level
RN cs ee +
Making BS 6399-2 work for you | 143
1. Published design guidance from authoritative sources such as:
@
(b)
©
other British, European and 150 Standards
Building Research Establishment Digests and other
publications
ESDU International Data Skeets (ESDU International,
27 Corsham Street, London N} GUA. Tel: 0171 490
5151).
2. Research papers published in pecr-roviewed journals,
You may expect that data from the sources listed under point 1
above will automatically comply with the provisions of he
Standard, but you should always check to make sure, BS 6399-2
cites the BRE Designer's Guide" as a source and thi
includes data
for domes, hyperboloid roofs and skew-hipped roofs deemed tou
rare for inclusion, While most other British Standards on wind loads
are more likely to take data from BS 6399.
, some will include
additional information, such as the lattice frame and truss
information in BS 8100.
Research data in peer-reviewed journals wi
always need to be
assessed for compatibility with BS 6399-2 and for compliance with
the provisions for wind tunnel tests given in Annex A (see below). It
would be safest to seek expert guidance.
3.1.7 Wind tunnel tests
‘The note to Clause 1.1 Scope recommends wind tunnel tests when:
© the form of the
fing is not covered by the data in the
Standard
‘© the form of the building can be changed in response to the
test
* londing data are required
sults to give un optimized design or
more detail than the Standard can
provide.
Annex A. Necessary provisions for wind tunel testing lists the
principles that should be followed for wind tunnel tests on static and
dynamic buildings. Note that Annex A is marked as being normative,
o that these provisions must be met for such tests to be vali.
The provisions of Aner A are given as a set of general principles
‘chat will he understood hy the wind tunnel laboratory, but the
designer has no simple way of knowing whether they have been
met, In selecting a laboratory, the first thing you should do is to ask
how they intend to meet these requirements and ask to be shown the
facilities that will be used in your test. The following four main
o cl he seat lé
rer
omen a
}
30% toa
—Wind—> 4
eos rh
<n afte ET
IT 0 D 0 043350 0
ooo» O
Bm)
3/0
Pig 66, Typical urban wind anne! simulation
1
The wind tunnel. Wind tunnels capable of meeting the pro-
s of Annex A are specially developed for representing
buildings in their environment and are called ‘houndary-
layer wind tunnels". They have a working section of constant
cross-section that is at least six times longer than it is his
order to develop the proper flow characteristics from a
representation of the ground roughness upwind of the site.
Short-section wind tunnels developed for aircraft testing will
not be able to meet the provisions.
The simulation method. This is the physical method of
representing the building in its environment at mudel scale.
The elements you should expect for an urban simulation are
shown in Fig. 66. Check that the following points apply.
@
0)
©
@)
The height of the model of your building is less than
50% of the height ofthe tunnel. Flow simulation quality
is best near the ground and may not be sufficiently good
in the upper half of the wind tunnel.
“the neighbouring huildings are represented on the same
scale as your building. The number of neighbouring
buildings will depend on the size of the turntable and the
scale of the model, but you will require all ofthe imme
ate neighbours and all of their immediate neighbours: that
is, at least two blocks around your building. Any tall
building that may cast a wake on to your building may
nee (0 be separately represented off the turntable,
‘The general uiban roughness is represented for at least
five wind tunnel heights upwind and one wind tunnel
ight downwind of your model.
That the initial turbulence intensity and ground-level
RN cs ee +
144 1, Wind loading: a practical guide to BS 6399-2
Making BS 6399-2 work for you | 145
Vise ven <i
tunnel area
Lt)
Fig. 67. Wind mime! blockege
‘Tapping square and
Hush with surtaco
Pressure
ransducer
Fig. 68. Tarping=tubing-trarsducer sytem
momentum wall, or equivalent devices, at the upwind
end of the tunnel section,
3. Wind tunnel blockage. This is related to the arca of the wind
tunnel that the building model occupies. Ifthe model is too
large, the wind must squeeze between the model and the
walls of the tunnel, increasing the loads und distorting the
pressure distributions. Knsure that the area of your building
that protrudes above the average level of rooftops, ie. above
Ho, which is ‘seen’ by the wind is less than 59% of the total
tunnel area. This is illustrated in Fig. 67.
Note: A few wind tumels have special provisions to negate the effects of
blockage, allowing larger models to be used.
| | 4. The acquisition method. This controls how the pressures on
the surface of a model are measured. Figure 68 shows a
typical system.
(2) The tapping holes in the surface ul the model must be
square and flush with the surface.
(b) The total length of tubing connecting tap to pressure
transducer must not be longer than about 400mm,
otherwise the smallest gusts will be suppressed.
(©) The tubing should incorporate a restrictor that is
‘optimized to the frequency response of the system, to
prevent acoustic resonance in the tubing.
In practice, there may be a single pressure transducer for each lap.
as shown but, more likely, the transducer will be shared between a
large number of taps using an clectromechanical pressure-scanning
switch
‘The data analysis method must ‘enable the peak wind loads with
the required annual risk of being exceeded to be predicted’, This
may be achieved in 4 number of equivalent ways.
(a) From extreme-value analysis of the peak values (Cook=
Mayne method),
(b) From N highest values (Poterkw method, N=: 100).
(©) By extrapolating the probability distribution (Lawson
method)
However, measuring à single peak value will not be adequate
because its risk of exceedance will nat be known,
Tf the provisions of Annex A are not properly met in all respect
very likely that the results will he sufficiently distorted for them to
bbe unsuitable for application to design. The surest way to be
confident in any wind tunnel test is to commission it from a
laboratory that specializes in this field and has a proven record
3.2_ Full dynamic classification procedure
The full classification procedure in Annex C should be used
whenever there is reason to expect that the building does not have
typical dynamic characteristics. Annex C is also suitable as a poste
design check when good estimates of the fundamental natural
frequency and structural damping are known,
The method is applied in (wo stages: Equation C.3 followed by
Equation C.). Equation C.3 gives a value for the product Ky x Ky.
Annex C describes Kj, as the “building height factor’, but it is never
used separately from the building-type factor, Ky,
To apply Equation C.3 you will necd to know the following details.
M The terrain and building factor for the mean wind speed, Sy
This is Sy in Equation 29 with the peak factor set to #0.
9 The natural frequency of the fundamental mode of vibration,
m (in Hz).
© The diagonal dimension of the whole building, a (in m).
Making BS 6399-2 work for you | 145
Vise ven <i
tunnel area
Lt)
Fig. 67. Wind mime! blockege
‘Tapping square and
Hush with surtaco
Pressure
ransducer
Fig. 68. Tarping=tubing-trarsducer sytem
momentum wall, or equivalent devices, at the upwind
end of the tunnel section,
3. Wind tunnel blockage. This is related to the arca of the wind
tunnel that the building model occupies. Ifthe model is too
large, the wind must squeeze between the model and the
walls of the tunnel, increasing the loads und distorting the
pressure distributions. Knsure that the area of your building
that protrudes above the average level of rooftops, ie. above
Ho, which is ‘seen’ by the wind is less than 59% of the total
tunnel area. This is illustrated in Fig. 67.
Note: A few wind tumels have special provisions to negate the effects of
blockage, allowing larger models to be used.
| | 4. The acquisition method. This controls how the pressures on
the surface of a model are measured. Figure 68 shows a
typical system.
(2) The tapping holes in the surface ul the model must be
square and flush with the surface.
(b) The total length of tubing connecting tap to pressure
transducer must not be longer than about 400mm,
otherwise the smallest gusts will be suppressed.
(©) The tubing should incorporate a restrictor that is
‘optimized to the frequency response of the system, to
prevent acoustic resonance in the tubing.
In practice, there may be a single pressure transducer for each lap.
as shown but, more likely, the transducer will be shared between a
large number of taps using an clectromechanical pressure-scanning
switch
‘The data analysis method must ‘enable the peak wind loads with
the required annual risk of being exceeded to be predicted’, This
may be achieved in 4 number of equivalent ways.
(a) From extreme-value analysis of the peak values (Cook=
Mayne method),
(b) From N highest values (Poterkw method, N=: 100).
(©) By extrapolating the probability distribution (Lawson
method)
However, measuring à single peak value will not be adequate
because its risk of exceedance will nat be known,
Tf the provisions of Annex A are not properly met in all respect
very likely that the results will he sufficiently distorted for them to
bbe unsuitable for application to design. The surest way to be
confident in any wind tunnel test is to commission it from a
laboratory that specializes in this field and has a proven record
3.2_ Full dynamic classification procedure
The full classification procedure in Annex C should be used
whenever there is reason to expect that the building does not have
typical dynamic characteristics. Annex C is also suitable as a poste
design check when good estimates of the fundamental natural
frequency and structural damping are known,
The method is applied in (wo stages: Equation C.3 followed by
Equation C.). Equation C.3 gives a value for the product Ky x Ky.
Annex C describes Kj, as the “building height factor’, but it is never
used separately from the building-type factor, Ky,
To apply Equation C.3 you will necd to know the following details.
M The terrain and building factor for the mean wind speed, Sy
This is Sy in Equation 29 with the peak factor set to #0.
9 The natural frequency of the fundamental mode of vibration,
m (in Hz).
© The diagonal dimension of the whole building, a (in m).
146 | Wind loading: a practical guide to BS 6399-2
Making BS 6399-2 work for you | 147
DD The structural damping as a fraction of critical, £ This
depends on the material and structural form.
O The site wind speed, Y, (in m/s).
The correction factor for terrain roughness, Ki. Instead of the
simple criteria in Annex C, refer to Figure 4 for the size
effect factor C, and:
(if ine “A” applies, then Ki = 1-33 (value at the cous!) oF
O if line “B' applies, then K,=1 (value inland in open
county), but
© if line °C? applies, then Ki
Note: When using the "Wind Loading Ready-Reckoner (see 3.1.4 Desi
aids), ‘A’ ‘B’ or C'is indicated in Tables C.1 to C.13.
Equation C.1 gives the required value of dynamic augmentation
factor C,. To apply Equation C.J you need to know the following
parameters.
M The product Kj, x Ky determined from Fgnation C..
© The gust factor appropriate to the building size and terrain,
given by:
O S¿= lus, for sites in open country
© Sg= L+g,5,7, for sites in towns, obtained from Tables
22, 23 and 24.
he resulting value of dynamic augmentation factor Cy is used to
determine whether BS 6399-2 is applicable and to give equivalent
dynamic loads, exactly as if it had been determined by the signpost
procedure of $/.6.
+75 (value in town terrain).
3.3. Effective wind speed and dynamic pressure
3.3.1 Available options
The options for assessing the effective wind speed and dynamic
pressure were introduced in 2.1.3 and described in detail in 2.1.5 as
Option 1, 2 or 3. The ‘Application’ sections of the commentary Chapte
2 take you through the steps for implementation, Each of these options is
demonstrated later in 42, The main question now is, Which is the best
tone to choose in any particular circumstance?
33.1.1 Option I— Irrespective of direction
Option 1 represents the single worst case irrespective of direction. This
is directly equivalent to CP 3-V-2 and is useful for initial back-of:
envelope ealeulations. It isthe most conservative choice — particularly on
east-facing coasts or on the eastern boundary of a town-- as we shall see
below. However, if you can justify the structure using this simple
approach, no further calculation is necessary.
3.3.1.2 Option2— Orthogonal load cases
Option 2 is directly equivalent to using CP 3-V-2 with direction factors.
It is the option recommended by BRE Digest 436° as being the optimum
balance hetween conservatism and complexity for use with hand
calculations, and itis used for the worked examples in the Digest.
We will sec in 4.25 that Option 2 still has the potential to be quite
conservative on the boundaries of towns. Inside towns, the degree of
conservatism depends on whether you obtain the terrain and building
factor Sy from Table 4 of the standard method or from Equarion 29 of the
directiomal method.
33.1.3 Option 3—Directional method
Option 3 removes all conservatism hy determining the wind speeds in
each 30°-wide sector using the exposure parameters for that sector, On
average this option gives dynamic pressures that are 14% less than those
resulting from Option 2, but this varies from site to site in the range
0% ~30%, depending on the expos
Option 3 is over-complex for typical hand calculations, but if you are
having difficulty just ‘ular component, you may wish to
choose this approach to check a critical wind direction. With ts (welve
repeated calculations, this choice is suitable fur spreadsheet-based
calculations, although you will still need to determine the twelve
distances to sea and in town, However, if you have the BREVe design
aid, there is no reason to use the more conservative options because
BREVe removes all the effort required for Option 3 and avoids the
possibility of calculation errors.
33.14 Hybrid methods
Clause 3.4.2 permits the ‘hybrid’ combination of directional wind
speeds with standard pressure coefficients to recover most of the
‘conservatism of the standard method us follows,
© Determine the dynamic pressure for euch uf the twelve 30"-wide
sectors of wind direction p.
I Select the most oneroux value of dynamic pressure for each of the
ranges —45"< 0 < 445" either side of the orthogonal cuses.
Making BS 6399-2 work for you | 147
DD The structural damping as a fraction of critical, £ This
depends on the material and structural form.
O The site wind speed, Y, (in m/s).
The correction factor for terrain roughness, Ki. Instead of the
simple criteria in Annex C, refer to Figure 4 for the size
effect factor C, and:
(if ine “A” applies, then Ki = 1-33 (value at the cous!) oF
O if line “B' applies, then K,=1 (value inland in open
county), but
© if line °C? applies, then Ki
Note: When using the "Wind Loading Ready-Reckoner (see 3.1.4 Desi
aids), ‘A’ ‘B’ or C'is indicated in Tables C.1 to C.13.
Equation C.1 gives the required value of dynamic augmentation
factor C,. To apply Equation C.J you need to know the following
parameters.
M The product Kj, x Ky determined from Fgnation C..
© The gust factor appropriate to the building size and terrain,
given by:
O S¿= lus, for sites in open country
© Sg= L+g,5,7, for sites in towns, obtained from Tables
22, 23 and 24.
he resulting value of dynamic augmentation factor Cy is used to
determine whether BS 6399-2 is applicable and to give equivalent
dynamic loads, exactly as if it had been determined by the signpost
procedure of $/.6.
+75 (value in town terrain).
3.3. Effective wind speed and dynamic pressure
3.3.1 Available options
The options for assessing the effective wind speed and dynamic
pressure were introduced in 2.1.3 and described in detail in 2.1.5 as
Option 1, 2 or 3. The ‘Application’ sections of the commentary Chapte
2 take you through the steps for implementation, Each of these options is
demonstrated later in 42, The main question now is, Which is the best
tone to choose in any particular circumstance?
33.1.1 Option I— Irrespective of direction
Option 1 represents the single worst case irrespective of direction. This
is directly equivalent to CP 3-V-2 and is useful for initial back-of:
envelope ealeulations. It isthe most conservative choice — particularly on
east-facing coasts or on the eastern boundary of a town-- as we shall see
below. However, if you can justify the structure using this simple
approach, no further calculation is necessary.
3.3.1.2 Option2— Orthogonal load cases
Option 2 is directly equivalent to using CP 3-V-2 with direction factors.
It is the option recommended by BRE Digest 436° as being the optimum
balance hetween conservatism and complexity for use with hand
calculations, and itis used for the worked examples in the Digest.
We will sec in 4.25 that Option 2 still has the potential to be quite
conservative on the boundaries of towns. Inside towns, the degree of
conservatism depends on whether you obtain the terrain and building
factor Sy from Table 4 of the standard method or from Equarion 29 of the
directiomal method.
33.1.3 Option 3—Directional method
Option 3 removes all conservatism hy determining the wind speeds in
each 30°-wide sector using the exposure parameters for that sector, On
average this option gives dynamic pressures that are 14% less than those
resulting from Option 2, but this varies from site to site in the range
0% ~30%, depending on the expos
Option 3 is over-complex for typical hand calculations, but if you are
having difficulty just ‘ular component, you may wish to
choose this approach to check a critical wind direction. With ts (welve
repeated calculations, this choice is suitable fur spreadsheet-based
calculations, although you will still need to determine the twelve
distances to sea and in town, However, if you have the BREVe design
aid, there is no reason to use the more conservative options because
BREVe removes all the effort required for Option 3 and avoids the
possibility of calculation errors.
33.14 Hybrid methods
Clause 3.4.2 permits the ‘hybrid’ combination of directional wind
speeds with standard pressure coefficients to recover most of the
‘conservatism of the standard method us follows,
© Determine the dynamic pressure for euch uf the twelve 30"-wide
sectors of wind direction p.
I Select the most oneroux value of dynamic pressure for each of the
ranges —45"< 0 < 445" either side of the orthogonal cuses.
148 | Wind loading: a practica! guide to RS 6399-2
Making BS 6399-2 work for you | 149
O Apply these values with the standard method pressure coefficients
for each orthogonal case.
For sites in towns, the simplest and most effective hybrid method is to
replace the terrain and building factor Si, of the standard method in Table
4(§2.2.3) with the values given by Equation 29 in the ditectional method
(§3.2.3.2.3). This hybrid approach works with each of the three Options,
removing all the conservatism that occurs because of the step change at
2km into the town,
3.3.2 Sites near the coast, estuaries or inland water
The effect of the direction factor Sa is stronger than the effect of
distance to sea on the terrain and building factor Sy. Example 16
demanstrates the most extreme cases that can vecur for onshore and
offshore winds at a site on the coastline in open country. The most
onerous combination of Sy and 5, occurs when the onshore case includes
4 =240?, where Sa=1, Le. on a west-facing coast, case (a). On an cast
facing coast, the worst combination is not the onshore case (c), as might
be expected, but always occurs for the offshore case (b), even when the
distance lo the western coast is greater than 100 km. It follows (hut if you
wish to use Optiont, irrespetive of direction, and the site is in open
‘country, then it is always safe to use the closest distance to sea in the
range 210° < ip < 270 as shown in Fig. 69.
Example 16. Onshore and offshore winds for coastal sites in
country terrain
Wid > wind > Wind —
nennen AA Ein
menu Ce > Cou Fe Se
Dito soa ken Destarcato sea 1098 Dance ta sak
5,2100, 5.0178 Set 5,188, 5.7074 Sara
On Sa cire
(9) Westcoast ste. (D) East coast site, (6) East coast site,
‘onshore wind ofshore wind ‘onshore wind
270" y Site
Fig. 69. Distances to sea required for Option in open country
Fig. 70, Distance to sea ln estuaries
It can be difficult to determine the effective distance to sea for sites
close to estuaries. The simplest and safest advice is to use the closest
distance lo any water or open country respeetively, but this will often be
very conservative. The ‘I km mule’ in §/.7.2 for inland lakes (bigger than
Tkm and closer than (km), works reasonably well, hut Fig. 70 gives
better rules for estuaries.
O For site A, where the distance to the estuary a and the extent of
land after the estuary € are both greater than the width of the
estuary b, the fetch is taken to the Far coast.
Making BS 6399-2 work for you | 149
O Apply these values with the standard method pressure coefficients
for each orthogonal case.
For sites in towns, the simplest and most effective hybrid method is to
replace the terrain and building factor Si, of the standard method in Table
4(§2.2.3) with the values given by Equation 29 in the ditectional method
(§3.2.3.2.3). This hybrid approach works with each of the three Options,
removing all the conservatism that occurs because of the step change at
2km into the town,
3.3.2 Sites near the coast, estuaries or inland water
The effect of the direction factor Sa is stronger than the effect of
distance to sea on the terrain and building factor Sy. Example 16
demanstrates the most extreme cases that can vecur for onshore and
offshore winds at a site on the coastline in open country. The most
onerous combination of Sy and 5, occurs when the onshore case includes
4 =240?, where Sa=1, Le. on a west-facing coast, case (a). On an cast
facing coast, the worst combination is not the onshore case (c), as might
be expected, but always occurs for the offshore case (b), even when the
distance lo the western coast is greater than 100 km. It follows (hut if you
wish to use Optiont, irrespetive of direction, and the site is in open
‘country, then it is always safe to use the closest distance to sea in the
range 210° < ip < 270 as shown in Fig. 69.
Example 16. Onshore and offshore winds for coastal sites in
country terrain
Wid > wind > Wind —
nennen AA Ein
menu Ce > Cou Fe Se
Dito soa ken Destarcato sea 1098 Dance ta sak
5,2100, 5.0178 Set 5,188, 5.7074 Sara
On Sa cire
(9) Westcoast ste. (D) East coast site, (6) East coast site,
‘onshore wind ofshore wind ‘onshore wind
270" y Site
Fig. 69. Distances to sea required for Option in open country
Fig. 70, Distance to sea ln estuaries
It can be difficult to determine the effective distance to sea for sites
close to estuaries. The simplest and safest advice is to use the closest
distance lo any water or open country respeetively, but this will often be
very conservative. The ‘I km mule’ in §/.7.2 for inland lakes (bigger than
Tkm and closer than (km), works reasonably well, hut Fig. 70 gives
better rules for estuaries.
O For site A, where the distance to the estuary a and the extent of
land after the estuary € are both greater than the width of the
estuary b, the fetch is taken to the Far coast.
150 | Wind loading: « practical guide to BS 6399-2
Making BS 6399-2 work for yeu 1 151
© Forsite B where a < band site C where a < ¢, the fetch is taken to
the near coast.
‘These rules are summarized as:
a+b+e ita>bande>b
% (26)
xl a ia<borc<b
‘The rules for estuaries may also be applied to the edges of towns
irregular boundaries. In both cases they introduce an unwanted step
change. This step may be avoided by using the multiple fetch exposure
method inthe BRE Designer’ Guide that simplemente y the BREVE
program (see 3.14).
3.3.3 Sites in towns or permanent woodland
333.1 Distance in town
The effect of the direction factor Sa is also stronger than the effeet of
distance in town on the terrain and building factor Sy. Example 17
demonstrates the most extreme cases that can occur for onshore and
offshore winds at a site on the boundary of an inland town.
Example 17. ‘On-town’ and ‘off-town’ winds for sites on the
boundary of inland towns
Wind > Wind > < Mod —
NN CLO BT QO GT som
mn RA Ca
mener nee Gare
Rene SEM er
moque | Meme Aa
ran ES az
Sonnen Mgmt (Sinnen
ne EEE ONE
BEE RE EEE
As expected, the most onerous combination of Sy and Sy vccurs whi
the ‘on-town’ case includes 2 = 240, Le. on the western boundary of the
town, case (a).
On the eastern boundary, the worst combination always oecurs for the
‘offutown’ case (b). This occurs even when the neighbours give the
min elfective height He=0-4H, for the offtown case
forthe on-town case (e). It follows that i is always safe
to use the shortest distance in town in the range 210°< yp <270° when
using Option 1, irrespective of direction, at sites in inland towns.
3.3.3.2 Coastal towns
The situation for coastal towns is a littke more complex. This is
illustrated in Example 18. The offshore case for a building on the seafront
(4) is identical to the off-town case (b) of Example 17 hecause the ground
roughness downwind of the site has no effect. But the onshore case for a
10 m-high building on the seafront (h) is more onerous because the effects
of distance to sea, distance in town and effective height, acting together
on Sy, overcome the reduced directional factor Sy. However, for buildings
Further than 100 m inland (e), the reduced effective height makes this case
loss onerous again. I follows thatthe rule for inland towns also applies to
‘eoastal towns except for buildings on the exposed seafront.
Example 18. Onshore and offshore
— Wind Wind —
ala ATA ¡A
a Tone See tome Sea
Dance to aa > 14m ian sea m Sim
"Dao tn = Su Bitar na Qu Dane ion = io
5.2074, 5.- 170 Suan
ES Se
(@) Site on seatront of east (b) Ste on sealront of eust (<) Sto 100m inside cost
cost town, ofshore, ‘coast town, onshore, coast town, onshare,
Fon tom" wind "orion" wind
We have already noted in 24.3.3 that the shelter provided by
neighbouring buildings reduces loads by a maximum of 30%, This is
discussed in more detail below. ‘the benefit of this shelter reduces wi
increasing height above ground as the displacement height H becomes a
smaller proportion of the reference height H,. Figure 71 compares the
offshore profiles ofS, for east-coast towns of various sizes with the onshore
profiles at the seafront and at 100m inland, assuming that Hy — 10m. We
can see that the onshore case for Ihe seafront is the most onerous only For
reference heights below about H,= 15m.
Making BS 6399-2 work for yeu 1 151
© Forsite B where a < band site C where a < ¢, the fetch is taken to
the near coast.
‘These rules are summarized as:
a+b+e ita>bande>b
% (26)
xl a ia<borc<b
‘The rules for estuaries may also be applied to the edges of towns
irregular boundaries. In both cases they introduce an unwanted step
change. This step may be avoided by using the multiple fetch exposure
method inthe BRE Designer’ Guide that simplemente y the BREVE
program (see 3.14).
3.3.3 Sites in towns or permanent woodland
333.1 Distance in town
The effect of the direction factor Sa is also stronger than the effeet of
distance in town on the terrain and building factor Sy. Example 17
demonstrates the most extreme cases that can occur for onshore and
offshore winds at a site on the boundary of an inland town.
Example 17. ‘On-town’ and ‘off-town’ winds for sites on the
boundary of inland towns
Wind > Wind > < Mod —
NN CLO BT QO GT som
mn RA Ca
mener nee Gare
Rene SEM er
moque | Meme Aa
ran ES az
Sonnen Mgmt (Sinnen
ne EEE ONE
BEE RE EEE
As expected, the most onerous combination of Sy and Sy vccurs whi
the ‘on-town’ case includes 2 = 240, Le. on the western boundary of the
town, case (a).
On the eastern boundary, the worst combination always oecurs for the
‘offutown’ case (b). This occurs even when the neighbours give the
min elfective height He=0-4H, for the offtown case
forthe on-town case (e). It follows that i is always safe
to use the shortest distance in town in the range 210°< yp <270° when
using Option 1, irrespective of direction, at sites in inland towns.
3.3.3.2 Coastal towns
The situation for coastal towns is a littke more complex. This is
illustrated in Example 18. The offshore case for a building on the seafront
(4) is identical to the off-town case (b) of Example 17 hecause the ground
roughness downwind of the site has no effect. But the onshore case for a
10 m-high building on the seafront (h) is more onerous because the effects
of distance to sea, distance in town and effective height, acting together
on Sy, overcome the reduced directional factor Sy. However, for buildings
Further than 100 m inland (e), the reduced effective height makes this case
loss onerous again. I follows thatthe rule for inland towns also applies to
‘eoastal towns except for buildings on the exposed seafront.
Example 18. Onshore and offshore
— Wind Wind —
ala ATA ¡A
a Tone See tome Sea
Dance to aa > 14m ian sea m Sim
"Dao tn = Su Bitar na Qu Dane ion = io
5.2074, 5.- 170 Suan
ES Se
(@) Site on seatront of east (b) Ste on sealront of eust (<) Sto 100m inside cost
cost town, ofshore, ‘coast town, onshore, coast town, onshare,
Fon tom" wind "orion" wind
We have already noted in 24.3.3 that the shelter provided by
neighbouring buildings reduces loads by a maximum of 30%, This is
discussed in more detail below. ‘the benefit of this shelter reduces wi
increasing height above ground as the displacement height H becomes a
smaller proportion of the reference height H,. Figure 71 compares the
offshore profiles ofS, for east-coast towns of various sizes with the onshore
profiles at the seafront and at 100m inland, assuming that Hy — 10m. We
can see that the onshore case for Ihe seafront is the most onerous only For
reference heights below about H,= 15m.
152 1 Wind loading: a practical guide 10 BS 6399-2
Making BS 6399-2 work for you | 153
pi E
ol 4 |
Es im
1. A
I» 1 | u
Fe = Se
Lo room mie em nay is
el} — [oline ly nae
nee
pres
Rd
SS AS Dee dm m an m
Tu arg, 5
hy oc apis mn fo ie ie
3333 Urban shelter o
‘The maximum 30% shelter provided by neighbouring buildings is
caused by the upward displacement of the wind speed profile. Neither BS
{6399-2 nor CP 3-V-2 include any allowance for (he direct shelter caused
by the wakes of individual buildings. Accordingly, the effective wind
speeds should not be affected by the demolition of any individual
neighbouring building,
“The ground roughness rules of both BS 6399-2 and CP 3-V-2 imply that
urbanization is imeveisible in the long term, and that demolition of large
areas will be follawed hy re-development, Where this is not the ease, the
‘wind loading on buildings around the edge of the demolition zone is likely
to be increased. It would thus be prudent to re-assess the stability of
vulnerable buildings ur seek specialist advice. Similar advice is given in the
‘code for tall buildings adjacent to low-rise buildings (§1.7.3.4).
“The steps for determining the displacement height Ha in 2.4.3.2 require
you to measure the average height and spacing of permanent up
obstacies. You are advised to concentrate only on the rooftops of the
principal buildings and to ignore any ground-level chuter. But what do
you do if there is a distinct step change in the obstruction height?
Compare the definition of displacement height Figure £./ with Fig. 72
showing upwind buildings of two distinct heights. You necd to determine
the displacement height set by both sets of upwind buildings and use the
larger value. In the case shown in Fig, 72, the more distant,
buildings set the displacement height
Fig, 72. Dispacemen high for ro diferent ses of min bitinge
333.4 Permanent woodland
‘Note 1 la §1.7.2 allows you to treat permanent forest and woodland as
town category. The key word here is ‘permanent’. You need to be quite
sure that the woodland will not be clear felled in the Future. Obviously,
you must expect commercial plantations to be felled eventually. ‘The one
exception where it may be safe to account for the direct shelter af non-
permanent woodland is in the design of temporary works or a building
during construction,
33.5 Hybrid options for sites In towns
Table 4, the table of terrain and building factor Si in the standard
method, was compiled by applying the standard value of gust peak factor
50-344 and (opographic increment $,=0 in Equation 29 of the
directional method (Equation 10). The Wind Loading Keady-Reckoner.”
described in 31.44, was compiled in the same way. Tahle 4 is
e for all sites in towns, except for those less than 100m
the town or exactly at 2km inside the town,
Replacing Table 4 with Equation 28 or 29 isa hybrid option permitted
by 83.4.2. Table 4 shows the effect of applying this option for an effective
height of #,— 5 m at an inland town, typical of a 12:5m high building
surrounded by neighbours of the same height.
Table 4 ect of hybrid method at I, Sm for an inland nen (3100 80 fom sea)
Distance in town: 5, kom 5 from Equations 28 or 29 Effect on a,
km Table 4
o 145 OBBRACTES 44%0-192)— 1465 2%
o1 145 038240 H6x(1+3- A) — 1%
1 145 O882x0754x(148-44x0-192163) RI 9%
199 145 OBS2O TIO AAN IAASO192X 1-03) EME 140%,
2 136 03S20030x(113:48x0192x1.63)—1-348 42%
10 6 0882x0701 x(143-44%0192%1-63)= 1-286 115
30 36 OSORIO 19216) LOST 136%
Making BS 6399-2 work for you | 153
pi E
ol 4 |
Es im
1. A
I» 1 | u
Fe = Se
Lo room mie em nay is
el} — [oline ly nae
nee
pres
Rd
SS AS Dee dm m an m
Tu arg, 5
hy oc apis mn fo ie ie
3333 Urban shelter o
‘The maximum 30% shelter provided by neighbouring buildings is
caused by the upward displacement of the wind speed profile. Neither BS
{6399-2 nor CP 3-V-2 include any allowance for (he direct shelter caused
by the wakes of individual buildings. Accordingly, the effective wind
speeds should not be affected by the demolition of any individual
neighbouring building,
“The ground roughness rules of both BS 6399-2 and CP 3-V-2 imply that
urbanization is imeveisible in the long term, and that demolition of large
areas will be follawed hy re-development, Where this is not the ease, the
‘wind loading on buildings around the edge of the demolition zone is likely
to be increased. It would thus be prudent to re-assess the stability of
vulnerable buildings ur seek specialist advice. Similar advice is given in the
‘code for tall buildings adjacent to low-rise buildings (§1.7.3.4).
“The steps for determining the displacement height Ha in 2.4.3.2 require
you to measure the average height and spacing of permanent up
obstacies. You are advised to concentrate only on the rooftops of the
principal buildings and to ignore any ground-level chuter. But what do
you do if there is a distinct step change in the obstruction height?
Compare the definition of displacement height Figure £./ with Fig. 72
showing upwind buildings of two distinct heights. You necd to determine
the displacement height set by both sets of upwind buildings and use the
larger value. In the case shown in Fig, 72, the more distant,
buildings set the displacement height
Fig, 72. Dispacemen high for ro diferent ses of min bitinge
333.4 Permanent woodland
‘Note 1 la §1.7.2 allows you to treat permanent forest and woodland as
town category. The key word here is ‘permanent’. You need to be quite
sure that the woodland will not be clear felled in the Future. Obviously,
you must expect commercial plantations to be felled eventually. ‘The one
exception where it may be safe to account for the direct shelter af non-
permanent woodland is in the design of temporary works or a building
during construction,
33.5 Hybrid options for sites In towns
Table 4, the table of terrain and building factor Si in the standard
method, was compiled by applying the standard value of gust peak factor
50-344 and (opographic increment $,=0 in Equation 29 of the
directional method (Equation 10). The Wind Loading Keady-Reckoner.”
described in 31.44, was compiled in the same way. Tahle 4 is
e for all sites in towns, except for those less than 100m
the town or exactly at 2km inside the town,
Replacing Table 4 with Equation 28 or 29 isa hybrid option permitted
by 83.4.2. Table 4 shows the effect of applying this option for an effective
height of #,— 5 m at an inland town, typical of a 12:5m high building
surrounded by neighbours of the same height.
Table 4 ect of hybrid method at I, Sm for an inland nen (3100 80 fom sea)
Distance in town: 5, kom 5 from Equations 28 or 29 Effect on a,
km Table 4
o 145 OBBRACTES 44%0-192)— 1465 2%
o1 145 038240 H6x(1+3- A) — 1%
1 145 O882x0754x(148-44x0-192163) RI 9%
199 145 OBS2O TIO AAN IAASO192X 1-03) EME 140%,
2 136 03S20030x(113:48x0192x1.63)—1-348 42%
10 6 0882x0701 x(143-44%0192%1-63)= 1-286 115
30 36 OSORIO 19216) LOST 136%
1541 Wind loading: a practical guide to BS 63992
Making BS 6399-2 work for you | 155
3.3.4 Complex topography
Unfortunately, not many real features have the simple shapes
shown in Figure 8, Figure 73 gives guidance on several common
problems.
(a) For sites on the upwind slope or on the flat top. the first
(upwind) erest is the correct crest position and the Feature is
treated as an escarpment. For sites on the downwind slope,
the second (downwind) crest is the correct erest position and
the feature is treated as a hilVridge. The length of the
upwind slope Za; is the horizontal distance between the
points where the line through the slope crosses horizontal
lines through the base and crest, while the height Z is the
PX er fattop
+X or downwind slope
x
tra
ba
mors
a
Li
At! |
da) ttopped ge, wth curved oo andere
i
Wind xq
—
sx
range st
range ol sta
(b) ridge with berm on upwind slope
Fig. 28. Bing rilges of ineonsenien shape
Wind
S
ne range of sto, —
LORS, —_
Wind tango of sto —1
an
m
dos
Fig, 74. General rate for maple te
canespunding vertical distance. Note that the lines du not
necessarily cross exactly at the crest as suggested by Figure
8, but this does not matter because the method depends on
gelting the correct upwind slope.
(b) À berm or other break in the upwind slope will form a new
rest that should be adopted for sites upwind or on the her.
‘On the remainder of the feature the true crest is used and the
line through the upwind slope should be drawn to represent
à reasonable average slope.
“The example of Fig. 7A(b) leads to a more general principle for
multiple hills and ridges. The topography method only works for
on single features, so a multiple feature must be simplified, as
icated by Fig, 74. Retain any the hilVsidge that the site
but extend the profile of this feature down to the level of the has
shown by the dashed curves. This is more important on the uy
slope because this sets the values of slope length Ey and slope height
Z. here may be a range of possible profiles, as shown by (a) and
(b), giving different values of Ly and Z. These differences are
largely offset by the corresponding values of base altitude Ar. Just
pick the one that seems most reasonable. Example 2 demonstrated
much of the advice given above.
The topography method is two-dimensional, so does not predict
the neccleration and changes in wind direction that occur around the
side flanks of hills or funnelling into valleys. Better Ihrec-
dimensional models are available that make use of the digital
terrain models available from Ordnance Survey. If the topography
around your site is complex, you should consider taking specialist
advice or using the service described in 3.1.4.7.
Making BS 6399-2 work for you | 155
3.3.4 Complex topography
Unfortunately, not many real features have the simple shapes
shown in Figure 8, Figure 73 gives guidance on several common
problems.
(a) For sites on the upwind slope or on the flat top. the first
(upwind) erest is the correct crest position and the Feature is
treated as an escarpment. For sites on the downwind slope,
the second (downwind) crest is the correct erest position and
the feature is treated as a hilVridge. The length of the
upwind slope Za; is the horizontal distance between the
points where the line through the slope crosses horizontal
lines through the base and crest, while the height Z is the
PX er fattop
+X or downwind slope
x
tra
ba
mors
a
Li
At! |
da) ttopped ge, wth curved oo andere
i
Wind xq
—
sx
range st
range ol sta
(b) ridge with berm on upwind slope
Fig. 28. Bing rilges of ineonsenien shape
Wind
S
ne range of sto, —
LORS, —_
Wind tango of sto —1
an
m
dos
Fig, 74. General rate for maple te
canespunding vertical distance. Note that the lines du not
necessarily cross exactly at the crest as suggested by Figure
8, but this does not matter because the method depends on
gelting the correct upwind slope.
(b) À berm or other break in the upwind slope will form a new
rest that should be adopted for sites upwind or on the her.
‘On the remainder of the feature the true crest is used and the
line through the upwind slope should be drawn to represent
à reasonable average slope.
“The example of Fig. 7A(b) leads to a more general principle for
multiple hills and ridges. The topography method only works for
on single features, so a multiple feature must be simplified, as
icated by Fig, 74. Retain any the hilVsidge that the site
but extend the profile of this feature down to the level of the has
shown by the dashed curves. This is more important on the uy
slope because this sets the values of slope length Ey and slope height
Z. here may be a range of possible profiles, as shown by (a) and
(b), giving different values of Ly and Z. These differences are
largely offset by the corresponding values of base altitude Ar. Just
pick the one that seems most reasonable. Example 2 demonstrated
much of the advice given above.
The topography method is two-dimensional, so does not predict
the neccleration and changes in wind direction that occur around the
side flanks of hills or funnelling into valleys. Better Ihrec-
dimensional models are available that make use of the digital
terrain models available from Ordnance Survey. If the topography
around your site is complex, you should consider taking specialist
advice or using the service described in 3.1.4.7.
156 | Wind loading: «practical guide to BS 6399-2
Making BS 6399-2 work for you À 157
3.3.5 Temporary buildings
“The concept of “temporary” is not applicable tu any building exposed to
(tc wind For more than one year.
‘The mean recurrence interval is used in Fig. 19 because it was
used previously in CP 3-V-2. I is frequently interpreted as the
“design lifetime”, but this is not correct. Buildings, particularly
houses, are regarded as permanent assets that increase in value and
last forever provided they are adequately maintained. The UK
Building Regulations require that permanent structures to which th
public have access use a minimum value S,— 1:0. Consider a
“temporary” building with a “design lifetime’ of only five years (Sp =
0485) sited next to an identical permanent building (Sp= D. Both
would be subject to identical wind loads. After five years the
temporary building could be torn down and replaced by another and
this process could he repeated forever. In any storm, the risk to
‘ovcupants of the temporary buildings would be ten times greater
than the risk to occupants of the permanent building. In a storm of
the standard design risk, the safety factor provided by a typical
partial factor 7e= 1-4 would be intact for the permanent building,
y
but entirely eroded for the temporary building. This is el
‘unacceptable in terms of public safety.
The concept of “temporary” is applicable to temporary works
when separate protocols are in place to ensure public safety. An
obvious example is falsework to support a building façade after
removal of the roof, since access to the work site will be restricted (0
construction workers and the site may be evacuated in strong winds.
‘Another example is a temporary grandstand or marquee at à public
event that may be cancelled if strong winds are forecast (0 be
imminent. In both these cases the choice of design risk is controlled
by economic considerations because safety of the public is
maintained through actively managed protocols.
it is sometimes assumed that large public buildings, such as
theatres, shopping centres or hotels, should be designed for a
smaller risk because they may contain large numbers of peuple.
"This assumption is erroneous because people are twice as likely to
be in their homes than clsewhere at any given time und even more
likely to be at home when a severe storm is forecast, The type oF
building for which a smaller design risk should be considered are
any that provide vital services in the event of natural disasters, €..
hospitals, police and fire stations, communication centres and the
like,
3.36 Serviceability limit
3.3.6.1 Active safety management
In 23.62, describing application of the probability factor 5, for
temporary buildings with active safety management you were advised to
choose a level of annual risk Q appropriate for the use of the building’
You will probably need to negotiate the value of risk with whoever
approves the safety management protocol. For buildings during
construction and temporary structures for licensed public events, this is
likely to be the Health and Safety Executive. For temporary grandstands
and other similar structures you may find the guidance given in the
Institution of Structural Engineers publication Temporary Demountable
Structures —Guidance on Procurement, Design and Use"? to be helpful,
3.3.62 Passive safety management
In 2.5.5.3, describing the effect of dominant openings, you were
advised to take a suitable value of probability factor for ‘elective’
‘openings, ie. when the opening is normally closed.
Previously, CP 3-V-2 determined the probability factor in terms
of the probability of occurrence P in a given period of exposure 7.
‘The minimum value of probability factor permitted by CP 3-V-2 for
temporary buildings was S3=0-77, corresponding to a probability
of P=0:632 in T=2 years, which corresponds to the modal (most
likely) maximum value in a two-year period.
Figure 19 shows that the BS 6399-2 risk model is slightly more
‘onerous than the previous CP 3-V-2 model for recurrence intervals
shorter than the stendard 50 years. Now the most likely (modal)
value in a two-year period corresponds to 5, = 0-852. However, BS
63992 uses the concept of annual risk Q, which we can relate to its
reciprocal, the mean recurrence interval R= 1/Q. It is therefore
lempling lo take a mean recurrence interval of two years as
corresponding to Q=0-5, because this gives S,=0-776 which ix
closely similar to the minimum value in CP 3-V-2. Unfortunately,
the
in recurrence interval becomes increasingly unreliable for
periods of less than ten years! and is meaningless for
(R= 1/1, gives @
BRE Digest 436° recommends using 5, =0-80 ‘to be compatible with
previous practice’. This is a compromise between the “previous risk" of
.632 in T=2 years and the "previous value" of 0-77. I corresponds
to the modal, or most likely, maximum valuc in a period of between one
and two years, This means that you should only use this value if you
expect the opening to remain closed in typical winter storms. Otherwise
you need to implement an active safely management protocol to ensure
year
1 which requires In0— —o in Equation D.)
Making BS 6399-2 work for you À 157
3.3.5 Temporary buildings
“The concept of “temporary” is not applicable tu any building exposed to
(tc wind For more than one year.
‘The mean recurrence interval is used in Fig. 19 because it was
used previously in CP 3-V-2. I is frequently interpreted as the
“design lifetime”, but this is not correct. Buildings, particularly
houses, are regarded as permanent assets that increase in value and
last forever provided they are adequately maintained. The UK
Building Regulations require that permanent structures to which th
public have access use a minimum value S,— 1:0. Consider a
“temporary” building with a “design lifetime’ of only five years (Sp =
0485) sited next to an identical permanent building (Sp= D. Both
would be subject to identical wind loads. After five years the
temporary building could be torn down and replaced by another and
this process could he repeated forever. In any storm, the risk to
‘ovcupants of the temporary buildings would be ten times greater
than the risk to occupants of the permanent building. In a storm of
the standard design risk, the safety factor provided by a typical
partial factor 7e= 1-4 would be intact for the permanent building,
y
but entirely eroded for the temporary building. This is el
‘unacceptable in terms of public safety.
The concept of “temporary” is applicable to temporary works
when separate protocols are in place to ensure public safety. An
obvious example is falsework to support a building façade after
removal of the roof, since access to the work site will be restricted (0
construction workers and the site may be evacuated in strong winds.
‘Another example is a temporary grandstand or marquee at à public
event that may be cancelled if strong winds are forecast (0 be
imminent. In both these cases the choice of design risk is controlled
by economic considerations because safety of the public is
maintained through actively managed protocols.
it is sometimes assumed that large public buildings, such as
theatres, shopping centres or hotels, should be designed for a
smaller risk because they may contain large numbers of peuple.
"This assumption is erroneous because people are twice as likely to
be in their homes than clsewhere at any given time und even more
likely to be at home when a severe storm is forecast, The type oF
building for which a smaller design risk should be considered are
any that provide vital services in the event of natural disasters, €..
hospitals, police and fire stations, communication centres and the
like,
3.36 Serviceability limit
3.3.6.1 Active safety management
In 23.62, describing application of the probability factor 5, for
temporary buildings with active safety management you were advised to
choose a level of annual risk Q appropriate for the use of the building’
You will probably need to negotiate the value of risk with whoever
approves the safety management protocol. For buildings during
construction and temporary structures for licensed public events, this is
likely to be the Health and Safety Executive. For temporary grandstands
and other similar structures you may find the guidance given in the
Institution of Structural Engineers publication Temporary Demountable
Structures —Guidance on Procurement, Design and Use"? to be helpful,
3.3.62 Passive safety management
In 2.5.5.3, describing the effect of dominant openings, you were
advised to take a suitable value of probability factor for ‘elective’
‘openings, ie. when the opening is normally closed.
Previously, CP 3-V-2 determined the probability factor in terms
of the probability of occurrence P in a given period of exposure 7.
‘The minimum value of probability factor permitted by CP 3-V-2 for
temporary buildings was S3=0-77, corresponding to a probability
of P=0:632 in T=2 years, which corresponds to the modal (most
likely) maximum value in a two-year period.
Figure 19 shows that the BS 6399-2 risk model is slightly more
‘onerous than the previous CP 3-V-2 model for recurrence intervals
shorter than the stendard 50 years. Now the most likely (modal)
value in a two-year period corresponds to 5, = 0-852. However, BS
63992 uses the concept of annual risk Q, which we can relate to its
reciprocal, the mean recurrence interval R= 1/Q. It is therefore
lempling lo take a mean recurrence interval of two years as
corresponding to Q=0-5, because this gives S,=0-776 which ix
closely similar to the minimum value in CP 3-V-2. Unfortunately,
the
in recurrence interval becomes increasingly unreliable for
periods of less than ten years! and is meaningless for
(R= 1/1, gives @
BRE Digest 436° recommends using 5, =0-80 ‘to be compatible with
previous practice’. This is a compromise between the “previous risk" of
.632 in T=2 years and the "previous value" of 0-77. I corresponds
to the modal, or most likely, maximum valuc in a period of between one
and two years, This means that you should only use this value if you
expect the opening to remain closed in typical winter storms. Otherwise
you need to implement an active safely management protocol to ensure
year
1 which requires In0— —o in Equation D.)
158 | Wind loading: a practical guide to BS 6399-2
Making BS 6399-2 work for you | 159
that the opening is always closed when a storm is forecast by the UK
Meteorological Office as being likely Lo cause structural damage.
3.3.63 Non-permanent and direct shelter
‚The one exception where it would be safe to account for non-
permanent shelter or for the direct shelter of a nearby building
the design of temporary works or a building during construction,
in cases where you are certain that the shelter will exist for the
duration of the works. This design will normally be at the
crviccability Tmt, using both the seasonal and probability Factors
$, and Sp. Provided that the sheltering building is at least as tall as
the works to be sheltered, the shelter factor fur free-standing walls in
Figure 27 may be used to determine the reduced overall lols,
3.4 Pressures and loads
3.4.1 Tributary areas across zones
When a tributary area spans across more than one pressure coefficient
zone, NS 6399-2 prediets à step change of loading. You should remember
thatthe steps are not real, but are a simplified representation of a smooth
transition between the zones as shown earlier in Fig. 34. The zones aro
fended to give u good estimate of the overall forces and moments on the
panel
Sometimes this
in be useful. For example, let us assume that the 2
panel shown in Fig. 75(a) is a window hinged on one side edge and
fastened on the other side edge. The pancl lies across twa zones where the
net pressure difference across each zone is denoted by A and B,
respectively, in Pa, The forces on the hinge and the fastening, shown as Pi
and Pin (b), can be determined by taking moments about either edge. So,
for the proportions of A and B zones shown:
(Ax 0-8 x 2) x 1-64 (Bx 122 x 2} x 06
(y (AAA
and
(A068 62) 008 4 (BX 1-22) LA
—
At times, this may be inconvenient for the design of the component. If
the component is conventionally rated for a uniform load, for example in
Pa
{
@
Py
P,
o
Fig. 75. Glaxing panel spaming A and B zonex. A and Pin the text represen the net
pressure difference fin Pa) aurons ench zone
tables of permissible net pressure for glazing pancls of various
dimensions, you may use the pressure averaged over the tributary area.
In the example of Fig. 75, the nct pressure averaged over the whole
window is given by:
Ax08+8x12
E a
‘Whether it is appropriate to Keep the step change in loading or to take
the average over the tributary area depends on the form of the structure:
and the design methods used, The choice is a question of engineering
judgement
3.4.2 Diagonal dimension with load sharing
Guidance on the diagonal dimension a was given in 2.6.2 for
plane components (Fig. 59) and for components loaded by mare
than one face of the building (Fig. 60). But this did not include
advice on load sharing, except to note thal the size effect factor is
reduced by only about 5% when the diagonal dimension is doubled.
‘This means that the degree of load sharing needs to be large 10
justify the effort of assessment,
ess of components since a component must deflect
road in onder In shed some ofits load onto another component.
Consider a portalframed building with lightweight purlins and
Profiled metal sheet cladding. The portal frames are very stiff
Making BS 6399-2 work for you | 159
that the opening is always closed when a storm is forecast by the UK
Meteorological Office as being likely Lo cause structural damage.
3.3.63 Non-permanent and direct shelter
‚The one exception where it would be safe to account for non-
permanent shelter or for the direct shelter of a nearby building
the design of temporary works or a building during construction,
in cases where you are certain that the shelter will exist for the
duration of the works. This design will normally be at the
crviccability Tmt, using both the seasonal and probability Factors
$, and Sp. Provided that the sheltering building is at least as tall as
the works to be sheltered, the shelter factor fur free-standing walls in
Figure 27 may be used to determine the reduced overall lols,
3.4 Pressures and loads
3.4.1 Tributary areas across zones
When a tributary area spans across more than one pressure coefficient
zone, NS 6399-2 prediets à step change of loading. You should remember
thatthe steps are not real, but are a simplified representation of a smooth
transition between the zones as shown earlier in Fig. 34. The zones aro
fended to give u good estimate of the overall forces and moments on the
panel
Sometimes this
in be useful. For example, let us assume that the 2
panel shown in Fig. 75(a) is a window hinged on one side edge and
fastened on the other side edge. The pancl lies across twa zones where the
net pressure difference across each zone is denoted by A and B,
respectively, in Pa, The forces on the hinge and the fastening, shown as Pi
and Pin (b), can be determined by taking moments about either edge. So,
for the proportions of A and B zones shown:
(Ax 0-8 x 2) x 1-64 (Bx 122 x 2} x 06
(y (AAA
and
(A068 62) 008 4 (BX 1-22) LA
—
At times, this may be inconvenient for the design of the component. If
the component is conventionally rated for a uniform load, for example in
Pa
{
@
Py
P,
o
Fig. 75. Glaxing panel spaming A and B zonex. A and Pin the text represen the net
pressure difference fin Pa) aurons ench zone
tables of permissible net pressure for glazing pancls of various
dimensions, you may use the pressure averaged over the tributary area.
In the example of Fig. 75, the nct pressure averaged over the whole
window is given by:
Ax08+8x12
E a
‘Whether it is appropriate to Keep the step change in loading or to take
the average over the tributary area depends on the form of the structure:
and the design methods used, The choice is a question of engineering
judgement
3.4.2 Diagonal dimension with load sharing
Guidance on the diagonal dimension a was given in 2.6.2 for
plane components (Fig. 59) and for components loaded by mare
than one face of the building (Fig. 60). But this did not include
advice on load sharing, except to note thal the size effect factor is
reduced by only about 5% when the diagonal dimension is doubled.
‘This means that the degree of load sharing needs to be large 10
justify the effort of assessment,
ess of components since a component must deflect
road in onder In shed some ofits load onto another component.
Consider a portalframed building with lightweight purlins and
Profiled metal sheet cladding. The portal frames are very stiff
160 | Wind loading: a practical guide to BS 6399-2
Making RS 6399-2 work for you | 161
compared with the purlins, so the purlins are unlikely to share load.
‘past the frames, even if they are structurally continuous. The profile
‘oF the metal sheet cladding gives it significant süffness normal to
the purlins, Le. down the slope of the roof. This meuns that there
vil be significant load sharing between the purlins across each bay.
A similar argument can be used to estimate whether load sharing
will occur in other types of structures.
‘Given that load sharing is likely, he influence function method
may be used to estimate the increased diagonal dimension. This is
illustrated in Fig. 76 for the purlins of the portal-framed building
described above and shown in (a), where we expect load sharing
between adjacent purlins, The method is as follows.
M Apply a nominal uniformly distributed load (UDL) to the
target purlin as shown in (b).
O Determine the deflected shape of the adjacent purlins and
roof sheeting along the line normal to the purlins midway
between frames as in (e).
Determine the size of the rectangle of equal height to the
maximum deflection that has the same area as the deflected
shape, as shown in (d),
D The length of this rectangle gives the extent of load sharing
with the adjacent purlins. Accordingly, the diagonal
dimension is given by
a = (length of rectangle)? + (spacing of frames)!
When load sharing is expected in two orthogonal directions, the
UDL should be applied to the tributary area of the target component,
the deflected shape plotted along bath orthogonal axes. A rectangle
valent (d) is drawn for each orthogonal axis and
length of X rectangle)? + (length of Y rectangle)
Note: I is possible for this method to indicate a diagonal dimension that
is smaller than the diagonal of the tributary area. The larger
diagonal should always be used.
The deflected shape may be calculated or determined trom
‘measurements on a representative structure,
3.43 Internal pressure and dominant openings
343.1 Permeability and porosity
The standard internal pressure coefficients in able 16 are given in
terms of the permeability of the building faces. It is important to
recognize the difference between:
(6) Dafleciod shape of adjacent purins
<—_Enxtent of load sharing — +
Déllected
Equal to detected area
(0) Entont of oad sharing
Fig. 76. Influence funtion method for esimatin local sharing
Making RS 6399-2 work for you | 161
compared with the purlins, so the purlins are unlikely to share load.
‘past the frames, even if they are structurally continuous. The profile
‘oF the metal sheet cladding gives it significant süffness normal to
the purlins, Le. down the slope of the roof. This meuns that there
vil be significant load sharing between the purlins across each bay.
A similar argument can be used to estimate whether load sharing
will occur in other types of structures.
‘Given that load sharing is likely, he influence function method
may be used to estimate the increased diagonal dimension. This is
illustrated in Fig. 76 for the purlins of the portal-framed building
described above and shown in (a), where we expect load sharing
between adjacent purlins, The method is as follows.
M Apply a nominal uniformly distributed load (UDL) to the
target purlin as shown in (b).
O Determine the deflected shape of the adjacent purlins and
roof sheeting along the line normal to the purlins midway
between frames as in (e).
Determine the size of the rectangle of equal height to the
maximum deflection that has the same area as the deflected
shape, as shown in (d),
D The length of this rectangle gives the extent of load sharing
with the adjacent purlins. Accordingly, the diagonal
dimension is given by
a = (length of rectangle)? + (spacing of frames)!
When load sharing is expected in two orthogonal directions, the
UDL should be applied to the tributary area of the target component,
the deflected shape plotted along bath orthogonal axes. A rectangle
valent (d) is drawn for each orthogonal axis and
length of X rectangle)? + (length of Y rectangle)
Note: I is possible for this method to indicate a diagonal dimension that
is smaller than the diagonal of the tributary area. The larger
diagonal should always be used.
The deflected shape may be calculated or determined trom
‘measurements on a representative structure,
3.43 Internal pressure and dominant openings
343.1 Permeability and porosity
The standard internal pressure coefficients in able 16 are given in
terms of the permeability of the building faces. It is important to
recognize the difference between:
(6) Dafleciod shape of adjacent purins
<—_Enxtent of load sharing — +
Déllected
Equal to detected area
(0) Entont of oad sharing
Fig. 76. Influence funtion method for esimatin local sharing
162 | Wind loading: a practical guide to BS 6399-2
Making BS 6399-2 work for you | 163
0.3
TA ge
(6) Equally porous
Fig. 72. Pesmesdili; and porsiry
1. permeability is the open area in a face (in m?)
2. porosity is the ratio of the open area to the total arca.
Figure 77(a) shows a square-plan and a rectangular-plan building that
have equal permeability because the number and sizes of the openings are
the same on every face, leading lo Cu = — 03.
Figure 77(b) shows a square-plan and a rectangular-plan building that
are equally porous. The faces of the square-plan building are also equally
msc cuch face is the same size, again leading to
03. However, the long faces of the rectangular building are
Ace as permeable as the short faces because they are twice the length
and therefore twice the area. In this last case, CpizÉ — 0-3 and the internal
pressure should he determined by the balance of flows as described in
2.5.5. Worked examples of this procedure are given in 4.5.5.
permeable be
3
3.4.3.2 Response time
343.2.) Adiabatic response
We noted in 2.5.5.1.2 that the balance of flow establishes very
quickly, typically faster than the duration of the equivalent static
gust, We can deduce this by considering that atmospheric pressure is
around 100 kPa, so that a typical I KPa change in pressure requires
only 1% of the building volume (0 pass through the openings.
Because the response is so fast, the process is adiabatic, and the
response time is given by the equation:!
MOV
CoA pas
where the density of air pa = 1-225 kg/m”, the ratio of specific heats
% =I-4, the discharge cucificient Cp=0-6 and the atmospheric
pressure Pasas 100 kPa are all constants.
If we use Equation F.J in BS 6399-2 to convert the response time
110 the diagonal dimension a, the value of Vo is eliminated. If we
then take the difference between external and intemal pressure
coefficients as being typically Cyo—Cyi=, this leaves only the
building volume O and the opening arca A as parameters. Equation
27 then simplifies to:
a = 0.0019 OA as)
Figure 78 gives the equivalent diagonal dimension for internal
pressures over a wide range of areas of opening A and volume of
building O.
The BS 6399-2 equations for enclosed buildings, Equation 13,
and for dominant openings, Equation 15, are also shown on the
figure, These use only the building volume © as the parameter, so
that Fig, 78 gives the corresponding opening area A assumed by BS
6399-2 for each volume.
(Gre = Gy)! en
3.4322. Effectof building flexibility
When the building is flexible, as all are to some degree, its
volume will also expand! and contract in response to wind loads, and
Ran.
rs perro Ale!)
Fig. 78. Equivalent diagonal dimension for imernalprecuure
Making BS 6399-2 work for you | 163
0.3
TA ge
(6) Equally porous
Fig. 72. Pesmesdili; and porsiry
1. permeability is the open area in a face (in m?)
2. porosity is the ratio of the open area to the total arca.
Figure 77(a) shows a square-plan and a rectangular-plan building that
have equal permeability because the number and sizes of the openings are
the same on every face, leading lo Cu = — 03.
Figure 77(b) shows a square-plan and a rectangular-plan building that
are equally porous. The faces of the square-plan building are also equally
msc cuch face is the same size, again leading to
03. However, the long faces of the rectangular building are
Ace as permeable as the short faces because they are twice the length
and therefore twice the area. In this last case, CpizÉ — 0-3 and the internal
pressure should he determined by the balance of flows as described in
2.5.5. Worked examples of this procedure are given in 4.5.5.
permeable be
3
3.4.3.2 Response time
343.2.) Adiabatic response
We noted in 2.5.5.1.2 that the balance of flow establishes very
quickly, typically faster than the duration of the equivalent static
gust, We can deduce this by considering that atmospheric pressure is
around 100 kPa, so that a typical I KPa change in pressure requires
only 1% of the building volume (0 pass through the openings.
Because the response is so fast, the process is adiabatic, and the
response time is given by the equation:!
MOV
CoA pas
where the density of air pa = 1-225 kg/m”, the ratio of specific heats
% =I-4, the discharge cucificient Cp=0-6 and the atmospheric
pressure Pasas 100 kPa are all constants.
If we use Equation F.J in BS 6399-2 to convert the response time
110 the diagonal dimension a, the value of Vo is eliminated. If we
then take the difference between external and intemal pressure
coefficients as being typically Cyo—Cyi=, this leaves only the
building volume O and the opening arca A as parameters. Equation
27 then simplifies to:
a = 0.0019 OA as)
Figure 78 gives the equivalent diagonal dimension for internal
pressures over a wide range of areas of opening A and volume of
building O.
The BS 6399-2 equations for enclosed buildings, Equation 13,
and for dominant openings, Equation 15, are also shown on the
figure, These use only the building volume © as the parameter, so
that Fig, 78 gives the corresponding opening area A assumed by BS
6399-2 for each volume.
(Gre = Gy)! en
3.4322. Effectof building flexibility
When the building is flexible, as all are to some degree, its
volume will also expand! and contract in response to wind loads, and
Ran.
rs perro Ale!)
Fig. 78. Equivalent diagonal dimension for imernalprecuure
164 | Wind loading: a practica! guide 1o BS 6399-2
Making BS 6399-2 work for you | 165
all the displaced air must also flow through the openings,
the response time. BS 6399-2 makes no allowance for
but it can be substantial. You may assess the effective volume of a
flexible building by first determining the bulk modulus af the
building fy from:
(29)
where O is the volume of the building and AO is the change in
volume caused by a change in internal pressure Apı. This will be a
very big number, typically in the range from 10*Pa for buildings
with long-span flexible roofs to 10° Pa for small stiff buildings, The
effective volume Our is then given by:
Ou = 0 + à) (30)
where ky = 1-410" Pa is the bulk modulus of air. For long-span
arena-iype buildings, the ratio ky/ky can be as high as 5, Yeading to an
effective volume six times the actual volume.
3.43.3. Dominant openings
34.3.3.1 How small con a dominant opening be?
“This is question bogged by both CP 3-V-2 and BS 6399-2, since both
define dominant openings in terms of the size of the opening relative to
the sum of the other openings. The diagonal dimension representing the
response time is given by Equation 15 and, from this, the area of the
dominant opening is given for any volume by the corresponding Jine on
Fig. 78, If we represent Ihe volume as a cube and the opening as a square,
as shown in Fig. 79, Equation 15 implies a porosity af only 1%, ie, the
characteristic dominant opening is very small
Y
ET
Equation 15, equivalent porosity 1%
Fig, 79. Character duminant opening implied by Equation 15
‘This means that the building volume must be very lange indeed before
there is any benefit uf using Fig. 78 in place of Equation 15. It also means.
that the response time for any dominant opening larger than 1% of the
face area will be faster than that assumed by BS 6399-2, However,
Equation 15 also ensures that the diagonal dimension is never smaller
than the actual diagonal of the opening.
The importance of small dominant openings should not come as a
surprise. Here is the corresponding advice from the 1972 Wind Loading
Handbook
"The fact that a reltively small opening can he dominan aná ta ican be effective in
permiting a soletunial change i intra pressure in a shor te makes imperative
{hat the designers should consider the posiiity of window breakage during a severe
storm. Ordinary accidental breakage of windows is unlikely to coincide wih the peak
‘ind loading contin, but sum damage is Experience has shown (hu most window
breakage during stiong winds oxcars onthe windward face, sometimes asa result of
impact fon wind-honte dels... [is recommended that the possibly of occurrence
of à dominant opening shou be considered inthe ight ofthe intended sireng ofthe
cladding and lang ofthe ding and the permeshility that wil est ekewhere on
the bling fade.
‘This advice is as pertinent now as it was in 1972. A frequent cause of
structural damage to roofs of houses is due to failure of glazed patio doors
and windows on the windward face. It is important to ensure that all loose
hard objects that may be pieked up by the wind are removed from gardens
and patios when waming of an impending severe storm is issued.
Following this simple piece of advice will greatly reduce the risk of
structural damage
343.32 Multiple openings in one face
‘The rules for multiple dominant openings in one face were
described in 2.5.5.3 earlier, but two points deserve further clarification.
lis clear (from 2.5.5.3) that multiple large openings on different
faces do not result in a dominant opening unless one is twice the
total area of all the others, However, the footnote 2) to §2.6.2 states:
“Two or more openings in the same face will contribute to one
effestiv dominant opening equal to the combined arca and a
diagonal dimension equal to that of the largest opening.” This
diagonal dimension may be unnecessarily conservative,
Firstly, when the openings lie in different extemal pressure
coefficient zones, as shown in Fig. 80, the diffe
pressures will cause a flow through the openings in exactly the same
manner as if they were on different faces. In this case, the internat
pressure may be determined from the balance of flows using Equation
18.
Making BS 6399-2 work for you | 165
all the displaced air must also flow through the openings,
the response time. BS 6399-2 makes no allowance for
but it can be substantial. You may assess the effective volume of a
flexible building by first determining the bulk modulus af the
building fy from:
(29)
where O is the volume of the building and AO is the change in
volume caused by a change in internal pressure Apı. This will be a
very big number, typically in the range from 10*Pa for buildings
with long-span flexible roofs to 10° Pa for small stiff buildings, The
effective volume Our is then given by:
Ou = 0 + à) (30)
where ky = 1-410" Pa is the bulk modulus of air. For long-span
arena-iype buildings, the ratio ky/ky can be as high as 5, Yeading to an
effective volume six times the actual volume.
3.43.3. Dominant openings
34.3.3.1 How small con a dominant opening be?
“This is question bogged by both CP 3-V-2 and BS 6399-2, since both
define dominant openings in terms of the size of the opening relative to
the sum of the other openings. The diagonal dimension representing the
response time is given by Equation 15 and, from this, the area of the
dominant opening is given for any volume by the corresponding Jine on
Fig. 78, If we represent Ihe volume as a cube and the opening as a square,
as shown in Fig. 79, Equation 15 implies a porosity af only 1%, ie, the
characteristic dominant opening is very small
Y
ET
Equation 15, equivalent porosity 1%
Fig, 79. Character duminant opening implied by Equation 15
‘This means that the building volume must be very lange indeed before
there is any benefit uf using Fig. 78 in place of Equation 15. It also means.
that the response time for any dominant opening larger than 1% of the
face area will be faster than that assumed by BS 6399-2, However,
Equation 15 also ensures that the diagonal dimension is never smaller
than the actual diagonal of the opening.
The importance of small dominant openings should not come as a
surprise. Here is the corresponding advice from the 1972 Wind Loading
Handbook
"The fact that a reltively small opening can he dominan aná ta ican be effective in
permiting a soletunial change i intra pressure in a shor te makes imperative
{hat the designers should consider the posiiity of window breakage during a severe
storm. Ordinary accidental breakage of windows is unlikely to coincide wih the peak
‘ind loading contin, but sum damage is Experience has shown (hu most window
breakage during stiong winds oxcars onthe windward face, sometimes asa result of
impact fon wind-honte dels... [is recommended that the possibly of occurrence
of à dominant opening shou be considered inthe ight ofthe intended sireng ofthe
cladding and lang ofthe ding and the permeshility that wil est ekewhere on
the bling fade.
‘This advice is as pertinent now as it was in 1972. A frequent cause of
structural damage to roofs of houses is due to failure of glazed patio doors
and windows on the windward face. It is important to ensure that all loose
hard objects that may be pieked up by the wind are removed from gardens
and patios when waming of an impending severe storm is issued.
Following this simple piece of advice will greatly reduce the risk of
structural damage
343.32 Multiple openings in one face
‘The rules for multiple dominant openings in one face were
described in 2.5.5.3 earlier, but two points deserve further clarification.
lis clear (from 2.5.5.3) that multiple large openings on different
faces do not result in a dominant opening unless one is twice the
total area of all the others, However, the footnote 2) to §2.6.2 states:
“Two or more openings in the same face will contribute to one
effestiv dominant opening equal to the combined arca and a
diagonal dimension equal to that of the largest opening.” This
diagonal dimension may be unnecessarily conservative,
Firstly, when the openings lie in different extemal pressure
coefficient zones, as shown in Fig. 80, the diffe
pressures will cause a flow through the openings in exactly the same
manner as if they were on different faces. In this case, the internat
pressure may be determined from the balance of flows using Equation
18.
166 | Wind loading: a practical guide 10 BS 6399-2
Making BS 6399-2 work for you | 167
Fig. M. Tho large epenings in the same face
Secondly, the dingonal dimension must still be the larger of the
(diagonal of the opening and the effective diagonal calculated from
the volume using Equation 15. The diagonal of the largest opening
is conservative and should be used when openings arc on different
Faces. But when the openings are on the same face and are
approximately equal in size, the smallest gust that affects them all
‘equally is given by the diagonal between the furthest corners, as
shown in Fig, 80
3.44 Cladding loads
344.1 Corner zones
‘The critical surface pressure for cladding design is usually the high
‘suction zones that form behind upwind comers and edges —the “A” and
'B? zones in the standard method. BS 6399-2 gives guidance on these
zones for basic shapes and extends this to cover wall zones for complex
shapes, including irregular and inset Faces, and this was described in 2.5.2
to 2.54. However, we need some additional guidance on where high-
suction zones form on raofs in these cases.
Clause 3.3.2.2 describes how the zones in the directional method are
defined from the ‘upwind comer” and illustrates this in Figure 35. We
may assume equivalence in the standard method as the ‘outside corners”
and “inside comer’ shown in Fig, 81. The comer ‘A’ zones coeur only at
the outside comers.
3.442 Roof zones for irregular flush faces and Inset storeys
The effect of irregular faces and inset storeys on wall zones is covered
by BS 6399-2 in hoth standard ($2.4.4) und directional methods
($2.3.1.8), but further guidance is needed on the high-suction zones of
the adjacent roofs.
Figure 82(a) shows the zones on the lower roof next to an irregular
flush face, The A zone appears in both comers because both should be
considered to be ‘outside’ corners. The very high coefficient Cye= ~ 20
‘occurs in the extra E zone because the A zone on the adjacent roof
emhances the pre-existing A zone on the wall,
Fig. 81. “Owuido" and “inside” comera
is
—
LY? me 7
NE N
LL
| lia]
—— mee la
rune Bieten
Fig. 82. Higher annee for rf next to irregular and inst faces
We introduced the concept of ‘reflection’ in 2.5.2 to cover smaller
extensions. Figure 82(b) shows the zones on the lower extension roof.
Now the A zone appears only in the ‘outside’ corner because the other
comer is protected by the adjacent wall. There will still he an E zone on
the wall, but the coefficient will not be as high and itis suggested that you
use the value in the adjacent B zone of the roof.
The difference between (a) and (b) in Fig. #2 is the degree of
protection offered by the adjacent wall. A minimum set-back of 6/10 is
recommended.
Note that Fig, 82(b) also shows that the ‘reflection’ must not make the
breadth of the extension larger than the actual breadth of the building, i
following the principle of the ‘smallest enclosing rectangle’ as noted
earlier in 2.5.2.2.
34.5 Loads on internal partitions
Section 2.5.5.2 recommended C,—0-5 as the net pressure coefficient
seross partitions to be used at the serviceability limit. This-applics to
enclosed buildings without dominant openings.
Making BS 6399-2 work for you | 167
Fig. M. Tho large epenings in the same face
Secondly, the dingonal dimension must still be the larger of the
(diagonal of the opening and the effective diagonal calculated from
the volume using Equation 15. The diagonal of the largest opening
is conservative and should be used when openings arc on different
Faces. But when the openings are on the same face and are
approximately equal in size, the smallest gust that affects them all
‘equally is given by the diagonal between the furthest corners, as
shown in Fig, 80
3.44 Cladding loads
344.1 Corner zones
‘The critical surface pressure for cladding design is usually the high
‘suction zones that form behind upwind comers and edges —the “A” and
'B? zones in the standard method. BS 6399-2 gives guidance on these
zones for basic shapes and extends this to cover wall zones for complex
shapes, including irregular and inset Faces, and this was described in 2.5.2
to 2.54. However, we need some additional guidance on where high-
suction zones form on raofs in these cases.
Clause 3.3.2.2 describes how the zones in the directional method are
defined from the ‘upwind comer” and illustrates this in Figure 35. We
may assume equivalence in the standard method as the ‘outside corners”
and “inside comer’ shown in Fig, 81. The comer ‘A’ zones coeur only at
the outside comers.
3.442 Roof zones for irregular flush faces and Inset storeys
The effect of irregular faces and inset storeys on wall zones is covered
by BS 6399-2 in hoth standard ($2.4.4) und directional methods
($2.3.1.8), but further guidance is needed on the high-suction zones of
the adjacent roofs.
Figure 82(a) shows the zones on the lower roof next to an irregular
flush face, The A zone appears in both comers because both should be
considered to be ‘outside’ corners. The very high coefficient Cye= ~ 20
‘occurs in the extra E zone because the A zone on the adjacent roof
emhances the pre-existing A zone on the wall,
Fig. 81. “Owuido" and “inside” comera
is
—
LY? me 7
NE N
LL
| lia]
—— mee la
rune Bieten
Fig. 82. Higher annee for rf next to irregular and inst faces
We introduced the concept of ‘reflection’ in 2.5.2 to cover smaller
extensions. Figure 82(b) shows the zones on the lower extension roof.
Now the A zone appears only in the ‘outside’ corner because the other
comer is protected by the adjacent wall. There will still he an E zone on
the wall, but the coefficient will not be as high and itis suggested that you
use the value in the adjacent B zone of the roof.
The difference between (a) and (b) in Fig. #2 is the degree of
protection offered by the adjacent wall. A minimum set-back of 6/10 is
recommended.
Note that Fig, 82(b) also shows that the ‘reflection’ must not make the
breadth of the extension larger than the actual breadth of the building, i
following the principle of the ‘smallest enclosing rectangle’ as noted
earlier in 2.5.2.2.
34.5 Loads on internal partitions
Section 2.5.5.2 recommended C,—0-5 as the net pressure coefficient
seross partitions to be used at the serviceability limit. This-applics to
enclosed buildings without dominant openings.
168 1 Wind loading: a practical guide 10 BS 6399-2
Making BS 6399-2 work for you | 169
Nearly all the reported damage to intemal partitions was caused by
unintentional dominant openings that occurred after the extemal envelope
had been breached. In many cases this was due to failure of glazing oc
rolter-shulter doors. Ifthe use of a building is changed so that an elective
dominant opening becomes permanently open, then the intemal partitions
Of the room affected by the opening may need to be strengthened.
In existing buildings, i is usually obvious when significant wind loads
are being carried by internal partitions because internal doors hecome
difficult to open against the pressure of the wind. IF this happens, the
internal partitions are not three times more permeable than the external
envelope, so $2.6./.2 applies
The author has personally witnessed failure of an extemal glazing
panel, which cocurred when an intemal corridor door was opened in a
storm, Opening the door relieved the wind toad on the corridor wall, but
Iransferved it onto the external glazing of the leeward wall. It follows
itis good practice ta close all internal doors during a severe storm because
this shares the wind load between external and intemal walls. This
reduces the risk of failure in the extemal envelope at the expense of
increasing the risk to the internal walls
34.6 Structural loads
3.4.6.1 The overall load equation
“The overall loud equation, Equation 7 (Equation 22) in the standard
method and Equation 22 (Equation 21) in the directional method applies
16 any member that contributes to resisting the overall structural loads, Le.
any member in the main structural frame. However, if that member also
carries local cladding loads into (he main structure, then its ability to resist
these local loads should be assessed using Equation 6, that is without the
0:85 factor and without the dynamic augmentation,
For example, the windward column of a portal frame should be
assessed:
O for its capacity to resist the local lateral loads applied to the
‘column by the sheeting rails, using Equation 6,
and also
© for it capacity to resist the overall forces on the portal acting as a
whole using Equation 7.
Remember that $2.3.1.8 states thatthe frictional forces from Equation
7a are additional to the overall forces calculated from Equation 7 or 22,
so the 0:85 factor and C, do not apply to the friction component,
(a) $2.1.8.7 has no eflect (6) Base shear is increased whan
Pons 1 decreased to 60%
Fig. 83. Effect of $2.37 on base shear
34.6.2 Asymmetry
‘The 2002 amendment vepl
-d a complex rule with a requirement to
make ‘an allowance’ for asymmetry. In practice, hawever, most of the
extra work inthe original asymmetry clause ($2.1.2.7) was avoidable with
‘litle thought. The recommendation for torsion in Note 1 is easy to apply
+ and was discussed earlier in 2.6.4.
Application of Note 2 is shown for the base shear on a duopitch
building in Fig. 83. When the external pressure forces are resolved into
the horizontal direction we have a caleulation of the Form:
Px = Piront watt + Pupwind 1001 — Páownwind root — Prear walt
When the roof pitch is a =15° or steeper, as in (a), we obtain positive
values Of Pion win and Paging mer and negative values uf Pyowasind rot
and Pace wath 80 the sign of each of the component forces is positive. For
example, subtracting the negative (suction) value for the rear wall gives a
positive component. Factoring down any of these positive component
loads to 60% of the full load can only reduce the base shear and §2.1.3.7
Has no effect. In this case there are no beneficia loads.
‘When the roof piteh is less than a =15°, as in (b), we obtain a negative
value OF Popwind root that gives a negative component load which is
beneficial 10 the base shear. Factoring this component load te 60%
increases the buse shear, but factoring any of the other components again
reduces the base shear. This may not be a very good example because the
horizontal component from very low-pitched roofs is small, There may
very well be cases of bracing ties or struts that have little or no load in the
symmetrical load case, but are strongly loaded by the asymmetrical load
case. When there are a number of beneficial loads, the new Note 2
recommends factoring all of them to 60%. The old “complicated” rule
required you to factor only the most beneficial load.
‘The simplest way to apply §2.1.3.7 is as follows,
Lay out the calculation of load as a summation of 1
los from the external pressures on each build
component
face,
Making BS 6399-2 work for you | 169
Nearly all the reported damage to intemal partitions was caused by
unintentional dominant openings that occurred after the extemal envelope
had been breached. In many cases this was due to failure of glazing oc
rolter-shulter doors. Ifthe use of a building is changed so that an elective
dominant opening becomes permanently open, then the intemal partitions
Of the room affected by the opening may need to be strengthened.
In existing buildings, i is usually obvious when significant wind loads
are being carried by internal partitions because internal doors hecome
difficult to open against the pressure of the wind. IF this happens, the
internal partitions are not three times more permeable than the external
envelope, so $2.6./.2 applies
The author has personally witnessed failure of an extemal glazing
panel, which cocurred when an intemal corridor door was opened in a
storm, Opening the door relieved the wind toad on the corridor wall, but
Iransferved it onto the external glazing of the leeward wall. It follows
itis good practice ta close all internal doors during a severe storm because
this shares the wind load between external and intemal walls. This
reduces the risk of failure in the extemal envelope at the expense of
increasing the risk to the internal walls
34.6 Structural loads
3.4.6.1 The overall load equation
“The overall loud equation, Equation 7 (Equation 22) in the standard
method and Equation 22 (Equation 21) in the directional method applies
16 any member that contributes to resisting the overall structural loads, Le.
any member in the main structural frame. However, if that member also
carries local cladding loads into (he main structure, then its ability to resist
these local loads should be assessed using Equation 6, that is without the
0:85 factor and without the dynamic augmentation,
For example, the windward column of a portal frame should be
assessed:
O for its capacity to resist the local lateral loads applied to the
‘column by the sheeting rails, using Equation 6,
and also
© for it capacity to resist the overall forces on the portal acting as a
whole using Equation 7.
Remember that $2.3.1.8 states thatthe frictional forces from Equation
7a are additional to the overall forces calculated from Equation 7 or 22,
so the 0:85 factor and C, do not apply to the friction component,
(a) $2.1.8.7 has no eflect (6) Base shear is increased whan
Pons 1 decreased to 60%
Fig. 83. Effect of $2.37 on base shear
34.6.2 Asymmetry
‘The 2002 amendment vepl
-d a complex rule with a requirement to
make ‘an allowance’ for asymmetry. In practice, hawever, most of the
extra work inthe original asymmetry clause ($2.1.2.7) was avoidable with
‘litle thought. The recommendation for torsion in Note 1 is easy to apply
+ and was discussed earlier in 2.6.4.
Application of Note 2 is shown for the base shear on a duopitch
building in Fig. 83. When the external pressure forces are resolved into
the horizontal direction we have a caleulation of the Form:
Px = Piront watt + Pupwind 1001 — Páownwind root — Prear walt
When the roof pitch is a =15° or steeper, as in (a), we obtain positive
values Of Pion win and Paging mer and negative values uf Pyowasind rot
and Pace wath 80 the sign of each of the component forces is positive. For
example, subtracting the negative (suction) value for the rear wall gives a
positive component. Factoring down any of these positive component
loads to 60% of the full load can only reduce the base shear and §2.1.3.7
Has no effect. In this case there are no beneficia loads.
‘When the roof piteh is less than a =15°, as in (b), we obtain a negative
value OF Popwind root that gives a negative component load which is
beneficial 10 the base shear. Factoring this component load te 60%
increases the buse shear, but factoring any of the other components again
reduces the base shear. This may not be a very good example because the
horizontal component from very low-pitched roofs is small, There may
very well be cases of bracing ties or struts that have little or no load in the
symmetrical load case, but are strongly loaded by the asymmetrical load
case. When there are a number of beneficial loads, the new Note 2
recommends factoring all of them to 60%. The old “complicated” rule
required you to factor only the most beneficial load.
‘The simplest way to apply §2.1.3.7 is as follows,
Lay out the calculation of load as a summation of 1
los from the external pressures on each build
component
face,
170 | Wind loading: a practical guide to BS 6399-2
Making BS 6399-2 work for you | 171
Fig. 34. Pressure coefficients for thre-bay ducpitch rong with pitch eagle 30°
excluding all components of internal pressure that cancel out on
internal surfaces:
Py Pa + Ppt Pot Put Pots. on.
© Find the largest negative component in the sum and factor this
down to 60%, for example: B
Pro +41-44119-12.6x0:648:3-98-62KN
This reduces the asymmetty check to a single calculation, even when
there are a large number of component Toads.
The process is illustrated for total horizontal force on a Ihres-hay 30°-
luopitch building in Fig. 84. The faces of the building are labelled ‘a’ to
4". In terms of the buse shear Px, only the resolved farces on faces d and
have negative values and are beneficial. The d values on the second bay
are greater than the f values on the third bay due to the reduction factors
for multi-bay roofs in Table 12. Accordingly, $2.7.3.7 requires the
coefficients for the upwind pitch of the second bay, Cp= — 09 x 08 —
— 0.72 and C,= —0-7X0-8= -- 0-56, to be factored by 60% to become
Gy= — 0-43 and Cy = — 0-34 respectively. This increases the base shear
slightly. (The base shear is generally used to check the overall stability of
the building and to determine the maximum sway displacement.)
3.5 Additional tips and tricks
3.5.1 Minimizing effort
35.1.1 Sites on significant topography
When topography is indicated as significant by Figure 7, itis necessary
to determine the topographic dimensions, as described in 2.1.6.2, This
involves a great deal of effort in extracting information from maps,
drawing cross-sections through the topography and extracting the
dimensions. These dimensions are used to derive just two key topographic,
parameters: the effective slope 4, for each wind direction and the location
factor s for each wind direction and each height above ground (2.1.6.5).
Cid a Gene ne ds US A RAN
is set by the de in Equation 10 and not by the key topography
Parameters in Equation 1! (Equation 5). It would be most helpful (0 have
a simple indicator to avoid wasting this effort,
‘The effect of topography is largest near the ground at the crest. Here
As= Ar+Z and Z= tLe. Using these relationships, we may
solve Equations 10 and 11 ta give L=1200m as the value of effective
slope at which site altitude and topography have equal influence. If
L¿> 1200m the altitude factor is always set by the site altitude and the
topography dimensions are not needed. L, is given by:
[lo iO
cos esos e
This leads to the following useful tip.
35.12 Si
We recall the advice given in 2.4.3.3 on the minimum effective height
for sites in towns. This leads to the following useful tip.
35.13 The ‘unfactored load’ method
‘The major advantage of using the separate size effect factor ist
allows us to calculate a single set of pressures and loads with C,— 1. We
will call these the ‘unfactored” pressures and loads because the required
value of the size effect factor C has not yet been applied, so they are
AAA a pci à
Making BS 6399-2 work for you | 171
Fig. 34. Pressure coefficients for thre-bay ducpitch rong with pitch eagle 30°
excluding all components of internal pressure that cancel out on
internal surfaces:
Py Pa + Ppt Pot Put Pots. on.
© Find the largest negative component in the sum and factor this
down to 60%, for example: B
Pro +41-44119-12.6x0:648:3-98-62KN
This reduces the asymmetty check to a single calculation, even when
there are a large number of component Toads.
The process is illustrated for total horizontal force on a Ihres-hay 30°-
luopitch building in Fig. 84. The faces of the building are labelled ‘a’ to
4". In terms of the buse shear Px, only the resolved farces on faces d and
have negative values and are beneficial. The d values on the second bay
are greater than the f values on the third bay due to the reduction factors
for multi-bay roofs in Table 12. Accordingly, $2.7.3.7 requires the
coefficients for the upwind pitch of the second bay, Cp= — 09 x 08 —
— 0.72 and C,= —0-7X0-8= -- 0-56, to be factored by 60% to become
Gy= — 0-43 and Cy = — 0-34 respectively. This increases the base shear
slightly. (The base shear is generally used to check the overall stability of
the building and to determine the maximum sway displacement.)
3.5 Additional tips and tricks
3.5.1 Minimizing effort
35.1.1 Sites on significant topography
When topography is indicated as significant by Figure 7, itis necessary
to determine the topographic dimensions, as described in 2.1.6.2, This
involves a great deal of effort in extracting information from maps,
drawing cross-sections through the topography and extracting the
dimensions. These dimensions are used to derive just two key topographic,
parameters: the effective slope 4, for each wind direction and the location
factor s for each wind direction and each height above ground (2.1.6.5).
Cid a Gene ne ds US A RAN
is set by the de in Equation 10 and not by the key topography
Parameters in Equation 1! (Equation 5). It would be most helpful (0 have
a simple indicator to avoid wasting this effort,
‘The effect of topography is largest near the ground at the crest. Here
As= Ar+Z and Z= tLe. Using these relationships, we may
solve Equations 10 and 11 ta give L=1200m as the value of effective
slope at which site altitude and topography have equal influence. If
L¿> 1200m the altitude factor is always set by the site altitude and the
topography dimensions are not needed. L, is given by:
[lo iO
cos esos e
This leads to the following useful tip.
35.12 Si
We recall the advice given in 2.4.3.3 on the minimum effective height
for sites in towns. This leads to the following useful tip.
35.13 The ‘unfactored load’ method
‘The major advantage of using the separate size effect factor ist
allows us to calculate a single set of pressures and loads with C,— 1. We
will call these the ‘unfactored” pressures and loads because the required
value of the size effect factor C has not yet been applied, so they are
AAA a pci à
172 Y Wind loading: a practical guide 10 US 6399-2
Making BS 6399-2 work for you À 173
unfactored pressures and loads represent the maximum loads that would
apply to any component with a diagonal dimension less than «a =$ m. We
may then apply the relevant sic effect factor to these loads with the size
effect factor applicable to each part or element of the building at the end
of the assessment. This is what is meant by the ‘unfactored oad” method.
Wiıy is this an advantage?
+ ‘There are usually many sizes of element in a building project.
© Only one set of unfaciored pressure and lond calculations
required and this is used for all sizes of elements.
+ Ihe unfactored pressures may be calculated before the sizcs of
individual elements are known.
+ The designer can provide this set of unfactored pressures and loads
to al] sub-contractors.
© The size effect factor is used to determine the effective pressures and
loads for cach component once its diagonal dimension is known.
® This method is robust because the required value of size effect
factor never execeds unity. The design remains safe (but
if the sub-contractor forgets to apply it.
Usually this greatly reduces the amount of calculation required. In
terms of CP 3-V-2, it is equivalent to using Class A for all calculations,
then applying a factor to obtain the Class B and Class © loads.
Th order to encourage its use, the unfactored load method has been used
wsistently in all the design examples of Chapter 4, ax well as BRE
Digest 4362 even when only one size of component is considered.
35.14 Sloping and skewed surfaces ;
Tt is often necessary to resolve the’ forces on sloping mofs into
horizontal and vertical components. Less frequently in the directional
method, it may be necessary to resolve the forces that are skewed to the
ction. Because BS 6399-2 works from
the surface pressure and pressure is a scalar, the forces act normal to the
surface, hence the term “normal pressure’ used throughout this Guide.
It is therefore conventional to convert net surface pressures to loads
normal to the surface, then to resolve these loads into the required axes.
However, since pressure isa scalar, itis equally valid to resolve the sloping,
‘or skewed area into horizontal and vertical components, then to apply the
net surface pressure Lo the resolved area. Consider the monopitch root in
Fig. 85, where the net surface pressure averaged over the root is p.
By resolving forces, we have:
+ The true area of the roof, A =10x10/cos 13°= 102-6 m”
© The net force normal to the roof, P= pA = 102-6pN.
7 mar)
Fig. 85. Resolved horiromal and vertical forces on a monito ro
+ The resolved horizontal force, Py=Psin 13° 23-1 pN.
+ The resolved vertical force, Py = Pos 13°= 100pN,
By resolving area we have
+ The vertical resolved area, Ay = 10x10 = 100m}.
+ The horizontal resolved arca, Ay = Av tan 13°=23-1 m?
© The resolved horizontal force, Py =Ayp N=23-1 pN.
+ The resolved vertical force, Py =AypN— 100pN.
It is generally quicker to apply the pressures to the resolved areas
because roof zones are defined in terms of plan dimensions and henec
plan area Ay. ‘the corresponding area in clevation is obtained from
Ay=Aylana.
3.5.2 When the exact site details are unknown
We have already noted that, in town terrain, provided that the bi
not significantly taller than its neighbours and there is not an open space
in front of the building, the effective height is always set by the lower
limit 44, =0:44f,. If the building spacing is unknown, but is typical of
suburban or urban arcas, lake the spacing to be Xp-= 20m.
s, you may be quoted the previous grow
categories for CP 3-V-2:
1 For Category 1or2, use the ‘country’ ground roughness. BS 6399.
2 assumes that there will be no significant obstructions to give a
roughness
1 Fur Category 3 oF 4, use the ‘town’ ground roughness. In Category
3 the equivalent displacement height is Hy= 10m, but we have
already noted that this category represents a town centre with four-
storey buildings and that Hy=7-5 m is more appropriate for two-
storey housing, As a minimum, you should assess the obstruction
height from the number of storeys, taking the storey height as 3m
(§1.7.3.3), and adding half a storey for pitched roofs. CP 3-V-2
assumed that neighbouring buildings were always close enough for
the fall effect to apply, giving Ha—08 Ho=2-4 x (number of
Making BS 6399-2 work for you À 173
unfactored pressures and loads represent the maximum loads that would
apply to any component with a diagonal dimension less than «a =$ m. We
may then apply the relevant sic effect factor to these loads with the size
effect factor applicable to each part or element of the building at the end
of the assessment. This is what is meant by the ‘unfactored oad” method.
Wiıy is this an advantage?
+ ‘There are usually many sizes of element in a building project.
© Only one set of unfaciored pressure and lond calculations
required and this is used for all sizes of elements.
+ Ihe unfactored pressures may be calculated before the sizcs of
individual elements are known.
+ The designer can provide this set of unfactored pressures and loads
to al] sub-contractors.
© The size effect factor is used to determine the effective pressures and
loads for cach component once its diagonal dimension is known.
® This method is robust because the required value of size effect
factor never execeds unity. The design remains safe (but
if the sub-contractor forgets to apply it.
Usually this greatly reduces the amount of calculation required. In
terms of CP 3-V-2, it is equivalent to using Class A for all calculations,
then applying a factor to obtain the Class B and Class © loads.
Th order to encourage its use, the unfactored load method has been used
wsistently in all the design examples of Chapter 4, ax well as BRE
Digest 4362 even when only one size of component is considered.
35.14 Sloping and skewed surfaces ;
Tt is often necessary to resolve the’ forces on sloping mofs into
horizontal and vertical components. Less frequently in the directional
method, it may be necessary to resolve the forces that are skewed to the
ction. Because BS 6399-2 works from
the surface pressure and pressure is a scalar, the forces act normal to the
surface, hence the term “normal pressure’ used throughout this Guide.
It is therefore conventional to convert net surface pressures to loads
normal to the surface, then to resolve these loads into the required axes.
However, since pressure isa scalar, itis equally valid to resolve the sloping,
‘or skewed area into horizontal and vertical components, then to apply the
net surface pressure Lo the resolved area. Consider the monopitch root in
Fig. 85, where the net surface pressure averaged over the root is p.
By resolving forces, we have:
+ The true area of the roof, A =10x10/cos 13°= 102-6 m”
© The net force normal to the roof, P= pA = 102-6pN.
7 mar)
Fig. 85. Resolved horiromal and vertical forces on a monito ro
+ The resolved horizontal force, Py=Psin 13° 23-1 pN.
+ The resolved vertical force, Py = Pos 13°= 100pN,
By resolving area we have
+ The vertical resolved area, Ay = 10x10 = 100m}.
+ The horizontal resolved arca, Ay = Av tan 13°=23-1 m?
© The resolved horizontal force, Py =Ayp N=23-1 pN.
+ The resolved vertical force, Py =AypN— 100pN.
It is generally quicker to apply the pressures to the resolved areas
because roof zones are defined in terms of plan dimensions and henec
plan area Ay. ‘the corresponding area in clevation is obtained from
Ay=Aylana.
3.5.2 When the exact site details are unknown
We have already noted that, in town terrain, provided that the bi
not significantly taller than its neighbours and there is not an open space
in front of the building, the effective height is always set by the lower
limit 44, =0:44f,. If the building spacing is unknown, but is typical of
suburban or urban arcas, lake the spacing to be Xp-= 20m.
s, you may be quoted the previous grow
categories for CP 3-V-2:
1 For Category 1or2, use the ‘country’ ground roughness. BS 6399.
2 assumes that there will be no significant obstructions to give a
roughness
1 Fur Category 3 oF 4, use the ‘town’ ground roughness. In Category
3 the equivalent displacement height is Hy= 10m, but we have
already noted that this category represents a town centre with four-
storey buildings and that Hy=7-5 m is more appropriate for two-
storey housing, As a minimum, you should assess the obstruction
height from the number of storeys, taking the storey height as 3m
(§1.7.3.3), and adding half a storey for pitched roofs. CP 3-V-2
assumed that neighbouring buildings were always close enough for
the fall effect to apply, giving Ha—08 Ho=2-4 x (number of
174 | Wind loading: a practical guide to BS 6399-2
Making BS 6399-2 work for you | 175
storeys) in metres, This assumption remains reasonable, provided
that there is not a clear space bigger than 20m directly upwind of
the building.
‘The final problem is the value of site alitude Ay. This will nearly
always be marked on the site plans. Taking the highest value in the
locality as found from an OS 1:50000 map will generally be conservative,
but uneconomic in hilly areas. Topography has the potential to almost
double the wind loads that occur on flat ground
3.5.3 When the orientation of the building is unknown
Clause 2.2.2.3 directs you to use a direction factor Sy—1-0 when the
orientation of he building is unknown. This is our Option 1 —Inrespective
of direction, standard method, ie. the most conservative option.
For sites in upon country, we have already seen in 3.3.2 and Fig. 69 that
we only need the distance to sca in the sectors 9 =210*, 240° and 270°,
and dl reduce conservatism when the closest distance lo sea
outside this range.
For sites in towns, we may still he able to eliminate some more
conservatism by determining the worst wind speed in all possible
ircotions using Option 3—Twelve 30%wide sectors, directional method.
Use the standard value of gust peak factor gy=3-44 and topographic
increment S,=0 in Equarion 29 as described in 3.3.35, select the highest
wind speed and use this in conjunction with the size effect factor C, of the
standard method.
However, this requires twelve times more effort if the calculations are
done by hand. ‘There are several compromises belwoen these two
‘extremes, The choice depends on the local exposure of the building,
© If surrounded by buildings of a similar height, so that the effective
height Ho is the same in all directions, then you again need only
the worst combination of SiS, for the three critical sectors
1p =210?, 240° and 270°, Le. only three out of the twelve possible
sectors. This follows from Example 18.
1 Ion the boundary of the town, then calculate the ‘on-town’ case
using the corresponding value of direction factor $, and the value
of Si, for open country, in addition to the three critical sectors
«e =210", 240° and 270°, ic. only four out of the twelve possible
sectors. This follows from Example 17.
Use the most onerous wind speed found in any of these cases.
As you gain experience, you should find that significant conservatism
inated when the site is near an cast-fucing cuast, near the eastern
boundary of a town, or both.
3.5.4 Automating the balance of flow calculation for internal
pressure
‘The iterative process to obtain the balance of flow described in
255.1 is straightforward, as demonstrated in Example 12 and
Example 13, but is tedious. Most spreadsheet programs, such as
Microsoft Excel™ allow automatic iteration, The balance af flows
may be determined in a single line by adding a small fraction of the
net inflow into a room 10 the internal pressure for that room,
creating a ‘circular reference”, The calculation will continue in a
Joop until the net inflow into every room is zero ur the maximum
umber of iterations is reached. If the fraction of the net inflow
added to the internal pressure is 100 large, the calculation will be
unstable and the calculated internal pressure will oscillate between
ever-larger positive and negative values. If this occurs, reduce the
fraction added and increase the maximum number of iterations. The
steps to implement this procedure are demonstrated. lat
Example 37 (in 4.5.5)
3.5.5 Minimizing wind loads
355.1 Changing the external shape
Provided itis considered early enough in the design, it may be possible
to change the shape of u building to minimize the wind loads, I is sensible
to avoid the forms of building that produce high loading, e.g. to substitute
2 duopitch roof for a monopitch roof, ur a hipped roof for a gabled rao. I.
is also sensible to align the building, if possible, so that the most onerous
pressure coefficients are not coincident with the most onerous wind
speeds, eg. the high eaves of a monopitch root should not be on the
Western sid
Given a building form, loads may be reduced by judicious use of local
features, With low buildings, most ofthe flow passes over the roof and the
loads can be minimized by changing the form of the eaves. Table 8 shows
that curved or mansard eaves are more effective than parapets in reducing
the high-suction loads on flat roofs. With tall buildings, most of the flow
passes around the sides and the high suctions in the "A" zone of the side
walls may be reduced by chamfering the comers to exploit the reduction
factors for polygonal buildings in Table 27.
Finally there are some specific aerodynamic features that can be
incorporated, particularly along the eaves, such as ‘vented caves’ or
“vortex gutters”! that can reduce suctions around the periphery of low-
pitched roofs by up to 25%, These should only be adopted after having
taken expert guidance and should be verified by wind tunnel testing
Making BS 6399-2 work for you | 175
storeys) in metres, This assumption remains reasonable, provided
that there is not a clear space bigger than 20m directly upwind of
the building.
‘The final problem is the value of site alitude Ay. This will nearly
always be marked on the site plans. Taking the highest value in the
locality as found from an OS 1:50000 map will generally be conservative,
but uneconomic in hilly areas. Topography has the potential to almost
double the wind loads that occur on flat ground
3.5.3 When the orientation of the building is unknown
Clause 2.2.2.3 directs you to use a direction factor Sy—1-0 when the
orientation of he building is unknown. This is our Option 1 —Inrespective
of direction, standard method, ie. the most conservative option.
For sites in upon country, we have already seen in 3.3.2 and Fig. 69 that
we only need the distance to sca in the sectors 9 =210*, 240° and 270°,
and dl reduce conservatism when the closest distance lo sea
outside this range.
For sites in towns, we may still he able to eliminate some more
conservatism by determining the worst wind speed in all possible
ircotions using Option 3—Twelve 30%wide sectors, directional method.
Use the standard value of gust peak factor gy=3-44 and topographic
increment S,=0 in Equarion 29 as described in 3.3.35, select the highest
wind speed and use this in conjunction with the size effect factor C, of the
standard method.
However, this requires twelve times more effort if the calculations are
done by hand. ‘There are several compromises belwoen these two
‘extremes, The choice depends on the local exposure of the building,
© If surrounded by buildings of a similar height, so that the effective
height Ho is the same in all directions, then you again need only
the worst combination of SiS, for the three critical sectors
1p =210?, 240° and 270°, Le. only three out of the twelve possible
sectors. This follows from Example 18.
1 Ion the boundary of the town, then calculate the ‘on-town’ case
using the corresponding value of direction factor $, and the value
of Si, for open country, in addition to the three critical sectors
«e =210", 240° and 270°, ic. only four out of the twelve possible
sectors. This follows from Example 17.
Use the most onerous wind speed found in any of these cases.
As you gain experience, you should find that significant conservatism
inated when the site is near an cast-fucing cuast, near the eastern
boundary of a town, or both.
3.5.4 Automating the balance of flow calculation for internal
pressure
‘The iterative process to obtain the balance of flow described in
255.1 is straightforward, as demonstrated in Example 12 and
Example 13, but is tedious. Most spreadsheet programs, such as
Microsoft Excel™ allow automatic iteration, The balance af flows
may be determined in a single line by adding a small fraction of the
net inflow into a room 10 the internal pressure for that room,
creating a ‘circular reference”, The calculation will continue in a
Joop until the net inflow into every room is zero ur the maximum
umber of iterations is reached. If the fraction of the net inflow
added to the internal pressure is 100 large, the calculation will be
unstable and the calculated internal pressure will oscillate between
ever-larger positive and negative values. If this occurs, reduce the
fraction added and increase the maximum number of iterations. The
steps to implement this procedure are demonstrated. lat
Example 37 (in 4.5.5)
3.5.5 Minimizing wind loads
355.1 Changing the external shape
Provided itis considered early enough in the design, it may be possible
to change the shape of u building to minimize the wind loads, I is sensible
to avoid the forms of building that produce high loading, e.g. to substitute
2 duopitch roof for a monopitch roof, ur a hipped roof for a gabled rao. I.
is also sensible to align the building, if possible, so that the most onerous
pressure coefficients are not coincident with the most onerous wind
speeds, eg. the high eaves of a monopitch root should not be on the
Western sid
Given a building form, loads may be reduced by judicious use of local
features, With low buildings, most ofthe flow passes over the roof and the
loads can be minimized by changing the form of the eaves. Table 8 shows
that curved or mansard eaves are more effective than parapets in reducing
the high-suction loads on flat roofs. With tall buildings, most of the flow
passes around the sides and the high suctions in the "A" zone of the side
walls may be reduced by chamfering the comers to exploit the reduction
factors for polygonal buildings in Table 27.
Finally there are some specific aerodynamic features that can be
incorporated, particularly along the eaves, such as ‘vented caves’ or
“vortex gutters”! that can reduce suctions around the periphery of low-
pitched roofs by up to 25%, These should only be adopted after having
taken expert guidance and should be verified by wind tunnel testing
176 1 Wind loading: a practical guide to BS 6399-2
355.2. Controlling the internal pressure
Although it makes no difference to the overall loads, the internal
pressure can be used to distribute these loads more cquitahly through the
structure and to minimize the net pressure on cladding, The internal
pressure may be controlled by the provision of vents in specified
locations. This is often done for other reasons, usually to provi
ventilation or suppress condensation, but can be exploited with a litle
thought. For example, roof spaces urc usually cross-ventilated. IF à
duopitch roof is ventilated only through the soffit under the caves, the
internal pressure will vary from Cy, = +02 for wind normal to the eaves,
10 Cy= —03 for wind parallel 0 the eaves, from the values in Table 16.
However, if equal permeability is provided in the soffit along the ven
as well as the eaves, the intemal pressure enefficient will be maintained at
Cy —0 in alt wind directions.
"The internal pressure may be made more negative by providing vents in
the zones of consistently high suction. The line of the ridge is the best
compromise position, The area of the ridge ventitator should exceed the
permeability of the rest of the envelope by a factor of two, in order to
establish a dominant opening. The penalty for making the intemal
pressure more negative is an increase in the net pressure aeruss the
‘windward wall. If this causes the glazing in the windward wall to fail, the
xemnal pressure will become positive and the benefit will become a
Aiabitity.
Internal walls may be used to compartmentalize the internal volume so
thatthe internal pressure is more positive on the windward side and more
negative on the leeward side, relieving the net pressure on the external
cladding, The penalty is that the internal walls need to be designed to take
the resulting wind loads and automatic mechanisms to close the internal
doors must be installed.
4. Worked examples
4.1 Introducing the examples
4.1.1 Roleofthe examples
‘The role of these examples is more than simply to demonstrate the basic
procedures of BS 6399-2. That is more like the role of BRE Digest 4362
These examples are intended to bring together the yuidance and
idual examples of Chapters 2 and 3 in the context of typical
applications and to extend the expertise of the user beyond the basic
procedures.
While the worked examples have been chosen to illustrate as many
aspects as possible within the space available, they cannot cover
everything and each example is not « complete design calculation. For
these reasons, the examples are not suitable as templates to be copied.
ach individual design has unique site and building characteristics that
require you to make the correct choices of clause. The advice in this
Guide is intended to give you the confidence to do this correctly. In fact it
is much harder to make significant errors in applying BS 6399-2 than it
was with CP 3-V-2,
cry effort has been made lu execute the examples in a consistent
fashion, in particular:
® The displacement height Ha of Annex E has been used throughout
to facilitate the determination of multiple values of effective height
He (sce 2.4,3 and Equation 6).
+ The calculations are performed in terms of unfactored pressures
and loads Lo facilitate the design of components of different sizes
and the size effect factor is applied at the final stage.
It is good practice to adopt these two conventions as standard, even in
the eases where only one effective height or one sizeof loaded aca
355.2. Controlling the internal pressure
Although it makes no difference to the overall loads, the internal
pressure can be used to distribute these loads more cquitahly through the
structure and to minimize the net pressure on cladding, The internal
pressure may be controlled by the provision of vents in specified
locations. This is often done for other reasons, usually to provi
ventilation or suppress condensation, but can be exploited with a litle
thought. For example, roof spaces urc usually cross-ventilated. IF à
duopitch roof is ventilated only through the soffit under the caves, the
internal pressure will vary from Cy, = +02 for wind normal to the eaves,
10 Cy= —03 for wind parallel 0 the eaves, from the values in Table 16.
However, if equal permeability is provided in the soffit along the ven
as well as the eaves, the intemal pressure enefficient will be maintained at
Cy —0 in alt wind directions.
"The internal pressure may be made more negative by providing vents in
the zones of consistently high suction. The line of the ridge is the best
compromise position, The area of the ridge ventitator should exceed the
permeability of the rest of the envelope by a factor of two, in order to
establish a dominant opening. The penalty for making the intemal
pressure more negative is an increase in the net pressure aeruss the
‘windward wall. If this causes the glazing in the windward wall to fail, the
xemnal pressure will become positive and the benefit will become a
Aiabitity.
Internal walls may be used to compartmentalize the internal volume so
thatthe internal pressure is more positive on the windward side and more
negative on the leeward side, relieving the net pressure on the external
cladding, The penalty is that the internal walls need to be designed to take
the resulting wind loads and automatic mechanisms to close the internal
doors must be installed.
4. Worked examples
4.1 Introducing the examples
4.1.1 Roleofthe examples
‘The role of these examples is more than simply to demonstrate the basic
procedures of BS 6399-2. That is more like the role of BRE Digest 4362
These examples are intended to bring together the yuidance and
idual examples of Chapters 2 and 3 in the context of typical
applications and to extend the expertise of the user beyond the basic
procedures.
While the worked examples have been chosen to illustrate as many
aspects as possible within the space available, they cannot cover
everything and each example is not « complete design calculation. For
these reasons, the examples are not suitable as templates to be copied.
ach individual design has unique site and building characteristics that
require you to make the correct choices of clause. The advice in this
Guide is intended to give you the confidence to do this correctly. In fact it
is much harder to make significant errors in applying BS 6399-2 than it
was with CP 3-V-2,
cry effort has been made lu execute the examples in a consistent
fashion, in particular:
® The displacement height Ha of Annex E has been used throughout
to facilitate the determination of multiple values of effective height
He (sce 2.4,3 and Equation 6).
+ The calculations are performed in terms of unfactored pressures
and loads Lo facilitate the design of components of different sizes
and the size effect factor is applied at the final stage.
It is good practice to adopt these two conventions as standard, even in
the eases where only one effective height or one sizeof loaded aca
178_| Wind loading: «practical guide to BS 6399-2
Worked examples | 179
Fig. 86. Distance in ton for SK400877
Table 5, Parameers for ste at SKAO0877
Direction: des 0 30 60 90 120 150 180 210 240 270 300 230
Distance to sea: km 200 120 90 98 200 200 200 200 200 200 158 200
Disnee in town: km 9 9 0 0115 15 45 35 95 10095 75
Ostucton height: m 0 0 0 0 8 8 8 8 8 8 8 8
Ohatrcton separation: m 40 4 2 2 20 % 6 0
Note: The particular choice of examples does not imply that the
calculations shown are required for those elements thal are
permitted to he sized using the prescriptive methods given in
Approved Document A to the Building Regulations and in the
“small buildings code" BS 8103-2 and BS 8103-3.
4.1.2. The site at SK 400877
The examples will demonstrate the three options for effective wind
speed and dynamic pressure for a site on the north-eastern boundary of
Sheffield at grid reference SK400877. This site was chosen because
climination of conservatism can be better demonstrated at sites on the
‘eastern side of towns and placing the site on the town boundary allows
the effect of permanent obstructions to he compared with an open
‘exposure.
The site altitude is 55m and the topography is not significant as
defined by Figure 7. The distances from the site to the town boundary
in each 30°-wide sector are illustrated in Fig. 86. These values,
together with the distances to sea, are listed in Table 5. The reader is
invited to confirm these values by reference to the appropriate maps.
For comparison, Fig. 86 also shows the location of the site used in
BRE Digest 4367 which lies in the south-western outskirts of
Sheffield.
We shall need to know the heights and positions of neighbouring.
buildings. These will be assumed, depending on the target building in
each example, so will always be notional values. Table 5 also lists the
values of obstruction height and spacing that we shall use for the first
example building, a timber-framed house. These values cannot be
checked by reference to maps because the site layout is notional and the
buildings do not actually exist
4.13. The example buildings
Example calculations of surface pressures and loads are given for four
buildings:
1. a two-storey timber-framed house with attached garage
2. long-span portal-framed building
3. a five-storey office block
4. a square-section tower on a podium.
4.1.4. Similarities to BRE Digest 436
Readers may note similarities 10 the site and buildings used for the
example calculations in BRE Digest 436 Paris 2 and 3 The examples
given here are alittle more complicated in order to illustrate more of the
‘extensive guidance given in this Guide. For example, our site is also in
Sheffield, but it is on the north-eastern boundary, whereas the
site is inside the city. Our house has an attached garage, the
house docs not, The similarities and differences have been deliberatcly
introduced to benefit the reader. The examples in Digest 436 demonstrate
the basic fundamentals. These cxamples are complementary and will
extend your expertise to intermediate and advanced in your use of the
Standard
42 Effective wind speed and dynamic pressure
42.1 Reference parameters
order lo derive effective wind speeds at the site, we need to know the
heights and spacing of the surrounding buildings as well as the reference
heights of the target building and its orientation. We shall take the 1w0-
storey house as the target for this example and assume the site layout
shown in Fig, 87. The target building is the one at the centre of the
concentric rings. The ridge is orientated at y —45°, the height to the ridge
Worked examples | 179
Fig. 86. Distance in ton for SK400877
Table 5, Parameers for ste at SKAO0877
Direction: des 0 30 60 90 120 150 180 210 240 270 300 230
Distance to sea: km 200 120 90 98 200 200 200 200 200 200 158 200
Disnee in town: km 9 9 0 0115 15 45 35 95 10095 75
Ostucton height: m 0 0 0 0 8 8 8 8 8 8 8 8
Ohatrcton separation: m 40 4 2 2 20 % 6 0
Note: The particular choice of examples does not imply that the
calculations shown are required for those elements thal are
permitted to he sized using the prescriptive methods given in
Approved Document A to the Building Regulations and in the
“small buildings code" BS 8103-2 and BS 8103-3.
4.1.2. The site at SK 400877
The examples will demonstrate the three options for effective wind
speed and dynamic pressure for a site on the north-eastern boundary of
Sheffield at grid reference SK400877. This site was chosen because
climination of conservatism can be better demonstrated at sites on the
‘eastern side of towns and placing the site on the town boundary allows
the effect of permanent obstructions to he compared with an open
‘exposure.
The site altitude is 55m and the topography is not significant as
defined by Figure 7. The distances from the site to the town boundary
in each 30°-wide sector are illustrated in Fig. 86. These values,
together with the distances to sea, are listed in Table 5. The reader is
invited to confirm these values by reference to the appropriate maps.
For comparison, Fig. 86 also shows the location of the site used in
BRE Digest 4367 which lies in the south-western outskirts of
Sheffield.
We shall need to know the heights and positions of neighbouring.
buildings. These will be assumed, depending on the target building in
each example, so will always be notional values. Table 5 also lists the
values of obstruction height and spacing that we shall use for the first
example building, a timber-framed house. These values cannot be
checked by reference to maps because the site layout is notional and the
buildings do not actually exist
4.13. The example buildings
Example calculations of surface pressures and loads are given for four
buildings:
1. a two-storey timber-framed house with attached garage
2. long-span portal-framed building
3. a five-storey office block
4. a square-section tower on a podium.
4.1.4. Similarities to BRE Digest 436
Readers may note similarities 10 the site and buildings used for the
example calculations in BRE Digest 436 Paris 2 and 3 The examples
given here are alittle more complicated in order to illustrate more of the
‘extensive guidance given in this Guide. For example, our site is also in
Sheffield, but it is on the north-eastern boundary, whereas the
site is inside the city. Our house has an attached garage, the
house docs not, The similarities and differences have been deliberatcly
introduced to benefit the reader. The examples in Digest 436 demonstrate
the basic fundamentals. These cxamples are complementary and will
extend your expertise to intermediate and advanced in your use of the
Standard
42 Effective wind speed and dynamic pressure
42.1 Reference parameters
order lo derive effective wind speeds at the site, we need to know the
heights and spacing of the surrounding buildings as well as the reference
heights of the target building and its orientation. We shall take the 1w0-
storey house as the target for this example and assume the site layout
shown in Fig, 87. The target building is the one at the centre of the
concentric rings. The ridge is orientated at y —45°, the height to the ridge
180 | Wind loading: a practical guide to BS 6399-2
Worked examples | 181
Fig, 87. Site layout for house at SK400877
is 8m and the height to eaves is 6-3 m. The obstruction heights and spacing
are given, together with the distances to sea and in town, in Table 5.
Note: These worked examples use the corrected factors in Appendix B. If
‘you work through the examples using the values in the tables of BS
6399-2: 1997, yuu should obrain effective wind speeds within 2%
and dynamic pressures within 4% of the results of these examples.
4.2.2 Option | — Irrespective of direction
The effective wind speed and dynamic pressure for Option 1 is
calculated for the ridge of 8m and the caves height of 6-3m in Example
19. Remember that this option selets the worst case for each parameter
so, on the edge of the town, the worst exposure is open country with no
obstructions. The calculations are shown using:
® full logarithmie interpolation in Table 4, and
+ the Ready-Reckoner tables without interpolation, Le. taking the
next worst value.
Using the Ready-Reckoner without interpolation adds about 2% to the
dynamic pressure. Using the hybrid method of Equation 29 (see 33:38)
gives no advantage in this instance because the distance in town is zero
And the worst exposure for Option 1 is open country.
Example 19. Dynamic pressure by Option 1 —Inrespe
direction
Table 4 with Reads-Reckoner, Notes
logarithmic interpolation
interpolation
Basie wind spew, iz mls 226 26 rom Figure 6
Altitude factor, Sa 105$ 1055
Site wind speed, Y zu 28
Tights Hos o o Worst case, on
Okstriction separation, X: m0 o edge of town
Displacement height, Hy:m 0. 0
Closest distance to sea: km 90 9 East cons is closest,
‘Closest distance in town: km 0 o age of town
Reference eight, Mim 8 8 Ridge Height
fective height, Heim 30 8
Terrain and building factor, Sy 1:5K 160
fective wind speed, Ya m/s 377 a
Dynamic pressure: q; Pa 8691 3850 For roof and
‘able walle
Reference height, Hm 63 Eaves height
Efectivo height, Hm 63
153 154
364 367
su2 8256 For walls with eaves
4.2.3 Option 2—Orthogonol load cases
‘To apply Option 2 fully, the orientation of the building must be known,
e =45° in this example. The orthogonal cases for the house are shown in
Fig, 88 and are denoted by NW, NE, SE and SW. The worst coi
of parameters must be found over the range 445 cither sid
orthogonal direction, ie. from 4? =0° to = 90° for the NE case as shown
by the light hatching.
Note: This example includes four directions in euch urthogonal case
because the orientation af the house lies exactly on a boundary
berucon the 30 wide sectors. À more Ispical orientation would
only include tree ofthe standard directions
Worked examples | 181
Fig, 87. Site layout for house at SK400877
is 8m and the height to eaves is 6-3 m. The obstruction heights and spacing
are given, together with the distances to sea and in town, in Table 5.
Note: These worked examples use the corrected factors in Appendix B. If
‘you work through the examples using the values in the tables of BS
6399-2: 1997, yuu should obrain effective wind speeds within 2%
and dynamic pressures within 4% of the results of these examples.
4.2.2 Option | — Irrespective of direction
The effective wind speed and dynamic pressure for Option 1 is
calculated for the ridge of 8m and the caves height of 6-3m in Example
19. Remember that this option selets the worst case for each parameter
so, on the edge of the town, the worst exposure is open country with no
obstructions. The calculations are shown using:
® full logarithmie interpolation in Table 4, and
+ the Ready-Reckoner tables without interpolation, Le. taking the
next worst value.
Using the Ready-Reckoner without interpolation adds about 2% to the
dynamic pressure. Using the hybrid method of Equation 29 (see 33:38)
gives no advantage in this instance because the distance in town is zero
And the worst exposure for Option 1 is open country.
Example 19. Dynamic pressure by Option 1 —Inrespe
direction
Table 4 with Reads-Reckoner, Notes
logarithmic interpolation
interpolation
Basie wind spew, iz mls 226 26 rom Figure 6
Altitude factor, Sa 105$ 1055
Site wind speed, Y zu 28
Tights Hos o o Worst case, on
Okstriction separation, X: m0 o edge of town
Displacement height, Hy:m 0. 0
Closest distance to sea: km 90 9 East cons is closest,
‘Closest distance in town: km 0 o age of town
Reference eight, Mim 8 8 Ridge Height
fective height, Heim 30 8
Terrain and building factor, Sy 1:5K 160
fective wind speed, Ya m/s 377 a
Dynamic pressure: q; Pa 8691 3850 For roof and
‘able walle
Reference height, Hm 63 Eaves height
Efectivo height, Hm 63
153 154
364 367
su2 8256 For walls with eaves
4.2.3 Option 2—Orthogonol load cases
‘To apply Option 2 fully, the orientation of the building must be known,
e =45° in this example. The orthogonal cases for the house are shown in
Fig, 88 and are denoted by NW, NE, SE and SW. The worst coi
of parameters must be found over the range 445 cither sid
orthogonal direction, ie. from 4? =0° to = 90° for the NE case as shown
by the light hatching.
Note: This example includes four directions in euch urthogonal case
because the orientation af the house lies exactly on a boundary
berucon the 30 wide sectors. À more Ispical orientation would
only include tree ofthe standard directions
182 1 Wind loading: a practical guide 10 RS 6399-2
Worked examples | 183
ANNE
NW 8 RTS
YO SS
do ER
RR
pen WN
Fig, 88. Orthogonal cases for hone
‘The effective wind speed and dynamic pressure for Option 2
calculated for both heights in Example 20. Remember that this option
selects the worst ease for each parameter in the 90"-wide sector. ‘The open
country exposure with no obstructions applics in every case except SW.
The calculations are shown using full logarithmic interpolation in Table 4,
and using the Ready-Reckamer tables interpolation, as in Option 1
above.
Example 20, Dynamic pressure by Option 2— Four orthogonal
load cases
Table 4 with logaitwnic Ready- Reckoner. no
interpotaton Interpolation
Basic wind speed, Va m/s 226 26
Altitude factor. Sy 1055 1055
: NW
agen direction NE SE SW NW NK SR SW
ere dios tac Ss 075 085 100 0M 078 04 10 09
Site wind spoed, Ve ms 1860 2027 2344 2360 18-60 20.27 2386 2480
Obsirction height im 00800080
Obsiruetion separation, Xm O0 2% 0 O0 0 20 0
Displacement height, Hm 0 0 56 D 9 0 6 0
‘Closest distance o sca: km 90 98 20D 158 90 98 200 158
Closest divunse in tower: Km 0 0 45 0 0 0 45 0
Orthogonal direction ME SE SW NW NE SE SW NW
Reference height. Wom = «$8 8 8 ON 8 8
Ellecive height, Him 80 80 32 80 8 8 1 8
Terrain and bing factor, Sa 1-579 1-570 1-222 1-568 1.60 160 192 1:57
Fective wind speed. Vo m/s 294 31-8 291 370 208 324 315 374
Dynamic pressure: qu l'a 528-7 6206 5204 893 544 644. CB 248
Reference height, Hm 6363 63 61 63 63 63 63
Electve height Hem 63 63 25 63 7 7 3 7
‘Terrain and building factor, Sy 1526 1317 1147 1518 157 1:57 129 150
Effective wind speed. Y m/s 284 107 273 357 292 318 108 364
Dynamic pressure: gy: Pa 4935 $79.2 458-1 7814 523 620 582 812
qq > > E >
The design dynamic pressure is less than Option 1 for all the
orthogonal cases. The smallest value occurs for the SW case, even though
this includes $a= 1-0, because of the shelter provided by the city. The
largest value occurs for the NW case because this includes the direc
= 0" that has an open country exposure,
Using the Keady-Reckoner ‘without interpolation adds 17% 10 the
dynamic pressure uf the SW case at Hy = 8m and 27% at H,=-6.3m. This
is becaus
+ th effective height is rounded up and this has the greatest effect on
conservatism when close to the ground
© the 4:5 km distance in town is just short of the next table for Skm in
town, so is rounded down to the lable for 2km in town.
Most of this conservatism may be recovered by using the exact
effective height and interpolating for height and for distance in town
between the tables for 2km in town and $ kin in town
The hybrid method of Equation 29 (see 33.3.5) will give S,=1-191 for
the SW case by allowing for the exact 4-5km distance in town, reducing
the dynamic pressure of the value from Table 4 by 5%. The procedure is
illustrated in the = 180° column of the next example.
4.24 Option 3—Directional method
We do not necessarily need to know the orientation of the building in
order to apply Option 3. We use the directional method to determine the
effective wind speed and dynamic pressure in each 30°-wind sector, using
the datum gust peak factor g = 3-44 of the standard method, as described
in 353,
J we know the orientation of the building, we can select the worst
dynamic pressure +45° either side of each orthogonal direction and apply
Worked examples | 183
ANNE
NW 8 RTS
YO SS
do ER
RR
pen WN
Fig, 88. Orthogonal cases for hone
‘The effective wind speed and dynamic pressure for Option 2
calculated for both heights in Example 20. Remember that this option
selects the worst ease for each parameter in the 90"-wide sector. ‘The open
country exposure with no obstructions applics in every case except SW.
The calculations are shown using full logarithmic interpolation in Table 4,
and using the Ready-Reckamer tables interpolation, as in Option 1
above.
Example 20, Dynamic pressure by Option 2— Four orthogonal
load cases
Table 4 with logaitwnic Ready- Reckoner. no
interpotaton Interpolation
Basic wind speed, Va m/s 226 26
Altitude factor. Sy 1055 1055
: NW
agen direction NE SE SW NW NK SR SW
ere dios tac Ss 075 085 100 0M 078 04 10 09
Site wind spoed, Ve ms 1860 2027 2344 2360 18-60 20.27 2386 2480
Obsirction height im 00800080
Obsiruetion separation, Xm O0 2% 0 O0 0 20 0
Displacement height, Hm 0 0 56 D 9 0 6 0
‘Closest distance o sca: km 90 98 20D 158 90 98 200 158
Closest divunse in tower: Km 0 0 45 0 0 0 45 0
Orthogonal direction ME SE SW NW NE SE SW NW
Reference height. Wom = «$8 8 8 ON 8 8
Ellecive height, Him 80 80 32 80 8 8 1 8
Terrain and bing factor, Sa 1-579 1-570 1-222 1-568 1.60 160 192 1:57
Fective wind speed. Vo m/s 294 31-8 291 370 208 324 315 374
Dynamic pressure: qu l'a 528-7 6206 5204 893 544 644. CB 248
Reference height, Hm 6363 63 61 63 63 63 63
Electve height Hem 63 63 25 63 7 7 3 7
‘Terrain and building factor, Sy 1526 1317 1147 1518 157 1:57 129 150
Effective wind speed. Y m/s 284 107 273 357 292 318 108 364
Dynamic pressure: gy: Pa 4935 $79.2 458-1 7814 523 620 582 812
qq > > E >
The design dynamic pressure is less than Option 1 for all the
orthogonal cases. The smallest value occurs for the SW case, even though
this includes $a= 1-0, because of the shelter provided by the city. The
largest value occurs for the NW case because this includes the direc
= 0" that has an open country exposure,
Using the Keady-Reckoner ‘without interpolation adds 17% 10 the
dynamic pressure uf the SW case at Hy = 8m and 27% at H,=-6.3m. This
is becaus
+ th effective height is rounded up and this has the greatest effect on
conservatism when close to the ground
© the 4:5 km distance in town is just short of the next table for Skm in
town, so is rounded down to the lable for 2km in town.
Most of this conservatism may be recovered by using the exact
effective height and interpolating for height and for distance in town
between the tables for 2km in town and $ kin in town
The hybrid method of Equation 29 (see 33.3.5) will give S,=1-191 for
the SW case by allowing for the exact 4-5km distance in town, reducing
the dynamic pressure of the value from Table 4 by 5%. The procedure is
illustrated in the = 180° column of the next example.
4.24 Option 3—Directional method
We do not necessarily need to know the orientation of the building in
order to apply Option 3. We use the directional method to determine the
effective wind speed and dynamic pressure in each 30°-wind sector, using
the datum gust peak factor g = 3-44 of the standard method, as described
in 353,
J we know the orientation of the building, we can select the worst
dynamic pressure +45° either side of each orthogonal direction and apply
1841 Wind loading: a practical guide to BS 6399.2
Worked examples | 185
this tothe respective orthogonal case. If we do not know the orientation of
the building, we simply take the largest value of dynamic pressure found.
Whether this largest value is significantly lower than the value from
Option 1 will depend on the relative exposure of the site by direction.
Sites newr a westfacing coast will not benefit much, but significant
conservatism will be recuvered at sites near an east-facing coast or on the
astern boundary of a town.
The dccional method calculation is given in Example 21. The design
values for the orchogonal load cases are selected from these results in
Example 22.
Example 21. Dynamic pressure by Option 3— Twelve 30°-wide
sectors, directional method
e in mv 26
Ber vee
reci de ann m ow
Dino ei a 01 on 01 ON om PES on 149 0m on 0
Free thao ast au toa ar ana EN 28428603170 8
fine we Sop a0 90, sa a0 tin o an me 20) in am
assi a") 0 tg ts 43 38 38400 98S
AR
ani eps om eliana»
ERES
oes wi oe reese ses ess
hee me à FR IN 8 et apap od ak an ae
vac 109650 0 oes os om sont oan 055 0805080
ren fina puezoregbinvo om 0701 bat mr D dan EL
row 7 Do an en oer aes as or me
actor, 7 VOOR ON ARTO 12 TED 1 782 1:742 1.742 1-742 1.655.
fcr Sag sae san sat san cat age ag da ae 348 Da
Teva ad dig fit So 1368168 L879 199014561208 149111391407 1.168 1.167 1271
Elo sm yet Venn” 292-223 278 277 253 283 241 263 28 TS 253 209
Pan peewee gehn SALAM A AGE A703 5937 33983873 42994749 454390333788
een ee em 63 6263 62 62 6363 63 € 63 63 67
ete bn Mem 6) 6) 6) 63 47 29 an 292 an am am 27
Pret 5 1992 09240930 092403720764 97630764 076:0764 07640776
1460 16110 018001920208 0208 038 028 020602080206.
DEU NUTWMESSKAGISDEN 0623 06220623 D487
Ss HAS
mc, ki 244 21 ade ua eer eer san na a
ral nd bal fc. Sa SS 154515961517 1387118 AUS LUE 1091.93 (0981128
lve wid speck Vem” 292 264 206 20% 206 221 227 26 261 258 DA 220
yuu PO 46342604123439036202M1 MO PB AIRE 10811622084
Example 22. Dynamic pressure for orthogonal load cases from
Option 3
Orthogonal direction NE SE SW NW NE SE SW NW
Reference height, Hy 63 67 63 63 8 & 8 08
Effective wind speed, Vii mvs 282° 268 261 282 292 277 278 292
Dynamic presu qu l'a 4862 439.0 418.1 ANA 521.0 4703 4749 5210
-_—_— [7
Note that the dynamic pressure for easterly winds y =90", where the
directional factor is smallest (S4=0-74), is only just smaller than at
«e 240%, where Su=1. This indicates that the effect of the direction.
factor almost exacily balances the combined effects of exposure (distance
in town and effective height) in this case.
The values for our example on the NE boundary of Sheffield
{SK400877) for H,=8m are compared with the corresponding values
in the SW outskirts (SK320810) from Digest 436° in Table 6.
The largest value on the NE boundary comes from the north,
corresponding to the sector of open exposure with the largest direction
factor. The largest value in the SW outskirts occurs at 49 = 240", where
Sa=1, as expected when the effective height is similar in all directions,
responding values for our site at p=240" are significantly
smaller for two reasons: (he altitude of our site (As=$ m) is less than
fetch of town is larger (see
42.5 Comparing options
We are now in a position to see the conservatism inherent in the
three options on the easter boundary of an inland town, The results
Table 6, Comparison of eflecive wind speeds and dmamic pressures at mo sites in
Shoes
INE boundary of Sheet, SK40087
Vien: deg 010 90 120 150 Ham 210 240 270 0 am
Flletivewin speed emis 292 273 275 277 253 235 24 263 278 278 253 209
Dele prosere.quPa $210455.4 468-1 4703 3997 3908 39734029 424946033993 TEA
SW outshes of Shel, ICAO:
ec: dog 0 20 8 où 10 150 180 210 240 270 KO a0
Flleinomndspce Yo mí 261 244 244 248 226 276 292 288 109 206 MA 278
rc pes. q: Pa_ AR 1662 3662 37633125 46995271 $07 258625748 6081 4622
Worked examples | 185
this tothe respective orthogonal case. If we do not know the orientation of
the building, we simply take the largest value of dynamic pressure found.
Whether this largest value is significantly lower than the value from
Option 1 will depend on the relative exposure of the site by direction.
Sites newr a westfacing coast will not benefit much, but significant
conservatism will be recuvered at sites near an east-facing coast or on the
astern boundary of a town.
The dccional method calculation is given in Example 21. The design
values for the orchogonal load cases are selected from these results in
Example 22.
Example 21. Dynamic pressure by Option 3— Twelve 30°-wide
sectors, directional method
e in mv 26
Ber vee
reci de ann m ow
Dino ei a 01 on 01 ON om PES on 149 0m on 0
Free thao ast au toa ar ana EN 28428603170 8
fine we Sop a0 90, sa a0 tin o an me 20) in am
assi a") 0 tg ts 43 38 38400 98S
AR
ani eps om eliana»
ERES
oes wi oe reese ses ess
hee me à FR IN 8 et apap od ak an ae
vac 109650 0 oes os om sont oan 055 0805080
ren fina puezoregbinvo om 0701 bat mr D dan EL
row 7 Do an en oer aes as or me
actor, 7 VOOR ON ARTO 12 TED 1 782 1:742 1.742 1-742 1.655.
fcr Sag sae san sat san cat age ag da ae 348 Da
Teva ad dig fit So 1368168 L879 199014561208 149111391407 1.168 1.167 1271
Elo sm yet Venn” 292-223 278 277 253 283 241 263 28 TS 253 209
Pan peewee gehn SALAM A AGE A703 5937 33983873 42994749 454390333788
een ee em 63 6263 62 62 6363 63 € 63 63 67
ete bn Mem 6) 6) 6) 63 47 29 an 292 an am am 27
Pret 5 1992 09240930 092403720764 97630764 076:0764 07640776
1460 16110 018001920208 0208 038 028 020602080206.
DEU NUTWMESSKAGISDEN 0623 06220623 D487
Ss HAS
mc, ki 244 21 ade ua eer eer san na a
ral nd bal fc. Sa SS 154515961517 1387118 AUS LUE 1091.93 (0981128
lve wid speck Vem” 292 264 206 20% 206 221 227 26 261 258 DA 220
yuu PO 46342604123439036202M1 MO PB AIRE 10811622084
Example 22. Dynamic pressure for orthogonal load cases from
Option 3
Orthogonal direction NE SE SW NW NE SE SW NW
Reference height, Hy 63 67 63 63 8 & 8 08
Effective wind speed, Vii mvs 282° 268 261 282 292 277 278 292
Dynamic presu qu l'a 4862 439.0 418.1 ANA 521.0 4703 4749 5210
-_—_— [7
Note that the dynamic pressure for easterly winds y =90", where the
directional factor is smallest (S4=0-74), is only just smaller than at
«e 240%, where Su=1. This indicates that the effect of the direction.
factor almost exacily balances the combined effects of exposure (distance
in town and effective height) in this case.
The values for our example on the NE boundary of Sheffield
{SK400877) for H,=8m are compared with the corresponding values
in the SW outskirts (SK320810) from Digest 436° in Table 6.
The largest value on the NE boundary comes from the north,
corresponding to the sector of open exposure with the largest direction
factor. The largest value in the SW outskirts occurs at 49 = 240", where
Sa=1, as expected when the effective height is similar in all directions,
responding values for our site at p=240" are significantly
smaller for two reasons: (he altitude of our site (As=$ m) is less than
fetch of town is larger (see
42.5 Comparing options
We are now in a position to see the conservatism inherent in the
three options on the easter boundary of an inland town, The results
Table 6, Comparison of eflecive wind speeds and dmamic pressures at mo sites in
Shoes
INE boundary of Sheet, SK40087
Vien: deg 010 90 120 150 Ham 210 240 270 0 am
Flletivewin speed emis 292 273 275 277 253 235 24 263 278 278 253 209
Dele prosere.quPa $210455.4 468-1 4703 3997 3908 39734029 424946033993 TEA
SW outshes of Shel, ICAO:
ec: dog 0 20 8 où 10 150 180 210 240 270 KO a0
Flleinomndspce Yo mí 261 244 244 248 226 276 292 288 109 206 MA 278
rc pes. q: Pa_ AR 1662 3662 37633125 46995271 $07 258625748 6081 4622
186 1 Wind loading: a practical guide ta BS 6399-2
Worked examples | 187
Fig. 89, Dynamic pressure for ee Am at SK400S77
from Examples 19, 20 and 21 are plotted together in Fig. 89 for the
reference height of H, = 8m. We can immediately sec that a great deal
‘of conservatism is eliminated between Options 1 and 3, which is
typical for sites on eastern boundaries of towns. The Option 1 values
are so high because taking the worst of all cases combines the
maximum direction factor for «2 — 240° with the open exposure to the
est.
Option 2 climinates most of the conservatism for the NE, SE and
SW cases, but not the NW case. In this case it is the combination
Sa 0:99 at y = 270" with the open exposure at — 0° that cuuses th
problem. The site exposure was deliberately chosen to illustrate this
point. Option 2 will perform better at more typically exposed sites
Further dividing the NW case into ‘on-town’ and ‘off-lown’ exposures
will el most of the remaining conservatism, as explained eatlier
Fig. 90, Plan of haute and immediate neighbours
43 Timber-framed house
4.3.1 The house example
‘The timber-ramed house comprises two Em square and 3m high
timber-frame storeys sitting on a 300 mm high foundation. Tt has a 22.5"
duopiteh roof formed from 15 timber trusses at 600 mm centre, providing
a 200mm overhang at cach gable, It has a 3m wide, 3m high garage
attached along the full length of the NW face, flush withthe front of the
house. Figure 90 shows the house in plan and the spacing to its immediate
neighbours
We shall calculate the wind loads for coses NE and NW, corresponding
16 the highest dynamic pressures from Example 22.—the orthogonal load
eases derived from the directional method calculation uf Example 21.
The other two cases are left as an exereise for the reader.
Note: There is eme degree of symmetry to the example house because of
the attached garage. Scaling dimensions, zones sizes and
Coefficients will be the same for the NE and SW cases, but
different for the SE and NW cases.
43.2. Scaling parameters
The scaling lengths and reference heights for the house are shown in
Example 23, together with the span ratios D/H and gap ratios gap that
are needed to derive Ihe pressure coefficients, The rules for the scaling
length b were explained in 2.5.2,
Worked examples | 187
Fig. 89, Dynamic pressure for ee Am at SK400S77
from Examples 19, 20 and 21 are plotted together in Fig. 89 for the
reference height of H, = 8m. We can immediately sec that a great deal
‘of conservatism is eliminated between Options 1 and 3, which is
typical for sites on eastern boundaries of towns. The Option 1 values
are so high because taking the worst of all cases combines the
maximum direction factor for «2 — 240° with the open exposure to the
est.
Option 2 climinates most of the conservatism for the NE, SE and
SW cases, but not the NW case. In this case it is the combination
Sa 0:99 at y = 270" with the open exposure at — 0° that cuuses th
problem. The site exposure was deliberately chosen to illustrate this
point. Option 2 will perform better at more typically exposed sites
Further dividing the NW case into ‘on-town’ and ‘off-lown’ exposures
will el most of the remaining conservatism, as explained eatlier
Fig. 90, Plan of haute and immediate neighbours
43 Timber-framed house
4.3.1 The house example
‘The timber-ramed house comprises two Em square and 3m high
timber-frame storeys sitting on a 300 mm high foundation. Tt has a 22.5"
duopiteh roof formed from 15 timber trusses at 600 mm centre, providing
a 200mm overhang at cach gable, It has a 3m wide, 3m high garage
attached along the full length of the NW face, flush withthe front of the
house. Figure 90 shows the house in plan and the spacing to its immediate
neighbours
We shall calculate the wind loads for coses NE and NW, corresponding
16 the highest dynamic pressures from Example 22.—the orthogonal load
eases derived from the directional method calculation uf Example 21.
The other two cases are left as an exereise for the reader.
Note: There is eme degree of symmetry to the example house because of
the attached garage. Scaling dimensions, zones sizes and
Coefficients will be the same for the NE and SW cases, but
different for the SE and NW cases.
43.2. Scaling parameters
The scaling lengths and reference heights for the house are shown in
Example 23, together with the span ratios D/H and gap ratios gap that
are needed to derive Ihe pressure coefficients, The rules for the scaling
length b were explained in 2.5.2,
188 | Wind loading: a practical guide to BS 6399-2
Worked examples 1 189
——— en
Example 23, Scaling dimensions for timber-framed house, NE
and NW cases
On account of the garage, we must consider thee parts uf the house in ether case
Inthe NE case the breadth is 2-1 m für the lower storey and Kim for the
‘upper strey, while 4 Ki forthe house and # =3 m for the garage, giving b =8m
for par 1, b 11m, for par 2 and & =6m for par 3. Note tha this gives two values of
ap rato forthe Am wide gap 10 the neighbour on the lft
the NW case the heights of the house eaves and ridge are measured from tho
oof of the garage, bul Ihe reference heights H, are measured from the ground, as
always
Un am ont
ITS t 1
E
ve
Fee vah=tm Patt tomar, 0e
Dm. 01 EN
Tim Paiz cromo vals, D - 27 66m
=
Pad guage, b= 2H= em
(a) Case NE (0) Case NW
For the NE case the windward wall forms an irregular face, so that
$244.1 and §2.4.4.2(b) apply.
Part 1. For the house walls above the garage and the house 100%,
8m and H =8m, giving b= 8m
Part 2. For the house walls below the garage roof, B= 11m and
m. giving b = 11m.
Part 3. For the garage, B = 11m and H
For the NW ease, the house wall above the garage is an inset storey, so
that §2.4.4.2 applies.
Part 1. For the house roof, B 8m and H =Sm, giving b= 8m,
Part 2. For the inset house walls above the garage roof, 8 =8m and
H =33m, giving b=66m.
Part 3. For the garage, B= 11m and H —3m, giving b=6m.
3m, giving b =6m,
4.3.3 Pressure coefficients
“The sizes of the pressure coefficient zones and the values uf Cpe are
derived in Example 24 for the NE case and Example 25 for the NW case.
The SW case is a mirror image of the NE case. The SE case will he qui
different from the NW case because the garage is now on the leeward wall,
resulting in u different value of H and b, 50 different zone sizes and Cy.
values through the span ratio 1/11. This i left as an exercise forthe read
Example 24, Pressure coefficient zones for timber-framed house
NE case
as | es
202
08 os
¡a
or
ta
ale a
Pan ppm |
seca [ian
rome LE
e 1
+073 0.90
NE (eee E NW taco
Worked examples 1 189
——— en
Example 23, Scaling dimensions for timber-framed house, NE
and NW cases
On account of the garage, we must consider thee parts uf the house in ether case
Inthe NE case the breadth is 2-1 m für the lower storey and Kim for the
‘upper strey, while 4 Ki forthe house and # =3 m for the garage, giving b =8m
for par 1, b 11m, for par 2 and & =6m for par 3. Note tha this gives two values of
ap rato forthe Am wide gap 10 the neighbour on the lft
the NW case the heights of the house eaves and ridge are measured from tho
oof of the garage, bul Ihe reference heights H, are measured from the ground, as
always
Un am ont
ITS t 1
E
ve
Fee vah=tm Patt tomar, 0e
Dm. 01 EN
Tim Paiz cromo vals, D - 27 66m
=
Pad guage, b= 2H= em
(a) Case NE (0) Case NW
For the NE case the windward wall forms an irregular face, so that
$244.1 and §2.4.4.2(b) apply.
Part 1. For the house walls above the garage and the house 100%,
8m and H =8m, giving b= 8m
Part 2. For the house walls below the garage roof, B= 11m and
m. giving b = 11m.
Part 3. For the garage, B = 11m and H
For the NW ease, the house wall above the garage is an inset storey, so
that §2.4.4.2 applies.
Part 1. For the house roof, B 8m and H =Sm, giving b= 8m,
Part 2. For the inset house walls above the garage roof, 8 =8m and
H =33m, giving b=66m.
Part 3. For the garage, B= 11m and H —3m, giving b=6m.
3m, giving b =6m,
4.3.3 Pressure coefficients
“The sizes of the pressure coefficient zones and the values uf Cpe are
derived in Example 24 for the NE case and Example 25 for the NW case.
The SW case is a mirror image of the NE case. The SE case will he qui
different from the NW case because the garage is now on the leeward wall,
resulting in u different value of H and b, 50 different zone sizes and Cy.
values through the span ratio 1/11. This i left as an exercise forthe read
Example 24, Pressure coefficient zones for timber-framed house
NE case
as | es
202
08 os
¡a
or
ta
ale a
Pan ppm |
seca [ian
rome LE
e 1
+073 0.90
NE (eee E NW taco
190 | Wind loading: a practical guide to BS 6399-2
Worked examples | 191
0.84
os 1
03
3
[o
ET SE lace
Fonnelling (424.14) applies to the side walls of the house und garage. Note that
each sorey of the NW face has a diferent sealing dimension 6. leading to differen
‘ves of zoncs and, though the different gap ratios, o different coefficients. Inset
Storey 82.44.20) apps to the house wall onthe NW face, requiring the extra zune E
fn the comer above the garage roof, Note that zane B does not appear on the garage,
root because b= 24 and the A zones occupy the whole length of the eaves.
Example 25. Pressure coefficient zones for timber-framed house
NW case
‘Two special considerations apply
1. polar Mosh face 62.449, applies to NE and SW faces
3. Inser stores $244.2(0) applies to windward wall of house on NW face,
realucing the windward and leeward pressure coefficients Uhough the span
ratio DAL
‘As the 225° pitch of the house roof lies in the range 19° a <4S" Table 9 gives
both positive and negative values of Cy ard we must use the more onerous depending
‘on application. Similarly, §2.5.1.7 requires the positive pressure on the windward wal
fof the house to be applied to the roof ofthe garage.
E
3 El
E [les
lal EE
E
on
? 08
NE ace A FRS
eo 1 | oe woz
Wace ET
For the NW and SE cases, our house will lic directly in the wake of the
upstream neighbour, It is untikely that these pressure coefficient values
will ever he reached, unless the neighbour is demolished, ‘The allowance
for shelter from the upwind neighbours comes as a reduction of dynamic
pressure through the effective height (see 2.4.3).
4.3.4 Racking forces in timber panels
To select suitable timber panels for the house we need to know the
horizontal shear forces at the mid-height of each storey, as shown in Fig.
91. These are calculated for the NE and NW cases in Example 26. It is
Si Im LS
(a) Cases NE & SW
big. 9. Racking forces in timber panels
(b) Cases NW & SE
Worked examples | 191
0.84
os 1
03
3
[o
ET SE lace
Fonnelling (424.14) applies to the side walls of the house und garage. Note that
each sorey of the NW face has a diferent sealing dimension 6. leading to differen
‘ves of zoncs and, though the different gap ratios, o different coefficients. Inset
Storey 82.44.20) apps to the house wall onthe NW face, requiring the extra zune E
fn the comer above the garage roof, Note that zane B does not appear on the garage,
root because b= 24 and the A zones occupy the whole length of the eaves.
Example 25. Pressure coefficient zones for timber-framed house
NW case
‘Two special considerations apply
1. polar Mosh face 62.449, applies to NE and SW faces
3. Inser stores $244.2(0) applies to windward wall of house on NW face,
realucing the windward and leeward pressure coefficients Uhough the span
ratio DAL
‘As the 225° pitch of the house roof lies in the range 19° a <4S" Table 9 gives
both positive and negative values of Cy ard we must use the more onerous depending
‘on application. Similarly, §2.5.1.7 requires the positive pressure on the windward wal
fof the house to be applied to the roof ofthe garage.
E
3 El
E [les
lal EE
E
on
? 08
NE ace A FRS
eo 1 | oe woz
Wace ET
For the NW and SE cases, our house will lic directly in the wake of the
upstream neighbour, It is untikely that these pressure coefficient values
will ever he reached, unless the neighbour is demolished, ‘The allowance
for shelter from the upwind neighbours comes as a reduction of dynamic
pressure through the effective height (see 2.4.3).
4.3.4 Racking forces in timber panels
To select suitable timber panels for the house we need to know the
horizontal shear forces at the mid-height of each storey, as shown in Fig.
91. These are calculated for the NE and NW cases in Example 26. It is
Si Im LS
(a) Cases NE & SW
big. 9. Racking forces in timber panels
(b) Cases NW & SE
192 | Wind loading: a practical guide to BS 6399-2
Worked examples | 193
curious that the critical racking load for the lower storey Py occurs for
‘case NE and the critical racking load for the upper storey P2 oecurs for
fase NW (although the differences are quite small. This is due to the
different proportion of storey loads to roof/gable Toads in each case.
—— —K—ɣɣíKÉ2
Example 26. Racking forces for timber-framed house
“Our frst action must be to apply the dynamic clasficaion to confirm the señal
‘of BS 6399-2 and obtain the dynamic augmentation factor C,. As expected, C, tune
‘ut o be very small,
Action Notes
Filing eight above ls base M 6+ 3m $13.32, Using height o emest
Read valve of Ry ttom Table Y KyseO-S $1.6.) When in doubt toke next
larger value
Using H and Ky read Cs front Fig? GO $161 IC,>ON ge beter value
from Amex C
1 C,> 025 HS 6309.2 is
or applicable
Check G< 225 Yes $162 BS 6390 can be used
"This sw mate o fadgemen. ls assumed tb he root does no cont Lo te
‘aces, Ung the height of the ridge worl be mre eonscrative Dt, a Cf 0 sul, the
‘ference le signa
‘Our next action is to determine the loaded areas and the corresponding diagonal
dimensions a, The racking shear load. for the lower panels Pa is the sum of the
horizontal loads on the root or gable, te upper storey and the top half ofthe lower
storey. The racking load for Ihe upper panels P is the sum of the horizomal Ins on
the roof or gable and the top half of the upper storey, The loaded areas À are shown
taiched in elevation below, together with the corresponding diagonal dimensions 2
Cases NE & SW "Cases NW & SE
“The presence ofthe ganige is an addtional complication. We shall assume that it
does nok contribute tothe racking loads for the NE cose, bot that its fay propped by
the house in the NW cose. This leads to the loading patterns shown below, The
“dynamic pessutes from Option 3 in Example 22 have been used, bt derivalion of the
dima value for M=3m is let as an exercise for the reader, The intemal
Pressures are not relevant as they cancel out on opposite walls,
aso
y . EE
E am
08 03 asamara
b 08 =]
Case NE Case NW
“Te atan for the NE ese i elatvely sine, ies nly the gle walle
cont hse cet a uno ove ah abe nd hee ly on
tec ig ad es cod Oran se Te are ee
ened he gait nd an sach ey rom in ie prove propio
five und Pa Ne thal he se ele at Cs pled ad al sation
"age together wi he dynam augmenatin cir € In Epon 7. Te so etc
Haar ie ten fm line" of Figure 4 zas the li open cory fortis
Ces. Po the ut some SW uo, you may tae Cy fm ne fot own
tein, eter with ce ler Ayramie pesa to Example 22
NE Case
An Notes
Untastored® horizontal for un gable
stat CynudA=S2LOKO8408)X8x 172 = DIN 2.135
Untaciored" horizontal free on each storey:
Coton = CqnaA~S20x(0B+03)XBXI 137540
Diagonal dimension or Pray (8945) 92m $244
Diagonal dimension for Pasas (4415972 Em Defined above
Sie effect factor fr Pi: Cy 0954 $213: Figure 4
Sie effect factor for Pa Ca = 0963 line 8 ste in country
Racking load: Py -ORSCD Pim —E ln) Call 4G
ORS (A897 197544372)x0.98831 1 ‘ow includes Cy site
Racking load: Py 085 (5 Pénal) IAG efect and C, dynamic
ORS RYT 19754 1/2)x0963201 = SION mention
92.136
= Apply Ger (ee 344)
Worked examples | 193
curious that the critical racking load for the lower storey Py occurs for
‘case NE and the critical racking load for the upper storey P2 oecurs for
fase NW (although the differences are quite small. This is due to the
different proportion of storey loads to roof/gable Toads in each case.
—— —K—ɣɣíKÉ2
Example 26. Racking forces for timber-framed house
“Our frst action must be to apply the dynamic clasficaion to confirm the señal
‘of BS 6399-2 and obtain the dynamic augmentation factor C,. As expected, C, tune
‘ut o be very small,
Action Notes
Filing eight above ls base M 6+ 3m $13.32, Using height o emest
Read valve of Ry ttom Table Y KyseO-S $1.6.) When in doubt toke next
larger value
Using H and Ky read Cs front Fig? GO $161 IC,>ON ge beter value
from Amex C
1 C,> 025 HS 6309.2 is
or applicable
Check G< 225 Yes $162 BS 6390 can be used
"This sw mate o fadgemen. ls assumed tb he root does no cont Lo te
‘aces, Ung the height of the ridge worl be mre eonscrative Dt, a Cf 0 sul, the
‘ference le signa
‘Our next action is to determine the loaded areas and the corresponding diagonal
dimensions a, The racking shear load. for the lower panels Pa is the sum of the
horizontal loads on the root or gable, te upper storey and the top half ofthe lower
storey. The racking load for Ihe upper panels P is the sum of the horizomal Ins on
the roof or gable and the top half of the upper storey, The loaded areas À are shown
taiched in elevation below, together with the corresponding diagonal dimensions 2
Cases NE & SW "Cases NW & SE
“The presence ofthe ganige is an addtional complication. We shall assume that it
does nok contribute tothe racking loads for the NE cose, bot that its fay propped by
the house in the NW cose. This leads to the loading patterns shown below, The
“dynamic pessutes from Option 3 in Example 22 have been used, bt derivalion of the
dima value for M=3m is let as an exercise for the reader, The intemal
Pressures are not relevant as they cancel out on opposite walls,
aso
y . EE
E am
08 03 asamara
b 08 =]
Case NE Case NW
“Te atan for the NE ese i elatvely sine, ies nly the gle walle
cont hse cet a uno ove ah abe nd hee ly on
tec ig ad es cod Oran se Te are ee
ened he gait nd an sach ey rom in ie prove propio
five und Pa Ne thal he se ele at Cs pled ad al sation
"age together wi he dynam augmenatin cir € In Epon 7. Te so etc
Haar ie ten fm line" of Figure 4 zas the li open cory fortis
Ces. Po the ut some SW uo, you may tae Cy fm ne fot own
tein, eter with ce ler Ayramie pesa to Example 22
NE Case
An Notes
Untastored® horizontal for un gable
stat CynudA=S2LOKO8408)X8x 172 = DIN 2.135
Untaciored" horizontal free on each storey:
Coton = CqnaA~S20x(0B+03)XBXI 137540
Diagonal dimension or Pray (8945) 92m $244
Diagonal dimension for Pasas (4415972 Em Defined above
Sie effect factor fr Pi: Cy 0954 $213: Figure 4
Sie effect factor for Pa Ca = 0963 line 8 ste in country
Racking load: Py -ORSCD Pim —E ln) Call 4G
ORS (A897 197544372)x0.98831 1 ‘ow includes Cy site
Racking load: Py 085 (5 Pénal) IAG efect and C, dynamic
ORS RYT 19754 1/2)x0963201 = SION mention
92.136
= Apply Ger (ee 344)
194 | Wind loading: a practical guide to BS 6399-2
Worked examples | 195
NW Case
Action Notes
Unfsctored® horizontal farce on roof (Cyan — Cera
S2NO5x08+03x32+09X084 048 x3-2)x lan22 5x84 = GON
Unfactored® horizontal fore on upper storey: mas
GC Cyn) = 486-3071 #021) 283 Nor
Untactored” horizontal force on lower story:
pie Ca) A=BSSKOBHBEIDMIKEKT — WIEN
Diagonal dimension for Pyiay =182 +017 +491 = 103m $21.34
Diagonal dimension tor 03 TASA 88m Defined above
effect factor Foe Pr Car 0935 2134
Sie fc far fr Pa 099. Figured te:
Aching Hd 2088 Pa he) CNG) panne
AIDE TST PROS ON = COSN Now incr
can ad POS Pira D Pr CHC) em
45 (6880 LFP ON nn
Apply Ca late (ste 244)
“The calenlation for Une NW case is more complex for the following reasons,
© There is different reference height and corresponding dynamic pressure for
the roof and walls of the house and the Front wall of the garage.
‘© The horizontal component from each roof zone is required, using Ihe more
‘onerous positive values on the upwind pitch
© The upper and lower sureys have different toads. The upper storey loads come.
from Ihe windward and leeward hause walls. The lower storey Toads are taken
to come from the windward wall ul the garage and the leeward wall of the
house.
Note: Fur simplicity, the garage Is assumed 10 accupy the whole first storey.
‘covering dhe 200mm high scrip thar exists between the top of the first stares
‘and the roof of he garage
© The size effet factor C, i taken from line
appropriate forall but 15° ofthe 90°
of Figure 4 since town terrain is
sector,
_ -_ _ o _—_— _—.
Fig. 92. Overall forces en roof truss
4.3.5 Overall forces on raof trusses
‘The overall forces uro defined in Fig, 92 as the uplift force un either
end of the truss and the horizontal shear. These are calculated for the
highest-loaded trusses of the NE and NW cases in Example 27, assuming
no load sharing between trusses (worst case).
——
Example 27. Overall forces on roof trusses of the timber-framed
jouse
The figure below shows the iributary areas A (shaded), together with the
corresponding diagonal dimensions « for 8 verge and a central mus, Here we use
the simplest definition of tributary area —extending forthe length of the iss and
haltway towards each of the adjacent trusses. he load on the tros is taken as the
pressure load on this trihatary area, without any load sharing betwecn truses. While
this is a convenient practice, it is not strictly accurate when the loading is non
uniform, Le. when there isa change of zone within the tributary sre,
pecan > room
|
im
(0) or P, and P, (pian)
TESTS
(©) tor, (elevation)
Worked examples | 195
NW Case
Action Notes
Unfsctored® horizontal farce on roof (Cyan — Cera
S2NO5x08+03x32+09X084 048 x3-2)x lan22 5x84 = GON
Unfactored® horizontal fore on upper storey: mas
GC Cyn) = 486-3071 #021) 283 Nor
Untactored” horizontal force on lower story:
pie Ca) A=BSSKOBHBEIDMIKEKT — WIEN
Diagonal dimension for Pyiay =182 +017 +491 = 103m $21.34
Diagonal dimension tor 03 TASA 88m Defined above
effect factor Foe Pr Car 0935 2134
Sie fc far fr Pa 099. Figured te:
Aching Hd 2088 Pa he) CNG) panne
AIDE TST PROS ON = COSN Now incr
can ad POS Pira D Pr CHC) em
45 (6880 LFP ON nn
Apply Ca late (ste 244)
“The calenlation for Une NW case is more complex for the following reasons,
© There is different reference height and corresponding dynamic pressure for
the roof and walls of the house and the Front wall of the garage.
‘© The horizontal component from each roof zone is required, using Ihe more
‘onerous positive values on the upwind pitch
© The upper and lower sureys have different toads. The upper storey loads come.
from Ihe windward and leeward hause walls. The lower storey Toads are taken
to come from the windward wall ul the garage and the leeward wall of the
house.
Note: Fur simplicity, the garage Is assumed 10 accupy the whole first storey.
‘covering dhe 200mm high scrip thar exists between the top of the first stares
‘and the roof of he garage
© The size effet factor C, i taken from line
appropriate forall but 15° ofthe 90°
of Figure 4 since town terrain is
sector,
_ -_ _ o _—_— _—.
Fig. 92. Overall forces en roof truss
4.3.5 Overall forces on raof trusses
‘The overall forces uro defined in Fig, 92 as the uplift force un either
end of the truss and the horizontal shear. These are calculated for the
highest-loaded trusses of the NE and NW cases in Example 27, assuming
no load sharing between trusses (worst case).
——
Example 27. Overall forces on roof trusses of the timber-framed
jouse
The figure below shows the iributary areas A (shaded), together with the
corresponding diagonal dimensions « for 8 verge and a central mus, Here we use
the simplest definition of tributary area —extending forthe length of the iss and
haltway towards each of the adjacent trusses. he load on the tros is taken as the
pressure load on this trihatary area, without any load sharing betwecn truses. While
this is a convenient practice, it is not strictly accurate when the loading is non
uniform, Le. when there isa change of zone within the tributary sre,
pecan > room
|
im
(0) or P, and P, (pian)
TESTS
(©) tor, (elevation)
196 | Wind loading: a practical guide to BS 6399-2
Worked examples | 197
Internal pressure
‘The internal pressure of the upper storey sets on the underside of each truss,
‘contributing to the hold-dowo forces. The calcuations assume thatthe gable walls
have opening windows (permeable) and doors and that the side walls are
impermeable, giving Cu +02 for the NE esse and Cyy= — 03 for the NW case.
Along each verge, the pressure acting on the soffit ofthe 200 mun wide root overhang,
is taken to be the pressure on the adjacent wal ($2.54)
¿Case NR, wind parallel to ridge
For wind parallel tothe idge the verge truss ies In the edge zones A and 8, bul has
half the loaded area ofthe other misses. "The next russ ics pat in ie A and B zones
and party in the C zone and has the full oaded area, The other trusses tie inthe C or D
‘anes, The eiicl, highest loaded truss will be ether the verge or the next truss. By
symimety, Py and Pa are equal 10 the uplift on half the width of the root and Py is
zero. Vaïnes of P and forthe NB cuse are derived next
‘Action Mars
Internal pressure
From Table 16. neral pressure confins Cn 202 F6) Wind aural o
permeable face
From Equation 13, diagonal inmnsionte SIT 42234 Figure A ne,
sd cou
Sie fet flor: Ca = 0815
Uno internal presen pa 5210402 = 120Pa— $24.32
sternal pressure
nfotoed exeral pressure pe Ce sens
Zone À euemal pressure pea SAO MI) Pa =
Zone B extemal prete: poy = S521 03-13) - =677 Pa
Zone © extemal press: foc S21 0-00) —313 Pa
Soft exten pressure: pag = 5210%C408) =417 Pa
Size ee Verge toss Other esses
ingomt dimension,
BR
Sum IPN 82134 Figured fine
0961 sie cowry
Verge wuss | Neat ss € reve 1
Fat oak Zone A rer = 0 320.6 (08-09)92. one
In rok, Zane ares An = 02422 06m UR-0A)x2 one
lé rok, Zone C ses: be = On (0.9-08)4 =04n? 2400
mem are A “Olsed 04m ETS
Solana: y= 0232 = Oe Ont One
Forces
PP A Cr Con Aa Ann te As ADC
Verge wus Next wuss
pln Feres Py = Py TSN uN
Anny G later
Case NW, wind normal to ridge
For wind normal o he ridge al russes are loaded identienlly. apart from the verge
trusses which have a half-size tuibutary area and a 200 m overhang, From the pressure
coefficients for he upssind slope, shown in Example 25, the negative values ere
‘most onerous for uplift forces 7} and P, while the positive values are the most
tenerous for the horizontal shear, Py, The values of Py and Pa are found by taking
moments around each eaves. The moment arm from either eaves tothe centroid Uf
ech of the externa pressure coefficient zones is shown in the figure below.
The valve of Pi is found by resolving the horizontal components of Un zone areas.
e
se
2
Die
a if
=
Seales 30 dat
ome
cats ame
tree
Dies II
Piagonal dimensos: a [9 + (0:61 = 8.001 POST 42214
Sr ie o, m
Zune Au Scam —0 109
O A Moon and
Zone Ewer Apm00.08-040 m —O5c08r ne zo 19 EY
Zom Gas ApnBGx32= 100m? 06 Dr ma D ane "ed
Fores
Up fore a winder exer: 82235 tata moment ar dominar
PA MCS =D Aa 8 + ASE AA o AC) SACAN BEN
pce Icod ee {ELT ing mens aba ad ever
Prey AC ADE ne ACTA AM we AG Nele SIN
orion eet SEE rating Equation 24 (Ear 7)
PaO A Pide Bo As parte) CLG) SOIN
pen Eier
Worked examples | 197
Internal pressure
‘The internal pressure of the upper storey sets on the underside of each truss,
‘contributing to the hold-dowo forces. The calcuations assume thatthe gable walls
have opening windows (permeable) and doors and that the side walls are
impermeable, giving Cu +02 for the NE esse and Cyy= — 03 for the NW case.
Along each verge, the pressure acting on the soffit ofthe 200 mun wide root overhang,
is taken to be the pressure on the adjacent wal ($2.54)
¿Case NR, wind parallel to ridge
For wind parallel tothe idge the verge truss ies In the edge zones A and 8, bul has
half the loaded area ofthe other misses. "The next russ ics pat in ie A and B zones
and party in the C zone and has the full oaded area, The other trusses tie inthe C or D
‘anes, The eiicl, highest loaded truss will be ether the verge or the next truss. By
symimety, Py and Pa are equal 10 the uplift on half the width of the root and Py is
zero. Vaïnes of P and forthe NB cuse are derived next
‘Action Mars
Internal pressure
From Table 16. neral pressure confins Cn 202 F6) Wind aural o
permeable face
From Equation 13, diagonal inmnsionte SIT 42234 Figure A ne,
sd cou
Sie fet flor: Ca = 0815
Uno internal presen pa 5210402 = 120Pa— $24.32
sternal pressure
nfotoed exeral pressure pe Ce sens
Zone À euemal pressure pea SAO MI) Pa =
Zone B extemal prete: poy = S521 03-13) - =677 Pa
Zone © extemal press: foc S21 0-00) —313 Pa
Soft exten pressure: pag = 5210%C408) =417 Pa
Size ee Verge toss Other esses
ingomt dimension,
BR
Sum IPN 82134 Figured fine
0961 sie cowry
Verge wuss | Neat ss € reve 1
Fat oak Zone A rer = 0 320.6 (08-09)92. one
In rok, Zane ares An = 02422 06m UR-0A)x2 one
lé rok, Zone C ses: be = On (0.9-08)4 =04n? 2400
mem are A “Olsed 04m ETS
Solana: y= 0232 = Oe Ont One
Forces
PP A Cr Con Aa Ann te As ADC
Verge wus Next wuss
pln Feres Py = Py TSN uN
Anny G later
Case NW, wind normal to ridge
For wind normal o he ridge al russes are loaded identienlly. apart from the verge
trusses which have a half-size tuibutary area and a 200 m overhang, From the pressure
coefficients for he upssind slope, shown in Example 25, the negative values ere
‘most onerous for uplift forces 7} and P, while the positive values are the most
tenerous for the horizontal shear, Py, The values of Py and Pa are found by taking
moments around each eaves. The moment arm from either eaves tothe centroid Uf
ech of the externa pressure coefficient zones is shown in the figure below.
The valve of Pi is found by resolving the horizontal components of Un zone areas.
e
se
2
Die
a if
=
Seales 30 dat
ome
cats ame
tree
Dies II
Piagonal dimensos: a [9 + (0:61 = 8.001 POST 42214
Sr ie o, m
Zune Au Scam —0 109
O A Moon and
Zone Ewer Apm00.08-040 m —O5c08r ne zo 19 EY
Zom Gas ApnBGx32= 100m? 06 Dr ma D ane "ed
Fores
Up fore a winder exer: 82235 tata moment ar dominar
PA MCS =D Aa 8 + ASE AA o AC) SACAN BEN
pce Icod ee {ELT ing mens aba ad ever
Prey AC ADE ne ACTA AM we AG Nele SIN
orion eet SEE rating Equation 24 (Ear 7)
PaO A Pide Bo As parte) CLG) SOIN
pen Eier
198 1 Wind loading: a practical guide to BS 6399-2
Worked examples | 199
“The equation for surface loads, Fqwarion 6, is uncd forthe windward uplift forces
because they are generated from the difference between internal and external
pressures, On the other hand, the equation for overall loads, Equation 7, is used for the
horizontal force because ts generated from the difference between external pressures
‘on windward and leeward faces.
‘The dynamic augmentation factor in Equation 7 applies to the main structural
members that resist the mverll lateral loads. I is included in the ealeuatlen fr the
horizontal force Pu on the basis that this will be enhanced by any lacral dynamic
response ofthe building. Anyway, C,=0.01 is very small, whichis typical of housing
4.4 Long-span portal-frame building
4.4.1 The portal-frame building example
‘The dimensions of the portal-frame building are defined in Fig. 93. The
aligned, lke the house, with the idge at yp — 45°. This dir
from the example building in BRE Digest 436° only in the fact that the
{gable wall that faces north-east is permanently open.
We shall assume that the building is among other portalframe
buildings of the same height with typical spacing, Le. a light industrial
estate. As the spacing is not known, we shall take the typical value of
Xo—20m (see 3.6.2). This leads to values for the required parameters
given for our site at SK400877 given in Table 7, The calculation of
dymamic pressure at the required reference heights is lll as an exercise
om
en
om À won
Fig. 93. Dimension of the nomatfrae bring
Table 7. Parameters for the poral frome building
Dynamic pressure, ay Pa
SE
NW NE sw
Ride 590 86 sm
Faves sors
rinogonel ne Sealing length, b= 2H m ‘Span ratio, DH
NESW 13 62
NW SR: wal 14 43
NW SE: root 193
for the reader. The values of dynam
are the same because both are
© pressure for the NE and NW cases
el by @ = 0° sec Example 22.
44.2 Pressure coefficients
‘The sizes of the pressure coeff
derived in Example 28
—— ———
Example 28, Pressure coefficient zones for th x
ae r the portal-framed
nts ond the values of Ce and Cy e
‘The pressure coulent zones and values for the pote bi
time blog ae given
low We donot know the exact loeation of cin bulking, cry thal te
soscing is ype. Here we assume tht the neighboring balling no cute
fuonstng. bu i melting I suspected then the wor case values elven In Tale 8
snl be The wa cen ee len om Tale nd ne un rico
rom Table 10. The grey sing dens the mera pest usd bythe open
taken from fable 18. a Dear
Wind angle @ = 0, case NM and SE
Wind angio @ = 0 casas NE an SW
parts
£ ste
ly $ =Hos| 035
Fr. "te
B Eos 1
CE
e =a TES
Wind aca. Smt
Cp
« , +06 04 os |?
Leorenttce n 7
ae E Br
Ho eue Tun
Suns m Fos +088]
— 2 jas Solid ‘Open
Ein
Scan Leeward a SSS
ore (28 E 21-016
Sois "Open
4.43 Internal pressures
‘The internal pressure, which is controlled by the orientation of the
permanently open side, is calculated for each orthogonal load case
Worked examples | 199
“The equation for surface loads, Fqwarion 6, is uncd forthe windward uplift forces
because they are generated from the difference between internal and external
pressures, On the other hand, the equation for overall loads, Equation 7, is used for the
horizontal force because ts generated from the difference between external pressures
‘on windward and leeward faces.
‘The dynamic augmentation factor in Equation 7 applies to the main structural
members that resist the mverll lateral loads. I is included in the ealeuatlen fr the
horizontal force Pu on the basis that this will be enhanced by any lacral dynamic
response ofthe building. Anyway, C,=0.01 is very small, whichis typical of housing
4.4 Long-span portal-frame building
4.4.1 The portal-frame building example
‘The dimensions of the portal-frame building are defined in Fig. 93. The
aligned, lke the house, with the idge at yp — 45°. This dir
from the example building in BRE Digest 436° only in the fact that the
{gable wall that faces north-east is permanently open.
We shall assume that the building is among other portalframe
buildings of the same height with typical spacing, Le. a light industrial
estate. As the spacing is not known, we shall take the typical value of
Xo—20m (see 3.6.2). This leads to values for the required parameters
given for our site at SK400877 given in Table 7, The calculation of
dymamic pressure at the required reference heights is lll as an exercise
om
en
om À won
Fig. 93. Dimension of the nomatfrae bring
Table 7. Parameters for the poral frome building
Dynamic pressure, ay Pa
SE
NW NE sw
Ride 590 86 sm
Faves sors
rinogonel ne Sealing length, b= 2H m ‘Span ratio, DH
NESW 13 62
NW SR: wal 14 43
NW SE: root 193
for the reader. The values of dynam
are the same because both are
© pressure for the NE and NW cases
el by @ = 0° sec Example 22.
44.2 Pressure coefficients
‘The sizes of the pressure coeff
derived in Example 28
—— ———
Example 28, Pressure coefficient zones for th x
ae r the portal-framed
nts ond the values of Ce and Cy e
‘The pressure coulent zones and values for the pote bi
time blog ae given
low We donot know the exact loeation of cin bulking, cry thal te
soscing is ype. Here we assume tht the neighboring balling no cute
fuonstng. bu i melting I suspected then the wor case values elven In Tale 8
snl be The wa cen ee len om Tale nd ne un rico
rom Table 10. The grey sing dens the mera pest usd bythe open
taken from fable 18. a Dear
Wind angle @ = 0, case NM and SE
Wind angio @ = 0 casas NE an SW
parts
£ ste
ly $ =Hos| 035
Fr. "te
B Eos 1
CE
e =a TES
Wind aca. Smt
Cp
« , +06 04 os |?
Leorenttce n 7
ae E Br
Ho eue Tun
Suns m Fos +088]
— 2 jas Solid ‘Open
Ein
Scan Leeward a SSS
ore (28 E 21-016
Sois "Open
4.43 Internal pressures
‘The internal pressure, which is controlled by the orientation of the
permanently open side, is calculated for each orthogonal load case
200 | Wind loading: a practical guide 10 BS 6399-2
Worked examples | 201
Example 29. Compare this with the Digest 436* example, where the
internal pressure is controlled by a roller-shutter dour that is taken as
closed at the ultimate limit and open at the serviceability limit
— m
Example 29. Internal pressures for the portal-frame building,
‘The dimensions and orientation of Ihe open side control the value uf internal pressure.
Acton Naver
Diagonal of open ce, a (0 +7" Diegonal of opening
26
Ontogenal eue NE SESW NW NE ise sot
Wind ange. 8 do non
Internat pressure uen, Cy, 085 06 -016 Frs Esa
Dynamic pressure ar open Wall Pa SIDO 4956 5221 Fco Tole
Unfaiors” itomal pressive, Pa 4667 2974895
Sie et (not. Cu 0863 0835 085
Figo 4 nes 8°
and ©, R134
Apply Cy tater
If we compare this example of an open face with the example in Digest
4362 of a dominant opening, we see that the range of internal pressure
coefficients is smaller, Dominant openings are generally more onerous
than a completely open face, however they can be treated as a
serviceability case when they are usually closed.
4.4.4 Highest-loaded purlin and rail
We expect the highestloaded purlin to be the first or second purlin
from the eaves in the bay at the upwind end, At wind angle 0 = 0° the first
purlin lies within the A zone but has half the tributary area of the second
purlin which lies partly in the A and partly in the C zone. Similarly, we
expect the highest Inaded rail to be a middle rail in the end bay. The
tributary areas for these purlins and rails are shown in Fig, 94.
‘The calculations for the highest purlin and rail loads are given in
Example 30, using the pressure coefficient averaged over the loaded area
when it includes a zone boundary (see 34.1). The highest loading (data in
bold face) occurs to the rails and purlins of the windward end bay in the
NE case, with the wind blowing into the open face,
fon |
pate La |
Zn
som KT
er
as a
Rank
(a) Wind angle a= 0° (6) Wind angle 4= 90°
Pig. 9%. Tia oras for puros and rief te portal frame bling
Example 30. Highest-loaded purlin and rail for the portal-
framed building
ogee EW = 90) won _ Au
Porn Mi Pate Porn Ra
Middle Middle Ist od Mile
Paterna presse
Disgorldimceson. cm 64 BAGG
Sree con Cu 0942 Om 0982 US 0978 NE Cam. Mo
dons rs cn TR 1898 GA 1908. 1398, lod dune
es ester pesao
Bullen. Ce coms on 13366 Average gorse ones
Dam pin ge 580. 005 3190 S013 From Table?
wes, po Pa 02-5180 1548 21099 Men ovr nat are
Internal pressure
Ske ft fear, Cy DA OSE OAS ORS 083
Unfcoredintemal presen, Fra Frame 29
ms AT 4667-3294 2204-2004
Look PPC RN IR 127 AM 05 196 NE cere orar
4.4.5 Loads on portal frames, NW cose
‘The main frames resist the wind loads as portal frames when the wind
is normal to the eaves, 0 =0", so that the NW case is the relevant design
case. The highest-louded frame is Frame 2, the first frame after the end
Worked examples | 201
Example 29. Compare this with the Digest 436* example, where the
internal pressure is controlled by a roller-shutter dour that is taken as
closed at the ultimate limit and open at the serviceability limit
— m
Example 29. Internal pressures for the portal-frame building,
‘The dimensions and orientation of Ihe open side control the value uf internal pressure.
Acton Naver
Diagonal of open ce, a (0 +7" Diegonal of opening
26
Ontogenal eue NE SESW NW NE ise sot
Wind ange. 8 do non
Internat pressure uen, Cy, 085 06 -016 Frs Esa
Dynamic pressure ar open Wall Pa SIDO 4956 5221 Fco Tole
Unfaiors” itomal pressive, Pa 4667 2974895
Sie et (not. Cu 0863 0835 085
Figo 4 nes 8°
and ©, R134
Apply Cy tater
If we compare this example of an open face with the example in Digest
4362 of a dominant opening, we see that the range of internal pressure
coefficients is smaller, Dominant openings are generally more onerous
than a completely open face, however they can be treated as a
serviceability case when they are usually closed.
4.4.4 Highest-loaded purlin and rail
We expect the highestloaded purlin to be the first or second purlin
from the eaves in the bay at the upwind end, At wind angle 0 = 0° the first
purlin lies within the A zone but has half the tributary area of the second
purlin which lies partly in the A and partly in the C zone. Similarly, we
expect the highest Inaded rail to be a middle rail in the end bay. The
tributary areas for these purlins and rails are shown in Fig, 94.
‘The calculations for the highest purlin and rail loads are given in
Example 30, using the pressure coefficient averaged over the loaded area
when it includes a zone boundary (see 34.1). The highest loading (data in
bold face) occurs to the rails and purlins of the windward end bay in the
NE case, with the wind blowing into the open face,
fon |
pate La |
Zn
som KT
er
as a
Rank
(a) Wind angle a= 0° (6) Wind angle 4= 90°
Pig. 9%. Tia oras for puros and rief te portal frame bling
Example 30. Highest-loaded purlin and rail for the portal-
framed building
ogee EW = 90) won _ Au
Porn Mi Pate Porn Ra
Middle Middle Ist od Mile
Paterna presse
Disgorldimceson. cm 64 BAGG
Sree con Cu 0942 Om 0982 US 0978 NE Cam. Mo
dons rs cn TR 1898 GA 1908. 1398, lod dune
es ester pesao
Bullen. Ce coms on 13366 Average gorse ones
Dam pin ge 580. 005 3190 S013 From Table?
wes, po Pa 02-5180 1548 21099 Men ovr nat are
Internal pressure
Ske ft fear, Cy DA OSE OAS ORS 083
Unfcoredintemal presen, Fra Frame 29
ms AT 4667-3294 2204-2004
Look PPC RN IR 127 AM 05 196 NE cere orar
4.4.5 Loads on portal frames, NW cose
‘The main frames resist the wind loads as portal frames when the wind
is normal to the eaves, 0 =0", so that the NW case is the relevant design
case. The highest-louded frame is Frame 2, the first frame after the end
202 | Wind loading: a practical guide to BS 6399-2
Worked examples | 203
frame, ic. the only frame having a full-sized tributary area that also lies
fully within the verge A and E zones. We shall also determine the loads
‘on the middle frames that lie within the B and F zones, represented by
Frame 4, for a reason that will become: apparent later.
445.1 Loads on individual members
The wind loads are transmitted 10 the portal frames as a series of point
loads through the rail and purlin fixings as shown in Fig. 96, We could sum
the individual unfactored rail and purlin loads, then apply the appropriate
size effect factor for the frame, but instcad we shall adopt the quicker and
commonly used practice of summing the load over the tributary area of
cladding 10 give an equivalent uniformly distributed load (UDL). The
calculations for the loads on the individual members of the Frame highest
Jonded frame in the NW case are demonstrated in Example 31
Fig. 95, Rail and purl ads on a frame
Example 31. Loads on members of frames of the portal-frame
building in the NW case
Treat rca
Une ral pese. pam 3094 From sample 29
Se fet ete Ca Lois
Estena priare Pro Upwindrfer Dunning ter
Frans 24 2
Dymmi pee. se SS soo Fron Te 7
argc rame
men 06 ous From Bsn 29
Den
ap MOS 1596 3088 2008
tacon dino, cam 92 164 Jed 164 ind 92
See ete ctr Co ONS 0802 02 OF 042 OBS Sue om
Land
tt a int.
(mCumCuis6Nin 3366 274 236 266 66 GS Lol perece
“Apply oa ie
Although the overall drag of the building is the same as if it wore
enclosed, Ihe high internal suetion enuscd hy the open face changes the
distribution of load between the members. This produces a very high load
on the front column and a load on the rear column that acls into the wind.
Also, the high uplift expected on the rafters is greatly reduced, becoming
‘a downvrard load on the downwind rafters and on the upwind rañer of the
central frames. Compare this with the typical enclosed building case in
the corresponding example in Digest 436.2
We have not included the dynamic augmentation factor €, because
these are component loads. Following the advice in 3.4.6.1, we should
include C, and the factor 0:85 in Equation 22 (Equation 7) when these
member loads are used to calculate stresses in the whole frame acting as a
portal, ie. where the loads on both windward and leeward members
contribute to the stress. However, we must remember 10 apply the
asymmetric load provisions of $2.1.3.7 only on the extemal pressure
component of these loads, as explained earlier in 2.6.4, and to change the
size effect factor C from the values for the members tu the value for the
overall frame,
4452 Base shear
In this example, we cal shcar force at the base of the
columns in order to demonstrate overall toads using Equation 22
(Equation 7). Figure 60, earlier, showed the tibutary area (hatched) for
the horizontal hase shear of a frame viewed in elevation and the
corresponding diagonal dimension for base shear. This principle is
applied to our highest-loaded frame in Fig. 96. We assume no load
sharing between frames.
We can accumulate the unfactored base shear from the wnfactored
external pressures and then apply the appropriate size effect. We have
alway obtained the relevant values in Example 31 and can carry these
ig, 96. Tribunary area for base shear of «frame jn elevaron
Worked examples | 203
frame, ic. the only frame having a full-sized tributary area that also lies
fully within the verge A and E zones. We shall also determine the loads
‘on the middle frames that lie within the B and F zones, represented by
Frame 4, for a reason that will become: apparent later.
445.1 Loads on individual members
The wind loads are transmitted 10 the portal frames as a series of point
loads through the rail and purlin fixings as shown in Fig. 96, We could sum
the individual unfactored rail and purlin loads, then apply the appropriate
size effect factor for the frame, but instcad we shall adopt the quicker and
commonly used practice of summing the load over the tributary area of
cladding 10 give an equivalent uniformly distributed load (UDL). The
calculations for the loads on the individual members of the Frame highest
Jonded frame in the NW case are demonstrated in Example 31
Fig. 95, Rail and purl ads on a frame
Example 31. Loads on members of frames of the portal-frame
building in the NW case
Treat rca
Une ral pese. pam 3094 From sample 29
Se fet ete Ca Lois
Estena priare Pro Upwindrfer Dunning ter
Frans 24 2
Dymmi pee. se SS soo Fron Te 7
argc rame
men 06 ous From Bsn 29
Den
ap MOS 1596 3088 2008
tacon dino, cam 92 164 Jed 164 ind 92
See ete ctr Co ONS 0802 02 OF 042 OBS Sue om
Land
tt a int.
(mCumCuis6Nin 3366 274 236 266 66 GS Lol perece
“Apply oa ie
Although the overall drag of the building is the same as if it wore
enclosed, Ihe high internal suetion enuscd hy the open face changes the
distribution of load between the members. This produces a very high load
on the front column and a load on the rear column that acls into the wind.
Also, the high uplift expected on the rafters is greatly reduced, becoming
‘a downvrard load on the downwind rafters and on the upwind rañer of the
central frames. Compare this with the typical enclosed building case in
the corresponding example in Digest 436.2
We have not included the dynamic augmentation factor €, because
these are component loads. Following the advice in 3.4.6.1, we should
include C, and the factor 0:85 in Equation 22 (Equation 7) when these
member loads are used to calculate stresses in the whole frame acting as a
portal, ie. where the loads on both windward and leeward members
contribute to the stress. However, we must remember 10 apply the
asymmetric load provisions of $2.1.3.7 only on the extemal pressure
component of these loads, as explained earlier in 2.6.4, and to change the
size effect factor C from the values for the members tu the value for the
overall frame,
4452 Base shear
In this example, we cal shcar force at the base of the
columns in order to demonstrate overall toads using Equation 22
(Equation 7). Figure 60, earlier, showed the tibutary area (hatched) for
the horizontal hase shear of a frame viewed in elevation and the
corresponding diagonal dimension for base shear. This principle is
applied to our highest-loaded frame in Fig. 96. We assume no load
sharing between frames.
We can accumulate the unfactored base shear from the wnfactored
external pressures and then apply the appropriate size effect. We have
alway obtained the relevant values in Example 31 and can carry these
ig, 96. Tribunary area for base shear of «frame jn elevaron
206 | Wind loading: a practical guide to BS 6399-2
Worked examples | 205
forward into the caleulation of base shear in Example 32 to avoid
unnccessary duplication of effort. The external pressure load on the
‘upwind rafter is a beneficial component of the horizontal base shear, so
we need to apply the asymmetric load provisions of $2.1.3.7. Although
Frame 2 atthe verge is more highly loaded than Frame 4 in che middle of
the building, a larger proportion of this higher loading is beneficial. After
applying $2.1.3.7 we find that the middle frames (Frame 4) have a
marginally higher overall base shear than the verge frames (Brame 2).
m
Example 32. Base shear on the frames of the portal-frame
building in the NW case
Mes
Dynami casifcation
"acne eg, me 1 BAIE Ure het re amer
Flia pe ete 2 BEAT Fm Tae?
Dane augmentant, G DO From Fee 3
Check C2028 Yo 62 BS O82 con be
ed
Eternal resure Fm Upwind tater Uownaad fer Rest
Fromss M0 2 4 2 4 2
‘Averaged union”
sur, 209-3596 AMÓ 2608 —2999 -502 Frou Example 30
Dot
uen À Aldi
stator Ina
Poa: kN 1204-572 AM AIS ARI ZN Inematpresure
Amen
Diagonal dimension. cm asa
Sie ec ctr Figure nese in on
Dase tir, POSSE prey Di) C4) Frame? Frame à $2.46 Apr Cy
OOO Poe Podio ISR IDISRN BAS Pap Denia
Apply, ter
The base shear that we obtain in Example 32 is the sum of the
individual shears at the foot of each column, hut we have already
discovered in 4.4.5.1, that this will not be distributed evenly between the
two columns, The shear athe base of each individual column will depend
om the distribution of loads on the individual members, the stiffness of the
portal frame and its ability to share load with other frames, €.g. through
racking of the cladding,
4.4.6 Overall horizontal load, NE case
e ore tn al ocean
‘The wind bracing shares this load with the fixings at the base of the
columns, and is usually assumed to bear half. The relevant diagonal
dimension for the gable wall is shown in Figure 5(d)
al sum dita DONG à
pressure coefficient +085-{-0-1)- +095. In addition to the
normal pressure loads on the gable, there is a component from friction
áiLáAAAAAA—m—
Example 33. Overall horizontal load for the portal-frame
building in the NE case
Noes
Dynamic pressure a ridge, q. $490Pa From Example 31
Internal pressure coefficient, Gy 085 $263, shorter face open
Unfactored inerral pressure, py M667P $21.32
Size effect factor, Cu, 0863 From Example 29
Fstemal pressure coefficient, Cpe 01 Leeward wall
Unfnetored external pressure, pe -549Pa
Diagonal dimension for gable, a AK Ligure Sta)
Size effect factor, Coe 0868 Figured.Mne'B), country
‘Area of gable wall, À 249-8 m?
Dynamic augmentation factor, CIS From Example 32
‘Overall load from normal pressures. P 98:9kN Front Equation 24 (7)
Frictional drag coeiclemt, Cr 00 Table 6 Corrugations
Arca of € zones on walls, Ac
7x (60-19.3) 560 m7 Aroa swept by wind
Area of D zones on roof. Ao,
= 30/eus 10° (601-9465) 15138m* Roof efect dominates
Diagonal dimension fer frito 586m From roof D zones on plan
Size effec factor for Faction, C, 0814
Load from friction, Py NEBEN 19% of normal pressure load
‘Total horizontal load =P + Py 17780
Worked examples | 205
forward into the caleulation of base shear in Example 32 to avoid
unnccessary duplication of effort. The external pressure load on the
‘upwind rafter is a beneficial component of the horizontal base shear, so
we need to apply the asymmetric load provisions of $2.1.3.7. Although
Frame 2 atthe verge is more highly loaded than Frame 4 in che middle of
the building, a larger proportion of this higher loading is beneficial. After
applying $2.1.3.7 we find that the middle frames (Frame 4) have a
marginally higher overall base shear than the verge frames (Brame 2).
m
Example 32. Base shear on the frames of the portal-frame
building in the NW case
Mes
Dynami casifcation
"acne eg, me 1 BAIE Ure het re amer
Flia pe ete 2 BEAT Fm Tae?
Dane augmentant, G DO From Fee 3
Check C2028 Yo 62 BS O82 con be
ed
Eternal resure Fm Upwind tater Uownaad fer Rest
Fromss M0 2 4 2 4 2
‘Averaged union”
sur, 209-3596 AMÓ 2608 —2999 -502 Frou Example 30
Dot
uen À Aldi
stator Ina
Poa: kN 1204-572 AM AIS ARI ZN Inematpresure
Amen
Diagonal dimension. cm asa
Sie ec ctr Figure nese in on
Dase tir, POSSE prey Di) C4) Frame? Frame à $2.46 Apr Cy
OOO Poe Podio ISR IDISRN BAS Pap Denia
Apply, ter
The base shear that we obtain in Example 32 is the sum of the
individual shears at the foot of each column, hut we have already
discovered in 4.4.5.1, that this will not be distributed evenly between the
two columns, The shear athe base of each individual column will depend
om the distribution of loads on the individual members, the stiffness of the
portal frame and its ability to share load with other frames, €.g. through
racking of the cladding,
4.4.6 Overall horizontal load, NE case
e ore tn al ocean
‘The wind bracing shares this load with the fixings at the base of the
columns, and is usually assumed to bear half. The relevant diagonal
dimension for the gable wall is shown in Figure 5(d)
al sum dita DONG à
pressure coefficient +085-{-0-1)- +095. In addition to the
normal pressure loads on the gable, there is a component from friction
áiLáAAAAAA—m—
Example 33. Overall horizontal load for the portal-frame
building in the NE case
Noes
Dynamic pressure a ridge, q. $490Pa From Example 31
Internal pressure coefficient, Gy 085 $263, shorter face open
Unfactored inerral pressure, py M667P $21.32
Size effect factor, Cu, 0863 From Example 29
Fstemal pressure coefficient, Cpe 01 Leeward wall
Unfnetored external pressure, pe -549Pa
Diagonal dimension for gable, a AK Ligure Sta)
Size effect factor, Coe 0868 Figured.Mne'B), country
‘Area of gable wall, À 249-8 m?
Dynamic augmentation factor, CIS From Example 32
‘Overall load from normal pressures. P 98:9kN Front Equation 24 (7)
Frictional drag coeiclemt, Cr 00 Table 6 Corrugations
Arca of € zones on walls, Ac
7x (60-19.3) 560 m7 Aroa swept by wind
Area of D zones on roof. Ao,
= 30/eus 10° (601-9465) 15138m* Roof efect dominates
Diagonal dimension fer frito 586m From roof D zones on plan
Size effec factor for Faction, C, 0814
Load from friction, Py NEBEN 19% of normal pressure load
‘Total horizontal load =P + Py 17780
206 | Wind fondling: a practical guide to BS 6399-2
Worked examples | 207
oe je am
L rm
| antes ae on |]
Lan se
+ il me
DH =
(a) NE face (0) NW face (0) SW face
ig. 97. Dimensions ofthe five-stures fie bullng
= 25
‘Table & Dynamic pressures 4, in Pa) at SK400877 for he five-store office example
Orthoganal direction
NE se sw NW
We BRL NGS 233
5 78
45 Five-storey office building
4.5.1. The five-storey office example
“The dimensions of the five-storey office example are defined in Fig.
97. The building has the same alignment, with the minor axis aligned to
(p 45°. There is a small single-storey extension with a 4m wide by 3 m
high roller-shutter door that forms an elective dominant opening. We shall
use this example to derive the loads on wall cladding panels for ultimate
and serviceability limits.
"This time we shall assume that the urban development to Ihe west and
south of our site at SK400877 is typical of Category 3 in CP 3-V-2,
leading to a displacement height of Hg = 10 m. We shall need the dynamic
pressure for the reference heights #,= 4m and 15m and these values are
provided in Table 3 for the four orthogonal cases, obtained using Option
3. At H,=4m the largest value which applies to both the NE and NW
‘cases is set by 9 = 0°, as before, ut at H, = 15 m, the largest value is set
by y = 240°. This occurs because the shelter given by the zero-plane
displacement is less effective at height, as noted earlier in 3.3.3.2 and
demonstrated in Fig. 71.
4.5.2 Scaling length, b
“The scaling lengths for the five-storey office are shown in Example 34,
“The rules für scaling length were explained in 2.5.2.
—__ Un
Example 34, Scaling dimensions for the five-storey office
For the NE case, the main
10m,
ice building is a simple rectangular block, with
For the NW a NE cae we mus cone he pat ash el, Pat
jul anoto ed wo pant ur 42232 242 Bate tp
per mad o) he we sl ve LS pu li enh D
ye Lape for Parts 1 and 2 and by the height for Part 3. 7
See mm ope shown bow The sei o
bis set hy the height for both these parts. ™ ee
Cpr ths ee ith he oso Example 24
tm
$ ske Kol
] 127m) Sir
En
1m Pan2,b=15m, H= 15m Pan, b= 8m, H, = 4m
¡«— 20m 200 —»
RCD
A ZA À
21,
£ | 12227
Par 1. b= 20m, H,= 15m Pat2, b= 6m, H, = 4m
For the NE case we have a single value for b, based on the main fa
de ingl based on the main face of
For the NW and SE cases, the windward wall forms an inegular face,
so that §2.4.4.1 and §2.4.4.2(b) apply.
1. For the office walls above the roof level of the extension, B= 10m
and 115m, giving b= 10m.
2. For the office walls below the roof level of the extension, B— 15m
and H= 15m, giving b= 15m.
3. For the extension, B~ 15m and H=4m,
ving b= 81
For the SW case, the office above the extension Forms an
so that $2.4.4.2 applies.
set storey,
L For the office, B=20m and H= 11 m, giving b= 20m.
2. For the extension, B= 20m and H—4m, giving b— 8m
Worked examples | 207
oe je am
L rm
| antes ae on |]
Lan se
+ il me
DH =
(a) NE face (0) NW face (0) SW face
ig. 97. Dimensions ofthe five-stures fie bullng
= 25
‘Table & Dynamic pressures 4, in Pa) at SK400877 for he five-store office example
Orthoganal direction
NE se sw NW
We BRL NGS 233
5 78
45 Five-storey office building
4.5.1. The five-storey office example
“The dimensions of the five-storey office example are defined in Fig.
97. The building has the same alignment, with the minor axis aligned to
(p 45°. There is a small single-storey extension with a 4m wide by 3 m
high roller-shutter door that forms an elective dominant opening. We shall
use this example to derive the loads on wall cladding panels for ultimate
and serviceability limits.
"This time we shall assume that the urban development to Ihe west and
south of our site at SK400877 is typical of Category 3 in CP 3-V-2,
leading to a displacement height of Hg = 10 m. We shall need the dynamic
pressure for the reference heights #,= 4m and 15m and these values are
provided in Table 3 for the four orthogonal cases, obtained using Option
3. At H,=4m the largest value which applies to both the NE and NW
‘cases is set by 9 = 0°, as before, ut at H, = 15 m, the largest value is set
by y = 240°. This occurs because the shelter given by the zero-plane
displacement is less effective at height, as noted earlier in 3.3.3.2 and
demonstrated in Fig. 71.
4.5.2 Scaling length, b
“The scaling lengths for the five-storey office are shown in Example 34,
“The rules für scaling length were explained in 2.5.2.
—__ Un
Example 34, Scaling dimensions for the five-storey office
For the NE case, the main
10m,
ice building is a simple rectangular block, with
For the NW a NE cae we mus cone he pat ash el, Pat
jul anoto ed wo pant ur 42232 242 Bate tp
per mad o) he we sl ve LS pu li enh D
ye Lape for Parts 1 and 2 and by the height for Part 3. 7
See mm ope shown bow The sei o
bis set hy the height for both these parts. ™ ee
Cpr ths ee ith he oso Example 24
tm
$ ske Kol
] 127m) Sir
En
1m Pan2,b=15m, H= 15m Pan, b= 8m, H, = 4m
¡«— 20m 200 —»
RCD
A ZA À
21,
£ | 12227
Par 1. b= 20m, H,= 15m Pat2, b= 6m, H, = 4m
For the NE case we have a single value for b, based on the main fa
de ingl based on the main face of
For the NW and SE cases, the windward wall forms an inegular face,
so that §2.4.4.1 and §2.4.4.2(b) apply.
1. For the office walls above the roof level of the extension, B= 10m
and 115m, giving b= 10m.
2. For the office walls below the roof level of the extension, B— 15m
and H= 15m, giving b= 15m.
3. For the extension, B~ 15m and H=4m,
ving b= 81
For the SW case, the office above the extension Forms an
so that $2.4.4.2 applies.
set storey,
L For the office, B=20m and H= 11 m, giving b= 20m.
2. For the extension, B= 20m and H—4m, giving b— 8m
208 | Wind loading: a practical guide to BS 6399-2
Worked examples | 209
4.5.3 Pressure coefficients for the walls
The sizes of the pressure coefficient zones and the values of Cpe are
derived in Example 35, The SE case is a mirror image of the NW case. It
may be helpful to compare the NW face for the NE and SW cases with the
definition figure for irregular flush faces, Figure 14, in BS 6399-2.
Note the position of the roller-shutter door on the SW face of the
extension relative to the zones. For the NE case it lies on the leeward face
where the external pressure is controlled by the leeward face of the office,
ie. Cye= —03 and H,= 15m. In the NW and SE cases, it lies just inside
the C zone on the side face and for the SW case itis on the windward face,
where H,=4m. This controls the intemal pressure of the lower storey
when the door is open.
Example 38. Pressure coefficients for the walls of the
storey office
“Two special considerations apply.
© regular Mush face §2.44.1 applies to the NW and SE faces, but only the NW
face is shown, Note thal the diferent scaling lengths b for Pans 1 and 2 (see
Example 34) cause different sizes of zones an the NE face.
© Inset storey 82.44.21) applies to the SW face and the additional E zone with
Gye = — 20 is required forthe NW and SB cases. Note that the reference height
or this B zone is the height ofthe extension root, H,= 410, and tha the roller
shutter dour les euiely within the C zone with Cye= — 04.
one 1 1 Det
+08 |! “alos 03
L | ls
He ace ton ÊTES
(6) Cose NE
¡o
04 [08 oa | |
il
Eyes] Eu
NE face ET SW face
(0) Case Nw
sn,
cn T 1
03 os [à
sr [as
NE face NW face
(6) Caso sw
4.54 External pressures
Example 36 gives the unfactored external pressures e
zones of Example 35. The highest suction of 1837 Pu or Loly la
the onerous E zone for the NW case, even though this is not the sven ect
wind direction and the dynamic pressure is taken at the beight of ihe
extension. The mex highest suction bus inthe À zone for MO ane
in the strongest wind direction us expected. eres
Example 36. Unfactored external pressures for the walls of the
five-storey office
The unsre etai pressure or each Zone in y
oni asin moine the presse
Cocer Example 38 hy th evan dynamic sue le Table A
The zone of exo peste ar shown below, wie haine Whee the
‘evan dj restr he eg te oc and Nach where te ae
wie, au y ay
pen E $ -185Pa
L
Nein rs
(cone
er nz
8 | 5 ww] | E |:
cero] sors Tefstarn; [estra TA
RE Re 7
NE heu Nice Sas
(0) Caro NW
Worked examples | 209
4.5.3 Pressure coefficients for the walls
The sizes of the pressure coefficient zones and the values of Cpe are
derived in Example 35, The SE case is a mirror image of the NW case. It
may be helpful to compare the NW face for the NE and SW cases with the
definition figure for irregular flush faces, Figure 14, in BS 6399-2.
Note the position of the roller-shutter door on the SW face of the
extension relative to the zones. For the NE case it lies on the leeward face
where the external pressure is controlled by the leeward face of the office,
ie. Cye= —03 and H,= 15m. In the NW and SE cases, it lies just inside
the C zone on the side face and for the SW case itis on the windward face,
where H,=4m. This controls the intemal pressure of the lower storey
when the door is open.
Example 38. Pressure coefficients for the walls of the
storey office
“Two special considerations apply.
© regular Mush face §2.44.1 applies to the NW and SE faces, but only the NW
face is shown, Note thal the diferent scaling lengths b for Pans 1 and 2 (see
Example 34) cause different sizes of zones an the NE face.
© Inset storey 82.44.21) applies to the SW face and the additional E zone with
Gye = — 20 is required forthe NW and SB cases. Note that the reference height
or this B zone is the height ofthe extension root, H,= 410, and tha the roller
shutter dour les euiely within the C zone with Cye= — 04.
one 1 1 Det
+08 |! “alos 03
L | ls
He ace ton ÊTES
(6) Cose NE
¡o
04 [08 oa | |
il
Eyes] Eu
NE face ET SW face
(0) Case Nw
sn,
cn T 1
03 os [à
sr [as
NE face NW face
(6) Caso sw
4.54 External pressures
Example 36 gives the unfactored external pressures e
zones of Example 35. The highest suction of 1837 Pu or Loly la
the onerous E zone for the NW case, even though this is not the sven ect
wind direction and the dynamic pressure is taken at the beight of ihe
extension. The mex highest suction bus inthe À zone for MO ane
in the strongest wind direction us expected. eres
Example 36. Unfactored external pressures for the walls of the
five-storey office
The unsre etai pressure or each Zone in y
oni asin moine the presse
Cocer Example 38 hy th evan dynamic sue le Table A
The zone of exo peste ar shown below, wie haine Whee the
‘evan dj restr he eg te oc and Nach where te ae
wie, au y ay
pen E $ -185Pa
L
Nein rs
(cone
er nz
8 | 5 ww] | E |:
cero] sors Tefstarn; [estra TA
RE Re 7
NE heu Nice Sas
(0) Caro NW
10 BS 6399-2
210 | Wind loading: a practical gui
Worked examples | 211
arte, za assez
£
-199Pa ER oe
oro ERES _
Ne toe iw taco swe
19) Caso SW
pressure is athe height of he extension. The SR exe is not shown, but the zones are
the mios Image of the NW case and lower dynamic pressures apply
[Note that these are unfuctored pressures (C,= 1-0) and so apply 10 the smallest
areas of cladding. You will need 10 apply the size effet factor Ca appropriate to the
size ofthe tributary area when you accumulate the load.
“The zone dimensions have beer omited for cloiy, but are given in Example 36
45.5 Internal pressures
45.5.1 Upper storeys and ground floor, ultimate limit
Tf we assume that the cladding of the building has the same porosity on
all faces, we have a situation very similar to Example 12. However, the
permeability is not equal on each face because the area of the main faces
is twice the area of the small faces. ‘This was explained in 3.4.3. In
Example 37 we determine the internal pressure of the upper storeys and
of the ground floor for the SW and NW eases from the balance of flow
Equation 18, using the unfactored external pressures we obtaincd in
Example 36. Following §2.6./.3, we assume that the large roller-shutter
door in the SW face of the extension is closed at the ultimate limit.
Example 37. Unfactored internal pressure for the five-storey
office
‘We shall take each storey to he open-plan and determine the intemal pressure forthe
[NW and SW cases from the balance of flows. This is equivalent to Example 12 except
thal the area ofthe openings is taken as proportional ro the widih of each zane, making
the flow through each opening equal to = zone wii x (pe—p,)" for inflow and
@=zane width x Gp)? Tor outflow.
For the upper storeys of the office and the ground floor in the SW case, both sie
walls have the same zones, sn the zone widihs have been added together. For the
ground floor inthe NW ease, the zones differ between the office and extension walls,
s0 have been tented separately. In this case the subscripts "0" and “ES refer Lu the
office and extension respectively.
Upper re =
sw Fret Sie al ne Rew
nr =
bem su sare
Se wth 3 onde
Be Que
wo In tah n= 030
Dm red
ee penn
we Trot Si wale
var
mur AS M6 er rom type 35
Fa a ae 0 eb ie
BR Que Oe Que FO
33 “ion Sms ir OS 7386 on ini n= 030
era er un me E nen
Mr e o
Fort Seile heu
= or
pm as cm = Fr Barge 38
PN er) » tes od se
BP a On Qu De
Sie et aaa “2 Ae oro Int y= 036%
D uni “387 Lisa 3 Teor
Fe re Aa Ge mein
nw Frets Pr Be M Gn Ce Reim Ree
pete aes Sse Se Sw Sta Se SPS
Le wir m à Be Be Ys
RE" as Oran Go Un A do Oe Ee
Ts ST er Miro ra Son Bi GF
CRT ais so Sr
CE 3160 Se Bsa
With a spreadsheet program that automatically iterates "circular references, such as
‘Microsoft Excel, you can obtain the results in a single step as shown in the following
table
A BG D 8 6 6 OH OU OU ka
1 tome Pom Pro An Ar Me Me es Roue
O
Date ET SE “Ish Me
IE Givers Ona Dan Où Ou Om Go Ge Oar Oe TC
ES Sie “Hse Sha “Ae MS MR M4 0
210 | Wind loading: a practical gui
Worked examples | 211
arte, za assez
£
-199Pa ER oe
oro ERES _
Ne toe iw taco swe
19) Caso SW
pressure is athe height of he extension. The SR exe is not shown, but the zones are
the mios Image of the NW case and lower dynamic pressures apply
[Note that these are unfuctored pressures (C,= 1-0) and so apply 10 the smallest
areas of cladding. You will need 10 apply the size effet factor Ca appropriate to the
size ofthe tributary area when you accumulate the load.
“The zone dimensions have beer omited for cloiy, but are given in Example 36
45.5 Internal pressures
45.5.1 Upper storeys and ground floor, ultimate limit
Tf we assume that the cladding of the building has the same porosity on
all faces, we have a situation very similar to Example 12. However, the
permeability is not equal on each face because the area of the main faces
is twice the area of the small faces. ‘This was explained in 3.4.3. In
Example 37 we determine the internal pressure of the upper storeys and
of the ground floor for the SW and NW eases from the balance of flow
Equation 18, using the unfactored external pressures we obtaincd in
Example 36. Following §2.6./.3, we assume that the large roller-shutter
door in the SW face of the extension is closed at the ultimate limit.
Example 37. Unfactored internal pressure for the five-storey
office
‘We shall take each storey to he open-plan and determine the intemal pressure forthe
[NW and SW cases from the balance of flows. This is equivalent to Example 12 except
thal the area ofthe openings is taken as proportional ro the widih of each zane, making
the flow through each opening equal to = zone wii x (pe—p,)" for inflow and
@=zane width x Gp)? Tor outflow.
For the upper storeys of the office and the ground floor in the SW case, both sie
walls have the same zones, sn the zone widihs have been added together. For the
ground floor inthe NW ease, the zones differ between the office and extension walls,
s0 have been tented separately. In this case the subscripts "0" and “ES refer Lu the
office and extension respectively.
Upper re =
sw Fret Sie al ne Rew
nr =
bem su sare
Se wth 3 onde
Be Que
wo In tah n= 030
Dm red
ee penn
we Trot Si wale
var
mur AS M6 er rom type 35
Fa a ae 0 eb ie
BR Que Oe Que FO
33 “ion Sms ir OS 7386 on ini n= 030
era er un me E nen
Mr e o
Fort Seile heu
= or
pm as cm = Fr Barge 38
PN er) » tes od se
BP a On Qu De
Sie et aaa “2 Ae oro Int y= 036%
D uni “387 Lisa 3 Teor
Fe re Aa Ge mein
nw Frets Pr Be M Gn Ce Reim Ree
pete aes Sse Se Sw Sta Se SPS
Le wir m à Be Be Ys
RE" as Oran Go Un A do Oe Ee
Ts ST er Miro ra Son Bi GF
CRT ais so Sr
CE 3160 Se Bsa
With a spreadsheet program that automatically iterates "circular references, such as
‘Microsoft Excel, you can obtain the results in a single step as shown in the following
table
A BG D 8 6 6 OH OU OU ka
1 tome Pom Pro An Ar Me Me es Roue
O
Date ET SE “Ish Me
IE Givers Ona Dan Où Ou Om Go Ge Oar Oe TC
ES Sie “Hse Sha “Ae MS MR M4 0
212 | Wind loading: a practical guide to BS 6399-2
Worked examples | 213
L. Layout the spreadsheet as above,
2. tn cell AS insert the initial value for pi = — 188 (Pa)
3. In cell BS insert the Formula =1F(B3>$A5,R2°SORTNI-SAS)-BA"SORT
SAD).
4. Copy this formula to cells CS to KS.
5. Sum the net inflow by inserting the formula =SUM(BS..KS) in cell LS.
{6 Now start the iteration by inserting the formula — AS+LS°001 in cell AS.
Step 6 insets ‘circular reference" that uds u small uen of the inflow to itself
and this will continue unlit there is 2210 net inflow or the maximum number if
iterations set forthe spreadsheet Is reached, IF you. not achieve balance, increase the
maximum member of iterations and reduce the value of the factor 001
® Upper storeys. With wind normal to the main fuce, cases SW and
NE, we expect the intemal pressure coefficient to lic between the
values in Table 16 of Cy; +0-2 for ‘two opposite walls equally
permeable, other faces impermeable, wind normal lo permeable
face” and'Cy=—0-3 for ‘four walls equally permeable’. The
internal pressure we obtain for the SW case, 7,= — 171 Pa, implies
an internal pressure coefficient of Cyj== — 0-27. With wind parallel
(0 the main face, cases NW and SE, we expect the internal pressure
coefficient to be more negative than Gi — 03 for “four walls
equally permeable’ because the windward wall is smaller and su
contributes less inflow from the positive pressure. ‘The internal
pressure we obtain for the NW case, p,= —250Pa, implies an
internal pressure coefficient of Co — 0-40.
+ Ground floor. We assume that the extension is connected through
10 the ground floor of the office. he SW case is similar to the
upper storeys. except that the zone sizes differ and the external
pressures are less onerous. Now we have two dynamic pressures,
because the reference height for the rear wall is = 15m, while
1, = 41m for the remaining walls (sec Example 35). The NW case
is more complex, with two zones of external pressure on the front
wall and rear walls, and each side having different zones.
45.5.2 Ground floor, serviceability limit
Following $2.6.1.3, we assume that the large roller-shutter door in the
SW face of the extension is closed al the ultimate limit. Following the
Table 9, Unfacored internal pressure on the ground four of the fue storey ofie with
3p 080
Cue NF Cue NW ‘Cue SW
intemal pre pap, SIRO IAE OR
= 107 --9 er
advice in 3,3.4 from BRE Digest 4367 we will use a probability factor of
Se 0-80 at the serviceability limit
The arca of the open door is Ágoge— 12m and this is more than three:
times the remaining openings. The internal pressure is therefore 90% of
the external pressure at the door from Example 36. Values are given for
the thrce principal cases in Table 9.
4.5.6 Loads on cladding panels
45.6.1 Size effect factor
1. External pressure. The office is clad in panels that are 2-5m wide
by 3m high and the extension is clad in panels 2:5m wide by 4m
high. In both cases the diagonal dimension is less than Sm and
2. Internal pressure, door elased. We assume that the volume of
the floor is given by the dimensions of the extension
(20mx5mx4m) and the ground floor of the office
(20m x 10m x 3m), giving O= 100m. From Equation 13 the
diagonal dimension a =10 x(1000)'"*— 100m, giving
Ca 70-775 for the NE case (line "B') and Cy =0:730 tor the
other cases from Figure 4.
3. Internal pressure, door open. With the door open, we require the
‘greater diagonal dimension of:
o sonal of the dour: a — (4? + 32)"
O or from Equation 15: a
which gives Ca= 1-0, ie. there is no reduction from the unfactored
internal pressure.
45.6.2. Highest-loaded panels, ultimate limit
1.” Office. The highest-loaded 2.5m by 3 m pancl on the upper storeys
of the office is either the panel that spans the 2m wide E zone in
the NW case where the suction is the most onerous but part of the
panel is in the less oncrous B zone:
Po=(-847x3x2--5U2X3x0:5) N —(-251x0-730:3x2-5)N
= —5835KN+1:374KN =4-46kN
Worked examples | 213
L. Layout the spreadsheet as above,
2. tn cell AS insert the initial value for pi = — 188 (Pa)
3. In cell BS insert the Formula =1F(B3>$A5,R2°SORTNI-SAS)-BA"SORT
SAD).
4. Copy this formula to cells CS to KS.
5. Sum the net inflow by inserting the formula =SUM(BS..KS) in cell LS.
{6 Now start the iteration by inserting the formula — AS+LS°001 in cell AS.
Step 6 insets ‘circular reference" that uds u small uen of the inflow to itself
and this will continue unlit there is 2210 net inflow or the maximum number if
iterations set forthe spreadsheet Is reached, IF you. not achieve balance, increase the
maximum member of iterations and reduce the value of the factor 001
® Upper storeys. With wind normal to the main fuce, cases SW and
NE, we expect the intemal pressure coefficient to lic between the
values in Table 16 of Cy; +0-2 for ‘two opposite walls equally
permeable, other faces impermeable, wind normal lo permeable
face” and'Cy=—0-3 for ‘four walls equally permeable’. The
internal pressure we obtain for the SW case, 7,= — 171 Pa, implies
an internal pressure coefficient of Cyj== — 0-27. With wind parallel
(0 the main face, cases NW and SE, we expect the internal pressure
coefficient to be more negative than Gi — 03 for “four walls
equally permeable’ because the windward wall is smaller and su
contributes less inflow from the positive pressure. ‘The internal
pressure we obtain for the NW case, p,= —250Pa, implies an
internal pressure coefficient of Co — 0-40.
+ Ground floor. We assume that the extension is connected through
10 the ground floor of the office. he SW case is similar to the
upper storeys. except that the zone sizes differ and the external
pressures are less onerous. Now we have two dynamic pressures,
because the reference height for the rear wall is = 15m, while
1, = 41m for the remaining walls (sec Example 35). The NW case
is more complex, with two zones of external pressure on the front
wall and rear walls, and each side having different zones.
45.5.2 Ground floor, serviceability limit
Following $2.6.1.3, we assume that the large roller-shutter door in the
SW face of the extension is closed al the ultimate limit. Following the
Table 9, Unfacored internal pressure on the ground four of the fue storey ofie with
3p 080
Cue NF Cue NW ‘Cue SW
intemal pre pap, SIRO IAE OR
= 107 --9 er
advice in 3,3.4 from BRE Digest 4367 we will use a probability factor of
Se 0-80 at the serviceability limit
The arca of the open door is Ágoge— 12m and this is more than three:
times the remaining openings. The internal pressure is therefore 90% of
the external pressure at the door from Example 36. Values are given for
the thrce principal cases in Table 9.
4.5.6 Loads on cladding panels
45.6.1 Size effect factor
1. External pressure. The office is clad in panels that are 2-5m wide
by 3m high and the extension is clad in panels 2:5m wide by 4m
high. In both cases the diagonal dimension is less than Sm and
2. Internal pressure, door elased. We assume that the volume of
the floor is given by the dimensions of the extension
(20mx5mx4m) and the ground floor of the office
(20m x 10m x 3m), giving O= 100m. From Equation 13 the
diagonal dimension a =10 x(1000)'"*— 100m, giving
Ca 70-775 for the NE case (line "B') and Cy =0:730 tor the
other cases from Figure 4.
3. Internal pressure, door open. With the door open, we require the
‘greater diagonal dimension of:
o sonal of the dour: a — (4? + 32)"
O or from Equation 15: a
which gives Ca= 1-0, ie. there is no reduction from the unfactored
internal pressure.
45.6.2. Highest-loaded panels, ultimate limit
1.” Office. The highest-loaded 2.5m by 3 m pancl on the upper storeys
of the office is either the panel that spans the 2m wide E zone in
the NW case where the suction is the most onerous but part of the
panel is in the less oncrous B zone:
Po=(-847x3x2--5U2X3x0:5) N —(-251x0-730:3x2-5)N
= —5835KN+1:374KN =4-46kN
214 | Wind loading: a practical guide to BS 6399-2
Worked examples | 215
er any of the pancls within the 4 m wide A zane of the SW case, where the
‘external suction is less oncrous but the internal pressure is less beneficial:
171 0-730%3x2-5)N,
33KN
Po=(-8373:2:5)N |
263kN +0-936KN
2. Extension. The highest-loaded 2-5 m by 4 m panel on the extension
spans the 1-6 m wide A zone and is partly in the B zone on the SW
face in the NW ease:
Pym (-550xdx1-6-339x4x09)N — (1380-7304 X25) N
4740 KN + 1-007 KN — 3-73 KN
Note: This is not the highest-Inaded panel on the ground floor. That
distinction goes to the 2.5m by 3m office panel in the A zone on
the NE face in the NW case (pe= —816 Pa).
4563 Highest-loaded panels, serviceability limit
‘Now we need to check (0 see whether the highest pane! loads at the
ultimate limit are exceeded at the service limit when the roller-shutter
door is open. The SW case is the only case where the door is in a zone of
positive external pressure.
1. Suction. The highest-loaded 2-5 m by 4m panel on the extension
spans the 1-6m wide A zone and part of the B zane:
Pa (-447x4x1.6-293x4x09) x 0-8? N -(HISBXIXZSIN
2-504 KN + 1-580kN 4.08 KN
‘The highest-loaded 2-5 m hy 3m panel on the office wall lies in the
B zune (pe= -293 Pa}:
Poy= (-293:3x2:5) x 0-8? N (158% 32:5) N
= HAOGKN= 1-185 KN =2:23 KN
We see thatthe service limit controls the design against suction of
the ground oor panels on the extension, but not on the office.
2. Pressure. For completeness, we should also check the maximum
positive panel loads:
Po=(+465x3x2:5)x0.82N -{-169x3x25)N
= 4 2232KN + 1:268KN=3:50kN
Pa =(+254x4x2:5)x08N —(—169:4x2:5)N
—+1-626KN + 1-690 KN=3-32KN
£-20m weten
47m
He 6m
Cn A
Fig. 98. Dimensions ofthe tower and prdium
4.6 Tower and podium
4.6.1 The tower and podium example
‘The dimensions of the building are given in Kig. 98. For consistency,
we use the same site location at SK800477 for this example, even though
it is unlikely that a development of this size would be built on the
boundary of a town, We shall assume that:
the tower has a central core of reinforced concrete
storey height is 3 m for tower and podium
the major axis of the tower is aligned to 9 =4
the surrounding buildings are a mixture of two and three storeys
(2:5 storeys on average) and that their spacing is unknown, but
‘ypical
ves an obstruction height Hy=2:5x3=7-5m and a spacing
10m, We shall further assume, since the spacing is unknown, that
funnelling is maximum along the sides of the podium.
We shal use this example to demonstrate the pressure coefficient zones
on the tower and podium and the shear force at each storey level in
conjunction with the division-by-parts rule of §2.2.3.2 (see 2.4.4). As this
shear force depends on the response of the tower, which may be
sufficiently dynamic to be out of the scope of BS 6399-2, we will start
with the dynamic classification,
4.6.2 Dynamic classification
‘The dynamic classification of the tower is shown in Example 38, Note
that the classification usos the height of the tower above ground level
because the concrete core of the tower is assumed to continus through the
podium to a foundation at ground level. The tower is shown to be 5%
dynamic, 50 is well within the scope of BS 6399-2.
Worked examples | 215
er any of the pancls within the 4 m wide A zane of the SW case, where the
‘external suction is less oncrous but the internal pressure is less beneficial:
171 0-730%3x2-5)N,
33KN
Po=(-8373:2:5)N |
263kN +0-936KN
2. Extension. The highest-loaded 2-5 m by 4 m panel on the extension
spans the 1-6 m wide A zone and is partly in the B zone on the SW
face in the NW ease:
Pym (-550xdx1-6-339x4x09)N — (1380-7304 X25) N
4740 KN + 1-007 KN — 3-73 KN
Note: This is not the highest-Inaded panel on the ground floor. That
distinction goes to the 2.5m by 3m office panel in the A zone on
the NE face in the NW case (pe= —816 Pa).
4563 Highest-loaded panels, serviceability limit
‘Now we need to check (0 see whether the highest pane! loads at the
ultimate limit are exceeded at the service limit when the roller-shutter
door is open. The SW case is the only case where the door is in a zone of
positive external pressure.
1. Suction. The highest-loaded 2-5 m by 4m panel on the extension
spans the 1-6m wide A zone and part of the B zane:
Pa (-447x4x1.6-293x4x09) x 0-8? N -(HISBXIXZSIN
2-504 KN + 1-580kN 4.08 KN
‘The highest-loaded 2-5 m hy 3m panel on the office wall lies in the
B zune (pe= -293 Pa}:
Poy= (-293:3x2:5) x 0-8? N (158% 32:5) N
= HAOGKN= 1-185 KN =2:23 KN
We see thatthe service limit controls the design against suction of
the ground oor panels on the extension, but not on the office.
2. Pressure. For completeness, we should also check the maximum
positive panel loads:
Po=(+465x3x2:5)x0.82N -{-169x3x25)N
= 4 2232KN + 1:268KN=3:50kN
Pa =(+254x4x2:5)x08N —(—169:4x2:5)N
—+1-626KN + 1-690 KN=3-32KN
£-20m weten
47m
He 6m
Cn A
Fig. 98. Dimensions ofthe tower and prdium
4.6 Tower and podium
4.6.1 The tower and podium example
‘The dimensions of the building are given in Kig. 98. For consistency,
we use the same site location at SK800477 for this example, even though
it is unlikely that a development of this size would be built on the
boundary of a town, We shall assume that:
the tower has a central core of reinforced concrete
storey height is 3 m for tower and podium
the major axis of the tower is aligned to 9 =4
the surrounding buildings are a mixture of two and three storeys
(2:5 storeys on average) and that their spacing is unknown, but
‘ypical
ves an obstruction height Hy=2:5x3=7-5m and a spacing
10m, We shall further assume, since the spacing is unknown, that
funnelling is maximum along the sides of the podium.
We shal use this example to demonstrate the pressure coefficient zones
on the tower and podium and the shear force at each storey level in
conjunction with the division-by-parts rule of §2.2.3.2 (see 2.4.4). As this
shear force depends on the response of the tower, which may be
sufficiently dynamic to be out of the scope of BS 6399-2, we will start
with the dynamic classification,
4.6.2 Dynamic classification
‘The dynamic classification of the tower is shown in Example 38, Note
that the classification usos the height of the tower above ground level
because the concrete core of the tower is assumed to continus through the
podium to a foundation at ground level. The tower is shown to be 5%
dynamic, 50 is well within the scope of BS 6399-2.
216 | Wind loading: a practical guide to BS 6399-2
Worked examples | 217
m
Example 38. Dynamic classification of the tower
Action ‘Notes
rg height above its base:
81.332 Use height top of
paraperfrom ground
51.51. When in doubt tae nest
Read value of Ka from Table 1
larger value
Using H and Ky read C; from Figure 2 C 5161 1C,>04 get beter value
rom Annex €.
Check C,<025 os 5162 DN 6399 can be uned
Thames mas the toner car extends through he podium othe foundations at ground level
Ee
4.6.3. Scaling length b and division-by-parts
"The sealing lengths for the tower and podium are shown in Example
39. Following §2.2.3.2 allows the tower to be split into the parts shown For
‘the purposes of determining overall structural loads (sec 2.4.4). When the
middle part exists, it may be split into as many horizontal slices as
desired, but we shall align the slices to the 3m storey heights,
Ne eS
Example 39. Scaling dimensions for the tower and podium.
Wind angle, 0 des Mates
oo
Tower
Heigl of tower, hm 41 4 813.32 From roof of podium
Hresdih, Bam 20 15 HA
Depth, 15 20 51344
Aspect ratio, HB 21 27 92242 Bosh HIB>2, See Figure 11
Span rato, DA 037 049 82.412 DH see Table 5
Sealing length. :m 20 15 82413 Smaller of B or 207
Parapet height 0 2 2 325.14 Assume 2m high parapet
Prapet height ratio, WO 043 See Table 8
Podium
Heigit of podium, Him 6 6 $1332 To roof af podium
Bresdth, Bem 4 40 3
Depth, Dem 40 40 81344
Aspect cali, HB 945 015 82232 Both H/B< I, one pan,
‘span ratio, D/H 10 10 82412 Di >4, sce Table 5.
Scaling length, 1212 $2404 Smaller of B or 26
lowing $22.32 allows the tower to be split into the parts shown below for the
purposes of determining overall structural lads. When the middle part exists, e may
be ph int us many horizontal lies as desired, We shall lig the ses tothe 3m
storey helt.
h = 47m 4,
7m,
en
= pon
IT 12
(a) Wind angio @ = 0° (6) Wind angle 9 = 0"
4.6.4 External pressure coefficients
‘The extemal pressure coefficient zones and values on the walls and
roof of the tower and podium are shown in Example 40. Clause 2.5.1.7
requires that we consider the effect af the presence of the tower on the
pressure coefficients for the podium roof. The zones of additional
pressure coefficients from $2.5.1.7 must be used with the dynamic
pressure fur the lower walls, Because the height of the tower gives a large
dynamic pressure and the pressure coefficients are larger than fur the D
zone in Table 8 that would otherwise occupy most of the podium roof,
these atonal zones wl cool e design press on he podi
1001
Example 40. Pressure coefficients for the tower and podium
‘The external pressure coefficient zones and values on the walls of the tower and
podium are shown in the following figures. The tower is shown truncated to save
Space, but the zones span the full height to the top of the parapet. Although we
‘assuming maximum funnelfing along the sides of the podium, this will not occur
along the NE face because the exposure is open and there are no neighbouring
buildings
Worked examples | 217
m
Example 38. Dynamic classification of the tower
Action ‘Notes
rg height above its base:
81.332 Use height top of
paraperfrom ground
51.51. When in doubt tae nest
Read value of Ka from Table 1
larger value
Using H and Ky read C; from Figure 2 C 5161 1C,>04 get beter value
rom Annex €.
Check C,<025 os 5162 DN 6399 can be uned
Thames mas the toner car extends through he podium othe foundations at ground level
Ee
4.6.3. Scaling length b and division-by-parts
"The sealing lengths for the tower and podium are shown in Example
39. Following §2.2.3.2 allows the tower to be split into the parts shown For
‘the purposes of determining overall structural loads (sec 2.4.4). When the
middle part exists, it may be split into as many horizontal slices as
desired, but we shall align the slices to the 3m storey heights,
Ne eS
Example 39. Scaling dimensions for the tower and podium.
Wind angle, 0 des Mates
oo
Tower
Heigl of tower, hm 41 4 813.32 From roof of podium
Hresdih, Bam 20 15 HA
Depth, 15 20 51344
Aspect ratio, HB 21 27 92242 Bosh HIB>2, See Figure 11
Span rato, DA 037 049 82.412 DH see Table 5
Sealing length. :m 20 15 82413 Smaller of B or 207
Parapet height 0 2 2 325.14 Assume 2m high parapet
Prapet height ratio, WO 043 See Table 8
Podium
Heigit of podium, Him 6 6 $1332 To roof af podium
Bresdth, Bem 4 40 3
Depth, Dem 40 40 81344
Aspect cali, HB 945 015 82232 Both H/B< I, one pan,
‘span ratio, D/H 10 10 82412 Di >4, sce Table 5.
Scaling length, 1212 $2404 Smaller of B or 26
lowing $22.32 allows the tower to be split into the parts shown below for the
purposes of determining overall structural lads. When the middle part exists, e may
be ph int us many horizontal lies as desired, We shall lig the ses tothe 3m
storey helt.
h = 47m 4,
7m,
en
= pon
IT 12
(a) Wind angio @ = 0° (6) Wind angle 9 = 0"
4.6.4 External pressure coefficients
‘The extemal pressure coefficient zones and values on the walls and
roof of the tower and podium are shown in Example 40. Clause 2.5.1.7
requires that we consider the effect af the presence of the tower on the
pressure coefficients for the podium roof. The zones of additional
pressure coefficients from $2.5.1.7 must be used with the dynamic
pressure fur the lower walls, Because the height of the tower gives a large
dynamic pressure and the pressure coefficients are larger than fur the D
zone in Table 8 that would otherwise occupy most of the podium roof,
these atonal zones wl cool e design press on he podi
1001
Example 40. Pressure coefficients for the tower and podium
‘The external pressure coefficient zones and values on the walls of the tower and
podium are shown in the following figures. The tower is shown truncated to save
Space, but the zones span the full height to the top of the parapet. Although we
‘assuming maximum funnelfing along the sides of the podium, this will not occur
along the NE face because the exposure is open and there are no neighbouring
buildings
218 | Wind loading: a practical guide to WS 6399-2
Worked examples | 219
.
a
e e, E
HA
(8) Wind angle @
[EN
L.£ 51 L
Leeward face
[
‘Windward face
(b) Wind angle @= 90°
“The principal zones and values forthe external pressure cuctficiens on ne mals of
the tower and podium are shown below. Note that Ihe parapet of the tower reduces
values in the A and B zones by a diferent amount in each wind direction because the
ratio Mb dife i
1) Wind angle = oF (0) Wind angle 8 = 90"
Clause 2.5.17 requires tat we consider the effect ofthe presence of ths tower on
the pressure coefficients for the podium roof. The zones nf aukitional pressure
Les Ei
(a) Wied angle 0 20 (9) Wind ange = 80"
coetriclents from 42.5.7 ae elven next and must he used with the dynamic pressure
for the tower walls. Because the height of the tower gives a large dynar
and the pressure coelcions are larger than for the principal Ha roof zones, these
additional zones will control the design pressures on the podium oo.
4.6.5 Dynamic pressures
‘The dynamic pressures corresponding to the top of each of the parts of
each of the tower and podium are shown in Table 10, with a description
of each part. As before, these are the values you should obtain using
Option 3— Directional method, for the site at SK800477 with the
obstruction heights and spacing assumed above.
‘As noted in 4.5.1 for the five-storey office example, the shelter given
hy the zero-plane displacement in the SW case reduces the dynamic
pressure more at the lower levels than al the higher levels. Consequently,
the SW case has the least dynamic pressure at H, 6m, but the greatest
dynamic pressure at all other heights. The critical cases for cladding
pressures on the walls and the principal roof zones on th podium are the
NE and NW cases, The crtical case for the tower and the additional zones
on the podium roof is the SW cas
4.6.6 Horizontal shear forces at the base of each storey of the tower
Calculations of the horizontal shear force at the hase of each storey of
the tower are given in Example 41. The tower is divided into the parts
defined in Example 39. The dynamic pressure for the top of each part is
taken from Table 10. Figure Ste) uf BS 6309-2 shows that the diagonal
dimension for the horizontal shear at any level is the diagonal of the
loaded area above that level in a horizontal projection.
Table 10. Dynamic pressures q, (in Pa) ar SKAUOS7? far parts ofthe tower and ptm
3
NE SE SW NW Application
6 473 4327 4070 4793 Pod, scr and cladding
a 008 9409 Lowest part of unser. 0 — 97
2 1 10035 Ist storey of ie pat of ame, © 9
26 7829 10174 Lancet pot of unser 0-0
27 1093 7961 10581 10344 — Middle pan of unser. 0-0". 2nd stores of
mite part of waver, 090°
0 7266 11065 rd storey af mile par of tower, 0 =
32 768 11360 Top of middle par of tower, 3 = 90"
4
JOT] 9282 12778 12824 Upper par af over, 807 and 90" amd
lading of wc
‘Notes Values tar are not needed for storey shear culeetions have been left Blank
Worked examples | 219
.
a
e e, E
HA
(8) Wind angle @
[EN
L.£ 51 L
Leeward face
[
‘Windward face
(b) Wind angle @= 90°
“The principal zones and values forthe external pressure cuctficiens on ne mals of
the tower and podium are shown below. Note that Ihe parapet of the tower reduces
values in the A and B zones by a diferent amount in each wind direction because the
ratio Mb dife i
1) Wind angle = oF (0) Wind angle 8 = 90"
Clause 2.5.17 requires tat we consider the effect ofthe presence of ths tower on
the pressure coefficients for the podium roof. The zones nf aukitional pressure
Les Ei
(a) Wied angle 0 20 (9) Wind ange = 80"
coetriclents from 42.5.7 ae elven next and must he used with the dynamic pressure
for the tower walls. Because the height of the tower gives a large dynar
and the pressure coelcions are larger than for the principal Ha roof zones, these
additional zones will control the design pressures on the podium oo.
4.6.5 Dynamic pressures
‘The dynamic pressures corresponding to the top of each of the parts of
each of the tower and podium are shown in Table 10, with a description
of each part. As before, these are the values you should obtain using
Option 3— Directional method, for the site at SK800477 with the
obstruction heights and spacing assumed above.
‘As noted in 4.5.1 for the five-storey office example, the shelter given
hy the zero-plane displacement in the SW case reduces the dynamic
pressure more at the lower levels than al the higher levels. Consequently,
the SW case has the least dynamic pressure at H, 6m, but the greatest
dynamic pressure at all other heights. The critical cases for cladding
pressures on the walls and the principal roof zones on th podium are the
NE and NW cases, The crtical case for the tower and the additional zones
on the podium roof is the SW cas
4.6.6 Horizontal shear forces at the base of each storey of the tower
Calculations of the horizontal shear force at the hase of each storey of
the tower are given in Example 41. The tower is divided into the parts
defined in Example 39. The dynamic pressure for the top of each part is
taken from Table 10. Figure Ste) uf BS 6309-2 shows that the diagonal
dimension for the horizontal shear at any level is the diagonal of the
loaded area above that level in a horizontal projection.
Table 10. Dynamic pressures q, (in Pa) ar SKAUOS7? far parts ofthe tower and ptm
3
NE SE SW NW Application
6 473 4327 4070 4793 Pod, scr and cladding
a 008 9409 Lowest part of unser. 0 — 97
2 1 10035 Ist storey of ie pat of ame, © 9
26 7829 10174 Lancet pot of unser 0-0
27 1093 7961 10581 10344 — Middle pan of unser. 0-0". 2nd stores of
mite part of waver, 090°
0 7266 11065 rd storey af mile par of tower, 0 =
32 768 11360 Top of middle par of tower, 3 = 90"
4
JOT] 9282 12778 12824 Upper par af over, 807 and 90" amd
lading of wc
‘Notes Values tar are not needed for storey shear culeetions have been left Blank
220 | Wind loading: a practical guide to BS 6399.2
nn
Example 41. Horizontal shear force at the base of each storey of
the tower
“the calculation of horizontal shear Py a the base of cach storey (including the height
‘of the parapet) above the level of the podium roof is as follows:
+ q. s the dynamie pressure foreach storey from ‘Fable $0
Pasay i the anfactored horizontal joa on each storey, Le. with C,
+ "ne diagonal dimension of the loaded area above the base ofthe storey
© ais be sizeelfect factor of the loaded ara, using line “B' of Figure 4 above
10 m and line °C’ below 10m (see Key to Figure 6)
+ Paris the overall horizontal shear atthe base of each storey from Equation 7 of
tie Standard ($2.1.2.0) using the dynamic pressure q, for the appropriate part
and C005 from the dynamic classification in Example 36
Case NW, 0 =0" Caso SW, 8= 90
Storey ele Pa AN Ge PAN GP Pl
Is ma 1978 206 0993 110 12778 a 158 0913 M
li usa 827 215080 5 12778 020 110 0908 14
13 24 827 228 0885 29 1218 620 186 0901 IM
[2 124 827 244 080 30 1278 020 205 038 21
1 {2524 827 262 0875 366 1218 020 27 086 2%
lo 12524 827 292 0860 427 11065 62 250 087% 3%
9 aa 683 305 0866 477 10581 512 275 OST 361
R 1017s 671 323 0258 526 10035 504 200 0865 397
7 na G71 252 0853 575 9198 504 326 OTK 433
Ela 671377 0847 G2 DIS S04 353 0852 468
5
4
3
Jord 871 403 0942 668 DIR 504 381 0847 50%
504 409 ot SIS
50% 417 0804 550
10174 611 429 0805 686
174 671 456 0800 730
Loads on the parapets should be determined as if they were frec-
standing walls (§2.5.1-4.2) using the provisions of $2.8.1. The dyna
augmentation factor C, does not apply 1 the parapets.
Appendix A. Lattice structures
A.l Lattice frames and trusses
Al Basis
Clause 1.1 Scope of BS 6399-2 permits the use of “other methods’,
provided that they are equivalent, The Standard frequently directs the user
lo Reference 6 for further information. This is the BRE Designer's Guide
10 Wind Loading of Building Structures, Part 2,' from which most of the
pressure cucfficient data were derived.
Like BS 8100, the Designer's Guide! method for lattice str
based on the solid area and solidity ratio as defined in the key
Note that when the latte springs from ground level, the effective length
is twice the height, L=2H, following the principle of reflection
uroxluved cerlicr. By using the windward face as reference, as shown
Fig, A2, the method also extends to three and Fou-boom trusses.
AJ.2 Dynamic pressure
__ BS 6399-2 permits the individual elements of a lattice to be divided
into small parts and the dynamic pressure determined at the centre of each
part. When assessing the lattice as a whole, this is equivalent to taking the
average dynamic pressure over the projected envelope of the lattice, as
shown in Fi vs the solidity ratio of the lattice to be
Fig. A. Kes fur plane frames
nn
Example 41. Horizontal shear force at the base of each storey of
the tower
“the calculation of horizontal shear Py a the base of cach storey (including the height
‘of the parapet) above the level of the podium roof is as follows:
+ q. s the dynamie pressure foreach storey from ‘Fable $0
Pasay i the anfactored horizontal joa on each storey, Le. with C,
+ "ne diagonal dimension of the loaded area above the base ofthe storey
© ais be sizeelfect factor of the loaded ara, using line “B' of Figure 4 above
10 m and line °C’ below 10m (see Key to Figure 6)
+ Paris the overall horizontal shear atthe base of each storey from Equation 7 of
tie Standard ($2.1.2.0) using the dynamic pressure q, for the appropriate part
and C005 from the dynamic classification in Example 36
Case NW, 0 =0" Caso SW, 8= 90
Storey ele Pa AN Ge PAN GP Pl
Is ma 1978 206 0993 110 12778 a 158 0913 M
li usa 827 215080 5 12778 020 110 0908 14
13 24 827 228 0885 29 1218 620 186 0901 IM
[2 124 827 244 080 30 1278 020 205 038 21
1 {2524 827 262 0875 366 1218 020 27 086 2%
lo 12524 827 292 0860 427 11065 62 250 087% 3%
9 aa 683 305 0866 477 10581 512 275 OST 361
R 1017s 671 323 0258 526 10035 504 200 0865 397
7 na G71 252 0853 575 9198 504 326 OTK 433
Ela 671377 0847 G2 DIS S04 353 0852 468
5
4
3
Jord 871 403 0942 668 DIR 504 381 0847 50%
504 409 ot SIS
50% 417 0804 550
10174 611 429 0805 686
174 671 456 0800 730
Loads on the parapets should be determined as if they were frec-
standing walls (§2.5.1-4.2) using the provisions of $2.8.1. The dyna
augmentation factor C, does not apply 1 the parapets.
Appendix A. Lattice structures
A.l Lattice frames and trusses
Al Basis
Clause 1.1 Scope of BS 6399-2 permits the use of “other methods’,
provided that they are equivalent, The Standard frequently directs the user
lo Reference 6 for further information. This is the BRE Designer's Guide
10 Wind Loading of Building Structures, Part 2,' from which most of the
pressure cucfficient data were derived.
Like BS 8100, the Designer's Guide! method for lattice str
based on the solid area and solidity ratio as defined in the key
Note that when the latte springs from ground level, the effective length
is twice the height, L=2H, following the principle of reflection
uroxluved cerlicr. By using the windward face as reference, as shown
Fig, A2, the method also extends to three and Fou-boom trusses.
AJ.2 Dynamic pressure
__ BS 6399-2 permits the individual elements of a lattice to be divided
into small parts and the dynamic pressure determined at the centre of each
part. When assessing the lattice as a whole, this is equivalent to taking the
average dynamic pressure over the projected envelope of the lattice, as
shown in Fi vs the solidity ratio of the lattice to be
Fig. A. Kes fur plane frames
222 | Wind loading: a practical guide to BS 6399-2
Appendix A. Latice structures | 223
A = solid area normal to one face
b= ABD
ig. Ad. Reference face of trusses
need to be assessed in sections that have only a small variation of solidity
across their area,
A1.3. Netpressure coefficients
‘The net pressure coefficient for plane frames with fat-faced, suberitical
and supereritical circular members are given in Fig. A.3, Fig, AA and
Fig. AS, respectively. The net pressure coefficients for the reference face
of square and triangular trusses are given in Figure A.6. Where a frame
‘composed of a mixture of fla, sub- and supercritical elements, the overall
load may be determined from the area and pressure coefficient of each
type:
À Cp= An Cp at Aso pot + Ar Cpsuper AD
Er
Fig. A.3 Nor pressure cocficien for plane frames ith fa elements
esure enefclet for plane frames 1
sete a,
Fig, AS. Net preuure coefficient for plane frenos with supercritical circular elements
ALA Shielding
The drag force on a lattice removes un equal amount of momentu
from the wind flowing through it, producing a ‘wind shadow’. This,
turn, reduces the load on other lattice frames or trusses immediately
downwind, The effect of the wind shadow can be expressed as a shielding
factor given by:
«ext
1-(' = =) (42)
where the solidity ratio € and net pressure coefficient C, refer to Ihe frame
producing the wind shadow, the shielding frame.
Appendix A. Latice structures | 223
A = solid area normal to one face
b= ABD
ig. Ad. Reference face of trusses
need to be assessed in sections that have only a small variation of solidity
across their area,
A1.3. Netpressure coefficients
‘The net pressure coefficient for plane frames with fat-faced, suberitical
and supereritical circular members are given in Fig. A.3, Fig, AA and
Fig. AS, respectively. The net pressure coefficients for the reference face
of square and triangular trusses are given in Figure A.6. Where a frame
‘composed of a mixture of fla, sub- and supercritical elements, the overall
load may be determined from the area and pressure coefficient of each
type:
À Cp= An Cp at Aso pot + Ar Cpsuper AD
Er
Fig. A.3 Nor pressure cocficien for plane frames ith fa elements
esure enefclet for plane frames 1
sete a,
Fig, AS. Net preuure coefficient for plane frenos with supercritical circular elements
ALA Shielding
The drag force on a lattice removes un equal amount of momentu
from the wind flowing through it, producing a ‘wind shadow’. This,
turn, reduces the load on other lattice frames or trusses immediately
downwind, The effect of the wind shadow can be expressed as a shielding
factor given by:
«ext
1-(' = =) (42)
where the solidity ratio € and net pressure coefficient C, refer to Ihe frame
producing the wind shadow, the shielding frame.
224 | Wind loading: a practical guide to BS 6399-2
Appendix A. Lanice structures | 225
ES IM El
= Fume
ect €
Fig. Ad. Net pressure efficient for reference face of tresses
1 2 3
Fig. AZ. Principle of wind shadow behind frames and trusses
‘The concept of wind shadow applics to multiple frames. Figure A.7
shows three plane frames, cach frame producing a wind shadow that adds
lo previous wind shadows, Note that frame 3 is shown only partially,
shielded by the smaller frame 2. The shielding factor for the mh frame of
à series of multiple frames is given by summing over the upwind #1
Frames:
n-(1- Km au) (aay
where for the first upwind frame, ny = 1-0 (no shielding). The shielding.
factor is given more conveniently as a design chart in Figure AB.
ect ag lat tare IC
Fig, AB. Shielding factor n or frames and trues
ALS Application without shielding
© For each circular element, determine whether it is suberitical or
supercritical at the design effective wind speed from the diameter
of the element d in metres:
O subcritical if d V, < 6 m/s
O. supereritical if d Ve > 6 mis
O, if supercritical, determine the effective wind speed at which
sd ms. If this is more
the element is still suberitical from V-
than 70% of the design value Ve you should check the loads at
this wind speed taking all elements to be subcritical.
(1 Determine the area of each kind of clement, flat-faced, subcritical
and supercritical circular, and the total area A of all elements:
© fora plane frame this is the projected arca of all elements
© for a square or triangular truss this is the area of elements in
one face,
Appendix A. Lanice structures | 225
ES IM El
= Fume
ect €
Fig. Ad. Net pressure efficient for reference face of tresses
1 2 3
Fig. AZ. Principle of wind shadow behind frames and trusses
‘The concept of wind shadow applics to multiple frames. Figure A.7
shows three plane frames, cach frame producing a wind shadow that adds
lo previous wind shadows, Note that frame 3 is shown only partially,
shielded by the smaller frame 2. The shielding factor for the mh frame of
à series of multiple frames is given by summing over the upwind #1
Frames:
n-(1- Km au) (aay
where for the first upwind frame, ny = 1-0 (no shielding). The shielding.
factor is given more conveniently as a design chart in Figure AB.
ect ag lat tare IC
Fig, AB. Shielding factor n or frames and trues
ALS Application without shielding
© For each circular element, determine whether it is suberitical or
supercritical at the design effective wind speed from the diameter
of the element d in metres:
O subcritical if d V, < 6 m/s
O. supereritical if d Ve > 6 mis
O, if supercritical, determine the effective wind speed at which
sd ms. If this is more
the element is still suberitical from V-
than 70% of the design value Ve you should check the loads at
this wind speed taking all elements to be subcritical.
(1 Determine the area of each kind of clement, flat-faced, subcritical
and supercritical circular, and the total area A of all elements:
© fora plane frame this is the projected arca of all elements
© for a square or triangular truss this is the area of elements in
one face,
226 1 Wind loading: a practical guide to BS 6399-2 Appendix A. Latice structures | 227
D Determine the ratio L/B and the solidity ratio € = 4/(BL) from
the dimensions £. and B of the lattice envelope, using Figs A.1 and Example A.1. Net pressure coefficient for a whole lattice frame
A.2 as reference.
Look up the value of net pressure coefficient €, from Fig. A.3,
a
Fig. A4, Fig. AS or Fig. A6,
Apply the values of C, to the respective areas of the elements Ap, 50]
For the laico Example 14, shown below:
Ava and Anger: ER
The overall horizontal loud on cach frame P is given by: Bram. _ l
P = (Con An + Cour os + Cong Anne) (aa) ¿20m || I
Bu fl
I l
A.1.6 Application with shielding I l
Perform the steps without shielding given in A.1.5 above. 1 ILE
FT If the frame or truss is composed of differently shaped elements, | iF
calculate Ihe average net pressure coefficient from Cy =(CpawAnst N
+ Crean Art + Cp per Aspe VA: !
M The value of En tor the first, upwind frame doce not change N [
because it is unshielded, | fl
Calculate € Cp; for the first frame and look up the shielding factor N
‘is forthe next frame in Fig. AA. [
O Apply this shielding factor to the net pressure coefficient of the SSS] I)
second frame, m Ca
7 Sum Cr Gy for the first two frames, DE Gp) = Cı Cy Cam Ca
and look up the shielding factor na for the third frame in Fig. A.B. The solid eto
Apply this shielding factor to the net pressure coefficient of the Sr SR, To
third frame, 13 Cp remera 1 pa” = 0652 20-48
Continue this sequence to the downwind frame,
© The overall horizontal load on the nth frame P is given by: 2 The nilio =
° 3 From Figure A.3 the mt pressure coefficient €, 1:43, This fs 139 ess than
Pa = Gi Cyndy as the vale from summing individual clemente in Example 14.
Nore: I the example of Fig. A.7, the third frame is partly shielded by 4. For a secon frame behind this fame
frame one only (12) an frames I an
frame on by (m2) and partly by frames I and 2 (ns). (a) FCC) =0348x1x1-41=0:50,
(b) From Figure A.8 the shiclding factor is» = 0-56,
(© The net pressure coefficient 2 Cy = 1430.56 0.80
5. Fora thin! fans behind these Frames:
9 DUC nCp)= 10-3481 1-43) 40548056 1-43)
=050+028-0.78
(9) From Figure A.8 the shielding factor is x = 0:36,
(© The net pressure coefficient 13 Cy = 1-4320:35 — 031.
A _A _ AA<NIIIMN 0000
Example A. demonstrates the application of these rules for the plane
lattice frame used in Example 14.
D Determine the ratio L/B and the solidity ratio € = 4/(BL) from
the dimensions £. and B of the lattice envelope, using Figs A.1 and Example A.1. Net pressure coefficient for a whole lattice frame
A.2 as reference.
Look up the value of net pressure coefficient €, from Fig. A.3,
a
Fig. A4, Fig. AS or Fig. A6,
Apply the values of C, to the respective areas of the elements Ap, 50]
For the laico Example 14, shown below:
Ava and Anger: ER
The overall horizontal loud on cach frame P is given by: Bram. _ l
P = (Con An + Cour os + Cong Anne) (aa) ¿20m || I
Bu fl
I l
A.1.6 Application with shielding I l
Perform the steps without shielding given in A.1.5 above. 1 ILE
FT If the frame or truss is composed of differently shaped elements, | iF
calculate Ihe average net pressure coefficient from Cy =(CpawAnst N
+ Crean Art + Cp per Aspe VA: !
M The value of En tor the first, upwind frame doce not change N [
because it is unshielded, | fl
Calculate € Cp; for the first frame and look up the shielding factor N
‘is forthe next frame in Fig. AA. [
O Apply this shielding factor to the net pressure coefficient of the SSS] I)
second frame, m Ca
7 Sum Cr Gy for the first two frames, DE Gp) = Cı Cy Cam Ca
and look up the shielding factor na for the third frame in Fig. A.B. The solid eto
Apply this shielding factor to the net pressure coefficient of the Sr SR, To
third frame, 13 Cp remera 1 pa” = 0652 20-48
Continue this sequence to the downwind frame,
© The overall horizontal load on the nth frame P is given by: 2 The nilio =
° 3 From Figure A.3 the mt pressure coefficient €, 1:43, This fs 139 ess than
Pa = Gi Cyndy as the vale from summing individual clemente in Example 14.
Nore: I the example of Fig. A.7, the third frame is partly shielded by 4. For a secon frame behind this fame
frame one only (12) an frames I an
frame on by (m2) and partly by frames I and 2 (ns). (a) FCC) =0348x1x1-41=0:50,
(b) From Figure A.8 the shiclding factor is» = 0-56,
(© The net pressure coefficient 2 Cy = 1430.56 0.80
5. Fora thin! fans behind these Frames:
9 DUC nCp)= 10-3481 1-43) 40548056 1-43)
=050+028-0.78
(9) From Figure A.8 the shielding factor is x = 0:36,
(© The net pressure coefficient 13 Cy = 1-4320:35 — 031.
A _A _ AA<NIIIMN 0000
Example A. demonstrates the application of these rules for the plane
lattice frame used in Example 14.
228 | Wind loading: a practical guide to BS 6399-2
Appendix A. Lattice structures | 229
A2 Unclad building frames
AZI Basis
Unchad frames of rectangular buildings are effcetively two arrays of
plane frames set normal to each principal axis, X and Y. The orthogonal
cases give relatively simple shielding. wind parallel to the X axis,
the projected area ofthe columns and the primary beams normal to the X
axis form one array of frames, giving a drag force in the X direction. With
wind parallel to the Y axis, the projected area of the columns and the
primary beams normal tothe Y axis form the other array of Frames, giving
à drag force in the X direction. So X frames shield other X frames and Y
frames shield other Y frames that lie downwind.
We are often interested only in the tutal forces on the frame for stability
purposes. The sum )°(CyC,) which we ubiain during the shiclding
Calculation represents the coefficient of total drag based on the envelope
area of the orthogonal case Ag BH, giving:
P=4BH Cac) (4.6)
where Js the mcan dynamic pressure over the height of the building frame.
Ifthe solidity of the building frame is not reasonably uniform over its height,
it should be assessed in sections that have only a small variation in solidity
‘Shielding becomes more complex when the wind is skewed, as shown
in Fig, A. We can see that the wind shadows of the X or Y frames shield
parts of both X and Y frames that lie downwind. With only a small wind
angle from normal, the columns down une side of the frame become fully
‘exposed to the wind and this will increase Ihe X component of load above
the orthogonal value. Por typical rectangular building frames, the ratio of
total X force al wind angle 9 to the orthagonal case is given! by:
Pe _ (@ iz
on) cos 20 (A?)
Wind ado x
A
“
x = 5 E
Fig, A, Wind shadows from X and Y frames of weld Bling frome
Wind ani
Fig, 40. Factor an ortogonal leads far uaclod building frames in skewed winds
‘When this is plotted as Fig. A.10, we can see that the maximum X force.
30” when the building frame is long in the wind
re A.10 may be used to obtain the design loads on typical
unelad building frames at any wind angle from the loads calculated for the
‘orthogonal cases,
A2.2. Secondary beams
So far we have considered only the principal heams and columns for
the main structural frame and we have assumed that the momentum lost to
the drag of the upwind frames is fully mixed. Typical building frames will
clude secondary beams spaced al intervals between either the X or Y
principal beams to carry the floors. Because of their close spacing, the
momentum will not be fully mixed and each secondary beam will lie
directly in the wakes of the upwind beam, as indicated in Fig. A.11. These
wakes widen as their momentum is mixed into the flow. An additional
wake-shielding factor 7 for secondary heams is given! by:
118 1-2(4/a)"?
fi
where x is the spacing between beams and d is the depth of the upwind,
shiclding beam, This is plotted as a design chart in Fig. A.12.
‘The relevant reference height for the dynamic pressure is the height of
the secondary beam above the ground, However, where each floor has the
same size, number and spacing of secondary beams, the mean dynami
pressure is still appropriate for the total load.
Aw smaller of { (4.8)
Appendix A. Lattice structures | 229
A2 Unclad building frames
AZI Basis
Unchad frames of rectangular buildings are effcetively two arrays of
plane frames set normal to each principal axis, X and Y. The orthogonal
cases give relatively simple shielding. wind parallel to the X axis,
the projected area ofthe columns and the primary beams normal to the X
axis form one array of frames, giving a drag force in the X direction. With
wind parallel to the Y axis, the projected area of the columns and the
primary beams normal tothe Y axis form the other array of Frames, giving
à drag force in the X direction. So X frames shield other X frames and Y
frames shield other Y frames that lie downwind.
We are often interested only in the tutal forces on the frame for stability
purposes. The sum )°(CyC,) which we ubiain during the shiclding
Calculation represents the coefficient of total drag based on the envelope
area of the orthogonal case Ag BH, giving:
P=4BH Cac) (4.6)
where Js the mcan dynamic pressure over the height of the building frame.
Ifthe solidity of the building frame is not reasonably uniform over its height,
it should be assessed in sections that have only a small variation in solidity
‘Shielding becomes more complex when the wind is skewed, as shown
in Fig, A. We can see that the wind shadows of the X or Y frames shield
parts of both X and Y frames that lie downwind. With only a small wind
angle from normal, the columns down une side of the frame become fully
‘exposed to the wind and this will increase Ihe X component of load above
the orthogonal value. Por typical rectangular building frames, the ratio of
total X force al wind angle 9 to the orthagonal case is given! by:
Pe _ (@ iz
on) cos 20 (A?)
Wind ado x
A
“
x = 5 E
Fig, A, Wind shadows from X and Y frames of weld Bling frome
Wind ani
Fig, 40. Factor an ortogonal leads far uaclod building frames in skewed winds
‘When this is plotted as Fig. A.10, we can see that the maximum X force.
30” when the building frame is long in the wind
re A.10 may be used to obtain the design loads on typical
unelad building frames at any wind angle from the loads calculated for the
‘orthogonal cases,
A2.2. Secondary beams
So far we have considered only the principal heams and columns for
the main structural frame and we have assumed that the momentum lost to
the drag of the upwind frames is fully mixed. Typical building frames will
clude secondary beams spaced al intervals between either the X or Y
principal beams to carry the floors. Because of their close spacing, the
momentum will not be fully mixed and each secondary beam will lie
directly in the wakes of the upwind beam, as indicated in Fig. A.11. These
wakes widen as their momentum is mixed into the flow. An additional
wake-shielding factor 7 for secondary heams is given! by:
118 1-2(4/a)"?
fi
where x is the spacing between beams and d is the depth of the upwind,
shiclding beam, This is plotted as a design chart in Fig. A.12.
‘The relevant reference height for the dynamic pressure is the height of
the secondary beam above the ground, However, where each floor has the
same size, number and spacing of secondary beams, the mean dynami
pressure is still appropriate for the total load.
Aw smaller of { (4.8)
230 | Wind loading: a practical guide to BS 6399-2
Appendix A, Lattice structures | 231
Wind Wehe of secondary naar
“ — u EN
Example A.2. Wind loads on a typical building frame
Im this example we calculate the total horizontal loads on the building Frame shown
below. This fume isa sitmplitied version of an office building with composite floors
and masonry dado walls around each storey. The primary and secondary Y beams
are spaced to carry Im span cold-rolled permanent shntering. The outer Y heams,
VU and ¥4, are deeper than the other Y beans Lo carry the external cladding
We shall calculate the total horizontal Toads on the primary frame in both
orthogonal directions, the maximum X load on the primary frame in skewed winds,
And the orthogonal Y load on the complete frame.
Pe
42m
2051305
Y Y
ec pompa uo x
ig. Ad2. Wake shlelding factor ny Jor horkomual beans
A23 Application
7 Perform the steps in AWS and A.1.6, above, for the orthogonal
ceases of the primary X and Y frames.
O Determine the toral orthogonal Toads on the primary frames in the
X and Y directions from Equation A.6
© Determine the spicing ratio wl of the secondary beams and look.
up the value of the wake-shiclding Factor ny in Fig. A.12.
Determine the net pressure coefficient adjusted for length Cy for
each secondary hcam using the procedure of $2.8 described in
2581
O Apply the shielding factor from the upwind frames, together with
ie wake-shielding factor mw to the net pressure coefficient
adjusted for length xC, 10 give the effective net pressure
cocificient for the secondary beam nn w% Cpe
Add the total load accumulated on the secondary beams tothe total
load on the primary frame for the relevant orthogonal case.
‘This procedure is illustrated in Example A.2.
3650171 poses
> $“
LUN: z
Y
v
MA
Nore: Direct wake-shielding may be applied to primary beams that tie-in
the wakes of secondary beams, hut this requires vou to calculate
the loads on the primary beams separately from the leads an the
cols Th eta benef probably ot wort the extraer
Note: Direct wake-shielding should nor be applied to the primary beams
when secondary beams are not present, Although the primary
beams may lie in wakes, the columns may lie in faster flow between
the wakes, cancelling out any additional wake-shielding,
Appendix A, Lattice structures | 231
Wind Wehe of secondary naar
“ — u EN
Example A.2. Wind loads on a typical building frame
Im this example we calculate the total horizontal loads on the building Frame shown
below. This fume isa sitmplitied version of an office building with composite floors
and masonry dado walls around each storey. The primary and secondary Y beams
are spaced to carry Im span cold-rolled permanent shntering. The outer Y heams,
VU and ¥4, are deeper than the other Y beans Lo carry the external cladding
We shall calculate the total horizontal Toads on the primary frame in both
orthogonal directions, the maximum X load on the primary frame in skewed winds,
And the orthogonal Y load on the complete frame.
Pe
42m
2051305
Y Y
ec pompa uo x
ig. Ad2. Wake shlelding factor ny Jor horkomual beans
A23 Application
7 Perform the steps in AWS and A.1.6, above, for the orthogonal
ceases of the primary X and Y frames.
O Determine the toral orthogonal Toads on the primary frames in the
X and Y directions from Equation A.6
© Determine the spicing ratio wl of the secondary beams and look.
up the value of the wake-shiclding Factor ny in Fig. A.12.
Determine the net pressure coefficient adjusted for length Cy for
each secondary hcam using the procedure of $2.8 described in
2581
O Apply the shielding factor from the upwind frames, together with
ie wake-shielding factor mw to the net pressure coefficient
adjusted for length xC, 10 give the effective net pressure
cocificient for the secondary beam nn w% Cpe
Add the total load accumulated on the secondary beams tothe total
load on the primary frame for the relevant orthogonal case.
‘This procedure is illustrated in Example A.2.
3650171 poses
> $“
LUN: z
Y
v
MA
Nore: Direct wake-shielding may be applied to primary beams that tie-in
the wakes of secondary beams, hut this requires vou to calculate
the loads on the primary beams separately from the leads an the
cols Th eta benef probably ot wort the extraer
Note: Direct wake-shielding should nor be applied to the primary beams
when secondary beams are not present, Although the primary
beams may lie in wakes, the columns may lie in faster flow between
the wakes, cancelling out any additional wake-shielding,
232 | Wind loading: a practical
‘ates
Main structural frame
Side elevation
X tramos
Ham 26
Bm ix
DÉTENU
Frames X1-x6
Am 96
Ge ana) = 0158
Go 169
rame m 0
x 1 16
x 0% 128
x 05 008
Ya 0 081
x 09066
om
x 7
Oahogomai load, Pe = gBH LEGEN 5617
Peay = Pet
Maximum lon
ya vs 0%
Grogonal oad, Py —q BH ECG EN 947
¡Secondary beams om
0305m
25m
12
»s
10
40
Fiat
x= 82
ca 976
Berween ¥ frases 7
nn tay
nn 06
por 056
E mwrG
057
Front eration
snm
ns
5
7
Weve an
tae” Tha
Coe ons
tn tn
np SNC
ST ax
on D
aie van
de kn
Mar 08
toes 00
ay
inc, Den,
bus O2
vis om
Bian 0304
012
025
Gens
ou
08
Forts
070
oa
LE
#57
Pr Coowondary beamo)= NA Dom CoN HOG
I frames and secondary hans: N
10574
eight of envelope
Breadth of envelope
Biene proportions
Projected area of elements
Solty vaio,
From Vigure AS
Application stop in BAG
Soo frames
Earaion 4
Figure 40,019=25
Applica steps in MAS
jor fee
uation A6
Een
Bl
Le
Application steps In A23
Sum uf all secondary Bees
Total Y ou
Appendix B. Corrected factor tables
‘The small errors in Tables 4, 22, 23 and 24 of RS 6399-2: 1997 were
explained in 3.1.2. Corrected values are given in Tables B.I-B.4. These
are consistent with the model from which the Standard was constructed,
given in reference I, the computer program DREVe and the Ready-
Reckoner (set 31.4).
In an effort to keep the 2002 ammendments simple, not all these
conections were made where the differences were small. Accordingly, the
corrected tables in this Appendix remain relevant
Table B.1 Corrected values for Table 4, Sy
‘Site in country, or less than 2m inside town Sie in town, extending > 2km upwind
from the site
Elune Cloves distance wo searkm Effective Closest distance in sex km
height eight
Ham <O1 2 10 100 Hem 2 000m
2 14 ras ZAS 10
5 165 157 E
wm 1 1 14 16 158
15 136 182 15 tee 182 m
et") 10 2010 10 17
50 197 196 20 1 196 185
50 204 20 50 20 2m 105
wo an 212 wo zu 211 20
‘ates
Main structural frame
Side elevation
X tramos
Ham 26
Bm ix
DÉTENU
Frames X1-x6
Am 96
Ge ana) = 0158
Go 169
rame m 0
x 1 16
x 0% 128
x 05 008
Ya 0 081
x 09066
om
x 7
Oahogomai load, Pe = gBH LEGEN 5617
Peay = Pet
Maximum lon
ya vs 0%
Grogonal oad, Py —q BH ECG EN 947
¡Secondary beams om
0305m
25m
12
»s
10
40
Fiat
x= 82
ca 976
Berween ¥ frases 7
nn tay
nn 06
por 056
E mwrG
057
Front eration
snm
ns
5
7
Weve an
tae” Tha
Coe ons
tn tn
np SNC
ST ax
on D
aie van
de kn
Mar 08
toes 00
ay
inc, Den,
bus O2
vis om
Bian 0304
012
025
Gens
ou
08
Forts
070
oa
LE
#57
Pr Coowondary beamo)= NA Dom CoN HOG
I frames and secondary hans: N
10574
eight of envelope
Breadth of envelope
Biene proportions
Projected area of elements
Solty vaio,
From Vigure AS
Application stop in BAG
Soo frames
Earaion 4
Figure 40,019=25
Applica steps in MAS
jor fee
uation A6
Een
Bl
Le
Application steps In A23
Sum uf all secondary Bees
Total Y ou
Appendix B. Corrected factor tables
‘The small errors in Tables 4, 22, 23 and 24 of RS 6399-2: 1997 were
explained in 3.1.2. Corrected values are given in Tables B.I-B.4. These
are consistent with the model from which the Standard was constructed,
given in reference I, the computer program DREVe and the Ready-
Reckoner (set 31.4).
In an effort to keep the 2002 ammendments simple, not all these
conections were made where the differences were small. Accordingly, the
corrected tables in this Appendix remain relevant
Table B.1 Corrected values for Table 4, Sy
‘Site in country, or less than 2m inside town Sie in town, extending > 2km upwind
from the site
Elune Cloves distance wo searkm Effective Closest distance in sex km
height eight
Ham <O1 2 10 100 Hem 2 000m
2 14 ras ZAS 10
5 165 157 E
wm 1 1 14 16 158
15 136 182 15 tee 182 m
et") 10 2010 10 17
50 197 196 20 1 196 185
50 204 20 50 20 2m 105
wo an 212 wo zu 211 20
234 | Wind loading: a practica guide 10 BS 6399-2 opa cnc neon | 295
Table 82 Comet ves o able 22.5 ad Jable 3 Creed sles for Table 24 ont
A
Bree Faser prima dance fom vn ven Ene Rue ren
A tom wa
ps | st
tom Fuer re) | Km “= | a +
TS OST 00 Us? 0m 074 ons 2 O0 06 pany er
& om ons On am 02 aie er ee nes ve
SS du om os ot ou SH oo ons om on
Yo om om am om ort Hota tem ah Om 0
w lan sr 1 ane 1008 1 7 MO so om or
SG où os GPS OM ONE aim ROSE UN te OMB 07m
sos ia van nam MAS 1076 BH 0% 0% 007 oui oùé
oe om où om om omo E E
DS dm a war 1060038 om oes on
SO GET Gin OMR OT OMS OS Ones He US mr ti vin
5 SH da dus Ia a 120 MH IG os 059 om omis
Ÿ Gi ont on out oise O7 O1 CC Nae an Tan le
a BO 1000 100009420028 os
Ÿ om Ole ole om 016 a 07 Him Ie ENE
OL OT Em o om
SO Ge mms 0 om m O7 018 FH 00 100 1 La
ee u Le 20 1001000 om mn on
Some 006 om ou mH 0m ai FO ton Ie Mion LR LT
30 à UNS DES IS I 20 10 100 hm om
Ÿ où US OMS OS 0068 OU Dam Fi LIN io om, ‘tow te
Table BA Corrected values for Table 2,
recive Diagonal dimension. aim
height Hem ______-_ —
sn 40 100 20 300
10 a0 2m 2
20 14 324295266228 1
50 3 30 2m 20 20 197
100 34 30 26 28 20 192
200 a ad 0 28 242 213 19
300 34 41 32 28 24 216 19
Table 82 Comet ves o able 22.5 ad Jable 3 Creed sles for Table 24 ont
A
Bree Faser prima dance fom vn ven Ene Rue ren
A tom wa
ps | st
tom Fuer re) | Km “= | a +
TS OST 00 Us? 0m 074 ons 2 O0 06 pany er
& om ons On am 02 aie er ee nes ve
SS du om os ot ou SH oo ons om on
Yo om om am om ort Hota tem ah Om 0
w lan sr 1 ane 1008 1 7 MO so om or
SG où os GPS OM ONE aim ROSE UN te OMB 07m
sos ia van nam MAS 1076 BH 0% 0% 007 oui oùé
oe om où om om omo E E
DS dm a war 1060038 om oes on
SO GET Gin OMR OT OMS OS Ones He US mr ti vin
5 SH da dus Ia a 120 MH IG os 059 om omis
Ÿ Gi ont on out oise O7 O1 CC Nae an Tan le
a BO 1000 100009420028 os
Ÿ om Ole ole om 016 a 07 Him Ie ENE
OL OT Em o om
SO Ge mms 0 om m O7 018 FH 00 100 1 La
ee u Le 20 1001000 om mn on
Some 006 om ou mH 0m ai FO ton Ie Mion LR LT
30 à UNS DES IS I 20 10 100 hm om
Ÿ où US OMS OS 0068 OU Dam Fi LIN io om, ‘tow te
Table BA Corrected values for Table 2,
recive Diagonal dimension. aim
height Hem ______-_ —
sn 40 100 20 300
10 a0 2m 2
20 14 324295266228 1
50 3 30 2m 20 20 197
100 34 30 26 28 20 192
200 a ad 0 28 242 213 19
300 34 41 32 28 24 216 19
References
1 Cook, NA. 19NS and 1990). The Designer's Guide so Wind Loeding of Building
Siructares: Part I and Part 2, Butsrworh, London, ISBN 0408-0870. and 40
vos.
2 Buiking Research Establishment (1999), Wind Londing on Buildings: Brief Guidance
Jor Using BS 6399 2:1997, BRE Digest 436. Puts 1-3. CRC LU, London,
3 Construction Research and Informations Assocation 1981). Wind Engineering in the
Eigivies, CIRIA, London.
4 Building Research Establishment (1989). The Assessment of Wind Lande, BRE Digest
246. Pats 1-7. HMSO. London
5 Newbery, CA. and Baton, KI. (1974). The Wind Loading Handbook. UMSO,
London. ISBN 0 11 670529 9,
6 International Atomic Energy Agency (1981. Eureme Meteorological Evene in
clear Poner Plan Sing, INEA Safety Serie, No SO-SG-SI1A. IAEA, Vienna
7 Cook, NJ. and Prior, MJ. (1987). Extreme wind climace of the United Kingdom
Jounal of Wind Engineering & industrial Aeradswantes, 26. 365-372.
8 Cook, NJ. (1982). Towards beter estimation of extreme winds. Journal of Wind
Engineering & Industrial Aerodmanier 9, 23-32.
9 Gaton, VG. (1976). Maps of hourly mean wind speed over the United Kinadom,
1965-1973. Climatological Memorandum 79. Meteorological Otic, Bracke
10 Cook, NJ. (1983). Note un the dieconal and seasonal assessment of extreme winds
for design. Journal of Wind Engineering & Industal Aerodwamics 12. 65-372.
11 Cook, NJ. (1997). The Deaves and Haris ABL model applied (u heterogeneous train
Jounal of Wind Engineering & Industrial Acrodyvanies, 68, 197-214
12 Sure. D. and Djakovieh, D. (1995), Flutustng pressures on nels of tall buildings
Journal of Wind Engineering & Industriel Aeralynaes, 58, 81-112,
13 Wilford, MR. and Allsop, A.C. (1990). Design Guide for Wind Loads on Unciad
Building Sucher Daring Construction, Raiding Resemch Paahlihment Repo
"BR 173. Construction Research Communications Lid, London.
14 Blessmann, J, (1987). Acdo do vento em cobertura curras, La Parte, Cadena Tecnico
CT-86 Universidude Federale do Rio Grande do Sul, Porto Alegre.
15 Blessmam 1. (1987). Vento em coberturas curvas — pavidhdes vignhes. Caderno
Tecnico CT 88, Universidade Federale do Rio Grande do Sul, Porto Alegre
16 British Standards Institution (1997). Eurorode 1: Bass of Design and Actions on
Sımeimes, Part 24 Actions on Stuctwrer — Wind action, DD ENV 1991-24 1007,
London,
17 Bish Standards tnstitution (1997). ISI, Wind Loading Rendy-Reckoner for BS 639%
Part 2 1997. MSI Specialist Hook LR 10139. RSI, London,
1 Cook, NA. 19NS and 1990). The Designer's Guide so Wind Loeding of Building
Siructares: Part I and Part 2, Butsrworh, London, ISBN 0408-0870. and 40
vos.
2 Buiking Research Establishment (1999), Wind Londing on Buildings: Brief Guidance
Jor Using BS 6399 2:1997, BRE Digest 436. Puts 1-3. CRC LU, London,
3 Construction Research and Informations Assocation 1981). Wind Engineering in the
Eigivies, CIRIA, London.
4 Building Research Establishment (1989). The Assessment of Wind Lande, BRE Digest
246. Pats 1-7. HMSO. London
5 Newbery, CA. and Baton, KI. (1974). The Wind Loading Handbook. UMSO,
London. ISBN 0 11 670529 9,
6 International Atomic Energy Agency (1981. Eureme Meteorological Evene in
clear Poner Plan Sing, INEA Safety Serie, No SO-SG-SI1A. IAEA, Vienna
7 Cook, NJ. and Prior, MJ. (1987). Extreme wind climace of the United Kingdom
Jounal of Wind Engineering & industrial Aeradswantes, 26. 365-372.
8 Cook, NJ. (1982). Towards beter estimation of extreme winds. Journal of Wind
Engineering & Industrial Aerodmanier 9, 23-32.
9 Gaton, VG. (1976). Maps of hourly mean wind speed over the United Kinadom,
1965-1973. Climatological Memorandum 79. Meteorological Otic, Bracke
10 Cook, NJ. (1983). Note un the dieconal and seasonal assessment of extreme winds
for design. Journal of Wind Engineering & Industal Aerodwamics 12. 65-372.
11 Cook, NJ. (1997). The Deaves and Haris ABL model applied (u heterogeneous train
Jounal of Wind Engineering & Industrial Acrodyvanies, 68, 197-214
12 Sure. D. and Djakovieh, D. (1995), Flutustng pressures on nels of tall buildings
Journal of Wind Engineering & Industriel Aeralynaes, 58, 81-112,
13 Wilford, MR. and Allsop, A.C. (1990). Design Guide for Wind Loads on Unciad
Building Sucher Daring Construction, Raiding Resemch Paahlihment Repo
"BR 173. Construction Research Communications Lid, London.
14 Blessmann, J, (1987). Acdo do vento em cobertura curras, La Parte, Cadena Tecnico
CT-86 Universidude Federale do Rio Grande do Sul, Porto Alegre.
15 Blessmam 1. (1987). Vento em coberturas curvas — pavidhdes vignhes. Caderno
Tecnico CT 88, Universidade Federale do Rio Grande do Sul, Porto Alegre
16 British Standards Institution (1997). Eurorode 1: Bass of Design and Actions on
Sımeimes, Part 24 Actions on Stuctwrer — Wind action, DD ENV 1991-24 1007,
London,
17 Bish Standards tnstitution (1997). ISI, Wind Loading Rendy-Reckoner for BS 639%
Part 2 1997. MSI Specialist Hook LR 10139. RSI, London,
238 | Wind loading: a practical guide 19 BS 6399-2
14 Miler, CA. Cook. and Bamerd, RH. (1998), Caliration of the exposure of UK
ler Journal of Wind Engineering & Industrial Aerodynamics 74-76, 153
ñ
19 Irscion of Structural Engineers (1999). Temporary Demvamable Snuctures —
Cxidonce on Procurement. Design and Use. Inttution of Siuctural Engineers,
London
Index
2002 Amendments. 15, 54, 78, 169
Altitude factor, 37
Annual risk, 35
mean recurrence interval, 157
probabitity factor, 43
seasonal factor, 42
Asymmerie loads, 15, 127, 169
Asymnavetry
roofs, 85
walls, 77
Balance of flow, 106
automating, 175
example, 210
Bamel-vauli wots, 98
fiction, 87
Basic wind speed,
Blockage
free-standing canopies, 114
wind tunnel, 144
Boundary wall, 116
BRUCH 137
BREVe2, 134
BREWS, 134
Building height, 18
Building Regulations, 15
Approved Documents, 129
minimum annual risk, 42
Buildingetype factor, 33
5
Calibrations, 3
Circular-plan buildings
fat roofs, 90
pitched roofs, 97
walls, 82
Coastal sites, 148
Component lords, A
Conservatisin
coastal sites, 21
design wind specds, 8
dynamic classification, 34
examples, 186
Tunnelling, 78
hybrid options, 22, 131
internal pressuren, 111
lattice frames, 120
Toad sharing, 126
‘optimum compromiso, 13, 22
pressure coefficients, 67
Ready- Reckoner, 1
reducing, 131
standard method, 18
Country, 49
CP 3-V-2
categories, 173
changes from, 6, 54, 64, 68, 70,
120, 124, 157
conservatism, Y
dlivision-by-parts, 10, 56
engineering model, 5
equivalent options, 146
fonafing, 78
minimum fetch, 8, 49
misinterprettion. 8
sie exposure calibration, 50)
size effect, 64
step changes, 3
temporary huilé
Curved eaves, 88
14 Miler, CA. Cook. and Bamerd, RH. (1998), Caliration of the exposure of UK
ler Journal of Wind Engineering & Industrial Aerodynamics 74-76, 153
ñ
19 Irscion of Structural Engineers (1999). Temporary Demvamable Snuctures —
Cxidonce on Procurement. Design and Use. Inttution of Siuctural Engineers,
London
Index
2002 Amendments. 15, 54, 78, 169
Altitude factor, 37
Annual risk, 35
mean recurrence interval, 157
probabitity factor, 43
seasonal factor, 42
Asymmerie loads, 15, 127, 169
Asymnavetry
roofs, 85
walls, 77
Balance of flow, 106
automating, 175
example, 210
Bamel-vauli wots, 98
fiction, 87
Basic wind speed,
Blockage
free-standing canopies, 114
wind tunnel, 144
Boundary wall, 116
BRUCH 137
BREVe2, 134
BREWS, 134
Building height, 18
Building Regulations, 15
Approved Documents, 129
minimum annual risk, 42
Buildingetype factor, 33
5
Calibrations, 3
Circular-plan buildings
fat roofs, 90
pitched roofs, 97
walls, 82
Coastal sites, 148
Component lords, A
Conservatisin
coastal sites, 21
design wind specds, 8
dynamic classification, 34
examples, 186
Tunnelling, 78
hybrid options, 22, 131
internal pressuren, 111
lattice frames, 120
Toad sharing, 126
‘optimum compromiso, 13, 22
pressure coefficients, 67
Ready- Reckoner, 1
reducing, 131
standard method, 18
Country, 49
CP 3-V-2
categories, 173
changes from, 6, 54, 64, 68, 70,
120, 124, 157
conservatism, Y
dlivision-by-parts, 10, 56
engineering model, 5
equivalent options, 146
fonafing, 78
minimum fetch, 8, 49
misinterprettion. 8
sie exposure calibration, 50)
size effect, 64
step changes, 3
temporary huilé
Curved eaves, 88
240 1 Index
Index | 241
Demolition, 152
Design aids, 132
Diagonal dimension, 17, 31, 124
‘extemal pressures, 124
fiction loads, 124
“internal pressures, 126, 163
oad sharing, 159
loaded area, 39
Direction factor, 19, 21. 40
towns, 150
mear coast, 148
known orientation, 174
Directional method, 18
altude factor, 39
conservaisit, 9
direction factor, 41
dynamic pressure, 64
efectivo wind speed, 64
externa pressure coefficient, 69
flat roots, 87
normal pressures, 121
obstructions, 54
pitched rook, 92
roofs, 85
site wind speed, 45
terrain and building factor, 62
walls, 77. 78
wind direction. 19
Displacement height, 53, 55
choice of, 152
‘examples, 177
Divergence distance, 71
Division-by-parts, 56, 58
calibration, 10
elements, 118
example, 216
Dominant openings, 112
‘diagonal dimension, 126, 160
lective, 113
example, 206
multiple. 165
serviceability limit, 157
size, 164
Duopitch roots, 93, 94
augmentation Factor, 34
‘example, 215
full procedure, 145
limits, 34
‘Dynamic pressure, 29, 64
examples, 178
options, 146
Efective height, 15, 55
Effective wind speed, 63, 64
accuracy, 131
BREVE, 23
conservatism, $
examples, 178
minimum value, 54
options. 146.
Equivalent static gust, 17
Faluaries, 148
External pressure coefficients, 29, 69
‘alternative sources, 141
roofs, 85
walls, 77
Face fonds, 31
Fences, 116
Fetch, 47
fetch factor, 49
Five-sorey office building, 206
‘ized dimensions, 18
Flat roofs, 87
fiction, 87
Free-standing canopies, 114
walls, 15, 116
Friction loads. 123
roofs, 87
Ground roughness, 48, 51
ipped roofs. 93, 96
Hybrid options, 22. 131
in towns, 153
Influence function, 160
Inland water, 148
Inset storeys
roof zones, 166
roofs, 90, 97
sealing length, 75
Internal partitions, 167
Internal pressure coefficients, 30,
106
Internal pressures
asymmeiric loads, 128
controlling, 176
diagonal dimension, 126, 163
dominant openings, 112.
enclosed buildings, 111
response time, 110, 162
Internal wells
tooßs, 91, 97
walls, 84
Interpolation
precision, 130
terrain and building Factor, 60
Irregular Aush faces
roof zones, 166
scaling length, 74
Lattice structures, 221
elements, 119
exclusion, 70
Loud sharing
diagonal dimension, 124, 159
introduced, 32
Long-span roots
pil, 68
Mansard eaves, 88
Mansard roofs, 97
Minimizing wind loads, 175
Monopitch roofs, 93, 94
Multi bay roofs, 100
Malt-piteh roofs, 97
Net pressure coefficients, 30
elements, 117
Non-ventcal walls, 81
directional method, 79
Normal pressures, 29, 121
Obstruction high, 18, 5
18, 53,58
ox tam, 152 3.
rion spacing, 1
er sed an 019
sample, 198
Open topped bul
Option 21
8, 113
direction factor, 40
example, 180
ground roughness, 52
site wind speed, 45
Option 2, 21
direction factor, 41
example, 181
example, 183
round roughness, $2
site wind speed, 45
Overall loads, 31, 122
structural loads, 168
‘Overseas sites
site wind speed, 45
Parapeis. 88, 116
pitched roofs, 97
Peak factor, 17
terrain and building factor, 60.
Permanent buildings
probability factor, 44
seasonal factor, 42
Permeability, 160
walls, 112
Pitched roofs, 15, 92
comments, 102
friction, 87
Polygonal-plan buildings, 82
pitched roofs, 96
roots, 89
walls, 79
Porosity, 111
ws permeability, 160
Portal frames
diagonal dimension, 125
Index | 241
Demolition, 152
Design aids, 132
Diagonal dimension, 17, 31, 124
‘extemal pressures, 124
fiction loads, 124
“internal pressures, 126, 163
oad sharing, 159
loaded area, 39
Direction factor, 19, 21. 40
towns, 150
mear coast, 148
known orientation, 174
Directional method, 18
altude factor, 39
conservaisit, 9
direction factor, 41
dynamic pressure, 64
efectivo wind speed, 64
externa pressure coefficient, 69
flat roots, 87
normal pressures, 121
obstructions, 54
pitched rook, 92
roofs, 85
site wind speed, 45
terrain and building factor, 62
walls, 77. 78
wind direction. 19
Displacement height, 53, 55
choice of, 152
‘examples, 177
Divergence distance, 71
Division-by-parts, 56, 58
calibration, 10
elements, 118
example, 216
Dominant openings, 112
‘diagonal dimension, 126, 160
lective, 113
example, 206
multiple. 165
serviceability limit, 157
size, 164
Duopitch roots, 93, 94
augmentation Factor, 34
‘example, 215
full procedure, 145
limits, 34
‘Dynamic pressure, 29, 64
examples, 178
options, 146
Efective height, 15, 55
Effective wind speed, 63, 64
accuracy, 131
BREVE, 23
conservatism, $
examples, 178
minimum value, 54
options. 146.
Equivalent static gust, 17
Faluaries, 148
External pressure coefficients, 29, 69
‘alternative sources, 141
roofs, 85
walls, 77
Face fonds, 31
Fences, 116
Fetch, 47
fetch factor, 49
Five-sorey office building, 206
‘ized dimensions, 18
Flat roofs, 87
fiction, 87
Free-standing canopies, 114
walls, 15, 116
Friction loads. 123
roofs, 87
Ground roughness, 48, 51
ipped roofs. 93, 96
Hybrid options, 22. 131
in towns, 153
Influence function, 160
Inland water, 148
Inset storeys
roof zones, 166
roofs, 90, 97
sealing length, 75
Internal partitions, 167
Internal pressure coefficients, 30,
106
Internal pressures
asymmeiric loads, 128
controlling, 176
diagonal dimension, 126, 163
dominant openings, 112.
enclosed buildings, 111
response time, 110, 162
Internal wells
tooßs, 91, 97
walls, 84
Interpolation
precision, 130
terrain and building Factor, 60
Irregular Aush faces
roof zones, 166
scaling length, 74
Lattice structures, 221
elements, 119
exclusion, 70
Loud sharing
diagonal dimension, 124, 159
introduced, 32
Long-span roots
pil, 68
Mansard eaves, 88
Mansard roofs, 97
Minimizing wind loads, 175
Monopitch roofs, 93, 94
Multi bay roofs, 100
Malt-piteh roofs, 97
Net pressure coefficients, 30
elements, 117
Non-ventcal walls, 81
directional method, 79
Normal pressures, 29, 121
Obstruction high, 18, 5
18, 53,58
ox tam, 152 3.
rion spacing, 1
er sed an 019
sample, 198
Open topped bul
Option 21
8, 113
direction factor, 40
example, 180
ground roughness, 52
site wind speed, 45
Option 2, 21
direction factor, 41
example, 181
example, 183
round roughness, $2
site wind speed, 45
Overall loads, 31, 122
structural loads, 168
‘Overseas sites
site wind speed, 45
Parapeis. 88, 116
pitched roofs, 97
Peak factor, 17
terrain and building factor, 60.
Permanent buildings
probability factor, 44
seasonal factor, 42
Permeability, 160
walls, 112
Pitched roofs, 15, 92
comments, 102
friction, 87
Polygonal-plan buildings, 82
pitched roofs, 96
roots, 89
walls, 79
Porosity, 111
ws permeability, 160
Portal frames
diagonal dimension, 125
242 1 Index
Index 1 243
Portal-frame building. 198
Principal changes
summary, 6
roof zones, 167
Probability factor, 43
serviceability limit, 157.
Probability model
change, 43
Ready-Reckoner, 133
Recessed hays
wo, 91, 97
sealing length, 74
walls, 84
Re-entrant corners
roofs, 91, 97
sealing length, 72
walls, 83
Reference height, 18, 52, 54
Roof overhangs, 87
Roots
general rules, 86
Sealing length, 71, 72
comments, 77
imroduced, 68
pitched roofs, 94
walls, 78
Sea, 49
Seasonal factor, 42
Serviceabilty limit, 157
Shape factors. 66
commentary, 120
Shelter
boundary walls und fences, 116
complex terrain, 138
direct, 158
in towns, 53, 152
minimum effective height, 56
parapets, 117
Signboards, 116
Site altitode
cliffs and escarpments, 28
Site exposure, 54.
commentary, 64
comments, 55
dynamic pressure, 63
effective height, 52
effective wind speed. 63
round roughness, 48
model, 46
{errain and building factor, 59
Site wind speed, 44, 45
overseas sites, 45, 139
Size effect
“diagonal dimension, 124
directional method, 64
dynamic pressure, 64
introduced, 41
terain and b
fest factor, 127
friction Jos, 124
introduced, $9
normal pressures, 121
overall loads, 122
standard method, 64
Smaller extensions
scaling length. 75
Smallest enclosing recta
roof zones. 167
walls, 82
Special considerations
roofs, 90, 96
walls, 83
Standard method. 18
allitude factor, 39
direction factor, 40
ddynamie pressure, 64
effective wind speed, 64
‘external pressure coefficient, 70
flat roots. 89
normal pressures, 121
obstructions, 54
pitched roofs, 94
roots, 85
site wind speed, 45
terrain and building factor, 62
walls, 78, 81
wind direction, 19
de, 77
Step changes
effect in CP 3-V-2, 3
in BS 6399.2, 64
Structural sections. 117
Surface friction, 30
Surface lod, 122
Surface pressure, 29
‘Temporary build
1 factor, 44
and building factor, 59, 60
‘comments, 63
in towns, 150, 153
Ready-Reckoner, 133
‘Timber framed house, 187
“Topography, 23
cliffs and escarpments. 28
complex, 138, 154
dimensions, 25
bills and sidges, 27
Key values, 28
minimizing effort, 170
significance criteria, 24
“Tower and podium, 215
Towns, 49
‘coastal, 151
hybrid options, 153
‘minimizing effort, 170.
reclucing conservatism, 131
sites in, 150
vtr areas. 122,
across zones, 158
Ti
Unfactored loads, 122, 174, 177
Upwvind wings, 72 ’
roofs, 91, 97
walls, 83
Variable dimensions, 18
Walls,
pressure coefficients, 77
Wind climate
altitude factor, 37
basic wind speed, 35
directional factor, 40
model, 35
factor, 42
site wind speed, dd
Wind direction, 19
Wind tunnel tests, 142
Woodland, 150, 153
Zones. 67
‘comer zones, 166
Index 1 243
Portal-frame building. 198
Principal changes
summary, 6
roof zones, 167
Probability factor, 43
serviceability limit, 157.
Probability model
change, 43
Ready-Reckoner, 133
Recessed hays
wo, 91, 97
sealing length, 74
walls, 84
Re-entrant corners
roofs, 91, 97
sealing length, 72
walls, 83
Reference height, 18, 52, 54
Roof overhangs, 87
Roots
general rules, 86
Sealing length, 71, 72
comments, 77
imroduced, 68
pitched roofs, 94
walls, 78
Sea, 49
Seasonal factor, 42
Serviceabilty limit, 157
Shape factors. 66
commentary, 120
Shelter
boundary walls und fences, 116
complex terrain, 138
direct, 158
in towns, 53, 152
minimum effective height, 56
parapets, 117
Signboards, 116
Site altitode
cliffs and escarpments, 28
Site exposure, 54.
commentary, 64
comments, 55
dynamic pressure, 63
effective height, 52
effective wind speed. 63
round roughness, 48
model, 46
{errain and building factor, 59
Site wind speed, 44, 45
overseas sites, 45, 139
Size effect
“diagonal dimension, 124
directional method, 64
dynamic pressure, 64
introduced, 41
terain and b
fest factor, 127
friction Jos, 124
introduced, $9
normal pressures, 121
overall loads, 122
standard method, 64
Smaller extensions
scaling length. 75
Smallest enclosing recta
roof zones. 167
walls, 82
Special considerations
roofs, 90, 96
walls, 83
Standard method. 18
allitude factor, 39
direction factor, 40
ddynamie pressure, 64
effective wind speed, 64
‘external pressure coefficient, 70
flat roots. 89
normal pressures, 121
obstructions, 54
pitched roofs, 94
roots, 85
site wind speed, 45
terrain and building factor, 62
walls, 78, 81
wind direction, 19
de, 77
Step changes
effect in CP 3-V-2, 3
in BS 6399.2, 64
Structural sections. 117
Surface friction, 30
Surface lod, 122
Surface pressure, 29
‘Temporary build
1 factor, 44
and building factor, 59, 60
‘comments, 63
in towns, 150, 153
Ready-Reckoner, 133
‘Timber framed house, 187
“Topography, 23
cliffs and escarpments. 28
complex, 138, 154
dimensions, 25
bills and sidges, 27
Key values, 28
minimizing effort, 170
significance criteria, 24
“Tower and podium, 215
Towns, 49
‘coastal, 151
hybrid options, 153
‘minimizing effort, 170.
reclucing conservatism, 131
sites in, 150
vtr areas. 122,
across zones, 158
Ti
Unfactored loads, 122, 174, 177
Upwvind wings, 72 ’
roofs, 91, 97
walls, 83
Variable dimensions, 18
Walls,
pressure coefficients, 77
Wind climate
altitude factor, 37
basic wind speed, 35
directional factor, 40
model, 35
factor, 42
site wind speed, dd
Wind direction, 19
Wind tunnel tests, 142
Woodland, 150, 153
Zones. 67
‘comer zones, 166
Tags
Categories
GeneralDownload
Download Slideshow Get the original presentation fileQuick Actions
Statistics
- Views 5
- Slides 131
- Age 508 days
Related Slideshows
22
Pray For The Peace Of Jerusalem and You Will Prosper
RodolfoMoralesMarcuc
35 views
26
Don_t_Waste_Your_Life_God.....powerpoint
chalobrido8
38 views
31
VILLASUR_FACTORS_TO_CONSIDER_IN_PLATING_SALAD_10-13.pdf
JaiJai148317
34 views
14
Fertility awareness methods for women in the society
Isaiah47
31 views
35
Chapter 5 Arithmetic Functions Computer Organisation and Architecture
RitikSharma297999
30 views
5
syakira bhasa inggris (1) (1).pptx.......
ourcommunity56
31 views