474 Wind Loading of Structures Cook, N.J. (1990) The designer's guide to wind loading of building structures. Part 2 Static structures. Building Research Establishment and Butterworths, London Davenport, A.G. (1977) The prediction of risk under wind loading. Proceedings. 2nd International Conference on Structural Safety and Reliability, Munich Germany, September 19-21, pp. 511-596. Davenport, A.G. (1982) Chapter 12: The interaction of wind and structures. I Engineering Meteorology. ed. E. J. Plate, pp. 527-572. Elsevier, Amsterdam, Dorman, C.MLL. (1984) Tropical cyclone wind speedsin Australia. Civil Engineering Transactions, Institution of Engineers, Australia, CE26: 132-139. ESDU International. (1983) Strong winds in the atmospheric boundary layer. Par 2: discrete gust speeds, ESDU Data Item 83045, ESDU International, London, UK, (revised 2002). Highways Agency (UK). (2001) Design rules for aerodynamic effects on bridges, BD 49/01. Holmes, J.D. (2002) A re-analysis of recorded wind speeds in Region A. Australia Journal of Structural Engineering, 4: 29-40. Holmes, J.D. (2014) Along-wind and cross-wind response ofa generic tall building + comparison wind-tunnel data with codes and standards. Journal of Win Engineering and Industrial Aerodynamics, 132: 136-141. Holmes, J.D. and Ginger, ].D. (2012) The gust wind speed duration in AS/NZ. 1170.2. Australian Journal of Structural Engineering, 13: 207-217. Holmes, J.D., Baker, C.J., English, E.C., and Choi, E.C.C. (2005a) Wind structure and codification. Wind and Structures, 8: 235-250. Holmes, J.D., Kasperski, M., Miller, C.A.,Zuranski, J.A.and Choi .(2005b) Extreme wind structure and zoning. Wind and Structures, 8: 269-281. Holmes, J.D., Tamura, Y. and Krishna, P. (2009) Comparison of wind loads calcu lated by fifteen different codes and standards, for low, medium and high-ris buildings. 11h Americas Conference on Wind Engineering, San Juan, Puert Rico, 22-26 June. Holmes, J.D., Allsop, A. and Ginger, J.D. (2014) Gust durations, gust factors and gust response factors in wind codes and standards. Wind and Structures, 19. 339-352. Hong Kong Buildings Department. (2004) Code of practice on wind effects in Hong Kong. Government of the Hong Kong Special Administrative Region, Buildings Department, Mongkok. International Standards Organization. (2009) Wind actions on structures, 1S International Standard. ISO 4354. Kasperski, M. (1993) Aerodynamics of low-rise buildings and eodification»Journal of Wind Engineering and Industrial Aerodynamics, S0#253=2630 Kasperski, M. and Geurts, C. (2005) Reliability and code level. Wind an Structures, 8: 295-307. Letchford, C.W., Holmes, J.D., Hoxey, R.P. and Robertson, A.P, (2005) Wind pres sure coefficients on low-rise structures and codification. Winid and Structures, 8: 283-294, Mehta, K.C. (1998) Wind load standards. Proceedings, Jubileum Conference on Wind Effects on Buildings and Structures, Porto Alegre, Brazil, 25-29 May. Ministry of Land, Infrastructure and Transport, Japan: (2000) Building Standard Law of Japan, Enforcement Orders Regulations and Notifications.454 Wind Loading of Structures Wind climate Aerodynamic (basic or regional shape factor Design criteria for| wind velocity, (pressure & force strength & directional factors) coefficients) serviceability Adjustment Gust or dynamic Sof exposine response factors (topography, shielding) and height Figure 15.1 The wind loading ‘chain’. (Adapted from Davenport, 1977, 1982) 15.2 GENERAL DESCRIPTIONS The provisions of following five codes and standards are summarized in] this Chapter: ISO 4354:2009 — Wind actions on structures — published in 2009, EN 1991-1-4.6:2005, Eurocode 1: Actions on Structures—Part 1.4: General Actions - Wind Actions — published in 2005, ASCE Standard ASCE/S! 16. Minimum Design Loads = Associated Criteria for Buildings and Other Structures — published in 2016, * AIJ Recommendations for Loads on Buildings — published in 2004, Australian/New Zealand Standard. Structural design actions. Part 2: Wind actions. AS/NZS1170.2:2011 - published,im)2011 with) Amendments from 2012 to 2016. The documents reviewed are those current at the time of writing. Althoug! there are many other wind codes and standardsin the world, many: of thes« have been derived from, or are closely related to, one or the other of th above documents, so that the following comments have wide and general application. 15.2.1 ISO 4354-wind actions on structures The current version of ISO International, Standard 4354, vag Actions on Structures, published by the International, Organizatioy for Standardization (ISO), was issued in 2009. The 2009 edition wa:Wind loading codes and standards 455} completely re-formulated and bears little resemblance to the earlier (1997, edition. As described in the introduction to ISO 4354, the document i: intended ‘for use by countries without an adequate wind-loading stated and to bridge between existing standards. Most of the technical infor- mation in ISO 4354 is provided in a series of informative annexes. Tw general methods of calculation of wind forces are given: one based on peak velocity (nominally with a 3-second duration), and the other on al mean velocity (nominally with a 10-minute averaging time). However, fo: structures for which dynamic response effects are not important, the pea! velocity method should be used. The main part of the document is quite short, and consists largely o} definitions of the terms in the expression used to calculate wind pressure: P= (uie)(Cp)(Con) (15.1) The site peak dynamic pressure, q,,.. is derived from the peak site win speed, which, in turn, is derived from a reference peak wind speed (in the standard exposure of 10m height in open country terrain): gue — 0.5P (Voc)? (18.2) Vate = Veet “Coop (15.3) Cy, in Equation (15.1) allows for the effects of fluctuating pressures due t upwind turbulence, in the wake of a structure, as well as forces resulting, from resonant dynamic response. Section 9 of ISO 4354 provides several methods to be used to determing pressure and force coefficients, including the use of wind-tunnel tests. Basic requirements for suitable wind-tunnel test procedures are given ii Annex H in ISO 4354. Interestingly, Section 9 also apparently allows th use of ‘computational based data’. However, this is somewhat contra- dicted by Annex I in ISO 4354, which cautions against the use of compu: tational fluid dynamics (CFD) techniques for this purpose, at the present stage of their development. The comments in Annex on the applicabilit of CFD techniques are generally compatible with those in Chapter 16 off this book. Some basic aerodynamic pressure and force Coefficients for simple shape: are provided in an informative Annex D. This annex contains some inter- esting and innovative features: a. A clear specification of the target probability level associated with specified pressure and force coefficients» This:-has:beensetat the 80% fractile of the extremes, and can be estimated as follows: Gy sor = Cy £0.7C poms (s.4)456 Wind Loading of Structures when a reference time of Thour (or equivalent in a wind-tunnel test is adopted, Cie G26; (15.5) when a reference time of 10 minutes, or equivalent, is used. Cyan iS the root-mean-square value determined from the set of the extreme pressure coefficients determined from multiple samples with the same reference period. (Note that this is not the same as the r.m.s. fluctuating pressure coefficient, C,', for a complete pressure-time his tory, as defined in Sections 8.3.1 and 9.4.1 of this book.) . The use of the load-response correlation (LRC) method (see Section) 5.4.3) to determine some effective pressure coefficients for overall lift and drag forces for some simple shapes of low-rise buildings. These are given in Section D10 of ISO 4354. s In Sections D6 and D7 of ISO 4354, external and internal pressure coef- ficients that originate in the Australian/New Zealand Standard, AS/NZ: 1170.2, have been specified. However, these should be used with caution in conjunction with ISO 4354, as they have not been adjusted for the dif. ferent gust durations in the two Standards (0.7 seconds in AS/NZS 1170.9) versus 3 seconds in ISO 4354). This could be achieved through the dynamid response factor, C4yqy Which should be greater than 1.0 for small structures. The suggested value of Cy, of 0.85 in Section DS of ISO 4354 is clearl un-conservative for small low-rise buildings, when used with 3-second gust velocities and quasi-steady pressure coefficients (see Section 15.3 in this chapter, and Holmes et al., 2014). Annex E in ISO 4354 provides detailed informative advice on thi dynamic response factor for dynamically sensitive structures such a: tall buildings and towers. Methods are given for estimating along-wind, cross-wind and torsional response, and these are closely related to thosd given in the Recommendations of the Architectural Institute of Japan| (Al). 15.2.2 EN 1991-1-4.6 Eurocode I. Part I-4 wind actions Eurocode 1 on wind loads, issued in 2005, is a European Standard (EN) which is intended for use in most European countries. The version in eachi member country contains a ‘National Annex’, applicable only to that par- ticular country. EN 1991-1-4.6 represents several years of. work. by repre sentatives from many countries of the European Union and two separate committees, and is the nearest document to a truly multi-national wind: loading standard currently in existence. This is a lengthy document with comprehensive methods of Stati¢ and, dynamic design for wind loads. The code is applicable to buildings and|Wind loading codes and standards 457) other structures, with heights up to 200m, and to bridges with spans lesy than 200m. No basic wind speeds are provided ~ these are provided sepa rately in each National Annex. However, the basic wind velocity in each) country is a 10-minute mean velocity at 10m height in open country ter. rain, with an annual probability of exceedance of 0.02 (50-year return} period). The mean wind velocity v,(z) at a height, z, above the terrain is given by: A(z) + Co(Z)- Ye (15.6)| Um (&) = where v; is the basic wind velocity at 10 m height over open country terrain; ¢,(z) is a roughness factor, which varies with both height and terrain type and c¢,(z) is the orography (i.e. topography) factor. For the roughness fac tor, five different terrain types are defined with roughness lengths, z, (seq Section 3.2.1), varying from 0.003 to 1.0m. The peak velocity pressure, q,(2), is given by: 4p (2) = cel2) “4 (15.7) where q,=% p v2. is the density of air, given in the National Annexes, and ¢la) is an ‘exposure factor’ given by: c(2)=[1+ 7112} [e@)-col0)P (15.8) J, 2) is the turbulence intensity at height, z. The term [1+71,(z)] is effectively a gust factor for the velocity pressure, and is an approximation to [1+g, 1,(2), with the peak factor g, taking a value of 3.5. Thus, the exposure factor, ¢,(z), combines gusting effects, terrain, height and topographic effects into a single height- and terrain-dependent factor, and enables the 10-minute mean wind velocity to be effectively converted t a gust wind velocity, and velocity pressure. The use of a peak factor of 3.5 with a 10-minute reference period corresponds to an equivalent gust dura tion of about 0.1 seconds. ‘The number of shape factors presented in Section Jf EN 1991-146 is extensive, with the number of cases covered exceeding those in most other codes and standards, External pressure coefficients on buildings, are given! for loaded areas of 1m’, and 10m”, denoted by C, 4 and C, 19, respectively. For areas between | and 10 m*, linear interpolation is applied. A ‘structural factor’ that allows for reductions in effective loading due to correlation effects on large structures, and for possible increases due t resonant dynamic effects, is defined in Section 6. This is discussed in morg detail in Section 15.8. Iwo alternative and independent procedures tor calculating a for along-wind dynamic response (B? and R2) are given in separate Annexe (Annex B and Annex C, respectively). A separate Annex/D_provide458 Wind Loading of Structures graphical information on the structural factor as a function of height and width of the structure, with separate graphs given for concrete and stee buildings, and for chimneys, with or without liners. The several methods given for calculation of the structural factor is poten: tially confusing for the user, and may have legal implications. However, thi multiple alternative options may be resolved within the National Annexe: for each participating country. Annex E contains detailed information on vortex-induced response of slender structures, such as chimneys. Two different and independent approaches for calculation of the cross-wind amplitude are given in Section: E.1.5.2 (Approach 1’) and E.1.5.3 (‘Approach 2’), respectively. Approach 1 is based on a sinusoidal excitation model (see Section 11.5.1 in this book]s Approach 2 is derived from a random excitation model (see Section 11.5.2} Work for a revision of EN 1991-1-4.6 was in progress in 2019-2020. 15.2.3 ASCE/SEI standard ASCE 7-16 ASCE/SEI 7-16, ‘Minimum Design Loads and Associated Criteria for) Buildings and Other Structures’, is a complete loading standard covering all types of loads. The wind-loading part consists of Chapters 26-31 From 1995 onwards, ASCE 7 has incorporated a number of significant changes in the wind load provisions from the 1993 and earlier editions. This includes the use of a 3-second gust wind speed instead of the ‘fastest-mile-o! wind’ as used in the past, a new zoning system for basic wind speeds, th use of high average recurrence intervals (300, 700 and 1,700 years) for ulti mate limit states design, the incorporation of topographic factors, som new data on pressure coefficients, a simplified procedure for buildings les than 9m in height, and a revised method for along-wind dynamic response calculation. The wind-loading provisions of ASCE 7-16 comprise six chapters as} follows: Chapter 26 gives ‘General Requirements’, including wind hazard maps, exposure categories, topographic multipliers, gust-cffeer,fac tors and internal pressure coefficients. Chapter 27 provides a ‘Directional Procedure’; the “main wind-forcel resisting system’ (MWFRS) for buildings of all heights. Chapter 28 gives an ‘Envelope Procedure’ (non-directional) for low. rise buildings (defined as having a mean roof height less than 18 m). Chapter 29 describes a directional procedure for_building. appurte- nances (such as rooftop structures and equipment) and other struc: tures (such as freestanding walls and signage, chimneys tanks, lattic frameworks and towers). Chapter 30 provides procedures for assessing wind loads on ‘compo: nents and cladding’.Wind loading codes and standards 459 * Chapter 31 describes the main requirements for the ‘Wind-Tunnel Procedure’. This may be required in earlier chapters for certain wind, sensitive structures, and is available as an option for all structures or, parts of structures covered by Chapters 27-30. The ASCE Standard has no legal standing of its own, but its provision: are cited by many of the regional, city and county building codes. Anj ‘International Building Code’ in the United States draws on the ASCE Standard for wind load provisions. 15.2.4 AlJ recommendations for loads on buildings The Recommendations of the Architectural Institute of Japan were revised) in 2004 and 2015, and form a comprehensive loading code including th effects of dead, live, snow, seismic, temperature, earth and hydraulic pres. sure, as well as wind loads. Chapter 6 on wind loads comprises 54 pages, with 114 pages of Commentary. The derivation of the wind-loading section| of the 2004 edition of the AlJ, and revisions from the 1993 version, werd described in detail by Tamura et al. (2004). Like the ASCE Standard, this is a comprehensive and advanced wind} loading document, although the Recommendations have no legally bind+ ing standing in Japan. The Building Law of Japan has a separate set o} wind-loading rules - BSL]-2000 (Ministry of Land, Infrastructure and) Transport, 2000). Since the latter does not have a comprehensive set of rules for cross-wind and torsional dynamic response, the AIJ is com| monly used by structural designers for buildings greater than 60m inl height. 15.2.5 Australian/New Zealand standard AS/NZS 1170.2 The joint Australian and New Zealand Standard for Wind Loads, AS/NZS, 1170.2, was issued in 2002, and revised in 2011, as a combined Standard, replacing separate documents from the two countriesvIt is a comprehensive document of about a hundred A4 pages, and is’supported by a separate} Handbook (Australasian Wind Engineering Society; 2012): AS/NZS1170.2 has an indirect legal status in Australia by being’ called} up in the National Construction Code (Australian Building Codes Board, 2019). This document, or Part 0 of the joint Standards on Structural design} actions, AS/NZS 1170.0 (Standards Australia, 2002) must be consulted ‘4 obtain the appropriate annual probability of exceedance for the importanc and use of the structure, before use of AS/NZS 1170.2, Wind actions. The nominal basic wind speed in AS/NZS11/0.2 1s a U.2-second gus! measured at 10m height in open country tertain, and. values are wind for a range of annual recurrence intervals from 1 to 10,000years, for fou!460 Wind Loading of Structures regions. The gust duration was re-defined in 2012 from the value of 3 sec onds given in earlier editions of the Standard. The justification for this wa: given by Holmes and Ginger (2012). For most buildings, excluding thon with large numbers of occupants, and important post-disaster facilities, th annual risk of exceedance for ultimate limit states wind speeds is specified. in the Building Code of Australia, as 1/500. Tall buildings generally ar assessed to have a higher ‘importance level’ and an annual risk of exceed. ance of 1/1,000 is adopted. Other structures such as temporary ones and cyclone shelters may have lower and higher values, respectively. A draft revised edition of AS/NZS 1170.2 was issued in 2020 (Standard: Australia, 2020), and is expected to be published by 2021. This wil include some revision of regional boundaries for the basic regional win speeds and directional multipliers in both Australia and New Zealand. The inland region in Australia (at least 200 km from the coastline) is dom: inated by non-synoptic winds generated by thunderstorms (see Sectiong 1.3.3 and 1.3.5), and terrain-height and topographic multipliers appropri ate to these events will be specified. In the regions of Australia affecte: by tropical cyclones, a linear reduction in wind speed with distance rea the coastline to reflect the weakening of the storms, and a ‘climate chang: multiplier’ to incorporate the apparent increase in stranger severe cyclones (see Figure 1.18) will be introduced. A number of changes to shape factors, with additional data for on-ground arrays of solar panels (solar farms), will also be included. 15.3 BASIC WIND SPEEDS OR PRESSURES Codes and standards for wind loading are currently based on extreme wins speeds with a variety of nominal averaging times. These variations hav occurred for a number of reasons, such as the type of recording system: used by meteorological services to record winds, or the type of extrem wind event that dominates designs for wind in a particular jurisdiction. Some codes are based on wind speeds averaged over relativelydong peri ods, such as ten minutes or one hour. However, the'wind speed is ofte effectively converted to a gust speed within the format of the code, bt calculating building pressures or forces. Gust factors, being the ratio between the expected maximum gust and the mean/val¢, in an averag ing time such as | hour, are therefore important for these conversions (seq Section 3.3.3). The use of a gust wind speed as a basis is particularly rec- ommended for jurisdictions with extreme winds caused. by.transient, non- synoptic wind events, for which hourly-averages arc not appropriate. Most national wind codes and standards are.based on a maximum gust wind speed, with a defined gust duration; it is mostcommon to find it state as a ‘3-second’ gust. The reason for this value/ispartly historical, wiel a perception of this being a typical averaging time of anemometers useWind loading codes and standards 461 to record historical data on gust wind speeds. However, as discussed by Holmes et al. (2014), the effective frontal area associated with a gust of this duration, at typical design wind speeds in synoptic wind events, is equiva- Jent to that of a tall building. For smaller structures, a code or standard} based on a 3-second gust should therefore incorporate a ‘gust effect factor’ or amplification factor, somewhat greater than 1.0, to allow for the lack of reduction due to correlation effects over small frontal areas. Table 15.1 summarizes the basic wind speed characteristics used, or rec ommended, in the five documents surveyed in this chapter. In all cases, the standard meteorological reference position of 10m height in flat, open} country is used. The ISO Standard, as previously discussed, does not give basic wind speeds or dynamic pressures. Annex B in ISO 4354 provides peak factor and gust factors for conversion of wind data with various other averaging times, to the 3-second gust and 10-minute velocities used as a basis fo calculation of wind loads in ISO 4354. For synoptic winds, more compre. hensive terrain- and wind speed-dependent conversions are given in ESDU| $3054 (ESDU International, 1983), and by Holmes et al. (2014). The Eurocode EN 1991-1-4.6, also does not give basic wind speeds. althongh a previons 1994 draft (CEN, 1994) gave ‘reference wind velocities’ for 18 countries in Europe in an informative Annex. National Annexes no’ provide basic wind speed information for individual countries in Europe. The US Standard (ASCE 7-16) contains maps with wind-speed contours. with closely spaced ones for Alaska and the coastal regions adjacent to the} Gulf of Mexico and the Atlantic Ocean. In the latter case, the effects of hur- ricanes are of particular concern. Maps of basic wind speeds for the island of Hawaii incorporate topographic effects. The values of basic wind speed} given on these maps, are peak gust wind speeds, with mean recurrence intervals (MRI) of 300, 700, 1,700 and 3,000 years. The MRI required ra be used depends on the importance of the building, or other structure. Thi basic wind speed values, for the non-hurricane regions of the continental] United States, have changed significantly in the 2016 edition of ASCE-7) Table 15.1 Definitions of basic wind speeds Retum periods/annual Code Averaging time recurrence intervals* 1SO 4354:2009 3 seconds, (IO minutes) Not specified EN 1991-1-4.6 10 minutes Soyears ASCE 7-16 3 seconds 350-700-1,700-3,000 years) Al) (2004) 10 minutes \00years ASINZS| 170.2:201 1 0.2 seconds 500-1 000 years* * For ultimate limit states design. ® A.wide range of annual recurrence intervals are provided:in ASINZS 1170.2 for various limits states, and importance levels.462 Wind Loading of Structures compared with those in the 2010 version, following extensive re-analyses, of anemometer data. The Recommendations of the Architectural Institute of Japan (AIJ gives a detailed map showing contours of the basic wind speed (10 minute mean with 100-year return period). Single values are given for thi outlying territories such as Okinawa. A map of 500-year return rerio values is also given to enable users to interpolate for intermediate retur: periods. In the Australian/New Zealand Standard, the 500-1,000years returt periods shown in Table 15.1 apply to the majority of buildings in Australia (Importance Level 2 or 3 in the National Construction Code of Australia) for assessment of ultimate limit states design criteria. Basic wind speed: are given in the form of maps with five regions, denoted by A, B, C, D ond W. Two of these regions (C and D) comprise a coastal strip exposed to th effects of tropical cyclones (Section 1.3.2). Regional wind speeds are speci fied for each Region as a function of annual probability of exceedance. Thi analysis of extreme wind speeds for Region A, covering most of Australia. in the 2002 Australian Standard was described by Holmes (2002). Some analysis of wind speeds for Regions C and D was discussed by Dorman (1984), 15.4 MODIFICATION FACTORS All the documents include modifiers for the effect of terrain/height and topography, although in the case of ASCE 7, these act on the dynamic pres sure, rather than wind speed. In the Eurocode, the mean wind speed is modified for terrain and height (roughness factor c,), and for topography (described as ‘orography’), c,, ther converted into a gust dynamic pressure at the height of interest, by a facto1 involving turbulence intensity (i.e. a gust factor acting on the dynamic pres sure). The exposure coefficient c,(z) includes terrain/heightand topographic (orographic) effects within Equation (15.8). EN 1991-1-4.6 and AS/NZS1170.2 use a logarithmic law (or a modifica: tion for gust speeds) to define the terrain/height variation, ASCE'7 and ay use a power law variation, and ISO 4354 gives parameters for both. AS) NZS1170.2 allows for averaging of terrain roughness upwind of the site with an interpolation of terrain/height multipliers. All five documents provide factors or multipliers for topographic, or oro: graphic, speed-up effects on wind speeds. None of the documents allo for any shielding effects of topography. However, there are significant dif ferences in the magnitudes of the speed ups predicted by the various docu: ments for the same topographic geometry, as. discussed by Holmes ef al. (2005Sa).Wind loading codes and standards 46: Table 15.2 Calculation formats for velocity, dynamic pressures and building pressure Code Velocity Dynamic Pressure Building PressureiForce| 1804354 Vece= Veet Corp ie= (N12) p (Ve)? P=dute Cp Cyn EN 1991-1-4.6 v,=Car Gessentio l2)= ce) (1/2) p ve We= (2) Spe ASCE 7 v 2) p KK KV? 1 p=4 (GG) All Uy=UskoE kin qv=(122) 2 Usd Weg GGa? ASINZSI170.2Voey=VMMecxy MaMa 4e= (1/2) ParMien P=4 Cig Ciyn * The subscript f denotes D (for walls) or R for roofs in the All. The Australian/New Zealand Standard, AS/NZS1170.2, is unique inj having a ‘Shielding Multiplier’, which allows for reductions in velocit; when there are buildings upwind of greater or similar height Table 15.2 summarizes the formats for calculation of design wind veloci- ties and dynamic pressures in the various documents. 15.5 BUILDING EXTERNAL PRESSURES Table 15.2 also shows the general format for calculation of external pres- sures on wall or roof surfaces of enclosed buildings. The formulae (in the right-hand column) appear to be quite different from each other, but they all contain quasi-steady or mean pressure coefficients (Cy, per Cy Egg) and factors to adjust the resulting pressures to approximate peak values. In the case of ISO 4354 and AIJ, there are gust factors on pres sure (C4,, and G,); in the case of EN 1991-1-4.6, the gust factor is incorpo, rated into the exposure coefficient, c,(z), as discussed previously. In ASCE 7, the quantities G and C, are usually combined together as (GC,) in tables. In AS/NZS1170.2, the ‘aerodynamic shape factor’, Cp, consists of pressure coefficients, multiplied by factors for area reduction. combination of roof and wall surfaces, local pressure effects and porou: cladding. The local pressure factor K, is always greater'than 1, and the arcal reduction factor K,, which allows for correlationseffects over large areas in} separated flow regions, is less than one. AS/NZS1170:2 is'alone'in having’al factor (K,) for porous cladding. The tables of shape factors and pressure coefficients of exterior surfaces of buildings given in the various documents are also sources of significant differences. In all cases, the nominal wind directions are normal, to the} walls of buildings of rectangular plan. ASCE 7 and AS/NZS1170.2 require alternative positive roof pressure} coefficients to be considered. These are important values for the design of frames, especially for those in colder climates where dead loads are often} high, as pointed out by Kasperski (1993).464 Wind Loading of Structures EN 1991-1-4.6 gives tables of external pressure coefficients c,,. which are, comparable to those in ASCE-7 and AS/NZS1170.2, since they are effec- tively applied to gust dynamic pressure through the use of the exposure coef. ficient c,(z). The tables give two values: ¢,.., intended for tributary areag less than 1 m2, ie, local cladding design, and c,...9 intended for major struc tural members. It appears that the numerical values for flat and gable (‘duo- pitch’) roofs in EN 1991-1-4.6 are comparable to those in ASCE-7 and AS/ NZS1170.2, and alternative (positive or lower negative) values are given for most roof pitches. However, no variation with height/width ratio is given. The factors incorporated into the shape factor in the Australia/Ne' Zealand Standard AS/NZS1170.2 for flat and gable-roofed buildings have already been discussed. However, it should also be mentioned that the, effect of tributary area and correlation effects are dealt with by the use of the three factors: K, (area reduction factor), K, (action combination factor and K, (local pressure factor). The action combination factor, K., in th Australian and New Zealand Standard, allows for a reduction when wind) pressures from more than one building surface, e.g. walls and roof, contrib: ute significantly to a load effect. The AJ Recommendations also separate the specification of loads on| the structural frames and on the ‘components and cladding’ of hnildings The specification of pressure coefficients is separated from the specifica- tion of the gust factor. Unlike any of the other documents, the gust factor, Gp, for the loads on the roofs of low-rise buildings has a dependency on| natural frequency. Buildings are classified as those with heights less than, or greater than 45 metres, a somewhat greater height than used in the other documents. 15.6 BUILDING INTERNAL PRESSURES The treatment of internal pressures varies considerably from one document to-another. For buildings with nniformly distribuited openings g#@1991-1 Wregives a graph of c,, varying from +0.35 to -0.5, asa funetion of an opening ratio, x. For a dominant opening, the internal presstffevetfivlenteexprenvedan a fraction of the external pressure coefficient ofthe face with the" opening This document also gives fairly detailed puidanée off petssuregon walls shd roofs, with more than one skin. ASCE 7-16 (in Clause 26:13), specifies four different situations; open, partially open, partially enclosed, and enclosed BUfGitns, and'specities vit ues of GC,, between +0.55 and —0.55. A feature, not found in the other standards, is a reduction factor, R,, for large biilldihg volumes. 'AS/NZS1170.2 gives two tables with varios péshegSad Gogabve Ptud of internal pressure coefficients, G,,,. For one of these tables/the valuedWind loading codes and standards 465, depend on the ratio of dominant openings on the windward wall to th total open area on other walls and roof. [SO 4354 gives similar interna pressure coefficients to those in AS/NZS 1170.2. The AIJ Recommendations do not specify a positive internal pressure, dic. the: possibility of dominant openings is tiot-considered, For building) without dominant openings, values of C,, of 0 or 0.5 are specified. 15.7 OTHER SHAPES AND SECTIONAL FORCE COEFFICIENTS Apart from the AIJ Recommendations, which is intended exclusively foi buildings, all the surveyed documents contain shape, or force, coefficient: for a variety of structure shapes and cross sections. Table 15.3 sainerizel the data given. The data in all these documents appear to be based on modern wind: tunnel measurements for the most part. EN 1991-1-4.6 clearly contains th most comprehensive set of data. ISO 4354 only provides a limited amount of data on shape factors in Annex D, but suggests that other sources may| be used in conjunction with standard, but with appropriate adjustment fo! gust averaging time and exposure. These sources include other codes 01 standards. 15.8 DYNAMIC RESPONSE CALCULATIONS The five standards contain procedures for the calculation of dynami response for wind-sensitive structures, such as slender, flexible, lightly damped tall buildings. ISO 4354, in Clause E.2.2, has a relatively comple: set of numerical criteria to determine whether a structure is ‘dynamically sensitive’ in the along-wind, cross-wind and torsional modes. EN 1991. 1-4:2005, in Clause 6.2, also has a set of empirical criteria to'determin: whether a structure need not be treated as dynamics however, generally buildings with height to along-wind depth (ie: h/d) greater than four, require calculation of the structural factor. ASCE 7 and“AS/NZS1170: classify wind-sensitive structures as those with a first-mode natural fre quency less than 1 Hz; ASCE 7 also requires a height-to-breadth (or -depth ratio greater than four. In ISO 4354, information for calculation of along- wind response is given in two alternative formats — one for use with a (10-minute) mean dynam pressure (Clause E3), and a second for use with the (3-second) pea dynamic pressure (Clause £4). These are respectively in the gust respons’ factor discussed in Section 5.3.2 of this book, and the dynamié responsd factor format of Section 5.3.4. However, the user should be/aware that,466 Wind Loading of Structures Table 15,3 Shape factors contained in five documents (excluding rectangular enclosed buildings) Type 180.4354 EN 1991 ASCET A) ASINZSI170.2 Stepped roofs No No Yes No No Free-standing walls, Yes (walls only) Yes Yes No Yes hoardings Parapets No Yes? Yes No No Free-standing roofs Yes Yes Yes Yes Yes (canopies) Attached canopies No No No No Yes Multispan roofs No Yes Yes Yes* Yes (enclosed) Multispan canopies No Yes No No No Arched roofs No Yes Yes Yest Yes Domes No Yer Ye: Yes? Yes Bins, silos, tanks Yes Yes Yes No Yes Circular sections Yes Yes Yes Yes Yes Polygonal sections No Yes Yes. No Yes Structural angle Yes Yes No No Yes sections Bridge decks No Yes No No No Lattice sections Yes Yes Yes No Yes Flags No Yes No No Yes Sphere No Yes No No Yos Roof-mounted solar No No Yes No Yes panels * Given in commentary section. © Treated as free-standing walls. © Treated as arched roofs. in the latter formulation, the specified peak factor, g,, of 3.4 in Equatio: (E.4.3) of ISO 4354 is too high when used with gust pressures derived fro a 3-second gust, and is incompatible with Table B.1 in Annex B of ISO435: (see also Table 3.4 in this book) — potentially leading to un-conservati values of Caja In Section E6.1 of ISO 4354, some recommended values of structural damp: ing are given as a function of construction material and building height. Thes values are apparently for ultimate limit states design criteria, as 75% of the specified values are recommended for ‘habitability to horizontal vibrations’. ISO 4354 recognizes the importance of vortex. shedding in. causing dynamic cross-wind effects in slender prismatic.and cylindrical structures. For rectangular cross-sections, the critical wind speed at which large ampli tude motions may result, can be calculated, as.a function of side ratio and Scruton number (see Sections 5.5.2 and 11.5.1).Wind loading codes and standards 467] A ‘structural factor’, in EN 1991-1-4.6, denoted by ¢,¢4, incorporates th combined effects of size (c,) and resonant dynamic response (c,) for overal structural loads, or loads on major structural elements. The structural fac- tor is equivalent to the ‘gust effect factor’ for dynamic structures in ASCE] 7, and the ‘dynamic response factor’ discussed in Section 5.3.4. This factor is given in the Eurocode by: 14 2kply (<<) VB? +R T+7-I(e) (15.9) z, is a reference height. For buildings and towers, this is normally taken] as 60% of the roof height. k, is a peak factor for the response. I,(z,) is the longitudinal turbulence intensity at the reference height. B? is a background factor. R? is a resonance response factor. As discussed in Clause 6.3.1 in EN 1991-1-4.6, the structural factor actually specified as a combination of a size factor, ¢, with a dynamic factor, ¢,, to form the combined structural factor, ¢,c,. The size factor separatel takes account of correlation effects. Alternative methods of calculation of the factors, k,, B* and R*, are given] in Ammexes B and C, with a simplified graphical method for estimation of ¢,¢ys given in Annex D. Alternative expressions for the standard deviation| of along-wind acceleration, for serviceability limit states, are also given in| Annexes B and C. Comprehensive information, including working equations, regarding vortex excitation and other aeroelastic effects such as vortex-induced larg. amplitude lock-in type vibrations, galloping (Section 5.5.2 in this book). various types of interference excitations, and flutter (Section 5.5.3) are, included in Annex E of EN 1991-1-4.6. Recommended calculation pro} cedures for dynamic structural properties, including natural frequencies) mode shapes, equivalent masses and logarithmic decrement, are given in} Annex F of EN 1991-1-4.6. In ASCE 7-16, an analytical procedure for determination OF a "gust effect factor’, G, for the along-wind vibrations of flexible buildings and other structures is presented in the commentary section of that Standard, Thi development of this factor was described by Solari and Kareem (1998) The gust effect factor is, in fact, a dynamic response factor (Section 5.3.4), defined in the same way as the structural factor, cc, in EN 1991-1-4.6, ie. it is based on a 3-second gust wind speed, as. adopted by AS Th calculation procedure is nearly identical to thatin EN 1991-1-4.6, makin use of the closed-form equations of Solari (1983): Expressions for maxi- mum _along-wind displacement and standard deviation and’ maximu:468 Wind Loading of Structures along-wind acceleration are also given. However, no analytical procedurd for cross-wind response is given. In the AIJ Recommendations, a detailed procedure is applied to esti mate the dynamic response of wind-sensitive structures. For along-wind response, a standard gust response factor approach along the lines o! Equation (15.3) is used to determine a gust effect factor Gp. A mode shap correction for prediction of peak base bending moments for buildings withl nonlinear mode shapes is provided (Tamura et al., 2004). Vortex-induced cross-wind vibration and wind loads can also be deter} mined from the AIJ, based on r.m.s. cross-wind base moment data obtained from wind-tunnel tests. Expressions for effective cross-wind load distribu: tions, displacement and acceleration are given. However, the cross-wind response calculations are restricted to prismatic cross-sections with height-to-breadth ratio no greater than six, and to wind directions norma toa face of the building. Expressions for torsional angular acceleration and torsional wind load distribution are also given. Guidelines for assessing potential aeroelastic instabilities including lock-in type vortex resonance and galloping instabilities are presented. The dynamic along-wind response of tall buildings and towers is dealt with, throngh a dynamic respanse factor, Cy,,,, in Section 6 of ASINZS1170.? this corresponds with the definition of dynamic response factor given irl Section 5.3.4 of this book. The methodology is a greatly modified version} of the description given by Vickery (1971). Cross-wind base overturning moment and acceleration can be deter- mined from cross-wind force spectrum coefficients, derived from wind: tunnel test data for a series of square and rectangular section buildings with the incident wind normal to a face. Suggested values of damping fo: a range of steel and concrete structures under different stress levels are given. The importance of aeroelastic instabilities, such as lock-in, gal loping, flutter and interference are discussed separately in the Handbool for AS/NZS1170.2. However, a ‘diagnostic’ method for the crosswind response of chimneys, masts and poles of circular cross*ection is pro: vided in the Standard itself. 15.9 INTER-CODE COMPARISONS Several numerical comparisons have been made of wind load calculation: by various international codes and standards. The results from.some o! those comparisons are summarized here. General comparisons between major wind-loading codes and standard: have been made by Cook (1990), Mehta (1998),.and Zhou et al. (2002) fo: dynamic effects. A special issue of the journal. Wid and Structures com: prised five papers in which all aspects of codification for wind loads werdWind loading codes and standards 464 reviewed (Holmes et al., 2005a, b; Tamura et al., 2005; Letchford et al. 2005; Kasperski and Geurts, 2005). Holmes et al. (2009) described an extensive comparison of calculation: of wind loads on three buildings from fifteen codes and standards in thd Asia-Pacific region. The three buildings, comprising a generic low-rise, al medium-rise and a high-rise building, are shown in Figure 15.2. The low- rise building was a typical steel-framed portal frame structure located in a] rural area; the medium-rise building was a 48 m high office building locate inan urban area, and the high-rise building was 183m tall and also inate in an urban area. In all three cases, design wind speeds at the top of eacl building were pre-specified; wind speeds with averaging times of 3 seconds. 10 minutes and 1 hour were all specified, and the participants selected an| appropriate time according to the stated averaging time in their own cod or standard at the time of the comparison. The comparisons for the low-rise building showed coefficients of varia. tion of 20%-31% in the net pressures across the building surfaces. Thesd relatively high values were partly caused by differences in the treatment o} internal pressures, since a large opening in the windward wall of the build ing was specified. Somewhat smaller coefficients of variation of 13%-26%| were obtained for the wind loads on a large door and a small window in} the building. For the medium-rise building, the coefficients of variation for the pre- dictions of base shear and bending moments, and cladding pressure: were consistent at 22%-23%. The coefficients of variation for along: wind and cross-wind base shears and moments on the 183m high rise building were relatively small at 14%-17%. This was attributed to th common origin of many of the code provisions for dynamic response t wind. Holmes (2014) described a comparison of the responses of a generic tal building as calculated by three codes and standards, with the consensus along-wind response determined by several wind-tunnel groups using th high-frequency base balance technique (see Section 7.6.2). The Bayarisons of Holmes (2014) are interesting because of the direct Comparison of cod values with wind-tunnel data. For wind acting normal to the wide face ot the tall building in question (similar in dimensions to the high-rise buildin shown in Figure 15.2), AS/NZS 1170.2:2011 produced along-wind gd bending moments that are within -4% to +8% of the averages of the wind. tunnel data (for three different wind speeds). ASCE 7-10 produced value 8%-17% below the average of the wind-tunnel data, andthe Hong Kony Code of Practice at the time (HK Buildings Department, 2004) produce values 27%-33% below the average of the wind-tunnel data. The pape: gave explanations for the discrepancies; in the case of the Hong Kong Codg, of Practice, the differences were directly associated with low drag cef- ficients specified in the code. That discrepancy has been addréssed in the470 Wind Loading of Structures Q Ass m 5 bays @5 m=25m ‘Span = 15m. (a) 48m (Holmes et al., 2009): (a) low-rise industrial building, (by medium-rise offic Figure 15.2 Three buildings used for an inter-code comparison in the Asia-Pacific regior building and (c) high-rise building.Wind loading codes and standards 47] 2019 revision of the Hong Kong Code (Buildings Department, Hong Kong, 2019). 15.10 OTHER CODES, STANDARDS AND DESIGN GUIDES The codes and standards discussed in earlier sections of this chapter are primarily intended for wind-loading design of habitable buildings, which, in most developed countries, is controlled by legislation. The extent t which other structures are covered in the major codes and standards var- ies greatly. For example, shape factors for the wind loading of bridges ar included in Eurocode 1, but not in the other documents. Another difference is the extent to which wind-tunnel testing has been standardized. In ASCE 7, a separate chapter (Chapter 31) covers the ‘Wind-Tunnel Procedure’. This mainly restricts the reduction in design loads that can be obtained from| wind-tunnel testing compared to those from the main procedures of thd standard. In many cases, other standards and design guides have been issued) for structures apart from buildings. Some of these are discussed in the following. ‘As discussed in Chapter 12, the aerodynamics of long-span bridges is al specialist topic and normally involves extensive wind-tunnel testing, which| only a relatively few wind tunnels are capable of undertaking. As mentioned in Section 12.6, a basic standard published in the United Kingdom, BD 49/01] (Highways Agency, 2001), sets out some design rules for the aerodynamid design of bridges with useful limiting parameters for dynamic instabilities such as flutter and galloping of bridge decks. Requirements for wind-tunnel testing, including boundary-layer modelling, topographic effects and sec; tion, and full aeroelastic modelling are given in this document. Eurocode 3 for ‘Design of Steel Structures’ (CEN, 2006) includes wind: loading information for steel communication towers and masts, such as sectional drag coefficients for lattice tower sections of yatious:solidities| A similar document, AS 3990 (Standards Australia, 1994), was issued in} Australia, although the wind-loading information in that document = now been updated and included in the main wind-loading ‘standard, AS) NZS 1170.2. Transmission line structures — both high-voltage and local distribution! towers and poles — have particular vulnerability to small local non-synoptiq storms. This is reflected in Australian/New Zealand StandardAS/NZ: 7000:2016, (Standards Australia, 2016) a standard for design of overhead lines, which provides special rules for extreme winds from convective down- drafts, including increased span reduction factors (see Section 13.2.3). Ihé, ASCE Guidelines (ASCE, 2020) perform a similar function, although tech- nically the document is not an enforceable standard.472. Wind Loading of Structures The proliferation in a short time of large-scale energy harvesting by arrays of solar panels on inclined frames has left wind codes and standards, struggling to provide up-to-date wind-load information for the supporting structures, the design of which is governed by wind loads. For panels on flai or near-flat roofs, the guidelines of the Structural Engineers Association o} California have provided very useful design information (SEAOC, 2017), This document also includes loads for arrays of panels on ground — i.e! ‘solar farms’ (see also Section 14.7). Wind-tunnel testing, although widely used throughout the world as al method of determining wind loads (see Chapter 7), has been lightly reg: ulated. Two documents that provide minimum recommended criteria for modelling and processing of data are an American Standard, ASCE/SEI 49-12 (ASCE, 2012), and a Quality Assurance Manual of the Australasia‘ Wind Engineering Society, AWES-QAM-1-2019 (AWES, 2019). Thes documents are widely used and quoted, but are, up to now, not mandatory, 15.11 GENERAL COMMENTS AND FUTURE DEVELOPMENTS This chapter has reviewed the provisions of five major and current (at th time of writing) standards for wind loading. Considerable differences exist in both format and the type of information presented in these documents. At present, the international standard on wind loading, ISO 4354:2009, is not generally used for structural design. It is difficult to use it as an oper: ating standard for design, without an accompanying set of wind climate data — i.e. the first link in the wind-loading chain of Figure 15.1. ISO 4354) has been adopted by Ethiopia, but without the necessary basic design wind speeds. Eurocode | (EN 1991-1-4.6), however, has been adopted in nearly all o! the European countries. It has also been adopted in South Africa and i parts of Asia. ASCE 7 is also widely used outside of the United States, such! as in the Middle East, and earlier versions have been adapted for certain| Asian jurisdictions (such as the Philippines and Taiwan). AS/NZS 1170. is widely used in the islands of the South Pacific, and an earlier version hat been adapted for Malaysia. It seems that a true international standard, although desirable for wind loads, remains an unachieved objective so far. The first requirement is aj common format and notation. As indicated by Table 15.2, there are signifi cant differences in the notation used across the codes, for what are essen} tially the same parameters. For wide international acceptance in tropical and sub-tropical, as well as temperate, climates, the special requirements of regions affected \by typhoons (tropical cyclones or hurricanes) and thunderstorms will need tWind loading codes and standards 47: be incorporated. Some greater consistency in the treatment of effects such| as directionality, topography and internal pressures could be achieved, as there is general agreement amongst researchers on the underlying physical and statistical principles behind these effects. One problem appears to be that in some jurisdictions, loading codes arq the responsibility of government agencies, and regulators without the neces sary wind engineering expertise, or practical experience that comes from] using the documents for design. In other cases, the users of the document (i.e. practicing structural engineers) are the dominant ‘players’ in the devel- opment of a document. In further cases, academic wind engineers may bq dominant, possibly leading to a document that is generally technically ‘cor rect’, but may contain ambiguities and is generally user-unfriendly. Ideally all three groups should play a role in the development of a wind code o1 standard to achieve a satisfactory user-friendly, non-ambiguous document, that is also technically acceptable. REFERENCES American Society of Civil Engineers, (2020) Guidelines for electrical transmission, line structural loading, 4th Edition. ASCE Manual of Practice 74, ASCE, Reston, VA. American Society of Civil Engineers. (2012) Wind tunnel testing for buildings and other structures. ASCE Standard ASCE/SEI 49-12. ASCE, Reston, VA. American Society of Civil Engineers. (2016) Minimum design loads and associated criteria for buildings and other structures. ASCE/SEI 7-16. ASCE, Reston, VA. Architectural Institute of Japan. (2004) AI) recommendations for loads on build. ings. AlJ, Tokyo. Australian Building Codes Board. (2019) National Construction Code of Australia, ABCB, Canberra. Australasian Wind Engineering Society, (2012) Wind loading handbook fo Australia and New Zealand ~ Background to AS/NZS 1170.2 Wind actions, AWES-HB-001-2012. Australasian Wind Engineering Society. (2019) Wind engineering studies of build: ings. AWES-QAM-1-2019. Buildings Department, Hong Kong. (2019) Code of practice on wind effects ‘a Hong Kong, Hong Kong Special Administrative Regiony-Chinay Septembe 2019. CEN (European Committee for Standardization), (1994) Eurocode 1: basis of design and actions on structures, Part 2-4: Wind actions (draft). ENV. 1991-2-4, CEN, Brussels, Belgium. CEN (European Committee for Standardization). (2005) Eur6é6der1?"Aetions on} structures - Part 1-4: General actions - Wind:actions»EN-1991-1=4yCEN, Brussels, Belgium. CEN (European Committee for Standardization); (2006), Furocode 3: Design of steel structures - Part 3-1: Towers, masts and chistineys. EN 1993-3-1:2006, CEN, Brussels, Belgium.