Comparison and harmonization of building wind loading codes among the Asia-Pacific Economies

Abstract

This paper reviews wind loading codes and standards in the Asia-Pacific Region, in particular in the 15 countries and areas. A general description of wind loading model is given as a famous wind loading chain described by four variables including velocity pressure, exposure factor, pressure coefficient, and gust response factor. Through the APEC-WW Workshops and the extensive calculations for three examples of low, medium and high rise buildings, these four important variables of wind loads are evaluated and compared with statistical parameters, mean values and coefficients of variation. The main results of the comparison show some differences among the 15 economies, and the reasons and further incorporation are discussed and suggested.

Keywords

wind loading / codification / velocity pressure / exposure factor / pressure coefficient / gust response factor

Introduction

After John Smeaton of England originated a formula for wind pressure loads in 1759, wind actions on structures and structural elements have to be considered in the design as one partial load among various design loads. To determine wind actions on structures, each country needs to have appropriate codification to specify wind loading and to determine wind induced responses in structural design, which results in numerous wind loading codes and standards in the world, for example, the ASCE Code [1], the Australian and New Zealand Standard [2], the National Building Code of Canada [3], the Japan Recommendations [4], the European Standard [5], the International Organization for Standardization [6], and so on. Under the globalization of construction industry and the development of unified international codes and standards, it is necessary to better understand and compare the underlying differences among international or regional wind loading standards in order to further incorporate for future alignments of wind loading and even wind resistance design codes and standards.

The previous studies on the major international standards mentioned above have found that the dominant contributions to the scatter in wind loading were the varying definitions of wind field characteristics, including mean wind velocity profile and some turbulence wind parameters [7,8]. Some other published papers and reports for the Asia-Pacific Economic Cooperation (APEC) countries and areas have shown the significant importance on extreme wind speeds of tropical cyclones and the other extremes of benign monsoons and local thunderstorm downdrafts for design [9,10].

With the support of the Centre of Excellence (COE) and the Global COE in Wind Engineering at Tokyo Polytechnic University in Japan, a new practical outcome of comparative study on wind loading codes and standards among a regional area composed of a group of bordering countries or areas have been launched through six Workshops on Regional Harmonization of Wind Loading and Wind Environmental Specifications in the Asia-Pacific Economies (APEC-WW) since 2004. At the 2nd APEC-WW in Hong Kong in 2005, three particular examples were purposely assigned for each country or area representing three typical building models, including a low-rise building, a medium-rise building and a high-rise building. In the subsequent two Workshops, the design wind loads on three building examples have been evaluated and compared in accordance with the wind loading codes and standards of 15 Asia-Pacific Economies [11]. The basic results of three examples and the obvious reasons for differences were summarized by J. Holmes, Y. Tamura and P. Krishna [12]. A series of papers related to benchmark analysis of these three typical buildings were published and presented in the 7th Asia-Pacific Regional Conference on Wind Engineering [13-19], and the further discussion were made on regional wind velocity map, unified terrain categories and model code for low-rise buildings in the 5thand 6thAPEC-WW Workshops at Chinese Taipei in 2009and at Gangwondo Korea in 2010.

With the background of the APEC-WW Workshops, this paper is going to make further quantitative and statistical comparison and contrast of wind loading components based on various codes and standards in 15 Asia-Pacific Economies, including the Australian and New Zealand Standards (AN) [2], the National Building Code of Canada (CC) [3], the China National Standard (CS) [20], the Code of Practice on Wind Effects of Hong Kong (HK) [21], the Indian Standard Code (IC) [22], the Standard National Indonesia (IS) [23], the Recommendations for Loads on Buildings of Japan (JR) [4], the Korean Building Code (KC) [24], the Malaysian Standard (MS) [25], the National Structural Code of Philippines (PC) [26], the Singapore Standard (SS), the Taiwan Building Code (TC) [27], the Wind Load Code of Thailand (TC) [28], the United States’ ASCE Code of [1] and the Loads and Actions Norm for Design of Vietnam (VN) [29].

A general description of wind loading model can be given by a well known process, a wind loading chain, proposed by A.G. Davenport [30], and consisted of four components

(1)W=q{C}_e{C}_p{C}_g,

in which q is a reference wind pressure or velocity pressure mainly depending on wind velocity, C_e is an exposure factor to adjust for the terrain conditions and the height, C_p is a pressure coefficient related to structural shape and C_g is a gust response factor (GRF) due to turbulent wind actions (gust loading factor GLF) or structural dynamic response (dynamic response factor DRF). These four components are numerically calculated and statistically analyzed through three building examples and based on the 15 wind loading codes and standards of Asia-Pacific Economies in this paper.

Velocity pressure

Reference velocity pressure q can be simply described by the square of reference wind velocity U as follows:

(2)q={\displaystyle \frac{1}{2}}\rho U^{\mathrm{2}},

in which ρ is a air density and has the value of 1.16 to 1.25 kg/m³ with the average of 1.22 kg/m³ and the coefficient of variation of 2% among 15 codes or standards listed in Table 1 [11], and the value of reference wind velocity U basically depends upon three conditions, including reference height, averaging time and return period, which are discussed as follows.

Wind velocity varies with height above the ground in the atmospheric boundary layer, which can be described by wind velocity profiles with either Power Law or Logarithmic Law in most wind loading codes or standards. The values of design wind velocity are different at different levels, and are used to be calculated from the basic level or the reference height, at which the reference or basic wind velocity is defined. Table 1 shows that the unified reference height of 10 m is used in all 15 economies’ codes or standards.

The values of wind velocity are largely controlled by the averaging time since natural wind velocity fluctuates with time. The shorter averaging time is the higher value of wind velocity will be since the maximum wind velocity sample is always chosen in consideration. There are generally three kinds of averaging time adopted in wind loading codes and standards, that is, 3 s, 10 min and 1 h. Theoretically, the averaging time should be determined by dominant extreme wind events, for example, 3 s for thunderstorm downdrafts or outflows, 10 min for tropical cyclones or typhoons and 1 h for extratropical gales. In practice, each code or standard provides only one out of three kinds of averaging time shown in Table 1. If the ratio of wind velocity values is assumed to be 1.5:1.0:0.92 for the averaging time of 3 s:10 min:1-h, the mean value and coefficient of variation of wind velocity values due to averaging time were computed and listed in Table 1 based on the 15 economies’ codes or standards.

Design wind velocity is also governed by the return period in the way that its value increases with the increase of a return period. The return period for the design wind velocity is 50 years, 100 years or 500 years among the 15 economies’ codes or standards in Table 1. Although there are good reasons for selecting one or another, the most popular return period is 50 years. With the assumption of Gumbel Distribution of wind velocity, the wind velocity ratio of R-year to L-year return periods can be expressed as follows:

(3){\displaystyle \frac{U_R}{U_L}}=a_L-b_L\ln \left[-\ln \left(\mathrm{1}-{\displaystyle \frac{\mathrm{1}}{R}}\right)\right],

in which U_R and U_L are the wind velocities for R-year and L-year return periods, respectively, a_L and b_L are constants for L-year return periods, and R is a return period. Taking the result from the statistical study for wind velocity in Shanghai [31], a_L and b_L are equal to 0.625 and 0.096 for 50 year return period, respectively, and accordingly the ratio of wind velocity values is 1:1.07:1.22 for the return periods of 50-years:100-years:500-years. The statistical analysis results of different wind velocity values due to return period are summarized in Table 1, but the mean value and the coefficient of variation of wind velocity do not include the value due to Australian & New Zealand Standards

The velocity pressure ratio η_q is defined for representing relative velocity pressure as follows:

(4){\eta }_q＝{\displaystyle \frac{q}{q_{\mathrm{0}}}}={\displaystyle \frac{\rho U^{\mathrm{2}}}{{\rho }_{\mathrm{0}}{U}_{\mathrm{0}}^{\mathrm{2}}}},

in which ρ₀ is a reference density of 1.20 kg/m³, which is the most popular value among the 15 economies’ codes and standards, and U₀ is a basic wind velocity defined in the conditions of the reference height of 10 m, the averaging time of 10 min and the return period of 50 years. The velocity pressure ratios η_q for the 15 economies’ codes or standards are computed and listed in Table 1, and the mean value and the coefficient of variation of the pressure ratios excluding the value due to Australian & New Zealand Standards were calculated to be equal to 1.347 and 43.4%. Among the relative differences of air density, averaging time and return period, the most dominant difference comes from averaging time from Table 1.

Exposure factor

It has been well recognized that the variation of wind velocity above the ground can be described by wind profile, which is related to drag on the wind as it blows over upstream terrain. Since the drag mostly depends upon the surface roughness of the upstream terrain different roughness effects produce different types of terrain categories or different wind profiles. To represent these varying roughness conditions, different wind profiles are specified in different wind loading codes with the pattern of profile law and the number of terrain categories. Although there are three patterns of wind profile, including Power Law, Logarithmic Law and Deaves & Harris Model, currently used in the 15 economies’ codes and standards, the most popular one is Power Law, for example, 13 out of 15 economies’ codes and standards shown in Table 2, described as

(5)U_z=U_{\mathrm{0}}{\left({\displaystyle \frac{z}{z_{\mathrm{0}}}}\right)}^{\alpha },

in which U₀ is the reference wind velocity at the reference height of z₀, U_z is the particular wind velocity at the height of z, and α is the exponent of Power Law due to surface roughness of upstream terrain.

The number of terrain categories specified in the 13 economies’ codes and standards is also very different. For example, Hong Kong has only one terrain category whereas Japan has five categories as shown in Table 2. The maximum and the minimum values of exponent α and the corresponding gradient heights δ are collected in Table 2 (through personnel communication with the participants of APEC-WW). To make the comparison and contrast of exposure factors, the basic values of α and δ are also provided for transferring wind pressure from the basic terrain roughness, in which basic wind velocity is defined, to the maximum or the minimum terrain roughness. This transformation can be done only through the condition that wind velocity always keeps in the same value at the gradient height. The exposure factor ratio η_e, therefore, can be defined as follows

(6){\eta }_e={\left({\displaystyle \frac{{\delta }_b}{z_0}}\right)}^{\mathrm{2}{\alpha }_b}{\left({\displaystyle \frac{z_{\mathrm{0}}}{{\delta }_s}}\right)}^{\mathrm{2}{\alpha }_s},

in which α_b and δ_b are the exponent value of Power Law and the gradient height (m) related to the basic terrain roughness, α_s and δ_s are the exponent value of Power Law and the gradient height (m) related to the specific terrain roughness, and z₀ is the reference height assumed to be 10 m in Table 2. The exposure factor ratios η_e in the maximum and the minimum values for the 14 economies’ codes or standards were computed and listed in Table 2, and the mean values and the coefficients of variation of the exposure factor ratios were calculated to be equal to 1.23 and 14% in the minimum terrain category, and 0.356 and 65% in the maximum terrain category, respectively. The CoV of exposure factor ratio in the maximum terrain category is much larger than that in the minimum terrain category. The main reason can be attributed to the fact that the exponent α_max is much more scattered than the exponent α_min shown in Table 2.

Pressure coefficient

The comparison of pressure coefficients related to structural shape were made among the 15 economies’ codes and standards through two typical models, including a low-rise building and a medium-rise building. The low-rise building is a typical steel-framed warehouse located in a rural area with open terrain all around in Fig. 1. Design wind speeds at the top of the frame, 6m, are assumed to be equal to 39m/s, 26m/s and 23m/s for the averaging times of 3 s, 10 min and 1 h, respectively. The main calculations include the net pressure coefficients at A, B, C and D of gable walls or roof of the frames at the end of the building and the maximum and minimum wind pressures on the 3 m × 4 m roller door on SW wall and the 1 m × 1 m window on NE wall, in which internal pressures from a large opening were considered for some wind directions [11]. Tables 3 and 4 respectively list pressure coefficients on the north-west frame and cladding pressures(kPa) on the roller door on the SW wall and on the windows on the NE wall of the low rise building based on the 15 economies’ codes and standards [11,12]. For the frame structure, the coefficients of variation of pressure coefficients are about 45% to 47% except 178% in Part A in wind SW and between 45% and 63% in wind NW, respectively. The maximum or minimum cladding pressures have better statistical results, the CoV being 25% for maximum value and 32% for minimum one on the roller door on SW wall, and13% for maximum value and 21% for minimum one on the window on NE wall, respectively.

The medium-rise building is a 48 m high, 60 m long and 30 m wide office building located in a tropical city with suburban terrain for all directions in Fig. 2. The building is of reinforced concrete frame construction with a facade consisting of mullions spaced at 1.5 m. The building is assumed to be air-conditioned with non-opening windows, and can be considered effectively sealed with regard to internal pressures. Design wind speeds at the top of the building, 48 m, are assumed to be equal to 56, 36 and 33m/s for the averaging times of 3 s, 10 min and 1 h, respectively, and a turbulence intensity of 0.20 at the top is assumed. Table 5 compares the cladding pressures on window elements near the corners at the top level. Although the pressure coefficients are scattered with the CoV of 48% for the maximum values and 40% for the minimum values [11], respectively, the comparison is better in the cladding pressures, the CoV being about 21% to 22% [12].

Gust response factor

Among the 15 economies’ codes or standards, gust response factor (GRF) is specified to take into account of turbulent wind actions on stiff structures with gust loading factor (GLF), such as the above-mentioned low-rise building and medium-rise building, and structural dynamic response of very flexiblestructures with dynamic response factor (DRF), such as the high-rise building in Fig. 3. Table 5 also shows the comparison of gust loading factors and base forces of the medium-rise building due to the 15 economies’ codes [11]. The mean value and the CoV of GLF are equal to 1.31 and 39%, which shows quite large differences. The calculated values of base forces, however, reached to quite small CoV, 22% in base shears and 21% in base bending moments [12], respectively, which demonstrates no significant correlation between GLF and base force.

The high-rise building, shown in Fig. 3, was 183 m high, with the rectangular cross section of 46 m by 30 m located in urban terrain. The building was assumed to have an average density of 160 kg/m³, and linear mode shapes in both sway directions with natural frequencies of 0.20 Hz. The structural damping ratio was specified to be 0.012 for base force calculation. Design wind speed at the top of the building, 183 m, was assumed to be 59 m/s for 3 s averaging time, 41 m/s for 10 min and 37 m/s for 1 h, respectively, and a turbulence intensity of 0.17 at the top was also assumed. For wind direction normal to the 46 m wall, only along-wind loading and base force are discussed in the following.

Table 6 shows the characteristics of wind loading and base force of the high-rise building. Design wind speeds were provided in three kinds of averaging time among the 13 economies’codes or standards excluding Singapore and Vietnam standards. The four standards representing Australian & New Zealand, Indonesia, Malaysia and Philippines adopted design wind speed of 3 s averaging time. Other four standards of China, Japan, Korea and Taiwan, China used 10 min design wind speed, and the rest five standards of Canada, Hong Kong, China, India, Thailand and United States used 1h values. These design wind speed values resulted in quite large CoV, 43%, of the velocity pressure q_H at the top of the building [11]. The exposure factor C_e was defined as

(7)C_{\mathrm{e}}={\displaystyle \frac{{\displaystyle {\int }_{\mathrm{0}}^H{\left({\displaystyle \frac{z}{H}}\right)}^{\mathrm{2}\alpha }\mathrm{d}z}}{H}}={\displaystyle {\int }_{\mathrm{0}}^H{\left({\displaystyle \frac{z}{H}}\right)}^{\mathrm{2}\alpha }\mathrm{d}\left({\displaystyle \frac{z}{H}}\right)}={\displaystyle \frac{\mathrm{1}}{\mathrm{1}+\mathrm{2}\alpha }},

in which H is the height of the building and α is the exponent of Power Law. The calculated mean value and CoV of C_e are equal to 0.717 and 9% in Table 5. The pressure coefficient C_p was specified in each code or standard, and its mean value and CoV are 1.28 and 6%. Both C_e and C_p have very small value of CoV. The GLF or DRF C_g was provided in each standard or code, and the mean value and CoV are 1.80 and 35%, respectively. It is interesting to see that the CoV of the wind loading W is only about 14%, which is much smaller than that of q_H or C_g. To find out the reason of this difference, the coefficient of variation of the product of q_H⋅C_g is purposely calculated, and the CoV of the product is tremendously reduced to 14%, which shows significant correlation between velocity pressure and GLF or DRF. Furthermore, the calculated values of base forces supported to similar CoVs, 14% in base shears and 13% in base bending moments [12].

Conclusions and incorporation

This paper examines the differences and similarities of wind loading codes or standards in 15 Asia-Pacific Economies. Following wind loading chain, four variables including velocity pressure, exposure factor, pressure coefficient and gust response factor were evaluated and compared with mean values and coefficients of variation. From the comparison and contrast of wind loading calculations of three typical buildings, the conclusions and further harmonization can be reached as follows:

1) Velocity pressure q mainly depends on four parameters including air density, reference height, averaging time and return period. Since both air density and return period have very small coefficients of variation of 2% and 3%, respectively, and the reference height of 10 m is uniformly adopted in all 15 economies’ codes or standards, the only major variation comes from averaging time, which has the CoV of 23% and contributes over 40% CoV to velocity pressure. To harmonize the calculation of velocity pressures, further incorporation in the Asia-Pacific Region should be considered in developing agreed regional wind velocity maps for 3 s, 10 min and 1 h extremes with 50-year return period.

2) The number of terrain categories varies from one in Hong Kong to five in Japan, and the exponent values of Power Law are between 0.07 and 0.15 in the minimum category and between 0.11 and 0.36 in the maximum category, respectively. These scattered values resulted in the exposure factor CoVs of 15% in the minimum category and 67% in the maximum category. Future harmonization should begin with simplification and unification of terrain categories for surface roughness exposures, in particular for basic or reference terrain category.

3) Pressure coefficient has rather large coefficients of variation, for example, 20% to 61% in the low-rise building and 40% to 48% in the medium-rise building, and cladding pressure has relatively smaller CoVs, between 13% and 26% in the first building and between 21% and 22% in the second building. Although the CoV differences between pressure coefficient and cladding pressure need to be identified, the main cause of quite large CoVs would seem to be on the fact that different standards have different wind tunnel testing sources on which the coefficients have been based. This could be resolved by benchmark study through both wind tunnel testing of scaled models and site measurement of prototype structures in the future.

4) Gust response factor is generally specified to take into account of structural dynamic response and turbulent wind actions. The former is totally governed by structural flexibility, and can be called as dynamic response factor (DRF), which has no correlation with velocity pressure. The latter includes the main account for varying averaging time, and can be defined as gust loading factor GLF, for example, GLF being 1.92 and 2.06 from 3s wind velocity pressure to 10 min and 1h velocity pressure. Accordingly, GLF has significant correlation with velocity pressure related to averaging time, and results in quite large CoV among the 15 economies’ codes and standards. The future incorporation should be conducted on not only gust response factor itself but also the combination of gust response factor and velocity pressure.

Future alignments of wind loading codes or standards in the Asia-Pacific Region are very much necessary and optimistic.

References

[1] ASCE. Minimum Design Loads for Buildings and Other Structures. Reston, VA, ASCE/SEI 7–05, New York, 2006

[2] Standards Australia/Standards New Zealand, Structural Design Actions. Part 2 Wind Actions, AS/NZS1170.2: 2002

[3] National Research Council of Canada. National Building Code of Canada. NRCC, Ottawa, 2005

[4] Architectural Institute of Japan. Recommendations for Loads on Buildings. AIJ-RLB-2004, Tokyo, 2004

[5] Eurocode 1. Actions on Structures, Part 1–4: Wind Actions. London, British Standards Institute, 2004

[6] ISO 4354. Wind Actions on Structures. Switzerland, 2009

[7] Zhou Y, Kijewski T, Kareem A. Along-wind load effects on tall buildings: Comparative study of major international codes and standards. Journal of Structural Engineering, 2002, 128(6): 788–796

[8] Tamura Y, Kareem A, Solari G, Kwok K C S, Holmes J D, Melbourne W H. Aspects of the Dynamic Wind-Induced Response of Structures and Codification. Wind and Structures, 2005, 8(4): 251–268

[9] Holmes J D, Melbourne W H. Design wind speeds in the West Pacific. In: Proceedings of the 4th Asia-Pacific Conference on Wind Engineering. Golden Coast, Australia, July 14–16, 1997

[10] Holmes J D, Weller R. Design Wind Speeds for the Asia-Pacific Region. Standards Australia, Handbook HB 212–2002, Sydney, NSW, Australia, 2002

[11] Gairola A, Mittal A. Part 2: Work Examples. In: Proceedings of the 3rd APEC-WW Workshop. New Delhi, India, November 2–3, 2006

[12] Holmes J D, Tamura Y, Krishna P. Wind loads on low, medium and high-rise buildings by Asia-Pacific codes. In: Proceedings of the 4th International Conference on Advances in Wind and Structures. Jeju, Korea, May 29–31, 2008

[13] Holmes J D. Developments in codification of wind loads in the Asia Pacific. In: Proceedings of the 7th Asia-Pacific Conference on Wind Engineering. Taipei, China, November 8–12, 2009

[14] Holmes J D, Ginger J D. Codification of internal pressures for building design. In: Proceedings of the 7th Asia-Pacific Conference on Wind Engineering. Taipei, China, November 8–12, 2009

[15] Bashor R, Kareem A. Comparative study of major international standards. In: Proceedings of the 7th Asia-Pacific Conference on Wind Engineering. Taipei, China, November 8–12, 2009

[16] Kasperski M. Incorporation of the LRC-method into codified wind load distributions. In: Proceedings of the 7th Asia-Pacific Conference on Wind Engineering. Taipei, China, November 8–12, 2009

[17] Choi E C C. Proposal for unified terrain categories exposures and velocity profiles. In: Proceedings of the 7th Asia-Pacific Conference on Wind Engineering. Taipei, China, November 8–12, 2009

[18] Krishna P. Comparisons of building wind loads and proposal for an A-P model code. In: Proceedings of the 7th Asia-Pacific Conference on Wind Engineering. Taipei, China, November 8–12, 2009

[19] Tamrua Y, Holmes J D, Krishna P, Guo L, Katsumura A. Comparison of wind loads on medium-rise building according to Asia Pacific Codes/Standards. In: Proceedings of the 7th Asia-Pacific Conference on Wind Engineering. Taipei, China, November 8–12, 2009

[20] GB 50009–2001 (revised). Load Code for the Design of Building Structures. Beijing: China Architecture and Building Press, 2006 (in Chinese)

[21] Code of Practice for Wind Effects in Hong Kong. Buildings Department of Hong Kong, 2004

[22] Indian Standard Code of Practice for Design Loads (other than earthquake) for Buildings and Structures, Part 3: Wind Loads, IS: 875 (Part 3), Bureau of Indian Standards, 1987

[23] Standard National Indonesia, SNI 03–1727. Agency for National Standardization (BSN), Jakarta, 2003

[24] KGG-KBCS-05. Korean Government Guidelines of Korean Building Code–Structures, 2005

[25] Code of Practice on Wind Loading for Building Structures. Malaysian Standard, Ms 1553, 2002

[26] National Structural Code of the Philippines. 5th ed. NSCP-2001, Association of Structural Engineers of the Philippines, 2001

[27] Specifications for Building Wind-Resistant Design (Wind Load Provisions of Taiwan Building Code). Taiwan Architecture and Building Research Institute, 2006

[28] Wind Loading Code for Building Design. EIT Standard 1018–46, Engineering Institute of Thailand, 2003

[29] Loads and Actions Norm for Design, TCVN 2737–1995, TieuChuan Viet Nam, 1995

[30] Davenport A G. The wind loading Chain–2004 update. In: Proceedings of the International Workshop on Wind Engineering and Science. New Delhi, India, October 29–30, 2004

[31] Ge Y J, Xiang H, F. Statistical study for mean wind velocity in Shanghai area. Journal of Wind Engineering and Industrial Aerodynamics, 2002, 90(12–15): 1585–1599

[32] Cheung J C K, Holmes J D. APEC-WW Australia 2010 Report: codes/specifications. In: Proceedings of the 6th APEC-WW Workshop. Kwandong, Korea, October 21–23, 2010, 1–8

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg