comparative study of major international standards - IAWE

COMPARATIVE STUDY OF MAJOR INTERNATIONAL STANDARDS Rachel Bashor1 and Ahsan Kareem2 Ph.D. Candidate, NatHaz Laboratory, University of Notre Dame, Notre Dame, IN, USA, [email protected] 2 Robert M. Moran Professor, NatHaz Laboratory, University of Notre Dame, Notre Dame, IN, USA, [email protected] 1

ABSTRACT Globalization of the construction industry and the development of unified international codes and standards intensifies the need to better understand the underlying differences between the major international wind loading standards. A comprehensive comparison of the wind loads and their effects on tall buildings is conducted utilizing six major international codes and standards: ASCE 2005 (USA), AS/NZ 2002 (Australian and New Zealand), NBCC 2005 (Canadian), AIJ 2004 (Japanese), Eurocode 2004 (EU), and ISO 2009. The key areas of comparison include the provisions for strength design as well as the serviceability requirements in the alongwind, acrosswind, and torsional directions. As the standards utilize a common theoretical framework, the equations are re-written in a general format in order to compare the individual parameters. KEYWORDS: WIND CODES AND STANDARDS, TALL BUILDINGS, WIND LOADING

Introduction Globalization of the construction industry and the development of unified international codes and standards, i.e. ISO [2009], intensifies the need to better understand the underlying differences between the major international wind loading standards. Previous studies have found that the varying definitions of wind field characteristics, including mean wind velocity profile, turbulence intensity profile, wind spectrum, turbulence length scale, and wind correlation structure, were the primary contributors to the scatter in predicted response quantities [Zhou et al. 2002; Tamura et al. 2005]. As nearly every major building code has been updated in the last few years, it is necessary to update previous code comparison work. A comprehensive comparison of the wind loads and their effects on tall buildings is conducted utilizing six major international codes and standards. These codes/standards are: the American Society of Civil Engineers’ Minimum Design Loads for Buildings and Other Structures [ASCE 2005], the Australian and New Zealand Standard [SAA 2002], the National Building Code of Canada [NRC 2005], the Architectural Institute of Japan Recommendations [AIJ 2004], the European Standard [Eurocode 2004], and the International Organization for Standardization 4354 [ISO 2009]. All utilize the traditional displacement-based gust loading factor for assessing the dynamic along-wind loads and their effects on tall structures but incorporate different provisions for the acrosswind and torsional loads [Tamura et al. 2005]. The key areas of comparison include the provisions for strength design in the alongwind, acrosswind, and torsional directions as well as the serviceability requirements. As the standards utilize the same basic theory, the equations are re-written in a general format in order to compare the individual parameters. These parameters are investigated and several examples are presented. Finally, the deviations in the predictions are discussed and suggestions are made to improve agreement between the standards. Wind Characteristics in Codes and Standards Although these standards determine wind loading in the along-wind direction using a random-vibration-based gust factor approach, the parameters are defined differently. These parameters are re-written in a consistent format and compared with each other. Some of the difficulties in using international standards is the use of different terminology and the

The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan

incorporation of factors within other terms, making it hard for designers to work in a global environment. Rewriting the basic equations in a general format will help designers decipher the nuances of the different codes/standards and understand the resulting differences in the response. Note that the scope of this analysis is limited to dynamically sensitive buildings of regular shape. All the standards recommend that extremely tall and irregular shaped structures be designed using wind tunnels. Alongwind Wind Loads In all six standards, the alongwind loads are determined by multiplying the wind pressure by the tributary area of the building. The general expression for pressures on a building for all the standards can be expressed as: p = qGC p (1) where q = velocity pressure; G = gust factor; and C p = pressure coefficient. The following investigates both internal pressures and external pressures, acting in the windward and leeward directions. The loads are then determined by combining the pressures acting on a wall and the corresponding tributary area. Moments are determined by multiplying the load at a given height by the corresponding height. Base shear forces and moments are then determined by the sum of the loads and moments at each level. The velocity pressure can be expressed as: q = 12 ρV02 ⋅ C exp osure ⋅ Cterrain ⋅ C direction ⋅ C importance ⋅ C other (2) where ρ = air density; V0 = basic wind velocity; Cexp osure = velocity profile or exposure factor; Cterrain = terrain and topography factor; C direction = directionality factor; Cimportance = building importance factor; and C other = a factor accounting for other things such as hurricane zone, shielding, or mean recurrence interval. The effects of terrain, directionality, building importance, and other factors are not considered in this study. However, the definitions of velocity profile are analyzed in detail and compared between the standards. Averaging times for wind velocity vary between the standards and within the standards. For example, in Eurocode, the velocity is adjusted from 10 minute to one hour for calculations of response. In addition, the reference height at which the gust factor and other parameters are calculated is different between the codes/standards, as summarized in Table 1. These differences between averaging time and reference heights affect the intermediary parameters and resulting responses, making a simple comparison between the standards challenging. Throughout this analysis, the effect of differing averaging times has been minimized as much as possible. Table 1: Averaging Times and Reference Heights ASCE AS/NZ NBCC AIJ

Eurocode ISO Averaging time for basic 3-s 3-s 3-s 1-hr 10-min 10-min wind velocity 10-min Averaging time for design 1-hr 3-s 1-hr 10-min 1-hr 10-min velocity at reference height Reference height for gust 0.6h h h h 0.6h h factor NOTE: ISO provides two procedures for determining loads: one for peak response and one for mean response.

The wind velocity in each code is described by a profile law, either power or logarithmic. Eurocode, AS/NZ and ISO both use the logarithmic law, whereas the others use a power law. ISO also provides power law fits to the profiles. The velocity profiles are dependent on the exposure category. Each standard uses three to five exposures categories, and can be described by six general exposure categories. For the purposes of this paper,


exposure category (EC) 4 is defined as open land, category 3 is defined as suburban, and category 2 is defined as urban. To compare the profiles, it is important to account for the differences in averaging time. In the following, the velocity profiles are compared based on their respective averaging times and exposure categories. In Figure 1, the profiles are compared in terms of the general exposure categories in either mean-hourly or 10-minute averaging time. It is noted that the standards have similar EC3 and EC4 profiles while the “urban” profiles differ considerably. The divergence in the profiles at upper heights can be attributed to the effects of using a power law versus the logarithmic profile [e.g., Simiu and Scanlan 1978]. Mean (1-hr or 10-min) Velocity Profiles EC3 - Suburban

Height (m)

EC2 - Urban

EC4 - Open

400

400

400

350

350

350

300

300

300

250

250

250

200

200

200

150

150

150

100

100

100

50

50

50

0

0 0

0.5

1

1.5

2

0

Velocity Profile AIJ

0.5

1

1.5

2

Velocity Profile Eurocode

NBCC

0 0.5

1

1.5

2

Velocity Profile ASCE

AS/NZ

ISO

Figure 1: Mean Velocity Profiles of All Codes/Standards for Exposure Categories 2, 3, and 4 The gust factor for the six standards may be written in terms of a general format as: GLF G= (3) Gq where GLF is the gust loading factor originally defined by [Davenport 1967] which is expressed as: GLF = 1 + r g 2B B + g R2 R

(4)

In the above equations, gB,gR are the peak factors for response, r describes the turbulence intensity, Gq is the gust factor for the wind velocity pressure, B is the background factor and R is the resonant factor. A summary of the peak and gust factors for the standards is provided in Table 2. The parameter r in Table 2 describes the turbulence intensity of the wind in terms of the turbulence intensity profile and a multiplier. The general form of r is: r = 2I h (5) where Ih is the turbulence intensity profile. The AS/NZ, Eurocode, ISO and NBCC standards utilize the turbulence intensity as defined in Equation (5). ASCE uses a factor of 1.7 instead of 2, and AIJ uses a factor ranging from 2.13 to 2.418 based on the exposure category. As shown in Figure 2, there is significant variation in the turbulence intensity profiles, especially for the Urban exposure category. This leads to variations in the resulting gust factor [Zhou et al. 2002].


Table 2: Comparison of Peak and Gust Factors gR

g = 2 ln (νT ) +

ASCE

0.577 2 ln(νT )

gB

gv

Gv

T

ν

3.4

3.4

1+ gvr

3600

f0

AS/NZ

g = 2 ln (νT )

3.7

3.7

1+ gvr

600

f0

AIJ

g = 2 ln (νT ) + 1.2

gR

–

1

600

f0C

gR

3.5

1+ gvr

600

f0C

gR

–

1

3600

f0C

0 .6 2 ln (νT ) 0 . 577 2 ln (ν T ) + 2 ln (ν T )

g = 2 ln (νT ) +

Eurocode

g=

NBCC

0.5772 1 + g v r (peak) 600 3.4 3.4 f0 2 ln(νT ) 1 (mean) NOTE: gR is peak factor for resonance, gB is peak factor for background, gv is peak factor for wind velocity, f0 is the natural frequency of the building and C refers to a factor which is a function of the background and resonant responses. g = 2 ln(νT ) +

ISO

EC3 - Suburban

Height (m)

EC2 - Urban

EC4 - Open

500

500

500

400

400

400

300

300

300

200

200

200

100

100

100

0 0

0.5

1

Turbulence Intensity Profile AIJ

0 0

0.2

0.4

0.6

0.8

Turbulence Intensity Profile Eurocode

AS/NZ

ASCE

0 0

0.2

0.4

0.6

0.8

Turbulence Intensity Profile NBCC

ISO

Figure 2: Turbulence Intensity Profiles of for Exposure Categories 2, 3, and 4 As discussed in [Tamura et al. 2005], the energy factor of the standards takes one of three forms: von Karman, Kaimal, or Davenport. Specifically, ASCE and Eurocode use a form of the Kaimal normalized wind velocity spectrum, AIJ and AS/NZ both use the von Karman definition, and NBCC uses the Davenport definition. The choice of spectra definition impacts the value of the energy factor and the resulting resonance response factor. The final parameter in Equation (1) is the pressure coefficient. For wind design, there are both external and internal pressures. While the external pressure coefficients are reasonably consistent at building height, the internal pressure coefficients vary considerably between the codes/standards. Acrosswind and Torsional Loads Although the standards are fairly consistent with respect to alongwind loading, the treatment of the acrosswind and torsional loading differs amongst the codes/standards. For example, ASCE, Eurocode, and NBCC utilize partial loading to account for acrosswind and


torsional loads, although ASCE provides an alternative method in the Commentary (Aerodata). The partial loading technique simply applies fractions of the alongwind pressures in different combinations. Torsion is introduced either by asymmetric loading, as in NBCC and Eurocode, or by an applied moment defined as a combination of the alongwind load multiplied by a defined eccentricity (ASCE). Aerodata, AS/NZ, AIJ, and ISO provide procedures for determining the acrosswind and torsional loads as a function of the base bending moment of the resonant response. As these procedures typically rely on databases, the results vary to a higher degree than the alongwind comparisons. Accelerations In addition to strength requirements, the serviceability requirements, in terms of acceleration, are assessed. All codes and standards provide equations for defining alongwind accelerations, however acrosswind and torsional accelerations are not included in every code. The alongwind acceleration can be generally defined as: q G C bhK &xˆ&( z ) = h R fx (6) φ1 ( z ) m1 where qh is the velocity pressure at the reference height, GR is the resonant component of the gust factor, Cfx is the force coefficient, b is the building width, h is the building height, K is the mode shape correction factor, m1 is the generalized mass in the first mode, and φ1 ( z ) = ( hz )k is the first mode shape evaluated at height z. Note that both NBCC and ISO define the alongwind acceleration in terms of the displacement.

Example Analyses of Tall Buildings To compare the wind loading standards, several example buildings are analyzed with each code. The first example building is a square building with height of 200 m, width and depth of 33 m, natural frequency for alongwind and acrosswind of 0.2 Hz, damping of 1% in all directions, linear mode shapes, building density of 180 kg/m3, air density of 1.22 kg/m3, basic wind velocity for strength of 40 m/s (3 second) and 35 m/s (3 second) for serviceability. To convert the velocity to different averaging times, the relationship developed by Durst [ASCE 2005] is utilized. The building is analyzed using first an urban exposure then an open exposure. Factors accounting for wind direction, importance, etc. are assumed to be 1. While the parameters are re-written in terms of a general form so as to accurately compare the various parameters, the analysis performed calculates the loading in the same manner stated by the codes/standards. The comparison of the individual parameters reveals several areas of disagreement between the codes/standards, leading to differences in the wind loads. Results from the analysis of Example 1 are presented in Table 3. This example highlights important similarities and differences between the standards. For instance, ASCE and AS/NZ both determine the response using peak velocity pressure thus yield very similar results despite the variations in the intermediate parameters. While the GLF for Eurocode B and C agrees with AS/NZ and ASCE, respectively, the large velocity pressure contributes to the increase in the overall loads for Eurocode. Note that although Eurocode defines the velocity as 10-min, the velocity pressure is based on gust response. The velocity pressure for ISO using the peak method is smaller than that for Eurocode, thus the larger GLF contributes to larger overall loads. For the standards based on mean response (AIJ, NBCC, and Mean ISO), the ISO velocity pressure is the smallest leading to the smallest overall load while AIJ has the largest velocity pressure and overall load. The coefficient of variation (CoV) for the resulting loads is 0.10. The acceleration predictions follow similar trends as compared to the strength results, with Eurocode providing the largest accelerations.


Table 3: Alongwind Results for Example 1 using Urban Exposure ASCE

Eurocode

Aerodata

AS/NZ

ISO AIJ

B

NBCC

C

Peak

Mean

V0 (m/s)

40

40

40

28.1

28.1

26.4

40.0

28.1

href (m)

120

200

200

120

200

200

200

200

27.5

32.6

46.4

31.5

36.4

33.1

50.8

31.6

Vhref

(m/s) 2

q h (kN/m )

1.41

–

1.31

0.81

0.76

1.57

0.61

B R GLF Base Shear (MN) Base Moment (MN-m) σ &x& (milli-g)

0.583 0.526 2.69 9.95

– – – 8.10

0.633 0.601 2.49 9.65

0.512 0.735 2.50 10.81

1.73 0.409 1.015 2.61 11.44

0.491 0.810 2.18 11.82

0.422 1.67 2.84 11.23

0.663 0.853 3.085 10.59

0.701 0.853 3.112 9.87

1,084

1,112

1,049

1,257

1,328

1,303

1,271

1,268

1,279

3.44

3.98

3.3

5.37

6.39

3.96

–

–

–

13.03 15.06 10.44 17.21 20.49 12.72 – – &xˆ& (milli-g) NOTE: Aerodata refers to Commentary section of ASCE. Eurocode and ISO have two procedures. V0 , Vh are basic and design wind velocity; href is reference height; q h is velocity pressure; B, R are

–

ref

background and resonance factor, GLF is gust loading factor; and

σ &x& , &xˆ& are rms and peak acceleration.

As shown in Table 4, the results for Example 1 using the open exposure category EC4 follow similar trends to the results using exposure category EC2, with a CoV of 0.2. However, ISO using the peak methods agrees very well with ASCE. Also, the NBCC loads are considerably higher than the other loads. As indicated in Figure 1, the large variation in the velocity pressure, which is a function of the velocity profile, contributes to the scatter in the resulting loads. To reduce the scatter and dependence on velocity pressure, Example 2 analyzes the same building with a modified velocity pressure distribution, external and internal pressure coefficients for all the codes/standards. In order to obtain a velocity pressure consistent between the codes/standards, the basic velocity is adjusted so that the velocity pressure at building height is the same. To account for varying averaging times, the gust factor for wind velocity pressure is determined in accordance with [Zhou et al. 2002] and [Solari 1993] and defined as: V  G τq (T ) = 2 τ  − 1 (7)  VT  where τ ,T are averaging times of interest. Thus, Gq3s (10min ) = 1.92 and Gq3s (1hr ) = 2.06 . As shown in Table 5, the modifications made to the velocity pressure and pressure coefficients brought the wind load amongst the codes/standards closer together. The CoV was reduced to 0.07 and the mean velocity pressure based codes/standards are now more in line with the peak based codes/standards. The discrepancies can be contributed to turbulence intensity variances and the differences in determining the GLF. For the third example presented in Table 6, a rectangular building is analyzed. For this example the following properties are assumed: h = 180 m, b = 30 m, d = 60 m, natural frequencies in both sway directions = 0.25 Hz, building density = 180 kg/m3, damping of 2%, and basic wind velocity of 60 m/s (3 second) for strength and 45 m/s (3 second) for serviceability. The building is assumed to be in EC3. Based on the discussions above, velocity pressure and pressure coefficients are modified to be consistent. The results are similar in


trend to those noted previously, with a CoV of 0.07. Note that NBCC is markedly smaller than the other results. This is likely attributed to the lower value of turbulence intensity for NBCC as compared to the others. When the turbulence intensity for NBCC is modified to correspond to the other turbulence intensities, the resulting wind loads are more consistent. Table 4: Alongwind Results for Example 1 using Open Exposure ASCE 40

V0 (m/s) href (m) Vhref

(m/s)

Aerodata

AS/NZ

40

40

Eurocode B

AIJ

28.1

28.1

ISO

NBCC

C 26.4

Peak 40

Mean 28.1

120

200

200

120

200

200

200

200

38.1

41.2

52.6

41.6

44.0

40.2

54.4

43.9

2 q h (kN/m )

1.84

–

1.62

1.18

0.98

1.81

1.177

B R GLF Base Shear (MN) Base Moment (MN-m) σ &x& (milli-g)

0.624 0.889 1.854 14.88

– – – 13.78

0.633 1.107 1.939 14.07

0.524 1.177 2.083 16.82

0.428 1.536 2.171 17.67

0.528 1.274 2.062 17.94

0.422 2.466 2.347 20.13

0.625 1.513 2.028 14.86

0.636 1.513 2.031 14.78

1,572

1,705

1,468

1,862

1,955

1,918

2,136

1,683

1,846

3.85

6.63

4.51

7.28

8.46

6.06

–

–

–

14.57

25.09

13.95

23.53

27.34

19.57

–

–

–

&xˆ& (milli-g)

2.21

Table 5: Alongwind Results for Example 2 using Suburban Exposure ASCE

AS/NZ

V0 (m/s)

40

42

href (m)

120

Eurocode

NBCC

25.8

26.0

25.7

Peak 39.1

Mean 27.1

200

120

200

200

200

200

33.5

52.1

33.3

37.6

36.2

52.0

37.6

1.65

1.65

1.65

0.86

0.80

1.65

0.86

B

Vhref

(m/s)

2 q h (kN/m )

ISO

AIJ

C

B 0.607 0.633 0.517 0.417 0.513 0.422 0.637 0.653 R 0.729 0.967 0.801 1.095 0.907 2.010 1.147 1.147 GLF 2.228 2.177 2.231 2.329 2.055 2.491 2.385 2.392 Base Shear (MN) 12.79 13.66 11.81 12.33 12.54 11.76 12.89 12.08 Base Moment (MN-m) 1,370 1,450 1,312 1,370 1,360 1,317 1,457 1,512 NOTE: Aerodata database only has Urban and Open Exposures and is therefore removed from this example. Table 6: Alongwind Results for Example 3 using Suburban Exposure ASCE

AS/NZ

V0 (m/s)

60

62.6

href (m)

108

Eurocode

NBCC

38.5

39.2

38

Peak 58.7

Mean 40.8

180

108s

180

180

180

180

48.9

76.9

48.8

55.5

53.5

76.9

55.4

3.61

3.61

3.60

1.88

1.75

3.61

1.87

B

Vhref

(m/s)

2 q h (kN/m )

ISO

AIJ

C

B 0.617 0.651 0.523 0.427 0.519 0.459 0.645 0.660 R 0.530 0.737 0.633 0.820 0.738 1.590 0.931 0.931 GLF 2.153 2.158 2.193 2.249 2.037 2.389 2.370 2.377 Base Shear (MN) 18.03 19.66 17.29 17.73 18.05 16.14 19.19 18.22 Base Moment (MN-m) 1,759 1,900 1,731 1,775 1,789 1,656 1,942 2,044 NOTE: Aerodata database only has Urban and Open Exposures and is therefore removed from this example.


Concluding Remarks This paper examines the differences and similarities in major international wind codes/standards. Although many parameters were examined, the scope is limited to dynamically sensitive, regular-shaped buildings with flat roofs that are classified as enclosed. To accurately compare the parameters, the various equations in the codes/standards are written in a general format. While significant discrepancies are apparent in the comparison of the intermediary parameters, the overall loads are reasonably consistent. However, with a few modifications to specific parameters, the discrepancies between the standards are further reduced. The parameters contributing to the most differences in the resulting wind load are those associated with the wind velocity characteristics. Ultimately, the standardization of wind loading standards is achievable with an understanding of the similarities and differences between the major codes/standards. Acknowledgements The support for this paper was in part provided by the grant # CMMI 0601143 from the National Science Foundation and the Global Center of Excellence, Tokyo Polytechnic University funded by the Ministry of Education, Culture, Sports, Science and Technology (MEXT). References AIJ (2004), RLB Recommendations for Loads on Buildings, Tokyo, Structural Standards Committee. ASCE (2005), Minimum Design Loads for Buildings and Other Structures, Reston, VA, American Society of Civil Engineers. Davenport, A. G. (1967), "Gust Loading Factors," Journal of the Structural Division 93: 11-34. Eurocode (2004), Part 1-4: Wind Actions, Eurocode 1: Actions on Structures, London, British Standards Institute. ISO (2009), 4354: Wind actions on structures, Switzerland, ISO. NRC (2005), National Building Code of Canada, Ottawa, Associate Committee on the National Building Code, National Research Council. SAA (2002), AS1170.2 Part 2: Wind Forces, Australian Standards AS1170.2:2002, Sydney, Standards Association of Australia. Simiu, E. and R. Scanlan (1978), Wind Effects on Structures: An Introduction to Wind Engineering, New York, John Wiley & Sons. Solari, G. (1993), "Gust buffeting I: Peak wind velocity and equivalent pressure," Journal of Structural Engineering 119(2): 365-382. Tamura, Y., A. Kareem, G. Solari, K. C. S. Kwok, J. D. Holmes and W. H. Melbourne (2005), "Aspects of the Dynamic Wind-Induced Response of Structures and Codification," Wind and Structures 8(4): 251-268. Zhou, Y., T. Kijewski and A. Kareem (2002), "Along-Wind Load Effects on Tall Buildings: Comparative Study of Major International Codes and Standards," Journal of Structural Engineering 128(6): 788-796.