Abstract

Conventional building design wind pressure p can be expressed as a product of the site's reference wind speed pressure q , exposure factor Ce , building pressure coefficient Cp and aerodynamic gust factor Cg . Although the evaluation of Cp and Cg values has progressed significantly in wind engineering, the Ce specifications, especially those for inhomogeneous terrain, were usually given by meteorologists who are mainly interested in predicting higher boundary-layer level winds that sense terrain roughness variations in a scale that may be so large that is not appropriate for building dimensions. Therefore, it is not surprising to see that a number of issues/discrepancies exist in the terrain-related provisions in the wind standards and codes of practice. Most recently, wind engineering studies found that terrain roughness patches of different distance to the site have different strengths of influence, and the building wind loads are very sensitive to small-scale roughness variations as well. Within this context, a systematic study on the development of the terrain exposure factor on building wind loads has been carried out. Variations of wind speed and turbulence intensity ( Iu ) profiles, as well as low-rise building loads, above fetch with roughness changes have been investigated by using systematic wind tunnel measurements and numerical simulations (of a set of classic flow-motion equations with a simple eddy-viscosity closure), with particular attention to the small-scale roughness changes close to the site. The numerical simulation is developed in order to ensure that the present wind-tunnel findings can be better than the standard numerical results that have been adopted into a number of wind standards and codes of practice. The velocity data are formulated into a new speed model and a new Iu model, and the low-rise building pressure measurements are analyzed in order to clarify a number of pertinent code issues/discrepancies. The proposed models and design classification have been compared with previous findings, including a limited amount of full-scale data, with satisfactory results. Thereafter, a number of conclusions are made as follows: (1) The proposed speed model should be better than, or at least alternative to, the ESDU 82026 model, which has been adopted by the current British, Australian and American wind standards. The shortcoming of the ESDU 82026 model is due to the oversimplified homogeneous terrain assumptions that prevent this model from properly applying for fetch with multiple small-scale roughness changes. (2) The proposed Iu model is more practical and predictive than the ESDU 84030 standard data sets that belong to the few data provisions for predicting Iu profile above inhomogeneous terrain. (3) For low-rise buildings, it is found that the Suburban exposure factor should be increased from the ASCE 7-02 specification up close to the ASCE 7-95 level (the former is approximately 25% lower than the latter). In addition, the present results can be used to clarify the discrepancy among different national/international wind standards and codes of practice on the minimum fetch length requirement for a patch to qualify as homogeneous terrain. The present research findings will be suggested for updating the terrain-related provisions of the current wind standards and codes of practice