Non-Gaussian peak factor model for windward walls of rigid buildings of different aspect ratios

Abstract

Windward walls, which are critical for setting the base shear loads, have been observed to follow non-Gaussian features due to the upstream horseshoe vortex, which nature is more prominent for low-rise buildings. Regarding the non-Gaussian process, a translation framework from a Gaussian to a non-Gaussian process has been widely used to approximate the non-Gaussian peak pressures. However, there is a lack of knowledge regarding the connection between the fluid mechanism (i.e., horseshoe vortex around buildings) and the translation framework. To address this, the paper uses two aerodynamic databases to identify the key parameters affecting the upstream horseshoe vortex, by examining the dependence of statistics of windward wall pressures on the potential parameters influencing the physical characteristics of horseshoe vortex. Aside from the height-to-breadth ratio, the turbulence intensity strongly affects the skewness and kurtosis of area-averaged windward wall pressures. The results suggest that turbulence intensity also affects the structure of the horseshoe vortex and subsequently amplifies the peak pressures, especially for lower height-to-breadth ratios. The statistical trend based on the height-to-breadth ratio and the turbulence intensity are then used with existing translation frameworks to approximate the peak factors of the windward wall pressures. The translation framework using cumulative distribution function mapping with a generalized Pareto distribution shows a good performance. An empirical model with parameters of height-to-breadth ratio and turbulence intensity is then developed for the generalized Pareto distribution parameters and up-crossing rate. This model is able to predict peak factors based on the height-to-breadth ratio and turbulence intensity alone. The results show that the non-Gaussian peak factors may range from 1.4 to 1.7 times its Gaussian counterpart for high turbulence intensity, particularly for low-rise buildings.