Simulation of Non-Gaussian Fluctuating Wind Pressure through Spectral Representation Method

Abstract

In the simulation of wind loads, it is mostly hypothesized that wind loads are stationary Gaussian stochastic processes. But, inasmuch as the wind pressure fluctuations measured at corners as well as other separation flow regions of roofs of buildings are found to be highly non-Gaussian, the Gaussian assumption is not appropriate for the generation of the wind pressure time history in these local regions. In order to simulate the non-Gaussian fluctuating wind pressure, a new framework is proposed in the present paper. In addition, the power spectral density function of fluctuating wind pressure is deduced rigorously by resorting to the relationship between the wind pressure and velocity in this paper. In order to corroborate the feasibility and correctness of the present methodology, a numerical example of simulating one-dimensional univariate non-Gaussian fluctuating wind pressure is taken into consideration.