Analysis and simulation of large wind pressures due to strong gusts on structures

Abstract

In this study, the proper orthogonal decomposition (POD) and the autoregressive (AR) process is applied to simulate large wind pressures due to strong gusts on a tall structure. The POD is employed to identify the coherent structure characterizing strong wind pressures of the random pressure field. The eigenfunctions and the principal coordinates are computed using the covariance of the ensemble of large wind pressures in the POD analysis. The parametric autoregressive modeling technique is next employed to investigate and model the POD principal coordinates, and the fitted AR models are used to generate the new principal coordinates. The generated principal coordinates are used to simulate large wind pressures. It is shown that time series with large wind pressures are well reconstructed by the dominant modes obtained from the POD analysis, and that wind pressures with large peak values are simulated very well using both the principal coordinates generated by the AR models and the dominant POD modes.