The effects of wind forces are of primary importance in the design of tall buildings. Since dynamic wind forces have spatial amd temporal randomness and correlation, it is necessary to study the wind-induced response from statistical view point. This paper presents an efficient and practical method of simulating a set of stationary random processes with specified auto- and cross-spectral densities. The component processes are generated by the finite Fourier series with correlated random phase angles relevant to the coherence characteristics of wind forces. Also numerical examples of wind response analysis to the simmulated random processes are demonstrated.
