SIMULATION OF MULTIVARIATE NONSTATIONARY RANDOM-PROCESSES - HYBRID DFT AND DIGITAL FILTERING APPROACH

A numerical simulation scheme is presented that combines the advantage s of the discrete Fourier transform algorithm and a digital filtering scheme to generate nonstationary multivariate random processes. The re sulting time histories provide piecewise continuous evolutionary spect ra and the proposed simulation technique offers significant computatio nal efficiency. The effectiveness of the proposed technique is demonst rated with examples. The simulated records are in excellent agreement with the prescribed probabilistic characteristics. The proposed techni que has immediate applications to the simulation of ground motions, ev olutionary sea states, and fast-moving gust fronts.