Fully Nonstationary Analysis of Linear Structural Systems Subjected to Multicorrelated Stochastic Excitations

The stochastic analysis of structural systems subjected to seismic excitations requires the spectral characterization of both the excitation and the structural response. For large multisupported structures, the spatial variability of ground motion affects significantly the response so that in these cases the assumption of uniform ground motion is inappropriate and only the multicorrelated model of the ground motion acceleration is able to catch the actual effects of seismic waves on the structure. To reproduce the typical characteristics of real earthquakes ground-motion time history, such as the time and frequency varying content, the zero mean Gaussian evolutionary and the sigma-oscillatory stochastic processes have been proposed in literature for multicorrelated input too. In this study, a closed-form solution of the evolutionary power spectral density of the response of linear classically damped structural systems subjected to fully nonstationary multicorrelated excitations for both evolutionary and sigma-oscillatory stochastic processes is given. Finally, an analysis of a four-span continuous deck bridge, and validation with a very efficient Monte Carlo Simulation algorithm, is also performed.