Abstract
Response characteristics of a single-degree-of-freedom linear system subjected to non-Gaussian random excitation are investigated. The excitation is assumed to be a stationary stochastic process characterized by the non-Gaussian probability density and the power spectrum with the bandwidth and dominant frequency parameters. As the probability density of the excitation, bimodal distribution and Laplace distribution are used. The excitation is generated by solving an Itô stochastic differential equation numerically. Monte Carlo simulation is performed to obtain the stationary response distribution of the system. The response characteristics are discussed by using the maximum value of the absolute value of the cross-correlation function between the excitation and the response. It is observed that the response distributions become similar to each other when the maximum values are almost the same value. The shape of the stationary response distribution becomes the shape of the distribution of the excitation when the maximum value is close to one. When the maximum value is around 0.85, the stationary response distribution becomes the middle shape between the distribution of the excitation and the Gaussian distribution. When the maximum value is less than 0.6, the stationary response distribution becomes almost Gaussian distribution. The shapes of the stationary response distributions can be predicted roughly from the maximum value of the absolute value of the cross-correlation function without numerical simulation, because the maximum value of the cross-correlation can be calculated from the frequency response function of the system and the power spectrum of the excitation regardless of the distribution of the excitation.
