Random vibration analysis of stochastic time-varying systems

The analysis of the free and forced vibration of a randomly time-varying system is the subject matter of this paper. This is a complicated problem which has received relatively little discussion in the literature. Herein two methods are presented, apart from the digital simulation technique, of finding the response moments. The first one is a series technique which can be considered as a generalization of the well known Galerkin method. The second method belongs to the class of closure techniques. Upon presuming some of the joint distributions to be Gaussian, equations are derived for the first two response moments. It is shown further that the non-Gaussian output density can be approximately predicted by a simple transformation. Detailed numerical results are obtained and compared with computer simulated response statistics. It is demonstrated that the methods developed here are highly efficient. In particular it is found that the Gaussian closure approximation has a wide range of application.