ABSTRACT
An extensive list of publications addressing the topic of simulations of random processes and fields has appeared in the literature. In 1955, Housner [227] stated that “It is assumed that an accelerogram is formed by the superposition of a large number of elemental [one-cycle sine-wave] acceleration pulses random in time. It is shown that this agrees with recorded accelerograms, and an accelerogram composed in this fashion is shown to have the characteristics of actual recorded accelerograms.” This statement is, essentially, the basis for the generation (or synthesis) of artificial accelerograms. Housner’s concept was utilized by, among others, Bycroft [85], Hudson [230] and Rosenblueth [428], who modeled seismic ground motions by a series of pulses distributed randomly in time. Housner and Jennings [228] further generated Gaussian random time series based on a power spectral density compatible with the average (undamped) velocity spectra of recorded accelerograms. Bogdanoff et al. [64] generated non-stationary acceleration time histories as the sum of cosine functions with random phase angles uniformly distributed between [0, 2π ) modulated with an exponential time function. Amin and Ang [26], [27], Cornell [117], Goto and Toki [177], Jennings et al. [244], and Shinozuka and Sato [467] also generated nonstationary time series.
