Numerical Simulation of Non-Gaussian Random Fields

The non-Gaussian random fields are used to modelling some dynamic loads generated by wind turbulence, ocean waves, earthquake ground motion etc. These fields also represent the uncertain properties of different materials (reinforced concrete, composite, soils etc.). This paper presents some methods and the corresponding algorithms to the numerical simulation of stationary non-Gaussian random fields characterized by power spectral density or equivalently autocorrelation function and by the marginal probability distributions. The considered methods include the generation of stationary Gaussian random fields based on the spectral representation theorem and their transformation of these fields in stationary non-Gaussian random fields.