2014 年 19 巻 p. 187-196
In this study the stochastic properties of the response of a square steel plate under in-plane compressive load subject to random initial distortions have been characterized using Polynomial Chaos Expansions (PCE). At first the random characteristics of maximum initial distortions of a steel plate have been approximated using PCE. Subsequently an analytical procedure has been developed based on elastic large deflection analysis which can be used to predict the stochastic characteristics of the plate response. A significant feature of the proposed method is the avoidance of the generally time-consuming Monte Carlo Simulation for evaluation of response statistics. Moreover, this methodology requires knowledge about the governing equations of the physical system. It is also theoretically reasonable as it has a strong mathematical basis. It has been found that the proposed formulation can predict the statistical characteristics of the plate response with sufficient accuracy, when compared to those obtained from Monte Carlo Simulation.