Probabilistic local buckling strength analysis of compressive plates for normal and bridge high performance steels

Probabilistic analysis of local buckling strengths (LBS) for compressive plates of normal and bridge high performance steels was conducted. The Monte Carlo simulation method as well as the response surface method was employed to obtain the statistical distribution of LBS. As sources of variability of LBS, the initial out-of-plane displacement and the residual stress are considered as the stochastic variables. The response surface methods are employed to approximate LBS by a simple polynomial function of the initial displacement and the residual stress. For each value of the width- thickness ratio parameter R, a response surface of the normalized LBS is identified based on 114 finite element analysis results with different residual stresses and initial displacement. The response surface models are used in the Monte Carlo simulation to evaluate probabilistic distribution of LBS.