LAUSR

ABSTRACT The main objective of this study is to investigate the effect of non-uniform thickness reduction and model selection on the predicted ultimate strength of stiffened panels under longitudinal compression. Random fields theory is used to describe the two-dimensional variation of thickness over the plates’ surface. Non-Gaussian sample functions are thus generated using the spectral representation method in combination with translation fields, whereas Monte Carlo simulation is employed to compute the response statistics. The paper provides an insightful guide for the implementation of stochastic finite element method (SFEM) into a deterministic FE software such as Abaqus. The evaluation of stiffened panel's ultimate strength variation due to the described inherent uncertainties of thickness and model selection, namely, the CSR load-end shortening curves analytical relations versus SFEM, is implemented. The results revealed that the uniform thickness approach of CSR using the net-50 scantlings provides a reasonable conservatism on the predicted ultimate strength.