Probabilistic finite element analysis of sheet molding compound composites with an extended strength distribution model

Abstract

Sheet Molding Compound (SMC) composites exhibit large scatters in mechanical properties and a strong correlation between the strength and stress distribution. For example, SMC demonstrates a much higher strength in the 3-pt bending test compared to that demonstrated in the uniaxial tension test. As a result, finite element (FE) simulations with the mean mechanical properties underpredict the flexural response of the SMC structures. Also, probabilistic finite element (PFE) analysis with a unimodal statistical strength model built from the uniaxial tensile data did not solve this problem. To account for the dependency of strength on stress distribution, an extended strength distribution model (ESDM), in the form of a trimodal Weibull distribution consisting of unimodal Weibull distribution models of tension and flexural strengths, has been proposed. The ESDM was subsequently tested in PFE simulations of SMC tensile and flexural experiments following a randomization algorithm which considers the tensile strength as a random variable and assigns it according to ESDM across finite element models. Simulations with this approach successfully reproduced both the tensile and flexural responses with the predicted mean peak load, post-peak behavior, and energy absorption similar to experimental results. • A computational efficient method to include the scatter and size effect in FEA. • A trimodal Weibull strength model based on both tension and flexure strength data. • Simulations reproduced the tensile and flexural responses in experiments.