Effect of Material Variability on Multiscale Modeling of Rate-Dependent Composite Materials

Abstract

The effects of material variability on the mechanical response and failure of composites under high strain rate and impact loading are investigated in this paper. A previously developed strain rate–dependent, sectional micromechanics model is extended to account for the variability in microstructure and constituent material properties. The model presented in this paper also includes a three-dimensional damage law based on a work potential theory and a microscale failure criterion. Microstructural characterization of the composite is performed to obtain the statistical distributions needed for the stochastic methodologies. A Latin hypercube sampling technique is used to model the uncertainties in fiber volume fraction and viscoplastic material constants. A comparison of general Monte Carlo simulation and Latin hypercube–based Monte Carlo shows that the Latin hypercube technique converges using fewer simulations. The modulus and failure strain obtained using the developed methodology show good correlation with the experimental data. This novel stochastic sectional model is shown to correlate better with the available experimental data compared with the deterministic sectional model. A laminate level, parametric study is also conducted to investigate the effect of uncertainty on the residual energy of a composite laminate during impact.