Modeling of microstructure for uncertainty assessment of carbon fiber reinforced polymer composites

This presentation deals with the mathematical modeling of composite microstructures for uncertainty quantification of composite structural parameters. Multiple designs of unidirectional fiber reinforced composite materials with arbitrary ply orientations are investigated. We consider a homogenization approach from microscopic to macroscopic scales for the prediction of mechanical properties of the composites. An uncertainty assessment of the effective structural modulus of composite materials consisting of an elastic matrix reinforced with fibers as functions of the phase volume fractions and the structural properties of the constituents is conducted. We consider the global sensitivity analysis (GSA) methods based both on the Fourier Amplitude Sensitivity Test (FAST) and on the Sobol global sensitivity index (GSI). The proposed approach makes it possible to quantify the effective structural parameters of the material based on the variance in the constituents. Numerical results of the GSI and FAST computed for composite materials reveal significant dependence of the macroscopic composite on the probabilistic properties of the fiber volume fraction. The GSA is performed to quantify the influence of fiber volume fraction variation, lamina thickness variation, etc. A nonlinear stage for composite failure prediction based on the Tsai-Wu failure theory was considered. The GSI quantify the relative contribution of variances in material constituents to the total variance of the material under a critical load.