Convergence rates for computational predictions of stiffness loss in metal matrix composites

Micromechanics models for metal matrix composites must include both matrix inelasticity and damage evolution in order to accurately predict homogenized macroscopic properties of the composite. Due to these nonlinearities, approximate methods are generally employed to obtain solutions. The current paper discusses an ongoing effort by the authors to construct a computational algorithm which utilizes the finite element method. The model accounts for matrix inelasticity by either J2 plasticity theory or currently available viscoplasticity models. Microcracking, such as fiber-matrix debonding, is incorporated via the use of a nonlinear interface element. The emphasis in the current paper is on determining the accuracy of solutions obtained by the method. It is found that the inclusion of matrix inelasticity increases the computational requirements significantly. Furthermore, the addition of damage evolution creates still further computational requirements.

Convergence rates for computational predictions of stiffness loss in metal matrix composites Conference Paper