Towards designing composites with stochastic composition: Effect of fluctuations in local material properties

This article presents a numerical study of the mechanical behavior of particulate composites with stochastic composition. Two types of such materials are considered: composites with homogeneous elastic-plastic matrix and randomly distributed inclusions of stiffness sampled from a distribution function, and composites with matrix having spatially varying elastic-plastic material parameters with no inclusions as well as with randomly distributed identical inclusions. We observe that the presence of fluctuations in either inclusions or matrix material properties leads to smaller effective modulus, smaller strain hardening and a reduction of the yield stress of the composite. Fluctuations of the yield stress of the matrix leads to a significant reduction of the mean yield stress of the composite. Fluctuations of the elastic modulus and of the strain hardening are associated with the reduction of the mean of the distributions of elastic modulus and strain hardening of the composite. For the range of parameters considered, fluctuations lead to maximum principal stress fields with narrow distribution of values, which implies enhanced resistance to damage initiation. Increasing the variance of the distribution functions from which local material properties are sampled, while keeping the mean constant, renders these effects more pronounced. This study is motivated by the growing interest in additive manufacturing technologies which open new possibilities for designing composite materials.