A Data-Driven Approach to the Prediction of Plasticity in Composites

The work is devoted to the application of the data-driven approach to determining the yield criterion of unidirectionally reinforced composite materials. This condition has the form of a function of the average coordinate stresses. In contrast to the traditional phenomenological approach, where the form of this function is postulated with the inclusion of a limited number of material parameters, the new approach involves the construction of a grid form of the yield criterion. A numerical determination of the yield strengths of the composite based on the known properties of the matrix and fibers leads to a typical problem of analysis of big data. Based on the finite element analysis of the representative cell, a technique has been developed for determining the yield strengths of the composite under various load paths. Illustrative results are given for special cases of combinations of coordinate stresses.