Multi-Scale Modeling of Curing Residual Stresses in Composite with Random Fiber Distribution into Consideration

During curing process of composites, residual stresses inevitably develop and play an important role in the final mechanical properties of composites. Therefore, the consideration of residual stresses when designing composite structure is necessary. The causes of residual stresses are well known on the part and fiber-matrix level. However, the influence of the part level factors on micro residual stresses is less known and rarely investigated. This work aims at a better understanding of the effects of macro-level factors, including temperature variation and mechanical strains on micro-scale stresses. To this end, a multi-scale method is developed, which consists of a macro-scale model to capture temperature variation and mechanical strains field under a given cycle condition, and a RVE (Representative Volume Element) to predict residual stresses in matrix. The results demonstrate that the maximum micro residual stress in matrix presents about almost 52% reduction in the model with part-level information into consideration as compared with the results without considering the multi-scale effect. It can be also seen that with multi-scale effect into account, the matrix in the RVE experienced tensile residual stresses in the matrix-poor locations and compressive residual stresses in the matrix-poor locations.