Probabilistic progressive damage modeling of hybrid composites

Abstract

A novel analytical probabilistic progressive damage model (PPDM) is introduced for multiphase composites to predict the damage behavior of hybrid composites. The PPDM is based on effective field methods and the stochastic nature of fiber damage is captured by including weakest link theory and Weibull statistics. Three additional models are developed to compare with the PPDM. A stochastic model analogous to the PPDM (called SPDM), and two finite element models, one stochastic (SFEM) and one probabilistic (PFEM). All models are developed in a thermodynamically consistent framework and are extended to include residual thermal stresses. Finally, the four models are compared with models from the open literature for an AS4-M50S hybrid carbon–carbon composite with different hybridization ratios of high to low elongation fibers. The comparison reveals a great agreement between all models and indicates that the stochastic nature of fiber damage is the most influential parameter leading to damage.