Prediction of composite failure is of critical importance to the design of composite structures. However, because failure of composite structures is typically sudden and catastrophic, understanding the stochastic behavior of failure is critical to accurate prediction of structural reliability. The research reported here focuses on quantifying the length scales of microstructural variability, specifically variation of fiber volume fraction, and their relationship to stress fluctuations in the bulk material. From this study, a stochastic multiscale progressive failure simulation of a compressive test is developed to demonstrate the feasibility of incorporating realistic stochastic information in a multiscale model. Unlike other stochastic studies that randomize strengths or stiffnesses, this approach randomizes an entire microstructure on the basis of volume fraction distributions computed directly from SEM images. Furthermore, the distributions are specific to mesh size. The predicted failure modes are correct and the scatter in the compressive strength predicted by the simulation is similar to scatter in strengths measured experimentally.Nomenclature = sampling cell size f = fiber volume fraction
In this study, a novel multiscale combined creep strain and creep rupture model is proposed. By comparing to experimental data it was shown that the model provides accurate creep rupture predictions for unidirectional off-axis specimens. The creep strain model provides accurate predictions within the confines of the linear elastic restrictions used in its development. This creep strain and creep rupture model was incorporated into a progressive failure finite element simulation so that the effects of load redistribution could be considered, which tended to increase the life of the part. The finite element implementation also allowed for the consideration of realistic geometries resulting in complex stress states. By considering a perfectly flat specimen and one containing worst case thickness variation the experimental open hole creep rupture data was bounded. This suggests that by quantifying material defects realistic lifetime predictions can be made and an estimate of scatter can be acquired.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.