ANALYTICAL MODEL FOR COMPOSITE TRANSVERSE STRENGTH BASED ON COMPUTATIONAL MICROMECHANICS

Abstract

The transverse strength of fiber-reinforced composites is a matrix-dominated property whose accurate prediction is crucial to designing and optimizing efficient, lightweight structures. State-of-the-art analytical models for composite strength predictions do not account for fiber distribution, orientation, and curing-induced residual stress that greatly influence damage initiation and failure propagation at the microscale. This work presents a novel methodology to develop an analytical solution for transverse composite strength based on computational micromechanics that enables the modeling of stress concentration due to representative volume elements (RVE) morphology and residual stress. Finite element simulations are used to model statistical samples of composite microstructures, generate stress-strain curves, and correlate statistical descriptors of the microscale to stress concentration factors to predict transverse strength as a function of fiber volume fraction. Tensile tests of thin plies validated this approach for carbon- and glass-reinforced composites showing promise to obtain a generalized analytical model for transverse composite strength prediction.