A Micromechanical Model for Predicting Fatigue Response of Metal Matrix Composites Subjected to Environmental Degradation

Abstract

It has been observed in titanium matrix continuous fiber composites that structural life at elevated temperatures is significantly shorter in air than in inert gas. The physical reasons for this degradation in life are dependent on the titanium alloy considered. In this paper the metastable Ti-ß 21S matrix alloy is considered, and is embedded with SCS-6 fibers to produce a four ply unidirectional laminate.

Experimental results are briefly reviewed for laminates subjected to cyclic loading to failure at 482°C. It is demonstrated that life is reduced by a factor of about five when a specimen is first degraded by placing it in an oven for 24 hours at 700°C, and that this reduction is not due to oxidation. Rather, it is due to the development of a graded structure of brittle alpha grains that form due to oxygen diffusion into the matrix along the boundaries of the beta grains. These alpha grains induce microcracks near the surface of the composite that propagate into the interior of the composite and ultimately lead to premature failure of the structure.

A computational model is employed herein based on the above experimental observations. This model utilizes the finite element method to predict the thermomechanical response of a unit cell of the laminate subjected to cyclic loading. The algorithm accounts for phase heterogeneity, matrix thermoviscoplasticity, and damage evolution due to surface cracking and fiber-matrix debonding. Crack growth is modelled via the inclusion of cohesive zones wherever experimental observations indicate that cracks may grow. Environmental degradation is accounted for by degrading the properties of the cohesive zone near the composite free surface, in accordance with experimental determinations of fracture toughness of the degraded material.

Predictions obtained with the model indicate several useful observations. First, the effect of matrix viscoplasticity is important and should not be neglected in modelling. Second, there are significant differences between model predictions for the degraded and as-received material. Thus, with more advanced computers it may be possible to utilize the approach described herein to predict the effect of environmental degradation on structural life in continuous fiber metal matrix composites.