Life Predictions of Metal Matrix Composite Laminates under Isothermal and Nonisothermal Fatigue

Abstract

A remarkably simple model is proposed in this paper to predict the fatigue life of titanium matrix composite (TMC) laminates under thermomechanical fatigue. It is shown that the maximum applied stress, together with the associated temperature, represents a critical state of loading, of the fatigue process. This critical state of loading dictates a unique state of damage represented by stiffness reduction of the off-axis plies and viscoplastic deformation of the total laminate. As far as the fatigue life prediction is concerned, it is not necessary to know how the damage evolves, and what the stress-strain response is in each loading cycle. After a relatively small number of loading cycles, the stress of 00 fibers saturates to a stable level, and further damage is dominated by the 0° plies of the laminate. The fatigues life is determined by the saturated stiffness, cumulated nonelastic strain and the maximum applied stress. The model is based on a firm understanding of the fundamental damage mechanisms in the TMC laminates under various mechanical and thermal loading. Damage mechanisms and fatigue lives of isothermal, in-phase and out-of-phase fatigue are unified and examined using the concept of critical state of loading. The predictions of fatigue life are shown to be in a good agreement with experimental data of [0/90]2 s, [02/±45] s and [0/±45/90] s made of SCS-6/Titanium for room temperature fatigue, as well as in-phase and out-of-phase thermomechanical fatigue. The most important feature that separates the present research from the existing approaches is its simple and clear representation of the physics involved in the fatigue damage and failure of TMC. The key to the success of the model is the simple fact that failure is controlled by fibers in 0° plies for any laminates containing 00 plies. The data needed to predict the life of any laminates, containing 0° and off-axis plies, under various thermomechanical fatigues are minimal and obtained essentially from a unidirectional laminate. No fatigue life data for the laminates themselves are used to calibrate the model or to determine curve fining parameters. The model makes it possible to establish a data base for different unidirectional laminate and predict the fatigue life of a laminate based on the data of its laminate and the laminate stacking sequence.