Innovative method for improving composite fatigue life estimations

Approximately 70% of structural failures are due to fatigue. The fatigue mechanism in composites is much more complex than that of metals. The damage may be in one or more forms, such as a failure in a fiber-matrix interface, matrix cracking, delamination(s), and/or fiber breakage. Matrix cracking and delamination both reduce stored energy and stiffness. Damage from these mechanisms can be detected very early in the fatigue life of a composite. This damage produces a reduction of elastic properties such as stiffness. There is always a correlation between damage and stiffness reduction. Fatigue in multidirectional laminates is a well-understood phenomenon. It may be said that the damage process of multi-directional laminates has been clarified, at least qualitatively, if not quantitatively. Even though fatigue behavior is similar in woven composites, it is incorrect to apply the same concepts of the damage mechanism of multi-directional laminates to woven composites. This is because each ply in woven composites is itself bi-directionally reinforced by fiber bundles in the different directions. Recently, the use of woven composites for aerospace and other industrial applications has grown exponentially. These composites are typically manufactured using the low cost vacuum assisted resin transfer molding process (VARTM). With the large variations in fiber volume fraction, weave angle, thickness variations, an accurate estimation of the fatigue life prediction curve becomes a challenging task. It is common knowledge that the ultimate strength of composite materials is dependent on the fiber volume fraction, thickness of the laminate and the warpweft angle. Most of the fatigue analyses today are based upon the average ultimate tensile strengths conducted prior to fatigue testing. Variations in these ultimate strengths of the composites can lead to extensive scatter in the fatigue data. This paper presents an innovative technique for the accurate prediction of the fatigue life in woven and braided composites. The technique uses a multi-variant analysis in conjunction with experimental strength data. This procedure first predicts an accurate tensile/compressive strength of a specimen and then uses these results to develop the fatigue life setup criteria. Results indicate that this technique produces minimal scatter in the fatigue data, accurately predicts the tensile/compressive strengths as a function of the weave angle, the fiber volume fraction and the thickness of the specimen. This paper illustrates the application of this newly developed technique in the estimation of the fatigue life in woven and braided composites.