The durability of fiber composites - The case for mechanism-based models

Long-duration durability (particularly fatigue) testing of composites is ti me consuming and expensive. There is, therefore, a stronger incentive to re duce reliance on such testing than in other areas of design. Mechanism-base d models offer the potential to reduce the reliance on test programs. Such models have been derived for several fatigue damage mechanisms, notably: de lamination, fiber-bridged cracking, and off-axis ply cracking. There has be en less success at modeling fatigue processes at higher levels, such as at notches. Examples of existing models are presented from the literature, and their capabilities and deficiencies are discussed in the context of improv ing the processes of materials selection and design for durability. Key pri nciples are advocated, including modeling damage propagation utilizing Pari s-type expressions, the importance of length scales in modeling fatigue pro cesses, and the need to improve the capability to model multiple, interacti ng, damage processes.