Time-dependent failure criteria for lifetime prediction of polymer matrix composite structures

Abstract

The use of fibre -reinforced polymers in civil construction applications originated structures with a high specific stiffness and strength. Although these structures usually present a high mechanical performance, their strength and stiffness may decay significantly over time. This is mainly due to the viscoelastic nature of the matrix, damage accumulation and propagation within the matrix and fibre breaking. One serious consequence, as a result of static fatigue (creep failure), is a premature failure which is usually catastrophic. However in civil engineering applications, the structural components are supposed to remain in service for 50 years or more in safety conditions.One argument used to replace steel by polymer-matrix composites is its superior corrosion resistance. Yet stress corrosion of glass fibre s takes place as soon as moisture reaches the fibre by absorption. This phenomenon accelerates fibre breaking. In most 2 civil engineering applications, glass-fibre reinforced polymers (GRP) are the most common especially because raw material is less expensive.The lack of full understanding of the fundamental parameters controlling long-term materials performance necessarily leads to over-design and, furthermore, inhibits greater utilization. In this context, lifetime prediction of these structures is an important issue to be solved before wider dissemination of civil engineering applications can take place. As an example, standards dealing with certification of GRP pipes require at least 10000 hours of testing for a high number of specimens. Even though these strong requirements may be foreseen as reasonable, concerning the safety of civil engineering applications, they severely restrict the improvement and innovation of new products.The present chapter reviews some theoretical approaches for long-term criteria. Timedependent failure criteria will be presented and developed for practical applications and illustrated with experimental cases.