Modeling the Fatigue Behavior of Fiber-Reinforced Composite Materials Under Constant Amplitude Loading

Abstract

AbstractThe fatigue behavior of fiber-reinforced composite materials under constant amplitude and variable amplitude loading depends on the type of material and a number of loading parameters. This chapter presents the concepts and theoretical formulations used in order to model the fatigue behavior of fiber-reinforced composite materials, independent of their nature. The commonly used S–N curve types are presented and their applicability to the examined material system dataset is evaluated. Curves estimated by novel computational methods are also presented and compared to traditional ones. The concept of the constant life diagrams (CLDs) used for the quantification of the effect of the mean cyclic stress on the fatigue life of the examined materials is also described. The commonly used CLDs and those most recently introduced are methodically presented. Their performance is evaluated based on their ability to predict S–N curves under “unseen” loading conditions, covering all possible R-ratios of tension-tension, tension-compression and compression-compression fatigue. An alternative approach for the modeling of the fatigue behavior of composite materials under constant amplitude loading is also introduced in this chapter. This approach, according to which stiffness is conceived as the damage metric, whose degradation dominates the fatigue behavior of the material, is based on stiffness degradation measurements during the fatigue life of the examined material. When a certain stiffness limit is reached, the material is considered as having failed. Use of stiffness degradation as a measure for the modeling of the constant amplitude fatigue life of the examined material leads to the derivation of S–N curves corresponding to a predefined level of stiffness degradation and not to ultimate material failure.KeywordsGenetic ProgrammingFatigue BehaviorStress AmplitudeUltimate Tensile StressStiffness DegradationThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.