an unrealistic one. In reality, the stress field in a cracked laminate is quite far from being uniform. The multiple transverse cracks shield each other resulting in a very scattered stress field as can be observed experimentally using thermal wave image or photoelastic technique. In essence, this results in a large scatter in composite stress-strain response. Figure 1 – Strain amplitude evolution under constant stress amplitude fatigue loading in unidirectional composite laminates. Bonora and Newaz [6] proposed an approach to stiffness reduction in composites in term of the effective applied strain. They found that using strain as independent variable, the damage curve obtained from several specimens is characterized by a very low scatter confirming the fact that damage D as a function of strain, e, is a reliable material property degradation indicator in fatigue. Here, damage evolution with strain , defined as the complementary normalized stiffness loss for the laminate, can be accurately described as: [ ] ( ) ( ) D D D D cr th CDS th M ( ) ln / ln / e e e e e = + - é ë ê ù û ú 0 0 1 (1) where D0 is the initial damage, Dcr is the value of damage at CDS, eth is the threshold strain for which damage process, i.e. matrix cracking, is activated; eCDS is the strain at CDS and M is a material constant. This equation has been successfully verified for a large number of brittle composite systems both ceramic and polymeric. In addition, it has been found that, if the effective accumulated strain is correlated with cycles, the damage curve resulting from fatigue is the same as for quasi-static loading, as given in figure 2, suggesting the absence of new damage mechanisms due to fatigue. If the key parameter to understand damage accumulation in composite laminates is the effective applied strain it follows that damage evolution observed under constant stress amplitude is not a material feature or property. Let us assume to apply a stress amplitude Ds on a given laminate. If the resulting maximum strain associated with the applied maximum stress overcomes the material strain threshold (0.025 in Figure 2), some damage is produced in the laminate in the form of transverse matrix cracking. Since the damage curve is very steep in the first part (lower strain), small strain changes generates large damage amounts. At the end of the first cycle, the material stiffness is changed as a result of the damage and a residual strain can also be measured. In the following cycle, in order to reach the previous maximum stress with a lower stiffness, a larger maximum strain value will be reached and consequently new damage will be generated. This accumulation process will continue until CDS stage is reached. It follows that crack density accumulation has to increase faster in the early stage of fatigue life and slow down with the Ds De cycles cycles
RkJQdWJsaXNoZXIy MjM0NDE=