Aging of composites - Property evolution and modeling

Monolithic or sandwich organic matrix composites (OMCs) are now a must for engineers, and establishing criteria and models to predict their aging behavior is a major challenge. This requires not only characterizing the evolution of properties, but also understanding the degradation phenomena involved (physico-chemical, physical, mechanical). In the "Aging of composites. Mechanisms and study methodology", the study methodology to be adopted to predict their long-term behavior as well as the phenomena responsible for property degradation were detailed in their entirety.

To define a behavior prediction model, it is necessary to acquire as much information as possible on the material's constituents and structure, to identify as accurately as possible the stresses (thermal, hydric, mechanical, etc.) to which the composite is subjected in operational conditions, and to use inspection methods (optical, non-destructive testing: NDT) to monitor the formation and propagation of damage. It requires the use of realistic accelerated aging tests, for which it is necessary to ensure that they are representative. The greatest difficulty encountered with these tests is defining stress cycles that are representative of the stresses experienced in service.

Using various examples of structural UCOs, this dossier attempts to apply the methodology defined above. It describes the various means of behavioral analysis to be implemented, and indicates to the reader the changes in properties likely to be encountered in service, so that he or she can integrate them into long-term behavioral prediction.

Several "models" dependent on the CMO and its conditions of use, based on the evolution of the property limiting use and resulting either from physico-chemical analysis or from monitoring of mechanical properties are presented. As temperature and humidity are the two major factors influencing the initial and aging characteristics of materials, the prediction of their behavior as a function of these parameters is looked at systematically. From a mechanical point of view, impact resistance and an understanding of crack propagation mechanisms are systematically investigated to certify in-use behavior. Finite element modelling, which takes account of structural, stress and damage singularities, and a multi-scale approach to behaviour prediction are also presented. However, these models do not allow us to dispense with feedback on in-use behavior, which is the only way to validate them.