Identification of interfacial parameters in a particle reinforced metal matrix composite Al6061-10%Al2O3 by hybrid method and genetic algorithm

The mechanical behavior of interfaces between matrix and inclusions in composite materials has a strong influence on their mechanical properties such as the strength and the toughness of these materials. To effectively predict the mechanical behavior and investigate the effects of interface properties on the composites, a novel hybrid/inverse numerical method is proposed in conjunction with experimental measurements of real microstructure. This method is based on a combination of hybrid/inverse analysis, finite element method and an improved genetic algorithm (GA). A non-continuum four-node interface element is adopted to simulate the interface behavior of a metal matrix composite whose displacement field has been measured by experiment. By way of the observed failure occurring on the interface in the experiment, a hybrid/inverse analysis for estimating the four unknown mechanical parameters of the interface is carried out by using an improved GA and the interface element model mentioned. Approximate interfacial parameters obtained from the proposed method can reasonably simulate interfacial failure which is in agreement with that observed experimentally. It is found that the proposed hybrid/inverse method is simple and robust for solving complex interfacial problems in composites.