Optimizing thermal-elastic properties of multi-phase and multi-layer composites by using iterative MapReduce guided genetic algorithm

The objective of this work is to optimize the thermal-elastic properties of multi-phase and multi-layer (MPML) composites by controlling the interfaces and matrix layers thicknesses. Aiming at the traditional genetic algorithm (GA) including parallel genetic algorithm (PGA) faces efficiency, scalability, and programming difficulty to solve this kind of optimization problems, we propose an iterative MapReduce guided genetic algorithm (IMGA). This method brings bond into MapReduce, and reasonably allocates variousstages of GAto the map and reduce operator, then completes target optimization through multi-step iteration of map and reduce. The IMGA is interfaced with finite element code to find an optimal design for minimizing the coefficient of thermal expansion (CTE) of the MPML unidirectional fiber reinforced composite with constraints of elastic modulus and fiber volume fraction. Satisfactory results are obtainedby comparing IMGA and GA.