Optimization of al-base composite using genetic algorithm

This paper provides six tests were carried out for the base alloy (BA) (Al2%Mg) and the three composite samples ((AR1R(Al-2%Mg-2 %CKD), AR2R (Al-2 %Mg-8%CKD) & AR3R (Al- 2 %Mg-16 % CKD))) which were prepared by using powder metallurgy technique. As a results, it was found an optimum composite material using the hybrid method represented by genetic algorithms by using through carry out two ways of crossover (1X, 2X), basing on statistical data obtained from experimental results. The basic data were built, depending on their properties, to describe the composite.Then, the evolution algorithm is to make procedure for the genetic clustering process and provides a number of required clusters; to avoid the overlapping between clusters with the other. One of the clustering validity measures called "Davies-Bouldin index" as fitness function of that algorithm that used. Then, the two types of properties for each cluster: mechanical properties (hardness, thermal conductivity, wear rate, friction coefficient) and machining properties (surface roughness, tool life) were extracted. This paper concludes that composite (43&33) represented optimum composite material by using one point and two point crossover operators (1X,2X) respectively.