Implementation of Genetic Programming Technique for Simulating the Material Properties of Sintered Al-Fe Composite

The variation of sintered aluminium iron composites with respect to parameters of load, density ratio, aspect ratio and percentage of iron are analyzed. The material properties analyzed are axial stress, axial strain, hoop stress, hoop strain, Poisson's ratio and hydrostatic stress. Symbolic regression equations were generated to accurately predict the effect of parameters on the material properties. The limitations of conventional parametric equation fitting were overcome with the use of symbolic multivariate regression fitting. To achieve this objective, the novel approach of Genetic Programming has been used. The data sets have been taken from the Hydraulic Testing Machine using the sintered component. Data for aspect ratios of 0.25, 0.5, 0.75, 1.00; density ratios from 0.8-0.96 and % of iron from 0-8 have been included to find the effect of each parameter on the material properties. The predicted values using Genetic Programming coincided well with the experimental results. The results indicate the efficacy of GP to predict the effect of multiple input parameters on material properties. Further GP can be extended to predict the effect of variation of input parameters on the properties of any composite. The computational model developed will be used to predict the material properties for various iron composition by simple substitution without conducting experiments.