Experimental investigations on silicon carbide mixed electric discharge machining

Surendra Singh Thakur,Sharad K. Pradhan,Shankar Sehgal,Kuldeep K. Saxena
DOI: https://doi.org/10.1007/s12633-022-02022-w
IF: 3.4
2022-08-01
Silicon
Abstract:In this study, silicon carbide mixed electrical discharge machining (SCMEDM) process has been developed and later on modelled also using an artificial neural network (ANN) based technique as well as response surface methodology (RSM). Experiments were conducted with Al LM-25/SiC metal matrix composites as per Box Behnken design (BBD). Discharge current, pulse-on-time, servo-voltage, powder concentration, tool material and varying reinforcement levels were considered as machining input parameters. Material removal rate, tool wear rate and surface roughness were taken to be the response parameters. Analysis of variance (ANOVA) method was used to investigate the significant effect of parameters on the response measures. The experimental data was trained using a back-propagation ANN technique. Research shows that the influence of current, pulse length and tool material on the machining characteristics of Al LM-25 MMCs is significant. Surrogated models were also developed for proposed process using RSM. However, the accuracy of ANN models was found to be better than that of RSM models.
materials science, multidisciplinary,chemistry, physical
What problem does this paper attempt to address?