Comparison of artificial neural model and response surface model during EDAG of metal matrix composite

Metal matrix composites (MMCs) faces machining challenges due to its superior mechanical properties. Hybrid machining processes (HMPs) are gaining popularity for machining of MMCs and newly developed advanced materials. Electrical discharge abrasive grinding (EDAG) is such an HMP combines unconventional electrical discharge machining and conventional grinding. In present research the experimental investigation of the copper-iron-graphite MMC has been presented for one of the important quality characteristics; average surface roughness (ASR) during EDAG. The artificial neural network (ANN) and regression modeling have been used to develop the predictive models for ASR. Both the models have been compared for their suitability to predict ASR.