INTELLIGENT MODELLING AND MACHINING CHARACTERISTICS OF HYBRID MACHINING FOR HYBRID METAL MATRIX COMPOSITES

With the invention of advanced engineering materials, the shaping of such materials became a challenging issue. Two or more reinforcements are added to the metal matrix and form hybrid metal matrix composites (HMMCs) with preferable material properties, but shaping became quite challenging. Hybrid Surface grinding Electrical Discharge Diamond Face Surface Grinding (EDDFSG) is a suitable hybrid machining process capable of machining complicated HMMCs. With the help of EDDFSG, adverse effects of both traditional and non-traditional techniques are overcome while combining their benefits for a better machining results. A hybrid composite machined with a hybrid machining process requires an advanced technique for modeling and the performance prediction of complex machining characteristics. Material removal rate (MRR) and (R-a) rely on the process parameters, their influence must be extensively studied. Machine learning, a subset of Artificial Intelligence, allows machines to learn, develop, and execute tasks like human beings based on data rather than explicitly programmed. In the present work, an attempt has been made to develop a Machine Learning (ML), K Nearest Neighbor (KNN) based model, to predict MRR and Ra for machining EDDFSG of Al/Al2O3p/B4Cp and Al/SiCp/B4Cp HMMCs. The KNN algorithm is one of the efficient ML models for regression. Our training data set is normalized using the Min-max scalar to avoid a biased algorithm towards one process parameter. The model's accuracy is validated by average standard error metrics on the test data set. The impacts of the process parameters like pulseon time, gap current, wheel speed, pulse-off time, grit num- ber, table speed over the response variables of the ML model is studied and analyzed in depth. The remarkable results are found pertaining to machining characteristics of the EDDFSG process over traditional modelling techniques.