Application of JAYA algorithm for the optimization of machining performance characteristics during the turning of CFRP (epoxy) composites: comparison with TLBO, GA, and ICA

With the widespread application of carbon fibre-reinforced polymer (CFRP) composites, mostly in defence, automotive, and aerospace industries, the machining of those materials has become a major concern today. As the machinability of those composites differs from the conventional metals, a proper understanding of process behaviour and identification of the favourable machining environment (optimal setting of process parameters) are indeed necessary to improve product quality. The present work highlights the application potential of a multi-response optimization route by integrating nonlinear regression modelling, fuzzy inference system (FIS) in combination with the JAYA optimization algorithm, for the selection of optimal process parameter setting during the machining (turning) of carbon fibre-reinforced (epoxy) composites. Experiments have been carried out in consideration with spindle speed, feed rate, and depth of cut as process control parameters, whereas material removal rate (MRR), roughness average (R (a)), and net cutting force have been treated as machining performance characteristics. Attempt has been made to identify the best setting of process parameters for optimizing aforesaid output responses, simultaneously. The result of the JAYA algorithm has also been compared to that of TLBO (teaching-learning-based optimization) algorithm. In addition to this, the result obtained thereof has also been compared to that of two evolutionary optimization algorithms viz., GA (genetic algorithm) and ICA (imperialist competitive algorithm). Good agreement has been observed amongst the obtained results. The aforesaid case experimental study thus exhibits the application potential of a newly developed JAYA algorithm in the context of machining performance optimization during the turning of CFRP composites. The JAYA algorithm is basically a parameter-less optimization algorithm which does not require any algorithm-specific parameter and hence easy to implement.