Hybrid ANN ‐Based Modeling and Optimization of Drilling Performance in Basalt Fiber Composites

ABSTRACT In this research, the parameters to be optimized are the machining parameters in drilling of the woven Basalt fiber‐reinforced epoxy composites. It focuses on the modeling and optimization of the spindle speed and feed rate of different laminate thicknesses with torque and delamination factor as some of the important output responses. Response Surface Methodology (RSM) and Artificial Neural Networks (ANN) will be used to analyze and model the effects and interaction of the control parameters on the drilling performance. In order to better predict the ANN model, the Artificial Bee Colony (ABC) algorithm is used in the training process. Also, RSM optimization technique is applied based on the desirability to prove the best combination of control parameters in the investigated range. The paper includes an elaborate discourse of the impact of process variables on the results of drilling performance. The best machining conditions identified were a feed rate of 0.1 mm/rev, a spindle speed of 1200 rpm to be used on a 2.7 mm thick laminate. RSM as well as ANN–ABC based predictive models were found to be in great agreement with the experimental results that can be used in sustainable application.