Modeling of thrust force in drilling of CFRP composites using adaptive neuro fuzzy inference system

Abstract

Carbon fiber reinforced plastic (CFRP) material is identified as an emerging material for solving critical problems such as light weight, corrosion resistance and environmental durability. CFRP suits these properties in various engineering applications that have structural variations. In order to join such structures, drilling is an essential operation. Several problems are encountered in drilling of composites which delamination poses a major threat. Thrust force are directly related to it, and hence this response is measured and modeled using ANFIS in this paper. Model adequacy check is carried out by calculating the R-squared values and other useful error definitions such as root mean square error, mean absolute error and mean square error. The significance of ANFIS is illustrated by the plot of membership functions before and after training. Two-Gaussian membership function provides a better model on comparing with other common membership functions such as triangular, gbell, Gaussian and two-Gaussian membership functions.