Experimental Investigations on Machining of CFRP Composites: Study of Parametric Influence and Machining Performance Optimization

Carbon Fiber Reinforced Polymer (CFRP) composites are characterized by their excellent mechanical properties (high specific strength and stiffness, light weight, high damping capacity etc.) as compared to conventional metals, which results in their increased utilization especially for aircraft and aerospace applications, automotive, defense as well as sporting industries. With increasing applications of CFRP composites, determining economical techniques of production is very important. However, as compared to conventional metals, machining behavior of composites is somewhat different. This is mainly because these materials behave extremely abrasive during machining operations. Machining of CFRP appears difficult due to their material discontinuity, inhomogeneity and anisotropic nature. Moreover, the machining behavior of composites largely depends on the fiber form, the fiber content, fiber orientations of composites and the variability of matrix material. Difficulties are faced during machining of composites due to occurrence of various modes of damages like fiber breakage, matrix cracking, fiber–matrix debonding and delamination. Hence, adequate knowledge and in-depth understanding of the process behavior is indeed necessary to identify the most favorable machining environment in view of various requirements of process performance yields. In this context, present work attempts to investigate aspects of machining performance optimization during machining (turning and drilling) of CFRP composites. In case of turning experiments, the following parameters viz. cutting force, Material Removal Rate (MRR), roughness average (Ra) and maximum tool-tip temperature generated during machining have been considered as process output responses. In case of drilling, the following process performance features viz. load (thrust), torque, roughness average (of the drilled hole) and delamination factor (entry and exit both) have been considered. Attempt has been made to determine the optimal machining parameters setting that can simultaneously satisfy aforesaid response features up to the desired extent. Using Fuzzy Inference System (FIS), multiple response features have been aggregated to obtain an equivalent single performance index called Multi-Performance Characteristic Index (MPCI). A nonlinear regression model has been established in which MPCI has been represented as a function of the machining parameters under consideration. The aforesaid regression model has been considered as the fitness function, and finally optimized by evolutionary algorithms like Harmony Search (HS), Teaching-Learning Based Optimization (TLBO), and Imperialist Competitive Algorithm (ICA) etc. However, the limitation of these algorithms is that they assume a continuous search within parametric domain. These algorithms can give global optima; but the predicted optimal setting may not be possible to adjust in the machine/setup. Since, in most of the machines/setups, provision is given only to adjust factors (process input parameters) at some discrete levels. On the contrary, Taguchi method is based on discrete search philosophy in which predicted optimal setting can easily be achieved in reality.However, Taguchi method fails to solve multi-response optimization problems. Another important aspect that comes into picture while dealing with multi-response optimization problems is the existence of response correlation. Existing Taguchi based integrated optimization approaches (grey-Taguchi, utility-Taguchi, desirability function based Taguchi, TOPSIS, MOORA etc.) may provide erroneous outcome unless response correlation is eliminated. To get rid of that, the present work proposes a PCA-FuzzyTaguchi integrated optimization approach for correlated multi-response optimization in the context of machining CFRP composites. Application potential of aforementioned approach has been compared over various evolutionary algorithms.