Optimization of Wire-cut EDM process parameters using TLBO algorithm

The novelty of the work is reflected in using Teaching Learning Based Optimization as metaheuristic algorithm in combination with Response Surface Methodology for achieving multi response optimization of two conflicting responses while machining Incoloy 800H on Wire cut EDM. It is used in materials and equipment for furnace, petrochemical furnace cracker tubes, pigtails and headers, and electric heating element sheathing. There are a number of process parameters of Wire cut EDM which affect the output responses. Here, the effects of Pulse On Time, Pulse Off Time, Peak Current and Servo Voltage on the Cutting Rate and Dimensional Deviation were studied. Experiments were performed according to Box Behnken design with 29 experiemental runs. Pulse On Time was found to be the most significant factor in affecting the Cutting Rate and Dimensional Deviation as envisaged by Analysis of Variance . First, single response optimization was performed using Teaching Learning Based Optimization Algorithm which resulted in an optimal factor level setting of Pulse On Time-120mu, Pulse Off Time-45 mu, Peak Current-200A, Servo Voltage-47V corresponding to Cutting Rate, whereas, Pulse On Time-120mu, Pulse Off Time-35 mu, Peak Current-160 A, and Servo Voltage-30V formed an optimal setting of the factors corresponding to the Dimensional Deviation. Multi-response optimization resulted in an optimal factor level setting of Pulse On Time-120 mu, Pulse Off Time-38 mu, Peak Current-190 A, and Servo Voltage-53 V corresponding to both Cutting Rate and Dimensional Deviation. Then, the optimal factor level settings for single response and multi-response optimizations were confirmed through three confirmation tests. The average of the confirmation tests was compared with predicted value and percentage error was calculated. The percentage error comes out to be less than 5% in all the cases.