Optimization Cutting Parameters on Turning Process to Increasing Surface Roughness SKT4 Material with Taguchi Method

This experimental is present the best parameter to get increasing surface roughness for JIS SKT4 material. Taguchi method was involved to combine parameter were used in turning process, namely cutting speed, feeding, depth of cut and tool nose radius. This experiment was conducted to find out best parameter combination for turning process which, the result is minimum roughness average JIS SKT4 material with carbide cutting tool material. The experimental used was Taguchi L27 with 3 times of replication. Backpropagation Neural network (BPNN) method is used to recognize relation between parameter process and experimental response, while Genetic Algorithm method is used to determine the best combination of process parameter that can optimize the surface roughness of JIS SKT4 material. BPNN have a 4-8-81 network architecture which consist of 4 input layers, 2 hidden layers with 8 neurons in the output layer. Tansig activation program and training program is used to process the data from taguchi metdod and experimental data. The optimum parameter recommendation from Genetic Algorithm are cutting speed 131.62 m/min, feeding 0.04 mm/rev, depth of cut 0.3 mm and nose radius is 0.39. The optimum parameter recommendation from Genetic Algorithm done with cutting experimental on turning machine with 3 repetation and the surface roughness average result is 1.5 μm. This experimental improve 301.03% surface quality from the product of guide pin JIS SKT4 material.