Optimization of Turning Nst 37.2 Steel with Uncoated Carbide Cutting Tools

Abstract

Selection of optimum machining parameters is an essential factor in process planning for efficient metal cutting operations. In this study, an artificial neural network- based tool wear predictive model and a genetic algorithm-based optimization model were developed to determine the optimum cutting parameters for turning NST 37.2 steel with uncoated carbide cutting inserts. Multi-layer, feed-forare, back -propagation network was used in predictive model, while maximum metal removal rate (MRR) was used as the objective function and tool wear as samples NST 37.2 steel bars with 25mm diameter and 400mm length workspiece and Sandvice Coromant(R) uncoated carbide inserts with International Standard Organization (ISO) designation SNMA 12406. Dry machining at different cutting conditions with cutting speed (v), feed rate (f) and depth of cut (d) ranging from 20.42-42.42 mm/min, 1.0-2.2mm/rev and 0.2-0.8mm, respectively were carried out. Eight passes of 50mm length of cut were machined at each cutting, the spindle power and tool wear (flank and nose) were measured during each cutting operation. Results have shown that the predictive model had acceptable accurancy and optimum cutting parameters obtained were: v =42.32mm/min, f = 2. I9 mm/rev and d = 0.8mm.