A new cost model for high speed hard turning of AISI 4340 steel

Abstract

Technology is developing from grinding to hard turning then to high speed hard turning (HSHT) in which three advanced turning processes are merged together: hard turning, high speed turning and dry turning. Among the advantages of this modern process are final product quality, reduced machining time, and lower machining cost. On the other hand, the high cost of the cutting tools must be minimized. Thus there is clearly a need to study HSHT economically to give the evidence to the industrial sector that this process is worth considering. In order to demonstrate its economic viability, it is of particular importance to develop a cost model to run in optimal conditions based on specified objectives and practical constraints. In this research, a scientific and systematic methodology to design cost model and the optimal cutting conditions is developed. A new procedure was proposed for structuring and developing the cost model. This procedure is a hybrid of different estimating methods supported by statistical and artificial intelligent techniques. In this study; the following parameters are used; cutting speed 175-325 m/min, feed rate 0.075 – 0.125 mm/rev, depth of cut 0.1 – 0.15 mm and negative rake angle from zero to (-12) degree. The work piece material was AISI 4340 hardened to 60 (HRC) and the cutting tool is mixed ceramics (AL₂O₃- TiC) under dry conditions. The experimental works were divided into two different designs; Full Factorial Design (FFD) for three independent variables (cutting speed, feed rate and negative rake angle) and Box-Behnken Design (BBD) in which the depth of cut was also considered. Surface roughness, flank wear length, cutting forces and feeding forces have been measured, and then two different methods have been used in modeling and predicting the measured parameters; BBD and ANN. It is found that the BBD gave the better results with lower deviation. Machining time and tool life were the main drivers of the cost in HSHT process. Machining time has been modeled by dividing the time into three main elements; operation time, non operation time and preparation time. The tool life was predicted by direct monitoring of the flank wear progress. It was found that the statistical approach gave better predicting to use in the cost models. The machining cost was measured by RM/cm² and broken down into three main parts; machine cost rate per time (labor, maintenance, material, energy, capital, and renting costs), tooling costs and energy cost then a mathematical relationship relating the cost of a process as a dependent variable to one or more independent variables were derived for three different scenarios; one shift, two shifts and three shifts with 300 working days. By using BBD as one of the RSM collections, new models for surface roughness, machine cost rate, tooling cost and energy cost have been developed using only four inputs variables. Two different optimization methods; Desirability function (DF) and genetic algorithm (GA), have been applied to obtain the minimum values of the total machining cost and surface roughness. The results show that the best values in the boundary design that can achieve the minimum roughness (0.31 μm) with machining cost 0.595 RM/cm³.