OPTIMIZATION OF FORCES AND SURFACE ROUGHNESS DURING TURNING OF EN31 STEEL

Abstract

Machining is one of the most common and important manufacturing process in the industry. The investigation of the machining process and development of models in order to optimize cutting force, feed force, thrust force and surface roughness is important when it comes to the cost of manufacturing and the quality of finished products. Effect of input parameters: Spindle Speed, Feed and Depth of Cut is determined by correlating these parameters with cutting force, feed force, thrust force and surface roughness. The main objective of this research was to study the effect of spindle speed, feed and depth of cut on cutting force, feed force, thrust force, surface roughness and tool wear during turning of EN 31 Steel using uncoated and coated titanium based cemented carbide insert. EN-31 Steel is high carbon low alloy steel giving good ductility & shock resisting properties combined with wear resistance. It is also known as bearing steel as it is used for production of bearings. The design matrix was prepared on the basis of 3 factors, 3 level Box-Behnken Design. Response Surface Methodology was used for the development of mathematical models correlating spindle speed, feed and depth of cut with cutting force, feed force, thrust force and surface roughness. Modelling of tool wear could not be done because of the limited readings as no tool wear was observed while machining with coated insert whereas a maximum tool wear of 0.05 mm was observed while machining with uncoated insert. The developed models were checked for adequacy using ANOVA. All the calculations were carried out using Minitab 17. Main and interaction effects were plotted and results were interpreted. The developed models can be suitably used for predicting the response parameters by selecting appropriate input parameters. Optimization was carried out using Response Optimizer Technique which performs joint optimization as well as using Genetic Algorithm Multi Objective Optimization. For the latter, the models developed using RSM were utilized as the fitness function and optimization was performed yielding a Pareto front. When the regression models are not fitting the data well, then only modelling using Artificial Neural Network is attempted [20]. Therefore ANN models for cutting force, feed force, thrust force and surface roughness were also developed. The results indicate that the ANN model results are satisfactory. iii A comparative study of Experimental results, RSM model results and ANN model results was done and it was concluded that ANN model has an edge over RSM model values in case of cutting force, feed force and thrust force. But in case of surface roughness, RSM has less absolute mean percentage error in comparison to ANN model values.