Research Article
Sarnobat SS and Raval HK
Abstract
The quality of machined surface is of vital significance due to its bearing of the in-service functionality of the component. In-service functionality of the machined parts like, tribological performance, fatigue life of the component etc.; are greatly dependent on the surface profile characteristic and the surface roughness generated after machining. However, the quality of surface is reliant on complexities of the numerous process parameters. The mechanics of metal cutting necessarily results into the dynamic instability of the process consequentially ensuing into cutting tool vibrations. Previous research indicates an association between the cutting tool vibrations and surface roughness. In this study the cutting tool vibrations in tangential and axial direction have been integrated with the input parameters; cutting speed, feed rate, depth of cut, work material hardness and tool edge geometry to develop prediction models for surface roughness from the experimentally obtained data by using Regression Analysis and artificial Neural Network methodologies. The results of the regression models and neural networks model are compared. A good agreement between the experimental and predicted values for both the models is seen, however neural networks approach outclasses regression analysis by a reasonable margin. Further it is also noted that the quality of surface is markedly influenced by the tool edge geometry and feed rate.