Formulation Techniques in Regression Analysis to Estimate the Roughness Value from Tribological Parameters in Hard Turning of AISI 52100 Steels

摘要

To machine any material the tribological parameters play an important role in the machining. The main important tribological parameters are speed, feed, and depth of cut. The hard material like EN-31 or AISI 52100 steels are having hardness in between the range of 41 HRC to 69 HRC on C scale of Rockwell hardness. To machine hard materials like this, the special purpose tools or the machinesare required. Most of the times the grinding process perform to machine these kinds of materials. But the alternatives like ceramics, cBN, PcBN tools are also available to machine these materials instead of grinding. The different tools are giving different values of roughness after machining, but the roughness value can be optimized by using the different statistical techniques by comparing the different combinations of tribological parameters while performing the machining operations on the hard materials. The various statistical software's and techniques are available in the market. But the most viable technique available for correlation is the regression analysis. The different types of models (i.e. linear, quadratic, and cubic) are available in regression analysis to correlate various tribological parameters. The operations perform on the AISI 52100 steel materials of hardness 58 HRC by varying the tribological parameters according to the L_9 Taguchi Design. The modeling is done through the linear regression analysis, quadratic regression analysis, and cubic regression analysis and the comparison between different models of estimated versus experimental roughness value. The result analysis shown and conclude that the estimated correlations would be able to predictswith accuracy of 99.50% to 99.90% and last but the least 6.19% uncertainty is present in the experimentations.