Paper Title
A PCA BASED QUALITY LOSS APPROACH FOR OPTIMAL GEOMETRY CHARACTERISTIC IN ECDM PROCESS

Abstract
Abstract - The present work uses a novel optimization technique combining PCA with quality loss concept for the selection of best setting of Electro Chemical Discharge Machining (ECDM) parameters for minimum geometric hole characteristic induced during drilling in drilling glass. Experiments were conducted with different drill, applied voltage, feed rate and machining time each at three levels using L27 setting of the orthogonal arrays. The responses considered to evaluate the geometric hole characteristic are Radial overcut (ROC), Circularity, and heat affected zone (HAZ). Principal component generated the uncorrelated linear combinations of original response which then used to derive combined quality loss i.e. hole geometry performance index (HGPI). Response table, response graph and the analysis of variance (ANOVA) were used to determine the optimal setting and the influence of ECDM parameters on the HGPI. The results obtained from the above analysis are validated by performing confirmation experiments. Keywords - Hole Geometry Performance Index (HGPI), Principal Component Analysis (PCA), Quality Loss, Taguchi Design, Co-Relation