Search Results

The concept of design for six sigma (DFSS) offers a framework to design a product and process right the first time. In general, Taguchi's robust design method has been widely adapted in design optimization, which is a critical phase in any DFSS projects. The objective of the paper is to develop an advanced strategy in selecting an optimized product design and manufacturing process that should be insensitive to various multivariate variation patterns of the multi-stage manufacturing system. A Monte Carlo variation simulation based method is presented that integrates Mohalanobis Distance (MD) method, a discriminant analysis technique, to analyze the manufacturing variation patterns detected by using the multivariate statistical tool, such as principal component analysis (PCA). The proposed method will be explained with an example of an automotive assembly.