Variation analysis for flexible sheet metal assembly with principal component analysis approach

Li Bing,Ran Tao,HuYing
2006-01-01
Abstract:Products made of flexible sheet metals are widely used in automotive, aerospace, appliance and electronics industries. One of the most important challenges for the assembly process with flexible parts is dimensional variation, which affects product functionality and performance. Since clamping and joining during assembly may introduce additional variation due to part deformation and spring-back. A new methodology is proposed for predicting flexible assembly variation considering the correlations of sources of variation. In this paper a new methodology is proposed for predicting flexible assembly variation considering the correlations of sources of variation. It incorporates the use of principal component analysis (PCA) and finite element analysis (FEA) in evaluating the effect of component variation on assembly variation. In addition, squared Mahalanobis distance is applied to detect the outliers from the whole sample data. The methodology is illustrated via a case study on cross-number assembly in BIW.
What problem does this paper attempt to address?