Published Online: 24 May 2013
Page Count: 8
Dept. of Industrial Management, National Taiwan Univ. of Science and Technology, Taipei,
(Received 18 April 2012; accepted 16 January 2013)
Measurement system analysis plays an important role of many continuous improvement initiatives in the manufacturing industry. Measurement systems are commonly evaluated by performing a gauge repeatability and reproducibility (GRR) study. In some circumstances, manufacturing tests in a GRR study may have multiple characteristics with a multivariate normal distribution. To determine the preferred method for a GRR analysis, multivariate analysis of variance (MANOVA) and principal component analysis (PCA) are applied to data taken from two real cases. Three performance measures [precision to tolerance ratio (PTRM), the percentage of repeatability and reproducibility (%RRM), and signal-to-noise ratio (SNRM)] are used to evaluate the adequacy of the measurement process. The results show that the MANOVA method is suitable for a GRR study with multiple characteristics.
Paper ID: JTE20120113