Analysis of Destructive Measuring Systems

World Conference on Quality and Improvement vol. 43 - May 1989

Abstract:For ongoing improvement in quality it is imperative that all sources of variation be identified to enable corrective action to be taken to reduce the variability of finished products. To provide this information at least one if not more measurement systems will be used to measure the variables being studied. These measuring systems have many sources of variation including the appraiser, the equipment, the methods used, and the environment. In some cases they involve destructive testing. Many of the more complex measuring systems being used in industry today do not fall neatly into the models generally used for measurement systems analysis. Destructive testing by its very nature does not lend itself to classical gauge repeatability and reproducibility (GRR) studies. The same samples cannot be used for repeat testing by the appraisers involved in the study. In such a case different samples must be used introducing variation due to the material. In the particular case of the tensile test more than one measuring device is involved compounding the complexity of the analysis. This paper will outline a statistical and graphical approach to analyzing a tensile strength measuring system. The same approach can be used for all destructive measurement systems.

Keywords: Gage Repeatability and reproducibility studies (GR&R) - Statistics - Destructive testing