| Abstract: |
This document reports on the findings from a collaborative effort by J. P. Hsu of the University of Texas at El Paso, T. W. Hsu (also of the same affiliation), James J. Filliben of the Statistical Engineering Division of the National Institute of Standards and Technology (NIST), and Theodore H. Hopp of the Machine Intelligence Group at NIST. The subject of the investigation was inspection sampling plans as related to coordinate measuring machines (CMM). The goal of the investigation was to compare various inspection plans, and, in particular, to make recommendations on minimal sample sizes to achieve pre-specified engineering tolerances on a CMM. The investigation is baseline in the sense that the recommendations flow from a part geometry which is--by design--fundamental (a plane) and a part characteristic which is fundamental (flatness). This report provides details on the engineering consideratiomns that went into setting up and running an appropriate experiment, and the several quantitative and graphical data analysis steps involved in the subsequent statistical analysis. We expect that the information included in this report will be of interest to CMM users, designers, and manufacturers; and we expect such procedural detail to serve as a guideline for other researchers involved on more complicated part geometries and part characteristics. |
Click here to retrieve PDF version of paper (784KB)