Statistical Issues in Geometric Tolerance Verification UsingCoordinate Measuring Machines

There is currently little or no guidance in the tolerancing standards for evaluating geometric variation of part feature using coordinate measuring machines (CMMs). Practitioners are faced with many important but unanswered questions, including the number of measurements that should be made on a part feature, and the location of such measurements. Since the standard requires knowledge of the entire feature while a CMM can take only a small sample in a reasonable time period, there are no simple answers to these questions. The objective of this research is to develop a rigorous procedure for the verification of standard geometric tolerance specifications, developing tradeoffs between accuracy and cost. The research will involve the following tasks: a detailed study of parts to allow identification of common profile variation patterns for specific processes; the development and evaluation of estimation techniques for the measurement of geometric variation; and the study of the relationship between the number of measurements and the resulting sampling risks. The impact of the research will be in several areas, including the development of a procedure to determine an appropriate number and location of measurements to satisfy specified risk levels. Prior information will be exploited regarding feature characteristics to reduce sample sizes while maintaining the same estimation accuracy. Educational impact will also be included in the form of classroom instruction, research students, seminars, and conferences.