Decision support for off-line gage evaluation and improving on-line gage usage

Abstract Many indices for evaluating gage capability have been developed and applied, but these methods are ambiguous to users. By considering both manufacturing capability and gage capability, a mathematical model for the measurement loss cost was developed for dinstinct off-line gage evaluation; however, some measurement loss cost still exists after the gage capability study. On-line usage of a gage is improved by reducing the remaining on-line measurement loss cost with the concept of guard limits. Feasible guard limits can be determined by considering the measurement loss cost resulting from both type I and type II errors. Instead of applying a series of gages to confirm the measurement on every part, unreliable measurement can be screened out in advance by the guard limits. Only the necessary precise inspection needs to be further performed. Rather than applying a better gage with higher cost, the concept of guard limits can also be applied to produce a better or similar effect when adopting a gage that is marginal for its applications. The unsatisfactory on-line measurement loss cost can be further reduced until it is acceptable.