5. GRR Methodology for Destructive Testing and Quantitative Assessment of Gauge Capability for One-Side Specifications
Teresa Mitchell, Victor Hegemann, K. C. Liu · Society for Industrial and Applied Mathematics eBooks · 1997
Gauge repeatability cannot be estimated by traditional methods when the measurement process is destructive in that the product is destroyed in the process of being measured. A methodology is presented for estimating gauge repeatability for the destructive bond pull strength tester used in semiconductor manufacturing. In addition, the traditional measurement system capability index Cp cannot be used to evaluate the impact of measurement variability on product quality for one-sided specifications such as those on pull strength, where bigger pull strength is better. A new definition of measurement system capability for the case of one-sided specifications is presented, with examples that demonstrate its usefulness. Executive Summary Problem Variability in the measurement system impacts the values of the quality characteristics of a process. To ensure meaningful analysis of the process, the measurement instrument must be accurate and repeatable.