Fuzzy Assessment Model To Judge Quality Level Of Machining Processes Involving Bilateral Tolerance Using Crisp Data

Ongoing pursuit of greater process quality has always granted manufacturers advantage in an increasingly competitive market place. This has prompted numerous researchers to develop quality assessment models for various machining processes. The Six Sigma quality management method, which determines process quality based on yield, is a useful standard of process quality which provides reference to both manufacturers for process improvements and consumers in product selection. This study has employed the loss-based Six Sigma quality index (SSQI) Q(pm) to analyze the quality level of a machining process with bilateral tolerance. The resulting evaluation with Q(pm) not only reflects loss and yield for the machining process; it also directly presents the attained quality level. In practice, Q(pm) must be estimated based on collected data to determine process quality. Unfortunately, uncertainty and imprecision are inevitable features of any data collection. This can lead to erroneous inferences of quality assessment using the crisp-based estimate (Q) over cap (pm). We, therefore, propose a fuzzy estimate of Q(pm) and develop the associated fuzzy statistical testing to increase the reliability of assessment and reduce miscalculation. The proposed fuzzy statistical test method is illustrated via a real-world example from a gear manufacturing plant.