PECASE: A Unified Methodology for Variation Management and Reduction in Multistage Manufacturing Processes

点击翻译按钮获取中文摘要

Proposal Title: PECASE: A Unified Methodology for Variation Management and Reduction in Multistage Manufacturing ProcessesInstitution: University of ArizonaThis PECASE grant sponsors research and teaching efforts in developing a unified methodology for variation management and reduction in Multistage Manufacturing Processes (MMPs). A MMP generally involves multiple operations to produce a product, which can be found in many industrial processes such as automotive body assembly, machining lines, progressive stamping, and semiconductor manufacturing. In a MMP, each operation adds inherent design variation when no fault occurs, and special assignable variation when a fault occurs, to the workpiece variation. The output workpiece of one operation is the input of the next operation. The final product variation is an accumulation of variation from all operations. Therefore, the characteristics of variation propagation is very complex in MMPs, depending on both product and process design. The research efforts in the project will develop methodologies to achieve on-target production with minimum variance by addressing both design and manufacturing concurrently. The research focus is to develop: (a) a math-based model to describe the variation propagation in a MMP; (b) an integrated methodology for variation management through design synthesis and optimization; and (c) statistical methods driven by engineering models for quick root cause identification and product defect prevention. In addition to research, new curriculums and courses will be developed. Various collaborative efforts with industry, international universities, and K-12 schools will be conducted in research and education in the project.If successful, the project results will enrich the science base and technologies in variation reduction and process control methodologies for MMPs, which include: (a) analytical (rather than empirical) modeling of MMP variation and its propagation; (b) synthesis (rather than analysis) of product/process tolerance for optimal management of inherent design variation to minimize their impact on final products; (c) real-time root-cause identification (rather than change detection) for manufacturing process control and continuous improvement; and (d) prognostics (rather than defect inspection) for defect prevention throughout the manufacturing lifetime. The developed unified methodology will generate significant economic impacts in industrial sectors where MMPs are used.This project was originally funded as a CAREER award, and was converted to a Presidential Early Career Award for Engineers and Scientists (PECASE) award in May 2004.