Font Size: a A A

Study On Six Sigma Tolerance Design Based On Concurrent Quality Engineering

Posted on:2006-05-25Degree:DoctorType:Dissertation
Country:ChinaCandidate:Z H ZhangFull Text:PDF
GTID:1119360212489378Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
This study aims to solve one of the difficult problems about Six Sigma critical techniques——that is how to design tolerance for Six Sigma based on concurrent quality engineering. The objective is to improve design quality and reduce cost in the up-front phases of the development cycle. Robust and optimizing methods on the basis of Response Surface Methodology (RSM) are used as the technical foundation of six sigma tolerance design. In this thesis, concurrent parameter and tolerance design models are constructed to realize the objectives of product robust design to process robust design.The paper opens with some basic concepts and traditional methods of tolerance design. The idea of concurrent quality engineering and the philosophy and practice model of Design for Six Sigma (DFSS) are presented. Subsequently technical methods are developed. Firstly, the method of six sigma mechanical design tolerancing is improved with design of experiment and Taguchi's tolerance design with variation coefficient and percentage of variance. Directly following on the discussion of the necessarity to concurrently design parameter and tolerance by dual responses of RSM, the models of concurrent six sigma tolerance design are introduced for various conditions. In addition, another model taking into account noise factors is presented with the restriction of minimum process variance confidence region of response model. Then on the basis of comparative analysis on Central Composite Design (CCD) which including Central Composite Circumscribed design(CCC),Central Composite Inscribed design(CCI) and Central Composite Face-centered design(CCF) in RSM, the paper proposes methods concerning how to choose experiment design of combined (single) array. Lastly, applied research of the proposed methods is demonstrated in order to validate and compare their validity. On the basis of comparison, some conclusions are derived as follows:(1) The methods of six sigma mechanical design tolerancing and Taguchi's tolerance design improved by DOE are more effective and sound than traditional methods. In each method, process capability index provides flexibility for designing for manufacturability.(2) The concurrent parameter and tolerance design is a more efficient manner than separate or sequential manner with a significant amount of total cost savings, which effectively proves the effect of concurrent quality engineering.(3) It is an effective strategy for six sigma tolerance design to reduce effect of variation through robust design of RSM. Furthermore the robustness and cost would be improved with the Response Model containing noise factors.(4) The design is more flexible with restriction of confidence region than rigid restriction. The economic settings of process mean and tolerance can be achieved while the process capability meets the criterion of six sigma.(5) The objectives of high quality and low cost can be achieved simultaneously by six sigma tolerance design. It is can be proved from the example that high quality is not conflicted with low cost. In fact, high quality would be result in low cost. The view is consistent with Deming's point.According to the idea of concurrent quality engineering and the mathematic models as the bridge of product design and process design, the manufacturability of concurrent product design and process design can be guaranteed. Six sigma tolerance design would provide a system methodology that can be implemented in practice in order to realize the goal of DFSS——economic and robust product design and process design.
Keywords/Search Tags:six sigma tolerance design, concurrent quality engineering, response surface methodology, robust design, concurrent parameter and tolerance design, cost function
PDF Full Text Request
Related items