Font Size: a A A

The Parallel Tolerance Robust Design Method Research

Posted on:2013-06-20Degree:MasterType:Thesis
Country:ChinaCandidate:F Z LiFull Text:PDF
GTID:2241330374986340Subject:Machinery and electronics engineering
Abstract/Summary:PDF Full Text Request
Tolerances connect the design and manufacturing of a product. Tolerances design has a great impact in manufacturing cost and product quality. Up to now, the research performed in tolerances design is focused on how to control the cost and quality, however, the study of robust tolerances design is less in amount. Firstly fundamental conception about tolerances allocation is introduced. Sensitivity robustness and feasibility robustness, which are two main segments of robust design, are incorporated into concurrent tolerances design in this paper.Better performance requires tighter tolerances which will lead to higher cost. In this paper, the minimum total cost is achieved by moving the nominal values of the non-critical dimensions to a less sensitive portion, which does not cause the decrease of the tolerances. In practical tolerances design problems, the process capability fluctuates in a small range, and then the uncontrollable variations change the constrained conditions and significantly affect the feasibility of an optimum solution. In order to overcome this problem, this paper proposed a method of feasibility robustness for tolerances allocation. The feasibility model is builded and the WCSR method is adopted to find the feasible solution.Traditional genetic algorithm suffers from many drawbacks such as premature convergence. Immune genetic algorithm, a novel genetic algorithm based on immunity, is applied to solve the model of tolerances sensitivity design. Although quality loss has been taken into account in many papers, the precision of the value of the quality loss coefficient can not be guaranteed. This paper proposed a method of double-objective optimization that the manufacturing cost and quality loss are considered as two objective functions. The quality loss coefficient is not the key of consideration. NSGA-II, a novel multi-objective optimization method, is used to deal with the double-objective optimization. The optimum results are a number of Pareto optimal solutions which can satisfy diversified market requirements.Oneway clutch assembly and piston-cylinder assembly is adopted as a demonstration. In the end, the paper discussed the future research about tolerances design.
Keywords/Search Tags:tolerances allocation, quality loss, robust design, IGA, NSGA-Ⅱ
PDF Full Text Request
Related items