Study On Computer-aided Statistical Tolerance Design

There have been significant breakthroughs and remarkable achievements in the field of CAD/CAM and CIMS in the last few years, but as an important element of mechanical product design and manufacturing process, Computer-Aided Tolerancing(CAT) lags far behind the CAD/CAM development, and becomes bottleneck that constraints development of CAD / CAM and CIMS. In order to improve this situation, domestic and foreign scholars have done a lot of research work.With referencing to a large number of domestic and international papers in the field, Some problems of computer-aided statistical tolerance design are studied in this paper.Distribution feature of processing error is the basis of statistical tolerance design. By investigating high-volume parts' error data, the nature and distribution feature of error is analyzed in this paper. Experimental results show that distribution of high-volume parts' processing error is basically the same as the normal distribution.Tolerance analysis is to study the impact of component links on closed link. As a statistical test method, Monte Carlo method has been applied in many experiments and simulation. In this paper, computer-aided statistical tolerance design based on Monte Carlo simulation is studied; flow of tolerance analysis with Monte Carlo methods is proposed; Monte Carlo simulation softwares programmed by MS EXCEL2003, Visual Basic6.0, Matlab7.0.1 is carried out to analyze and design tolerances. As an application of statistical tolerancing, the axial and radial tolerance of twin screw pulping extruder is analyzed, and . equations of linear and non-linear dimension chain are established, tolerance analysis and design are done by Monte Carlo method. The analysis results show that Monte Carlo methods have fast speed and high accuracy; and it can further reduce manufacture precision than RSS method under condition of the same assembly success rate.The common flow of tolerance optimization design is introduced; the tolerance optimization allocation with lowest-cost method is studied. Under the constraints of Worst Case method and statistical method and basing on the cubic polynomial model, the optimal radial tolerances of twin screw pulping extruder are allocated by using Matlab optimization toolbox. Optimization results show that: the use of lowest-cost method to optimize the allocation of tolerances can reduce the parts' tolerance requirement and processing costs, and can obtain good economic benefits.The study of this paper is of some importance in Monte Carlo tolerance analysis and tolerance optimization allocation, and of guiding significance in tolerance design of twin screw pulping extruder.