Virtual Assembly Tolerance Analysis Technology Research

Tolerance is a bridge of connecting performance requirement and machining of a product. It affects the performance, the production cost and machining process of a product. How to allocate rationally tolerance of parts and control stack tolerance is a key factor to gain a good cost-performance ratio and win market competition. At present, the tolerance design of products is far behind the development of CAD/CAM, the CAT (Computer Aidede Tolerancing) technique is an important problem to be solved in the development of external and internal technique of advanced manufacturing.Tolerance design includes tolerance analysis and tolerance allocation. From the point of view of the dimension chain, tolerance analysis is a calculation from dimensions and tolerances of composing loops to dimension and tolerance of close loop. Tolerance allocation is an inverse calculation from tolerance of close loop and nominal dimensions of composing loop to tolerances, upper and lower limits of composing loop in the condition of ensuring assembly requirement. In practical production and machining, many parts constitute an assembly product. So in virtue of the technique of virtual assembly and virtual prototype, tolerance analysis can validate whether the product assemblied is in accord with precision requirement, and analysis whether tolerances of part has influence upon product precision. Because of good check and feedback of analysis result, design quality can be improved.In this paper, tolerance analysis mainly relates to three methods, viz. "Extremum Method", "Statistics Method" and "method of Monte Carlo". The character of the "Extremum Method" is to calculate the limit dimension of the close loop according to every limit dimension of the composing loop in order to ensure an entire frogging character, and get the tolerance of close loop. The character of the "Statistics Method" is to ensure the frogging of the large sizes. According to the distribution of actual dimensions of the composing loop in the tolerance, it works out the actual distribution range of the close loop in terms of a believable probability, and calculates the tolerance of the close loop. When the relation between assembly tolerance and part tolerance is linear, the two metnods are applicable. Compared with "Extremum Method" and "Statistics Method", the method of Monte Carlo is suitable when assembly function is nonlinear, and the distribution of part parameter may be normal or nonnormal. So the method of Monte Carlo has a good precision and it is fit for practical production. It is applicable when a complicated problem...