Study On Concurrent Tolerancing Based On Machining Features And Process Factors

Tolerance design affects functional roles and manufacturing costs of products. It has become a key technique and choke of information integration of CAD/CAM. Aiming at the problem currently that the conventional tolerance design only can be applied to the single machining features and the ideal process factors, and can not closely associate with the component and manufacturing resources, further researches on key theories and methods related to concurrent tolerancing based on machining features and process factors are made. The main contents and creative points are as follows.A new method for automatic recognition of machining features is presented. Based on the machining features model described by eXtensible Markup Language (XML), the parameters of machining features are extracted, including sketch figure, dimension information, membership etc. These parameters are then translated to matrix vectors base the proposed rule of fuzzy translation and are treated as the training samples of Self-Organizing Feature Map (SOFM) neural network. Based on SOFM special characteristics of self-studying and self-organizing, the current machining features in tolerance design such as plane, hole, column and locating can be recognized automatically. Then the computation for machining features cost in tolerance design can be fixed easily. A practical example is also introduced to validate the presented method.Study on the relationship between manpower, methods, machinery, measurement, material, mother-nature and manufacturing cost is carried out in this paper. Based on analyzing and using the practice statistic data, the integrated influence coefficient of manufacturing cost based on process factors is presented to scale the influence of each process to the manufacturing cost. The machinability of tolerance can be assured after introduce the integrated influence coefficient in to tolerance design.The relationships between process and quality loss are investigated. Based on the process dimension and tolerance bounds, the limitation of standard non-dimension multivariate quality loss function and signal-noise ratio (SNR) are employed to found the SNR quality loss weight. The weight can be used to scale the quality loss of each machining process, and can be used into concurrent tolerancing optimization model to well computer the computation of quality loss.A method is proposed to present a concurrent tolerance optimization model based on machining features and process factors. From product assembly tolerance design to process tolerance design, optimization design model is presented to directly allocate assembly tolerances into optimal process tolerance through using of the machining features of pertinent component, process factors in process planning, and the information of each process. An engineering example is used to validate the efficiency of the method.The prototype system for modeling and representation of tolerance information is given. According to the input information including machining feature styles, process plane, dimension information and the required function, the prototype system can implement automatic optimization and can output the most economic tolerance allocation.This thesis is financially supported by National Science Foundation of China, named"Modeling and Methodology for Integration of Design and Manufacturing in Tolerance Design"(No: 50465001).