Study On The Methodology And Technology Of Robust Tolerance Design For Manufacture

On the basis of comprehensive analysis for recent research status of computer-aided Tolerancing both in and out of China, binding State Natural Science Foundation of China ("Study on Simultaneous Design of Blueprint Tolerance and Process Tolerance", No.59575076 -, "Study on Computer Aided Tolerancing System" No.59705022 and "Design theory of integration Tolerance based on Geometrical Product Specifications" No. 50275136) and State Natural Science Foundation of Zhejiang ("Study on Tolerance Design System Based on Robustness and Life Circle Cost" No.500115 and "Study on Process Tolerance design for CAPP" No.502016), the methodology and technology of robust tolerance design for manufacture is studied.In chapter 1, the important role of computer-aided Tolerancing in modern manufacture and research status on key theories and technologies of the region are summarized. The main contents, general structure scheme and innovation points of this dissertation are given.In chapter 2, the total cost model involved in tolerance design is studied. Firstly, a new cost-tolerance model, which closely links with real technological process, is built using fuzzy neural networks. The influence of process factor on the manufacturing cost is analysed and a fuzzy process factor is introduced using fuzzy mathematics principle. Then fuzzy neural network is applied to modelling of production cost-tolerance model. The inputs of the network are tolerance and fuzzy process factor and the output is manufacturing cost. Secondly, Taguchi's quality loss function is a simple quadratic symmetric style, the paper expands it a fuzzy quality loss model using fuzzy theory. A concept of fuzzy quality loss and fuzzy quality loss cost are first presented. Then a fuzzy quality loss model is put forward and a multiple characteristics' fuzzy quality evaluation method is given.In chapter 3, a robust tolerance design method is put forward according to fuzzy quality loss model. Firstly, the design model with fuzzy quality loss model is presented. Then tolerance design method utilizing two main kinds of DOE, i.e. Taguchi's method and response surface method, are given. The comparison and analysis between these methods is made at last.In chapter 4, design and evaluation method of feasible robustness and sensitive robustness in tolerancing is studied. Firstly, the concept of feasible robustness and sensitive robustness is given. Then the tolerance feasible robust design model and tolerance sensitive robust model are put forward. The optimisation methods are introduced to solve these models and some examples are given to verity the proposed method. The evaluation method is given at last.In chapter 5, with DFM (Design for manufacture) ideas, a framework of integrated system of robust tolerance design for manufacturing is established. A method for robust tolerance design in terms of Process Capability Indices (PCI) is presented. The component tolerance and the suitable manufacturing processescan be selected based on the real manufacturing context.In chapter 6, robust tolerance design for manufacture based on product data management system was build on MDT. Some examples are studied to demonstrate the proposed theory and method.In chapter 7, the important results and conclusions of the dissertation are summarized and the future research of compute-aided tolerancing and robust design is prospected.