Non-linear Assembly Deviation Analysis And Tolerance Co-allocation For Aircraft

Digital assembly coordination represents the development direction of the current and future aircraft assembly coordination technology, and it has great significances in improving the assembly precision and success rate, shortening the development cycle as well as reducing production cost. At present, the basic principle and system composition of digital assembly coordination technology have been widely recognized in China, but some basic theories in digital assembly coordination such as deviation propagation modeling, analysis and tolerance allocation have not been fully grasped, which is not conducive to playing the further application and effect of this technology. On the one hand, the complex aircraft product structure and shape, large quantities of parts or frocks, long assembly process and the highly nonlinear and cross propagating characteristics among deviations such as size, shape and position generated in the process of assembly coordination make the assembly deviation propagation modeling and analysis extremely difficult; on the other hand, as the coordination basis is three-dimensional model, tolerance allocation knowledge and experience accumulated in analog assembly coordination are no longer applicable, therefore, designers feel a great difficulty to accurately complete the tolerance allocation based on the three-dimensional model. To solve these problems, the nonlinear assembling deviation propagation modeling and analysis, as well as the assembling tolerance allocation method are studied in this dissertation based on the three-dimensional model and considering the generation and transfer of key characteristics as clues, which provides the theory and method for digital assembling coordination technology in China. Its main research contents and innovative works are as follows:1) Based on the nonlinear and span-order cross transfer characteristics of aircraft assembly deviation, the weighted graph expression method of nonlinear assembling deviation propagation process is proposed. First of all, the paper analyzes and concludes business process and information related to assembling deviation transfer modeling, and respectively establishes hierarchy relationship tree diagram, derivative coordination constraint graph,feature tolerance graph, assembly sequence expression method and deviation expression method; secondly, it proposes the weighted graph-tree model and construction rules of assembling deviation transfer process, and the integrated expression of realizing assembly deviation information-transfer, and it also constructs the instantiation process and method;finally, the nonlinear relationship graph of weighted graph-tree model is constructed, and the nonlinear relationship in the assembling deviation transfer is selected.2) In response to the defects that key characteristic identification at this stage depends on the engineering experience and the related technological process knowledge is not easy to express, quantization method of key characteristics and reasoning identification methods are put forward. Aiming at the key product characteristics(KPC) identification, it establishes an expanding derivative coordination graph which comprehensively reflects parts, geometric characteristics, coordination constraint and dimension relation, and its matrix expression. By adopting the search algorithm in graph theory and Taguchi quality loss theory, the KPC alternative characteristics building and its effect evaluation method are proposed; according to the key control characteristics(KCC) recognition, internal relevance between coordination constraints and KCC identification, and its polychromatic sets contour matrix expression are analyzed; the KCC reasoning identification method is presented.3) As there are difficulties in assembling deviation analysis such as multi-deviations,large distribution and nonlinearity, “Kriging-QMC” two-stage analysis method is put forward.First of all, sum it up to be two key problems including transferring coordination constraint to positioning constraint and parts positioning model solving; then, establish the theory method of location selection based on the geometric stability as the evaluation index, and analyze the positioning coupling deviation computing method in the plane-plane cooperation constraint and hole-pin cooperation constraint; finally, based on Kriging’s method, an approximate relationship between the positioning coupling deviation and parts pose deviation is established, making the assembling deviation analysis process change from the “implicit” to“explicit” and reducing the degree of nonlinearity; replacing the random sampling in the Monte Carlo method with low discrepancy sequence sampling, it proposes parts pose parameter statistical analysis method based on quasi-Monte Carlo method, which can improve the convergence efficiency of assembling deviation statistical analysis.4) A co-allocation model of assembly tolerance together with a solution is brought up with regard to the difficulty of massive non-linear optimization in the distribution of distribution of assembly tolerance. First, a framework of co-allocation model of assembly tolerance is established based on the non-linear relational views of the weighting graph-tree model to divide the problem into series of sub-problems. Then, data stream in the allocation of assembly tolerance is analyzed and a co-allocation model is built accordingly. Finally, a coordinating strategy of co-allocation mode and a calculation of the sub-problems on account of the genetic algorithm are set up on the basis of Augmented Lagrange formula.Taking the boarding gate of a certain type of airplane as an example, the thesis has realized the construction of weighting graph-tree model, the quantized identification of KPC,the reasoning identification of KCC, the analysis of assembly tolerance as well as its allocation. The result indicates the analysis of non-linear assembly deviation and the efficiency of co-allocation of assembly tolerance.