Assembly Accuracy Prediction Based-on CAD Model

CAT (Computer Aided Tolerancing) is the design and optimization on thetolerance of the product in the manufacturing within the use of computer technology,in order to achieve the precise control of the production cycle, manufacturing costsand product quality. Analysis of design tolerances and actual machining error of theproduct in the assembly design stage and product assembly, is to use computer-aided tools to predict the accuracy of product assembly, then the predictive value isused to evaluate the assembly quality and product performance, compared withassembly results to meet the design requirements and needs, in order to avoid themistake of the assembly program and reduce the number of the assembly adjustment.In the thesis, the structure of data of assembly information is established onanalysis of the characters of Pro/ENGINEER assembly model, and the method topick up information from CAD is developed. Then, on the base of the obtainedassembly data, the method of intelligent searching is designed to get correctassembly constraint relationship. With solving picking up dimension position andcalculation of other information, the automatic generation of assembly dimensionchain is completed.Taking accuracy calculation on assembly dimension chain as starting point,this paper comes up with3-D assembly dimension chain drawing. On synthesis ofdifferent tolerance analysis ways, some adjustment is made in the formulasaccording to3-D assembly dimension chain. At last, given the tolerance ofdimension, tolerance of shape and position, environment temperature and others thathave effects on assembly accuracy, the mathematical model of assembly accuracyprediction is established.Finally, based on Pro/ENGINEER the assembly accuracy prediction prototypesystem is developed with Microsoft Visual Studio. And experiments for theassembly precision are designed, the deviations between prediction accuracy in theCAD environment and measured accuracy, between anticipation accuracy anddesign requirement accuracy are within the allowable range, to meet therequirements of achieving the expected anticipation purpose.