Research On Precision Assembly Multi-process Oriented Process Quality Evaluation Method

Assembly is an important stage in the forming of the quality of precision products.The evaluation and control of the process quality of precision assembly can greatly improve the manufacturing quality and the performance of precision products.Precision assembly usually involves multiple processes,which makes the quality forming of multi-process precision assembly is featured by various influencing factors and complicated mechanism,and the process quality evaluation more difficult.It can be mainly reflected in three aspects.Firstly,it is quite difficult to comprehensively quantify various random factors influencing the assembly process.Secondly,it is difficult to describe the errors of form and position in the characteristic section containing a lot of micro-morphology information in a unified manner.Thirdly,it is quite difficult to deal with the error accumulation and complicated transmission mechanism in multiple processes.Aimed at those problems above,this paper has proposed the method of process quality evaluation for multi-process precision assembly based on relevant research projects.It can be elaborated as follows.1.An error model for single-process precision assembly has been built.The main error sources of precision assembly process are identified,including the error source of technological equipment,the error source of parts deviation and the error source of parts contact assembly.Moreover,how different error sources work in the process has been also researched.As to the error source of parts deviation,the method of evaluating the deviation error in the complicated characteristic section has been established.The point-cloud data of the precision surface has been measured and obtained.The ICP algorithm for the point-cloud registration has been adopted.Moreover,the method of evaluating the deviation error of complicated characteristic section has been also established.The method consists of three steps,namely preliminary registration,characteristic section fitting and point-cloud restructuring as well as secondary registration.It has been verified with the real cases for the deviation error evaluation of complicated characteristic section of blade parts.As to the error source of parts contact assembly,the method of evaluating the error of characteristic section contact assembly has been built.After elaborating on the features of spatial statistics on the characteristic section,this paper has innovatively utilized the variation function to describe the micro-morphology error on the surface.The model for contact assembly error on the local characteristic section has been built and the dimensions of global characteristic section have been also inferred.In addition,the method of calculating the contact assembly error of general characteristic section has been given.By means of surface simulation and simulated assembly experiment,the accuracy of calculation method has been inferred and also verified accordingly.2.The method of evaluating the multi-process precision assembly ability has been built.With the accumulation and transmission features of the errors between different processes taken into consideration,the single-process error model for precision assembly has been further applied to the multi-process dimension.Besides,the three indexes,namely the overall quality process ability,single quality process ability and former process influence,have been evaluated.The case of multi-process precision assembly for a certain optical component has been given to verify the effectiveness of the method.3.The method of evaluating the reliability of assembly equipment has been built.The data regarding the quality features of the product have been used to evaluate the reliability of technological equipment.Based on the multiple regression,the short-time assembly process model has been built.Based on the Markov chain,the long-time assembly process model has been also built.The formula for calculating the reliability of technological equipment has been proposed.As shown by the case of the assembly process for a certain part,the reliability evaluation method proposed in this paper is more accurate than the maximum likelihood method.