Research On Key Technologies Of Complex Product Assembly Accuracy Prediction Based On Digtial Twin

In recent years,as mechanical structure products continue to develop in the direction of higher quality,higher accuracy,higher performance,and higher integration complexity,the assembly requirements of complex products have become higher and higher,and the guarantee of assembly performance has also become more and more difficult.The assembly accuracy design is one of the important indicators to evaluate the assembly performance of complex products,and the related assembly accuracy analysis technology is an important means to ensure the product assembly performance.For the assembly accuracy analysis process of complex products,such as aerospace products and high-precision machine tools,it is necessary to consider not only the manufacturing error of parts and components,but also the error accumulation and transmission of the assembly process,and the actual assembly requirements and assembly process constraints.These factors will affect the accuracy of the complete product which make it unable to meet the design requirements and difficult to ensure the assembly quality and assembly performance.Therefore,before the implementation of the assembly,even during the implementation of the assembly process or even before the assembly is completed,the effective and accurate prediction of product assembly accuracy is the basic prerequisite for guaranteeing and improving the assembly performance of complex products.In this context,by making full use of the dual advantages of digital assembly technology and digital twin technology,the common basic key technical problems of complex product assembly accuracy prediction based on digital twin are deeply studied in this paper.It lays the foundation for theoretical methods,technologies and applications to implement the high-quality,highprecision and high-performance assembly of complex products.The dissertation is organized as follows:(1)The model expression and information management methods of digital twin-based assembly process model(DT-APM)for complex products are proposed.Combing the digital twin technology with the assembly process design for complex products,the overall implementation framework of DT-APM is presented,and the organization structure and formal expression of DT-APM are discussed in detail.Then,the model expression of DT-APM based on model-based definition(MBD)is realized through the object-related coupling mechanism,and a MBD-based expression model for DT-APM is established for different activity stages of assembly process.On this basis,the evolution mechanism and construction strategy of DTAPM for complex products are proposed.It provides the required “model-information-data”basis for the assembly accuracy prediction of digital twin-based assembly.(2)A novel modeling method of non-ideal surface model for digital twin is proposed.By introducing new-generation geometrical product specifications(GPS)and the idea of digital twin,a complete scheme for geometrical reference representation of digital twin-based assembly object model(i.e.,part digital twin model)is constructed using non-ideal surface model,and it is used for the representation of part geometric tolerance by the concept of skin model shape.Meanwhile,from the perspective of consistent expression of new-generation GPS standard chain,the model representation and creation methods of part skin model shapes based on specification and verification are respectively given,and the realization process of two methods are discussed in details.On this basis,the combination and correction methods of part skin model shapes are introduced,so as to realize the generation of the non-ideal surface model based on skin model shapes,which lays the foundation for the high-fidelity generation of digital twin-based assembly model of complex products.(3)A pre-analysis method of assembly success rate based on digital twin-based surface model is proposed.Firstly,the constraint expression method of assembly positioning of digital twin-based surface model based on assembly contact oriented graph is studied.And then,two accurate solution methods of assembly positioning between assembly features based on the element registration adjustment and the constrained surface registration are proposed,respectively.Secondly,by introducing the assembly joint surface and considering the assembly location priority in the analysis process of assembly error transmission in serial and parallel assembly,the calculation method of actual error transfer properties of joint surface groups in serial and parallel assembly is constructed.The geometric error variation relationship and constraint conditions between assembly features based on digital twin-based surface model are analyzed under the constraint of assembly location priority,and the calculation procedure of assembly success rate based on the principle of non-interference in the assembly positioning is established.Finally,by the calculation of the Monte Carlo simulation,the accurate calculation of assembly success rate based on digital twin-based surface model is realized.(4)A new accuracy prediction method of digital twin-based assembly for complex products considering multi-dimensional error sources is proposed.By introducing digital twin technology into the assembly process of complex products,the overall process and implementation steps of the digital twin-based assembly accuracy prediction considering multidimensional error sources are proposed.And then,the part manufacturing error and assembly process error based on digital twin-based surface model are respectively constructed by using small displacement torsor,specification or verification generation method,finite element analysis and assembly positioning method,which realizes the comprehensive error modeling and representation of multi-dimensional error sources for assembly.Moreover,the update and iteration mechanism of assembly error transmission based on dynamic dimensional chain and stream-of-variation theory is elaborated,and the assembly error calculation based on the Jacobian-torsor model and digital twin-based surface model is proposed,which realizes the digital twin-based assembly accuracy prediction considering multi-dimensional error sources.Finally,the accuracy of the proposed method is verified by the comparative analysis of case study.(5)An integrated demonstration system of assembly accuracy prediction for complex products based on digital twin is constructed.The experimental platform and application implementation process of the integrated demonstration system are introduced in detail.And then,various functional modules and actual application effects of the above-mentioned system are analyzed and verified,which provides a practical feasible solution for assembly accuracy prediction and assurance of complex products based on digital twin.