几何物理特性深度融合的装配误差传递机理及数字孪生模型构建与实现方法

The variation propagation analysis method based on ideal feature and rigid body hypothesis has been difficult to meet the increasing performance requirements of precision assembly process. Taking the rapid development of digital measurement and parallel computing as an opportunity, this project puts forward the ideal of using digital twin model with deep fusion of geometric and physical characteristics to accurate prediction of variation accumulation in complex and precision mechanical systems. This project explores the geometric error modeling method from tolerance design and geometric measurement. The three-dimensional measurement, three-dimensional point cloud (geometry), smooth particle (physics), error correction technology are integrated in this modeling method. Aiming at the typical assembly matching situations such as fixed connection, interference fit and clearance fit, this project proposes a digital twin model of variation propagation and a full dimensional chain model based on the fusion of geometric and physical aspects in the assembly process. Parallel computing technology is introduced for calculating small size features of assembly variation propagation considering geometric and physical characteristics to improve the computational efficiency. The model parameters are verified and modified by typical mating feature test to improve the prediction accuracy. This project is a deep investigation of digital and intelligent manufacturing technology in product assembly accuracy assurance, which provides theoretical support and technical realization scheme for promoting the application of digital twin technology.

机械系统装配性能不断提升，基于理想特征与刚体假设的误差传递分析方法已难以满足精量化装配要求。本项目以数字化测量与并行计算快速发展为契机，创新地提出了几何物理特性深度融合的数字孪生模型解决复杂、精密机械系统装配误差准确预测的构想。探究衔接设计与测量的几何误差表征方法，打通三维测量-三维点云（几何）-光滑粒子（物理）-误差修正技术路线。针对固定连接、过盈配合与间隙配合等典型配合特征，构建基于光滑粒子的装配过程几何物理要素融合的误差传递数字孪生模型与全尺寸链分析模型，揭示考虑形面变形和接触不确定性的装配误差传递机理，研究微小尺寸特征的几何与物理特性并行计算技术，提升复杂仿真模型计算效率。通过典型配合特征试验验证与修正模型参数，提高预测精度。本项目为数字化、智能化制造技术在产品装配精度保障中的深入结合，为推动数字孪生技术应用提供理论支撑与技术实现方案。

针对复杂、精密机械系统装配误差精准、快速预测目标需求，项目研究突破了目前装配精度预测中采用理想特征与刚体的假设。从微小尺寸下装配表面形貌与几何误差模型构建、综合几何与变形物理特性的配合特征误差传递模型构建、数字孪生模型构建及数值仿真实现与试验验证等几个方面开展研究工作，解决了机械系统装配数字孪生中几何误差微小尺寸的表征、模型构建、分析等关键技术问题，实现了针对生产现场实际的装配体几何误差分析数字孪生模型构建。项目关键技术为装配数字化、智能化、精量化提供了核心算法与使能技术，将进一步促进产品质量与生产效率提升，为数字孪生与生产实际融合，数字孪生思想与具体技术落地探索了一条可行的技术路线。