高速列车焊接构架抗疲劳断裂可视化设计软件研发

Fatigue and fracture are main reasons to lead to the failure of large-scale fundamental equipments. How to reliably predict and track the evolution feature of welded flaws is the key issue. Current design methods focus on the nominal stress method that cannot efficiently evaluate the amplification influence of welded cracks on fatigue damage, which is an essential reason to affect the running safety and lightweight design of MU trains. Based on self-developed ALOF structure and object/component oriented programming methods (VC++, UML, XML, MFC, VTK and Open CASCADE etc) together with the originally proposed polygonal element novel virtual node method (VNM) and improved double-interpolation extended finite element method (D-XFEM) suitable for fatigue cracking analysis by enhancing the shape function, the octree adaptive mesh multilayer refinement strategy, the fast successive over relaxation-preconditioned conjugate gradient (FastPCG), a novel visual fatigue life evaluation and fatigue cracking computational mechanical software with optimized preconditioning factor of large sparse linear systems and based-surface stress extrapolation method using triangle elements for high-speed railway welded frame as well as the tetrahedron-based polyhedral element virtual node method for the next generation bogie is therefore well designed and developed under the current safety standards and codes including the updated Association of American Railroads (AAR), the International Union of Railways (UIC), the International Institute of Welding (IIW), the JIS Railway Standards (JIS) and the China’s Railways Standards by adopting the fracture mechanics. By applying this safety evaluation system, engineers can efficiently, accurately and visually acquire the evolution characteristics of fatigue & fracture properties and residual fatigue for high-speed welded frame of bogies and other large-scale engineering structures under typical manufacturing flaws and loads, weld details and materials, which also provides the scientific support and theoretical basis for the round design, manufacturing and safe service.

重大装备服役的突出问题是疲劳断裂，能否可靠预测和追溯焊接缺陷的损伤演变是关键节点。当前转向架构架设计采用名义应力法，未考虑焊接缺陷致应力集中的放大性作用，是制约动车组安全服役及轻量化的根源。本项目引入断裂力学理论，在最新AAR、UIC、IIW、JIS和铁标下，改进网格分层加密算法、超松弛迭代预处理共轭梯度及考虑裂纹闭合和应力比的寿命模型，引入自主软件ALOF成熟框架和面向对象/构件的设计技术（VC++、UML、XML、MFC、VTK、OCC等），重点通过加强形函数开发高精度的单纯形网格D-XFEM开裂算法，探索矩阵迭代的最优预条件子实现百万级自由度求解规模，最终形成一套三维可视化抗疲劳断裂设计及仿真系统，确保能在典型制造缺陷和工况、焊缝细节及材料类型下高效准确评估焊接结构的疲劳断裂行为及剩余寿命，为我国高速列车焊接构架及其他大型工程结构的抗疲劳断裂设计和安全运行提供重要理论支撑和科学依据。

本项目针对重大装备服役的概率疲劳断裂这一突出问题，以准确、高效预测缺陷演化仿真计算的关键节点，开展了以下工作：阐明当前高速转向架构架基于名义应力的无限寿命设计方法的不足，提出把传统设计方法与基于断裂力学的损伤容限设计方法结合起来的两级服役行为评估技术，为此开发了高效高精度裂纹尖端断裂力学计算算法VP-XFEM和裂纹扩展速率模型iLAPS，集成构架用材料的基本力学、疲劳和断裂力学性能参数，统计焊接中可能的致命缺陷并开发专用接头细节菜单及兼容于ABAQUS软件等商业平台的通用数据接口，突破既有软件ALOF框架和设计技术，研制出高铁领域第一款损伤容限设计软件SinCrack。截至2019年底，发表正式期刊论文43篇（其中SCI检索24篇，单篇最高Google Scholar引用50次，并入选Eng Fract Mech近5年高被引论文）、会议论文15篇、会议报告33场次，国家专利授权10件、软件著作权3件，专著3部，部分成果获得2019年四川省科技进步二等奖（第一完成人），并直接应用于中国高速列车运维中。项目负责人入选第十二批四川省学术与技术带头人后备人选，唐立新优秀学者奖等荣誉称号，在国内外会议做特邀报告6次，受聘担任国际顶尖SCI期刊Int J Fatigue和Comput Sci Eng客座主编及《失效分析与预防》编委，受留学基金委全额资助赴英国曼彻斯特大学做高级访问学者半年。培养博士研究生4人、硕士研究生7人，在国内外铁路结构强度疲劳可靠性领域具有广泛影响力。在项目开展中，根据材料与结构的缺陷检测评定发展趋势，准确把握国际材料与力学研究前沿，开展实验力学理论研究，结合高速铁路技术、新材料技术和重大装备制造技术等国家重大需求开展工程应用研究，提出并在国内率先实践“基于同步辐射X射线成像的三维缺陷原位演化表征”这一全新研究方向，有助于为解决国家重大装备与工程中的关键力学问题提供新的思路和方法。