Simulation of vertical dynamic vehicle-track interaction using a three-dimensional slab track model
Simulation of vertical dynamic vehicle-track interaction using a three-dimensional slab track model
复制标题
DOI：
10.1016/j.engstruct.2020.110972
复制
发表时间：
2020-11-01
影响因子：
5.5
通讯作者：
Nielsen, Jens C. O.
中科院分区：
文献类型：
--
作者：
Aggestam, Emil;Nielsen, Jens C. O.
在改进轨道设计时，需要更好地了解轨道的损坏模式，然后铁路行业依赖于动态车辆-轨道相互作用的精确模拟的可用性。本文采用扩展的状态空间矢量法，在时域内模拟了运行中的车辆与板式轨道之间的垂向动力相互作用。一个三维的板式轨道模型，推出的轨道建模使用瑞利-Eschlishenko梁元素和混凝土面板和路基表示使用壳或固体有限元。从使用Python脚本在Abaqus中开发的参数化轨道模型中，系统矩阵被导出到Matlab中，在Matlab中执行动态车辆-轨道相互作用的仿真。采用复值模态叠加技术，降低了仿真的计算成本。在后处理步骤中，从动态分析中计算出的轮轨接触力被用作Abaqus轨道模型的输入，在该模型中评估各种轨道响应。特别是，在混凝土面板的关键位置确定的主应力的时间历程。此外，车辆的速度对轮轨接触力的影响，以及轨道下方的横向涵洞（模拟为基础垫层模量的局部增加）对轨道刚度变化的影响，在轨道水平，进行了研究。已进行了一系列轨道响应的网格收敛研究，包括何时使用线性或二次元素以及壳或实体元素的调查。最后，所提出的三维模型进行了比较，一个替代的二维模型，以确定在什么情况下，二维模型是足够的。
When improving track design, a better understanding of the track's damage modes is needed, and the railway industry is then dependent on the availability of accurate simulations of the dynamic vehicle-track interaction. In the present study, the vertical dynamic interaction between a travelling railway vehicle and a slab track is simulated in the time domain by using an extended state-space vector approach. A three-dimensional slab track model is launched where the rails are modelled using Rayleigh-Timoshenko beam elements and the concrete panel and roadbed are represented by using either shell or solid finite elements. From the parameterised track model, which is developed in Abaqus using Python scripts, the system matrices are exported to Matlab where the simulation of the dynamic vehicle-track interaction is performed. A complex-valued modal superposition technique is employed, which reduces the computational cost of the simulation. In a post-processing step, calculated wheel-rail contact forces from the dynamic analysis are used as input to the Abaqus track model where various track responses are evaluated. In particular, the time history of principal stresses is determined at critical locations in the concrete panel. Also the influence of the speed of the vehicle on the wheel-rail contact forces, and the influence of a transverse culvert below the track (modelled as a local increase of the foundation bedding modulus) on the track stiffness variation at the rail level, are investigated. A mesh convergence study for a range of track responses has been conducted including investigations of when to use linear or quadratic elements and shell or solid elements. Finally, the presented three-dimensional models have been compared to an alternative two-dimensional model to determine in what situations a two-dimensional model is sufficient.