Implementation of a comprehensive methodology for structural dynamics and NVH simulation incorporating MBS, FEM and ray tracing

To meet the fundamental requirements of environmental preservation, mitigating noise pollution stemming from railroad vehicles is imperative. Manufacturers and operators of rolling stock exhibit a keen interest in simulation tools tailored for calculating acoustic emissions from rails and vehicles. These tools not only furnish robust assistance and actionable insights for the early-stage development of resilient rails and vehicles but also offer cost-effective alternatives to prototype development and experimental endeavors. However, the paramount prerequisite for such a tool encompasses not only rapid computation but also the utmost accuracy achievable. Comprehensive acoustic simulation entails a highly intricate process, necessitating explicit simulation of Multi-Body Simulations (MBSs) alongside Finite Element Methods (FEMs) with millions of Degrees of Freedom (DOF). Such computations place substantial demands on computer performance. However, owing to significant advancements in computer capabilities, these once-complex models can now be readily computed on personal computers. This advancement in computational power offers an excellent assurance for numerical simulations of acoustics. In this thesis, a foundational model for calculating rolling noise is established. Initially, a comprehensive spatial vehicle-rail coupling model is developed to compute wheel-rail contact forces utilizing Timoshenko and Euler-Bernoulli beam theories, Multi-Body Dynamics (MBDs), and Hertzian nonlinear wheel-rail contact theory. This model not only corrects for the conformal contact of the wheel-rail to prevent discontinuities in contact points resulting from numerical calculations but also accounts for the influence of wheel and rail flexibility on the wheel-rail forces. Using this model, wheel-rail contact forces and Track Decay Rates (TDRs) are computed and validated through field experiments and in SIMPACK. The track parameters are predicted using an Artificial Neural Network (ANN). Subsequently, Equivalent Radiated Power Level (ERPL) of the wheel and rail under a unit force of 1N (also known as Noise Transfer Function (NTF)) as well as Panel Contribution Analysis (PCA) of the components are calculated in FEM models, providing the foundation for targeted acoustic optimization of the wheel and rail. Finally, a model of sound propagation outside the vehicle is constructed based on ray tracing theory. In this model, each wheel is treated as a blend of monopole and dipole sources, while each rail is regarded as a line source comprised of multiple monopole sources. The Sound Pressure Level (SPL) and the distribution of SPL on a reception plane, located at a distance of 7.5m, are computed. An analysis of the PCA of the rails revealed high sound radiation at the rail feet. Subsequently, the efficacy of reinforcing the UIC60 rail at this location in reducing radiated noise was explored. The results demonstrate a potential reduction in rail noise by 2.7 dB through this strategy. Additionally, the impact of rail pad stiffness and damping, as well as rail dampers, on rolling noise, was investigated. The results indicate that increasing rail pad damping can effectively compensate for the negative impact on acoustics resulting from the low stiffness of the rail pad. Moreover, employing suitable rail dampers proves to be an effective measure in reducing rolling noise. According to the calculations in this thesis, it is inferred that these two measures can reduce the noise generated by the 49E1 rail by 5.5 dB and 5.6 dB, respectively.

Author dc:creator

• Tian, Qiuyong

Advisor dc:contributor.advisor

• Hecht, Markus

Licence dc:rights.uri

• https://creativecommons.org/licenses/by/4.0/

Language dc:language.iso

en

Identifier URI

https://doi.org/10.14279/depositonce-22281

OAI identifier oai:identifier

oai:depositonce.tu-berlin.de:11303/23467