Comprehensive dynamic models of railway wheelsets and tracks for the prediction of rolling noise
- Autores: Víctor Tomás Andrés Ruiz
- Directores de la Tesis: José Martínez Casas (dir. tes.), Francisco David Denia Guzmán (dir. tes.)
- Lectura: En la Universitat Politècnica de València ( España ) en 2024
- Idioma: inglés
- Tribunal Calificador de la Tesis: Francisco Javier Fuenmayor Fernández (presid.), Publio Pintado Sanjuán (secret.), Jorge Patricio Arenas Bermudez (voc.)
- Programa de doctorado: Programa de Doctorado en Ingeniería y Producción Industrial por la Universitat Politècnica de València
- Materias:
- Enlaces
- Tesis en acceso abierto en: RiuNet
- Resumen
Rolling noise emission in railway systems presents significant challenges in urban and suburban environments, requiring a comprehensive understanding of its underlying mechanisms for effective mitigation. This Thesis offers a thorough investigation into rolling noise generation, considering both tangent and curved tracks, through the development of advanced high-frequency models and calculation tools to predict noise levels accurately.
Rolling noise arises from the interaction between railway wheels and tracks in the presence of roughness on their contact surfaces. The principal components contributing to the acoustic emission are the sleepers, rails, and wheels. Sleepers and rails are interconnected through rail pads, while wheels and rails are dynamically coupled due to their contact. Consequently, changes in any component might impact the dynamic and acoustic behaviour of all of them, outlining the necessity of a comprehensive model to address rolling noise radiation effectively.
Various models for the track and wheelset are explored and developed in this Thesis. The track, assumed to be infinite, is described using periodic structure theory and it is characterised by its wavenumbers and waveshapes. The wheelset is modelled through the Finite Element Method (FEM) and it is characterised by its natural frequencies and vibration modes. The wheelset and track interaction is described as a linearised relationship between the relative motion of both components and the contact forces. The models examined in this work are formulated in the frequency domain.
Identified track properties influencing rolling noise radiation are quantified using statistical techniques. While the rail geometry is found to have a limited impact on the total radiation, the viscoelastic properties of the track, particularly the rail pad stiffness, play a crucial role in noise generation. Variations up to 7.4 dB(A) were observed for different track designs.
Special attention is devoted to the wheelset modelling in this Thesis. Its rotation is considered using Eulerian coordinates, a convenient approach as the wheel contact point with the rail remains at a constant spatial position. Given the axial symmetry of its geometry, the wheelset dynamic response is expanded around the circumferential direction using Fourier series, which yields a two-dimensional (2D) formulation of the dynamic and acoustic behaviour of this three-dimensional (3D) system, with the circumferential coordinate being solved analytically. This methodology, denoted as axisymmetric approach, offers a significant reduction in the associated computational calculation time while preserving accuracy, making the model well suited for its integration into optimisation algorithms.
Lastly, a novel investigation into rolling noise when the vehicle negotiates a curve is conducted. While curved tracks are generally associated with squeal noise, this Thesis offers valuable insights into the importance of rolling noise as well. To achieve this, the inertial and gyroscopic effects associated with a wheelset running on a curve are modelled. Complex phenomena occurring at the wheel/rail contact, such as the relative motion between these two elements, are incorporated into the interaction model. The results indicate that the wheel/rail contact position serves as a reliable indicator of the impact of a curve on the rolling noise.
