The research presented in this thesis has identified and developed a sophisticated computer model for the analysis of railway track dynamic behaviour to be used by the Rail Cooperative Research Centre for Railway Engineering and Technologies (Rail CRC) in Australia. To be competitive railway track owners need to extract as much performance as possible from their asset without serviceability or catastrophic failure. Railway track designers therefore need to develop more knowledge of the static and dynamic loadings that track may be subjected to in its lifetime. This would be best undertaken using computer modelling capable of quantifying the effects of train speed, traffic mix, wheel impact loading and distribution of vehicle loads into the track. A comprehensive set of criteria for the selection of a model of track dynamic behaviour was developed. An international review of state-of-the-art models which represented the railway track structure under the loading of a passing train was undertaken. The models' capabilities were assessed and a number of potential models identified. A benchmark test was initiated to compare current models available throughout the international railway research community. This unique benchmark test engaged six researchers to compare their railway track models using a set of theoretical vehicle and track data. The benchmark results showed that significantly different results may be obtained by models, depending on the assumptions of the user in representing a particular track scenario. Differing complexities and modelling methods, the number of different input parameters required and the representation of the irregularities in the wheel and rail all have effect on the results produced. As a result of these initiatives, the DARTS (Dynamic Analysis of Rail Track Structures) computer model was chosen for use by the Rail CRC. A user-friendly interface was created for DARTS by the writer, which was readily interpretable by railway design engineers. At the time of writing, DARTS was found to be suitable for detailed investigations planned by the Rail CRC for future research and was provided for use through an Intellectual Property agreement with its author.

Orientador: Cassio Eduardo Lima de Paiva / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Civil, Arquitetura e Urbanismo / Made available in DSpace on 2018-08-24T08:58:58Z (GMT). No. of bitstreams: 1 Watanabe_JulianaSilva_M.pdf: 2195811 bytes, checksum: 47a972a34bbdd51a602e9d71d473ff40 (MD5) Previous issue date: 2013 / Resumo: As ferrovias de carga pesada (heavy haul) têm como objetivo, trabalhar com grandes volumes de carga, mantendo uma velocidade que possibilite um tempo de percurso cada vez menor, mas esse deslocamento deve ser feito com segurança e para isso é necessário que a via permanente esteja sempre em condições aceitáveis de circulação. No Brasil, a estrutura atual desse tipo de via férrea atualmente está dimensionada para suportar aproximadamente até 32 toneladas por eixo, mas para favorecer o maior escoamento e tornar a ferrovia mais eficiente intenciona-se chegar a 40 toneladas por eixo. Baseado nessas informações, o objetivo deste trabalho foi analisar uma solução técnica de via férrea adequada para uma carga de 40 toneladas por eixo. Foram avaliadas duas situações: na primeira utilizou-se uma camada de lastro apoiada diretamente sobre a plataforma; e na segunda foi adotada uma placa de concreto de cimento Portland entre a camada de lastro e a plataforma da via. Para a primeira situação foram calculadas as tensões no trilho e na plataforma através dos modelos numéricos propostos por Eisenmann e Zimmermann, e na situação da placa de concreto, dimensionou-se uma estrutura baseada no método de dimensionamento de pisos industriais da Associação Brasileira de Cimento Portland. Para os dois casos, as tensões atuantes na via permanente foram verificadas através do programa computacional FERROVIA. As conclusões foram obtidas através de comparações entre os valores das tensões calculados pelos métodos teóricos e pelo método computacional / Abstract: The heavy haul lines objective is to carry elevated freight, maintaining speeds that assure a shorter travel time. It must be done safely and, for this, it is required good conditions for the track. This Brazilian railway structure is designed to support 32 tonnes per axle, however to promote a better flow and to become more efficient there is an intention of change this freight to 40 tonnes per axle. Based on that, the aim of this study was to analyze a suitable technical solution t for 40 t/ axle load. Two situations have been evaluated: in one of them it was used a ballast layer over the formation; and in the other one it was adopted a concrete slab between the ballast layer and the formation. For the first situation stresses on rail and formation were calculated based on the theoretical models (Zimmermann and Eisenmann methods), and for the second one, a structure similar to a concrete pavement has been designed (ABCP industrial floor method). For both cases, the tresses on the track have been verified by the computational program FERROVIA. The conclusions were based on comparisons between the stresses values calculated by the theoretical and computational methods / Mestrado / Transportes / Mestra em Engenharia Civil

Orientador: Auteliano Antunes dos Santos Junior / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Mecânica / Made available in DSpace on 2018-08-16T12:15:36Z (GMT). No. of bitstreams: 1 Baruffaldi_LeonardoBartalini_M.pdf: 4986344 bytes, checksum: 359223cc9295f5b392b883e2841d037a (MD5) Previous issue date: 2010 / Resumo: Por 150 anos, os truques de três peças têm sido a estrutura padrão para o suporte de eixos e suspensões de trens de carga em diversos países. Embora sua robustez e facilidade de manutenção tenham conservado, em linhas gerais, a disposição e projeto dos componentes, novos requerimentos de carga e velocidade dos trens vêm mudando a forma como os projetistas enxergam o truque. Especial atenção tem sido dada ao modelo matemático da cunha de atrito, a peça fundida que é responsável pelo amortecimento dos vagões. A cunha de atrito promove a dissipação da energia mecânica por meio de contato de atrito seco com outros componentes do vagão. Devido às altas forças normais desenvolvidas nas superfícies de contato com características não suaves e, em geral, não lineares de atrito, as equações que regem o movimento da suspensão tornam-se de resolução difícil e surgem fenômenos como o de adesão escorregamento e o comportamento caótico típico de osciladores auto excitados. O presente trabalho tem como objetivo propor o uso de algoritmos de solução de problemas de complementaridade linear para resolver as forças de contato entre os corpos, visando a aprofundar a discussão sobre os modelos adotados para a cunha de atrito. Os resultados obtidos mostram que é possível modelar as forças de contato desse sistema utilizando um problema de complementaridade linear e que essa abordagem é, sob certas condições, mais eficiente do que o método das penalidades, normalmente aplicado para a resolução de problemas de contato / Abstract: For about 150 years now, three-piece trucks have been the standard axis' and suspensions' subframe used in freight railroad cars. The toughness and low maintenance costs of this system worked to maintain its basic design almost unchanged, but new requirements for loads and speed for freight cars are changing the way designers see the three-piece truck. Among the many interesting components of the three-piece truck, the friction wedge is getting some attention. The friction wedge is the main damping element in three-piece trucks and acts to dissipate mechanical energy via highly stiff contacts with friction. Due to the non-smooth and non-linear nature of frictional efforts, the equations of motion of the three-piece trucks become very awkward to deal with. Interesting phenomena of stick-slip, bifurcations, and limit cycle, typical of friction oscillators appear to some extent under normal operation. This work's main objective is to propose a new approach based on complementarity problems, used to solve for contact forces, to further extend the discussion on wedge dampers models. Results show that it is possible to model the problem using the linear complementarity problem and that, in some situations, this can be even more computationally efficient than the usual approach to solve contact problems: the penalty method / Mestrado / Mecanica dos Sólidos e Projeto Mecanico / Mestre em Engenharia Mecânica