Generation and Propagation of vibrations induced by high-speed railways João Manso Abstract In the years to come, Portugal has several challenges to overcome and one is to try to modernize its train network. Among other issues, vibrations are a major concern, specially due to high speed traffic. Therefore, a two-dimensional model has been designed, considering some hypotheses (straight tracks, constant longitudinal geometric and mechanic characteristics), to predict both generation and propagation of those vibrations. Although the phenomena is three-dimensional, a two-dimensional model is less time consuming and easier to implement, but there are some aspects to take into consideration, like, for example, geometrical soil damping. To overcome and comparing them to in situ values, a method that corrects the results has been proposed. To compare the in situ results with the results from the model, there are several filters that could be used to reduce random noise from a signal, while retaining a sharp step response, but, for time domain analysis, moving average filters are the premier filter to use. Both the train himself (weights and geometry) and track irregularities, are the most relevant vibration sources from high speed traffic. So, taking into consideration the train weight, the distances between axles and a stochastic process to model the rail roughness, those actions could be predicted. There are some conclusions, in this work, that should me mentioned referring to two-dimensional model results. One is that they don't consider geometrical damping correctly, so, surface wave vibrations should be corrected by a factor. Keywords High-speed trains, axle loads, track irregularities, vibrations, geometric and material damping, two-dimension numerical model Introduction The development of tracks of high-speed trains has increased in the last decades in countries of Europe, North America and Asia. Trains are now able to achieve velocities of 300 km/h and higher, so its consequences to people and structures became a major concern. Numerical models and in situ and laboratory tests can be made to study some of those consequences and to test new methods to decrease them. 1