Investigation of dynamic responses of high-speed railway bridges under consideration of moving load models and multi-body models

The investigation of dynamic responses of high-speed railway bridges is increasingly important due to the expanding stock of high-speed trains. Developments of new trains lead to insufficient coverage of dynamic effects by standards on bridge design. In this dissertation, the dynamic responses of railway bridges are studied based on extensive sets of moving load models as well as 2D multi-body models of high-speed passenger trains. Moving load models allow a relatively fast evaluation of dynamic effects during bridge crossings and only require information of static axle loads and axle distances, which is available for a high number of trains. The moving load models of the currently operating high-speed passenger trains in the European network are used for the investigation of resonance excitations of bridges. The consideration of vehicle-bridge interaction effects requires a more detailed modeling of the trains using multi-body models with additional information about springs, dampers, and masses of the train. Multi-body models allow to take into account dynamic axle loads due to the movement of the train itself and therefore lead to calculation results closer to reality. More than 3,000 moving load models for passenger trains are assembled and railway bridge responses are investigated using simplified calculation methods. Based on the dynamic train signature envelope of this train catalog, a new dynamic load model for high-speed trains is developed aiming for only 20 model trains to represent the dynamic effects of all operating passenger trains. Additionally, a preselection procedure for dynamically relevant trains is developed based on the simplified calculation methods of train signature and train spectra. The resulting 500 dynamically relevant trains can be used for more complex computations of dynamic bridge responses. Vehicle-bridge interaction effects are recognized by comparing bridge responses of multi-body models, which cover interaction effects inherently, and moving load models, for which interaction effects can only be considered by an additional damping value which has to be calibrated. In calculations with moving loads the bridge damping ratio is increased by the additional damping value to account for the vehicle-bridge interaction. For the investigation of the vehicle-bridge interaction, multi-body models for more than 30 passenger trains (including conventional trains, trains with articulated bogies, and trains with equidistant wheelsets) are created. The algorithm for interaction calculations is implemented in MATLAB and the investigations are based on a parametric bridge set of about 100 single-span bridges. As only a small number of multi-body model parameters is publicly available and the correct representation of the dynamic train characteristics is uncertain, the eigenfrequencies of the theoretical multi-body models are validated based on measurements on stationary and moving trains. Beyond this, the influence of uncertainties in the data for stiffness and damping parameters in the multi-body models on the vehicle-bridge interaction is studied using a sensitivity analysis. The results of the comprehensive computational series of more than 1,000,000 bridge crossings are used to assess the validity of the additional damping method included in Eurocode 1 as well as further approaches for the additional damping given in literature. This dissertation provides a new dynamic load model of 20 model trains based on the dynamic signature envelope of more than 3,000 operating passenger trains. The investigations of vehicle-bridge interaction effects lead to the conclusion that special attention should be paid to the primary spring parameter in the definition of multi-body models and an additional damping value needs to be defined based on detailed train characteristics.