Influence of foreign objects on derailment risks in railway switches

Licentiate thesis, 2024

Railway switches are essential for ensuring efficient train operations, and efforts to improve their safety standards have driven the introduction of so-called switch rail control contact (TKK in Swedish) sensors in Sweden. These sensors monitor the gap between the switch and stock rail, complementing
the function of the drives. Such gaps can be a result of undetected foreign objects such as ballast stones or ice, that are stuck between the switch and
stock rail. According to statistics from the Swedish Transport Administration (Trafikverket), faults in TKKs due to component failure, snow and foreign object interference, account for a substantial portion of the total faults occurring in switches for a year. This results in increased maintenance and operational costs. Therefore, the aim of the thesis is to evaluate the necessity of these sensors by performing simulations of switch operations and derailment risk to provide input to a technical risk analysis of derailments in switches.

For assessing the derailment risk in scenarios of interfering foreign object, field measurements, finite element (FE) and multibody simulations (MBS) are
carried out. Two physical measurement campaigns were performed to obtain the deformation pattern of the switch rail in the presence of a foreign object. These measurements are used to validate the numerical models. The validated FE model of a switch is employed to perform parametric
studies on the influence of foreign object size and position on switching operations. The purpose is to identify in which cases foreign objects can be detected by the drives and when additional switch rail position sensors are needed to ensure that all objects leading to excessive gauge narrowing are detected. Results indicate that TKKs are necessary to detect all cases of gauge narrowing above 15 mm in the current switch design.

The dynamic vehicle–switch interaction with an interfering foreign object is modeled using non-linear flextrack module in MBS software Simpack. The
results of the simulations indicate that the risk of derailment is low when the foreign object is modeled with a failure criterion resembling ballast stone
crushing. When infinitely strong objects are stuck, the chances of wheel flange climbing are high. Ballast stones seem to not pose a derailment risk if caught between rails as they are crushed by high lateral forces.