Utmattelsesanalyse av ståljernbanebro ved bruk av FEM-modellering

Abstract

The Norwegian railway network features numerous steel bridges that were constructed over 60 years ago. Progress in technology and an increasing population have resulted in different and larger traffic conditions than the conditions when they were designed. It is therefore vital to evaluate the fatigue damage of the railway bridges to determine which of them are at the highest risk and need repairing or replacement to avoid possible collapse of the structures.

The aim of the study was to investigate how sensors and FEM-models can be used to improve the assessment of fatigue damage on old railway steel bridges. A literature review was conducted together with and experiment. In order to make an assessment of the most critical components, a FEM-model was constructed based on an existing riveted Warren truss bridge. To find the historical fatigue damage, a conservative and consistent load model from NTNU was used in the calculations. The existing railway bridge was instrumented with strain gauges and accelerometers to provide sensordata for the train passages. The structural response from an NSB-73 train passage was compared between the FEM-model and the strain gauges.

The results show that several elements of the structure exceed the critical damage value, indicating that the components have no remaining fatigue life. The study demonstrates that it is possible to improve fatigue analysis by combining FEM-models and sensor data.