Calibration and validation of fatigue design models for railway car bodies considering uncertainty

Abstract

The fatigue design of railway vehicles is a requirement for resource- and energy-efficient lightweight design. A simulation-based design process makes high demands on the reliability of results. Therefore, this work focuses on the calibration of weld quality parameters for the simulation-based fatigue design of railway car bodies. Since the direct measurement of weld parameters is costly, an indirect Bayesian calibration approach is implemented, effectively combining limited testing with numerical simulations. Fatigue tests have been performed on the probe, component, and structure level for the load-bearing section of the Next Generation Train (NGT) car body within the SimbaCon (Simulation-based Conception) project at DLR. Based on the distribution of the measured life spans for component tests, the corresponding weld seam quality factors are calibrated using Bayesian calibration. The resulting weld seam qualities are used for the fatigue assessment of the complex car body structure.