Numerical and Experimental Fatigue Damage Evaluation of Railway Underframe Mounting Structures under Random Vibrational Loads

Abstract

The main objective of this study was to estimate the fatigue damage of railway underframe mounting structures subjected to under random vibration loading and to assess the feasibility of extending their operational life beyond the originally intended service life. These underframe-mounted railway structures are typically designed for a service life of 25 years. This study investigated whether this lifespan could be safely extended based on the predicted fatigue performance. Additionally, the study aimed to verify the accuracy of the Finite Element Analysis (FEA) predictions for fatigue damage under random vibration loading through experimental testing conducted in accordance with IEC 61373. Fatigue damage was assessed using both mean and 3σ S-N curves to account for statistical scatter in material behaviour. The use of the 3σ S–N curve resulted in higher cumulative fatigue damage estimates across all evaluation methods. First, a mode-superposition-based random vibration analysis was performed in the three principal train coordinate directions to determine the dynamic response of the suspension frame. Second, fatigue damage of the suspension frame was estimated using FEA based on the Steinberg method for all three loading directions. Finally, the FEA predictions were validated against experimental random vibration test results using the Steinberg method in the frequency domain, the rainflow counting method in the time domain, and the Dirlik method in the frequency domain. Goodman mean stress correction was applied to the rainflow counting results to evaluate the influence of mean stress. The effect of mean stress was found to be negligible under the investigated test conditions because of the absence of pre-stress, aerodynamic loads, and braking loads. Using the rainflow counting method as the reference, the Dirlik method produced closely matching fatigue damage estimates, whereas the Steinberg method yielded slightly more conservative predictions. The cumulative fatigue damage predicted by both FEA and experimental testing remained low, indicating that the operational life of the suspension frame may be extended beyond 25 years, provided that periodic structural inspections are conducted. This research method can be extended to estimate fatigue damage of other safety-critical underframe railway mounting structures.