The bolster is a key structure for the metro train. To evaluate its fatigue reliability under the fatigue loading scheme, the full-scale fatigue test and theoretical calculation are performed. The bolster is modeled as a series system of stress concentration zones, namely weak points, and their stresses are calculated using validated finite element method (FEM). Since no fatigue crack is observed after the test and the fatigue life predicted by the Goodman diagram is more than [Formula: see text] load cycles, the bolster can service for more than 30 years with 90% reliability. To learn more about its fatigue reliability during the real service, a fatigue reliability model is established based on stress-life interference model. 10 load spectra are generated by Poisson processes, and the bolster is also modelled as a series system of six weak points. As a result, its cumulative failure probability is less than 0.002% within its 30-year service under normal atmospheric conditions.

The structural safety of a subway bolster is studied by fatigue strength testing and numerical simulation. A novel approach is proposed to determine the loading scheme in the fatigue test. This method identifies the loads accurately and allows for efficient yet economical fatigue test. Full-scale fatigue test of the bolster is carried out up to 1 × 10 7 load cycles under normal atmospheric conditions. The finite element model is validated based on the stresses measured by strain gauges. The fatigue life of the bolster is evaluated by Goodman diagram and nondestructive test since it is modeled as a series system of weak points. The results demonstrate that this bolster satisfies the structural safety requirement and its fatigue life is more than 1 × 10 7 load cycles with 90% reliability.