Influence of train–bridge aerodynamic interference on train running performance under bridge vortex-induced vibration

Separated box girders are widely used in long-span bridges but remain susceptible to vortex-induced vibration (VIV). Trains can affect bridge VIV by altering aerodynamics and vortex shedding, while bridge VIV may influence train responses. Using a separated triple-box girder road–rail cable-stayed bridge, this study combines a fluid–structure interaction (FSI) framework with wind–vehicle–track–bridge coupled dynamic analysis to investigate VIV responses, train aerodynamics, and running performance under different train-placement cases. The results show that, in the present simulations, trains widen the VIV lock-in range and increase the vibration amplitude. Among the considered train-placement cases, the two-train case gives the largest response, with the maximum reduced vertical amplitude increasing by 36.2% relative to the bare-girder case. Train–bridge aerodynamic interference makes train aerodynamic loads sensitive to train placement, and the most pronounced asymmetry is observed in the two-train case: the shielding effect reduces the lift and drag coefficients of the leeward-track train, whereas those of the windward-track train become slightly larger. The coupled dynamic analysis shows that, after train–bridge aerodynamic interference is taken into account, the examined train response indicators increase; in particular, for the windward-track train in the two-train case, the peak carbody vertical acceleration increases from 1.132 m/s 2 to 1.282 m/s 2 . When train aerodynamic-force fluctuation time histories under VIV are further considered, the carbody vertical acceleration changes only slightly, whereas the peak value and stochastic fluctuation level of the wheel load reduction rate increase markedly; in the two-train case, the peak wheel load reduction rate of the windward-track train increases from 0.435 to 0.552. These findings suggest that both the bridge VIV-induced displacement excitation and the aerodynamic-force fluctuation time histories of the train under VIV should be considered when evaluating the wheel load reduction rate responses of trains running on long-span road–rail bridges under VIV conditions.