Aerodynamic response characteristics of high-speed trains under turbulent crosswind using a train-adapted consistency-improved random flow generation inflow scheme

Turbulent crosswind produces hazardous low-frequency aerodynamic loads that threaten the lateral stability of high-speed trains (HSTs). Nevertheless, the underlying eddy evolution physics and St-dependent load-amplification mechanisms remain insufficiently quantified under conventional steady uniform inflow conditions. This work establishes the HST-consistency-improved random flow generation (CIRFG) synthetic turbulence framework by coupling the CIRFG algorithm with velocity superposition for HST turbulent crosswind simulations. Unsteady improved delayed-detached-eddy simulations are performed under uniform inflow and 15% turbulence intensity turbulent inflow conditions. The synthetic turbulence undergoes a two-stage streamwise evolution and reaches statistical equilibrium at (x/H ≈ −4) upstream of the train. Incoming turbulence exerts limited influence on time-averaged aerodynamic quantities, with most mean-coefficient deviations below 4% except for the Head Car drag (−11.3%) and Tail Car lateral force (−12.3%). Turbulent inflow substantially amplifies force and moment fluctuations: the Head Car lateral force, drag, and lift root mean square (RMS) rise by 208.6%, 717.1%, and 61.3%, while the corresponding Tail Car increments are 118.1%, 496.5%, and 40.8%. The total train roll and yaw moment RMS increase by 115.3% and 100.4%. Spectral coherence results show strong linear coupling between upstream turbulence and local aerodynamic forces of the head and tail cars within (0.3 < St < 1.5), whereas nearly no detectable coherence is found for integrated total train moments. Turbulence driven load amplification is confined to low-Strouhal-number bands. This study indicates that uniform inflow simulations underestimate low-frequency overturning relevant loads, and HST-CIRFG provides a feasible tool for HST turbulent crosswind research.