Search for a command to run...
Wind-induced snow can significantly alter the geometric features and flow patterns around bridge structures, thereby affecting their aerodynamic performance. This study systematically investigates the aerodynamic characteristics and flow dynamics of a flat box girder under varying snow depths, utilizing wind tunnel experiments in conjunction with large eddy simulation within a Reynolds number range of 2.1 × 104 to 5.5 × 104. Three distinct flow regimes are identified as snow depth increases: a leakage-dominated mode at h0 = 0 mm, a Kelvin–Helmholtz vortex-dominated mode at h0 = 3.6 and 7.2 mm (h0 represents the snow depth in open, flat areas at the bridge site), and a Kármán vortex street-dominated mode at h0 = 10.8 and 14.4 mm. Correspondingly, the mean drag coefficient increases significantly, while the mean lift and moment coefficients gradually decrease. These changes are attributed to the substantial reduction in the railing ventilation rate caused by snow accumulation, which alters local flow separation characteristics. These flow regime classifications are further supported by power spectral density (PSD) and proper orthogonal decomposition analyses. For h0 = 0 mm, no distinct peak is observed in the PSD; a broad peak appears at h0 = 3.6 and 7.2 mm, while a pronounced single peak emerges at h0 = 10.8 and 14.4 mm. Additionally, the Strouhal number decreases with increasing snow depth, indicating a lower vortex shedding frequency but enhanced periodicity. On the lower surface, the mean pressure coefficient gradually decreases, while fluctuating pressure is significantly amplified. In contrast, on the upper surface, pronounced pressure fluctuations shift toward the trailing edge. These findings provide valuable insights for the aerodynamic performance assessment and design of bridges in snowy environments.