Effects of flood-induced transverse bed degradation on local scour around adjacent bridge piers using three-dimensional CFD modeling

Abstract Local scour around bridge piers can be strongly affected by flood-induced changes in channel-bed morphology, yet most existing studies assume an initially horizontal bed. This study investigates local scour around two side-by-side cylindrical piers ( $$D=0.04\hspace{0.17em}m, S=3D$$ ) subjected to a pre-existing localized transverse degradation flank using a three-dimensional FLOW-3D morphodynamic framework. The selected numerical configuration was evaluated against laboratory measurements of single-pier scour on a horizontal bed and reproduced the measured maximum scour depth within approximately 2%. Controlled parametric simulations were then conducted for transverse bed slopes of $$15^\circ , 30^\circ $$ and $$45^\circ $$ , flow depths of 0.30–0.60 m, approach velocities of $$1.5-2.5 m{s}^{-1}$$ , and different sediment conditions. At $$U=2.0\hspace{0.17em}m\hspace{0.17em}{s}^{-1}$$ , and $$H=0.45\hspace{0.17em}m$$ , the maximum quasi-equilibrium scour depth increased from 0.058 m for the flat-bed reference to 0.0982, 0.1065 and 0.2377 m for slopes of $$15^\circ , 30^\circ $$ and $$45^\circ $$ , respectively, corresponding to a maximum amplification ratio of 4.10. Over the same slope range, the scour-localization index increased from 1.08 to 2.84, indicating a transition from relatively distributed erosion to strongly concentrated scour at the trough-side monitoring location (G3). Once this critical zone developed under the steeper degraded-bed configurations, G3 remained the location of maximum scour across the investigated flow-depth and sediment-sensitivity series. On the flat-bed reference, increasing the approach velocity from 1.5 to $$2.5\hspace{0.17em}m\hspace{0.17em}{s}^{-1}$$ increased the maximum scour depth at all investigated flow depths, while the free-surface regime at $$H=0.45\hspace{0.17em}m$$ changed from subcritical to supercritical under live-bed conditions. The $$45^\circ $$ case is interpreted as an extreme transient morphology rather than a typical equilibrium slope, and the associated absolute scour depth is reported for the selected numerical configuration rather than as a mesh-verified design value. Supplementary reduced-order hydrodynamic diagnostics were consistent with increases and spatial concentration in near-bed forcing, sediment mobility, modelled turbulent shear and rotational activity, and are treated only as exploratory indicators. Overall, transverse bed morphology primarily controls where scour becomes concentrated, whereas hydraulic and sediment conditions predominantly control how strongly the critical zone responds. These findings highlight the need to account for post-flood transverse bed morphology in bridge-scour assessment.