Effects of submerged vegetation on flow structure and local scour around bridge abutments under ice-covered flow conditions

In cold regions, ice cover often appears on the water surface of rivers, posing significant challenges for hydraulic engineering and infrastructure safety, particularly around bridge piers and abutments. This study aims to understand the complex interactions between ice cover on water surface, submerged vegetation in channel bed, and hydrodynamics of flow around the abutment in channels. Through a series of laboratory experiments in a large-scale flume, the research investigates key hydraulic and scour parameters under various environmental conditions. The research examines the effects of submerged vegetation and ice cover on flow structures around bridge abutments, considering different vegetation densities and arrangements (square and staggered patterns) and varying ice cover roughness (smooth and rough). Key turbulence parameters, including turbulence intensity, Reynolds shear stress (RSS), and turbulent kinetic energy (TKE), were analyzed. Findings reveal that under ice-covered conditions, velocity profiles change from an S-shaped curve observed in open flow to a convex shape, indicating significant modifications in flow dynamics. The presence of a rough ice cover and vegetation led to unpredictable turbulence patterns and an expanded area of negative RSS downstream of the abutments. Moreover, TKE levels were substantially lower under ice-covered flow conditions compared to those of open flow scenarios. The study also investigates hydraulic resistance, a vital factor in river engineering that affects flow dynamics and sediment transport. The combined effects of submerged vegetation, ice cover, and bed sediment on hydraulic resistance were assessed by calculating bed and ice cover shear stress, vegetative drag, and the composite Manning’s roughness coefficient. Results demonstrate that the presence of an ice cover on water surface significantly increases the bed shear stress, particularly in vegetated beds under ice-covered flow, where it can contribute up to 60% of the total shear stress. Vegetation arranged in staggered patterns reduced bed shear stress but increased vegetative drag. Overall, the drag force was found to exceed shear force in all scenarios, comprising 85% of total resistance. The results of this study also showed a clear inverse relationship between Manning's roughness coefficient and flow Froude number, emphasizing the complex interplay between flow and resistance factors. To address the critical issue of local scour around bridge abutments, effects of various vegetation configurations, densities, ice cover conditions, abutment shapes, and sediment sizes on scour depth was also investigated. The research employed advanced machine learning models—Artificial Neural Networks (ANN), Support Vector Machines (SVM), Multiple Linear Regression (MLR), and Gene Expression Programming (GEP)—to predict the maximum scour depths around abutments. The GEP model outperformed others in accuracy, effectively capturing nonlinear relationships among variables. Sensitivity analysis using the Partial Mutual Information (PMI) and SHAP (SHapley Additive exPlanations) identified key variables impacting scour depth, including flow Froude number, sediment characteristics, vegetation density, and ice cover roughness. Two predictive formulas for the maximum scour depths were developed, tailored to different bed materials and abutment shapes. This study highlights the importance of accounting for the combined effects of ice cover and vegetation characteristics in hydraulic design to reduce scour risks and improve bridge stability. The integration of machine learning models demonstrated a significant potential in accurately predicting critical parameters, providing a robust tool for enhancing the safety and reliability of bridge designs under diverse environmental conditions.