Clear-Water Scour around Spill-Through Abutments in Compound Channels under Free, Pressure, and Overtopping Flows

Abstract High-intensity rainfall triggered by climate change can result in bridge submergence and overtopping, generating pressure flow through the bridge opening. Recently, these extreme events have become more frequent, and the resulting flow phenomena become more complex. In this study, a detailed experimental investigation has been conducted to develop a relationship for predicting the scour depth around a bridge abutment due to a flow contraction that produces lateral as well as vertical contraction scour in the case of submerged orifice flow (SO) and overtopping flow (OT) conditions. Experiments have been carried out on two different physical models of bridges with varying embankment lengths in compound channels at the Georgia Institute of Technology (GT) and the University of Auckland (UoA) under free, submerged orifice, and overtopping flows in clear-water scour (CWS) conditions. Detailed sets of equilibrium scour bathymetries and hydraulic parameters were measured, and a regression analysis informed by a dimensional analysis and supported by theoretical contraction scour parameters was applied. The resulting best-fit regression models are presented in the form of two separate equations for scour around a long-setback abutment (LSA) and a short-setback/bankline abutment (SSA/BLA) in compound channels. A comparison with the methodology in a current standard shows that the large bias of overestimation has been reduced up to <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" alttext="plus or minus 4 percent sign" display="inline" overflow="scroll"> <mml:mo form="prefix">±</mml:mo> <mml:mn>4</mml:mn> <mml:mtext>%</mml:mtext> </mml:math> from 20% to 30% for LSA and <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" alttext="plus or minus 3 percent sign" display="inline" overflow="scroll"> <mml:mo form="prefix">±</mml:mo> <mml:mn>3</mml:mn> <mml:mtext>%</mml:mtext> </mml:math> from 40% to 50% for BLA/SSA. Field examples from historical events show good agreement with the suggested model.