The engineering design of a hydraulic structure such as a river bridge requires consideration of the factors that affect the safety of the structure. Among them, one of the most important variables is bridge foundation scour. However, engineering experience seems to indicate that computation of scour depth using current scour formulas tends to overpredict scour in comparison to field measurements. The result can be an overdesigned bridge foundation that increases the cost of the bridge. One possible reason for the overprediction is the current practice of adding separate estimates of contraction scour and pier scour when in fact these processes occur simultaneously and interact. During the occurrence of a flood, velocities and depths increase but they are affected by changes in the distribution of discharge between the main channel and floodplain. In addition, the time history or time development of contraction scour and local pier scour is not the same. As a result, the influence of contraction scour on pier scour, for example, is time dependent. Laboratory experiments are proposed using a 1:45 scale hydraulic model of the Ocmulgee River bridge at Macon, Georgia. Initially, the contraction scour will be measured without the bridge piers in place. In this experiment, the time history of the scour and the velocity distributions at the equilibrium state will be measured. Then the piers will be placed at the bridge cross-section in the flume, and the same measurements will be made. The sensitivity of the measurements to small changes in depth at the same discharge will also be determined, and comparisons will be made with field measurements of scour depth. The results will be used to assess the relative contribution of contraction scour and local pier scour to the final design of the bridge foundation depth.

Bridge scour, which is the removal of bed materials from near the bridge foundations, is observed to be the most predominant cause of bridge failures in the United States. Scour in cohesive soils is greatly different from scour in cohesionless soils owing to the differences in critical shear stresses, scour extents and the time taken to reach the maximum scour depth in the scour process. The present solutions available for the cohesionless soils cannot be applied to cohesive soils because of the above crucial reasons. Also, a compound channel model with main channel and flood plain arrangement represents more closely the field stream conditions rather than a simple rectangular prismatic model. In this study, a systematic investigation of the scour process due to flow contractions in a compound channel with cohesive soil bed is made by conducting a series of flume tests representing typical field conditions. The effect of the most crucial factors causing contraction scour namely flow velocity, depth of flow and the shape of the abutment is examined. Correction factors are developed for changes in flow geometries incorporating simulation results from the one dimensional flow simulation model HEC RAS. Most importantly, a methodology to predict the depth of the deepest scour hole and its location in the vicinity of the contraction structure is developed for compound channels through an extension of the presently available methodology to predict maximum scour depths in simple rectangular channels. A prediction method to identify the extent of the uniform scour depth is also developed. Finally, an investigation of precision of the proposed methodology has been carried out on the field data from a number of real life contraction scour cases. The results obtained from this study indicate that depth of flow and geometry of the contraction section significantly influence final scour depth in cohesive soils with deeper flows and harsh contractions resulting in increased scour depths. However, corrections for different contraction inlet skew angles and long contractions need to be further explored in future studies.

Bridge scour, which is the removal of bed materials from near the bridge foundations, is observed to be the most predominant cause of bridge failures in the United States. Scour in cohesive soils is greatly different from scour in cohesionless soils owing to the differences in critical shear stresses, scour extents and the time taken to reach the maximum scour depth in the scour process. The present solutions available for the cohesionless soils cannot be applied to cohesive soils because of the above crucial reasons. Also, a compound channel model with main channel and flood plain arrangement represents more closely the field stream conditions rather than a simple rectangular prismatic model. In this study, a systematic investigation of the scour process due to flow contractions in a compound channel with cohesive soil bed is made by conducting a series of flume tests representing typical field conditions. The effect of the most crucial factors causing contraction scour namely flow velocity, depth of flow and the shape of the abutment is examined. Correction factors are developed for changes in flow geometries incorporating simulation results from the one dimensional flow simulation model HEC RAS. Most importantly, a methodology to predict the depth of the deepest scour hole and its location in the vicinity of the contraction structure is developed for compound channels through an extension of the presently available methodology to predict maximum scour depths in simple rectangular channels. A prediction method to identify the extent of the uniform scour depth is also developed. Finally, an investigation of precision of the proposed methodology has been carried out on the field data from a number of real life contraction scour cases. The results obtained from this study indicate that depth of flow and geometry of the contraction section significantly influence final scour depth in cohesive soils with deeper flows and harsh contractions resulting in increased scour depths. However, corrections for different contraction inlet skew angles and long contractions need to be further explored in future studies.

The bridge scour depths in cohesive soil have been predicted using the scour equations developed for cohesionless soils due to scarce of studies about cohesive soil. The scour depths predicted by the conventional methods will result in significant errors. For the cost effective design of bridge scour in cohesive soil, the Scour Rate In COhesvie Soil (SRICOS) for the singular circular pier in deep water condition was released in 1999, and has been developed for complex pier and contraction scour. The present study is the part of SRICOS-EFA method to predict the history of contraction scour, and local scours, such as abutment scour and pier scour. The main objective is to develop the prediction methods for the maximum and the uniform contraction scour depth, the maximum pier scour depth and the maximum abutment using flume test results. The equations are basically composed with the difference between the local Froude number and the critical Froude number. Because the scour happens when the shear stress is bigger than the critical shear stress, which is the maximum shear stress the channel bed material can resist from the erosion, and continues until the shear stress becomes equal to the critical shear stress. All results obtained from flume tests for pier scour have been conducted in Texas A&M University from 1997 to 2002 are collected and reanalyzed in this study. Since the original pier scour equation did not include soil properties. The effect of water depth effect, pier spacing, pier shape and flow attack angle for the rectangular pier are studied and correction factors with respect to the circular pier in deep water condition were newly developed in present study. For the abutment scour, a series of flume tests in large scale was performed in the present study. Two types of channel - rectangular channel, and compound channel - were used. The effect of abutment length, shape and alignment of abutment were studied and the correction factors were developed. The patterns of velocity and of scour were compared, and it was found that the maximum local scour occurred where the maximum turbulence was measured. For the contraction scour, the results obtained from a series of flume tests performed in 2002 and a series of flume tests for the abutment scour in the present study are analyzed. The methodologies to predict the maximum contraction scour and the uniform contraction scour in the compound channel was developed. Although all prediction methods developed in the present study are for the cohesive soils, those methods may be applicable to the cohesionless soils because the critical shear stress is included in the methods. All prediction methods were verified by the comparison with the databases obtained from flume test results and field data.