Checking the Accuracy of Selected Formulae for both Clear Water and Live Bed Bridge Scour

Abstract

Due to severe scouring, many bridges failed worldwide. Therefore, the safety of the existing bridge (after contrition) mainly depends on the continuous monitoring of local scour at the substructure. However, the bridge's safety before construction mainly depends on the consideration of local scour estimation at the bridge substructure. Estimating the local scour at the bridge piers is usually done using the available formulae. Almost all the formulae used in estimating local scour at the bridge piers were derived from laboratory data. It is essential to test the performance of proposed local scour formulae using field data. In this study, the performance of selected bridge scours estimation formulae was validated and statistically tested using field data for existing bridges in Canada, Iraq (Kufa, Najaf), Pakistan, Bangladesh, and India. The validated formulae were HEC-18, Forehlich, and Johnson. The validation was conducted by comparing the predicted local scour depths obtained from applying the above selected formulae with the local scour depths obtained from the field data. The comparison between them was presented using a scattergram. However, statistical tests were used to present the accuracy of the local scour predictions. The tests were conducted using three statistical indices, namely, Theil’s coefficient (U), Mean Absolute Error (MAE), and Root Mean Square Error (RMSE). Among the tested formulae, the Jonson formula gave satisfactory performance since the values of U, MAE, and RMSE were found to be 0.112, 1.351, and 1.650, respectively.