No abstract

No abstract

No abstract

Maximum depth of scouring around bridge abutments is a significant criterion in design of safe depth for abutment foundation. Many studies done on maximum scouring depth are limited to specific shape of abutment and there is no general equation in the estimation of time varying scour depth at short abutments. In this research, maximum depth and also the temporal variation of local scour at a vertical-wall, wing-wall and semicircular abutments, and spur dike as well, were investigated experimentally. About 1400 sets of experimental data were collected. The results indicated that 70-90% of the equilibrium scouring depths occurred during the first 20% of overall time of scouring tests. According to the collected data, Multiple Nonlinear Regression (MNLR), Artificial Neural Networks (ANN) and Adaptive Neuro Fuzzy Inference System (ANFIS) were adopted to predict the time variation of scour depth around abutments. The computer models were compared to the other empirical equations presented in the literature. The study showed that the conducted regression model is rather precise and practical (R 2 = 0.88). Also, the results of the numerical modeling indicated that ANFIS model produced the best results (R 2 = 0.98) in comparison with ANN models using feed forward back propagation (R 2 =0.96) and radial basis function (R 2 = 0.94).

The accuracy of support vector regression (SVR) procedure in modeling the percentage of shear force carried by walls in a rectangular channel with rough boundaries was investigated. The SVR model is extended, and the more appropriate kernel function and input combination are studied. Finally, the SVR model with an exponential kernel function and three influence parameters was selected as the best SVR model with the lowest error. The output of this more appropriate SVR model is presented as a program. Then, this most appropriate SVR model is compared with three equations presented by other researchers for rough and smooth channels. The SVR model with the highest accuracy and lowest statistical values (RMSE of 0.565) performed the best compared with the other equations.