Modeling local scour using k-? turbulence model and soft computing techiques

In this study, a hybrid numerical-analytical model is developed, which simulates the temporal evolution of local scour downstream of hydraulic structures, and simulates the flow patterns in the scoured zone. The hybrid model consists of an analytical model and a k-? turbulence model. The analytical model analyzes the external forces acting on a bed grain and models the instantaneous shape of the scour hole solving an integro-parabolic equation. The numerical model derives the instantaneous shape of mobile bed and simulates the flow patterns in the corresponding scour hole.The proposed model is verified by experimental data taken from the studies of others in the literature, predictions of commercial software and numerical studies in the literature. The proposed hybrid model predicted the temporal evolution of local scour downstream of hydraulic structures, and simulated the flow patterns in scoured hole with quite good agreements with the experimental ones. The proposed hybrid model proved to be efficiently used in complex hydraulic engineering problems. The use of a new approach to prediction of local scour, soft computing technique, is presented. Artificial Neural Networks and Genetic Programming based models are developed, which predict the local scour downstream of hydraulic structures. The results encourage the use of soft computing technique in very complex hydraulic engineering problems.