Journal of Waterway, Port, Coastal, and Ocean Engineering

The development of a numerical model for the simulation of water flow in and above a porous beach is described. The model, OTTP-1D, is based on a previous hydrodynamic model of the swash zone, OTT-1D, and builds on the numerical methods developed in the earlier model by the addition of a permeable layer representing a gravel beach. The porewater flow equations used in the present model are described along with the boundary layer approach used to determine the bed shear stress when the effects of exchange between the free-flow and porewater flow regions of the model are taken into consideration. These equations are transformed so as to make them amenable to solution by an upwind finite volume method in the same way as the earlier model. The model is verified against a simple analytical case of Darcy flow and an estimate of lower limit of applicable grain size made thereby. Validation against observed water depths and free-flow velocities recorded on a gravel beach at Slapton, U.K. in 2001 is also undertaken. The model accurately predicts both measured instantaneous water depths and velocities.

Laboratory experiments have been performed to investigate the recession of soft cliffs due to oblique wave actions. The experiments were conducted in a wave tank with irregular waves. A test cliff was constructed of wet sand and (in one test) a sand/clay mixture, which was found to be appropriate for simulating the principal failure mode of soft cliffs fronted by erodeable beaches. The recession process is characterized by the following sequence: toe erosion—cliff failure—erosion of debris—toe erosion. The recession of the cliff top was measured from still photographs taken by an overhead camera. The longshore and cross-shore extents of individual failures and the time interval between two successive failures are obtained.

The hydrodynamic behavior of fixed and heave motion floating breakwaters is studied in the present paper, using a finite-difference, mathematical model based on the Boussinesq type equations. The flow under the floating breakwater is treated separately as confined flow. The pressure field beneath the floating structure is determined by solving implicitly the Laplace equation for the potential Φ of the confined flow using the appropriate boundary conditions. The dynamic equation of heave motion is solved with the consequent adjustments of the continuity equation in the case of a heave motion floating breakwater. Numerical results, concerning the efficiency of fixed and heave motion floating breakwaters, are compared to experimental results satisfactorily. The ability of the numerical model to predict the pressure field beneath the floating structure and the vertical force acting on it is thoroughly examined by making comparisons of the numerical results with large-scale experimental data. The experiments were conducted in the CIEM flume of the Catalonia Univ. of Technology, Barcelona, Spain. The final goal is to study floating breakwaters efficiency in shallow and intermediate waters.

Real time wave measurements in Turkey are often interrupted because of operational difficulties encountered. Therefore, the lacking significant wave height, period and directions were simultaneously estimated from the dynamic Elman type recurrent neural networks. Their predictions were compared with the commonly applied static feed-forward multilayer neural networks and with the stochastic Auto Regressive (AR) and Exogenous Input Auto Regressive (ARX) models. Two distinct learning algorithms, the steepest descent with momentum and the conjugate gradient methods were employed to train the neural networks. It was concluded that, the recurrent neural network generally showed better performance than the feed-forward neural network in the concurrent forecasting of multiple wave parameters. Both artificial intelligence techniques demonstrated a good performance when compared to the predictions of AR and ARX models. Prediction methods are also compared using continuous artificial data generated with known properties by measuring their performance in predicting the removed segments of various lengths. The multivariate ENN model successfully predicted the removed segments of artificially generated wave data. Hence, the learning ability of artificial intelligence techniques was verified signifying the robustness and fault-failure tolerance of neural networks.

A new system for the confinement and recovery of oil slicks or other floating materials is tested in a physical model to evaluate its feasibility and effectiveness. It consists of encircling a polluted area by a specially designed boom using light-weight material, in such a way as to trap entirely the oil slick inside a circular reservoir, making its dispersion less widespread. The system allows the oil slick to be slowly towed to safer zones, where it can be pumped and recycled after separating the reservoir into several compartments in order to increase the efficiency of pumping and treatment. The experimental investigation was carried out in a small-scale model and demonstrated the effectiveness of the system in practical situations. In terms of oil loss, the limits for the critical waves and towing velocities were determined. Carefully selected polymer granulate layers were used to reproduce the characteristics of oil slicks.

Directional wave statistics can be used to optimize the design of offshore structures which are not symmetrical. Unfortunately there does not appear to be any consensus on how directional design criteria should be developed or used. Using simple examples, we show that commonly used methods are unconservative. In order to preserve the reliability given by designing to omnidirectional criteria, the product of the probabilities of nonexceedance from all of the direction bands must equal the omnidirectional probability. We suggest a simple method for assuring consistency between omnidirectional and directional criteria.