OS42B-01
Numerical Studies of Flow Over a Sill: Sensitivity of the Non-Hydrostatic Effects to the Grid Size
A non-hydrostatic terrain following model in cross sectional form is applied to study the tidal inflow near a sill in an idealized stratified fjord. There is a transfer of energy from the barotropic tide to internal waves and then to irreversible mixing. The range of length scales involved goes from the scale of the forced tide to scales associated with wave breaking. Methods for computing non-hydrostatic pressure in terrain following models will be discussed. The sensitivity of the non-hydrostatic pressure effects to the grid size, and in particular the effects on the time mean fields, will be addressed.
OS42B-02
Development And Application Of Non-Hydrostatic Model To The Coastal Engineering Problems
The 3D non-hydrostatic free surface model developed by Kanarska and Maderich (2003) for stratified flows was further improved and has been used to simulate coastal processes. In the model the surface elevation, hydrostatic and non-hydrostatic components of pressure and velocity are calculated at sequential stages. Unlike most non-hydrostatic models, the 2-D depth-averaged momentum and continuity equations were integrated explicitly, whereas the 3-D equations were solved semi-implicitly at subsequent stages. The RANS and subgrid- scale eddy viscosity and diffusivity parameterization were implemented in the model to parameterize small-scale mixing. The model was applied to three coastal engineering problems. First, we used the model coupled with a 3D Lagrangian sediment transport model to predict scour caused by propeller jets of slowly maneuvering ships. The results of the simulations show good agreement with laboratory experiments and field ADCP measurements with tug boats. Second, the model was applied, while nested into the hydrostatic far-field counterpart model, for near-field simulation of cooling water discharge through submerged outfalls. Third, laboratory experiments and simulations were performed to estimate effects of large-amplitude internal solitary waves (ISW) on submerged structures and coastal bottom sediments. In the first series of experiments and simulations, the interaction of ISW-depressions with a rectangular bottom obstacle was investigated. In the second series, the ISW-depression was studied passing through a smooth local lateral constriction. The third series of laboratory experiments and simulations was conducted to investigate the dynamics of ISW of depressions reflecting from a steep slope. Contribution of V. Maderych in this work was supported by Hankuk University of Foreign Studies Research Fund of 2007.
OS42B-03
Wet and Dry Algorithms for Modeling Hydrodynamics in Bays/Estuary/Coastal Areas
To adequately address inundation problems in hydrodynamic modeling over bays/estuaries/coastal areas, appropriate wet and dry algorithms hold the key to accurately capture the moving fronts of waters over land surfaces. From the point of view of computational methods, two approaches are available, namely the moving grid method and the fixed-grid approach. The moving grid method usually adopts either the Lagrangian or the arbitrary Lagrangian-Eulerian (ALE) representation. In the Lagrangian representation, the grid movement strictly follows the local velocity of the fluid. Therefore the grid may deform severely for complex flows involving circulations and vortices. In the ALE representation, arbitrary velocity can be assigned to node movement, thus the grid movement in the interior of the domain is independent of the flow velocity. Thus, it is possible for the ALE method to maintain reasonably shaped meshes and describe the moving fronts accurately at the same time. In the fixed-grid method, the grid used to solve the flow problem is entirely or quasi-entirely fixed and the equations are expressed using the Eulerian representation. The moving fronts are identified based on whether computational cells are wet or dry. Dry cells are either taken out of the computational domain or assigned a minimum depth and remained in the computational domain. A dry cell, if it has been taken out of the domain, can become wet again if its neighboring cells are wet. There may be severe mass conservation errors if the scheme of maintaining a minimum depth is employed. In this paper, we present the moving grid method used in the modeling of the three dimensional surface flow in large water bodies surrounded by mash flats. Three algorithms were developed to model inundation of surrounding land surfaces: (1) semi-Lagrangian moving grid, (2) hybrid scheme of storage and hydrodynamic cells, and (3) coupled scheme of three-dimensional and two- dimensional flows. These algorithms are demonstrated and their advantages and disadvantage are discussed. The inundation of the Loxahatchee Estuary over its flood plain is used to demonstrate the robustness of the coupled scheme of 3D and 2D flows. http://people.cecs.ucf.edu/yeh
OS42B-04
Numerical modeling of shoaling of internal solitons with respect to maximum velocities.
Internal waves and solitons are found in all oceans. Solitons and internal waves that have amplitudes over 100 m and wave lengths of 20 km are measured. Measurements prior to the development of a gas field located in the shelf slope outside mid Norway (Europe) showed sudden strong near seabed velocities. The high velocities were often connected to drop in temperature similar to shoaling of solitons. The temperature drops indicated pycnocline movements maybe as large as 100 m which might be connected to internal waves or solitons. 2D non-hydrostatic numerical simulations of internal solitons which shoal along a linear continental shelf slope were performed. The sensitivity of the results to the steepness of the slope and amplitude of the soliton is investigated. The main focus is on the maximum velocity generated during the breaking of the solitons but also the effect of the grid resolution is discussed.