Models for Hydraulic Structures

Three-dimensional hydro-numerical models are used to study and optimise hydraulic structures and to examine the interaction ship/waterway. The Nast3DGPF method is being developed by the University of Bonn and adapted to user requirements in cooperation with the BAW. In addition, the 3D-CFD method Star-CCM+ from the company cd-adapco, which is already established on the market, is also used, as it covers another area of application. In areas of overlap, both methods are used for purposes of verification and validation of results. Furthermore, the open source method OpenFOAM is at present in the test stages. Tests are being carried out to establish in what measure OpenFOAM could replace one of the previously mentioned methods. The following problems can be answered with these methods:

• hydraulic dimensioning of structures
• studies of flow in lock approaches
• optimisation of filling and emptying processes
• assessment of high water discharge at impoundment weirs
• study of the propeller race
• studies of the interaction of the propulsion unit with the bottom
• studies of wave propagation in broad reaches
• studies on wind attack on ships
• analysis of trim and incidences of meeting ships

The extremely small-scale modelling in the near field of structures is distinguished fundamentally from the methods for mesoscale modelling of river basins. If the linear extension of the model areas for river basin modelling is between about 5 and 25 km with a local resolution of around 100 metres to 1 metre (e.g. for depicting groynes), the model area lengths for the depiction of hydraulic structures and ships will be around 100 to a maximum of 1000 metres. The models used have a high lattice gradient with unit dimensions from, at most, a few metres at the boundaries of the model to centimetres in order also to depict the flow conditions at gates, flaps, filling streams, propellers and rudder blades with sufficient precision. In the mesoscale modelling of river basins one can often concentrate on preferred spatial directions for the modelling. If river basins are modelled three-dimensionally, the water level is described using exactly one function value of the riverbed and the free surface is defined as the boundary of the model. In the near field of hydraulic structures and ships, however, there is a high-grade three-dimensional penetration of the structures under study by the surrounding fluids water and air. Structural components extend into the fluid which flows over or under them in equal measure; ships are immersed in the fluid. At hydraulic jumps, filling streams, running waves on revetments and freely flying bodies of waters, water and air must be modelled alternately as components by means of a location coordinate. A description of this type of water levels as a model boundary would be very complicated. In such instances, a multi-phase modelling method has now become generally accepted. The flows that occur, for instance, during the filling of lock basins, in stilling basins of weir plants and in the near surrounding of ships’ propulsion units (consisting of propelling screw, propeller tunnel and rudder blade) are dominated by a high degree of turbulence. The selection of the suitable turbulence model and the calibration of the parameters are decisive for the reliability of the results. The behaviour at wall boundary layers influences the surrounding flow regime. The extent of these effects is critically influenced by the scale that is being considered. At present, it is assumed that wall influences have relatively little influence on, for example, the discharge over a weir field when modelling turbulences, but that, on the contrary, wall influence is of elementary significance in the modelling of the interaction between ship, propulsion unit and bottom. The need for a study of the dynamic interaction between structure and fluid in the area administered by the Federal Waterways and Shipping Administration (WSV) is huge. The BAW has therefore great interest in the possibilities of numerical modelling for the lowering process of a weir, water flowing into a lock basin while the gates are moving, the dynamics of a ship in a lock and modelling of the trim of a moving vessel in a waterway. Proved models for cost-effective production by the BAW may not be expected in this sector for another few years, although the models used at present demonstrate many promising approaches.

Contact person: Dr.-Ing. Carsten Thorenz

---

back to Mathematical Models for Inland Areas

---

Overview