The Current Deflecting Wall: mitigating harbour siltation; Set-up and integration of physical and numerical modelling techniques
Kuijper, C.; Winterwerp, J.C. (2003). The Current Deflecting Wall: mitigating harbour siltation; Set-up and integration of physical and numerical modelling techniques. WL | Delft Hydraulics: [s.l.]. 46 pp.
The report gives a description of the combined application of a physical scale model and various numerical models to assess the effect of a Current Deflecting Wall (CDW) on the rate of exchange of water between river and harbour (or access channel) and consequently on siltation. The physical model is a slightly distorted model (with vertical and horizontal scales of 150 and 300), which includes a section of the Beneden Zeeschelde in Belgium. The model takes into account the effects due to tidal filling, horizontal entrainment and density currents resulting from salinity differences. Calibration of the model was done on the basis of variations of water levels, velocities and salinities on the river near the access channel. The boundary conditions of the physical scale model were derived from an overall two-dimensional (depth-averaged) numerical model as well as from a three-dimensional model, which was nested in the overall model. The benefits of a CDW for the Beneden Zeeschelde, as determined from the model tests, will be described elsewhere. To describe adequately the three-dimensional flow, as induced by a CDW, numerical models needed to be extended with proper process formulations and newly developed numerical schematization techniques. These were implemented in DELFT3D and relate to: (i) a fixed layer schematisation of the numerical grid in vertical direction, (ii) inclusion of a non-hydrostatic pressure distribution and (iii) application of a smooth boundary approximation. Comparison with results from the physical scale model will have to indicate the potentials and limitations of numerical modelling in these matters. Vice versa, the numerical scale model allows the assessment of possible scale effects of the physical model due to the distortion and relatively low Reynolds numbers (viscosity). The approach thus combines the strong points of both modelling approaches and offers the opportunity to develop and validate numerical models for complicated three-dimensional flow phenomena.