Experimental research on the effects of surface screens on a mobile bed

Abstract

In 2000 the Dutch government chose a new point of view for the Dutch rivers: “Room for the River”. This viewpoint is the basis for a new approach of high water protection in the Netherlands. Instead of strengthening and raising the dikes, solutions must be based on space and spatial quality. One of the suggested measures is the addition of secondary channels. The purpose of these channels is enlarging the conveyance area and the ecological role of the river. Maintaining the profile of these channels involves substantial financial consequences. Finding a sustainable solution for undesired erosion or sedimentation is the main focus of this study. The research question is stated as follows: “How can the undesired erosion or sedimentation in secondary channels be corrected with a temporary but sustainable solution in the form of surface screens?”. The main part of this study is an experimental study on the effects of surface screens on a mobile bed. The design of the physical experiments requires choices about the geometry of the flume. The experiments were carried out with a straight flume and with a dividing wall. Preparing the experiments requires information about the flume facility. The experiments have been carried out in the Environmental Fluid Mechanics Laboratory of DUT. The upstream boundary conditions are discharge and velocity distribution. The downstream boundary condition consisted of a fixed water level. The water level was kept constant along the natural slope of the surface. The experiments consist of taking velocity and bed level. The angle of attack and the penetration depth were chosen to be variable. The angle of attack was varied between 15 and 25 degrees. With these relatively small angles the screen acts as guidance for the flow, instead of an obstruction. The penetration depth was varied between 20% and 60% of the water column. The initial test run determined the optimal measurement duration and the initial equilibrium. Four representative cases have been described in detail, giving support to the general conclusions. The flow pattern changes under influence of the surface screen. The main flow direction is guided by the screen, introducing a transverse velocity at the surface. As flow continuity in the flume has to be maintained, the water near the bottom is forced to have a transverse velocity in opposite direction. Redistribution of the suspended transport and the bottom transport was induced. This generated locations were the actual transport did not meet the transport capacity, which gives rise to morphological changes. Next to the spiral motion the screen had an effect on the longitudinal flow velocities. The attacked side of the flume experiences a higher velocity, thereby having a higher transport capacity. This higher capacity gives rise to local erosion of the bed. At the unattacked side, sedimentation occurs, thereby rising the bed level. In the B-series of the experiment a dividing wall was added. The screen in front of the bifurcation gave rise to the same two processes, but the wall introduced an extra effect. The screen influenced the bifurcation relationship. The bed level adapted to the new conditions. The upstream effect of the bifurcation is explained by changes in water level topography, thereby influencing the backwater curve. In general the wall amplified the morphological development of the bed. Finally some suggestions have been made for the practical application of surface screens. In general the screens can be applied in a (secondary) channel or in front of a bifurcation. The use of a screen inside a channel has an advantage not to interfere with the navigation channel. The advantage of a screen in front of a bifurcation is influencing two channels simultaneously. One of the main disadvantages of the latter is the possibility of disturbing the delicate bifurcation relationship. When carefully implemented this effect can simultaneously be the main advantage of this screen layout, as the morphological response increases.