Document Type : Research Article
Authors
Razi University, Kermanshah
Abstract
Introduction: Flow and sediment transport has an important role in entrance deformation of open channel junctions. As water moved through a drainage network, it forced to converge at confluence. Due to increasing of water discharge and collision of converging flows, a complex three-dimensional and most highly turbulent location were occurred in the vicinity of the junction. Therefore a deep scour hole and point bar has developed in this area that caused the change in rivers morphology. Despite the large amount of research carried out on flow patterns in river confluences, only a few researches have focused on sediment transport.
Materials and methods: In this research three dimensional model (SSIIM1) was used to study of flow pattern and sediment and erosion pattern at 60 degree Junction .the Navier-Stockes equation of turbulent flow in a general three-dimensional geometry are solved to obtain the water velocity:
, (1)
Where U is average velocity, ρ is density of water, is pressure, the Kronecker delta, which is 1 if i is equal to j and 0 otherwise and general space dimension. The last term is Reynolds stress, often modeled with the following equation:
(2)
Where and k are eddy viscosity and turbulent kinetic energy respectively. Van Rijn's relations were used to calculate sediment suspended and bed load transport.
Dirichlet and zero gradients boundary conditions were used at inflow and outflow boundary respectively. fixed-lid approach was used to computed free surface by using zero gradient for all variables. The wall law for rough boundaries was also used as a boundary condition for bed and wall.
In equilibrium situation, The sediment concentration for the cell closet to the bed was specified as the bed boundary condition. Specified value was used for sediment concentration of other boundary conditions at upstream boundary and zero gradients for the water surface, outlet, and the sides. the only simulation of local scouring and sedimentation at confluence area was also considered.
The SSIIM1 model used structured grid and computer program to provide the required meshand the experimental data was applied to validated model.
The experimental setup consisted of a main flume 9 m long with 75 cm depth for the first 2 m and 45 cm for remaining section and 35 cm wide, and a lateral flume 3m long, 45cm depth and 25 wide. Both flumes had a horizontal slope. An 11cm layer of uniform sediment (D50 = 1.95 mm) was also laid on both channel beds.
Results and discussion:The results showed that the ability of model is relatively good to predict the position of the erosion and sedimentation pattern. The values of maximum scour depth for experimental test and simulation were 0.052 and 0.047 m respectively. However the maximum error to predict scouring depth value was about 10%. This difference could be due to the weakness of Van Rijn's equation to sediment transport and probably measured error. It must be noted that SSIIM1 only used the Van Rijn's equation for bed load transport.
The result Also showed that simulation and experimental test were similar and no sediment transport occurred in the tributary and main channel before the confluence. To investigate the effect of angle 60, 90 and 135 degrees and also discharge ratios of 0.5 and 0.66, the model was applied. A direct relationship was observed between discharge ratio and scouring depth . There was a difference between scouring of discharge ratio 0.5 and 0.66 on a specified angle andthis difference was more obvious with increasing confluence angle. Figure 1 showed the effect of discharge and confluence angle on scouring depth.
Figure 1- The effect of discharge ratio and confluence angle on scouring depth
Keywords
)
Send comment about this article