Estimation of the Dispersion Coefficients and Sedimentation in Compound Channels with vegetated floodplains

Increasing attention to environmental issues in the management of rivers, causese the development of research in the field of hydraulic parameters and sediment transport in the rivers. Vegetation in many rivers, has a significant impact on the flow hydraulic, morphology and sediment transport processes. Understanding the hydraulics of flow in a compound channel with vegetated floodplains is very important for determining the stage-discharge curve and for supporting the management of fluvial processes. There are few models within CFD codes that conceptually or physically represent the hydrodynamic impact of a vegetation canopy on the velocity field and conveyance characteristics within an estuarine or riverine system. The present research investigates the numerical and experimental modelling aspects of flow in compound channels with vegetated floodplain, emphasis to the suspended sediment transport. This study, has proposed a method for predicting the depth-averaged velocity in compound channels with vegetated floodplains, based on an analytical solution to the depth-integrated Reynolds-Averaged Navier-Stokes equation (RANS) with a term included to account the effects of vegetation. Two dimensional advection-diffusion equation were used to calculate the dispersion coefficients and sedimentation rate, this equation were solved by using a finite volume method. The vegetation is modeled via an additional term in the momentum equation to account for the additional drag force. The method includes the effects of bed friction, drag force, lateral turbulence and secondary flows, via four coefficients f, CD, λ & Γ respectively. This model is able to estimate the velocity distribution in channels with different dimensions, different vegetation densities. This model is able to estimate the velocity distribution in channels with different dimensions, different vegetation densities. The wake interference model given by Nepf (1999) is used to estimate the bulk drag coefficient. Modified forms of the Colebrook–White equation can be used to calculate a local friction factor in compound channels with vegetated floodplains. Several eddy viscosity models have been used in the literature to predict transverse momentum exchange in the mixing layer. Once velocity profile is obtained, the dispersion coefficient is evaluated and the remarkable role of vegetation is shown. Both the velocity profile and the dispersion coefficient have been validated using our new experimental results. The model was applied in inverse mode to estimate the model parameters that quantified eddy viscosity coefficient, secondary flow, longitudinal dispersion, transverse dispersion and filtration in the experiments. In addition, the Elder’s and Neph’s relationships also calibrated to calculate the Longitudinal and transverse dispersion coefficients. A Levenberg-Marquardt algorithm was used to find the values of the model parameters that minimized the sum-of-squared differences between measured and modeled velocities and particle concentrations. Experiments performed in a flume of the Hydraulic laboratory. The experimental channel is employed in this investigation was 10 m long × 0.3 m wide and 0.6 m deep. The channel has a simple rectangular cross section; however, this was modified by using Plexiglas sheeting to produce an asymmetrical compound shape. The artificial stems were PVC rods of 5 mm outside diameter, arranged in staggered grid. The model substance used as suspended load was silica of mean particle diameter around 7µm and density 2143 kg/m3. Silica suspensions were injected to produce fully mixed inflow concentrations of about 6 g/l. In general, two series of experiments have been done in this study. In The first series of experiments the depth averaged velocity were measured, the second series have been conducted to estimate the rate of sedimentation. Depth averaged velocity measured by using micro-propeller and pitot tubes. Results showed that the use of constant coefficient of 0.0683 for eddy viscosity coefficient have not a significant impact on depth averaged velocity. The error about 1% have been observed between results of numerical modeling and experimental data, when the relationships 0.417δρgS0 HDr and 0.62δρgS0HDr have been used for secondary flow coefficients, to the main channel and floodplain, respectively. Results showed that dispersion coefficient in vegetated channels increased cross the main channel compared to no vegetated case, due to the gradient of velocity. Sedimentation rate coefficients calibrated by comparing the experimental and numerical sedimentation rate. The numerical model underestimates the experimental data about 8%. Experimental data of Zong(2011) were used to verification of numerical model, the results showed that the modified model overestimates the data with 22% difference. The results show that the quasi 2D model reproduces a reasonable simulation of the flow field and sediment transport in compound channels with vegetated floodplains.