• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.7 no.6
• /
• pp.61-74
• /
• 2004

Hydraulic engineers and scientists working on river restoration recognize the need for a deeper understanding of natural streams as a complex and dynamic system, which involves not only abiotic elements(flow, sediments) but also biotic components. From this point of view, the role played by riverine vegetation dynamics and flow conditions becomes essential. Hydro-mechanic interaction between flow and flexible plants covering a river bed is studied in this paper and some previous works are discussed. Measurements of turbulence and flow resistance in vegetated open channel were performed using rigid and flexible tube. Measuring detailed turbulent velocity profiles within and above submerged and flexible stems allowed to distinguish different turbulent regimes. Some interesting relationships were obtained between the velocity field and the deflected height of the plants, such as a reduced drag coefficient in the flexible stems. Turbulent intensities and Reynolds stresses were measured showing two different regions : above and inside the vegetation domain. In flexible vegetated open channel, the maximum values of turbulent intensities and Reynolds stresses appear above the top of canopy. Method to predict a flow resistance in flexible vegetated open channel is developed by modifying an analytical model proposed by Klopstra et al. (1997). Calculated velocity profiles and roughness values correspond well with flume experiments. These confirm the applicability of the presented model for open channel with flexible vegetation. The new method will be verified in the real vegetated conditions in the near future. After these verifications, the new method should be applied for nature rehabilitation projects such as river restorations.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2004.05b
• /
• pp.957-960
• /
• 2004

From a water-environmental point of view, with a change of understanding and concern about vegetation, it changes that vegetation acts as stability of channel and bed, providing habitats and feed for fauna, and means improving those with appreciation of the beautiful but resistant factor to the flow So, it becomes important concern and study subjects that turbulent structure by vegetation, shear stress and transport as well as roughness and average velocity by vegetation. But from a hydraulic point of view, vegetation causes resistance to the flow and can increase the risk of flooding, Therefore, this thesis concern the flow characteristics in vegetated open channels. According to the experimental results, $z_0$ was on an average $0.4h_p$ in a vegetated open channel. So, the elevation corresponding to zero velocity in a vegetated channel was the middle of roughness element. The limit for logarithmically distributed profile over the roughness element was from $z_0$ to $0.80h_{over}$ for a vegetated channel. Among the existing theory, the method of Kouwen et al.(1969), Haber(1982), and El-Hakim and Salama(1992) except Stephan(2001) gave a very good value compared to the measured velocity profile.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.29 no.3B
• /
• pp.247-257
• /
• 2009

In the present paper, turbulent open-channel flows with alternate vegetated zones are numerically simulated using threedimensional model. The Reynolds-averaged Navier-Stokes Equations are solved with the ${\kappa}-{\varepsilon}$ model. The CFD code developed by Olsen(2004) is used for the present study. For model validation, the partly vegetated channel flows are simulated, and the computed depth-averaged mean velocity and Reynolds stress are compared with measured data in the literature. Comparisons reveal that the present model successfully predicts the mean flow and turbulent structures in vegetated open-channel. However, it is found that the ${\kappa}-{\varepsilon}$ model cannot accurately predict the momentum transfer at the interface between the vegetated zone and the non-vegetated zone. It is because the ${\kappa}-{\varepsilon}$ model is the isotropic turbulence model. Next, the open channel flows with alternate vegetated zones are simulated. The computed mean velocities are compared well with the previously reported measured data. Good agreement between the simulated results and the experimental data was found. Also, the turbulent flows are computed for different densities of vegetation. It is found that the vegetation curves the flow and the meandering flow pattern becomes more obvious with increasing vegetation density. When the vegetation density is 9.97%, the recirculation flows occur at the locations opposite to the vegetation zones. The impacts of vegetation on the flow velocity and the water surface elevation are also investigated.

• Journal of Korea Water Resources Association
• /
• v.38 no.10 s.159
• /
• pp.871-883
• /
• 2005

This paper investigates the impacts of turbulent anisotropy on the mean flow and turbulence structures in vegetated open-channel flows. The Reynolds stress model, which is an anisotropic turbulence model, is used for the turbulence closure. Plain open-channel flows and vegetated flows with emergent and submerged plants are simulated. Computed profiles of the mean velocity and turbulence structures are compared with measured data available in the literature. Comparisons are also made with the predictions by the k-$\epsilon$ model and by the algebraic stress model. For plain open-channel flows and open-channel flows with emergent vegetation, the mean velocity and Reynolds stress profiles by isotropic and anisotropic turbulence models were hardly distinguished and they agreed well with measured data. This means that the mean flow and Reynolds stress is hardly affected by anisotropy of turbulence. However, anisotropy of turbulence due to the damping effect near the bottom and free surface is successfully simulated only by the Reynolds stress model. In open-channel flows with submerged vegetation, anisotropy of turbulence is strengthenednear the vegetation height. The Reynolds stress model predicts the mean velocity and turbulence intensity better than the algebraic stress model or the k-$\epsilon$ model. However, above the vegetation height, the k-$\epsilon$ model overestimates the mean velocity and underestimates turbulence intensity Sediment transport capacity of vegetated open-channel flows is also investigated by using the computed profiles. It is shown that the isotropic turbulence model underestimates seriously suspended load.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.34 no.3
• /
• pp.813-820
• /
• 2014

This study presents a numerical modeling of mean flow and turbulence structures of partly-vegetated open-channel flows. For this, Reynolds-averaged Navier-Stokes equations with vegetation drag terms are solved numerically using the non-linear k-${\varepsilon}$ model. The numerical model is applied to laboratory experiments of Nezu and Onitsuka (2001), and simulated results are compared with data from measurement and computations by Kang and Choi's (2006) Reynolds stress model. The simulation results indicate that the proposed numerical model simulates the mean flow well. Twin vortices are found to be generated at the interface between vegetated and non-vegetated zones, where turbulence intensity and Reynolds stress show their maximums. The model simulates the pattern of the Reynolds stress well but under-predicts the intensity of Reynolds stress slightly.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.30 no.2B
• /
• pp.179-189
• /
• 2010

This paper presents a numerical investigation of vorticity generation in fully vegetated open-channel flows. The Reynolds stress model is used for the turbulence closure. Open-channel flows with rough bed-smooth sidewalls and smooth bed-rough sidewalls are simulated. The computed vectors show that in channel flows with rough bed and rough sidewalls, the free-surface secondary currents become relatively smaller and larger, respectively, compared with that of plain channel flows. Also, open-channel flows over vegetation are simulated. The computed bottom vortex occupies the entire water depth, while the free-surface vortex is reduced. The contours of turbulent anisotropy and Reynolds stress are presented with different density of vegetation. The budget analysis of vorticity equation is carried out to investigate the generation mechanism of secondary currents. The results of the budget analysis show that in plain open-channel flow, the production by anisotropy is important in the vicinity of the wall and free-surface boundaries, and the production by Reynolds stress is important in the region away from the boundaries. However, this rule is not effective in vegetated channel flows. Also, in plain channel flows, the vorticity is generated mainly in the vicinity of the free-surface and the bottom, while in vegetated channel flows, the regions of the bottom and vegetation height are important to generate the vorticity.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.26 no.5B
• /
• pp.529-537
• /
• 2006

This paper investigates the impacts of the drag-related weighting coefficients on mean velocity and turbulence structures. The transport equations for the Reynolds stress of vegetated open-channel flows are derived by using the temporal- and horizontal-averaging scheme. It is found that the total Reynolds stress of vegetated open channel flows consists of the Reynolds stress due to temporally fluctuating velocities and the Reynolds stress due to spatially fluctuating velocities. The drag-related weighting coefficient $C_{fk}$ for the total Reynolds stress component is found to be unit, while the coefficient for the Reynolds stress due to temporally fluctuating velocities can be negligible. This is the reason why very small weighting coefficients in previous studies yield very good agreements with measured data. In other words, the Reynolds stress due to spatially fluctuating velocities remains still unknown, especially due to the large number of measuring locations. Through a developed Reynolds stress model, vegetated open-channel flows are simulated and compared with measured data from the literature. Comparisons reveal that the computed mean flow and Reynolds stress structures are hardly affected by the drag-related weighting coefficients. However, the computed turbulence intensity profiles are significant different with the drag-related weighting coefficients. A budget analysis of the transport equations for the Reynolds stress component is carried to investigate why turbulence intensity is affected by the drag-related weighting coefficients.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.16 no.9
• /
• pp.6305-6314
• /
• 2015

This study presents numerical simulations of mean flow and turbulence structure of an open channel with submerged vegetation. Filtered Navier-Stokes equations are solved using large-eddy simulation (LES). The immersed boundary method (IBM) is employed based on a Cartesian grid. The numerical result is compared with experimental data of Liu et al. (2008) and shows that simulated results coincided reasonably with experimental data within the average error of 10%. Strong vortices are generated at the interface between vegetated and non-vegetated regions with spanwise extent. The generation of turbulence induced by shear at the interface is interfered with wake turbulence, resulting turbulence intensity maximum. Turbulence produced by shear affects the flow in vegetated region and the penetration depth increases with an increase in the submergence ratio. This result can be used to understand sediment transport mechanisms in the vegetated region.

• Journal of Korea Water Resources Association
• /
• v.43 no.6
• /
• pp.517-524
• /
• 2010

In this study, the 1D apparent shear stress model for vegetated compound open-channel flows was suggested. To consider the effect of momentum exchange between main channel and floodplain, the eddy viscosity concept was used in the present model. The interfacial eddy viscosity in the interface of main channel and floodplain was determined from the 3D Reynolds stress model. The evaluated interfacial eddy viscosity appears to be good agreement with those proposed previously. To investigate the effect of interfacial eddy viscosity, sensitive analysis was carried out. the computed backwater profiles are nearly identical with respect to the value of the interfacial eddy viscosity. However, the discharge conveyed by the floodplain changes is proportional to the interfacial eddy viscosity. Finally, the changes of the interfacial eddy viscosity due to the vegetation density and vegetation height were examined. The computed results of interfacial eddy viscosity are in proportion to the vegetation density and vegetation height, and the interfacial eddy viscosity has a range of $(2-5)\;{\times}\;10^{-4}$.

• Journal of Korea Water Resources Association
• /
• v.43 no.9
• /
• pp.823-833
• /
• 2010

Open-channel flows with submerged vegetation show two distinct flow structures in the vegetation and upper layers. That is, the flow in the vegetation layer is featured by relatively uniform mean velocity with suppressed turbulence from shear, while the flow in the upper layer is akin to that in the plain open-channel. Due to this dual characteristics, the flow has drawn many hydraulic engineers' attentions. This study compares layer-averaged models for flows with submerged vegetation. The models are, in general, classified into two-layer and three-layer models. The two-layer model divides the flow depth into vegetation and upper layers, while the three-layer model further divides the vegetation layer into inner and outer vegetation layers depending on the influence of the bottom roughness. This study compares the two-layer model and the three layer-model. It is found that the two-layer model predicts better the average value of the velocity and the prediction by the three-layer model is sensitive to Reynolds shear stress. In the three-layer model, the mean flow in the inner vegetation layer does not affect the flow seriously, which motivates the proposal of the modified two-layer model. The two-layer model, capable of predicting non-uniform mean velocity, is based on the Reynolds stress which is linear and of power form in the upper and vegetation layers, respectively. Application results reveal that the modified two-layer model predicts the mean velocity at an accuracy similar to the two- and three-layer models, but it predicts poorly in the case of very low vegetation density.