Back to results list
Please use this identifier to cite or link to this item:
|Title:||Experimental and numerical studies of flow structure and alluvial processes in partially-obstructed open channel with vegetation canopy||Authors:||Yan, Xufeng||Advisors:||Wai, Wing Hong Onyx (CEE)
Li, Chi Wai (CEE)
|Keywords:||Channels (Hydraulic engineering)
|Issue Date:||2018||Publisher:||The Hong Kong Polytechnic University||Abstract:||The study described in this dissertation presents two key contributions: 1. an experimental investigation and numerical simulation of three-dimensional flow characteristics in a partially-obstructed channel with a vegetation canopy (POCVC); 2. an experimental investigation and two-dimensional numerical simulation of bed morphology evolution in the POCVC. By means of the measurement of flow characteristics in the POCVC, the adjustment of flow entering the vegetation canopy was determined and further examined. Regarding the depth-averaged longitudinal velocity, the flow adjustment distance in the vegetation region is approximately 70% of the canopy length, given the same vegetation density. However, in the neighboring open water region (NOWR), the flow adjustment distance is much shorter, particularly in areas of high-submergence canopy. It was also observed that the vertical profile of longitudinal velocity could become fully-developed within the same flow adjustment distance. Significantly, in the junction region, as the flow propagates downstream the vertical profile of longitudinal velocity is deflected in the near-bed region. Correspondingly, vertical Reynolds stress is negative in that region. It is suggested that velocity deflection and vertical Reynolds stress negativity are induced by the generation of horizontal coherent vortices, enabling the low momentum from the vegetation canopy to be transported into the neighboring open water. The bed resistance, however, plays a negative role in the generation of those vortices. This is confirmed by transverse Reynolds stress in the upper layer region being found to be larger than that in the near-bed region. This new finding encouraged a hydrodynamic model to be subsequently proposed to give a description of the vortex pattern arising from the flow-vegetation interaction in the POCVC. This model can be used to understand flow behaviors in POCVC areas. The turbulent kinetic energy (TKE) budget of fully-developed flow in the POCVC was evaluated. The shear production and turbulent transport both in the vertical and transverse directions are considered. It is shown that within the deeper vegetation canopy TKE is produced by the vegetation stem wake which is balanced by a combination of dissipation and pressure transport. The lower vegetation density promotes the role of vegetation stem wake as the larger stem space tends to support wake development. In the junction region, the shear production due to the vertical/transverse coherent vortices contributes positively to a large proportion of TKE together with the wake production (inside the vegetation canopy). Apart from the energy dissipation and pressure transport, the turbulent transport negatively contributes to TKE. In the NOWR further away from the canopy edges, the turbulent transport becomes dominant.
The Spalart-Allamrus (SA) model was improved to model the surface flow in the POCVC. The concept of the coherent vortices generated in the POCVC was introduced in the numerical model. The turbulence length scale of dominant local vortices is specified so that the transport equation of eddy viscosity becomes anisotropic due to spatial variations of vortices. Results show that the improved vortex-based SA model is able to simulate the flow characteristics of the inner vegetation region except at the entry of the vegetation canopy. In the junction region, the flow velocity deflection and the negative vertical Reynolds stress in the near-bed region can be well reproduced by the improved model, while the standard SA model fails to reproduce the above flow and turbulence properties due to being unable to characterize the horizontal coherent vortices in the POCVC. The parameter study shows that the parameters associated with the horizontal coherent vortices are critical to the particular flow behavior in the junction region. Clean water scour experiments were conducted to investigate the characteristics of bed-load transport and bed morphologic evolution in the POCVC. It was observed that the massive sediment transport that occurs in the NOWR produces a significant peak of the sediment transport rate. It appears that the sediment sorting must occur during the sediment transport process. As the equilibrium bed is reached, characteristic bedforms can be recognized in the POCVC. Such bed features include (A) water pool induced by flow convergence in the NOWR, (B) slightly-eroded bed protected by the vegetation canopy in the VR, (C) slightly-eroded bed at upstream of the vegetation canopy, (D) local bed scouring near the leading edge of the vegetation canopy, (E) sediment bar with surface fining, and (F) a riffle downstream of the vegetated reach. Of significance is that the scour pool and the downstream riffle compose the famous pool-riffle morphology beneficial for aquatic habitat. Over the generated bed morphology, flow characteristics were measured to gain a better understanding of the bed morphology formation mechanism in the POCVC. The relationship existing between flow characteristics and bed morphology was examined. Results showed that linear relations exist among the longitudinal bed profile, longitudinal flow propagation distance, mean flow velocity and transverse inner extension of bed erosion. At the vegetated reach, the transverse bed profile follows the curve of the tangent hyperbolic function. With the well-defined transverse extension lengths the proposed tangent hyperbolic formulas can well predict the transverse bed profile. A two-dimensional morphological model, namely Nays2DH-IRRIC, was employed to model the alluvial process in the POCVC. The validation of the numerical model was carried out with the experimental depth-averaged flow velocity and bed morphology. Then factors including angle of repose (RAV), vegetation density (VD), canopy length (CL) and canopy width (CW) (representing the canopy blockage effect) of the vegetation canopy were adopted in the parameter study. It is found that above factors all have significant influence on bed morphology evolution. Numerical Results show that RAV parameterized in the slope failure model only dominates the transverse bed profile in the junction. Larger RAV leads to steeper transverse bed profile. The increase in VD, CL and CW is able to induce deeper pool region at the vegetated reach. Besides above, larger VD tends to induce upstream expansion in pool and bed erosion inside vegetation canopy. Larger CL induces larger pool region along vegetation canopy in the longitudinal dimension. With the increase in CW, the location of the deepest core in the pool tends to shift upstream.
|Description:||xxxii, 253 pages : color illustrations
PolyU Library Call No.: [THS] LG51 .H577P CEE 2018 Yan
|URI:||http://hdl.handle.net/10397/80156||Rights:||All rights reserved.|
|Appears in Collections:||Thesis|
Show full item record
Files in This Item:
|991022168756403411_link.htm||For PolyU Users||167 B||HTML||View/Open|
|991022168756403411_pira.pdf||For All Users (Non-printable)||6.61 MB||Adobe PDF||View/Open|
Citations as of Feb 18, 2019
Citations as of Feb 18, 2019
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.