Scour Depth at the Edge of Different Submerged Vanes Shapes

INTRODUCTION

In river meanders especially at the base of outer bank sever erosion can occur (Javaheri et al., 2008; Samadi et al., 2011). This type of erosion is called bed scouring which can occur in other parts of rivers such as river confluence (Ghobadian and Bejestan, 2007; Bejestan and Hemmati, 2008). Many measures have been developed to control such erosion in the past. Among these are Spur dike (Masjedi et al., 2010; Nasrollahi et al., 2008). Vegetation also can increase the bed roughness and can be benefit for reduction of bank scouring (Fathi-Moghadam, 2007; Sabegh et al., 2011). One of the new technique which have been used recently is the submerged vane. Submerged vanes (Iowa vanes) are designed in order to modify the near-bed flow pattern and bed-sediment motion in transverse direction in the rivers. The vanes are installed vertically on the channel bed, at an angle of attack α with the approaching flow (Odgaard and Wang, 1991a). The vanes height, length, distance from banks and distance from each other as well as the vane angle are very important parameters for design. The vane should not be too closed to each other to act as a bed roughness. Since investigators have shown that bed roughness can dissipate more kinetic energy of flow (Bejestan and Neisi, 2009; Ezizah et al., 2012). All investigations indicate that the local scour depth, d_s, increases with an increase in α. At larger α, the vane will be subjected to a relatively larger drag force, with increased flow resistance. The major problem in the use of larger α is occurrence of great scour around the vane that may dislodge the vane. Because of these reasons, vanes with α>20° are not practical in field applications (Gupta et al., 2010). Therefore, presenting the practical methods to surmount the scour problem is very important. Regarding to this fact that the most of vanes are pre-constructed structures, modifying the vane shape can be considered as a practical approach for this purpose.

River bank protection is one of the applications of the vanes (Odgaard and Kennedy, 1983; Odgaard and Mosconi 1987; Wang and Odgaard 1993; Johnson et al., 2001; Marelius, 2001). The submerged vanes technique is used to deepening the rivers in order to navigation purposes and reduction of the shoaling problems, too (Odgaard and Spoljaric 1986; Odgaard and Wang 1991a). In addition, the vane system can be used to limit the influx of bed-load sediment into riverside intake structures (Nakato et al., 1990; Wang et al., 1996; Nakato and Ogden, 1998; Barkdoll et al., 1999; Michell et al., 2006; Emamgholizadeh and Torabi, 2008; Lila and AliReza, 2006).

In general, the cost of the submerged-vane technique compares favorably (Odgaard and Wang, 1991b) with that of the structural and conventional methods such as dike, gabion, riprap and etc. Thereby cutting the leading edge of the vanes after design of the baseline rectangular vane reduces the required construction material of the vanes. Ouyang (2009) investigated the effect of the shape and dimensions of a submerged vane, with constant area in all alternatives, for sediment management in alluvial channels by using a numerical model; however, the vane-induced local scour is not introduced in the model. In the present study, the local scour around the vanes is the major of the investigation and also the area of the vanes is not constant. The present study is focused on the four shapes of the vanes, including a simple rectangular plate and three flat-plate types of vanes with reduced area due to 30, 45 and 60 degrees (θ) chamfer at leading edge of the vanes, respectively. For the rectangular vane, θ = 0°. Cutting the leading edge of the vanes leads to reduction of the effective area of the vanes. Most of the existing studies have been restricted in to a range of 0<Fr = 0.25 (Gupta et al., 2010). This study’s experiments carried out at 0.14, 0.16 and 0.20 Froude numbers (corresponding to 30, 35 and 45 L sec^-1 flow rates).

LABORATORY SETUP

In the present study, a laboratory recirculating flume has been used, 7.3 m in length, 0.56 m width, 0.6 m height and constant slope of 0.0028, with side walls of transparent Plexiglas. The vanes were made of Plexiglas sheets with thickness of 10 mm. A 3.67 m-long reach of the flume bottom was covered with sand of relative density 2.65 having median grain size d₅₀ of 0.5 mm. The depth of the sediment bed layer of the test reach was fixed at 10 cm (Fig. 1).

A subsurface reservoir supplies the required water flow (Q). A centrifugal pump discharges the water into the stilling tank at the entrance of the flume. In order to uniform the inflow water, a screen is established at 1 m distance from the flume entrance and then a transition with Δz = 10 cm, connects the main flume bottom to the surface of the sediment bed layer. A tail gate is used to adjust the depth (d_o) of water in the flume to a constant value of 25 cm, without causing backwater effects in the flume. The discharge is measured and adjusted by using a standard 53° triangular weir which was installed at the outlet system of flume. In this study 12 experiments carried out with three Froude numbers of Fr = 0.14, 0.16 and 0.20 and four vane shapes.

In each experiment, the vane was installed at the centerline of the flume at an angle of attack of 20° with the approaching flow. After placement of the vane, the sediment bed of the flume was leveled using a spirit level. Water flow was introduced to the flume very slowly by closing the tail gate so that no scouring occurred around the submerged vane during the startup. After the experiment was over, the scour and sedimentation pattern was measured in 2x2 mesh grids with the help of a laser distance meter that could be moved along rails fitted at the top of the flume. The time taken for the scour to attain equilibrium was 3 h.

The dimensions of the vanes are determined using the Odgaard and Wang (1991a) design criteria: whit a vane-height-to-water-depth ratio of H_o/d_o = 0.3 and aspect ratio of H_o/L = 0.3, the mean flow depth of d_o = 25 cm yields H_o = 7.5, L = 25 cm and vane relative submergence of T/d_o = 1-H_o/d_o = 0.7. The perspectives of the designed vanes are presented in Fig. 2a-d.

EXPERIMENTAL OBSERVATIONS

In present study both clear water and live bed conditions have been considered. In Fr = 0.14 and Fr = 0.16 clear water scour is occurred, thereby the sediment movement from upstream of the vane is negligible and flow velocity is less than critical velocity. The critical velocity was calculated by empirical formulae. In Fr = 0.20 the sediment transport form upstream of the vanes is considerable. Placement of the vane with 30° chamfer is shown in Fig. 3a-d.

Fig. 1:	Sketch showing experimental set-up

Fig. 2(a-d):	Schematic showing four types of vanes that are used in this study: (a) Simple rectangular vane (θ = 0°), (b) θ = 30°, (c) θ = 45° and (d) θ = 60°

Fig. 3(a-b):	Sediment bed (a) Initial level bed and (b) final topography

Figure 3a shows the initial level bed before introducing flow to the flume. Figure 3b shows the final topography of the sediment bed. According to Fig. 3, the vane function to remove sediments form left portion of the flume and deposit sediments on the right-hand side of the flume. Scour hole develops at the high-pressure side of the vane that vertical velocity components are generated. After primary changes in the sediment bed of the flume, the dune bed form is visible at downstream of the vane. Another observation is that the induced changes in the bed topography decay slowly in the downstream direction.

ANALYSIS

Investigation on the effect of vanes shape for local scour around the vanes and sedimentation pattern at downstream of the vanes is performed by using the topography data of the sediment bed in the test reach. The sediment bed topography was measured in a 0.7 and 0.8 m reach at clear water and live bed conditions, respectively. In addition, the transverse sediment bed profile measured at distance of s = 8H_o downstream from the center of the vanes. H_o is the vanes initial height.

Figure 4a-c show the graphs of transverse bed profile at leading edge of the vane for chamfer angles of θ = 0°, 30, 45 and 60° and Froude numbers of Fr = 0.14, 0.16 and 0.20.

Fig. 4(a-c):	Transverse bed profile at leading edge of the vane: (a) Fr = 0.14, (b) Fr = 0.16 and (c) Fr = 0.20. The symbols and dashed line represent the different chamfer angles and primary level bed, respectively. The vertical solid line represents the position of leading edge of the vanes in cross section at planed coordinate system

Table 1:	Summary of scour depth measurements at leading edge of the vanes

Graphs in Fig. 4 indicate that by increasing θ, the scour depth decreases. Figure 4a and b show the local scour at clear-water condition and Fig. 4c is related to live bed condition. As shown in Fig. 4c, at θ = 60° in Fr = 0.20, there is not scour and also some deposition has been occurred. Experiment condition at Fr = 0.20 is similar to field conditions due to the fact that in natural rivers there is often sediment transport from upstream watershed or river sections to downstream sections. Therefore, cutting the leading edge of the submerged vanes has effective role in reduction of local scour around the vanes and also it is practical and easy for field applications. According to Fig. 4, the scour depth at leading edge of the vanes (d_s) at three Froude numbers is summarized in Table 1.

Last row in the Table 1 represents the sour reduction percent respect to the d_s values for baseline rectangular vane at different Froude numbers. There is no scour at leading edge of the vane for θ = 60° at Fr = 0.20 and also some deposition has been occurred. Figure 5a-d present 3D plots of the bed topography around the vane in a 0.7 m reach including scour hole for four shapes of vanes (i.e., θ = 0, 30, 45 and 60°) and Froude number of Fr = 0.14. As is illustrated in Fig. 5 the scour hole becomes smaller for larger values of θ. The depth of initial level bed adjusted to a value of 10 cm, so according to Fig. 5a, for θ = 0° the maximum scour depth of d_sm = 5.7 cm. d_s is reduced to the values of 5.1, 4.5 and 3.8 cm (i.e., 10.5, 21.1 and 33.3% to the baseline rectangular vane) for θ = 30, 45 and 60°, respectively (Fig. 5b-d).

Contour maps of scour hole in a 0.8 m long reach at Fr = 0.20 and for different θ values are shown in Fig. 6. The vanes orientation is illustrated in this Figure. For rectangular flat-plat (Fig. 6a) the maximum score depth has been occurred at the leading edge of the vane. As indicated in Fig. 6b and c by increasing θ to the values 30 and 45° the scour hole is moved toward the trailing edge, along high-pressure side of the vane. From Fig. 6d the scour hole is moved to downstream of the vane and there is some deposition at leading edge of the vane.

In order to investigate the sedimentation pattern at downstream of the vanes, the transverse bed profiles have been compared at the distance of s = 8H_o. i.e., 296 cm downstream from the center of the vanes. H_o is initial height of the vanes (Fig. 7).

Fig. 5(a-d):	Bed topography in 0.7 m long reach including scour hole, for Fr = 0.14 at: (a) θ = 0°, (b) θ = 30°, (c) θ = 45° and (d) θ = 60°

Fig. 6(a-d):	Contour maps in 0.8 m long reach including scour hole, for Fr = 0.20 at: (a) θ = 0°, (b) θ = 30°, (c) θ = 45° and (d) θ = 60°. The dashed-dotted line and narrow rectangular represent the center line of the flume and top view of the vanes, respectively

Fig. 7(a-c):	Transverse bed profile at the distance of s = 8Ho from the center of the vanes: (a) Fr = 0.14, (b) Fr = 0.16 and (c) Fr = 0.20. The symbols and dotted line represent the different chamfer angles and primary level bed, respectively

Transverse bed profiles at downstream of the vanes for different values of Froude number, indicating that Sedimentation pattern at downstream of the vanes are not significantly affected by cutting the leading edge of the vanes. There is some differences in transverse bed profile at clear water condition (Fig. 7a, b) but due to sediment movement from upstream of the vanes in live bed conditions, the similarity of the bed profile is more visible (Fig. 7c). Similar behavior is founded for bed profiles at other cross sections (not shown). Deposition in the right half bank and scour at other portion of the sediment bed is illustrated in Fig. 7c.