3.1 Experimental observations
The main reason for flushing is to evacuate deposited sediment from the reservoir. The pressure flushing through bottom outlets acts by two factors: the local effect of the bottom outlets, and the general flow movement in the reservoir. In reference tests, only the local effect of the bottom outlets erode sediment from the reservoir. This region which is near the bottom outlet is very limited and cannot contribute to sediment flushing from the bulk of the reservoir. In the reference tests in the current study, as the bottom outlet was opened, sediment scoured out close to the outlet. This scour at the opening formed; within a very short period of time after water flowing through the outlet was clear. These observations are consistent with the observations of White & Bettess [47]; Di Silvio [7] Powell & Khan [35] and Kondolf et al. [22].
However, in the presence of submerged vanes with different skew angles, this region extended, and sediment was evacuated with high intensity. Based on the features of flow around submerged vanes observed from video recording combined with some described by Odgaard & Kennedy [29]; Odgaard & Wang [32] and Odgaard [33] and the erosion pattern we observed in our experiments it was concluded that the flow downstream of the vane with high turbulence formed a pair of clockwise and counter- clockwise vortices. The circulating flows interacted with the secondary currents generated by the vanes, and rotational flow formed; thus, the sediments rotated with severe vortices. In this way, a groove was created between the bottom outlet and vanes, which directed the transfer of part of the sediments from upstream to downstream. Finally, deposited sediments collected in front of the bottom outlet and were evacuated. It should be mentioned that in this study the behaviour of flow around submerged vanes as a turbulent structure was similar to findings of previous studies which used submerged vanes in rivers.
Outflow discharge is an essential property of pressure flushing. In this study, the tests were carried out with different outflow discharges \({(\text{Q}}_{\text{o}})\) of 12.5, 15, and 18 L s− 1. As shown in Fig. 2, an increase in scour cone size with increasing discharge and Froude number was obtained. The increase in discharge through bottom outlets to flush sediments strengthens the pressure gradient and increases sediment removal. The vertical gradients of velocity and bed shear stress reinforce a pulling effect on sediment particles. Thus, a higher flushing rate and larger cones are produced [9]. This is consistent with the observations of several researchers [8, 11, 26]. For the range of outflow discharge in this study, an average increase of 1.44 times in discharge causes the flushing cone volume, length, width, and depth to increase by approximately 3.4, 1.6, 1.6, and 1.3 times, respectively.
3.2 The performance of vanes for skew angles of θ = 10°, 20°, 30°, 45°
To investigate the effects of different vane angles, the angles are classified into two groups: a) 0º < θ ≤ 45° and b) θ > 45゜. At θ = 10°, 20°, 30°, and 45°, which are in group (a), the reduction of θ resulted in better pressure flushing performance. It was observed that when the submerged vanes are at a slight angle to the flow direction, down flow and rotational currents are generated in front and at the sides of the vanes. Sediments move from the high-pressure zone in front of the vanes to the low-pressure zone behind the vanes. Additionally, due to the pressure difference on both sides of the vanes, rotational flow is created. The combination of rotational flow with flow velocity creates a spiral motion downstream of the vanes, which generates transverse shear stress in the bed. Due to the rotational flow downstream of the vanes, sediments are gathered and discharged from the bottom outlet. By reducing the angle of the vanes, the intensity of the vortices around the vanes is increased, which results in the displacement and transfer of more sediment from upstream to downstream of the reservoir.
By increasing the inclination angle of the submerged vanes, the longitude and transverse zones affected by the vanes are increased. The flow hits the front of the vanes and affects the back area less. For this reason, the sediments upstream of the vanes are picked up from the bed, and the sediments downstream of the vanes are transferred with less intensity. Finally, less sediment moves with spiral motion, and the dimensions and volume of the sediment flushing cone are reduced.
Figure 3 shows the variations in the length, width, depth and volume of the flushing cone at different skew angles of vanes with flow direction and minimum spacing of vanes. In this figure, the increase of flushing cone dimensions with the reduction of θ in group (a) experiments can be observed.
To better understand the increased amounts of sediment flushing and the cone dimensions at different skew angles of vanes, Table (2) is presented. It can be observed from this table that in all Froude numbers, in 0°< θ ≤ 45°with increasing angles, the dimensions and volumes of flushing cones decreased. Additionally, the sediment flushing cones generated by submerged vanes for θ = 10°, 20°, 30°, and 45° are illustrated with contour lines in Fig. 4.
Table 2
Percent increase of dimensions and volume of flushing cones compared to reference test for each discharge
\(Fr\) | Angle Classification | θ (°) | \(\frac{{L}_{hr}}{{D}_{o}}\) | ∆Lc (%) | ∆Wc (%) | ∆dc (%) | \(\varDelta \frac{{V}_{c}}{{{D}_{o}}^{3}}\) (%) |
1.6 | 0°< θ ≤ 45° | 10 | 1.5 | 155.1 | 58.333 | 77.78 | 578.41 |
20 | 1.5 | 144.9 | 50 | 77.77 | 575.68 |
30 | 1.5 | 104.08 | 16.667 | 55.56 | 188.86 |
45 | 1.5 | 83.673 | 4.1667 | 33.34 | 104.55 |
1.6 | θ > 45° | 70 | 0.5 | 104.08 | 66.667 | 77.78 | 549.09 |
110 | 0.5 | 221.43 | 166.67 | 155.55 | 2786.4 |
135 | 0.5 | 313.27 | 241.67 | 333.33 | 3850 |
150 | 0.5 | 313.27 | 239.58 | 388.88 | 4865.9 |
160 | 0.5 | 313.27 | 218.75 | 388.9 | 4695.5 |
170 | 0.5 | 277.55 | 175 | 277.77 | 2695.5 |
1.92 | 0°< θ ≤ 45° | 10 | 1.5 | 107.31 | 57.961 | 60 | 340.29 |
20 | 1.5 | 100.25 | 49.647 | 60 | 338.52 |
30 | 1.5 | 72 | 16.392 | 40 | 87.472 |
45 | 1.5 | 57.877 | 3.9216 | 20 | 32.75 |
1.92 | θ > 45° | 70 | 0.5 | 72 | 66.275 | 60 | 321.26 |
110 | 0.5 | 153.21 | 166.04 | 130 | 1773.3 |
135 | 0.5 | 216.76 | 240.86 | 290 | 2463.6 |
150 | 0.5 | 216.76 | 238.78 | 340 | 3122.9 |
160 | 0.5 | 216.76 | 218 | 340 | 3012.3 |
170 | 0.5 | 192.05 | 174.35 | 240 | 1714.3 |
2.32 | 0°< θ ≤ 45° | 10 | 1.5 | 104.78 | 54.92 | 34.27 | 238.14 |
20 | 1.5 | 96.59 | 46.77 | 34.27 | 236.78 |
30 | 1.5 | 63.83 | 14.15 | 17.4 | 43.98 |
45 | 1.5 | 47.44 | 1.92 | 0.53 | 1.95 |
2.32 | θ > 45° | 70 | 0.5 | 63.828 | 63.077 | 34.93 | 223.53 |
110 | 0.5 | 158.03 | 160.92 | 93.96 | 1338.7 |
135 | 0.5 | 228.21 | 234.31 | 228.9 | 1868.8 |
150 | 0.5 | 231.75 | 232.27 | 271.06 | 2375.2 |
160 | 0.5 | 231.75 | 235.9 | 271.06 | 2366.7 |
170 | 0.5 | 203.08 | 169.08 | 186.73 | 1293.3 |
To achieve the maximum size flushing cone, the proper minimum spacing of vanes (\({\text{L}}_{\text{h}\text{r}})\) needs to be considered. It was determined that with an increase in \(\frac{{\text{L}}_{\text{h}\text{r}}}{{\text{D}}_{\text{o}}}\), from 0.5 to 1.5, the dimensions and volumes of the flushing cones increased for skew angles of θ = 10°, 20°, 30°, and 45°. In this group, the vanes, due to the slight transverse section created between them, lonely have major interactions together, and getting closer to each other will disturb these interactions. At angles of 10°, 20°, 30°, and 45° with increasing minimum spacing of the vanes, the sediments are distributed on a larger surface, which causes the sediment to be transferred to the opposite side of each vane and thus changes the morphological position in the cross-section of the bed. Therefore, the sediment bed rises in one part of the cross-section and falls in the other part. By analyzing Fig. 3, it was determined that in the discharge of 18 L s− 1 and θ = 10°, the length, width, depth, and volume of the flushing cone increased 1.31, 1.3, 1.33 and 2.6 times increasing, \(\frac{{\text{L}}_{\text{h}\text{r}}}{{\text{D}}_{\text{o}}}\) from 0.5 to 1.5.
3.3 The performance of vanes for skew angles of θ = 70°, 110°, 135°, 150°, 160°, 170°
For skew angles of θ > 45°, scouring initiated between vanes; thus, flushing cones were produced and expanded upstream of the bottom outlet of reservoir and transverse sections. It can be concluded from the observations that the scouring was higher downstream than upstream of the vanes. Scour developed several seconds after the start of each experiment. By increasing the skew angle, the upstream side of each vane is entirely exposed to the approaching flow. Likewise, at smaller angles, the effect of the low-pressure side is reduced, but at higher angles, the interference of vortices on both sides of the vanes enhances the scouring rate around the bottom outlet.
Indeed, the role of the submerged vanes is to create a secondary circulation, and rotational flow is caused by changes in the vertical pressure on both sides of the vanes. Due to the effects on both sides of the vanes at higher vane angles, the combination of this rotation with the velocity in the flow direction causes a helical motion downstream of the vanes, which transfers sediment in a transverse direction [32]. Figure 5 shows an increase in the dimensions and volume of flushing cones from θ = 70° to θ = 160°. It is illustrated in this figure, the evolution of the flushing process was similar for skew angles of θ = 135°, 150°, and 160°. At these angles, due to the greater interactions of secondary currents with two zones of low pressure and high pressure, scouring originated downstream of the vanes, and flushing cones were generated, expanding to their upstream and downstream sections. The scouring rate increased significantly as soon as all the sediments around the vanes were washed away. The increases in the flushing cone dimensions for θ > 45°are given in Table 2. This table shows that in θ > 45°, increase of skew angles up to 160° increased the dimensions and volumes of flushing cones and the maximum variations of dimensions and volumes of flushing cones in tests with vanes compared to test without vanes were in 135° ≤ θ ≤ 160°.
Figure 5 highlights the influence of the skew angle on the flushing cones with the vanes. The vertical axis represents the length of the flushing cone, and the horizontal axis represents the width of the flushing cone. The analysis of Figs. 3 and 5 shows that with increasing vane skew angle, the sediment flushing cones increased for 135° ≤ θ ≤ 160°.
Also Fig. 3 shows that in group (b) reduction of \({L}_{hr}\) increases the dimensions of the sediment flushing cones. In θ > 45°submerged vanes act as individual vortex structures and generate vortices downstream. In fact, in this group, the vanes due to more created transverse sections than the vanes with θ ≤ 45°have fewer interactions with each other. So, the reduction of their distance can increase this interplay. When the vanes are placed near each other, vortices interact. These interactions develop interaction between the centrifugal force, the lateral pressure gradient, and circulation and consequently increase the strength of vortices. The intensity of the vortices increases the strength of the axial flow. Therefore, sediments are picked up with high intensity and carried towards the bottom outlet. Figure 3 shows that in discharge of 18 L s− 1, the length, width, depth and volume of the flushing cone increased 1.4, 1.5, 1.69 and 2.33 times with a reduction in \(\frac{{\text{L}}_{\text{h}\text{r}}}{{\text{D}}_{\text{o}}}\) from 1.5 to 0.5.
3.4 Effects of different skew angles of vanes on longitudinal and transverse sections of the flushing cone
To observe the performance of vanes with different angles on sediment flushing cone dimensions, according to the defined coordinate system shown in Fig. 6, longitudinal and transverse sections are drawn (Fig. 7). In this figure, the X, Y and Z -axes are standardized with the bottom outlet diameter (\({\text{D}}_{\text{o}}\)). By analyzing Fig. 7, it can be concluded that there are major differences in the scour cone size for 135°≤ θ ≤ 160° in both longitudinal and transverse directions compared to the reference test and test with different angles. This can be explained by the smaller angles having a narrower cross-section and, thus, the smallest influence in the transverse direction.
The bed slopes of the scour cones have also been studied extensively, and several comparisons have been made between the slope angles of the longitudinal and transverse axes. The general consensus is that for non-cohesive sediment, the scour cone slopes would be approximately the same as the submerged angle of repose of the deposits. Many researchers have noted that the transverse bed slope is slightly steeper than the longitudinal bed slope [10, 20, 27].
The longitudinal and transverse scour cone slopes for each test were determined from the deepest point in the scour profile up to its upstream end. The average longitudinal and transverse slopes in tests with different skew angles of vanes were obtained 29° and 33.8°, respectively. It was concluded that the transverse slope is steeper than the longitudinal slope, which is consistent with previous studies.