A laboratory study of the effect of asymmetric-lattice collar shape and placement on scour depth and flow pattern around a bridge pier

The reduction of scouring is inevitably needed to prevent the scour around hydraulic structures in the path of watercourses (Pandey et al. 2021). Local scouring is caused by structures in the path of water flow on an erodible bed. Built structures can apply additional erosive forces on the bed around the structure. As a result, sedimentation and erosion rates increase locally around these structures, leading to cavities around these barriers (Foti & Sabia 2011; MacBroom 2012). In this regard, the formation of scouring cavities around the piers and supports of a bridge, which occur due to non-observance of hydraulic and river engineering issues in the design, is one of the main causes for the destruction of bridges (Suib et al. 2006). Scour is a highly important problem for rivers as well as for bridge piers. Bridges are vital structures which must be designed to prevent failure against scour effect. A scour hole can have a deleterious effect on a pier which without warning can lead to the failure of a bridge (Darshan et al. 2020). In addition to undermining the bridge foundation and changing the shape and form of the bed, these cavities affect the life and ecosystem of the rivers (Singh et al. 2019). Therefore, controlling the conditions governing real systems is usually difficult, and measuring the governing parameters involves great complexity. Thus, studying this phenomenon is important from economic and environmental aspects (Ministry of Power 2012). It is therefore necessary to consider hydraulic criteria in design studies during bridge construction programs. Nowadays, with the advent of modern flow measuring devices such as those used in Acoustic Doppler Velocimetry (ADV) or Particle Image Velocimetry (PIV), studies have been conducted on the effect of bridge piers and supports on flow structure (Wang et al. 2008). The importance of these measurements is that we can investigate the effect of different causes on the flow structure around the piers and supports of bridges through the development of these studies. Ettema et al. (2017) showed that further research concerning the field in which water flows around the pier and bridge supports could be useful in estimating scour depth. Therefore, in reviewing the present scouring experiments, the effect of parameters on the amount of scouring around a bridge pier and the flow structure around it has been investigated. Placing a bridge pier in a flow path, the simple and uniform pattern that reaches the pier undergoes drastic and complex changes. Hence, the vertical current formed at the bridge's pier is divided into ascending and descending sections. The surface wave formed on the surface of the water infront of the pier is caused by the movement of the rising current towards the surface of the water. The downflow on the surface of the sedimentary bed causes the formation of a scouring hole at the front of the bridge pier, which eventually creates a small scouring hole around it. A horseshoe vortex is a vortex-shaped system created by the rotation of a stream inside a hole. In addition, the separation of the approaching current from the sides of the bridge creates wake vortices behind the pier. The simultaneous effect of these two parameters increases the scouring rate and the potential for sediment transfer downwards. The protective effect of a collar on the bridge pier reduces the power of the downflow and the horseshoe vortices. Thus, the scouring rate is lower than the initial rate around the pier (Figure 1) (Guo et al. 2012). Muzzammil & Gangadhariah (2003) measured the strength, direction, and amplitude of horseshoe vortices around a pier. They concluded that during scouring the amplitude of horseshoe vortices initially increases to a maximum and then takes a decreasing course afterward. When a collar is installed on a pier to protect against scouring, the downflow is deflected from the bed as soon as it hits the collar, and scouring is prevented (Khozeymeh Nezhad et al. 2012).

Hong et al. (2015) measured lines with the same velocity, the three-dimensional shape of velocities, and turbulence values using ADV. They showed that the contractile flow around the abutment and the local turbulent structures near the downward part of a pier are important features of the flow field, causing the maximum depth near the abutment. Most studies conducted on scouring have focused on the final depth of scouring and methods to reduce it (Zarrati et al. 2004; Ghorbani & Kells 2008; Heidarpour et al. 2010). On the other hand, research studies on turbulent and surface flow patterns (Dey & Nath 2010; Tafarojnoruz et al. 2010; Izadinia et al. 2013) have been conducted focusing only on a single cylindrical pier. Dargahi (1990) investigated the mechanism of scouring around a bridge pier and how a collar affects the performance of downflows and ultimately reduces the scour of the bridge pier. Dargahi (1990) used two circular and oval (asymmetrical) collars, and chose the position of the collars from the bed such that: ⁠. His experiments were carried out in the conditions of ⁠. He observed a decrease in scouring velocity in the case with a collar and the greatest decrease in the depth of the hole with an oval (asymmetric) collar under the bed. On the other hand, the maximum reduction in the depth of the hole, in this case, was 50 and 70% upstream and downstream of the bridge pier, respectively. He was able to reduce the scouring depth up to 40% with a circular collar. Memar et al. (2020) investigated the level of collar installation from the sedimentary bed surface in reducing scour and the impact of the intensity of flow on this phenomenon. By reducing the flow intensity from 0.95 to 0/9⁠, the maximum depth in the case with collars decreased by an average of 20 to 70%. They also found that installing the collar below the surface of the sediment bed compared to installing it 1 cm above the bed provides 5% protection of the scour hole. Pandey et al. (2020) investigated the effectiveness of a symmetrical collar on reducing scour holes around the pier of a cylindrical bridge. The results of their study showed that the performance of using the collar reduces the maximum scour around the pier compared to the case without the collar by about 60%.

In addition, their findings revealed that the effectiveness of the collar decreases as the level of crown placement increases on the bridge pier relative to the sedimentary bed surface. Bestawy et al. (2020) examined the performance of different shapes of circular (jagged) collars. The results showed that crowns have a significant effect on reducing scour by inhibiting the destructive downflows around the pier. They stated that a Sigma_Slot collar showed the most effective performance by reducing the amount of scouring upstream and downstream by 59.3% and 52.8%, respectively. Pandey et al. (2018) presented experimental results of time average velocity components measured around circular pier models during transient scour stage for flow pattern and turbulence, using ADV. Conditions in a model of a gravel bed stream with four circular pier models of diameter 6.6, 8.4, 11.5 and 13.5 cm were used for this study. Also, they investigated scour holes at 0° and 180° plane. in that case study: at 0°, 61% larger than that for the smallest diameter pier model, whilst 180° planes are also presented around each pier. Bakhshpuri & Yahyaei (2016) evaluated the performance of a collar in reducing the scour depth of the piers in cylindrical bridges. Circular and square-shaped collars with dimensions of one, two, and three times greater than the pier diameter were examined in a channel with dimensions of 6 meters in length and a width of 0.7 meters. Based on the above studies, it is inferred that installing a collar at lower levels and larger collar dimensions has the greatest effect on reducing the scour depth. Wang et al. (2019) selected river sand with a median particle size of 0.324 mm and used it as sediment. According to their experimental results, it can be concluded that: the application of an anti-scour collar alleviated the local scour at the pier; and the protection decreased with an increase in the collar installation height, but increased with an increase in the collar external diameter and the protection range. Jalili & Ghomeshi (2014) investigated the effect of lattice collars on scouring the foundations of cubic and cylindrical bridges with side lengths and diameters equal to 4 cm. Experiments were conducted using 3 Froude numbers of 0.19, 0.16, and 0.13 under clear water conditions. To simulate the collar, square Plexiglas plates with a side length of 3B and a circle with a diameter of 3D were used. In this study, the performance of four simple collars, and 15, 30, and 40% lattice collars for scour hole changes were evaluated. With a Froude number of 0.19 for the 30% lattice collar with a cubic pier, the highest efficiency was 47% in reducing scouring. By contrast, for the cylindrical pier with the same Froude number, the 40% lattice collar has the highest efficiency of 34% in reducing scouring compared with other lattice collars. Gogus & Dogan (2010) investigated the effect of collar installation level above the bed surface, level with the bed and below the bed, on reducing scouring of bridge piers, based on the results of 97 experiments performed in a canal 1.5 meters in width, 30 meters in length and 1 meter in height and a slope of 0.001; different collar sizes were examined at +1 and +2 levels above the bed surface and level with the bed and −1 and −2 below the bed. The results showed that increasing collar width below the bed surface reduces collar efficiency in resisting scouring by 30% compared to the control. Increasing the installation level leads to increased scouring holes. Since the shape of the bridge pier has an effect on the amount of scouring, and given that the scour pattern is, based on the scientific literature, definitely affected by the three-dimensional field of the flow pattern around the pier, in this study, the scour hole and patterns of the flow field have been investigated using different shapes of asymmetric lattice collars by placing them at different levels on the pier of a cylindrical bridge to control or reduce scouring. The role of collars on scour depth and investigation of reducing scour holes around bridge piers are the main purposes of this paper. So, one of the objectives studied in this case is the position of the collar above the sediment bed, and also the role of asymmetric-lattice collar shapes on decreasing scour around a cylindrical pier. So far, the effect of an asymmetric-lattice collar shape has not been tested. The motivation for this paper is to find the best collar shape to debilitate vortices. Also, the performance of collar shape is examined for maximum scour depth around a bridge pier.

To conduct the scour experiments, a rectangular canal with a glass wall was used in the physical and hydraulic models laboratory of the Faculty of Water & Environmental Engineering of the Shahid Chamran University of Ahvaz (Figure 2). The length of the flume was 6.0 m, and its width and height were 0.72 and 0.6 m, respectively. The slope of the laboratory canal was adjustable, and it was set to a slope close to zero for the experiments. The canal contains an inlet pond at both ends. To measure the flow rate, a magnetic ultrasonic flow meter with an accuracy of 0.001 liters per second was used, which was installed at one end of the canal. The required water is pumped from the main tank to the canal using a pump, and a joint netted mesh is used to eliminate the turbulence of the inlet current. A sliding valve is designed to control and regulate the water level at the end of the canal, through which water can be returned to the tank, drained, and rotated in the system. Figure 3(a) and 3(b) show illustrations of the flume and the laboratory equipment.

To conduct the experiments, a cylindrical wooden pier with a diameter of 40 mm and a height of 50 cm was used to prevent the pier from sinking. To ensure the installation of the collar at the desired height from the bed surface, the pier was calibrated using a laser cutting machine with an accuracy of 1 mm. By examining the scour on the collars, Dargahi found that the large thickness of the collar creates a barrier to the flow and increases the scouring (Dargahi 1990). Therefore, the collars were prepared in Plexiglas plates with a thickness of 2 mm and were attached to the pier using silicone glue. Considering the results of many studies conducted on simple collars, Singh et al. (2009) proposed that symmetric collars with could be used for the cylindrical piers. Therefore, asymmetric collars with protrusions three times larger than the pier diameter towards the downstream of the canal, which has the greatest effect on scour reduction, were experimented with. Thus, to investigate the effect of the asymmetric lattice collar in this study, all collars were prepared from elliptical plates with and dimensions in 15, 30, and 40% lattice openings, in two models of 4 and 6 times larger than pier diameter in the longitudinal direction to weaken the downward currents around the pier (Figure 4). In the current study, the experiments were conducted for three different heights: on the surface of the sediment bed Zc = 0, 1 cm above the bed Zc = 0.25D, and 2 cm above the bed Zc = 0.5D and exposed to three different flow rates including 35, 30 and 25 liters per second, and Froude numbers equal to with 0.37, 0.32 and 0.26, respectively.

At a distance of 2 meters from the beginning of the canal, a box equal to the width of the canal with a length of about 2 meters and a height of 15 cm was introduced. The box was filled with a uniform sediment sample with an average diameter of 0.73 mm so that, after filling, it became parallel with the channel bed level. To achieve the depth of scouring equilibrium around the bridge pier, it is necessary to perform the experiments over a relatively long time so that changes in scour depth are insignificant over time, and the slope of the scour time development diagram is inclined towards zero (Cardoso & Bettess 1999). To determine the equilibrium time of the experiments, a 24-hour experiment with a relative speed was performed. During the experiments, changes in the sediment bed surface were controlled. Equilibrium time (8 hours) was obtained based on the measurements according to the initial level and scouring cavity formation. Then, at the beginning of each experiment, the canal was first filled at a low-flow rate; this is to prevent erosion caused by the laminar flow at the beginning of the experiment. The rate of the water flow was increased slowly to reach the desired flow rate. The experiments were performed at a constant current depth of 0.12 m, and after the end of the experiment, water slowly flows out of the canal. Finally, the maximum scouring depth and the sedimentation pattern created around the bridge pier were collected using a laser meter.

Furthermore, to ensure that the shape and the geometric dimensions of the scour cavity do not change during the velocity data collection, a very thin layer of cement and rock powder, combined in a 1:3 aspect ratio, was sprayed on the bed to stabilize it. Velocity data were collected using an ADV velocity meter at 200 Hz (Figure 5(a)). To study the flow pattern, it is necessary to measure and collect data around the pier; therefore, a specific network was considered around it according to Figure 5(b). Furthermore, three-dimensional velocity components were removed from the initial sediment bed of the canal at depths of 5, 6, 8, and 10 cm. At each point, more than 500 continuous speeds were recorded in three dimensions using a device. Over all, more than one million three-dimensional components of the velocity were recorded in the experiments at this stage. Moreover, the average of the continuous velocities at each point was also calculated. Thus, in each of the desired points, three velocity components in three different dimensions were calculated and used to study the flow pattern.

To investigate the effect of the presence of the collar on the scour of a pier of a cylindrical bridge, the experiments were divided into two categories as follows.

The experiments in this section were carried out according to Figure 3 to observe the effects of vortical currents around the bridge pier and to determine the scour maximum depth to be compared with the results of the experiments in the presence of a collar. In this case, the scour started from the front of the pier with the formation of a downflow and symmetrically according to the axis of the pier. As the scour cavity deepens and the horseshoe vortices become stronger, sediment is washed from the front and around the sides of the pier and accumulates in the form of a stack behind the pier. Moreover, as the scour cavity expands and creates a low-pressure area behind the pier, wake vortices are formed transporting sediments that have been carried by the flow downstream. In this regard, the scour cavity became wider by increasing the flow intensity, and the maximum scour depth increased. Table 2 and Figure 6 show changes in the dimensionless scouring depth and the pattern of erosion and sedimentation without the presence of a collar in the 8-hour experiment.

By installing the collar around the pier, at first, scour was created due to the presence of the wake vortices downstream of the pier. In this case, unlike in a pier-without-collar state, the horseshoe vortex was not observed from the beginning, and over time, grooves were formed on both sides of the pier at the edges of the collar. These grooves gradually expanded upstream and downstream, and their depth increased as well. The results of the experiments are presented in Table 3 at the setting level of ZC = 0, ZC = 0.25D and ZC = 0.5D.

In all experiments performed in the presence of a collar, the rate of scouring cavity formation was reduced compared to the control experiment according to Table 3 and based on the relationship as the percentage of scouring reduction. According to the results, it can be said that asymmetric collars were more effective in reducing scour than collars (Figure 7). Investigation of the flow mechanism around the pier shows that the wake vortices are activated downstream of the bridge's pier along with low-pressure centers for the sedimentation of the sedimentary deposits. Therefore, it seems that collars prevent some of these vortices from working, while in the collars these vortices can be more active, thus increasing the performance of the asymmetric collar. The installation of the collar not only leads to the reduction of the maximum scouring depth in front of the bridge pier but also reduces the height of the deposited sediments.

Another point is that the sediments do not exist up to a considerable distance downstream of sedimentary hills. In this regard, the lack of formation of sedimentary hills around the pier in the collar installation position is one of the advantages of the collar installation. This is because the improper distribution of sediments and the presence of sedimentary hills would result in a flow blockage between bridge piers, which in itself causes a kind of scour, especially contraction scour. During the installation of the collar, the same pattern of erosion and sedimentation occurred similar to that of a collar. On the other hand, while not reducing the depth of the scouring cavity to the same extent, the installation of the collar also reduced the height of sediments deposited downstream. Figure 8 also shows the longitudinal profile of the sediments around the pier in the conditions with and without a collar with a Froude number of 0.37.

The installation level of the collar has a significant effect on reducing scour around the bridge pier. As the installation level above the bed of the collar decreases, the performance of the collar increases. In all three distances of the collar installation on the pier, the scour cavity development was reduced compared to the control experiments. By placing the collar on the surface of the bed and preventing the formation of horseshoe currents under the collar, and increasing the time required for the development of the scouring cavity from the sides of the collar to beneath it, the maximum depth, longitudinal and transverse development of scouring cavity showed a significant reduction compared to the control state. At this level, by increasing the collar opening percentage, the maximum scour depth increased as well.

According to Figure 9 and Equation (4), as the distance of the dimensionless parameter of the collar installation level increases from the sedimentary bed, the scour reduction percentage decreases in the presence of the collar. The percentage of scouring reduction is determined using the results of control experiments and based on the maximum scour depth in the presence of the obtained collar. If the installation level increases again, this so-called parameter has little effect on the scour depth. Figure 10 shows the collar model scour pattern with 40% lattice at a Froude number of 0.26 for the two levels of installation on the bridge pier.

To become familiar with the velocity distribution at the maximum scouring point, in the without collar state or in the control experiment, three-dimensional velocity components were measured at different depths from the water surface to near the scour cavity floor. In this series of experiments regarding the velocity, the velocity profile has been considered in the depth direction and for a Froude number of 0.32. Figure 11 shows the longitudinal, transverse, and depth velocity profile. According to Figure 11, the maximum speed belongs to the transverse velocity, which is about 30 cm/s, although this velocity decreases during its way towards the bottom, reaching 28 cm/s. In terms of the depth velocity, it can be said that a downward current is formed from the water surface to a depth of about 3 cm, and its velocity is gradually reduced. When the velocity reaches zero, its direction changes, and an upward current is formed. This vortex, which is known as the horseshoe vortex, exerts a lifting force on sediment particles and the transverse and longitudinal velocities lead to the transition of the sediment particles raised from the bed.

In both experiments, the flow moved down near the water surface after colliding with the pier. Then the flow is diverted toward the main waterway near the bed after colliding with the floor. Near the bed, the velocity fluctuates less than in a collarless profile. The resulting vortex and downstream currents are the main causes of scouring around the pier. Since the wake zone does not help transfer fluid downstream, the current in its adjacent area moves rapidly, transferring excess fluid (Ahmed & Rajaratnam 1998). However, what is clear is that the presence of the collar has caused the spatial displacement of this parameter. In the presence of the collar, the created boundary layer flow could have performed better by delaying the separation of the flow from the collar surface in comparison with the created vortical currents. In the current study, an experiment was performed to determine the three-dimensional components and draw the flow pattern around the optimal collar installed on the bed surface. As shown in Figure 12, the horseshoe vortex indicated the least contact with the bed surrounding the pier, delaying the scour for a longer time. Of course, the horseshoe vortex washes the sediments around the collar. It penetrates beneath it, causing scour to happen, with this obtained scouring delayed compared with the collarless state. Furthermore, the intensity of horseshoe vortices in the optimal collar experiment has decreased compared to the control experiment. The flow pattern in the profiles shows that the presence of a downward current and the intensity of the formed horseshoe vortex is reduced in the front part of the bridge pier compared with the control experiment if the collar is used, causing more bed protection. Figure 13 shows the field of flow around the pier of a cylindrical bridge.

In Figure 12(a), it is observed that during the passing of the current along the path of the laboratory canal, opposite longitudinal pressure occurs near the pier after the passing of the gradient flow due to the collision of the current with the pier, which is located vertically in the path of the flow. This may cause the current to be separated from the pier and the particles to accelerate in areas where the pressure increases after the separation points of the wake return current. According to Figure 12(b), the use of the collar leads the return flow area to be kept away from the laboratory pier; by doing so, the collar creates the conditions for protecting the bed. The flow around the pier of a cylindrical bridge with a circular cross-section may create a uniform pattern of the vortex downstream. The end vortices cause the lifting force on the cylinder to fluctuate in the direction perpendicular to the movement of the flow lines, leading to the displacement of the sediment particles of the bed and moving them downstream. On the other hand, as the currents pass through the pier, they are joined together again, creating uniform flow conditions in the canal. It is also observed that by moving away from the surface of the sedimentary bed near the canal sidewalls, the velocity of the flow layers approaches the velocity of the free-surface flow.

In the present study, the performance of two collar models, namely and collars, was investigated to reduce positional scour around a bridge pier. The results showed that increasing the dimensions of the collars leads to the improvement of their performance. In the case of collar installation, the sediments around the pier have a uniform distribution. With a collar, the difference between the deepest and highest points of the bed is reduced to more than 50% of the case without a collar. On the other hand, by increasing the Froude number and the installation level of the collar, the scour increases in the areas around the pier.

The collars installed at a distance of 2 cm above the bed showed less efficiency regarding scour control. The collar model with 15% lattice installed at a distance of 2 cm and with a Froude number of 0.37 indicated the highest efficiency with a 44% reduction. By contrast, a collar with installation at a distance of 2 cm indicated the lowest scour control efficiency, which is related to the 40% lattice model with a scour rate reduction of 26%.

By examining the velocities around the collar, it can be said that the flow of the boundary layer created around the collars causes more sublation of the vortical currents, resulting in a smaller cavity around the pier. The collars investigated reduced the volume of the scour by affecting the boundary layer flow by delaying the separation of the flow. The current study also indicated the effects of a downstream flow and the resulting vortices in a bridge pier with a circular cross-section on the onset and development of the scour around it.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

The authors declare that they have no conflict of interest.

None.

This article does not contain any studies with human participants or animals performed by any of the authors.

Data cannot be made publicly available; readers should contact the corresponding author for details.