**4. Results and Discussion**

The test results show that the maximum scouring depth can be replaced by the maximum value of the measuring points reading. A local scour test of the cylindrical pier model without an anti-scour collar (the unprotected case) was conducted first to obtain the reference scour characteristics for the pier without any protection. The rest of the testing cases of the pier model with different anti-scour collar designs were then carried out and compared with the unprotected case to show the scour protection efficiency (factor β in percent).

#### *4.1. Characteristics of Local Scour for a Cylindrical Pier without Protection*

In order to eliminate the effect of geometry of the pier on the results [46], the dimensionless scour depth *S*/*D* measured at twelve measuring points under different scouring times are plotted in Figure 7a, respectively. The shape of the scour hole around the pier model is symmetric around the current flow direction. Because the flow velocity and bed shear stress behind the pier were smaller than that in front of the pier due to the shadowing effect of the pier, the scour depth is decreasing from the upstream (0◦) to the downstream pier side (180◦). The scour hole developed continuously with the testing time. At 10 min of scour, the maximum scour depth occurred at the 60◦ side of the pier, while the minimum scour depth appeared at the 180◦ position, which is just in the back of the pier. After 10 min, the position of the measured maximum scour depth gradually moved from the side to the front of the pier, and the final maximum scour depth at 120 min occurred at the front of the pier.

**Figure 7.** Depth scour data of a single cylindrical pier: (**a**) the time-varied scour development at twelve measuring points; and (**b**) the maximum scour depth of a single pier.

The maximum dimensionless scour depth *S*/*D* around the pier as a function of scour testing time is given in Figure 7b. In the first 30 min, scour occurred quickly; the maximum scour depth and the length and width of the scour hole increased rapidly. The growth rate of the maximum scour depth slowed down with the increase of testing time and reached close to zero in the approximate equilibrium condition. Similar phenomena have also been reported in the references [31,43]. The maximum dimensionless scour depth *S* of the sediment hole around the pier is 0.58 at 30 min and increases to 0.72 at 60 min, which is more than 90% of the scour depth at 120 min. Although scour still occurred in the later period, the growth of both scour depth and hole's range are very small. The minimum scour depth occurred at 150◦ at 120 min. As the photos of the scour hole at 120 min shown in Figure 8, the minimum scour depth is approximately 30% of the maximum scour depth of the final scour hole. According to the development of scour depth as a function of the experimental time for all working conditions, although the scour depth will still increase after 120 min in the tests, the increase of depth is very slight and the increasing rate is significantly slower than that of the first 60 min, which can be treated as the time to approximate equilibrium condition.

**Figure 8.** The scour hole in the approximate equilibrium condition.

The maximum bed shear stress and the maximum scour depth appeared in front of the upstream side of the pier [40,43]. Due to the blocking effect of the pier, turbulent flow is generated around the pier and accelerated the development of scour in front of the pier. They all found that the maximum scour depth occurred in the upstream front of the cylindrical pier.

#### *4.2. E*ff*ect of Collar Installation Height*

Testing of Cases 2 to 5 was conducted to investigate the influence of collar installation height on the scour protection effect. In these cases, the external diameter *W*/*D* and protection angle *a* of the anti-scour collar equaled 2.5 and 360◦, respectively. The installation height of the anti-scour collar *h*/*H* was set to be 0.1, 0.04, 0 and −0.04 for Case 2 to 5, respectively. The negative height indicates that the collar was located under the riverbed. The development of maximum scour depth *Smax*/*D* at different collar installation heights is shown in Figure 9a. The anti-scour collar reduced the local scour depth and had a protective effect on the cylindrical pier regardless of the installation height. The results of scour depth as a function of collar installation height agree with the experimental results of Chiew [24] and Ettema [30].

**Figure 9.** Scour protection effect at different collar installation height *h*/*H*: (**a**) the maximum scour depth development; and (**b**) the protection efficiency.

In order to assess the efficiency of the collar, we define a protection efficiency factor β = (*St,max* − *Si*)/*St,max*, where *St,max* is the maximum scour depth of the unprotected case in 120 min, *Si* is the maximum scour depth of Case *i* in 120 min. The larger the factor β is, the better the protective effect is. The protection efficiency factor β versus the installation height is given in Figure 9b. The photos of scour holes for Case 2 to 5 at *t* = 120 min are shown in Figure 10. The development of scour depth around the pier model as a function of collar installation height is illustrated in Figure 11.

**Figure10.** Thescourholeformofdifferentcollarinstallationheight:(**<sup>a</sup>**–**d**).

**Figure 11.** *Cont.*

**Figure 11.** The dimensionless scour depth development of different collar installation height *h*/*H* at each measuring point.

When the collar was located above riverbed, the closer the collar was to the riverbed, the better the protective effect. The protective effect is increased with the decrease of the installation height for all the measuring points around the pier. The maximum scour depth occurred at the 0◦ point, which located in front of the pier. The location of the maximum scour is the same to the model without an anti-scour collar.

When the anti-scour collar was located at the riverbed surface (*h*/*H* = 0), there was not any scour of sediment found at all twelve measuring points in the first 120 min testing. The anti-scour collar protected the sediment under the collar from being washed away, but the sediment at the downstream edge of the collar was washed away. Although the scour hole was developed to the sediment under the collar with the increase of scour testing time after 120 min, the maximum scour depth always occurred at two side points outside the collar behind the pier.

When the anti-scour collar was embedded into the riverbed (*h*/*H* = −0.04), the sediment above the collar was quickly removed away in the beginning of the test. The maximum scour depth equaled to the value of *h*. With the increase of testing time, the maximum scour depth occurred at two side points outside the collar behind the pier, which was similar to that of the case *h*/*H* = 0. Therefore, the maximum scour depth of *h*/*H* = −0.04 condition at the beginning stage was larger, and then was smaller than that of *h*/*H* = 0. According to the comparison of final scour depth, the case with *h*/*H* = −0.04 has the best protective effect. Moreover, there was almost no scour around the collar edge except that the sediment on the upper part of collar was washed out at the beginning, when the anti-scour collar was embedded into the riverbed.

It should be noted that the anti-scour collar cannot be embedded too deep. Otherwise all the sediment above the collar will be removed. Because of the existence of general scour and natural evolution scour, it is suggested to install the anti-scour collar at the general scour line to prevent local scour as much as possible.

## *4.3. E*ff*ect of Collar External Diameter*

The effect of the collar external diameter on the scour protective effect was investigated experimentally at a certain collar installation height *h*/*H* = 0.04. Four anti-scour collars with different external diameters (*W*/*D* = 3.0, 2.5, 2.0 and 1.5) were considered. The development of maximum scour depth *Smax*/*D* as a function of the collar external diameter is plotted in Figure 12a. The protection efficiency factor β (defined as (*St,max* − *Si*)/*St,max* in above section) versus the external diameter is shown in Figure 12b. It is clear that the external diameter of anti-scour collar affects the scour protective effect significantly. The protective effect is increased with the increase of the collar external diameter for all twelve measuring points. The scour development related to the increase of collar external diameter has also been reported in the previous studies [33–35].

**Figure 12.** Scour protection effect at different collar external diameter *W*/*D*: (**a**) the maximum scour depth development; and (**b**) the protection efficiency.

The photos of the scour hole at 120 min for four cases are given in Figure 13, respectively. According to Figure 14, the increase of the collar external diameter could reduce not only the maximum scour depth, but also the scour hole range. When the external diameter *W*/*D* was set to be 3.0, the edge of the scour hole was located inside the range of the collar. The sediment deposited behind the pier. With the decrease of the collar external diameter, its protective effect was decreased. When the collar external diameter *W*/*D* was 2.5, the scour hole range was similar with the collar size, which was larger than the case of 3.0 value diameter. With the collar external diameter decreasing, the scour hole was becoming larger than the collar size, and its range was larger.

**Figure 13.** The scour hole form of different collar external diameters at 120 min.

**Figure 14.** The dimensionless scour depth development of different collar external diameters *W*/*D* at each measuring point.

The development of scour depth as a function of scouring time at twelve measuring points is shown in Figure 14 for different anti-scour collar external diameters, respectively. The maximum scour depth occurred in front of the pier regardless of the external diameter. Although anti-scour collars with larger diameters resulted in better protective effects, it is not possible to install a collar with infinite external diameter due to hydraulic and economic reasons. According to this study, an anti-scour collar with an external diameter that equals three times the diameter of the pier can reduce more than 50% of the scour for the pier without any protection. However, it should be noted that there should be a limit between the scour reduction and the cost due to the increase of collar diameter, especially when it is applied in the actual long-span bridge pier with a large diameter [47].

## *4.4. E*ff*ect of Collar Protection Range*

Considering downward flow in front of the pier is one of the reasons that causes local scour of sediment. In the existing literature, there are many studies on the installation height and external diameter of anti-scour collars [24,25,27,29–33], but there are few studies focused on the protection range of anti-scour collars. It is interesting to determine if we can protect the sediment by reducing the protective range of the collar. In order to discuss the effect of the protective range of the collar, three protective ranges: 180 (half circle), 270 (3/4 circle) and 360 (full circle) are considered on the basis of anti-scour collars with installation height *h*/*H* = 0.1 and external diameter *W*/*D* = 2.5.

The development of maximum scour depth *Smax*/*D* for different collar protection ranges are given in Figure 15a. The protection efficiency factor β (=(*St,max* − *Si*)/*St,max*) versus the protection range of collars is shown in Figure 15b. It can be concluded from the results that the protection range affects the protective effect as well. The scour protective effect of the collar with 360◦ protection range was the best. With the decrease of protection range, the protection effect was weakened.

**Figure 15.** Scour protection effect at different collar protection range: (**a**) the maximum scour depth development; and (**b**) the protection efficiency.

The scour holes formed at 120 min of different protection angles are shown in Figure 16. The photos of scour holes at 120 min for four collar protection ranges are given in Figure 15. The development of scour depth as a function of scouring time at twelve measuring points is shown in Figure 17. When the protection range of the collars decreased from 360◦ to 270◦, the protective efficiency decreased distinctly. When it continued reducing from 270◦ to 180◦, the decrease in protective effect did not greatly change. Therefore, it is recommended to use anti-scour collars with a full circle protection range.

**Figure 16.** The scour hole form of different collar protection angle at 120 min.

**Figure 17.** The scour depth development of different collar protection angles at each measuring point.
