Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
686.1 IntroductionThis chapter presents observations and data from sets ofbrief, preliminary laboratory experiments conducted to evalu-ate scour countermeasures for protecting abutments. Theexperiments were carried out using small-scale replicas of a sim-ple abutment form: a wing-wall abutment extending at a depthinto a cohesionless bed of a rectangular channel. As mentionedat the outset of this report, wing-wall abutments are commonlyused for short bridges, such as those that span relatively narrowchannels. Given the large number of small bridges, especially inthe U.S. Midwest, the great majority of abutment failures haveoccurred for small single- or double-span bridges that com-monly have wing-wall abutments. Accordingly, it was thoughtuseful to expend laboratory effort exploring the responses ofsuch abutments to various scour countermeasure concepts.In most cases of subsequent bridge repair, and increasinglyfor the design of new bridges, it is usual to consider use of aprotective armor layer placed to prevent erosion of the chan-nel bed and bank around abutments. Also, to a lesser extent,consideration is given to wing-wall angle and abutment align-ment (relative to the channel crossed) to minimize scour.Adjustment of angle and alignment would seek to minimizelocal flow velocities and turbulence in the vicinity of the abut-ment, thereby reducing scour. The experiments focusedchiefly on the use of armoring countermeasures and to a lesserextent on flow-altering countermeasures. The experimentsinvestigated the performance of armor elements, aprons ofriprap and geobags placed around pile-supported wing-wallabutments retaining erodible embankments, and subject toclear-water flow conditions. Also investigated were the influ-ences of wall angle and abutment alignment on scour depth.In particular, the exploratory experiments investigated thefollowing questions:⢠Are there simple configurations of large armor units thatcould be an effective scour countermeasure method forwing-wall abutments?⢠Can aprons of smaller armor units or riprap be used as ascour countermeasure for wing-wall abutments, and, if so,to what extent should a riprap apron extend around awing-wall abutment?⢠How do large geobags perform as an alternative to riprapor cable-tied blocks for preventing abutment scour?⢠How does the wing-wall angle of an abutment affect scourdepth?⢠How does the abutment alignment to a channel affectscour depth?The findings to these questions consist of general observa-tions and small-scale laboratory data about armor unit,riprap, and geobag performance at small bridges.6.2 Program of ExperimentsIn accordance with the set of questions enumerated above,the program of exploratory experiments consisted of the fol-lowing four series of experiments:⢠Experiments on the scour countermeasure effectiveness oflarge blocks,⢠Experiments on the use of large geobags,⢠Experiments on the scour influence of wing-wall angle,and,⢠Experiments on the influence of abutment alignment onscour depth.The experiments were heuristic (i.e., trial-and-error dis-covery) and exploratory in nature. They were carried outusing a simple wing-wall abutment to explore the efficacy ofusing large armor units as a scour countermeasure. The unitsconsisted of two sizes of concrete block, one or more largegeobags, and a combination of large geobag and riprap stone.The use of large armor units held particular practical ini-tial appeal because such large blocks would not be moved bythe flow and because their roughness and bulk would redirectC H A P T E R 6Lab Results I: Preliminary Experiments
flow partially. Also, placing and positioning blocks around anabutment would seem relatively practicable, even in flowingwater.The test armor units were tried in various combinationsand layout extents to gage the sensitivity of scour develop-ment and depth with respect to the placement and location ofindividual large armor units. The experiments were heuristic,involving considerable adjustment and exploration of armorunit placement. Only a representative overview of the exper-iment results need be mentioned herein. The experiments arefully documented by Martinez (2003).6.3 Use of Large Blocks 6.3.1 Experiment LayoutA simplified configuration of wing-wall abutment was usedfor the experiments,which were all conducted using a laboratoryflume at the University of Iowa. The overall layout and dimen-sions of the flume are given in Figure 6-1, which also indicatesthe location of the test region in the flume. A sand-roughenedfalse-floor approach conveyed flow to the sediment recess mak-ing up the test section. The test abutments were placed in thesediment recess region. An overall view of the flume is shown inFigure 6-2, which also depicts the sediment recess.The preliminary experiments were done under conditionsof clear-water scour, with u*/u*c 0.8, where u* is the shearvelocity and u*c is the critical value of the shear velocity asso-ciated with bed-particle movement. The main hydraulicparameters for the flume flow were the following: meanvelocity, V0 0.55 m/s; and flow depth, Y0 0.10 m. The sed-iment parameters were the following: median particle size,d50 0.45 mm; standard deviation of sediment size,g 1.4;specific gravity of particles 2.4; and the angle of sedimentrepose, r 30 degrees.Similitude between laboratory experiments and field scalewas satisfied by the use of the aforementioned u*/u*c ratio, ofwhich a value of just below 1.0 represents a condition calledâclear-water scour.â This condition is extreme for scouringbecause the velocity is as high as possible without the move-ment of the channel bed, which causes infilling of the sedi-ment hole.The layout and dimensions of the wing-wall abutmentused are given in Figure 6-3. The abutment was made of asimple approximate form, in keeping with the exploratorynature of the preliminary experiments. The figures indicatethe layout extents of the armor units placed around the testabutments.Also indicated in Figure 6-3 are the two locations wherescour depth was greatest. Point A is at the face of the abut-ment, and Point B is somewhat downstream of the abutment.A consequence of extensive armoring of the bed around theabutment was that the location of deepest scour was forceddownstream.69Figure 6-1. Layout and dimensions of the flume, including the false floor and sediment recess.
6.3.2 ObservationsThe experiments showed that single large individual armorunits, or ensembles of blocks (or such units as dolos andtetrapods), alone, are of limited effectiveness as a scour coun-termeasure. Scour of the bed sediment around the armorunits diminished armor unit effectiveness as a scour counter-measure. Figure 6-4 shows the scour that formed around thewing wall without a countermeasure. Figures 6-5 and 6-6show before-and-after photos of the 22-mm blocks and11-mm blocks, respectively. Table 6-1 lists the scour depthsfrom three of the experiments. Without the presence of theblocks, a scour depth (dsA0) of 140 mm developed at the faceof the abutment (Point A).Flume observations showed that edge erosion of bed sedi-ment occurred around the blocks, caused the formation of alocal scour hole around each exposed block, and that theblock subsequently slid into the scour hole. As water flowedpast the blocks, vortices were shed, which entrained bed sed-iment from around the blocks. Bed sediment particles werewinnowed through the gaps of the overlying concrete blocks,causing the local scour hole to expand in area and eventuallyenvelop the blocks.The placement of concrete blocks reduced scour depth atthe abutment. Ten large concrete blocks (of side length22 mm) that were placed around the abutment as shown inFigure 6-5 reduced the scour depth by 32.2 percent com-pared with the baseline scour depth at the unprotected70Figure 6-2. View of the sediment flume.Figure 6-4. View of scour hole formed at wing-wallabutment.Two sizes of blocks made of cement and sand were testedin the flume: blocks with 22-mm side lengths and blocks with11-mm side lengths. The specific gravity of the blocks wasestimated as 2.30. The blocks were placed in differentarrangements to investigate as a scour countermeasure.Figure 6-3. Layout and dimensions of simplewing-wall abutment used in preliminaryexperiments, L 160 mm, Ha 32 mm.
abutment. The large blocks, acting as exposed large ele-ments, produced locally increased flow velocities and tur-bulence, such that bed sediment readily scoured fromaround the blocks. In due course, the scour hole developedaround the abutment, and the blocks gradually slid towardthe base of the scour hole.An important finding is that the smaller concrete blocks(of side length 11 mm) covering the same area as the largeblocks performed essentially the same in reducing scourdepth. The equilibrium depth of scour was identical to thatconducted with large concrete blocks. In other words, pro-vided that the blocks were not entrained by the flow, blocksize is less important than the extent of bed covered andthe presence of a filter-cloth underlay to reduce the win-nowing of bed sediment. A critical consideration thatemerges from the experiments is that the size of the block71(a) At start of experiment(b) At end of experimentFigure 6-5. Experiment with large (22-mm) blocksplaced at front of wing-wall abutment.(a) At start of experiment(b) At end of experimentFigure 6-6. Experiment with small (11-mm) blocksplaced at front of wing-wall abutment.chosen must be large enough to resist shear erosion, yetsmall enough to substantially reduce any winnowing ofbed-form sediment.To demonstrate the influence of aerial coverage on scourdepth, two additional rows of the smaller (11-mm) blockswere placed upstream of the abutment and perpendicularto the flow direction, thereby increasing the coverage. Thearrangement is shown in Figure 6-6. This experimentshowed that dsA was reduced by 60 percent. The reduction
72Table 6-2. Local scour depths at wing-wall abutment with geobag.Table 6-1. Local scour depths at wing-wall abutment with concrete blocks.a flat profile and rounded edges, thereby reducing local accel-eration of flow velocities around the geobag. Also, a largegeobag essentially provides its own filter cloth base as well asacts as an armoring layer. A further possible advantage of ageobag is the prospect of making a geobag that conforms to adesired shape and size for particular abutment sites. Experi-ment-scale geobags of approximately equivalent weight wereused as the large blocks and were sized as 90 mm à 70 mm Ã18 mm. The geobags were densely filled with sand.6.4.1 ObservationsTable 6-2 summarizes the results of the test with a singlelarge geobag. While winnowing erosion did not occurbetween the geobag and the abutment, edge failure remainedan unresolved concern. Figure 6-7 shows the formation of alarge scour hole adjacent to the geobag, into which the geobagslid. Note that the geobag setup in this experiment is hingedto the abutment; otherwise, it would have slid completely intothe scour hole.The experiment showed that, though the geobag protectedthe abutment, scour continued at a location shifted awayfrom the abutment to a location downstream of theabutment. Accordingly, two values of scour reduction need tobe considered: one at the abutment, dsA, and the other at thein scour depth is attributable to the increased area of bedprotection around the abutment. The placement of tworows upstream of the abutment helped to minimize ero-sion of bed sediment from around the leading edge of theabutment, thereby resulting in an enlarged extent of scourhole, but a shallower depth of scour. Furthermore, whenthe scour hole eventually developed fully around the abut-ment, the larger number of blocks provided greater cover-age of the base of the scour hole, thereby reducing scourdepth.The two mechanisms of scour reduction explained aboveproduce a much shallower scour hole. These experimentalresults agree with prior observations on riprap stones as apier-scour countermeasure (e.g., Chiew 1995), where suffi-cient riprap stones could significantly reduce winnowingfailure.6.4 Use of Large GeobagsThe main problems concerning the use of large armorunits, such as concrete blocks, for scour reduction are thewinnowing of bed sediment around blocks and edge erosionaround the blocks. To reduce these problems, experimentswere carried out to examine the use of a large geobag formedfrom geotextile fabric and filled with sand. A large geobag hasLayout X +/L X â/L Z +/L Z â/L dsA (mm)dsA/dsA0(%) No blocks 0 0 0 0 140 100.0010 blocks a1/L = 0.13 0.33 0.33 0.27 0 94 67 40 blocks a1/L = 0.13 0.33 0.33 0.27 0 94 67 70 blocks a2/ L= 0.07 0.33 0.33 0.27 1 56 40 dsA = scour reduction at the abutment with scour countermeasure.dsAO = scour depth at the abutmnet without scour countermeasure.Layout X +/L X â/L Z +/L Z â/L dsA (mm)dsA/dsA0(%) dsB (mm)dsB/dsA0(%) No bag 0.00 0.00 0.00 0.00 140 100 140 100 bag 0.69 0.69 0.50 0.00 84 60 102 73 bag+rock 0.69 0.69 0.50 0.00 60 43 102 73 bag 0.69 0.69 0.50 1.00 52 37 100 71 bag+rock 0.69 0.69 0.50 1.00 58 49 104 74 bag 0.69 0.69 1.00 1.00 0 0 112 80
maximum deepest point of scour, dsB. The values for thesetwo locations were 0.40dsA0 and 0.27dsA0, respectively.6.4.2 Geobag and Riprap StoneIn an effort to control edge erosion, simulated riprapstones (median diameter d50 8 mm) were placed around thegeobag. Minor improvements resulted such that dsAR at theabutment was increased from the original 40 percent to 60 percent, though the dsB at the deepest point of the scour holeremained almost the same. Therefore, placing loose riprapstones around the geobag in order to prevent edge failure hadmarginal success. Figure 6-8 shows how the geobag was at riskof sliding into the scour hole.When the area of geobag protection was enlarged aroundthe abutment so that the geobag covered the bed beneath thelarge-scale turbulence structures generated by flow aroundthe abutment, the geobag completely prevented scour devel-opment at the nose of the abutment, but the scour holedownstream of the abutment persisted. A further experimentinvestigated whether the placement of riprap stones on thegeobags would reduce the depth of the scour hole down-stream of the abutment. The idea explored in this experimentwas whether the riprap on the geobag would roll into thedeveloping scour hole and consequently retard its deepening.The results from both tests show that the deepest point of thescour hole is about 0.48dsA0. No scour occurred at the nose ofthe abutment. While the formation of the scour hole down-stream of the abutment seems to be unavoidable, the presenttest shows that armoring the bed would be able to control itsdevelopment and protect the scour countermeasure.6.5 Wing-Wall Abutmentand Geobags Experiments with a wing-wall abutment and geobagsentailed the same flume conditions as those used for theexperiments described in Section 6.3. However, now the abut-ment was of wing-wall shape. The wing-wall abutment formused for the experiment replicated, at a scale correspondingto approximately 1:40, the width of abutments typical of two-lane roads in the United States when the road width is about12 m (40 ft). The abutmentâs wing-walls were set at an angleof 45 degrees. Figure 6-9 shows the dimensions of the modelabutment used.Table 6-3 shows the ratio among geometric variables as wellas scour depths for four of the experiments conducted. Thescour at the unprotected abutment is shown in Figure 6-10.Scour was deepest at the abutment face.When a large geobag was placed around the wing-wallabutment, scour did not occur at the abutment face; that is,73(a) At start of experiment(b) At end of experiment, view from above(c) At end of experiment, side viewFigure 6-7. Experiment with a single largegeobag placed at front of wing-wall abutment.
scour reduction was 100 percent (dsA 0). However, turbu-lence generated by flow around the abutment and over thegeobag eroded the sand bed immediately downstream of thegeobag, thereby shifting the scour and creating a deeper scourhole. Figure 6-11 depicts the initial state and the eventualscoured state of the bed. The erosion of the bed at the down-stream edge of the geobag gradually propagated upstreamaround the edge of the geobag. It is noteworthy to point outthat this process of edge erosion was observed to occur for allthe experiments with geobags. The deepest scour hole for thisexperiments was dsB 143 percent of dsA0. Its location isshown in Figure 6-11(b) and (c). As the scour hole reachedthe downstream edge of the geobag, an additional row ofgeobags was used to further reduce the scour. Figure 6-12shows the initial condition and the eventual scour condition.Although scour was eliminated at the abutment, the scourhole immediately downstream of the abutment and geobagsremained, though it was somewhat shallower. Figure 6-12shows that the extra row of geobags diminished the erosionattributable to wake vortices. The maximum deepest scourwas 119 percent of the scour depth at the unprotected abut-ment (dsA).In addition, when a fringe of riprap stone was placedaround the geobags in an effort to limit edge erosion, themaximum scour depth was reduced further to 97 percentof dsA. The stones provided partial armoring of the scourhole.Dune-bed conditions pose the severest test for the stabilityof an armor cover, such as riprap or geobags, because the pas-sage of dunes may dislodge portions of a cover. This certainlywas found in the present study, and it is amply shown forefforts at armoring beds around piers (e.g., Chiew, 2000). It isof interest to note that existing guidelines for riprap designare based on laboratory experiments performed exclusively inclear-water scour and do not account for the dislodgingeffects of bed forms passing the riprap.74(a) At start of experiment(b) At end of experiment, view from above(c) At end of experiment, side viewFigure 6-8. Experiment with a large geobag placedaround the wing-wall abutment and with stoneplaced along geobag edges.Figure 6-9. Dimensions of simple wing-wall abut-ment used in preliminary experiments; L 160 mm, b 160 mm, 45 degrees, ra 160 mm, thicknessof geobag layer 20 mm.
6.6 Influence of Wing-Wall Angle A series of experiments was conducted in which the wing-wall angle was varied. No additional scour countermeasure wasused in these experiments. The angle (Figure 6-13) was set at15, 30, 45, 65, and 90 degrees to the flow.Figure 6-13 and Table 6-4 show the resulting trends for thevariation with of equilibrium scour depth at the abutment,dsA. The values of dsA are normalized with dsA obtained for the90-degree wing wall (i.e., the vertical wall). Scour depthsreduced as decreased. As is to be expected, a smaller angleof wing-wall produces less velocity component normal tothe wall. Consequently, the strength of the horseshoe vortexin the scour hole was reduced. Also, the intensity of waketurbulence was reduced. Figure 6-13 shows that the reduc-tions in scour depth are substantial, at least for the length ofabutment used in the experiments; for instance, the scourdepth using a 15-degree wall angle was only 23 percent of thescour depth that developed for a 90-degree (vertical-wall)wing-wall abutment.The findings on wall angle presented here indicate thescour-reducing merit of (a) decreasing the bluffness of anabutmentâs upstream profile and (b) streamlining the down-stream profile to greatly weaken wake vortices. The findingsdo not necessarily imply that angling the approach of a wing-wall abutment produces the same extent of scour depthreduction, because the downstream side of an angled abut-ment may still produce strong wake vortices. Also, as pointedout by Dongol (1994), reducing the scour at one abutment byreducing its angle to the flow may aggravate scour at theopposite abutment on the river bank; the opposite abutmenthas an adverse angle to the approach flow. This concern, how-ever, applies to long abutments that substantially contract theflow at a bridge crossing. It is not a concern that affects shortabutments, such as wing-wall abutments.6.7 Influence of AbutmentAlignment A brief further set of exploratory experiments examinedthe influence of abutment alignment on scour depth. Theseexperiments, conducted for the present study but using a dif-ferent flume than that shown in Figure 6-1, are reported byMartinez (2003). The corollary question addressed by theseexperiments is whether scour depth is minimized or aggra-vated by aligning a bridge at some angle other than 90 degreesto a channel. The experiments were conducted with a thinwall replicating a simplified abutment.Figure 6-14 shows that the scour depth, dsA, increased asalignment angle increased from 15 to 90 degrees, and then thescour depth decreased as the angle further increased from 90 to 150 degrees. The variation of dsA with angle appears to bealmost symmetrical for alignments upstream or downstream.For all angles, the deepest scour occurred at the end of theabutment. Dye observations from the present experimentindicate that the downflow and horseshoe vortices around theend of the abutment weakened as the abutment pointedupstream, as they also did when the abutment pointed down-stream. These flow features play major roles in scour, andweakening them is one way to minimize scour.75Table 6-3. Local scour depths at wing-wall abutment with geobag.Figure 6-10. Scour development at the unprotectedwing-wall abutment.Test Layout dsA (mm)dsA/dsA45(%)dsB (mm)dsB /dsA(%) W1 no geobag 65 100 0 0 W2 1 geobag row 0 0 93 143 W4 2 geobag rows 0 0 77 119 W5 2 geobag rows plus stone at edge 0 0 63 97
76Figure 6-11. Experiment with a large geobag placedaround the wing-wall abutment.(a) At start of experiment (b) At end of experiment, view from above (c) At end of experiment, view from the sideFigure 6-12. Experiment with two rows of largegeobags placed around the wing-wall abutment. (a) At start of experiment (b) At end of experiment, view from above (c) At end of experiment, view from the side
77Figure 6-13. Influence of wall angle on scour depth at awing-wall abutment.Table 6-4. Influence of wing-wall alignment on scour depth.Figure 6-14. Influence of abutment alignment on scourdepth at a wing-wall abutment.α (degrees) dsA (mm) dsA /dsA0 (%) 90 140 100.00 65 75 53.57 45 65 46.43 30 46 32.86 15 32 22.86
786.8 Summary of Findings fromPreliminary Experiments The results from the preliminary experiments led to thefollowing findings in answer to the questions posed at theoutset of this chapter. The findings are of significance forthe more detailed sets of experiments that were conductedsubsequently for the project:⢠Large concrete blocks placed around an abutment areinsufficiently effective as a scour countermeasure forreducing scour depth at an abutment. The winnowing ofthe bed material from around the blocks enables scour toprogress, though possibly not as deep as may occur if theblocks were not present. Sediment winnowing, edge ero-sion, and local scour around the blocks are processes thatneed to be addressed in order for armoring to function asan effective scour countermeasure.⢠Once a critical block size is attained (with respect toresistance to entrainment by flow), increasing block sizedoes not result in reduced scour depth. Of greater impor-tance than block size is aerial coverage of blocks. Smallerconcrete blocks closely arranged were more effective thanthe larger blocks because they caused less winnowing ofsediment.⢠A large geobag or a continuous mat of relatively smallgeobags holds promise of functioning as an effective scourcountermeasure for wing-wall abutments when the matextends over an area defined approximately as ra/La 1,where ra is radial distance out from the end of the abutmentand La is abutment length. Edge erosion remains a concernbecause the geobag is thick. However, edge erosion likelycan be reduced by use of riprap stone, or smaller geobags,placed around the geobags.⢠The results obtained show that decreasing wall angle (from90 degrees) to flow reduces the scour depth under eitherlive-bed or clear-water scour conditions. Decreasing thewall angle at an abutment was observed to weaken down-flow and the horseshoe vortex. Accordingly, an approach-flow guide wall likely can be effective in reducing scourdepth at a wing-wall abutment.⢠The brief ancillary experiments on scour at various align-ments of abutment show that scour depth is a maximumwhen an abutment is perpendicular to the channel crossed.