Temporal Variation of Local Scour at Bridge Piers with Complex ...

ICSE6 Paris - August 27-31, 2012Giuseppe Oliveto - Local scour at bridge piers with complex geometriesnew methodology to predict local scour depth at a complex pier for varying pile cap elevation. The effects ofthe upstream pile cap extension were also considered, noting such an extension acts to reduce scour byintercepting the pier downflow and protecting the bed from the pier-generated horseshoe vortex. Ataie-Ashtiani et al. (2010) carried out 70 experiments considering a variety of configuration, including differentsizes and shapes of complex piers. They found that the pile cap protects the bed from scour and postpone thebeginning of the scour development when it is located at the bed level. Moreover, in such a condition, ittakes a long time to achieve equilibrium conditions. The outcomes of an experimental study on local scour ata pile cap are provided by Amini et al. (2011). Experiments were conducted under clear-water regime andthe main variables considered were the pile cap dimension and its location relative to the initial streambed.Based on the data collected and historical measurements a new predictive method is then proposed.However, the above studies mainly consider equilibrium scour conditions and focus on the maximum scourdepth or the scour depth at the front of the foundation. Based on a number of recent laboratory experiments,some of which of long duration, this paper aims to provide some insights on the temporal and spatialevolution of the local bed morphology around piers founded on piles when the top elevation of the pile cap isflush with the undisturbed bed level. Differences with local scour at uniform cylindrical piers are alsoexamined.IIEXPERIMENTSExperiments were carried out in a 1 m wide, 1 m deep, and 20 m long rectangular straight channel at theUniversity of Basilicata, Italy. A three-components pier model was tested which included a cylindrical pierstem of diameter D=0.12 m, a square-shaped pile cap of width B=0.24 m and thickness T=0.08 m, and 4cylindrical piles of diameter d=0.04 m, spaced each other by s=0.12 m. The bed material was a nearlyuniform sand with density ρ s =2650 kg/m 3 , median grain size d 50 =1.7 mm, and sediment gradationσ =(d 84 /d 16 ) 1/2 =1.5. The sediment bed was 15 m long and 0.4 m deep. The undisturbed bed surface washorizontal and the pier model always set by flushing the top of the pile cap with the initial bed level.All the experiments were performed under steady flow conditions and clear-water scour regime except forrun CP9 in which live-bed conditions were attained. Some runs were of long duration (up to 13 days) toallow an effective interaction between the flow and the complex pier-foundation and achieve conditions ofquasi-equilibrium. Figure 1 shows the typical local bed morphology observed at the end of run.Figure 1: Local bed morphology at the end of run CP5. Flow was directed from right to left.The water discharge was measured with an orifice plate with accuracy of ±3% placed in the feederpipework. The water surface was measured with a conventional point gauge with accuracy to the nearestmillimetre and the sediment surface with a shoe gauge having a 4 mm by 2 mm wide horizontal plate at itsbase. Comprehensive measurements in space and time of local bed morphology were made. The accuracy ofbed measurements was of the order of the grain size. The approach flow depth was controlled by anadjustable sharp-crested weir located at the channel outlet. Once the bed was accurately levelled, the channelwas slowly filled to avoid any sediment movement by submerging the working section with the weir. The168

ICSE6 Paris - August 27-31, 2012Giuseppe Oliveto - Local scour at bridge piers with complex geometrieswater discharge was then slowly increased to the preselected value. The experiment started when theapproach flow depth was set to the preselected value by lowering the weir. Table 1 provides the mainconditions for each run with Q=discharge, h o =approach flow depth, F d =V/(g′d 50 ) 1/2 densimetric Froudenumber at the approach section with V cross sectional flow velocity, g′=g(ρ s −ρ)/ρ modified gravitationalacceleration, ρ mass density of water, and g gravitationalacceleration, F di =inception densimetric Froudenumber (Hager and Oliveto, 2002), t=time, and z u =scour depth at the pile cap front, z d =scour depth at the rearof the pile cap, l u =(upstream) scour hole length from the pier vertical axis to the upstream edge of scour hole,l d =(downstream) scour hole length from the pier vertical axis to the downstream edge of scour hole, and W=scour hole volume at a given time t. The hole geometry was analyzed at the channel axis, but z u can beconsidered as maximum scour depth around the pier. Moreover, a uniform cylindrical pier of diameterD=0.12 m was tested in run CP0.Test Q [m 3 /s] h o [m] F d [-] F di [-] z u [m] z d [m] l u [m] l d [m] W [dm 3 ] t [h]0.171 0.015 - - - 12CP0 0.048 0.10 2.89 2.830.187 0.017 - - - 240.188 0.017 - - - 320.192 0.017 0.50 1.10 192 76CP1 0.048 0.10 2.89 2.83 0.163 0.064 0.40 1.00 147 19CP2 0.048 0.10 2.89 2.83 0.162 0.084 0.40 1.10 147 760.090 0.034 0.40 0.60 26 3CP3 0.042 0.11 2.30 2.860.147 0.074 0.50 0.85 61 230.172 0.140 0.50 1.10 138 1430.183 0.161 0.50 1.15 163 271CP4 0.070 0.15 2.81 2.98 0.097 0.020 0.30 0.32 22 1CP5 0.070 0.15 2.81 2.98 0.130 0.036 0.55 1.00 60 40.113 0.041 0.45 0.95 55 2CP6 0.060 0.15 2.41 2.980.132 0.067 0.50 1.00 61 50.128 0.069 0.50 1.00 65 100.055 0.000 0.30 0.00 9 2CP7 0.060 0.18 2.01 3.050.071 0.000 0.30 0.00 10 50.080 0.047 0.30 0.20 10 80.126 0.038 0.40 0.65 37 2CP8 0.090 0.21 2.58 3.110.144 0.061 0.40 0.85 51 50.148 0.085 0.45 0.90 53 70.124 0.069 0.40 0.60 75 2CP9 0.100 0.17 3.55 3.030.153 0.057 0.40 0.60 115 50.160 0.058 0.40 0.70 116 7Table 1: Test conditions and main characteristics of the scour hole.Figure 2 gives the axial scour profiles for the run CP8 at t=7 h showing the characteristics of the scour holeconsidered in Table 1.169

ICSE6 Paris - August 27-31, 2012Giuseppe Oliveto - Local scour at bridge piers with complex geometries5l ux [cm]0-150 -100 -50 0 50 100 150 200l dz u-5-10z d-15-20z [cm]Figure 2: Axial scour profile at t=7 h for run CP8 and definition of main characteristics of the scour hole.Figure 3 provides in the left side the contour maps for runs CP0 and CP2 at t=76 h. The scour volume andthe scour hole characteristics along the channel axis for run CP0 were found rather larger as compared tothose for run CP2, due to the shielding effect by the pile cap. In the right side of the figure the contour mapsfor runs CP4 (t=1 h) and CP5 (t=4 h) are given. Scour region elongates downstream of the pile cap as twodistinct and divergent ogival-shaped areas. At the earlier scour stages the maximum scour depth occurred atthe sides of the pile cap, and afterwards it located at the pile cap front. In case of exposure to the flow of thepile group, the upstream piles were exposed in the first place, while the downstream ones came into viewonly after very long time. Overall, relevant differences between scour hole shapes for simple and complexpiers were observed at the rear of the pier. In case of simple pier the deeper scour zones located around thechannel axis while for complex piers they occurred laterally, outside the prolongation of the sides of pile cap.1000100050050001000-1000 -500 0 500 1000 1500 200001000-1000 -500 0 500 1000 1500 20005005000-1000 -500 0 500 1000 1500 20000-1000 -500 0 500 1000 1500 2000Figure 3: (Left side) Contour maps for runs CP0 (top) and CP2 (bottom) at t=76 h. (Right side) Contour mapsfor runs CP4 at t=1h (top) and CP5 at t=4 h (bottom). The coordinates are in mm.IIIANALYSIS OF DATAThe analysis of data was performed considering the HEC-18 approach for the equilibrium scour depth andthe formula suggested by Oliveto and Hager (2002) for temporal progress of scour around uniformcylindrical piers.The HEC-18 approach implies the superposition of three scour components which include the scour depthscaused by the pier stem, the pile cap, and the pile group. In particular, the pier stem scour component, z pier , isgiven byzpierho= Khpier2.0K1K2K3K4Dho0.65Vgho0.43(1)170

ICSE6 Paris - August 27-31, 2012Giuseppe Oliveto - Local scour at bridge piers with complex geometrieswith K hpier coefficient to account for the height of the pier stem above the bed and the shielding effect by thepile cap and K 1 , K 2 , K 3 , and K 4 correction factors for pier nose shape, angle of attack of flow, bed condition,and armouring by bed material size, respectively. Moreover, the pile cap scour depth component, z pile cap ,when the pile cap is located on or below the bed after pier stem scour (as in present study), is given byzpile caphf= Kw2.0K1K2K 3K4Bhfwith h f the distance from the bed (after pier stem scour) to the top of the pile cap, V f the average velocity inthe flow zone below the top of the pile cap, K w the wide pile cap factor, and K 1 , K 2 , K 3 , and K 4 correctionfactors as defined above.With reference to the temporal variation of scour, Oliveto and Hager (2002) found that the maximum scourdepth z(t) depends on three main parameters, namely the reference length L R =D 2/3 h o 1/3 , the approachdensimetric Froude number F d , and the relative time T=[σ 1/3 (g’d 50 ) 1/2 /L R ]t as( T )0.65.5 1.5Z = 0.068σ −0 Fd log(3)where Z=z(t)/L R .Figure 4 shows the ratio Ψ HEC-18 of computed to observed scour depths according to HEC-18 approach.5.04.0ψ HEC-18Vfghf0.43(2)3.02.01.0T0.01.0E+03 1.0E+04 1.0E+05 1.0E+06 1.0E+07Figure 4: Ratio Ψ HEC-18 of computed to observed scour depths according to HEC-18 approach as function ofdimensionless time T. (•) Data for run CP0, (- - - -) line of perfect agreement, and (- . - . -) regression line.It can be observed as: (i) predictions are very accurate for the run CP0 with uniform cylindrical pier. Thisresult is consistent with the fact that the run was of long duration (76 h) and equilibrium conditions werepractically attained; (ii) predictions are in any case conservative for runs with complex pier and satisfactorywhen T>10 5 ; (iii) conversely, significant overestimations arise when T

ICSE6 Paris - August 27-31, 2012Giuseppe Oliveto - Local scour at bridge piers with complex geometries2.0ψ OH1.51.00.5T0.01.0E+03 1.0E+04 1.0E+05 1.0E+06 1.0E+07Figure 5: Ratio Ψ OH of observed to computed scour depths according to Oliveto and Hager (2002) as functionof dimensionless time T. (•) Data for run CP0, (- - - -) line of perfect agreement, and (- . - . -) regression line.Finally, the characteristics of the scour hole were analysed in clear-water conditions assuming forsimplicity L R =D 2/3 h 1/3 o as reference length (because the geometry of the pier model was kept constant in thisstudy) and considering F d and T as governing parameters. Regression analysis reveals that: (i) the upstreamscour length, l u , rapidly attains its equilibrium value and is independent of F d , as can also be seen easily fromTable 1; (ii) the downstream scour length, l d , depends on F d and T as l d /(D 2/3 h 1/3 o )∼F 0.5 d T 0.22 ; (iii) the scourdepth at pile cap front starts at T around 10 2 and grows with time according to the relationshipz u /(D 2/3 h 1/3 o )∼F 0.75 d T 0.18 (coefficient of determination r 2 =0.80), and (iv) scour at the rear of the pile cap startslater, at T around 3⋅10 3 , compared to the pier front, but develops at a faster rate as z d /(D 2/3 h 1/3 o )∼F 0.35 d T 0.33(r 2 =0.87).IVCONCLUSIONSLaboratory experiments on local scour around a pier with a complex geometry were recently carried out atthe Hydraulic Engineering Laboratory, University of Basilicata, Italy. Runs lasted from 0.75 up to 271 hoursto explore the temporal variation of the local bed morphology in comparison with the case of uniformcylindrical piers. The main results can be summarized as follows:(i) Observed data were compared to the predicted values according to HEC-18 approach for complexpier foundations. Results reveal that the HEC-18 approach provides a suitable method for scourprediction at the equilibrium stage. Conversely, it could lead to not negligible overestimations whenmore realistic conditions of unsteady flow are considered;(ii) Observed data were also compared with the predicted values according to the formula by Oliveto andHager (2002) for uniform cylindrical piers. Results reveal that the shielding effect by the pile capoccur for T around 10 5 ; and(iii)The axial upstream scour length rapidly attains its equilibrium value while the axial downstream scourlength depends on F d and T. Moreover, scour at the rear of the pile cap starts later compared to the pierfront, but develops at a faster rate.This paper aims to provide only a contribute in a problem with a great number of variables. Thus, moreresearches would be needed mainly considering different sediment beds, pier-foundation geometries, and thecase of foundation exposed to the flow owing to general scour.VACKNOWLEDGMENTSThe author thanks Maria Cristina Marino for her help in data collection and processing.VIREFERENCESAmini A.S., Mohammad T.A., Aziz A.A., Ghazali A.H. & Huat B.B.K. (2011). – A local scour predictionmethod for pile caps in complex piers. Proceedings of the ICE - Water Management, 164(2):73-80.172

ICSE6 Paris - August 27-31, 2012Giuseppe Oliveto - Local scour at bridge piers with complex geometriesAtaie-Ashtiani B., Baratian-Ghorghi Z., & Beheshti A.A. (2010). – Experimental investigation of clearwaterlocal scour of compound piers. Journal of Hydraulic Engineering, 136(6):343-351.Chabert J., Engeldinger, P. (1956). – Étude des affouillements autour des piles de ponts. LaboratoireNational d’Hydraulique. Serie A., Chatou, France.Coleman S.E. (2005). – Clearwater local scour at complex piers. Journal of Hydraulic Engineering,131(4):330-334.Hager, W.H., Oliveto G. (2002). – Shields’ entrainment criterion in bridge hydraulics. Journal ofHydraulic Engineering, 128(5):538-542.Jones S.J., Kilgore R.T., & Mistichelli M.P. (1992). – Effects of footing location on bridge pier scour.Journal of Hydraulic Engineering, 118(2):280-290.Martín-Vide J.P., Hidalgo C., & Bateman A. (1998). – Local scour at piled bridge foundation. Journal ofHydraulic Engineering, 124(4):439-444.Melville B.W., Coleman S.E. (2000). – Bridge Scour. Water Resources Publications, Highlands Ranch,Colorado.Melville B.W., Raudkivi A.J. (1996). – Effects of foundation geometry on bridge pier scour. Journal ofHydraulic Engineering, 122(4):203-209.Oliveto G., Hager W.H. (2002). – Temporal evolution of clear-water pier and abutment scour. Journal ofHydraulic Engineering, 128(9):811-820.Oliveto G., Onorati B., & Comuniello V. (2006). – Effects of pile caps on local scour at bridge piers.Proceedings of 3rd International Conference on Scour and Erosion, CURNET, Gouda, The Netherlands.Parola A.C., Mahavadi S.K., Brown B.M., & El Khoury A. (1996). – Effects of rectangular foundationgeometry on local pier scour. Journal of Hydraulic Engineering, 122(1):35-40.Richardson E.V., Davis S.R. (2001). – Evaluating scour at bridges. Report No. FHWA-IP-01-00, HydraulicEngineering Circular (HEC-18). Federal Highway Administration, Washington, D.C., USA.Sheppard D.M., Glasser T. (2004). – Sediment scour at piers with complex geometries. Proceedings of 2ndInternational Conference on Scour and Erosion, World Scientific, Singapore.Sheppard D.M., Glasser T. (2009). – Local scour at bridge piers with complex geometries. GeotechnicalSpecial Publication, 186:506-513.173