Tailor Made Concrete Structures – Walraven & Stoelhorst (eds)© 2008 Taylor & Francis Group, London, ISBN 978-0-415-47535-8

Experimental tests on repaired and retrofitted bridge piers

T. Albanesi, D. Lavorato, C. Nuti & S. SantiniDepartment of Structures, University of Roma Tre, Rome, Italy

ABSTRACT: This research aims to study the seismic performance of existing r.c. bridge piers specimensheavily damaged after previous pseudodynamic tests and actually repaired and upgraded by using self compactingconcrete, stainless steel rebars and CFRP wrapped strips. Pier specimens are representative of tall and squatcircular r.c. piers designed according to Eurocode 8 and Italian Code before 1986: EC8 columns are repairedand Italian ones retrofitted to enhance ductility and shear capacity. Innovative materials have been used and suchinterventions are described in details with particular attention to practical problems occurred. Experimental testresults carried out at the Laboratory of experiments on materials and structures of the University of Roma Tre onself compacting concrete and stainless steel bars are shown and discussed. Retrofitting design and equipment forpseudodynamic tests are also presented. In coming tests aim to evaluate the effectiveness of adopted repairingand upgrading techniques not only to increase ductility but shear strength too.

1 INTRODUCTION

This paper focuses on current experimental tests at theLaboratory of experiments on materials and structuresof the University of Roma Tre on column speci-mens representative of tall and squat circular r.c. piersdesigned according to Eurocode8 (1998-2) and ItalianCode (D.M. LL.PP. 24.01.86). In previous research (DeSortis & Nuti 1996) some columns were tested untilcollapse by pseudodynamic tests but others are stillentire. An accurate study to detect the level of degra-dation in materials, as the case of real structures afteran earthquake, is now performed. Then, based on theevaluation done, EC8 columns are repaired and theItalian ones retrofitted by mean of FRP jacket, longi-tudinal stainless steel and self compacting concrete.Actual tests aim to evaluate the effectiveness of thistechnique. Repairing and retrofitting operations, testequipment and planned tests are presented.

2 PIER SPECIMENS

Eight 1:6 scaled column specimens representative oftall and squat 2.500 m diameter circular r.c. piersof regular and irregular bridges (Figure 1) designedaccording to Eurocode8 Part 2 (1998-2) and ItalianSeismic Code (D.M. LL.PP. 24.01.86) are considered.

Reinforcement scaling criteria, based on similitudefulfillment of global quantities (flexural and shearstrength, confinement effect, post-elastic buckling and

pullout of rebars) were applied (De Sortis et al. 1994)allowing the use of ordinary concrete mixing andcommercial reinforcing bars. An improved anchoragedetail of the longitudinal bars to improve bond-slipscaling accuracy was also investigated. Geometryand reinforcement configurations of pier specimenstogether with relative design criteria are summarizedin Table 1.

Longitudinal and transverse reinforcement config-urations and concrete cover thickness have some dif-ferences with respect to design drawings as sometimes

Figure 1. Layout of designed bridges.

673

Table 1. Pier specimens characteristics (D = diameter[mm], H = height [mm], C = cover [mm]) and design criteria(*hoop spacing lager than EC8 prescriptions).

Pier # design D H C bars spiral

1 DM 420 2340 30 24 ∅ 10 ∅5/80Tall 2 DM 420 2340 30 24 ∅ 10 ∅5/100

3 DM 420 2340 30 24 ∅ 10 ∅5/1004 DM 420 2340 30 24 ∅ 10 ∅5/805 EC8 420 2340 30 24 ∅ 10 ∅6/40

Squat 6 DM 420 1170 30 12 ∅ 12 ∅6/1207 EC8* 420 1170 30 42 ∅ 12 ∅6/1208 DM 420 1170 30 12 ∅ 12 ∅6/120

Figure 2. Current specimens state: tall (1, 2, 3, 4, 5) andsquat (6, 7, 8) specimens.

it happens in practice. Concrete with Rck = 25 MPaand FeB44k steel were fixed in design. Tested con-crete strength was about the same of the design onewhilst steel yielding varied from 550 MPa to 600 MPa,therefore exceeding the design one.

3 CURRENT SPECIMENS STATE

Piers 2 (tall DM), 4 (tall DM), 5 (tall EC8), 6 (squatDM) and 8 (squat DM) were tested until collapsein previous pseudodynamic tests (De Sortis & Nuti,1996) thus they are damaged while piers 1 (tall DM),3 (tall DM) and 7 (squat EC8) have not been sub-jected to tests and are intact (Figure 2). Detecteddamages include cover spalling, longitudinal bar buck-ling and rupture, yielding in transverse reinforcementand concrete core crushing.

4 REPAIR AND RETROFITTING

Foreseen operations on seriously damaged piers 2, 4, 5,6, 8 are (for example piers 6 and 8 in Table 2): mechan-ical removal of damaged concrete cover and cleaning

Table 2. Repair and retrofitting of pier.

Phase 1 Phase 2 Phase 3 Phase 4 Phase 5

of substrate from residue particles (phase 1), concretecore repair with resin injections (phase 2), substitutionof damaged (yielded, buckled or broken) bars usingstainless steel ones (phase 3), restoration of damagedconcrete cover with self compacting concrete (SCC)(phase 4) and C-FRP strips application (phase 5).

4.1 Damaged concrete removal (phase 1)

Cracked and nearly detached part of concrete coverand core have been completely removed over an heightof 550 mm (plastic hinge length plus overlap splice ofrestored longitudinal bar) from the pier base. About20 mm of concrete around reinforcing bars have beenremoved and the latter accurately cleaned in order toguarantee optimal bond with the repairing material.

The quality of the interface between the existingconcrete and the repairing material is essential fordurability and effectiveness of restoration. In partic-ular, lack of surface roughness makes the interface apreferential plane for rupture: mechanical removal fol-lowed by cleaning of substrate from residual particleswas used to provide good bond (Table 2: phase 1).

4.2 Concrete core consolidation (phase 2)

Injection of concrete core with bicomponent expoxyresin (EPOJET LV with 70 MPa and 20 MPa com-pressive and flexural strength respectively at 7 daysand adhesion to concrete higher than 3.5 MPa) withvery low viscosity (140 mPa.s Brookfield viscositywith 1 hour workability time at 20˚ C) have been used.About 20–40 mm deep perforations have been per-formed to place small plastic tubes. Cracks were closedby spreading with thixotropic bicomponent expoxyplaster (MAPEWRAP12 with 30 MPa tensile strengthafter 7 days -ASTM D 638, adhesion to concrete higherthan 3 MPa and setting time at 23◦C of about 5 hours)in order to obtain a sort of pipe system inside the coreending with the tubes. ∅2–5˚ mm sand obtained bycrushing was fixed on the fresh plaster to make its sur-face rough. Compressed air was insufflated in orderto remove powder and to check the communicabilitybetween inner pipes. The injections were preformedfrom the bottom towards the top until inner crackscomplete filling (Table 2: phase 2).

674

Figure 3. Welded joints (a) top (b) bottom (c) detail.

4.3 Reinforcing bar restoration (phase 3)

Austenitic stainless steels combine very good corro-sion behavior with excellent mechanical properties(strength and toughness). Although initially moreexpensive, it offers cost savings in the long termbecause of reduced maintenance and protection oper-ations and increasing life of the structure. These char-acteristics make it suitable for restoration of existingdamaged reinforcing bars and for seismic retrofitting.Stainless steel (AISI304) ∅12 bars were used fordamaged reinforcing bar substitution. First, the spi-ral reinforcement was cut and removed; the ends ofembedded hoop were welded to remaining longitudi-nal bars.Yielded and buckled longitudinal rebars werecut and removed too. Remaining rebars were cleaned,by mean of a metallic brush, to eliminate rust layersand residual concrete, plaster and resin. Single newbars were placed aside the existing ones trying to keepinner lever arm and cover thickness equal to the orig-inal ones: it was impossible to apply them in pairs toavoid asymmetries because of the congested reinforce-ment of the foundation. Welding was performed awayfrom the potential plastic hinge (Figure 3): since theportion of damaged bar often extends into the foun-dation, it was necessary to dig it (at least 70–80 mmdeep) around the bar before welding.

Welding was the only practicable solution sincethe congested reinforcement of the foundation didnot allow to perform holes deep enough and in theright position to guarantee effective bond even usingresin. Practical difficulties in welding due to reducedspace around the bars made it necessary to realize two50 mm weld puddles (the design length of a singlepuddle was 35 mm only). Removed spiral reinforce-ment was replaced with circular stainless steel ∅5stirrups; spacing was calculated to assure the sameoriginal design ratio ftAs/s (As = hoop cross section,s = spacing, ft = peak strength). Stirrups were fixedwith intercrossed metallic fastenings to avoid weld-ing that could produce dangerous local deformation inbars with such a small diameter.Anyway, 20 mm lengthwelding was necessary to close stirrups (Table 2:phase 3).

Stainless steel reinforcing ribbed bars have beentested and their mechanical properties determined bothfor monotonic and cyclic behavior including post-elastic buckling (Albanesi et al. 2006). Monotonic

Figure 4. Monotonic tests in tension and compression(λ = L/∅ = 5, 11) for ∅12 stainless steel bars.

Table 3. Mechanical properties of stainless steel bars (meanvalues of strength and elongation).

∅ fsy fsm fsm/fsy Agt Asumm MPa MPa – % %

5 829 927 1.19 5 22.612 790 941 1.19 5 22.0

tests in tension and compression on ∅12 bars (Figure4) used in restoration have been performed accordingto the European Code UNI EN 10002-1 (2004).

Yielding strength fsy (stress corresponding to 0.2%residual strain), peak strength fsm, fsm/fsm ratio, uni-form elongation at maximum stress Agt and at ruptureAsu are summarized in Table 3. Notice that mechan-ical characteristics for Italian stainless steel will bechanged.

4.4 Concrete restoration (phase 4)

Good workability and resistance to segregation andremarkable filling and passing ability make Self Com-pacting Concrete (SCC) an optimal material to restoredamaged concrete parts thus restoring element con-tinuity and homogeneity (crossing thick new andexisting reinforcing bars without causing vacuums intothe element and discontinuity at contact surface). Arestoring material should also have low shrinkage, hightensile strength and similar elastic modulus to that ofthe substrate in order to reduce tensile stresses due toshrinkage causing disjunction at interface and to guar-antee continuity of the structural element after restora-tion. SCC with addition of an expansion anti-shrinkageagent is suitable to this purpose.

Cast-in-place SCC seemed the most adequate solu-tion for the piers under consideration; it includes, inaddition to substrate preparation, formwork construc-tion, concrete production (Figure 5a), its application

675

Figure 5. (a) SCC production (b) formwork (c) SCC jetprocess.

and ageing. Formwork originally used to build thespecimens was opportunely modified to include ametallic groove (Figure 5b) allowing SCC pouringwithout interruptions from a single hole in the form-work (Figure 5c). The final result was generally goodalthough superficial imperfections were present afterformwork stripping (Table 2: phase 4). SCC filled upall available spaces even the upper part, opposite topouring position. Concrete restoration results in a SCCjacket 50–90 mm thick and 550 mm high.

A 35 MPa mean cubic strength is required as themost similar value to that of the original piers achiev-able with SCC. The main aims of the SCC mixdesign were: high fluidity, small diameter (<12 mm)aggregate (to easily fill up all the spaces between rein-forcement and concrete core), shrinkage offset addingconveniently measured out expansive agent. The latterwas essential to guarantee complete filling and thuscontinuity at the interface between existing and newmaterial. Resulting SCC mix was made of grit, sand,calcareous filler, cement expansive agent, superplas-ticizing and water with water-cement content equal to0.48.

In case of pier 6, SCC was formed with the same mixdesign but required more water and superplasticizingto get acceptable slumps due to possible filler humiditynow under investigation. SCC compressive cubic meanstrength at failure, measured over six standard cubes,resulted in 48.19 MPa (1.60 MPa standard deviation).Tests were carried out on fresh material to verify CodeUNI 11040 (2003) prescriptions. Shrinkage behaviorwith time (Figure 6) was measured at the Laboratoryof the Research Center BUZZI UNICEM in Guidonia(Rome).

4.5 Retrofitting (phase 5)

Retrofitting of structural vertical elements likecolumns or bridge piers by FRP jacket usually aimsto increase load carrying and/or ductility capacity byimproving confinement action in plastic hinge zones.Confinement increases concrete compressive strengthand ultimate strain and provide longitudinal bars con-straint against buckling preventing cover spalling. Byanalyzing pier characteristics (Table 1), piers 2 and

Figure 6. Hydraulic shrinkage of SCC.

6 seem to be the most representative for retrofittingdesign of slender and squat piers respectively. Detecteddamage state clearly shows that tested squat andtall piers suffered shear and flexural failure respec-tively.These failure mechanisms were also analyticallyproved by comparing shear strength with shear atresisting bending moment. Retrofitting designs forpier 2 and 6 aim to increase ductility capacity and shearstrength respectively in order to match EC8 designprescriptions. For the ductility upgrade, the thicknessof the FRP jacket to provide the target displacementductility (EC8) was determined according to Montiet al. approach (2001). Ductility upgrade is defined interms of a section upgrading index Isez , (the ratio ofthe target sectional ductility and the initially availableone). Confining pressure due to FRP jacket neededto guarantee ductility upgrade Isez is evaluated, alsoincluding existing steel stirrup confining effects onconcrete strength and ultimate strain and confiningpressure reduction due to discontinuous wrapping.The thickness corresponding to FRP contribution toupgrade shear strength of existing piers to the valuerequired in EC8 design was assumed as the maximumvalue determined according to the most commonlyused approaches: Eurocode 8 (1998-3), OPCM 3274(2003), CNR (CNR-DT 200 2004), FIB (Fib Bullettin2001). This gap is due to the different contribution oftransversal reinforcement (Albanesi et al. 2006, 2007).

Commercial 0.167 mm thick C-FRP (MapewrapC UNI-AX 300/10 with ultimate strain εfu = 2.0%and elastic modulus Ef = 230 GPa) was wrapped in100 mm wide unidirectional fiber strips with inter-val of 60 mm with the exception of the first onewhich is only 30 mm (Table 2-last column). FRP stripconfigurations for tall and squat piers are shown in Fig-ure 7. The area of FRP strips application was brushedwith a metallic brush to remove the very superficialconcrete substrate. Then the primer (MAPEWRAPPRIMER 1) was spread with a brush to make thesurface uniformly wet and, over the fresh primer, a1–2 mm thick thixotropic bicomponent expoxy plas-ter (MAPEWRAP 12) to fill up and level present

676

2215

100

3060

6060

3060

6060

6060

100

1040

Figure 7. FRP strips configuration designed for (a) tall piers2, 4, 5 and (b) squat piers 6, 8.

vacuums. The primer enhanced gripping betweenplaster and concrete.The reinforcement was fabricatedwith a wet layup process: first, resin (MAPEWRAP31) was applied to the fresh platter, then layers of fabric(1.700 m strips to guarantee 300–350 mm lap splice)were hand-placed with fibers direction perpendicularto pier axis and impregnated in place using ribbed rollsand squeegees to avoid entrapment of air. In case ofdouble layers, two distinct overlapped FRP strips wereapplied.

5 TEST EQUIPMENT

The test equipment, schematically shown in Figure 8and Figure 9, is composed by:

#1: system for the application of vertical loads atthe top of the specimen: the specimen is placed withina testing frame realized using a transverse spreaderbeam which is linked to the ground by means oftwo ∅47 mm Dywidag rods, one on each side of thespecimen, with hinged connections. The axial load isapplied by prestressing the two rods by mean of two1000 kN Hollow Plunger Cylinders, which are insertedin a cage bordered by four ∅26.5 mm Dywidag rodsand two hollows steel plates, the upper reaction oneand the plate connected to the hinge.

#2: system for the application of horizontal loadsand displacements: a 250 KN MTS hydraulic actuator,connected on one side to the top of the pier and on otherside to a reaction wall, is used to impose displacementsor loads to pier;

#3: ground anchorage system: the specimen footingis restrained to the laboratory strong floor using twotransverse beams fixed to ground with steel tendons inorder to avoid any base horizontal displacements androtations;

#4: control system and data acquisition: horizon-tal displacements (or loads) are applied by usingthe built-in MTS control system composed by anLVDT with stroke ±125 mm (or a 250 kN load cell),the digital controller Testar II and the acquisitionsoftware TestWare. The same system is used to acquire

160 220 1904 1122

6020

041

022

1028

020

040

024

0

200

160

Ø47

Ø26.5

Ø30

Ø60

300

670

2327

200

330

21575

5050

024

050

1200 500

200 200

#2

#3

#1

Figure 8. Test equipment: #1 vertical load applicationsystem, #2 MTS hydraulic actuator, #3 ground anchoragesystem.

10(12)

2 (4)

1

19

16 (17)

18

7 (9)

11(13)

6 (8)

3 (5)

22

01

30

20

05

70

12

0

23

0 45

0

10

0

front(back)

11

70

/ 2

34

0(s

qu

at /

tal

l pie

r)

#2

#3

#1

#4

Figure 9. Test equipment: #4 displacement acquisitionsystem.

horizontal displacements and loads.A parallel acquisi-tion is performed by using an external data acquisitionsystem NATIONAL IN-STRUMENTS NI PCI 6281,High-Accuracy M series Multifunction DAQ 18-Bit,up to 625 kS/s, 3 modules NI SCXI 1520 Univer-sal Strain Gauge Input Module with 8 Channels, and2 modules NI SCXI 1540, 8 Channels LVDT InputModule, with software LabVIEW Full System. In par-ticular, the horizontal displacements of three point ofthe pier, placed respectively at the height 230 mm,450 mm and 1170 mm or 2340 mm for squat andtall piers respectively from the pier base section, arerecorded using 3 linear potentiometers with stroke±50 mm (channels 19, 18, 1), whereas other 12 linearpotentiometers are used for the acquisition of verti-cal displacements of other three different pier sections(channels from 2 to 13) in order to estimate theircurvature along the height. Two linear potentiome-ters (channels 16, 17) are placed on pier foundation inorder to check incidental rigid rotations. Two hollowload cells, placed between the axial load jacks and the

677

upper steel plate of the cage, measure the axial load intendons.

6 TEST PROGRAM

Elastic and pseudodynamic tests are foreseen in thenext future: the former aims to characterize the piers,the latter to evaluate their non linear behavior untilcollapse and to assess the efficiency of adopted repair-ing and retrofitting techniques. Comparisons with theresults of previous pseudodynamic tests on the originalpiers will be carried out.

7 CONCLUSIONS

This paper focuses on current experimental tests at theLaboratory of experiments on materials and structuresof the University of Roma Tre on column speci-mens representative of tall and squat circular r.c.bridge piers designed according to EC8 and ItalianCode before 1986. Damages in pier specimens pre-viously tested until collapse by pseudodynamic testsare accurately detected. Damaged EC8 columns arejust repaired whilst the Italian ones are first repairedusing stainless steel bars and self compacting con-crete and then retrofitted by mean of C-FRP layersdesigned to upgrade ductility and shear strength tomatch EC8 design requirements. Pseudodynamic testsare foreseen in the next future.

REFERENCES

Albanesi,T., Lavorato, D. & Nuti, C. 2006. Prove sperimentalimonotone e cicliche su barre di acciaio inox. Sperimen-tazione su materiali e strutture. Convegno nazionale,Venezia: 357–366.

Albanesi, T., Nuti, C., Paolacci, F. & Santini, S. 2006. Cyclictests of existing r.c. columns repaired or retrofitted bymean of jacketing technique, fib Congress, Naples, paper0521.

Albanesi, T., Lavorato, D., Nuti, C. & Santini, S. 2007. Com-portamento ciclico di colonne in c.a. riparate ed adeguatecon fasciature in frp. ANIDIS: XII Convegno L’IngegneriaSismica in Italia, Pisa, 10–14 giugno.

Albanesi, T., Nuti, C. & Vanzi, I. 2007. Closed form con-stitutive relationship for concrete filled FRP tubes under

compression. Construction and Building Materials, 21,409–427.

CNR-DT 200. 2004. Istruzioni per la Progettazione,l’Esecuzione ed il Controllo di Interventi di Consolida-mento Statico mediante l’utilizzo di Compositi Fibrorin-forzati.

Coppola, L. 2000. Concrete durability and repair technology.Proceedings of 5th CANMET/ACI International Confer-ence on “Durability of Concrete”, Barcellona, Spain:1209–1220.

De Sortis,A. & Nuti, C. 1996. Seismic response by pseudody-namic tests of RC bridges designed to EC8. Proceedingsof 11th World Conference on Earthquake Engineering,Acapulco, Mexico.

De Sortis, A., Monti, G. & Nuti, C. 1994. Problemi di scalanella sperimentazione pseudodinamica di pile da ponte inc.a.. Atti del Worksop su Danneggiamento, Prove Ciclichee Pseudodinamica, Dip. Analisi e Prog. Str., Università“Federico II”, Napoli, Italia.

D.M. LL.PP. 24.01.86. Norme tecniche per le costruzioni inzone sismiche.

Eurocode 8. 1998. Design of structures for earthquakeresistance, Part 2: Bridges. (Draft March 2005).

Eurocode 8. 1998. Design of structures of earthquake resis-tance, Part 3: Assessment and retrofitting of buildings.(Draft November 2004).

Fib Bullettin. 2001. Design and use of externally bondedFRP reinforcement (FRP EBR) for reinforced concretestructures. Bulletin no.14, prepared by sub-group EBR(Externally Bonded Reinforcement) of fib Task Group 9.3‘FRP Reinforcement for Concrete Structures’.

O.P.C.M. 20.03.2003 n. 3274. Primi elementi in materia dicriteri generali per la clas-sificazione sismica del territorionazionale e di normative tecniche per le costruzioni inzona sismica, G.U. 08.05.2003 n. 105.

Moléz, L. 2003. Comportement des reparations structuralesen béton: couplage des effets hy-driques et mécaniques.PhD thesis, École Normale Supérieure de Cachan, France.

Monti, G., Nisticò, N. & Santini, S. 2001. Design of FRPjackets for upgrade of circular bridge piers. Journal ofComposite for Construction, ASCE, 5(2): 94–101.

Talbot, C., Pigeon, M., Beaupré, D. & Morgan, D.R. 1994.Influence of surface preparation on long-term bonding ofshotcrete. ACI Material J. 91(6): 560–566.

UNI EN 10002-1. 2004. Materiali metallici – Prova ditrazione – Parte 1: Metodo di prova a temperaturaambiente.

UNI 11040. 2003. Calcestruzzo autocompattante – Speci-fiche, caratteristiche e controlli.

678