A Novel Technique of Seismic Strengthening of Nonductile RC

A Novel Technique of SeismicStrengthening of Nonductile RC Frameusing Steel Caging and AluminumShear Yielding Damper

Dipti R. Sahooa) and Durgesh C. Rai,b)M.EERI

A novel strengthening scheme for seismically-weak RC frames is proposedwhich utilizes external steel caging to improve flexural/shear strength ofcolumns and aluminum shear-yielding damper (Al-SYD) to further enhancelateral strength, stiffness and overall energy dissipation capacity of the frame.This paper describes the effectiveness of this scheme as evidenced in anexperimental study on a reduced scale (1:2.5) single-story, single-bay, gravity-only designed reinforced concrete (RC) frame. The strengthened frame wassimultaneously subjected to gravity loads and reversed cyclic lateraldisplacements as per ACI-374 loading protocol. An innovative connectionscheme was designed to transfer a portion of frame lateral load to the energydissipation device (Al-SYD). Besides the significant increase in lateralstrength and stiffness of the strengthened frame, RC frame members did notsuffer any major damage during the entire test protocol. This indicatessignificant reduction in force demand on existing RC members because ofenhanced energy dissipation through hysteretic shear yielding of aluminumpanels. Moreover, the simple connection scheme proposed in this study provedvery efficient in transferring the frame lateral load to strengtheningelements. �DOI: 10.1193/1.3111173�

INTRODUCTION

A reinforced concrete (RC) building must have adequate strength, stiffness, redun-dancy, and ductility to perform satisfactorily in a major earthquake. The behavior of con-ventional RC framed buildings subjected to seismic loading primarily depends on gooddetailing of its components in regions of expected plastic deformation. The nonductileframed buildings, generally constructed before the development of rigorous seismic de-sign and detailing provisions, are often severely damaged or collapsed during strongearthquakes. As a consequence of general socio-economic development, there has beena growing interest in strengthening of such deficient buildings in recent times.

Seismic strengthening usually involves an increase in strength, stiffness, ductility, orcombination of these parameters of frame members. Strengthening of deficient columnsis one of the widely accepted techniques used to enhance the seismic performance ofmoment resisting RC frames (Endo et al. 1984). Several conventional techniques of col-

a) Doctoral Scholar, Dept. of Civil Engineering, Indian Institute of Technology Kanpur, UP-208016, India.b)
Associate Professor, Dept. of Civil Engineering, Indian Institute of Technology Kanpur, UP-208016, India.
415Earthquake Spectra, Volume 25, No. 2, pages 415–437, May 2009; © 2009, Earthquake Engineering Research Institute

416 D. SAHOO AND D. C. RAI

umn strengthening are available in the literature, such as concrete jacketing (Bett et al.1988, Rodriguez and Park 1994, Abdullah and Takiguchi 2003), steel jacketing (Sakinoand Ishibashi 1985, Chai et al. 1990, Aboutaha et al. 1999), and composite jacketing(Xiao and Ma 1997, Tastani and Pantazopoulu 2004). Each strengthening technique hasits own advantages and limitations depending on practicability, cost-effectiveness, mini-mal reduction in useable floor area, and effectiveness in enhancing the desired structuralproperties.

The increase in lateral strength and/or stiffness of existing frames beyond a certainlimit may not be always feasible by column strengthening using conventional techniques.Passive energy dissipation devices can be successfully employed to enhance overall seis-mic performance of frames by minimizing the seismic demand imposed on deficientframe members. Different types of devices are available now-a-days using energy dissi-pation and damping characteristics due to metal hysteresis, friction, viscous fluid, andvisco-elastic materials. These devices can be successfully integrated with the steelframed structures for enhanced seismic performance (Zhang and Soong 1992, Martinez-Rueda and Elnashai 1995, Soong and Dargush 1997, Rai and Wallace 1998). However,the most challenging part is to integrate such devices with the existing RC frames foreffective utilization of their energy dissipation and damping potential. Hence, there is aneed to develop an effective, fast, and practical upgrading scheme for deficient RCframes using energy dissipation devices.

In the present study, a novel seismic strengthening scheme as shown in Figure 1 isdeveloped for the existing nonductile RC frames to reduce their primary deficiencies inlateral strength, stiffness, and energy dissipation potential. The deficient RC columnswere strengthened by external steel cage to enhance their lateral strength and plastic ro-tational capacity, whereas Aluminum Shear-Yielding Damper (Al-SYD) was used as thesupplemental energy dissipation device to increase the lateral stiffness of the system andto dissipate the seismic energy through hysteretic shear yielding of the Al-SYD. An in-

Figure 1. Schematic diagram of strengthened RC frame with test set-up (All dimensions are inmm).

A NOVEL TECHNIQUE OF SEISMIC STRENGTHENING OF NONDUCTILE RC FRAME 417

novative connection scheme was also designed to integrate strengthening elements withthe existing RC frame and to transfer a portion of frame lateral load to the energy dis-sipation system. This paper presents the details of strengthening scheme, the design ofconnections of strengthening elements with the existing RC members, and the perfor-mance of strengthened frame under combined action of gravity loads and reversed cycliclateral displacements. The main objectives of this study are: (a) to evaluate the overallperformance of strengthened frame, (b) to investigate the energy dissipating behavior ofAl-SYD, and (c) to assess the effectiveness of the proposed connection scheme in suc-cessful integration of strengthening elements with the existing RC frame.

BACKGROUND

Two primary components of the proposed strengthening technique are the steel cageand the Al-SYD and their performance under cyclic lateral loads has been previouslystudied experimentally and analytically by various researchers. A brief description of thepast research on these components is provided in the following sections.

STEEL CAGING OF RC COLUMNS

Steel caging technique, a variety of steel jacketing, is commonly used to strengthen/retrofit of square/rectangular RC columns in various countries now-a-days (Li et al.2005). Steel cage consists of four longitudinal steel angle sections at corners of columns,which are interconnected by steel battens at few places along their length. The interme-diate space between steel cage and RC column may be filled with cement or epoxy mor-tar and a cover of concrete or shotcrete reinforced with light welded fabric may be ap-plied over the steel cage for protection against corrosion or fire. This technique isgenerally regarded as practical, fast, and cost-effective, which improves the seismic be-havior of the whole structure by increasing the lateral strength and ductility of columns(Dritsos and Pilakoutas 1992).

Nagaprasad (2005) investigated the performance of steel cage strengthened rectan-gular RC columns under constant axial compressive load and gradually increasing re-versed cyclic lateral load. The external steel cage consisted of four Indian Standardrolled equal-angle ISA 35�35�[email protected] N/m sections (BIS 1964) and mild steel bat-ten plates of 6 mm thickness. Steel cage was sufficiently tightened to the RC column sothat steel angles at corners were in close contact with concrete surface. However, theintermediate gap between steel cage and column surface was not filled with any kind ofbinder material. Steel cage was connected to the RC footing by means of a moment con-nection using base plates, stiffeners, and anchor bolts as shown in Figure 2a.

Full and stable hysteretic loops with strain hardening behavior as shown in Figure 2bwere observed for each strengthened column. The significant improvement in lateralstrength, stiffness, displacement ductility, plastic rotational capacity, and energy dissipa-tion potential of the strengthened column was primarily due to the confinement of RCcolumn section by external steel cage. Further, the premature buckling of angle sectionswas delayed by using wider battens in the expected plastic hinge region of steel cage.Based on experimental results, a theoretical confinement model was developed to deter-


mine the confined compressive strength of column concrete using values of effective bi-axial lateral stresses in the model proposed by Mander et al. (1988) for the confinementof concrete due to rectangular hoops.

ALUMINUM SHEAR-YIELDING DAMPER

Aluminum panels of soft alloys designed to yield in shear can be effectively used assupplemental energy dissipation devices, which can minimize the seismic demand onprimary structural members (Rai and Wallace 1998, 2000). These I-shaped devices areformed by welding aluminum plates for flanges, web, and transverse stiffeners as shownin Figure 3a and annealed at a high temperature to relieve the stress induced duringwelding process prior to being used for energy dissipation. The device is so placed in astructure that under the action of lateral loads, its web plate undergoes large shear de-

Figure 2. (a) State of strengthened RC column before and after slow-cyclic test, and (b) hys-teretic response of strengthened column with wider end battens (Nagaprasad 2005).

Figure 3. Behavior of aluminum shear-yielding damper under cyclic shear load: (a) state ofdamper before and after cyclic shear test and (b) typical hysteretic response of damper (Jain
et al. 2008).


formation. Due to low-yield strength and excellent post-yield strain-hardening behaviorof aluminum panels, significant energy is dissipated through full and stable hystereticloops as shown in Figure 3b.

Jain et al. (2008) conducted a series of tests on aluminum shear panels with varyingweb-to-thickness ratio, aspect ratio, slenderness ratio, alloy type, and number of panelssubjected to cyclic shear load. Stable and full hysteretic loops without pinching was ob-served even up to shear strains of 20% indicating the excellent energy dissipation po-tential of these panels. It was concluded that the problem of inelastic shear buckling atoperating shear strains could be delayed by increasing the number of transverse stiffen-ers along the length of shear panels. A relationship between buckling stress and slender-ness ratio of aluminum shear panel was also developed to limit the inelastic web buck-ling at design shear strains.

EXPERIMENTAL INVESTIGATION

PROTOTYPE STRUCTURE AND TEST SPECIMEN

A four-bay, five-story RC moment resisting frame with open-ground story andmasonry-infilled upper stories was selected as a prototype structure. Story height at theground floor and upper floors was 3.6 m, and 3.0 m, respectively. The width of each baywas 6.0 m at each floor level, and the RC frames were placed at center-to-center spacingof 6.0 m in transverse direction of the building. The service live loads on floors and roofwere 3.0 kPa and 1.5 kPa, respectively, whereas the lateral wind pressure was consid-ered as 1.5 kPa as per Indian Standard IS: 875 (BIS 1989) provisions. Various framemembers were designed in accordance with Indian Standard IS: 456 (BIS 2000) provi-sions for gravity loads without considering the detailing requirements for seismic load-ing conditions.

Nonductile RC specimen was chosen to be 1:2.5-scaled single-story and single-baymodel representing an interior bay at bottom story of the prototype frame (Figure 1).Table 1 summarizes the geometric properties and reinforcement detailing of the testspecimen. RC columns of the test specimen were strengthened using steel cage consist-ing of four Indian Standard rolled equal-angle ISA 35�35�4 [email protected] N/m sec-tions (BIS 1964) at corners. These angles were held together by mild steel batten platesat few places along the height of steel cage. The size of intermediate and end battens

Table 1. Dimensions and reinforcement detailing of frame members of specimen

MemberOverall size�L�B�H�

Clear cover(mm)

Longitudinalreinforcement

Transversereinforcement

Footing 3200�250�300 24 12� (6 nos.) 8�@300 mm c/cColumn 1350�160�160 16 16� (4 nos.) 4�@100 mm c/cBeam 3200�160�180 10 10� (8 nos.) 5�@300 mm c/cSlab 3200�800�50 8 10� (8 nos.) 8�@300 mm c/c

�=Diameter of rebar in mm; L=Length in mm; B=Breadth in mm; H=Depth/Height in mm


were chosen as 150 mm�50 mm�5 mm, and 150 mm�150 mm�12 mm, respec-tively. The larger plates used as end battens serve two functions, namely, (a) to reducethe possibility of premature local buckling of angle sections in the potential plastic hingeregion, and (b) to facilitate the connection of braces to the RC footing at the bottom andof shear collector beam to the steel cage at the top.

Aluminum plates made of alloy-1100 of 6.5 mm thickness were used for flangeplates, web plates, and transverse stiffeners of the Al-SYD as shown in Figure 4a. Widthof flange plates and transverse stiffeners were 100 mm, whereas the length of web platewas 250 mm. Tungsten Inert Gas (TIG) welding was carried out at the interface of theseplates to form a two-paneled Al-SYD of each square panel of size 125 mm. The Al-SYDwas later annealed by heating to a high temperature �420 °C� and cooled gradually at arate of 30 °C per hour. This annealing process relieved any residual stress induced inpanels due to welding of plates.

A shear collector beam of Indian Standard I-section (ISMB [email protected] N/m) wasplaced between the existing RC beam and the Al-SYD (Figure 1). The main function ofthis collector beam is to transfer the frame lateral load to the Al-SYD, thereby, integrat-ing the aluminum shear panels with the existing RC frame. The collector beam was con-nected to the steel cage by means of a simple connection using high-strength bolts. Asmall gap was left between the collector beam and the existing RC beam to accommo-date possible bending deformation of the collector beam due to lateral load on the frame.The main advantage of this arrangement is that it does not require any intervention tothe existing RC beam, such as chipping, drilling, etc., which may adversely affect itsoverall load carrying capacity.

The Al-SYD was connected to the collector beam at the top and to the T-hanger atthe bottom using high strength bolts as shown in Figure 4b. The T-hanger was supportedby two braces made of steel tubular sections of size 50 mm�50 mm�3 mm placed in

Figure 4. (a) Side view and sectional view of the Al-SYD and (b) arrangement of the Al-SYDbetween collector beam and T-hanger in the test specimen.


an inverted-V (chevron) pattern. One end of braces was welded to web of the T-hangerand the other end was connected to the stiffener attached to the steel cage near RC foot-ing. These braces were not designed to dissipate seismic energy; instead, these were in-tended to provide an unyielding support for the Al-SYD, so that it can undergo sheardeformations due to lateral displacement of the test specimen.

MATERIAL PROPERTIES

Cement concrete of mix proportion 1.00(cement):1.62(sand):2.33(coarse aggregate)and of characteristic compressive strength as 25 MPa was used in RC frame members ofthe test specimen. The target cube compressive strength of design mix at 28-days was31.6 MPa and the values of water-cement ratio and compaction factor were taken as0.52 and 0.9, respectively. Actual compressive strengths of 150 mm size concrete cubesat different days of curing are reported in Table 2, along with tensile stress-strain prop-erties of other materials used in the test specimen.

Thermomechanically treated (TMT) bars were used as longitudinal reinforcement invarious frame members of test specimen. TMT bars were also used as shear reinforce-ment in the footing, whereas high-strength wires were used as shear reinforcement inbeam and columns. For wires, a higher yield strength and smaller ultimate strain wereobserved as compared to TMT bars, as shown in Figure 5a. Typical tensile stress-strainresponse of aluminum plates in both unannealed and annealed conditions is shown inFigure 5b. Yield strength of rebars and aluminum plates was considered as the proofstrength, i.e., stress corresponding to 0.2% strain level. The annealing of aluminumplates reduced the yield strength by 45% and increased the ultimate strain by 36.5%.Rolled steel sections used for steel cage and shear collector beam were made of mildsteel with specified yield strength of 250 MPa.

Table 2. Actual properties of materials used in the test specimen

MaterialCompressive

strength (MPa)Yield strength

(MPa)Tensile

strength (MPa)Ultimate

strain (%)

Concrete @ 7-days 21.7 — — —@ 28-days 33.9@ day of testing 41.0

Wire (4 mm dia.) — 699 795 7.5Wire (5 mm dia.) — 665 768 7.5Rebar (10 mm dia.) — 460 584 8.0Rebar (16 mm dia.) — 420 565 11.2Aluminum (Unannealed) — 106 147 11.0Aluminum (Annealed) — 58 108 14.6Brace — 460 516 17.9


TEST SET-UP AND LOAD-TRANSFERRING ASSEMBLIES

An MTS hydraulic actuator of 250 kN capacity and 250 mm stroke was used to ap-ply cyclic lateral load at RC beam level of the test specimen. The actuator was supportedby a reaction frame firmly held to the laboratory strong floor (Figure 1). Two rolled steelchannel sections, i.e., Indian Standard ISLC [email protected] N/m (BIS 1964), were placedon top and bottom of the slab to help transfer lateral load to the test frame. RC footingof the specimen was held firmly to the laboratory strong floor using 45 mm diameterhigh-strength studs. The specimen was also laterally supported to prevent any out-of-plane movement using four stiff triangular vertical frames firmly held to the laboratoryfloor. Angle sections connecting two vertical frames on each side of the specimen servedas guide beams for three roller bearings at the slab level. This arrangement not only re-strained the out-of-plane movement but also facilitated nearly friction-free in-planemovement of the test frame.

The connection of steel cage to the RC footing was designed for the maximum mo-ment expected to develop at collapse mechanism of the strengthened frame. Base platesof size 250 mm�250 mm were attached to the footing using 20 mm diameter anchorbolts (Figure 6). These bolts were perfectly bonded to the hardened concrete of footingusing 2:1 (hardener: resin) mix epoxy mortar. Rib plates of size 100 mm�100 mm�12 mm were welded to the end battens of steel cage and the base plates on both sidesof columns. In addition, a rib plate of larger size 150 mm�150 mm�12 mm connect-ing the end batten and base plate was placed between smaller rib plates on one side ofboth columns for the connection of braces to the RC footing. The main advantage of thisscheme is that all connections are steel-to-steel except the base plate to footing connec-tion, which makes it simpler and faster to install.

INSTRUMENTATION

The instrumentation used in the slow-cyclic testing of the test frame consisted of (a)load cell and displacement transducer installed in the actuator, (b) linear variable differ-ential transformers (LVDTs), and (c) quarter-bridge electrical resistance strain gauges.

Figure 5. Tensile stress-strain curves obtained from coupon tests: (a) rebars and (b) aluminumplates.


Locations of strain gauges and LVDTs in the test specimen are shown in Figure 7. Theload cell and displacement transducer in the actuator measured the applied lateral loadand lateral displacement of the west end of beam, whereas the displacement of the eastend of beam was measured by means of LVDTs. Three LVDTs were used to measure thelateral displacements of each column. In addition, two more LVDTs were used on bothsides of the footing to monitor any sliding of the base of specimen. Several strain gaugeswere fixed to the longitudinal reinforcements of RC members, the bracing members, andthe Al-SYD to monitor the state of strain at different drift levels.

Figure 6. Details of Steel cage-to-footing connections: (a) top view and (b) elevation.

Figure 7. Position of LVDTs and strain gauges in the test specimen (All dimensions are in mm,
Bold numerals show the position of LVDTs).


LOADING HISTORY

The gravity load on the test frame was simulated using sand bags and a pressure of7.3 kPa was applied on the slab. Displacement history considered in the present studyconsisted of reversed cyclic lateral displacements as per ACI 374.1-05 (2006) provi-sions. Story drift is defined as the ratio of story displacement to the story height mea-sured from the footing top to the centerline of RC beam. As shown in Figure 8, the dis-placement history consisted of three repetitive cycles of story drifts of magnitude 0.20%,0.35%, 0.50%, 0.75%, 1.10%, 1.40%, 1.75%, 2.20%, 2.75%, and 3.50%. One cycle ofsmaller story drift is applied before proceeding to repetitive cycles of the next higherdrifts. The 3.5% drift limit is considered as satisfying criteria for the test specimen assuggested in ACI 374.1-05 (2006) because, in general, conventional moment resistingRC frames failed to achieve story drift of 3.5% on a consistent basis.

EXPERIMENTAL RESULTS

NATURAL FREQUENCY AND INITIAL STIFFNESS

Forced-vibration tests were carried out to study variation in natural frequency andlateral stiffness of the test frame with various strengthening elements. An electro-dynamic shaker (Make: Electro-seis, USA) was placed at center of the slab to excite thetest frame at low-strain level. The response of test frame was measured using a rangerseismometer (Make: Kinemetrics, USA). The excitation frequency was varied from1 Hz to 20 Hz in an increment of 0.5 Hz. Figure 9a shows the velocity response of thetest specimen at different stages of strengthening. The test frame strengthened only withsteel caging showed an increase of natural frequency by 9% as compared to the RC(bare) frame, whereas the corresponding increase for the test frame strengthened withboth steel caging and Al-SYD was about 45% (Table 3). A minor increase in equivalentdamping was also observed at both strengthening stages of the test frame for low-strainexcitation levels caused by the electro-dynamic shaker.

The lateral stiffness of test frame at different stages of strengthening was predictedusing measured values of natural frequency and estimating the portion of total mass par-

Figure 8. Displacement history used in the present study.


ticipating in the excitation. The excited mass of RC frame was assumed as the sum of (a)one-third mass of columns, (b) total mass of slab and beam, and (c) applied load due tosand bags. Further, the additional masses of steel cage, collector beam, and Al-SYDwere considered for the test frame at the respective strengthening stages. Table 3 sum-marizes the excited mass considered in the present study and the values of lateral stiff-ness of test frame at different stages of strengthening. The increase in lateral stiffness ofthe test frame strengthened only with steel cage was about 15% as compared to the RC(bare) frame and this increase can be neglected for design purposes. On the other hand,the corresponding increase in lateral stiffness was about 120% for the frame strength-ened with both steel caging and Al-SYD.

In addition to forced-vibration tests, load-controlled slow-cyclic tests were also con-ducted for a peak lateral load of ±5 kN to study variation in initial lateral stiffness of thetest frame. As shown in Table 3, values of initial lateral stiffness of the test frame withvarious strengthening elements matched well with the respective values observed inforced-vibration tests.

Figure 9. (a) Amplitude-frequency response of test frame. (b) Details of mass considered toestimate the lateral stiffness of test frame (m1=Mass of sand bags; m2=Combined mass of slab-beam; m3c=One-third mass of RC column; m3s=One-third mass of steel cage; m4=Mass ofcollector; m5=Mass of Al-SYD).

Table 3. Properties of test frame at different stages of strengthening

Stages of testframe

Forced-vibration test Slow-cyclic test

Naturalfrequency (Hz)

Equivalentdamping (%)

Excitedmass (kg)

Stiffness(kN/mm)

Stiffness(kN/mm)

RC (bare) frame 8.9 2.4 2505 7.8 7.4RC frame+steelcaging

9.7 2.7 2529 9.4 8.4

RC frame+steelcaging+Al-SYD

12.9 3.1 2565 16.8 18.2


HYSTERETIC BEHAVIOR

A displacement-controlled slow-cyclic test at higher story drift levels was carried outfor the test frame strengthened with both steel caging and Al-SYD. As shown in Figure10a, both RC beam and strengthened columns did not show any major damage even upto 3.5% story drift. Initially, some minor shear and flexural cracks in the cover concretewere observed in beam-column joints on both sides of the frame at a story drift of 1.0%as shown in Figure 10b. However, these cracks did not extend at the higher drift levelsdue to shear yielding of web panels of the Al-SYD. As expected, braces and collectorbeam of the strengthened frame did not show either yielding in tension or buckling un-der compression. Moreover, the behavior of connections of steel cage and braces to theRC footing was satisfactory as the premature failure due to yielding of plates, failure ofwelding, pull out of anchor bolts, etc. was not observed even up to 3.5% story drift. Asdesired, shear yielding of web plates and the buckling of stiffeners of the Al-SYD wasobserved, which enhanced the energy dissipation potential of the strengthened frame.

Linear elastic behavior of the strengthened frame was observed up to 0.35% storydrift and the opening of hysteretic loops was observed at a lateral load of 75 kN at 0.5%story drift indicating the initiation of inelastic behavior. The higher lateral stiffness ofthe strengthened frame was observed in first cycle of each excursion levels at the largerstory drifts, which was stabilized in subsequent repetitive cycles after some initial ad-justments. However, the magnitude of lateral load at any excursion level did not changewith repetitions of displacement cycles and continued to increase at the higher story driftlevels indicating significant strain-hardening behavior of the strengthened frame, asshown in Figure 11a. The maximum values of lateral load carried by the frame at storydrift of 3.5% were 213.9 kN, and −218.1 kN in the pull (west) and the push (east) di-rections, respectively.

Figure 10. (a) Deformed state of strengthened frame at 3.5% story drift and (b) close view ofcracks in beam-column joints in east side of the frame.


It is reasonable to assume that total lateral load carried by the strengthened frame isthe algebraic sum of lateral loads carried by the Al-SYD and the test frame withstrengthened columns. The lateral load carried by Al-SYD can be considered as thecomponent of brace force along the direction of story drift. Thus, the contribution ofsteel cage to lateral load carried by the test frame at any drift level was computed bydeducting the horizontal component of brace forces from the overall lateral load carriedby the test frame strengthened with both Al-SYD and steel caging as shown in Figure11b. The maximum contribution of the steel cage was observed to be one-third (70 kNat 3.5% story drift) of the total resistance offered by the test frame strengthened withboth Al-SYD and steel caging. Due to this minor role played by the steel caging in loadsharing with the Al-SYD, the steel caging can be restricted only to critical regions, suchas column ends.

BEHAVIOR OF BRACES

Brace axial force due to lateral displacement of the strengthened frame was obtainedby multiplying the cross-sectional area of brace with average value of axial stress deter-mined from the strain gauge data at each story drift level. The yield strength of braceswas computed as 259.4 kN for sectional area of 564 mm2 and the actual yield stress of460 MPa, whereas the buckling strength was determined as 162.7 kN, for slendernessratio of 65.2 taking effective length factor as 0.9. Figure 12 shows axial load carried bythe braces at different levels of story drift. As expected, both braces of the strengthenedframe exhibited elastic behavior and the observed maximum axial force of 100 kN re-sisted at 3.5% story drift was smaller than their yield and buckling strengths.

BEHAVIOR OF COLLECTOR BEAM

The bending moment exerted on the collector beam was computed from the state ofstrain measured using strain gauges fixed at two sections on the bottom flange, each at adistance of 250 mm from the mid-section of Al-SYD. Sections to the left and the right of

Figure 11. Hysteretic response of strengthened frame. (a) Observed hysteretic response for thetest frame with steel cage and Al-SYD. (b) Contribution of steel cage to the overall hystereticresponse.


the Al-SYD as shown in Figure 6 are referred as the west and the east sections, respec-tively. Under the action of lateral load, the collector beam deformed in double curvatureand behaved elastically up to story drift of 3.5% as the maximum value of strain mea-sured in the flanges were less than the nominal yield strain corresponding to yield stressof 250 MPa.

As shown in Figure 13a, maximum values of bending moment in the collector beamwere computed as 6.8 kNm and 15.6 kNm, respectively, in push and pull directions ofcyclic loading at 3.5% story drift. The observed maximum bending moment for the col-lector was about 56% of its plastic moment capacity of 26.6 kNm. At smaller drift lev-els, bending moment on the collector beam was symmetric in both directions of lateralloading; however, significant difference was noted at larger drift levels beyond 1.4%,when the vertical displacement of collector beam was restrained by the RC beam afterfree movement through the initial gap, as shown in Figure 13b.

Figure 12. Variation of axial force in braces with story drifts levels.

Figure 13. (a) Variation of bending moment of the collector beam with story drift and (b) pro-
gressive bending of collector beam with increased drift levels.


LATERAL STIFFNESS

The initial lateral stiffness for the RC frame and the frame strengthened only withsteel caging were about 40% and 46%, respectively, of the strengthened frame with steelcaging and Al-SYD, indicating significant contribution of braces and Al-SYD to overalllateral stiffness of the strengthened frame. As shown in Figure 14, lateral stiffness of thestrengthened frame reduced at a very high rate up to story drift of 1.0% beyond whichthe reduction was very negligible. At the end of cyclic excursion level of 3.5% storydrift, stiffness of the strengthened frame was about 25% of its initial value of18.2 kN/mm.

The actual and residual stiffness of the strengthened frame at different drift levels isalso compared in Figure 14. Residual stiffness of the frame was obtained from the maxi-mum lateral load and the story displacement measured during small cyclic excursionlevels that follow three repetitive cycles of larger drift levels as per the displacement his-tory shown in Figure 8. At the initial drift levels, significant difference was noted be-tween actual and residual stiffness. However, this difference reduces with the increase instory drift levels and beyond 1.75% story drift, the residual stiffness of the strengthenedframe was nearly equal to its actual stiffness.

STATE OF STRAIN IN REBARS AND ANGLES

The stress levels in column rebars (i.e., SG #1 and #4) placed on the same face of thewest column were well below the yield limit of 420 MPa as shown in Figure 15a. On theother face of the west column, rebars (i.e., SG #2 and #3) experienced larger stress lev-els, and the rebar corresponding to SG #2 reached the yield limit. Furthermore, straingauges on the angle sections of the west column reached the yield strain level corre-sponding to nominal yield of 250 MPa for angles. The lateral restraint provided by thesteel cage to the dilation of column concrete, in addition to the general bending resis-tance, probably caused the steel cage to experience larger strains. Rebars in the east col-umn, away from the actuator, did not reach the yield limit in either direction of lateral

Figure 14. Normalized Stiffness vs. story drift response of the strengthened frame.


loading as shown in Figure 15b. However, steel angles of the east column reached theiryield strength limit at a story drift of 2.75%. Moreover, strain gauges in the steel cage ofeast column measured smaller strains as compared to those in the west column.

BEHAVIOR OF AL-SYD

Shear force transferred to the Al-SYD is equal to the algebraic sum of horizontalcomponents of brace axial forces, which divided by the total area of web plate providesan estimate of shear stress in Al-SYD panels. The shear stress-shear strain response ofthe Al-SYD at different story drift levels is shown in Figure 16. Full and stable hystereticloops were observed with significant strain-hardening behavior up to 2.75% story drift.The yielding of web plates of Al-SYD was observed at 0.25% story drift correspondingto a shear stress value of 26.7 MPa. The maximum values of shear stress and shearstrain exhibited by the Al-SYD were about 85.5 MPa and 19.5%, respectively, at 2.75%

Figure 15. State of strain in rebars and steel cage of strengthened columns: (a) west columnand (b) east Column (SG#1, 2, 3, 4, 23, and 24 are on the rebars; SG#27, 28, 29, 30, 31, and 34are on the steel cage).

Figure 16. Shear stress-strain response of the Al-SYD. (a) Using strain gauge data at smaller
story drift levels (0.2—1.75%). (b) Using LVDT data at larger story drifts (1.0-2.75%).


story drift. The maximum value of lateral load carried by the device at 3.5% story driftwas about 150 kN, which corresponds to shear stress of 92.5 MPa. However, shearstrain at 3.5% story drift level could not be determined due to slippage of attachmentsfor LVDTs.

The energy dissipation potential of Al-SYD primarily depends on inelastic shear de-formation of web panels without premature shear buckling. As expected, both web pan-els of the Al-SYD did not show any sign of buckling even up to story drift of 3.5%.However, bending of flange plates observed at larger story drift levels suggested somerocking of the Al-SYD. As shown in Figure 17a, fracture of flange plates was also ob-served near the bolt hole and along the weld line at the interface of plates. This type offailure may have reduced the effective restrain to the web plate and allowed its rigid-body movements. In addition, the bearing failure of bolt holes was noted in both flangesof the Al-SYD resulting in elongated holes as shown in Figure 17b. This may have oc-curred as the harder steel bolts ‘bite’ into soft aluminum flange of the Al-SYD. The useof thicker flange plates and proper welding of aluminum plates may delay such localfailures resulting in even better energy dissipation potential for the Al-SYD. Due to theseconnection-related problems at the story drift of 3.5%, the slow-cyclic testing ofstrengthened frame was stopped at this drift limit.

ENERGY DISSIPATION AND DAMPING

The total energy dissipated at any drift level by the strengthened frame and the Al-SYD was calculated as the average area of hysteretic loops enclosed in three cycles ofdisplacement excursion and plotted as shown in Figure 18a. The maximum value of en-ergy dissipated by the frame per cycle was found to be 8.5 kNm at story drift of 3.5%.The energy dissipation response of the strengthened frame closely followed that of theAl-SYD, which confirmed that the major portion of total dissipated energy was due toinelastic shear deformation of the Al-SYD as shown in Figure 18b.

At smaller drift levels, the contribution of Al-SYD to the overall dissipated energywas very small, but significant contribution was noted at larger story drifts. At 1% story

Figure 17. State of the Al-SYD after the end of testing at 3.5% story drift level: (a) fracture inflange plates and (b) elongation of bolt-hole due to bearing failure


drift, the contribution of Al-SYD to the overall dissipated energy was about 50%. How-ever, due to slippage in attachments for LVDTs and connection-related problems of theAl-SYD, the amount of energy dissipated by the Al-SYD appears to be somewhat over-estimated for story drifts greater than 2%.

The equivalent viscous damping potential of the strengthened frame was calculatedfrom the observed energy dissipation as per FEMA 368 provisions (BSSC 2001). Assummarized in Table 4, the equivalent damping of the strengthened frame varied from3.81% at 0.2% story drift to 12% at 3.5% story drift, which indicates significant increasein damping potential over that of the RC (bare) frame for which the equivalent dampingof 2.4% was obtained from the forced vibration test.

Figure 18. (a) Comparison of energy dissipation potential of device and strengthened frame atdifferent drift levels (b) Deformed configuration of the Al-SYD at 3.5% story drift

Table 4. Equivalent viscous damping of strengthened frame as per FEMA 368 (BSSC 2001)

Driftlevel(%)

Energydissipation,

Eloop

(kNmm)

Peak positivedisplacement,

�+ (mm)

Peak negativedisplacement,

�− (mm)

Peakpositive force,

F+ (kN)

Peaknegative

force,F− (kN)

Effectivestiffness,

Keq

(kN/mm)

Effectivedamping,�eq (%)

0.20 30.6 2.93 −2.90 40.64 −46.96 15.01 3.810.25 67.7 3.73 −3.71 48.68 −55.66 14.02 5.560.35 156.90 5.13 −5.12 60.63 −67.71 12.52 7.590.50 374.22 7.46 −7.43 75.19 −79.53 10.39 10.340.75 714.32 11.06 −11.04 91.11 −95.60 8.45 11.031.00 1036.6 14.76 −14.72 105.72 −110.88 7.35 10.331.40 1761.43 20.67 −20.63 131.54 −136.65 6.49 10.121.75 2495.07 25.78 −25.76 151.97 −155.90 5.97 10.012.20 3727.56 32.28 −32.35 174.55 −178.06 5.45 10.402.75 5650.61 40.58 −40.52 195.67 −200.81 4.89 11.193.50 8516.8 51.53 −51.43 213.93 −218.09 4.20 12.19

+ − + − + − 2
Keq= �F −F � / �D −D �; �eq= �2/�� Eloop /Kmax�D −D � �


COMPARISON WITH ANLAYTICAL PREDICTION

The observed lateral strength of strengthened frames was compared with expecteddesign values using a simplified analytical model as shown in Figure 19a. Section ca-pacities of RC beam and columns were determined using stress-strain relationships forreinforcement and concrete as specified in Indian Standard code IS:456 (BIS 2000) andshown in Figures 19b and 19c. Considering actual material properties, moment capaci-ties of RC column and beam section were found to be 19.3 kNm, and 24.7 kNm, re-spectively. However, the moment capacity of strengthened column was determined con-sidering the confinement of the concrete due to external steel cage using a theoreticalmodel proposed by Nagaprasad (2005). The procedure for determining confined com-pressive strength of column concrete is summarized in the Appendix and it was calcu-lated as 48.1 MPa for the concrete of unconfined compressive strength of 33.9 MPa.Using the confined concrete strength, the moment capacity of strengthened column wascomputed as 33.5 kNm, by summing moment capacities of the RC column section andthe steel cage.

The lateral strength of the frame strengthened with steel cage and Al-SYD is sum ofthe lateral strength of the frame with steel cage strengthened columns and the shearstrength of Al-SYD. For sway-type mechanism with plastic hinges in beams and col-umns, lateral strength of the RC bare and strengthened frames was calculated (Table 5).It should be noted that while the existing RC (bare) frame was strong beam-weak col-umn system, both strengthened frames became more desirable strong column-weakbeam system. The observed lateral strength at story drift of 3.5% for strengthenedframes closely matched with the expected design values. Since the complete collapse ofthe frame strengthened with both Al-SYD and steel cage was not observed at 3.5% storydrift, the frame was capable of carrying more lateral load because of strain-hardeningbehavior of the Al-SYD.

CONCLUSIONS

A novel strengthening scheme employing column steel caging and aluminum shearyielding damper (Al-SYD) as energy dissipation device was proposed to significantly

Figure 19. Assumed collapse mechanism and stress-strain response of steel and rebars:(a) sway mechanism for the RC frame, (b) design stress-strain curve of concrete, and (c) designstress-strain behavior of reinforcing bars (fck=Characteristic compressive strength of concrete,fy=Yield strength of reinforcement, Q=Lateral load).


improve the seismic performance of nonductile RC frames. The main advantages of theproposed scheme are: (a) the minimal disturbances to the existing RC frame members,(b) mostly steel-to-steel connections that can be easily implemented and (c) significantenhancement in strength, stiffness and energy dissipation potential of deficient frames. A1:2.5 reduced scaled model of nonductile single-story RC frame strengthened as per theproposed scheme was experimentally investigated under combined gravity load andreversed-cyclic lateral displacement.

The strengthened frame exhibited full and stable hysteretic loops with complete ab-sence of pinching even up to story drift of 3.5%, with significant increase in lateralstrength and stiffness of the nonductile RC frame. Moreover, full and stable hystereticloops associated with significant post-yield strain-hardening behavior was observed inthe Al-SYD panels indicating excellent energy dissipation through hysteresis. A fivetimes increase in equivalent damping potential of the strengthened frame was observed,which helped reduce inelastic demand and, hence, damage in primary RC members upto 3.5% story drift. Due to minor role played by the steel cage in load sharing with theAl-SYD, it is suggested that the steel caging can be limited only to critical regions, suchas, column ends. Further, the proposed load-transferring mechanism and related connec-tions showed no sign of failure and both braces and shear collector beam of the strength-ening system behaved elastically up to 3.5% story drift. A simple analytical model basedon collapse mechanism was able to predict the lateral strength of strengthened framewith acceptable accuracy.

APPENDIX

In this study, the compressive strength of concrete confined by external steel cagewas estimated using a procedure proposed by Nagaprasad (2005), which is quite similarto the confinement model proposed by Mander et al. (1988) for concrete confined byrectangular hoops. The effective confinement area in plan and elevation of the columnwas determined by assuming a parabolic arching action of confining stresses between

Table 5. Lateral strength of the frames at ultimate condition

Frame Section properties Design strength Observed strength

RC (bare)frame

Moment capacity of RC Column,Mrc=19.3 kNm Qrc=

4Mrc

1.44=53.6 kN

—

Beam capacity, Mpb=24.7 kNmRC frame+Steelcaging

Moment capacity of strengthenedcolumn, Mcs=33.5 kNm Qps=

2�Mpb+Mcs�

1.44=80.8 kN

70.0 kN[Figure 11b]

Beam capacity, Mpb=24.7 kNmRC frame+Steelcaging+Al-SYD

Shear stress at 0.2 strain (=1.86�tensile yield stress) =108 MPa(Rai and Wallace 1998)Area of Al-SYD panel=1625 mm2

Shear strength, Qsyd=175.5 kN

Qfs=Qps+Qsyd=256.3 kN 216.0 kN[Figure 11a]


the steel angles as shown in Figure 20. A systematic procedure for determining the con-fined compressive strength for the column concrete is presented as follows:

Properties of RC column: Width of RC column, bx=160 mm; Depth of RC column,by=160 mm; Compressive strength of unconfined concrete, fck=34 MPa; Gross area ofRC column, Ac=bxby=25,600 mm2; Area of main reinforcements, Ast=804 mm2; Areaof concrete section, Acc=Ac-Ast=24,800 mm2.

Properties of steel caging: Depth of batten, d=50 mm; Thickness of batten, t=5 mm; Area of intermediate battens, As=dt=250 mm2; Yield stress of batten, fy

=250 MPa; Center-to-center spacing of battens, s=180 mm; Width of steel angles, c=35 mm; Effective spacing of battens, se=s-�d+2c�=60 mm.

Confined compressive strength of concrete: Effective width of column section, bxe

=bx−2c=90 mm; Effective depth of column section, bye=90 mm; Area of column in-effective in confinement, Ainef= �bxe2 +bye2� /3=5,400 mm2; Area of column effective inconfinement, Aef=Ac-Ainef=20,200 mm2.Lateral confining pressures in column section,

�x = �y =2Asfy

bys= 4.3 MPa

Reduction in effectiveness in elevation, Reduction in effectiveness in plan,

�s =�bx − se/2��by − se/2�

bxby= 0.66

Figure 20. Effective area of confinement for the strengthened column (a) Plan (b) Elevation


�n =Aef

Acc= 0.82

Effective confining stresses,

�xe = �ye = �s�n�x = 2.3 MPa

The confined compressive strength of concrete was obtained using the general solu-tion of multi-axial failure criterion developed by Mander et al. (1988). For the computedvalues of effective confining stresses 2.3 MPa in both directions, the confining factor, kfor the column concrete was determined as 1.42. Thus, confined compressive strength ofconcrete for the strengthened column, fcc=kfck=48.1 MPa.

REFERENCES

Abdullah, and Takiguchi, K., 2003. An investigation into the behavior and strength of rein-forced concrete columns strengthened with ferrocement jackets, Cem. Concr. Compos. 25,233–242.

Aboutaha, S., Engelhardt, M. D., Jirsa, J. O., and Kreger, M. E., 1999. Rehabilitation of shearcritical concrete columns by use of rectangular steel jackets, ACI Struct. J. 96, 1, 68–78.

American Concrete Institute (ACI 374.1-05), 2006. Acceptance Criteria for Moment FramesBased on Structural Testing and Commentary—An ACI Standard.

Bett, B. J., Klinger, R. E., and Jirsa, J. O., 1988. Lateral load response of strengthened and re-paired reinforced concrete columns, ACI Struct. J. 85, 499–508.

Building Seismic Safety Council (BSSC), 2001. NEHRP Recommended Provisions for SeismicRegulations for New Buildings and Other Structures, Part 1-Provisions, Prepared for theFederal Emergency Management Agency, FEMA 368, Washington, D. C.

Bureau of Indian Standards (BIS), 1964. Handbook for Structural Engineers, Part 1: StructuralSteel Sections, SP: 6(1), New Delhi.

Bureau of Indian Standards (BIS), 1989. Indian Standard Code of Practice for Design Loads(Other than Earthquake) for Buildings and Structures, IS: 875, Part-3 (Wind Loads), 2ndrevision, Reaffirmed 2003, New Delhi.

Bureau of Indian Standards (BIS), 2000. Indian Standard Plain and Reinforced Concrete-Codeof Practice, IS: 456, 4th revision, New Delhi.

Chai, Y. H., Priestley, M. J. N., and Seible, F., 1990. Seismic retrofit of circular bridge columnsfor enhanced flexural performance, ACI Struct. J. 88, 572–584.

Dritsos, S., and Pilakoutas, K., 1992. Composite technique for repair/ strengthening RC mem-bers, in Proc. of 2nd International Symposium on Composite Materials and Structures,Beijing, China, 958–963.

Endo, T., Okifuji, A., Sugano, S., Ayashi, T., Shimizu, T., Takahara, K., Saito, H., and Yon-eyama, Y., 1984. Practices of seismic retrofit of existing concrete structures in Japan, inProc. of 8th World Conference on Earthquake Engineering, vol. 1, San Francisco (CA), pp.469–476.

Jain, S., Rai, D. C., and Sahoo, D. R., 2008. Post-yield cyclic buckling criteria for aluminumshear panels, J. Appl. Mech. 75, 021015.


Li, Y. F., Chen, S. H., Cheng, K. C., and Liu, K. Y., 2005. A constitutive model of concreteconfined by steel reinforcements and steel jackets, Can. J. Civ. Eng. 32, 279–288.

Mander, J. B., Priestly, J. N., and Park, R., 1988. A theoretical stress-strain Model for ConfinedConcrete, J. Struct. Eng. 114, 1804–1826.

Martinez-Rueda, J. E., and Elnashai, A. S., 1995. A novel technique for the retrofitting of re-inforced concrete structures. Eng. Struct. 17, 359–371.

Nagaprasad, P., 2005. Enhancement of Flexural Capacity of Undamaged RC Columns by aComposite Steel Caging Technique, M. Tech. thesis, Indian Institute of Technology Kanpur,India.

Rai, D. C., and Wallace, B. J., 1998. Aluminium shear-links for enhanced seismic resistance.Earthquake Eng. Struct. Dyn. 27, 315–342.

Rai, D. C., and Wallace, B. J., 2000. Aluminium shear-link for seismic energy dissipation, inProc., 12th World Conference on Earthquake Engineering, Auckland, New Zealand, Jan. 30-Feb. 4, Paper No. 0279.

Rodriguez, M., and Park, R., 1994. Seismic load tests on reinforced concrete columns strength-ened by jacketing, ACI Struct. J. 91, 150–159.

Sakino, K., and Ishibashi, H., 1985. Experimental studies on concrete filled square steel tubularshort columns subjected to cyclic shearing force and constant axial force, J. Struct. Constr.Eng. 353 (Transactions of the Architectural Institute of Japan), 81–89.

Soong, T. T., and Dargush, G. F, 1997. Passive Energy Dissipation Systems in Structural En-gineering, John Wiley & Sons, New York, 368 pp.

Tastani, S. P., and Pantazopoulu, S. J., 1994. Experimental evaluation of FRP jackets in upgrad-ing RC corroded columns with substandard detailing, Eng. Struct. 26, 817–829.

Xiao, Y., and Ma, R., 1997. Seismic retrofit of RC circular columns using prefabricated com-posite jacketing, J. Struct. Eng. 123, 1357–1364.

Zhang, R. R., and Soong, T. T., 1992. Seismic design of viscoelastic dampers for structural ap-plications, J. Struct. Eng. 118, 1375–1392.

(Received 12 September 2007; accepted 23 December 2008�

Documents

A Novel Technique of Seismic Strengthening of Nonductile RC