Performance-Based Seismic Retrofit Strategy for Existing ... - CiteSeerX

Performance-Based Seismic Retrofit Strategy for Existing 2 Reinforced Concrete Frame Systems Using Fiber-Reinforced 3 Polymer Composites 1

Stefano Pampanin1; Davide Bolognini2; and Alberto Pavese2

4 5

6 Abstract: The feasibility and efficiency of a seismic retrofit intervention using externally bonded fiber-reinforced polymer composites on 7 existing reinforced concrete frame systems, designed prior to the introduction of modern standard seismic design code provisions in the 8 mid-1970s, are herein presented, based on analytical and experimental investigations on beam-column joint subassemblies and frame 9 systems. A multilevel retrofit strategy, following hierarchy of strength considerations, is adopted to achieve the desired performance. The 10 expected sequence of events is visualized through capacity-demand curves within M-N performance domains. An analytical procedure 11 able to predict the enhanced nonlinear behavior of the panel zone region, due to the application of CFRP laminates, in terms of shear 12 strength 共principal stresses兲 versus shear deformation, has been developed and is herein proposed as a fundamental step for the definition 13 of a proper retrofit solution. The experimental results from quasistatic tests on beam-column subassemblies, either interior and exterior, 14 and on three-storey three-bay frame systems in their as-built and CFRP retrofitted configurations, provided very satisfactory confirmation 15 of the viability and reliability of the adopted retrofit solution as well as of the proposed analytical procedure to predict the actual sequence 16 of events. 17 DOI: XXXX 18 CE Database subject headings: Concrete, reinforced; Frames; Beam columns; Fiber reinforced polymers; Retrofitting. 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

Introduction Extensive experimental-analytical investigations on the seismic performance of existing reinforced concrete 共RC兲 frame buildings, primarily designed for gravity loads, as typically found in most seismic-prone countries before the introduction of adequate seismic design code provisions in the 1970s, have confirmed the expected inherent weaknesses of these systems 共Aycardi et al. 1994; Beres et al. 1996; Hakuto et al. 2000; Park 2002; Pampanin et al. 2002; Bing et al. 2002; Calvi et al. 2002a,b兲. As a consequence of poor reinforcement detailing, lack of transverse reinforcement in the joint region as well as absence of any capacity design principles, brittle failure mechanisms are expected. At a local level, most of the damage is likely to occur in the beamcolumn joint panel zone while the formation of soft-story mechanisms can greatly impair the global structural performance of these RC frame systems. An appropriate retrofit strategy is therefore required, which is capable of providing adequate protection to the joint region while modifying the hierarchy of strengths between the different components of the beam-column connections according to a capacity design philosophy. 1 Dept. of Civil Engineering, Univ. of Canterbury, Christchurch, New Zealand. E-mail: [email protected] 2 Dept. of Structural Mechanics, Univ. of Pavia, Italy. Note. Discussion open until September 1, 2007. Separate discussions must be submitted for individual papers. To extend the closing date by one month, a written request must be filed with the ASCE Managing Editor. The manuscript for this paper was submitted for review and possible publication on September 21, 2005; approved on February 6, 2006. This paper is part of the Journal of Composites for Construction, Vol. 11, No. 2, April 1, 2007. ©ASCE, ISSN 1090-0268/2007/2-1–XXXX/ $25.00.

Alternative retrofit and strengthening solutions for reinforced concrete buildings have been studied in the past and adopted in practical applications. A comprehensive overview of traditional seismic rehabilitation techniques was presented by Sugano 共1996兲. Conventional techniques which utilize braces, jacketing, or infills as well as more recent approaches including base isolation and supplemental damping devices have been considered. Most of these retrofit techniques have evolved in viable upgrades. However, issues of costs, invasiveness, and practical implementation still remain the most challenging aspects of these solutions. The results of numerical and experimental investigations on a noninvasive and economical retrofit solution based on metallic haunch connections have, for example, been recently presented by Pampanin and Christopoulos 共2003兲 and Pampanin et al. 共2006兲. In the past decade, an increased interest in the use of advanced nonmetallic materials, including shape memory alloys 共Dolce et al. 2000兲 or fiber reinforced polymers 共FRP兲 共FIB 2001, 2006兲, has been observed. In this contribution, the feasibility and efficiency of a retrofitting intervention using FRP composite materials, according to a multilevel performance-based approach, will be presented. Depending on the joint typology 共interior or exterior兲 and on the structural details adopted, alternative objectives, in terms of hierarchy of strength and sequence of events within the beam-column-joint system, can be targeted and achieved. After a summary of the experimental campaign on seismic vulnerability of existing underdesigned beam column subassemblies and frame systems, representing the basic as-built configuration 共benchmark兲 for this study and presented in previous publications 共Pampanin et al. 2002; Calvi et al. 2002a,b兲, the main focus will be given herein to the description of 共a兲 the principles and theoretical developments of the conceptual retrofit strategy, 共b兲 the main features of a simplified analytical model

JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007 / 1

41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73

Table 1. Specimen Reinforcement and Section Geometry Joint type

Specimen

Exterior

T1A,T1Bb

T2A T2B Interior

a

b

Section dimensions 共mm兲 Beam

330⫻ 200a

Columna Beam

200⫻ 200a

C1b, C3b

Beam

C2, C4

Beam

Longitudinal reinforcement

Transverse reinforcement

Top 2␾8 + 2␾12; Bottom 2␾8 + 2␾12 3␾8 + 3␾8a Top 2␾8 + 1␾12; Bottom 2␾8 + 1␾12 Top 2␾8 + 3␾12; Bottom 2␾8 + 1␾12 Top 2␾8 + 2␾12; Bottom 2␾8 + 1␾12

␾4@115 mma ␾4@13 mma

Equal reinforcement for all specimens. Reinforced with FRP.

b

74 adopted to evaluate the increase in the joint shear strength due to 75 the application of FRP, 共c兲 the assessment of the internal hierar76 chy of strength through M-N 共moment-axial load兲 performance 77 domains to account for the variation of axial load in the column, 78 共d兲 the implementation of the proposed solution, and 共e兲 the ex79 perimental validation of the intervention via quasistatic cyclic 80 tests on four beam-column joint subassemblies and one three81 storey three-bay frame system, 2 / 3 scaled, retrofitted with CFRP 82 sheets. Comparisons with the response of the benchmark 共i.e., 83 as-built configuration兲 subassemblies and frame specimen are car84 ried out to emphasize the enhanced behavior of the retrofitted 85 configurations as well as the general reliability of the overall 86 performance-based seismic retrofit strategy.

87 88

Seismic Behavior of Existing Poorly Detailed RC Frames

89 Experimental Investigations on As-Built Systems 90 and Subassemblies 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113

The first phase of the research project involved the assessment, through analytical and experimental investigations, of the seismic vulnerability of existing reinforced concrete frame systems, primarily designed for gravity loads as typically found in major seismic prone countries in the period between the 1950s and the 1970s, before the introduction of modern seismic design provisions in the mid-1970s. In order to facilitate the introduction of the proposed retrofit strategy as well as to provide a benchmark comparison for the experimental tests on the FRP retrofitted configurations, presented in the following paragraphs, a brief summary and overview of the experimental results on the as-built solutions is given herein. Further details on the response of the as-built specimens can be found in Pampanin et al. 共2002兲 and Calvi et al. 共2002b兲, while more extensive research reports and publications on the experimental and numerical investigations on existing pre-1970 frames are under preparation. The experimental program on existing 共as-built兲 RC frame subassemblies and systems comprised of quasistatic tests carried out in the Laboratory of the Department of Structural Mechanics of the University of Pavia on six, one-way beam-column joint subassemblies 共two exterior knee joints, two exterior tee joints, and two interior joints兲 as well as on a three-storey three-bay

frame system. Both beam-column subassemblies and frame systems were scaled at 2 / 3. Particular attention was given to the vulnerability of the panel zone region. The design recommendations provided by the current Italian national design provisions 共Regio Decreto, 1939兲 were followed and, where necessary, integrated by textbooks broadly adopted in the engineering practice and available in that period 共e.g., Santarella 1957兲. It is worth noting that typical plan configurations of existing buildings in Mediterranean seismic-prone countries would consist of frames running in one direction only with lightly reinforced slab 共perforated clay brick units with cast-insitu concrete topping兲 spanning in the orthogonal direction. Simple one-way beam-column joint subassembly specimens without transverse beam nor structural cast-in-situ slab are thus adequate representatives of such construction practice. Further experimental investigation on the seismic performance of alternative beam-column joint typologies including two-way exterior beam-column joints with and without cast-in-situ slab under bidirectional cyclic loading are currently ongoing at the University of Canterbury, Christchurch 共Hertanto, 2006兲. Table 1 reports the geometric and reinforcement details of the critical sections of the beam-column subassembly specimens. In particular, Figs. 1 and 2 show, respectively, the details of the beam-column joint specimens T1 and C2, used as benchmark configurations before the retrofit intervention, and of the two frames 共identical structural systems, tested in the as-built and retrofit configuration兲. Consistently with most of the old practice, no transverse reinforcement was placed in the joint region. Plain round bars, with mechanical properties similar to those typically used in older periods, were adopted for both longitudinal and transverse reinforcement. Beam bars in exterior joints were not bent into the joint region, but anchored with endhooks. Lap splices with hook anchorages were adopted in the beam bars crossing the interior joints 共except for the specimen C1 with continuum reinforcement兲 as well as in column longitudinal bars at each floor level above the joint region and at the column-to-foundation connection 共in the frame system兲.

114

Test Setup and Loading Protocols

152

115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151

In order to allow for a comparison between the response of the 153 as-built and the retrofitted configuration, some details in the test 154

2 / JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007

Fig. 1. Geometry and reinforcement details in exterior joint specimen T1 and interior specimen C2 共as-built solutions兲 155 setup and loading protocol were implemented for the tests on the 156 beam-column subassemblies and on the frame systems as dis157 cussed in the following paragraphs, The test setup 共Fig. 3兲 and loading history/regime 共Fig. 4兲 of 158 159 the beam-column joint subassemblies were intended to accurately 160 reproduce the actual configuration within a frame subjected to 161 reversed cyclic lateral loading. Beam and column elements were 162 extended between points of contraflexure 共assumed to be at mid163 span in the beams and at midheight in the columns兲 where pin 164 connections were introduced. Simple supports at the beam ends 165 were obtained connecting pin-end steel members to the floor. The loading protocol consisted of a series of three cycles at 166 167 increasing levels of interstorey drift applied to the top of the 168 column through a horizontal hydraulic actuator. In order to more 169 closely reproduce the actual stress level in the joint during the 170 lateral cyclic sway of a frame building, the column axial load was 171 varied during the experimental tests as a function of the lateral 172 load, by means of a vertical hydraulic jack, acting on a steel plate 173 connected to the column base plate by vertical external post174 tensioned bars. The axial-load versus lateral-force relationships for exterior 175 176 and interior joints, which are functions of the geometric charac177 teristics of the frame 共i.e., bay number and length, number of 178 storeys兲, were evaluated with preliminary pushover analyses on 179 the three-storey–three-bay RC frame system. Significant varia180 tions of the axial load up to 40–50% with respect to the value due 181 to gravity load only were expected. During the tests on the beam182 column specimens, a simplified bilinear relationship between 183 axial and lateral load was adopted as shown in Fig. 4共b兲. It is 184 worth underlining that the adoption of a variable axial load rep185 resents a fundamental, although unusual, improvement in the 186 loading protocol typically adopted for quasistatic tests on existing 187 beam-column subassemblies available in literature, where the 188 axial load is most likely maintained constant. The importance of a 189 proper estimation of the variation of the axial load, particularly 190 when dealing with assessment and retrofit strategies of poorly 191 detailed RC frame subassemblies of system, will be more evident 192 after the considerations given in the following sections, when 193 discussing the delicate process of evaluating the hierarchy of 194 strength and sequence of events.

The frame system was subjected to quasistatic cyclic loading at increasing levels of floor displacements, applied to the structure using three electromechanical actuators connected to the closest beam through a steel extension arm. The presence of gravity loads, which in existing underdesigned or gravity-load dominated frames represents a significant portion of the overall capacity, was simulated using concrete blocks as shown in Fig. 5. The lateral loading history consisted of a series of three cycles at increasing level of top drift 共±0.2%; ±0.6%; ±1.2%兲 with one conclusive cycle at ±1.6%. The application of simulated seismic loads was based on a hybrid force displacement control: the top floor displacement was directly controlled while maintaining a code-type lateral force distribution, proportional to the mass and to the floor level height 共more details in Calvi et al. 2002b兲.

195

Behavior of As-Built Beam-Column Joint Subassemblies

209 210

As reported in Pampanin et al. 共2002兲, the exterior tee-joint specimens showed a particularly brittle failure mechanism given by the combination of joint shear damage with the effects of slippage of the plain round beam longitudinal bars within the joint region, which led to a concentrated compressive force at the end-hook anchorage. As a result, a concrete “wedge” tended to spall off 共Fig. 6兲, leading to a brittle behavior with marked pinching in the hysteresis loop and loss of bearing-load capacity 共Fig. 7, dark dashed line, top and center兲. Conversely, the interior joint specimens showed significant resources of plastic deformation 共Fig. 7, dark dashed line, bottom兲, even without specific ductile structural details. A marked pinching was still observed, due to slip of the column longitudinal reinforcement bars. According to preliminary capacity design considerations, shear joint cracking and column hinging were predicted to be relatively close events. The concentration of flexural damage in the column at early stages thus acted as a structural fuse for the joint panel zone, which did not suffer significant cracking and damage. However, it should be recalled that the global frame system response could be seriously impaired if column hinging led to a soft-storey mechanism.

211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231

JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007 / 3

196 197 198 199 200 201 202 203 204 205 206 207 208

Fig. 2. Test frame: geometric and mechanical characteristic

232 Global Behavior of the Frame System in the As-Built 233 Configuration 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251

The results of the quasistatic tests on the three-storey–three-bay frame system 共briefly summarized in Calvi et al. 2002b兲 confirmed the high vulnerability of the panel zone region as observed at a subassembly level 共particularly in exterior joints兲 and the tendency to develop undesirable global mechanisms, due to the absence of an adequate hierarchy of strength. As shown in Fig. 8, most of the damage concentrated in the joint region 共exterior tee-joints兲 or at the beam-column interfaces through the development of a single wide flexural crack as expected from the slip of plain round reinforcing bars. In the interior joints no cracks were observed. The exterior tee-joints were subjected to a damage mechanism, analogous to that observed during the tests on beam-column subassemblies. The aforementioned tendency to develop a concrete wedge mechanism due to combined effects of an inefficient strut mechanism in the joint region after first shear cracking in the joint and the stress concentration at the beam bar end hooks, was observed, which could lead to severe damage and consequent loss of load-bearing capacity. The

test was, however, interrupted at relatively early stages 共1.6% drift兲 for safety issues, after the clear indication of a softening behavior as shown by the hysteresis behavior in Fig. 9 共dark dashed line兲. At a global level, an interesting peculiar mechanism was observed, when compared to a weak-column strong-beam mechanism 共which would lead to a soft storey mechanism兲, typically expected in an existing building. Based on the experimental evidence and numerical investigations, the concept of a shear hinge mechanism has been proposed as an alternative to flexural plastic hinging in the beams 共Pampanin et al. 2002, 2003兲. The concentration of shear deformation in the joint region, through the activation of a so-called shear hinge, could in fact result beneficial, by spreading the interstorey drift demand among two consecutive storeys, thus reducing the deformation demand onto the adjacent structural members 共columns in particular兲 and postponing the occurrence of undesirable soft-storey mechanism. As noted by Calvi et al. 共2002a兲, the drawback of this apparently favorable effect on the global response is, however, the increase in shear deformations in the joint region which can possibly lead 共depend-

4 / JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007

252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271

Fig. 3. Test setup of exterior and interior beam-column joints 272 ing on the joint typology and structural details adopted兲 to 273 strength degradation and loss of vertical load-bearing capacity. 274 The post-cracking behavior of the joint depends, in fact, solely on 275 the efficiency of the compression strut mechanism to transfer the 276 shear within the joint. Thus, while rapid joint strength degradation 277 after joint diagonal cracking is expected in exterior joints, a hard278 ening behavior after first diagonal cracking can be provided by an 279 interior joint. Damage limit states based on joint shear deforma280 tions have recently been defined and reported in Pampanin et al. 281 共2003兲 as a support to seismic assessment and retrofit strategy of 282 pre-1970s RC frame systems. It is evident how, based on a de283 tailed assessment of the local damage and corresponding global 284 mechanisms, a more reliable seismic rehabilitation strategy can be 285 defined.

286

Multilevel Retrofit Strategy

287 288 289 290

Regardless of the technical solution adopted, the efficiency of a retrofit strategy strongly depends on a proper assessment of the internal hierarchy of strength of the beam-column joints as well as of the expected sequence of events within a beam-column sys-

tem 共shear hinges in the joints or plastic hinges in beam and column elements兲. The effects of the expected damage mechanisms on the local and the global response should also be adequately considered.

291

Performance-Based Retrofit Strategy

295

An ideal retrofit strategy would not only protect the joint panel zone region, by identifying 共critically兲 weak point in older frames, but would further upgrade the structure to exhibit the desired weak-beam strong-column behavior which is at the basis of the design of new seismic resistant RC frames. However, due to the disproportionate flexural capacity, in gravity-load-dominated frames, of the beams when compared to the columns, a complete inversion of hierarchy of strengths is difficult to achieve in all cases and for all beam-to-column connections without major interventions. This is more evident for interior beam-to-column connections where the moment imposed on interior columns from the two framing beams is significantly larger than for exterior columns. As indicated in the previous paragraph, interior joints are less vulnerable than exterior joints and exhibit a much more stable hysteretic behavior with hardening after first cracking.

296 297 298 299 300 301 302 303 304 305 306 307 308 309 310

JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007 / 5

292 293 294

Fig. 4. Loading protocol for beam-column joint subassemblies: 共a兲 top drift control; 共b兲 axial load versus lateral load relationship 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328

Fig. 5. Frame test setup and elevation view

It is thus conceivable, in a bid to protect the interior columns from hinging, to tolerate some joint damage. According to a multilevel retrofit strategy approach suggested by Pampanin and Christopoulos 共2003兲, two levels of retrofits can therefore be considered, depending on whether or not the interior joints can be fully upgraded. A complete retrofit would consist of a full upgrade by protecting all joint panel zones and developing plastic hinges in beams while columns are protected according to capacity design principles. A partial retrofit would consist of protecting exterior joints, forming plastic hinges in beams framing into exterior columns, while permitting hinging in interior columns or limited damage to interior joints, where a full reversal of the strength hierarchy is not possible. The viability of the partial retrofit strategy must be investigated on a case-by-case basis to assure that the localized damage to interior joints does not severely degrade the overall response of the structure or jeopardize the ability of the interior columns to safely carry gravity loads.

329 Assessment of Sequence of Events: 330 Performance Domains 331 332 333 334 335

A simple procedure to compare the internal hierarchy of strengths within a beam–column-joint system is herein presented. The evaluation of the expected sequence of events is then proposed to be carried out through comparison of capacity and demand curves within a M-N 共moment-axial load兲 performance domain.

Fig. 6. Evaluation of hierarchy of strengths and sequence of events: M-N performance domain 共exterior tee-joint T1 in as-built configuration兲

6 / JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007

Fig. 7. Behavior of interior joint specimen C3: detachment of the carbon fibers from the nodal region at 4% drift level

336 337 338 339 340 341 342 343 344 345 346 347

Fig. 6 shows, as an example, the M-N performance domain adopted to predict the sequence of events and level of damage in the joint panel zone expected for the exterior specimen T1. The capacities of beam, column, and joints are referred to a given limit state 共e.g., for joints: cracking, equivalent “yielding” or extensive damage, and collapse兲 and evaluated in terms of the equivalent moment occurring in the column at that stage, based on equilibrium considerations within the beam–columnjoint specimen. While the evaluation of M-N curves for beams and columns is a relatively simple task, the definition of an “equivalent” curve to represent the joint panel zone can rely on the procedure described below.

Fig. 8. Observed damage in as-built frame at 1.6% top drift

Fig. 9. Comparison of global hysteresis behavior 共base shear top drift兲 beween as-built and retrofitted frame

The capacity or damage level of a joint is typically expressed in terms of nominal shear stress 共␯ jn兲 or principal compressiontensile stresses 共pc , pt兲. Although current codes 共e.g., ACI 318, AIJ, EC8, NZS3101兲 tend to adopt simplified provisions which limit the nominal shear stress ␯ jn expressed as a function of the concrete tensile strength, k1冑 f ⬘c , or the concrete compressive strength, k2 f ⬘c , where k1 and k2 are empirical constants, it is commonly recognized that principal stresses, by taking into account the contribution of the actual axial compression stress acting in the column, can provide more accurate indications on the stress state and thus damage level in the joint region. Typical strength degradation models, available in the literature and based on research on poorly designed joints 共e.g.,Priestley 1997, Pampanin et al. 2002; shown in Fig. 10兲 can be adopted to define limit states in a joint panel zone subjected to shear and axial load. According to the simplified analytical model proposed by Pampanin et al. 共2003兲 to describe the joint nonlinear behavior, based on a rotational spring within a concentrated plasticity ap-

Fig. 10. Strength degradation curves for exterior joints in terms of principal tensile stress versus shear deformation

JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007 / 7

348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366

Table 2. Sequence of Events for Exterior Specimen T1 共As-Built Configuration兲 Specimen T1 共as-built兲 Type of lateral force

Open joint F⬍0

Close joint F⬎0

Lateral force 共kN兲

No.

Event

1 2

Joint cracking and deterioration starting pt = 0.19冑 f e− Beam yielding

−16.59

3 4

Upper column yielding Lower column yielding

−20.50 −22.75

5 6

Joint failure Lower column yielding

9.37 13.50

7 8

Upper column yielding Beam yielding

14.50 16.59

−10.94

367 proach, the equivalent moment-rotation curve of the joint region 368 共i.e., monotonic characteristics of the spring model兲 can be de369 rived from the corresponding principal tensile stress-shear defor370 mation curve using equilibrium considerations: for any given 371 level of principal tensile 共or compression兲 stress in the joint, the 372 corresponding “joint moment” M j, which is either the sum of 373 the beam moments or the sum of the column moments at that 374 stage, can be evaluated. So doing, M-N capacity curves cor375 responding to the different joint limit states can be plotted within 376 a performance domain where “equivalent column” capacity are 377 represented. As previously shown in Fig. 6 共as-built exterior specimen T1兲, 378 379 demand curves 共V-shaped兲 should account for the variation of 380 axial load due to the effects of lateral forces in a frame system 381 共for either opening and closing of the joint兲. Incorrect and non382 conservative assessment of the sequence of events can otherwise

Fig. 12. Joint strength degradation curve: contributions of FRP and concrete 共exterior speciment T1兲

result, leading to inadequate design of the retrofit intervention. It is worth noting that, for simplicity, the sequence of events corresponding to negative and positive sign of the lateral force 共opening or closing of the joint兲, should better be independently evaluated 共i.e., the numbering 1–8 actually indicates a sequence 1–4 in the negative direction and a sequence 1⬘ – 4⬘ in the positive direction兲. In the case of specimen T1, in the as-built configuration, a pure shear hinge mechanism, with extensive damage of the joint, was thus expected 共using a proper demand curve兲 prior to any hinging of beams or columns 共Table 2兲, as confirmed by the experimental tests. However, the order and “distance” of the events strongly depends on the demand curve assumed. If a constant axial load curve was used 共as shown in Fig. 6 for N = −100 kN兲, only a minor increase in the column strength 共in addition to the joint strengthening兲 would have appeared necessary, leading to a

Fig. 11. Effects of FRP on the moment-curvature curve of a member 共beam兲 critical section 8 / JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007

383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398

Table 3. Properties of High Modulus Carbon Fiber with Unidirectional Fabric 共MBrace CFRP C5-30兲

Type of fiber High modulus carbon 共C5-30兲

Tow sheet type

Density 共kg/ m3兲

Effective thickness 共one layer兲 共mm兲

Unidirect. fabric

1820

0.165

399 column hinging occurring before the formation of a beam hinge 400 共i.e., high risk of a soft storey mechanism even after the retrofit 401 intervention兲. The concept of a performance domain could thus be extended 402 403 from the purpose of assessing as-built systems and adopted to 404 evaluate and control the feasibility and efficiency of any retrofit 405 strategy on beam-column joints, provided that the effects of the 406 retrofit solution on the single elements 共beams, column, or joint 407 panel zone兲 can be simply and independently evaluated as de408 scribed in the following paragraphs.

409 410

Evaluation of FRP Strengthening Effects: Analytical Model

411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426

The effects of a retrofit intervention with FRP composite material in the form of externally bonded reinforcement on a beam-column joint, in terms of flexural or shear capacity in beams, columns, and panel zone region, is carried out through a step-by-step procedure. The occurrence of defined limit states 共cracking, yielding, debonding, crushing, and spalling of concrete, failure within the adopted materials兲 corresponding to a given stress or strain value can thus be properly evaluated and controlled when designing the retrofit intervention. As mentioned and shown, an accurate prediction of the expected sequence of events can thus be obtained through M-N performance domains. Analytical procedures available in the literature are adopted and properly modified to account for debonding phenomena as well as, more importantly, for the effects of the variation of axial load onto the joint panel zone behavior 共critical issue typically neglected兲.

Tensile strength 共MPa兲

E-modulus 共MPa兲

Utimate strain 共%兲

3,000

39,000

0.8

Flexural FRP Retrofit of Beams and Columns

427

The enhanced flexural behavior of a FRP retrofitted beam or column critical section was evaluated though a fiber section analysis. The Bernoulli-Navier hypothesis on plane sections remaining plane was assumed, considering fully composite action 共bond兲 between the external FRP laminates and the concrete. Debonding was taken into account according to the model proposed by Holzenkämpfer 共1994兲 共and adopted by the FIB guidelines of FRP retrofit, FIB 2001兲, and thus expected to occur at a strain limit level ␧deb = c1 · 冑 f ctm / E f t f , where E f is the FRP E-modulus, f ctm the mean value of concrete tensile strength, s the thickness of the FRP laminate, and c1 an empirical coefficient taken as 0.64 for CFRP as suggested by Neubauer and Rostásy 共1997兲. The material behavior was defined through proper stress-strain relationships, as follows: Mander et al. 共1988兲 model for concrete; Dodd-Restrepo model 共1995兲 for steel and a linear-elastic rule for the FRP composite material, consistent with the properties supplied by the provider. The moment-curvature behavior of the critical section in the presence of externally bonded FRP laminate can thus be evaluated for different levels of axial load 共Fig. 11兲 using an iterative procedure as typically done for RC sections. The position of the neutral axis is estimated until both compatibility and equilibrium conditions are satisfied. M-N capacity curves for beams and columns corresponding to a given limit state can be derived and plotted in a performance domain to define the sequence of events. The confinement effects of the FRP on the section curvature ductility capacity can be taken into account following procedures available in the literature 共e.g., Spoelstra and Monti 1999兲.

428 429 430 431 432 433 434 435

Fig. 13. FRP-retrofit solution for the exterior joint specimen T1B JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007 / 9

436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456

Fig. 14. FRP-retrofit solution for the interior joint specimen C3

457 Increase of Joint Shear Capacity due to the FRP 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497

The evaluation through analytical models of the strengthening effects on the panel zone 共joint兲 shear is a more complex task with limited research available in the literature. An overview of alternative procedures has been given by Antonopoulos and Triantafillou 共2002兲. Typical 共oversimplified兲 approaches consider the contribution of the FRP equivalent to external “stirrups” 共analogy with steel transverse reinforcement兲. Upper limits of the maximum strain in the FRP material are used in the calculations, either corresponding to the declared ultimate tensile capacity 共Gergely et al. 1998兲 or to constant strain values depending on the preparation of the concrete surface 共Tsonos and Stylianidis 1999; Gergely et al. 2000兲. A more rigorous model, based on stress equilibrium and strain compatibility equations of the panel zone region 共idealized as a three-dimensional element兲 has been presented by Antonopoulos and Triantafillou 共2002兲 as an extension of the model for RC joint behavior without FRP proposed by Pantazopoulou and Bonacci 共1994兲. Satisfactory validation of the analytical model was obtained on the experimental results on a total of 15 beamcolumn exterior beam-column subassemblies, tested by the authors 共Antonopoulos and Triantafillou 2003兲 or available in the literature 共Gergely et al. 2000兲. It is, however, important to underline that, as typically done in most experimental tests on beam-column joints, no variation of axial load as a function of the lateral force during the lateral sway of a frame system was considered during the tests. The implications of assuming a constant load in the assessment of the sequence of events prior to or after a retrofit intervention has been briefly discussed in the previous paragraphs. In the present contribution the original step-by-step iterative procedure proposed by Antonopoulos and Triantafillou 共2002兲, in its simplified version 共where the direct shear strength of the composite sheet is neglected兲, is adopted as a general platform and adequately extended after a few simple modifications to account for the variation of the axial load on the joint region. Consistently with the analytical procedure proposed to visualize the joint shear contribution within a M-N performance domain starting from principal tensile or compression stresses considerations, the basic equations of equilibrium and strain compatibility of the joint panel zone are rearranged to evaluate an

equivalent strength degradation curve 共principal tensile stress versus joint shear deformation兲 corresponding to the FRP contribution only. The overall strength degradation curve for the FRP retrofitted joint would thus be given by the combination of the FRP and concrete contributions, as shown in Fig. 12. Such a curve forms the basis for the evaluation of the equivalent joint moment M j, within a performance domain M-N. It is worth noting that in terms of analytical-numerical modeling according to a plasticity-concentrated approach, two rotational springs 关with moment-rotation curves derived, as mentioned, according to the method proposed by Pampanin et al. 共2003兲兴 can be adopted to represent the two independent contributions. It is in fact expected 共later confirmed by the experimental tests兲 that the cracking and damage of the joint can still occur underneath the protection given by the FRP laminates, whose major effect is to increase the overall joint strength, avoiding local failure mechanism 共such as the “concrete wedge” mecha-

Fig. 15. Evaluation of hierarchy of strengths and sequence of events: M-N performance domain 共exterior tee-joint T1B joint after retrofit兲

10 / JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007

498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516

Fig. 16. Preparation of FRP retrofit intervention on the test frame 517 nism兲 and achieving an enhanced global behavior by developing a 518 more desirable sequence of events 共e.g., weak-beam strong519 column mechanism if a total retrofit strategy is followed兲. Details on the analytical procedure to evaluate the joint shear 520 521 strength contribution of FRP as well as on simplified design 522 methods can be found in Vecchietti 共2001兲 and Nassi 共2002兲 523 and will be reported in future publications currently under 524 preparation.

525

Design of the Retrofit Intervention

526 527 528 529

According to the proposed multilevel retrofit strategy, a full retrofit was adopted for the exterior joint, i.e., protection of the joint and plastic hinge in the beam, while a partial retrofit was adopted for the interior joint specimen, i.e., partial protection of the col-

umn hinging while some damage in the joint region can be accepted. Issues related to the expulsion of the concrete wedge in the exterior joints as well as to the premature debonding of the fibers were carefully considered as explained in the following sections.

530

Retrofit Solutions

535

A few alternative FRP retrofit solutions 共relying on different forms or properties of the composite material兲 have been recently proposed in literature for beam-column joints subjected to lateral cyclic loading 共e.g., Pantelides et al. 2000兲. Extensive experimental investigations on an exterior beam-column joint retrofitted with FRP 共in the form of laminates or strips兲 have been carried out by Antonopoulos and Triantafillou 共2003兲. Due to the scope of that investigation 共evaluation of the FRP contribution to the joint

536 537 538 539 540 541 542 543

Fig. 17. Comparison of damage mechanisms in beam-column joint subassemblies before and after retrofit; 共a兲 shear hinge 共as-built T1兲; 共b兲 relocated beam plastic hinge 共retrofitted T1B兲; 共c兲 joint panel zone conditions underneath the composite sheet after testing JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007 / 11

531 532 533 534

Fig. 18. Comparison of hysteresis response of as-built and FRP-retrofitted beam-column subassemblies: exterior joints T1 and T1B, T2 and T2B and interior joints C2 and C3

544 shear strength兲, the design of the retrofit strategy aimed at guar545 anteeing that the damage occurred in the joint region. A selective 546 seismic strengthening technique for gravity-load-designed frames, 547 relying on both FRP laminates and near mounted surface, has 548 been recently proposed by Prota et al. 共2002兲. In this contribution, unidirectional carbon fiber laminates 549 550 共high-modulus CFRP, Table 3兲 were adopted for both exterior and

interior joints in the configurations illustrated in Figs. 13 and 14. It is worth noting that after considering alternative FRP materials 共i.e., glass, aramid, or carbon fibers with lower modulus and strength兲, the choice of high modulus CFRP, with a relatively low ultimate strain capacity, was primarily dictated by the significant difference between column and beam moment capacity typical of the older practice in Italy as well as in other Mediterranean seismic-prone countries 共where minimum values of longitudinal reinforcement ratio as low as 0.8% were allowed兲. Vertical FRP laminates were used on the external side 共shear兲 face of the column in both interior and exterior joints 共two layers per side兲 rather than on the flexural 共tension and compression兲 side faces, in order to increase the column flexural capacity as well as the joint shear strength. In addition, in the exterior joint specimen, a U-shape horizontal laminate, wrapped around the exterior face of the specimen at the joint level, was used to increase the joint shear strength as well as prevent the expulsion of a concrete wedge. An adequately limited anchorage length within the beam was calculated in order to 共a兲 guarantee sufficient shear strengthening in the joint without excessively increasing the beam capacity 共as per Fig. 15兲 and 共b兲 relocate the plastic hinge region at a controlled distance from the beam-column critical interface. Although the evaluation of strengthening effects was carried out including debonding effects 共when nonconservative兲, additional smaller strips were used to wrap the main FRP laminates and provide proper anchorage. In the case of the interior joint, the FRP laminate crossing the joint was intentionally left unprotected from debonding in the joint panel zone region. According to a partial retrofit approach, the retrofit solution for the interior joint C3 共Fig. 14兲, intended to allow some debonding of the vertical FRP sheets to occur along the joint panel zone, in order to facilitate the development of a combined damage mechanism with limited cracking in the joint and subsequent flexural hinging of the adjacent beams. The target performance of the retrofit solution was controlled using the proposed procedure based on the M-N performancedomain as shown in Fig. 15 and Table 3 for the exterior specimen T1B. Prior to testing the beam-column specimens, a partial retrofit strategy was implemented on the frame system 共Fig. 16兲, with the final intent to favor a more desirable inelastic global mechanism, able to protect brittle failure mechanisms due to the excessive damage and collapse of an exterior joint or the development of a soft storey. This could be achieved by forming plastic hinges in the exterior beams while accepting minor damage in the interior joint prior to the development of a flexural behavior in the adjacent structural elements 共in this case, the interior beams兲. A similar approach, in principle, and detailed layout of the FRP retrofit was adopted as per the beam-column specimens. Appropriate M-N interaction curves, accounting for the effective geometry and demand curve for each beam-column joint within the frame systems, were used to verify the efficiency of the final solution, which, for simplicity of execution, was implemented for both the first and second floor joints, and followed the solution adopted for the T1B and C3 specimens. No intervention seemed to be required at the third-floor level where no or negligible damage was expected in the panel zone region with flexural cracks developing in the column top sections.

12 / JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007

551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608

Table 4. Sequence of Events for Exterior Specimen T1B 共Retrofitted Configuration兲 Specimen T1B 共strengthened兲 Type of lateral force Open joint F ⬍ 0

Close joint F ⬎ 0

609 610

Lateral force 关kN兴

N°

Event

1 2 3 4

Beam yielding Upper column yielding Joint cracking 共no strength degradation兲 Lower column yielding

−25.32

5 6 7 8

Lower column yielding Upper column yielding Beam yielding Joint failure

15.75 16.98 18.91 19.67

−18.91 −23.11 −24.15

Improvement of Structural Performance after Retrofit

611 Experimental Results on Retrofitted Beam-Column 612 Joint Subassemblies 613 614 615 616 617

The results of the experimental quasistatic tests on three beam column joints in the retrofitted configurations 共namely T1B, T2B, and C3兲 provided very satisfactory confirmations of the efficiency of the adopted retrofit solution as well as of the reliability of the analytical procedure developed to design the intervention

and assess the expected sequence of events and performance. A summary of the results is given herein, while more details are available in Nassi 共2002兲 and will be reported in future publications. In all cases, the retrofit objective based on a multilevel retrofit strategy was achieved, leading to a significant improvement in the behavior of the subassemblies, which ultimately imply an enhanced behavior of the frame system 共adequate global inelastic mechanism兲. As shown in Fig. 17, a properly designed FRP-retrofit solution for exterior beam-column joints can protect and avoid the formation of a brittle shear hinge mechanism and re-establish a more desirable hierarchy of internal strengths and sequence of events, enforcing a beam plastic hinge mechanism, relocated at a controlled distance from the beam-column interface 共total retrofit兲. As a result, an improved and more stable hysteresis behavior was observed with increased ductility and energy dissipation capacity 共Fig. 18兲. The values of lateral force corresponding to the occurrence of the critical events were well-predicted by the analytical methods 共presented in Fig. 15 and Table 4兲. Similar considerations can be derived for the enhanced response of the interior joint specimen C3, where the partial retrofit strategy led to a controlled debonding of the column vertical fibers crossing the joint 共Fig. 7兲. The formation of flexural damage in the column was thus postponed. In addition to the increased overall strength 共as shown by the hysteresis loop in Fig. 18兲, the FRP provided a favorable confinement effect in the column

Fig. 19. Damage observations in the retrofitted test-frame 共at 2% top drift兲 JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007 / 13

618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645

Fig. 20. Shear hinge damage mechanism in as-built exterior joint specimen T1: formation of a concrete wedge mechanism

646 plastic hinge region avoiding the premature crushing and spalling 647 of concrete cover, protecting from strength degradation, buckling 648 of the longitudinal bars, and consequent failure. 649 Experimental Results on Retrofitted Frame System 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679

The global behavior of the retrofitted frame systems followed the expectations and analytical predictions. As shown in Fig. 19, the frame response was characterized by the formation of plastic hinges in the exterior beams, with no damage in the beam-column joints, protected by the FRP. In the interior joints, flexural vertical cracks developed at the beam-column interface at an earlier stage 共0.5–0.6 % interstorey drift兲 and further extended within the joint panel zone, confirming that cracking was occurring in the joint. Due to the lower level of imposed drift, when compared to the tests on subassemblies, the debonding of the FRP sheets along the joint region did not develop up to a complete peeling-off phenomenon. As a consequence of the inverted hierarchy of strength, at least in the exterior beams 共partial retrofit兲, a more desirable inelastic mechanism occurred, leading to higher strength and dissipation capacity, as evident from the more stable global hysteresis loop shown in Fig. 9, up to higher level of drift 共2%兲 when compared to the as-built solution, before observing a softening behavior 共onset of strength reduction兲 mainly due to P-D effects. It is worth noting that the unloading global behavior of the retrofitted frame shows a loss of stiffness with some pinching phenomenon similar to, although less evident than, that observed in the as-built system. As anticipated for the beam-column subassemblies, this effect can be due to the shear cracking developing in both the exterior and the interior beam-column joint belonging to the frame system, underneath the layers of FRP. In line with the proposed analytical model, FRP and concrete contribute in parallel to the overall strength degradation curve of the joints 共Fig. 12兲. An appropriate retrofit strategy would thus protect the joint from excessive deformation 共concentrated in the beam

Fig. 21. Deformed shape of the retrofit frame during testing

plastic hinge兲, while, due to the alteration of the hierarchy of strength, higher nominal shear 共or principle tensile兲 stresses might develop in the joint, part of which still has to be taken by the concrete component. Furthermore, the reinforcement details of the exterior joints in the frame systems are in general more similar to those of the T1 specimen 共see Figs. 1 and 2兲 which showed 共consistently with the predicted performance-based M-N domain兲 a more remarkable pinching behavior than the T2 specimen either before or after the retrofit intervention 共Fig. 18兲. In the case of the T2 specimen, in fact, lower beam reinforcement was adopted, leading to an earlier formation of a plastic hinge in the adjacent beam, with less rotational demand, thus damage, in the panel zone region 共see Figs. 20 and 21.

680

Concluding Remarks

693

The experimental results of quasistatic tests on beam-column joint specimens and three-storey frame systems, designed for gravity load only and retrofitted with CFRP laminates, provided very satisfactory confirmation of the efficiency of similar solutions for existing buildings.

694 695 696 697 698

14 / JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007

681 682 683 684 685 686 687 688 689 690 691 692

699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733

A multilevel retrofit strategy has been proposed depending on the subassembly type and structural details to achieve the desired performance with a feasible intervention. Alternative FRP composite materials in terms of mechanical properties or type 共sheets, strips, or near mounted surface rods兲 should be appropriately selected depending on the target performance as well as on the original “distance” between events according to hierarchy of strength considerations. A simplified analytical procedure to evaluate and control the sequence of events using a M-N performance domain has been presented, after receiving promising confirmations from the satisfactory experimental results. In the exterior joints, the occurrence of a brittle joint shear mechanism was adequately protected and a more desirable hierarchy of strengths and sequence of events achieved, leading to a more ductile and dissipating hysteresis behavior. In the interior joints, a controlled minor cracking in the joint panel zone was accepted, in order to protect a column sway mechanism. At a global level, the implementation of a partial retrofit strategy on a three-storey three-bay frame system favored the development of a more appropriate global inelastic mechanism, preventing brittle failure in exterior joints or undesired events such as a soft storey mechanism. Ultimately, as discussed in the Introduction, issues of accessibility of the joint region and invasiveness will have to be faced in real applications. However, it is worth noting that a typical geometrical and plan configuration of existing buildings designed for gravity load only in the 1950s–1970s period consist of frames running in one direction only and lightly reinforced slab in the orthogonal direction, the latter being quite typical of the construction practice in Mediterranean countries. In these cases, the adoption of the proposed retrofit intervention can be somehow facilitated, when compared with more recently designed buildings with frames in both directions and cast-in-situ concrete slabs providing flange effects.

734

Acknowledgments

735 736 737 738 739 740 741 742 743 744

The financial support provided by the Italian Ministry of the University and the University of Pavia, under a coordinated national project 共PRIN 2001兲, as well as by the European Community 共Contract No. SPEAR G6RD-CT-2001-00525兲 is gratefully acknowledged. The writers wish to thank the MAC S.p.a. Treviso for providing the materials and technical assistance for the retrofit intervention. The assistance and cooperation, during different phases of the project, of postgraduate students Mr. A. Vecchietti and Mr. R. Nassi are also gratefully acknowledged.

745

References

746 747 748 749 750 751 752 753 754 755 756

Antonopoulos, C. P., and Triantafillou, T. C. 共2002兲. “Analysis of FRP-strengthened RC beam-column joints.” J. Compos. Constr. 6共1兲, 41–51. Antonopoulos, C. P., and Triantafillou, T. C. 共2003兲. “Experimental investigation on FRP-strengthened RC beam-column joints.” J. Compos. Constr. 7共1兲, 39–43. Aycardi, L. E., Mander, J. B., and Reinhorn, A. M. 共1994兲. “Seismic resistance of R. C. frame structures designed only for gravity loads: Experimental performance of subassemblages.” ACI Struct. J. 91共5兲, 552–563. Beres, A., Pessiki, S., White, R., and Gergely, P. 共1996兲. “Implications of

experimental on the seismic behaviour of gravity load designed RC beam-column connections.” Earthquake Spectra 12共2兲, 185–198. Bing, L., Yiming, W., and Tso-Chien, P. 共2002兲. “Seismic behavior of non-seismically detailed interior beam-wide column joints. Part I: Experimental results and observed behavior.” ACI Struct. J. 99共6兲, 䊏䊏䊏–䊏䊏䊏. Calvi, G. M., Magenes, G., and Pampanin, S. 共2002a兲. “Relevance of beam-column damage and collapse in RC frame assessment.” J. Earthquake Eng. 䊏共1兲, 75–100 Calvi, G. M., Magenes, G., and Pampanin, S. 共2002b兲. “Experimental test on a three storey R. C. frame designed for gravity only.” 12th ECEE, London, Paper No. 727. Dodd, L. L., and Restrepo, J. I. 共1995兲. “Model for predicting cyclic behavior of reinforcing steel.” J. Struct. Eng. 121共3兲, 433–445. Dolce, M., Cardone, D., and Marnetto, R. 共2000兲. “Implementation and testing of passive control devices based on shape memory alloys.” Earthquake Eng. Struct. Dyn. 29共7兲, 945–968. FIB 共Federation International du Beton兲 共2001兲. “Externally bonded FRP reinforcement for RC structures.” FIB Bulletin No. 14, Lausanne. FIB 共Federation International du Beton兲 共2006兲. “Retrofitting of concrete structures by externally bonded FRPs with emphasis on seismic applications.” FIB Bulletin No. 35, Lausanne. Gergely, J., Pantelides, C. P., Nuismer, R. J., and Reaveley, L. D. 共1988兲. “Bridge pier retrofit using fiber reinforced plastic composites.” J. Compos. Constr. 2共4兲, 165–174. Gergely, J., Pantelides, C. P., and Reaveley, L. D. 共2000兲. “Shear strengthening of RC T-joints using CFRP composites.” J. Compos. Constr. 4共2兲, 56–64. Hakuto, S., Park, R., and Tanaka, H. 共2000兲. “Seismic load tests on interior and exterior beam-column joints with substandard reinforcing details.” ACI Struct. J. 97共1兲, 11–25. Hertanto, 䊏 共2006兲. “Seismic assessment of pre-1970s reinforced concrete beam column joint subassemblies” M.E. dissertation, Dept. of Civil Engineering, Univ. of Canterbury, Christchurch, New Zealand. Holzenkämpfer, P. 共1994兲. “Ingenieurmodelle des verbundes geklebter bewehrung für betonbauteile.” Ph.D dissertation, TU Braunshwieg 共in German兲. Mander, J. B., Priestley, M. J. N., and Park, R. 共1988兲. “Theoretical stress-strain model for confined concrete.” J. Struct. Eng. 114共8兲, 1804–1826. Nassi, R. 共2002兲. “Seismic retrofit strategy for underdesigned beamcolumn subassemblies using FRP.” Laurea thesis, Dept. of Structural Mechanics, Univ. of Pavia 共in Italian兲. Neubauer, 䊏, and Rostásy, 䊏 共1997兲. “Design aspects of concrete structures strengthened with externally bonded CFRP-plates, concrete⫹composites.” 7th Int. Conf. Structural Faults and Repair, Vol. 2, 109–118. Pampanin, S., Calvi, G. M., and Moratti, M. 共2002兲. “Seismic behaviour of R. C. beam-column joints designed for gravity loads.” 12th ECEE, London, Paper No. 726. Pampanin, S., and Christopoulos, C. 共2003兲. “Non-invasive retrofit of existing RC frames designed for gravity loads only.” FIB2003 Symp. Concrete Structures in Seismic Regions, Federation International du Beton, Athens. Pampanin, S., Magenes, G., and Carr, A. 共2003兲. “Modelling of shear hinge mechanism in poorly detailed RC beam-column joints.” Proc. FIB Symp. Concrete Structures in Seismic Regions, Federation International du Beton, Athens, paper No. 171. Pampanin, S., Christopoulos, C., and Chen, T-H. 共2006兲. “Development and validation of a haunch metallic seismic retrofit solution for existing under-designed RC frame buildings.” Earthquake Eng. Struct. Dyn. 35共14兲, 1739–1766. Pantazopoulou, S. J., and Bonacci, J. F. 共1994兲. “On earthquakeresistance reinforced concrete frame connections.” Can. J. Civ. Eng. 21, 307–328. Pantelides, C. P., Clyde, C., and Reaveley, L. D. 共2000兲. “Rehabilitation

JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007 / 15

757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823

824 825 826 827 828 829 830 831 832 833 834 835

of R/C building joints with FRP composites.” 12WCEE Auckland, Paper No. 2306. Park, R. 共2002兲. “A summary of results of simulated seismic load tests on reinforced concrete beam-column joints, beams and columns with substandard reinforcing details.” J. Earthquake Eng. 6共2兲, 1–27. Priestley, M. J. N. 共1997兲. “Displacement-based seismic assessment of reinforced concrete buildings.” J. Earthquake Eng. 1共1兲, 157–192. Prota, A., Manfredi, G., Nanni, A., and Cosenza, E. 共2002兲. “Capacity assessment of GLD RC frames strengthened with FR.” 12th European Conference on Earthquake Engineering, London, Paper No. 241. Regio Decreto. 共1939兲. 16 November-XVIII, n. 2228 共in Italian兲. Santarella, L. 共1957兲. Il cemento armato, Vol. 2, Ed. U. Hoepli, Milan.

Spoelstra, M. R., and Monti, G. 共1999兲. “FRP-confined concrete model.” J. Compos. Constr. 3共3兲, 143–150. Sugano, S. 共1996兲 “State of the art in techniques for rehabilitation of buildings.” Proc., 11th World Conf. on Earthquake Engineering, Acapulco, Mexico, Paper No. 2175. Tsonos, A. D., and Stylianidis, K. A. 共1999兲. “Pre-seismic and postseismic strengthening of reinforced concrete structural subassemblages using composite materials 共FRP兲.” Proc., 13th Hellenic Concrete Conference, Rethymno, Greece, 1, 455–466 共in Greek兲. Vecchietti, A. 共2001兲. “Seismic rehabilitation of concrete frame systems designed for gravity loads only using FRP composite materials.” Laurea thesis, Dept. of Structural Mechanics, Univ. of Pavia 共in Italian兲.

16 / JOURNAL OF COMPOSITES FOR CONSTRUCTION © ASCE / MARCH/APRIL 2007

836 837 838 839 840 841 842 843 844 845 846 847