EXPERIMENTAL BEHAVIOUR OF STRAIGHT AND HOOKED SMOOTH BARS IN EXISTING R.C. BUILDINGS

Transcription

1 Published by Elsevier Science Ltd. All rights reserved 12 th European Conference on Earthquake Engineering Paper Reference 393 (quote when citing this paper) EXPERIMENTAL BEHAVIOUR OF STRAIGHT AND HOOKED SMOOTH BARS IN EXISTING R.C. BUILDINGS G. Fabbrocino, G. M. Verderame, G. Manfredi Department of Structural Analysis and Design, University of Naples Federico II, Naples 8125, ITALY ABSTRACT Evaluation of seismic vulnerability of existing reinforced concrete structures is a very important issue in structural engineering and needs continuous updates. The approach to the problem cannot neglect the specific performances of materials and detailing adopted in the past. The present paper deals with smooth reinforcing bars used up to 7, and therefore present in a very large number of existing constructions exposed to seismic risk. Behaviour of anchored smooth bars is examined referring to basic properties of bond and response of bar end details used in critical regions. The attention is focussed on typical circular hooks commonly used in existing buildings and on the stress-slip response under static and cyclic loads. This is an interesting topic, since assessment of existing r.c. frames and the related evaluation of the deformation capacity under lateral loads require refined models for each source of deformation. Experimental results on hooked and straight bars under pull-out and beam-test are discussed, pointing out interesting aspects of response under service and ultimate loads. Keywords: Seismic assessment, fixed-end rotation, smooth rebars, bar anchorages, bond. INTRODUCTION Protection of existing buildings against seismic actions is a very relevant problem in many countries, and particularly in Italy, where wide regions are exposed to seismic risk and sometimes are only recently classified as seismic [1], consequently in many cases design does not take into consideration any relevant horizontal load. Furthermore, a large number of existing buildings in Italy dates back to 5 s and 6 s, due to the fast growth of constructions after the last world war and is characterized by a reinforced concrete framed structure. At that time, smooth bars were commonly used for reinforcement; they are characterized by poor bond properties compared to current deformed bars and need appropriate anchoring end details to ensure a satisfactory interaction with concrete. Structural maintenance and seismic assessment of such buildings depends on the availability of reliable models for members, but also of critical regions, i.e. base columns and beam to column joints. In fact, influence of rotations of end sections in r.c. members can be relevant and depends on different and complex sources of deformation [2]. Experimental and

2 theoretical results on effects of smooth bars and anchoring devices on fixed-end rotations of members is basically poor, since extensive research on seismic behaviour of r.c. structures started in 7 s, and fundamental studies refer to deformed reinforcement. Review of past researches indicated that typical end anchorages for smooth bars were hooks with 9, 135 and 18 opening angles. Early experimental researches [3, 4] were aimed to evaluate the effectiveness of end details on smooth reinforcement pull-out. Contemporary research by Saliger [5] was very interesting due to large number and types of tests; these were aimed to evaluate strength of anchorages, without any consideration of performances in terms of deformation. Straight and hooked bars (18 opening angle) were tested; different bar diameters, curvature radius and transverse rebar arrangements were considered. Results on pull-out tests agreed with Bach s tests on beams [3] that increased flexural strength due to end anchorage and showed that shape of anchorage and transverse reinforcement can give beneficial effects. Later, research by Mylrea [6] took into consideration both strength and deformation of hooked smooth reinforcing bars with 18 opening angle and tried to outline the influence of end hook radius and transverse reinforcement. Results show a slight influence of hook radius on deformation and point out the role of transverse reinforcement on the type of failure. In particular, plain specimens showed a non-ductile behaviour due to concrete failure, compared to bar failure activated by appropriate transverse bar detailing. Approaching 5 s, early types of deformed bars were investigated and compared to force-slip response of straight and anchored smooth bars [7]. More than forty specimens were tested varying the hook radius, surface type (smooth or ribbed), development length and opening angle. The above remarks about available technical literature on response of hooked smooth rebars, however, cannot lead to the development of reliable modelling of force-slip behaviour for seismic assessment. In fact, more recent and comprehensive researches deal with deformed bars and former studies rarely took account of large post-yielding behaviour of smooth anchorages [8]. Thus, in the following the results of experimental tests on straight and hooked smooth bars are presented. The tests have been carried out considering different setup: beam test, for straight bars bond and service conditions of hooked ones; pull-out type tests to describe the behaviour of hooked rebars up to failure. Evaluation of experimental results points out interesting aspects of the mechanical response of smooth and hooked bars and confirm that hooks can play a fundamental role in the development of rotations at beam and column ends when smooth bars are concerned. FIXED-END ROTATIONS OF MEMBERS AND ANCHORED BAR MODELLING The beam to column region is characterized by complex mechanical interactions, since actions due to connected end sections lead to very localized stresses on concrete panel and steel rebars with very high gradients that trigger complex deformation mechanisms. In Figure 1, global joint deformation is divided into two main components: - the concrete panel mechanism that is related to cracking of concrete and arrangement of transverse reinforcement in the joint region; - the slippage of anchored reinforcement that is dependent upon bond properties and anchoring details; it leads to a rotation of member end section that is commonly addressed as fixed-end rotation.

3 The above sources of deformation can be generally neglected if structural analysis is aimed to evaluate the ultimate load capacity of r.c. frames with deformed bars, but have a key role when drift capacity is concerned, as confirmed by theoretical-experimental comparisons [9]. γ a) b) Figure 1: Mechanisms governing the joint region deformation: panel (a), bar slip (b). free end - ds/dx = ; s hooked end - ds/dx ; s F = σ π φ 2 4 rigid end - ds/dx ; s = rigid end σ u σ y σ u σ y free end Steel stress Steel stress s u χ u Loaded end slip Curvature s y χ y ε y ε sh Steel strain ε u ε y ε sh ε u Steel strain F hook / F Moment M y M u 1. Figure 2. Influence of bar anchorage on beam and column end section performances. If smooth bars are considered, the problem is more complex since end details of reinforcing bars are essential for the development of the strength and the deformation mechanisms, but the experimental background is definitively poor. The main aspects related to response of end sections of members are summarised in Figure 2 that reports in detail the force transfer governing the behaviour of the reinforcing bar under tension drawn in Figure 1.b. The anchored bar is divided into two components, the straight region and the anchorage, represented by a circular hook. From a theoretical point of view, the end anchorage results in a restraint for the inner end of

4 straight rebar. Therefore, two boundary conditions can be identified: - if anchorage is not present, straight rebar is characterised by a free end, thus a slippage occurs and the inner end of the bar does not bear any axial load; - if anchorage is rigid, slip at the inner end of the rebar is equal to zero, and a pull-out force, F hook, develops on the anchoring device. In the first case, pull-out strength is due to bond properties of smooth bars that allow very low bar deformations compared to yielding one; in the second case, high ratios between applied force and anchorage reaction can be reached, even 5-6% of the applied tensile force. Behaviour of anchorage strongly influences the static and deformation response of the member end sections. In fact, pull-out of smooth bars without anchors leads to the premature failure of the member end section, conversely rigid anchorage allows the full development of flexural strength, as shown in Figure 2.d, and produces the minimum value of slippage at the loaded end, reported in Figure 2.b. As a result, the key problem in a reliable modelling of r.c. frames for seismic assessment is the definition of the relationship between the axial force applied on the anchorage and the slippage of its loaded end in view of the development of a behavioural model of the anchored bar. The availability of a specific force-slip relation for anchorage allows to model the anchored bar on the analogy of a smooth bar embedded in a concrete matrix [1], taking the end detail into account using an appropriate non-linear stress-slip relationship representing localised mechanical interactions between concrete and anchoring devices; as a result, traditional analytical procedure to analyse development length can be easily extended to anchored bars using comprehensive libraries of mechanical force (stress)/slip relationships for end details. EXPERIMENTAL PROGRAM AND TEST SETUP The modelling of anchored bars, as above mentioned, requires the knowledge of the mechanical response of each component (straight bar and anchoring detail) in terms of behavioural constitutive laws. A comprehensive review of technical literature, design rules and practice [11, 12] have been carried out in order to define a typical end anchorage. The latter has been identified as a circular hook with 18 opening angle and a straight end. The diameter of the circular region and the length of the end segment depend on the bar diameter. In particular, hook diameter is 5 times the bar diameter and the straight end length is equal to 3 times the bar diameter. The experimental program is divided into two phases, depending on the type of tests: beamtest for evaluation of bond properties of straight bars and service performances of hooked anchorages; pull-out tests to analyse the response of hooked anchorages both under service and ultimate load. The first phase is characterised by 1 tests, 6 straight bars and 4 hooked anchorages, the main investigated parameter is the bar diameter, 12 and 16 mm are considered according to the results of the above mentioned review of design practice. The second phase is characterised by a more comprehensive set of investigated parameters: bar diameter, cast direction, concrete cover thickness, orientation of hooks and the type of loading (monotonic or cyclic); at the present stage more than 2 tests have been carried out. In the following an overview of the experimental results is reported in order to discuss the main aspects of load-slip behaviour of anchored smooth bars. Reinforcement used are smooth rebars that are still available for secondary purposes in r.c. structures and have mechanical

5 properties similar to steel classified as Aq42 according to Italian design Code of 6 s [13]; yielding stress is about 32 MPa, ultimate stress is equal to 43 MPa and ultimate uniform strain is about 2%. Concrete has been prepared according to typical mix rules of 6 s and tests on cubes 15 mm wide are used to define mean concrete strength. In the following, an overivew of experimental results is presented. Beam-tests The first phase of the experimental program consists of beam-test carried out according to the setup described in Figure 3. The specimens are composed of two concrete blocks that are connected by a reinforcing bar. Mean concrete cubic strength is about 34 MPa. The load transfer is slightly different respect to standard beam-tests, since a steel hinged beam, see Figure 3.d, is used to apply the load on the concrete using shear studs. Ties Φ 4 mm Bars Φ 8 mm Rebar Plastic tube Ties Φ 4 mm Bars Φ 8 mm Rebar Plastic tube 25 mm 25 mm Plastic pipe 12 mm 1 Φ a) Rebar Transducer Reference ring 5 mm 1 mm a = 9 mm 3 mm F Steel beam 6 mm a = 9 mm Hinge b) h 12 mm L strain gauges 12 mm L b) Rebar Concrete blocks Figure 3. Beam test type specimens and test set-up for straight and hooked smooth bars. The load is applied on each side of the cylindrical hinge located on the symmetry axis and axial load T on the rebar can be easily evaluated using the following equilibrium condition: d) T F a = (1) 2 h As shown in Figure 3.a and 3.b both straight and hooked rebars have been tested. In the first case, the embedment length is assumed equal to 1 Φ; in order to avoid any interaction with surrounding concrete plastic pipes are used. The given embedded length l b is required for the evaluation of the bond stress that is calculated as follows: T τ b = (2) Σ l b being Σ the bar perimeter. For hooked bar specimens, plastic pipes are located in a way that only circular branch and straight end of the rebar are embedded and the mechanical response of the anchoring device

6 can be evaluated, as shown in Figure 3.c. The load on the steel beam is applied using a mechanical actuator in displacement control; a load cell, inductive transducers and strain gauges are used to measure the load, slippage and strain of rebars respectively; a typical test setup is represented in Figure 3.d. Transducers give the slippage at the loaded and the unloaded end of rebar when straight bars are concerned, but allow only an indirect measure of the slip at the hook end, since a reference ring is installed outside the plastic pipe that avoids contact between concrete and steel. Thus measure of hook slip can be computed as follows: s hook = s ε L (3) meas meas where s hook is the slippage of hook, s meas is the measured slip at the reference section, ε meas is the measured bar strain and L is the distance between the reference ring and the end section of the hook. Pull-out tests The second phase of the program consists of modified pull-out tests on hook anchorages; the test arrangement and the geometry of specimens is described in Figure 4. The concrete specimen is a cube 3 mm wide that is restrained using shear studs on two opposite vertical faces. In this way, shear forces on connectors avoid compressive longitudinal forces applied on the top surface as happens in standard pull-out tests and thus is more representative of the behaviour of hooks in critical regions, see Figure 1.b. The concrete is characterised by a cubic mean strength of about 3 MPa. The main feature of the test setup adopted in the second phase of the research is the direct measure of the slip at the end section of the anchorage. For monotonic test, interaction of the straight branch is prevented using a plastic pipe. The slippage at the end of the circular branch is measured using a high performance draw-wire displacement sensor. In addition an extensometer is also used during all the load process in order to evaluate the stress-strain relationship of each tested bar. Plastic pipe F Steel rebar Plastic pipe F Steel rebar Concrete specimen 3 mm Concrete specimen 3 mm F 3 mm a) F 3 mm Figure 4. Test arrangement for pull-out type tests on hooked smooth bars. b) In Figure 4, some details about the specimens and the restraints are given both for monotonic and cyclic tests. The latter are characterised by a slightly different setup, since firstly buckling of rebar and then interaction at the bottom surface of concrete block have to be avoided. The first goal is reached using a thin Teflon cover on the rebar to prevent bond and ensure a adequate lateral restraint for the rebar, the second one changing the location of concrete block

7 to avoid the contact between base plate and bottom concrete surface on the analogy with typical external joint geometry. Furthermore, a reduced clear length of the rebar is allowed outside concrete, thus the extensometer has not been used, conversely three strain gauges are installed directly on the rebar to measure local strains. The tests have been carried out using a uniaxial testing system able to apply the load under displacement control and measuring slip of anchorage inside the concrete block. EXPERIMENTAL RESULTS Beam-tests In Figure 5 the results of beam tests on straight bars are summarised. The shape of the bond stress-slip relationship points out that mechanical interaction is characterised by different phases, like the initial adhesion, and the final residual strength related basically to friction mechanisms. The bond stress for 16 mm rebars is about 1.75 MPa reached at a slip of about.18 mm; for 12 mm the values are 1.5 MPa and.3 mm respectively bond stress (MPa) bars Φ 12 mm bars Φ 16 mm.9 Model Code slip (mm) Figure 5. Summary of experimental results and MC9 provisions for smooth bars. (MPa) s ref 12-1-l 12-1-r 12-2-l 12-2-r s ref (mm) a) 5 4 (MPa) 3 1 s ref σ s 16-1-l 16-1-r 16-2-l 16-2-r s ref (mm) (MPa) s hook yielding stress 12-1-l 12-1-r 12-2-l 12-2-r s hook (mm) b) 4 (MPa) yielding stress 3 1 c) Figure 6. Experimental results of beam-test on hooked bars. s hook 16-1-l 16-1-r 16-2-l 16-2-r s hook (mm) d)

8 In the same plot the theoretical bond stress-slip relationship suggested by Model Code 9 [14] is also presented; it is worth noting that peak bond stress values are higher that MC9 maximum stress, but the latter matches well with the residual experimental stress, particularly for 16 mm bars. Tests on hooked bars are reported in Figure 6; they have been carried out up to a bar strain of 7%, thus a large strain-hardening range is covered; in fact beyond this limit the reference ring for the slip measure was not able to follow the system deformation and a detachment occurred. However, slips calculated at the anchorage end are plotted until a bar strain of about 1%, that is the maximum strain value allowed by installed strain-gauges. If slips at the reference ring and at anchorage end section are compared, a very interesting aspect can be pointed out. In fact, it is easy to recognise that anchorage end slip curve does not exhibit a plastic plateau, despite the very high ductility of reinforcement and the large plateau at the reference ring location. Pull-out tests Some experimental results of modified pull-out tests are reported in Figure 7 and Figure 8 referring to monotonic and cyclic tests respectively. The first series of plots completely describe the mechanical behaviour of circular hooks and shows all the measured parameters. (MPa) (MPa) yielding s hook ε rebar (mm/mm) ε rebar (mm/mm) s hook (mm) s hook (mm) Figure 7. Summary of experimental results of pull-out tests, full type specimens. The tests are carried out up to bar failure, that is reaching strains higher than 2%, both bar strain-hook slip and stress-slip relationship for hooked anchorage show that the slip of the hook at yielding is constant and slip increases when strain-hardening is activated. Furthermore, initial stiffness of the hook stress-slip relationship is very high, and even a stress range with no slip exists. The comparison between the experimental curves obtained from beam-tests and pull-out tests points out a relative matching of the mechanical behaviour in medium and high stress levels, while a scatter is present for low stress levels. In both cases, a significative slip is obtained at bar yielding; it ranges between.8 an 1.2 mm

9 and cannot be easily neglected when the deformation capacity of the anchored bar is concerned. This value increases when strain hardening is triggered and becomes about 4 mm at failure. In the last set of plots, the results of a cyclic test are summarised. The first plot reports the load history that is chosen to lead the steel reinforcement on the plastic plateau. Figure 8.b reports the end displacement versus axial stress of the rebar relation. The yielding stresses are reported on the same plot highlithing an asymmetrical behaviour of end anchorage. In fact, for a given displacement, the stress level reached under compression is higher than the corresponding stress under tension. This behaviour occurs until the concrete spalling at the bottom of the concrete block, that is clearly shown in the plot by the sudden loss of load capacity s load (mm) 4 3 Tension yielding stress σ load (MPa) concrete spalling Compression s load (mm) yielding stress Tension yielding stress a) (MPa) b) s hook (mm) - concrete spalling Compression yielding stress Figure 8. Experimental results of cyclic test on hook Φ 12 mm. c) After this value, the capacity of dissipation is strongly reduced and compressive strength of anchorage deteriorates, enabling a pinching type effect. From a local point of view the asymmetric behaviour is enlarged, as shown in Figure 8.c where the monotonic behaviour is also plotted. In fact, a very stiff behaviour occurs under compression during the early cycles; as the slips under compression increases, a rotation counterclockwise of the stress-slip relation occurs due to the large permanent deformation related to bar pulling out. This effect is however counterbalanced by an increasing level of deformation under compression that leads to the failure of a concrete cone in the bottom region of the block. This phenomenon starts a sudden reduction of the carrying capacity and a strong increase of under compression deformation and that has been also observed during tests on frames.

10 CONCLUSIONS Modelling of existing reinforced concrete frames designed without specific seismic rules is a key problem for safeguard and protection against seismic risk. Furthermore, in many European countries, and particularly in Italy, a very large percentage of reinforced concrete buildings are 35 years old, or even older, thus reinforcement consists of smooth bars. In fact, only in 7 s early applications of deformed bar appeared and a strong research effort in the field of seismic performances of structures began. As a consequence, technical literature on mechanical performances of anchored smooth bars is not comprehensive, mainly from the deformation standpoint, despite the relevance of this aspect on the response of critical regions, i.e. beam to column joints and base column. The experimental tests that have been discussed in the present paper are aimed to describe in detail the force-slip relation of bond mechanism for straight bars and of anchoring end details, i.e. circular hooks with 18 opening angle. The results are interesting and point out some particular aspects of the behaviour under both monotonic and cyclic loading. The slippage due to anchoring devices is significant and cannot be neglected, especially in large post-yielding field; mechanisms governing stress-slip response of hooks enable a reduced yielding spreading in the anchoring device, so that at yielding, the hook slip does not show a plastic plateau and increases only when strain hardening starts. Cyclic behaviour is basically asymmetric due to high stiffness of hook under compression, bar pull in leads to concrete spalling that is a typical failure mechanism in external joint regions. REFERENCES 1. De Marco, R., Martini, M.G., Di Pasquale, G., Fralleone, A., Pizza, A.G., Italian Classification and Seismic Code from 199 to XI Forum of Public Administration,. (in italian). 2. Cosenza, E., Manfredi, G., Verderame, G.M.. Seismic assessment of gravity load designed r.c. frames: critical issues in structural modeling. Accepted for publication on special issue of Journal of Earthquake Engineering, Bach, C., Deutcher Ausschus fur Eisenbeton, Hefts 9 and 1, Abrams, D.A., Test of bond between concrete and steel, Bulletin n 71, Engineering Experiment Station, University of Illinois, Urbana, pp.238, December Saliger, R., Schubwiderstand und Verbund in Eisenbeton-balken, Mylrea, T.D., The carrying capacity of semi-circular hooks, ACI Journal, Proceedings Vol.24, pp , Fishburn, C.C., Strenght and slip under load of bent-bar anchorage and straight embedments in haydite concrete, ACI Journal, Proceedings Vol.44, n 4, pp , December Mains, R.M., Measurement of the distribution of tensile and bond stresses along reinforcing bars, ACI Journal, Proceedings, Vol.48, n 3, pp , November Cosenza, E., Manfredi, G., Verderame, G.M., A nonlinear model for underdesigned r.c.frames. Proc. of XII ECEE, London, Eligehausen, R., Popov, E.P., Bertero, V.V., Local bond-stress relationships of deformed bars under generalised excitations, UCB/EERC 83, 23, Santarella, L., Il cemento armato La tecnica e la statica, Hoepli - Milano, (in Italian) 12. CEB Bulletin d Information n 23, Commission Aciers, Adherence, Ancrages, Circolare 23 maggio 1957 n 1472, Armature delle strutture in cemento armato. 14. CEB-FIP Model Code Design Code- Comite Euro-International du Beton, 1991.