See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/223798979 Experimental behaviour of anchored smooth rebars in old type reinforced concrete buildings Article in Engineering Structures · August 2005 DOI: 10.1016/j.engstruct.2005.05.002 CITATIONS READS 33 470 3 authors: Giovanni Fabbrocino Gerardo Verderame 266 PUBLICATIONS 1,689 CITATIONS 109 PUBLICATIONS 898 CITATIONS Università degli Studi del Molise SEE PROFILE University of Naples Federico II SEE PROFILE Gaetano Manfredi University of Naples Federico II 532 PUBLICATIONS 4,870 CITATIONS SEE PROFILE Some of the authors of this publication are also working on these related projects: Progetto Reluis - Coordinamento Progetto speciale Osservatorio Sismico delle Strutture e Monitoraggio View project All content following this page was uploaded by Giovanni Fabbrocino on 13 December 2016. The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document and are linked to publications on ResearchGate, letting you access and read them immediately. Engineering Structures 27 (2005) 1575–1585 www.elsevier.com/locate/engstruct Experimental behaviour of anchored smooth rebars in old type reinforced concrete buildings Giovanni Fabbrocinoa,∗, Gerardo M. Verderameb, Gaetano Manfredib a Department S.A.V.A., University of Molise, Via De Sanctis – 86100 Campobasso, Italy b Department of Structural Analysis and Design, University of Naples Federico II, Via Claudio, 21 – 80125 Napoli, Italy Received 1 December 2004; received in revised form 2 May 2005; accepted 3 May 2005 Abstract Modelling of existing reinforced concrete (r.c.) frames designed without specific seismic rules is a key problem for maintenance, structural upgrading and seismic assessment. In many European countries a very large percentage of reinforced concrete buildings are 40 years old, or even older; thus reinforcement consists of smooth rebars, since only in the 1970’s did early applications of deformed rebars appear. Technical literature on mechanical performances of anchored smooth rebars is non-comprehensive, mainly from the deformation standpoint, despite the relevance of this aspect to the response of critical regions, i.e. beam to column joints and column bases. In the present paper a series of experimental tests on smooth rebars are presented; they are aimed at describing in detail the force–slip relation for the bond mechanism for straight rebars and for anchoring end details, i.e. circular hooks with a 180◦ opening angle. © 2005 Elsevier Ltd. All rights reserved. Keywords: Old type r.c. constructions; Seismic assessment; Smooth reinforcement; Anchorages; Bond 1. Introduction The first step in upgrading strategies for addressing existing reinforced concrete (r.c.) structures is the assessment of seismic performances of materials and structural systems. In fact, many existing constructions in seismic areas have been designed only for gravity loads or according to outdated seismic rules, resulting in low available ductility and lack of a strength hierarchy. The measure of global ductility for framed structures is the interstorey drift ratio, that for reinforced concrete frames is dependent upon different contributions like the beam plastic rotation, the column flexural behaviour and the beam to column joint region deformation [1]. The latter is generally divided into two components related to shear deformation of the panel zone and to fixed-end rotation that is predominant in underdesigned structures and depends on the bond properties of reinforcement and anchoring devices [2]. The present paper ∗ Corresponding author. Tel.: +39 0874 404779; fax: +39 0874 404855. E-mail address: giovanni.fabbrocino@unimol.it (G. Fabbrocino). 0141-0296/$ - see front matter © 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.engstruct.2005.05.002 deals with the smooth reinforcement widely used up to the 1970’s in a very large number of existing constructions; they exhibit poor bond performances resulting in mandatory anchoring end details able to ensure the required level of interaction. Thus, the behaviour of anchored smooth rebars is a key issue in the development of reliable procedures for the evaluation of available bearing and/or displacement capacities of buildings. In fact, advanced structural analyses are able to take account of actual element dimensions, critical details, laboratory and field test data on the concrete and reinforcement, as shown in [3]. In the following, the results of an experimental evaluation on the behaviour of straight smooth rebars and 180◦ circular hooks are discussed; some interesting features of the structural response under service loads and in the large post-yielding field are pointed out. 1.1. Literature review Early experimental studies [4,5] were aimed at evaluating the effectiveness of end details on smooth reinforcement 1576 G. Fabbrocino et al. / Engineering Structures 27 (2005) 1575–1585 Fig. 1. Hooked rebar model and its role in the deformation of critical regions. pull-out. In the same period, the research by Saliger [6] was very interesting due to the large number and varying kinds of tests; these were aimed at evaluating the strength of anchorages, without any consideration of performance in terms of deformation. Straight and hooked rebars (180◦ opening angle) were tested; different rebar diameters, curvature radii and transverse rebar arrangements were considered. Results on pull-out tests agreed with Bach’s tests on beams [4], which exhibited an increased flexural strength due to end anchorage and demonstrated that the shape of the anchorage and transverse reinforcement could give beneficial effects. Later, research by Mylrea [7] took into consideration both the strength and the deformation of hooked smooth rebars with a 180◦ opening angle and tried to outline the influence of the end hook radius and transverse reinforcement. Results showed a slight influence of the hook radius on the deformation and indicated the role of transverse reinforcement in the type of failure. In particular, plain specimens showed a non-ductile behaviour by concrete failure, compared to rebar failure induced by appropriate transverse rebar detailing. Approaching the 1950’s, early types of deformed rebars were investigated and compared as regards the force–slip response to straight and anchored smooth rebars [8]. More than forty specimens were tested, varying the hook radius, surface type (smooth or ribbed), development length and opening angle. During the 1960’s experimental works on smooth rebars carried out by [9,10] offered a series of data used as references in many later studies in the field of bonding. However, available technical and experimental data on smooth rebars as reinforcement for concrete structures are not fully satisfactory as regards the development of a reliable numerical modelling of anchored rebars placed in critical regions. In fact, the majority of data can be found in the framework of studies aimed at assessing the performances of deformed reinforcements and defining safe design provisions [14]; accordingly, smooth rebars are used as a reference but are not fully investigated, particularly if a large post-yielding phase is considered. This is the case also for more recent research on plain rebars that today are commonly used for precast concrete elements [12]. In summary, a more comprehensive approach is required in order to develop reliable numerical procedures and give consistent predictions of the r.c. construction response under seismic actions or of the residual bearing capacity of existing buildings [11,13]. 2. Research objective A review of experimental analyses aimed at the evaluation of smooth rebars and related anchoring end details reveals a lack of data in the large post-yielding field response. A number of studies on the subject were carried out, but this was many years ago, so the limitations of the testing equipment due to the technology available at that time lead to results being non-comprehensive. In fact, it has been found that results rarely refer to strain levels beyond yielding and that the use of equipment under force control prevents the reporting of descending branches that could be relevant for modern applications. From a behavioural perspective, the response of hooked smooth rebars results from the interaction between two distinct components: the straight rebar portion, where the behaviour is basically related to the bond interaction; and the anchoring device, made of a circular hook, where the specific rebar shape activates local interaction mechanisms that involve large volumes of concrete. Fig. 1 reports the hooked rebar model and the idealised constitutive relationship of the components identified for analysing the problem: the bond between the smooth rebars and the concrete in the straight rebars, and the anchoring end detail response. The latter can be given in terms of the pull-out force (and/or stress) and slippage at the end of the anchorage; a ‘global’ representation of localised interaction mechanisms involving the hooked rebar and the surrounding concrete can be obtained. Such an idealisation has a number of advantages, since it makes the analysis of the anchored rebar using consolidated numerical techniques reliable [14] and gives stable solutions even in the large post-yielding field. The present paper reports the results of an experimental programme aimed at calibrating proper constitutive G. Fabbrocino et al. / Engineering Structures 27 (2005) 1575–1585 1577 relationships for anchoring devices; specific tests for characterising the two components of the anchored smooth reinforcement (straight rebar and hooked end) are discussed. These tests represent the experimental background of reliable models of reinforced concrete joints where smooth rebars are used [15]. 3. Test programme The experimental programme described in this paper consists of 20 tests with distinct aims: • evaluation of smooth rebar bond properties with three beam tests and three pull-out tests; • evaluation of the response of hooked anchorages both in service and at ultimate load with fourteen pull-out tests. All the tests are carried out on both straight and hooked 12 mm rebars. The selection of steel rebars was based on the mechanical and surface properties of materials used in the decade 1960–1970 [16]. The hooked rebar geometry was defined after a comprehensive review of Italian and international design codes and manuals used as reference in the reference period [17]. In particular, the hook geometry can be described by referring to two dimensionless parameters: the ratio between the inner diameter of the hook and the rebar diameter, equal to 5, and the ratio between the straight end length and the rebar diameter, generally equal to 3. 3.1. Material properties The smooth rebars used in the context of the present work are hot rolled and classified as Feb22k [16]; in particular, tensile tests carried out on 12 mm reinforcements have shown a mean yielding stress σs,y = 320 N/mm2 , initial hardening under strain εsh = 3%, ultimate stress σs,u = 440 N/mm2 and ultimate strain εs,u = 23%. Stress–strain plots are reported in Fig. 2(a), where the significant ductility can be recognised together with a large strain hardening ratio (1.375). The concrete has been prepared according to typical mixing rules of the 1960’s [18] and tests on cubes 150 mm wide were used to define the mean concrete strength. Table 1 reports the concrete mix design data for both beam test and pull-out specimens that have been prepared in two distinct phases, characterised by different strength developments probably due to the different humidity of the coarse aggregates. Specimens and cubes for strength evaluation have been cast together and cured in the same open air environmental conditions for 28 days before testing. The beam test specimens exhibited a mean cubic compressive strength of 34.20 MPa; Table 2. The pull-out test specimens exhibited a mean cubic compressive strength of 29.34 MPa; Fig. 2(b). Fig. 2. Steel stress–strain plot (a); concrete compressive strength of pull-out specimens (b). Table 1 Concrete mix design Component Water/cement ratio Aggregate size (0–4 mm) Aggregate size (4–10 mm) Aggregate size (10–20 mm) (kN/m3 ) (kN/m3 ) (kN/m3 ) 0.45 10.14 3.13 5.16 Table 2 Compressive strength of beam test specimens Cube Cubic strength (MPa) (MPa) Cylindrical strength (MPa) (MPa) 1 2 3 4 5 6 33.60 34.00 33.60 35.70 33.60 34.70 26.90 27.20 26.90 28.60 26.80 27.70 34.20 (2.49) 27.30 (2.54) 1578 G. Fabbrocino et al. / Engineering Structures 27 (2005) 1575–1585 Fig. 4. Test arrangement for pull-out type tests on straight bars. Fig. 3. Beam test set-up (a); specimen type (b). 3.2. Test set-up The beam test has been carried out according to the setup described in Fig. 3. Specimens were composed of two concrete blocks connected by a reinforcing rebar. The load transfer is slightly different with respect to standard beam tests [19], since a steel hinged beam, Fig. 3(a), is used to apply the load on the concrete using shear studs. The load is transferred on each side of the cylindrical hinge located on the symmetry axis and the axial load T on the rebar can easily be evaluated using the equation of equilibrium between the moment due to external force and the resistance one due to tensile stresses in the rebar. The embedment length, L b , is assumed equal to 10Φ; in order to avoid any interaction with surrounding concrete, plastic pipes are used. The embedded length L b is used to evaluate the bond stress, which is calculated assuming a constant distribution of the bond stress along the rebar. 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 the rebars respectively; transducers give the slippage at the loaded and the unloaded ends of rebar — Fig. 3(b). The set-up for the pull-out tests is shown in Fig. 4; specimens were made of cubes, 300 mm wide, that incorporate the rebar to be tested. A plastic pipe is used to avoid interaction between the rebar and the surrounding concrete except in the embedded zone, 10Φ long. The testing equipment is completed by a bolted steel envelope that restrains the concrete block by means of threaded rebars on the lateral surfaces, as shown in Fig. 4. It is worth noting that special care has been devoted to avoiding any tensile force Fig. 5. Test arrangement for pull-out type tests on hooked smooth bars; Full type specimen. in the threaded rebars and keeping the concrete unconfined, without lateral compressive stresses. Measurements of the tensile force F, of the rebar strain and of the slips between the loaded and the unloaded ends and the concrete are taken; tests are carried out under displacement control, so that descending branches can be fully detected. The second phase of the programme consists of modified pull-out tests on hook anchorages. The main parameters investigated are: the concrete cover; the cast direction; the position of the circular branch with respect to the top surface of the specimen. In fact, two types of specimens have been designed in order to modify the hook concrete cover in compliance with the reinforcement detailing in base column/internal beam to column joints and external joints respectively. Therefore, three test set-ups have been considered: • ‘Full’ type specimens, shown in Fig. 5, that consist of a concrete cube 300 mm wide and the rebar centred in the cross section; this leads to a significant concrete cover that can be representative of the above-mentioned base column or the internal beam to column condition. • ‘End’ type specimens, shown in Fig. 6, that are representative of the typical location of rebars in external beam to column joint regions; in this case, the concrete cover is 22 mm, and the concrete block has a 180 mm thick by 300 mm wide cross section. • ‘Full-H’ type specimens; these are characterised by the same geometry as Full type specimens, but the cast direction is perpendicular to the rebar and the location of G. Fabbrocino et al. / Engineering Structures 27 (2005) 1575–1585 1579 Fig. 6. Test arrangement for pull-out type tests on hooked smooth bars; End type specimen. the circular branch with respect to the block top surface is changed. In all cases, the load is applied to the free end of the rebar and the reaction force is imposed by an external steel box restrained to the concrete block using bolts embedded in the concrete. This specific set-up has been chosen in order to avoid compressive stresses on the top surface of the concrete and to fit the real conditions of rebars under tension in cracked sections. The bolts used as shear connectors are the only reinforcement present in the concrete blocks and are characterised by zero pre-tension to avoid lateral confinement of concrete, similarly to the bond tests. The main aspect of the test set-up is the direct measurement of the slip at the end section of the anchorage; in fact, interaction of the straight branch is prevented using a plastic pipe, as shown in Figs. 5 and 6, and the slippage at the end of the circular branch is measured using a high performance draw-wire displacement sensor. Preliminary validation of measurements taken by draw-wire transducers has been performed in order to avoid incorrect data; the axial tensile force generated by the transducer and the very low flexural stiffness of the wires used for measurements of the slippage ensured the reliability of the system. In addition, an extensometer has also been used throughout the load process in order to evaluate the stress–strain relationship of each rebar tested. The tests have been carried out using a uniaxial testing system able to apply the load under displacement control and measuring the slip of the anchorage inside the concrete block. 4. Experimental results 4.1. Bond test Due to the nature of the reinforcement and the geometry of the specimen, splitting phenomena did not occur, so concrete blocks were not damaged macroscopically during the tests. Measurements of slippage at loaded and unloaded ends demonstrated that the differences between them are negligible, so the constitutive relationships have been plotted depending on the unloaded end slip. Fig. 7. Beam test results: (a) steel stress–slip plot; (b) bond stress–slip plot. Figs. 7(a) and 8(a) show both beam test and pull-out test results; evaluation of the bond stress τb has been carried out, depending on the bonded length L b = 10Φ and on the tensile reinforcement stress σs as follows: τb = σs · As Σ · Lb (1) where As and Σ are the area and the rebar perimeter respectively. Figs. 7(b) and 8(b) show clearly the different phases of the interaction phenomenon: adhesion with negligible slips, interlocking with increasing slip up to a peak value and then a friction based residual stress. The mean value of the bond stress is about 1.42 MPa, corresponding to a slip of about 0.04 mm for beam test type specimens, and 1.96 MPa, corresponding to a slip of about 0.14 mm for pull-out test type specimens; see Table 3 for details. In the same plot the theoretical bond stress–slip relationship suggested by Model Code 90 (MC90) [20] is also presented; it is worth noting that peak bond stress values 1580 G. Fabbrocino et al. / Engineering Structures 27 (2005) 1575–1585 Fig. 8. Pull-out test results: (a) steel stress–slip plot; (b) bond stress–slip plot. Table 3 Results for pull-out and beam test specimens Specimen 1 2 3 Mean value (COV) Pull-out test Slip τb max (mm) (MPa) Beam test Slip (mm) τb max (MPa) 0.14 0.14 0.15 2.30 1.90 1.67 0.05 0.03 0.05 2.16 1.06 1.05 0.14 (4.12) 1.96 (16.26) 0.04 (28.86) 1.42 (44.92) are higher than the MC90 maximum stress, but the latter matches well with the residual experimental stress. 4.2. Hook test Pull-out tests on circular hooks in the three different arrangements (Full, End, Full-H) are reported in the present section. The reinforcement has a diameter of 12 mm, in compliance with the previously discussed bond tests. It is worth noting that Full and End type specimens allow for analysing the influence of the concrete cover on the hook behaviour, while Full-H tests indicate the role of the concrete casting direction. Five Full type specimens, described in Fig. 9(c), are first examined. In particular, Fig. 9(a) reports the measured stress–strain relationships of the rebars, Fig. 9(b) represents the stress–slip relationship at the hook end and finally Fig. 9(d) gives the strain versus hook slip relationship plot. Analysis of experimental results indicates a strongly nonlinear behaviour of the anchoring device even for low stress levels in the rebar; in more detail, an initial high device stiffness is observed, since zero slips are measured for stress levels lower than 50 MPa. The slippage of the circular hook taken at the yielding stress (strain) exhibits a considerable variability ranging from 0.80 mm to an upper bound of 2.55 mm. The behaviour of the hook in the post-yielding phase of the rebar is characterised by an interesting phenomenon that can be recognised with reference to Fig. 9(d). In fact, slippage at the yielding stress remains basically constant over the whole plastic plateau of the rebars and increases only when strain hardening starts; this circumstance is also confirmed by a stress–slip relationship that seems to be continuous and does not show a sudden increase at the yielding stress. This phenomenon is common to all the tests, confirming that yielding spreading does not occur along the circular branch and the interaction is basically governed by a mechanical interlock, which leads to the concentration of the normal stress at the end of the hook without yielding penetration along the curved branch. On the other hand, evaluation of the stress–strain relationship of rebars show that the plastic deformation takes place over the whole unbonded straight branch of the rebar. The strain hardening phase, as already mentioned, is then characterised by an increase of the slippage at the hook end activated by the load increase, even if a progressive reduction of the stiffness is observed up to rebar failure. The slippage measured at rebar failure ranges between 2.66 and 7.86 mm. In Fig. 10(a), a Full type specimen during the test is reported on; Fig. 10(c) indicates the tensile failure of the rebar. Both pictures clearly show the devices used to measure the rebar strain and the anchoring device slippage. The specimen cross section is then reported in Fig. 10(b) for after a pull-out test; it is easy to recognise that the unloaded end of the hook is basically affected by slippage without visible concrete damage. The plots of Fig. 11 referring to the End test series show that the shape of the curve is similar to the previous ones, but the results are less scattered in terms of the hook slip at yielding; in fact measured values range from 1.62 to 2.00 mm. The stress/strain curves of the rebar, however, indicate that a sudden loss of load occurs in the large postyielding field; this loss is significant since it can reach even 60% of the maximum load. This phenomenon is due to a splitting type of failure that occurs in the concrete cover, G. Fabbrocino et al. / Engineering Structures 27 (2005) 1575–1585 1581 Fig. 9. Summary of experimental results of pull-out tests; Full type specimens. Fig. 10. Full type specimen. Set-up (a), final state of the anchorage after the pull-out test (b), anchorage failure (c). triggering a relevant increase of the slip at the hook end, as shown clearly by the rebar strain–hook slip relation. Nevertheless, the anchorage is still able to bear tensile stresses in the cracked state also, so in many cases the pre- cracking load can be recovered and a progressive pull-out of the rebar develops. On this subject, it is worth noting that the solution adopted to restrain the concrete block in End type specimens 1582 G. Fabbrocino et al. / Engineering Structures 27 (2005) 1575–1585 Fig. 11. Summary of experimental results of pull-out tests; End type specimens. did not affect the response of the hook due to the absence of lateral confinement and to the tolerances used for bolt installation (parallel to hooks), resulting in free relative displacements between the concrete and surrounding steel envelope. A review of experimental tests indicates that three different responses of the anchoring device in the postcracking phase occurred. The first behaviour is characterised at the crack formation by a sudden loss of bearing capacity that can be estimated as 40% of the peak load and then by a gradual reloading phase affected by large slips of the hook up to rebar failure (specimens 1 and 2). The second type of behaviour exhibits, similarly to the previous one, a sudden loss of load at crack formation, but the reloading branch is not able to trigger the rebar failure due to large slips of the hook, so anchorage failure can be recognised in Fig. 12(c) (specimen 4). The last type of behaviour is then characterised by a Full type specimen stress–slip response, without any clear loss of load (specimens 3, 5). Fig. 12 reports a number of pictures taken during and after the End type specimen tests. In particular, Fig. 12(b) shows the final state of a specimen after the pull-out test; an estimation of the slip at the unloaded end of the circular hook can be made with reference to the measuring device placed upon the cracked surface. Furthermore, a comparison between Figs. 12(b) and 10(b) indicates a large extension of the concrete damage in End type specimens due to the local stresses between the steel rebar and the surrounding medium. The last set of plots, in Fig. 13, show the response of hooks depending on the casting direction and on the position of the circular branch with respect to the top surface of the specimens (Full-H type). The results can be easily divided into two groups, showing the influence of the last parameter on the response of the hooks. In fact, hooks placed downwards are stiffer both at yielding and at collapse due to rebar failure. Table 4 reports slips measured at the yielding stress/strain, s y , and the one corresponding to the strain hardening start, ssh , for each group of tests (Full and End type specimens); mean values of the above parameters are also given. G. Fabbrocino et al. / Engineering Structures 27 (2005) 1575–1585 1583 Fig. 12. End type specimen. Set-up (a), final state of the anchorage after the pull-out test (b), anchorage failure of specimen 4 (c). Fig. 13. Summary of experimental results of pull-out tests; Full-H type specimens. It is worth noting that basically End type specimens are characterised by a larger deformability compared with Full type ones; the increase of deformation can be estimated as around 20%. This result is not necessarily related to cracking of concrete, since it occurs in the advanced strain hardening phase; conversely it can be related to a different level of confinement of the hook depending on the concrete cover thickness. 1584 G. Fabbrocino et al. / Engineering Structures 27 (2005) 1575–1585 5. Conclusions Table 4 Results of pull-out test: Full and End type specimens Type Specimen 1 2 3 4 5 Mean value (COV) Full sy (mm) ssh (mm) su (mm) End sy (mm) ssh (mm) su (mm) 1.58 2.55 1.45 0.96 0.80 1.58 2.55 1.45 1.07 0.97 4.26 7.86 4.45 3.40 2.66 1.96 2.00 1.62 1.86 1.62 2.10 2.03 1.71 1.91 1.71 – – – – – 1.47 (46.73) 1.52 (41.27) 4.53 (44.07) 1.81 (10.08) 1.89 (9.50) – – However, the different evolutions of phenomena in the strain hardening phase prevents a comprehensive comparison between slips at failure su , so Table 4 reports only the values measured for Full type specimens. Table 5 provides an estimation of the influence of the cast direction on the response of the anchoring devices. In fact, slips measured at the yielding stress/strain, s y , and those corresponding to the strain hardening start, ssh , and to rebar failure, su , are reported with reference to FullH type specimens; mean values of the above parameters are also given. It is easy to recognise that ‘up’ specimens exhibit a larger deformability compared with ‘down’ ones; the increase can be estimated as 80% in the case of yielding slip and about 20% in the case of ultimate slip. References Table 5 Results of pull-out tests: Full-H type specimens Type Full-H (down) sy ssh (mm) (mm) 1 2 0.72 1.27 Mean value 1.00 Specimen Seismic assessment of old type r.c. constructions is of interest for structural engineers; however, knowledge of basic interaction phenomena involving smooth rebars is not comprehensive, especially due to the lack of data ranging from yielding to the large strain hardening field. This circumstance represents a limitation on the use of advanced non-linear analysis procedures. The present paper gives an experimental contribution in this research area. Beam tests and pull-out tests allowed for describing in detail the force–slip relation of the bond mechanism for straight rebars and that of anchoring end details, i.e. circular hooks with a 180◦ opening angle. The results indicate some particular aspects of the behaviour under monotonic loading. The slippage due to anchoring devices is relevant and cannot be neglected, especially in the large post-yielding field; mechanisms governing the stress–slip response of hooks allow a reduced yielding spreading in the anchoring device, so at yielding, the hook slip does not show a plastic plateau and increases only when strain hardening starts. The concrete cover plays a role in the large post-yielding field, since splitting type failures have been observed in End specimens that fit the rebar embedment in the external beam to column regions. The casting direction seems to have an influence on the behaviour, together with the relative position of the hook with respect to the top surface of the concrete specimen. su (mm) Full-H (up) sy ssh (mm) (mm) su (mm) 0.73 1.36 2.80 3.75 1.83 1.83 1.99 1.93 4.16 4.25 1.05 3.28 1.83 1.96 4.21 A combined review of results given in Tables 4 and 5 demonstrates that Full-H “down” specimens, representative of beam top smooth reinforcement, are characterised by a stress–slip response that is stiffer that the Full-H “up” configuration, representative of beam bottom smooth reinforcement, and of the Full one, which is column reinforcement. Finally, comparison of experimental data obtained from modified beam tests and pull-out tests indicates the uncertainties of beam test data related to indirect measurement of slippage with respect to the alternative solution adopted for the pull-out tests; as a result, the improved reliability of the pull-out tests allows for the assessment of the actual response of hooked anchors. [1] Bonacci JF, Wight JK. Displacement-based assessment of reinforced concrete frames in earthquake. In: Mete A. Sozen symposium. ACI publication SP 162; 1996. p. 117–33. [2] Cosenza E, Manfredi G, Verderame GM. Seismic assessment of gravity load designed r.c. frames: critical issues in structural modeling. Journal of Earthquake Engineering 2002;6(1) [special issue]. [3] Cosenza E, Manfredi G, Verderame GM. A nonlinear model for underdesigned r.c. frames. In: Proceedings of XII ECEE. 2001. [4] Bach C. Deutcher Ausschus fur Eisenbeton. Hefts 9 and 10.1911. [5] Abrams DA. Test of bond between concrete and steel. Bulletin no. 71. Urbana: Engineering Experiment Station, University of Illinois; 1913. p. 238. [6] Saliger R. Schubwiderstand und Verbund in Eisenbeton-balken. 1913. [7] Mylrea TD. The carrying capacity of semi-circular hooks. ACI Journal, Proceedings 1928;24:240–72. [8] Fishburn CC. Strength and slip under load of bent-bar anchorage and straight embedment in haydite concrete. ACI Journal, Proceedings 1947;44(4):289–308. [9] Rehm G. Über die grundlagen des verbundes zwischen stahl un beton. DafStb, H. 138. Berlin: W. Ernst u. Sohn; 1961. [10] Rehm G. Kriterien zur beuterilung von bewehrungsstaben mit hoch wertigem verbund. In: Stahlbetonbau. Berlin: W. Ernst u. Sohn; 1969. p. S. 79–96. [11] Kankam LK. Relationship of bond stress, steel stress and slip in reinforcing concrete. Journal of Structural Engineering 1997;79–85. [12] Mo YL, Chan J. Bond and slip of plain rebars in concrete. ASCE Journal of Materials in Civil Engineering 1996;8(4):208–11. G. Fabbrocino et al. / Engineering Structures 27 (2005) 1575–1585 [13] Fabbrocino G, Verderame GM, Manfredi G, Cosenza E. Experimental response and behavioural modelling of anchored smooth bars in existing rc frames. In: International conference ‘Bond in concrete – from research to standards’. 2002. [14] Eligehausen R, Popov EP, Bertero VV. Local bond–stress relationships of deformed bars under generalized excitations. UCB/EERC 83, 23. 1983. [15] Fabbrocino G, Verderame G, Manfredi G, Cosenza E. Structural models of critical regions in old-type r.c. frames with smooth rebars. Engineering Structures 2004;26:2137–48. [16] Verderame GM, Stella A, Cosenza E. Mechanical properties of reinforcement used for r.c. constructions in 1960s. In: 10th [17] [18] [19] [20] 1585 national conference “L’Ingegneria Sismica in Italia”. 2001 [in Italian]. Fabbrocino G, Verderame G, Manfredi G, Cosenza E. Experimental behaviour of smooth bars anchorages in existing r.c. buildings. In: fib 2002 Osaka congress. 2002. Paper W-463. Verderame GM, Manfredi G, Frunzio G. Mechanical properties of concrete used for r.c. constructions in 1960’s. In: 10th national conference “L’Ingegneria Sismica in Italia”. 2001 [in Italian]. RILEM technical recommendations for the testing and use of construction materials. RC5: Bond test for reinforcing steel. 1. Beam test, 1982. UK: E & FN Spon; 1994. CEB Bulletin No. 213/214. CEB-FIP model code 90. 1993.