Seismic Strengthening of Beam-Column Joints Using Diagonal Steel Bars
†1Hande GÖKDEMİR1 Tuğrul TANKUT2
ABSTRACT
It is known that beam-column connections are subjected to high stresses during earthquakes.
Therefore, the design of these connections plays an important role for an earthquake resistant structure. Closely spaced stirrups should be used in beam-column connections to carry these stresses safely. Since the placement of ties required by the code in this region is rather difficult, this condition is ignored in many cases. Consequently, beam-column joints are sometimes deficient in terms of seismic behavior. In a beam-column connection which has no sufficient shear reinforcement, it is suggested that additional reinforcement arranged by placing steel bars into diagonal holes drilled from the outside and anchoring them by epoxy can carry principal tensile stresses in beam-column connections. This experimental study was carried out at the METU Structural Mechanics Laboratory by using T-shaped test specimens reflecting an exterior joint region of a frame having usual dimensions reduced to a scale of approximately 2/3. Total of eleven specimens were tested, where four were reference specimens and remaining seven were strengthened at the joint region with diagonal steel bars.
Keywords: Beam-column joints, steel bars, seismic strengthening, epoxy.
1. INTRODUCTION
Various techniques have been developed in the last few decades to improve the seismic performance of reinforced concrete framed structures through “member strengthening technique” by strengthening of all deficient structural members one by one. Individual columns and individual beams can now be effectively strengthened by using these techniques. The aim of “system behaviour improvement techniques” is to use frames having rigid vertical elements instead of insufficient lateral force load-bearing systems. To achieve this, either new elements are placed to relieve the existing building by taking a large amount of earthquake loads or base isolators and dampers are used to reduce the earthquake forces transferred to the buildings.
There are many techniques for improving the seismic performance of reinforced concrete frame structures of individual members such as columns, beams, shear walls, floors and
1 Eskisehir Osmangazi University, Eskişehir, Turkey - handeg@ogu.edu.tr 2 Middle East Technical University, Ankara, Turkey - ttankut@metu.edu.tr
† Published in Teknik Dergi Vol. 28, No. 3 July 2017, pp: 7977-7992
foundations. These are categorized under “member strengthening”. However, these techniques are not applicable to seismic strengthening of beam-column joints.
Additionally, beam-column joints of reinforced concrete framed structures are critical zones since they are subjected to severe stress reversals under seismic action. Although recent codes require considerable joint reinforcement, joints are not properly reinforced in most of the existing building structures that were built earlier. These structures are in need of seismic strengthening for their joints.
Earthquake forces lead to severe stress forces in beam-column joints. Principal diagonal tensile stresses cause cracks in concrete and gradually disintegrate the concrete in the beam- column joint zone. In order to carry these stresses, it is necessary to use stirrups in joint regions. Indeed, current seismic codes require ties used at the column ends to be continued in the beam-column joint zone. But this application is rather difficult in the joint zones. Most of the time, there are no ties in beam-column joints for most of the existing buildings. Various techniques are available for seismic strengthening of columns and beams. Although beam- column joints are very critical zones, since they are subjected to severe stress reversals under seismic action, a practical and effective joint strengthening technique has not yet been developed.
Especially in Turkey, because of the insufficiency of lateral rigidity, the use of “member strengthening” is not very successful, sufficient and cost-effective. Therefore, the "System Behavior Improvement" strategy is generally adopted. This approach aims to bring the defective building elements to an adequate state by receiving a good deal of the seismic load and thus relieving the deficient members. A limited number of elements may still require strengthening and they are treated as necessary. Besides, when this strategy is followed, the insufficiency of the beam-column connections as well as other inadequate elements loses its importance, since insufficient beam-column joints start to receive much smaller stresses and become satisfactory.
Even though it is not very common in Turkey, “member strengthening” methods for improving the seismic behaviour of buildings are widely used in many countries around the world. Thus, an effective, economic and practical strengthening technique is needed for beam column joints.
Beam-column joints need to be strengthened by effective and convenient methods, in order to prevent collaps. Another important point is that the strengthening should be done quickly without evacuating the buildings and without discomforting the people living in their buildings. The basic principle for preventing the collapses of buildings caused by the joint collapses is to ensure prior hinging in the beams without significant damage.
Antonopoulus [1], Ghobarah [2], Prota et al. [3] refer to beam-column joint strengthening techniques using CFRP (carbon fiber reinforced polymer) and GFRP (glass fiber reinforced polymer) in the literature. However, these studies can hardly be considered as joint strengthening. The proposed interventions could be applicable to simple and isolated test specimens or to strengthen the ends of the beams/columns, however, they are not practical in the presence of slabs and cross beams, and not applicable to three-dimensional joint zones.
The investigation reported by Gökdemir H. [5], in the present paper is probably the first realistic attempt to develop a genuine technique towards seismic strengthening of the joint
zone itself. The present study is an effort in this direction. The basic principle of the proposed technique is to implant epoxy anchored bars into the holes drilled through the joint zone in the two diagonal directions to take care of the principal diagonal tension developing in the joint under seismic loads. Tests have been performed using reinforced steel bars as the additional reinforcement. Considering the cases where it is not possible to drill a diagonal hole crossing the joint to enable the use of a single bar, the possibility of using two lap-spliced bars inserted from either sides was also studied. Material type, the amount of the material, anchoring conditions, and the case of lap-splicing were the main parameters investigated in this study. The aim of the study was to strengthen the joint to prevent premature failure of the zone prior to hinging in the beam. If this condition is provided, the more ductile behaviour will occur. Thus, a new, practical and effective joint strengthening technique is developed for beam-column joints, which are very critical zones subject to severe stress reversals under seismic action. Test results indicate that the proposed technique can effectively be used in seismic strengthening of the joints; if the required amount of steel bars are placed in the joint and the ends are properly anchored.
2. EXPERIMENTAL WORK 2.1. Test Specimens
A typical reinforced concrete framed structure of usual dimensions was considered as a prototype. Points of inflection were assumed to develop at the mid-points of the beams and columns under the seismic action. As inspired by the test specimen successfully used in an earlier study made in Canada by Seçkin M. [6], a sub-assemblage consisting of an external bent of such frame, bounded by three points of inflection, was isolated to define the test specimen.
Table 1. The compressive strength of concrete specimens in the experiment day
Specimen fcm (MPa)
PR 19.7
RL 20.8
RC 21.5
RU 21.4
S20B 21.3
S20W 22.2
S20 20.6
S16B 20.1
S20BL 21.4
S16BL 21.3
S14BL 20.9
Table 2. The properties of the reinforcing bars The place of the reinforcing
steel bars Number Diameter (mm) Property
Yield Strength
fy (MPa)
Ultimate Strength fu (MPa)
Beam longitudinal 5 20 Deformed 536 623
Column longitudinal 4 20 Deformed 536 623
Diagonal 4-8 16-20 Deformed 430-536 667-623
Ties 32-36 8 Plain 360 490
The prototype dimensions were reduced to 2/3 scale, and the length of the beam was further reduced to fit the capacities of the laboratory facilities available. An external joint was chosen, not only for the ease of application but also assuming that a strengthening technique successful on external joints will be even more successful on internal joints. With the same consideration, beam span was reduced so that the shear at the joint was increased.
Test specimens were reinforced according to the Turkish reinforced concrete code and the Turkish seismic code. No joint ties were used in series B specimens. The resulting test specimen geometry and reinforcement are illustrated in Figure 1.
Figure 1. Test specimen geometry and reinforcement
The specimens were made of a C20 grade of concrete shown in Table 1 and S420 grade of reinforcing steel shown in Table 2. The desired concrete strength could be obtained with an acceptable variation, but the yield strength of steel was somewhat higher than the specified value. However, this deviation did not cause any complications since the same steel was used in all specimens.
A two-component epoxy was used for anchoring the steel bars into the concrete. This was a commercially available epoxy known in the market by the name MBRACE Adesivo Saturant A & Saturant B.
Diagonal steel bars in the end of the joints were mechanically anchored either by welding or by nuts as shown in Figure 2.
Figure 2. Steel bar end anchorage by nuts
The joint strengthening bars were inserted from the two sides and were lap spliced in some test specimens, namely S20BL, S16BL, S14BL where the lap length was 400 mm (20 bar diameters) and the distance between the lapped bars was 20 mm.
The test programme consisted of two series;
(i) Reference tests,
(ii) Strengthening by steel bars.
Test series and specimen properties in each series are displayed in Table 3. As can be observed in the table, the test programme was designed to investigate the parameters below:
Effect of the amount of ties in the joint on the behaviour and strength
Effectiveness of joint strengthening by diagonal steel bars Effect of the anchorage at the ends; bolted, welded, straight Effect of the amount of steel implanted
Effect of the lap splicing
Table 3. Test series and specimen properties
Series
Test
Specimen Joint Ties
Joint
Strengthening Remarks/
Anchorage
A Reference
PR - - Pilot test
RL - - Reference, lower
bound
RC Ties, 8/80 mm - Reference, code
level
RU Ties, 8/50 mm - Reference, upper
bound
B Steel bar
implant
S20B - 2 220 diag. Bolt + epoxy
S20W - 2 220 diag. Weld + epoxy
S20 - 2 220 diag. Straight, just epoxy
S16B - 2 216 diag. Bolt + epoxy
S20BL - 2 220 diag. Bolt + epoxy,
lapped
S16BL - 2 216 diag. Bolt + epoxy,
lapped
S14BL - 2 120 diag. Bolt + epoxy,
lapped
2.2. Test Set-Up
The test set-up is schematically shown in Figure 3. Both at the top of the upper half column and at the bottom of the lower half column, the specimen was pin connected to the strong wall. At the free end of the beam, reversed cyclic load was applied in the vertical direction by a double acting hydraulic cylinder coupled with a load cell connected with pin connectors to the specimen at the top and to the strong floor at the bottom.
A column axial load equal to 20% of the column axial load capacity was applied and maintained throughout the test through a four-strand external prestressing set-up powered with another hydraulic ram.
The instrumentation system is schematically shown in Figure 4. Besides the beam tip deflection, shear deformations in the joint zone and the beam curvatures at the root were measured. Some secondary measurements such as rigid body rotation etc. were also taken.
2.3. Test Results
A variety of diagrams were plotted using the test data obtained from the study of Gökdemir H. [5]. and these diagrams were interpreted as regards strength, strength degradation, ductility, energy dissipation, stiffness, stiffness degradation etc. Detailed information concerning the test results can be found Gökdemir H. [5]. A brief summary is given in Table 4.
The satisfactory performance criterion has been established as “a joint strong enough to withstand the loads leading to hinging in the beam”. A strengthening intervention was considered successful only if hinging (or at least, yielding of the flexural steel) took place in the beam. The remarks given in the table should be interpreted with this understanding.
In Table 4, four performance indicators namely, capacity ratio, ductility ratio, initial stiffness and flexural steel yielding are presented and overall performance evaluation is made based on these four criteria.
The capacity ratio is expressed in terms of the ratio of the measured maximum load to the load leading to yielding of the flexural steel in the beam. (Pmax / Py)
Ductility is expressed in terms of the ratio of the tip deflection corresponding to 60% of the maximum load on the descending branch, to the tip deflection at which significant change of slope is observed. The authors are quite aware that this is rather unusual since normally the yield deflection and the deflection corresponding to 85% of the maximum load are taken into consideration for the calculation of ductility. However, the chosen ratio is expected to serve the purpose in the absence of a well-defined yielding point. On the other hand, the immediate post-peak drop in the load observed in many tests encouraged the authors to use the deflection corresponding to a lower percentage of the maximum load instead of 85%.
The initial stiffness indicator given in Table 4 is again rather approximate. Since a gradual slope change is observed in all tests even at rather low load levels, an approximate average slope of the pre-yield portion of the envelope curve is used as the stiffness indicator.
In most of the tests, strain gauges were placed on the flexural steel at the root of the cantilever;
and flexural steel yielding was verified by strain gauge readings in these tests. In other tests, the width of the major flexural crack was monitored as an indication of yielding.
Beam-column joint zone in the experiment (S20BL) is shown in Figure 5.
Figure 5. Beam-column joint zone in the experiment (S20BL)
3. DISCUSSION
The envelope curves of the reversed cyclic load-tip deflection diagrams for all tests are presented in Figure 6 and 7 in two different combinations. In those envelope curves, the positive load and tip deflection values show that the tip load was applied upwards to downwards and the beam tip deflection was downwards; moreover, the negative load and tip deflection values show that the beam tip load was applied downwards to upwards and the beam tip deflection was upwards.
Figure 6 includes the reference tests (Series A); Figure 7 displays the curves for the specimens strengthened by steel bar implants (Series B), the latter two lines provide the lower and upper bound reference tests.
A careful and critical study of these diagrams together with the performance indicators given in Table 4 lead to some interesting observations summarised and briefly discussed in the following two sub-sections.
3.1. Series A: Reference Specimens
The performance of the joint without any ties (RL) is very unsatisfactory, as expected.
Both the capacity and behaviour improve as the amount of joint reinforcement increases.
However, the most interesting observation is the behaviour of the specimen (RC) in which 1.25 times the joint reinforcement recommended by the Turkish code is used.
The performance of this specimen is far from being satisfactory. An acceptable performance could be obtained by using twice of the joint reinforcement recommended by the Turkish code.
3.2. Series B: Specimens Strengthened with Steel Bar Implants
Diagonal steel bar implants improve the performance effectively, given the required amount of steel is placed with appropriate end anchorages.
Both bolted and welded bar ends serve the purpose satisfactorily, whereas epoxy embedded implants without any end anchorages are far less effective.
The effectiveness of the strengthening is almost directly proportional to the amount of steel embedded. The amount of diagonal steel implant in each of the two diagonal directions was equivalent to three times the amount of the code required joint reinforcement in specimens S20B & S20W which performed satisfactorily. 1/3 reduction in the amount of steel (S16B & S16BL) embedded led to almost 1/3 reduction in the capacity.
Lap splicing of the implanted diagonal bars appears to be perfectly acceptable.
Specimens S20BL and S16BL present similar performance to those of S20B and S16B respectively.
4. CONCLUSIONS AND RECOMMENDATIONS
The conclusions drawn within the limitations of the results obtained from eleven tests, performed on 2/3 scale specimens representing an external joint of a typical reinforced concrete framed building structure, are presented below. In the light of these conclusions, a
few preliminary proposals are also made concerning the implementation of the proposed joint strengthening technique in practice.
4.1. Conclusions
The inadequate performance of the specimen carrying 1.25 times the code required joint reinforcement casts a shadow on the code requirement. It is true that a few unfavourable conditions have been combined in the test specimens. The code recommended joint steel may be sufficient in the case of an interior joint or in the case of a joint surrounded by the beams of the other direction. Nevertheless, further experimental research of rather detailed nature seems necessary to justify or modify the related code clause.
The proposed technique for joint strengthening appears to be effective and practical.
Steel bars placed in the joint in two diagonal directions serve satisfactorily as shear reinforcement and improve the performance. However, one should realise that they do not completely substitute the joint reinforcement in the form of ties, which also have a second and very desirable effect of confining the concrete in the joint.
Steel bars can effectively be used in the proposed technique.
Steel bars require proper end anchorages.
Lap splices have been shown to perform equally effectively.
The test specimen was isolated from a "no-sway" prototype frame in the present pioneering work. Results would have been more assertive if test specimens could reflect the “sway” frame behavior, which would have required a far more sophisticated test setup. However, the conclusions of the present study appear to be quite satisfactory to indicate the adequacy of the proposed joint strengthening technique.
Since the bending capacity of the beam will increase with the addition of the T-beam reinforcement, the hinging in the beam will be difficult and the joint area strength will have to be even higher. However, as noted above, this preliminary work will provide useful inputs for the planned future experimental studies. It is aimed that these experiments will also contribute to the experimental work of the T-beams in the future.
4.2. Recommendations for Practice
Generally there are two possible approaches in the seismic strengthening of reinforced concrete framed structures; (i) strengthening of all deficient structural members one by one and (ii) system behaviour improvement by introducing new components to reduce the seismic loads that the major components receive, thus relieving the existing deficient members. The proposed technique is necessary and suitable for the former case, but not so essential in the latter.
A thorough seismic analysis of the structure to be rehabilitated indicates the beam- column joints in need of strengthening. These joints can effectively be strengthened by the proposed technique using steel bars.
In the design of steel bars, smaller number of larger size bars should be preferred for ease of application. As a rule of thumb, the tension capacity of the diagonal bars in one direction (i.e., cross-sectional area of the implanted bars, times the yield strength of steel) should be equal to three times the tension capacity in the diagonal direction of all the code recommended joint ties (i.e., number of ties in the joint, times two legs of each tie, times the cross-sectional area of the tie bar, times the yield strength of tie steel, times 0.7, approximately Cos 450, times 3.0).
The geometry of each one of the joints to be strengthened should be carefully studied and the existing steel bars should be located and marked using a steel detector.
In some cases, it may not be possible to drill holes running from a corner to the opposite corner due to the amount of the existing steel bars. In such cases, two parallel holes can be drilled from the two opposite corners to accommodate lap-spliced bars.
The attachments required for the bolted or welded end anchorages can be mass-produced prior to application. Furthermore, they necessitate smaller diameter holes easier to drill between the existing steel bars.
When the preparations are completed, the steel bars are threaded through the holes and epoxy is injected to the remaining space in the hole. Then the end anchorages are applied.
In the case of bolted ends, nuts should be tightened to a reasonable torque. In the case of welding, cooling is essential not to burn the epoxy.
References
[1] Antonopoulos, C. P., Triantafillou, T. C., Exp Investigation of FRP Strengthened RC Beam-Column Joints, Journal of Composites for Construction, 344(7), 39-49, 2003.
[2] Ghobarah, A., Said, A., Seismic Rehabilitation of Beam-Column Joints Using FRP Laminates, Journal of Earthquake Engineering, 5(1), 113-129, 2001.
[3] Prota, A., Nanni, A., Manfredi, G., Cosenza, E., Selective Upgrade of Under-designed RC Beam-Column Joints Using CFRP, ACI Structural Journal, 101(11), 699-707, 2004.
[4] Taslıgedik, A.Ş., “Lap Splice Behaviour of CFRP Rolls”, MSc thesis, Middle East Technical University, Ankara, Turkey, 2008.
[5] Gokdemir, H., “Seismic Strengthening of Beam-Column Joints”, PhD thesis, Eskisehir Osmangazi University, Eskisehir, Turkey, 2009.
[6] Seckin, M., “Hysteretic Behaviour of Cast-in-Place Exterior Beam-Column Sub- Assemblies”, PhD thesis, Toronto University, Toronto, Canada, 1981.
