10NCEE Tenth U.S. National Conference on Earthquake Engineering Frontiers of Earthquake Engineering July 21-25, 2014 Anchorage, Alaska STRENGTHENING WITH WING WALLS FOR SEISMICALLY SUBSTANDARD R/C BEAM-COLUMN JOINTS Yuebing Li 1 and Yasushi Sanada 2 ABSTRACT A large number of reinforced concrete (RC) buildings designed for gravity loads according to older seismic design codes exist even in regions of moderate seismicity, and some of them contain no transverse reinforcement in the beam-column oints. In consideration of economic and technical conditions in developing countries, a seismic strengthening method using RC wing walls is proposed for this kind of seismically substandard beam-column oint. This paper presents a design procedure for applying the proposed method to exterior beam-column oints. In this study, three 3/4-scale substandard exterior beam-column oint specimens were constructed with the same structural details and tested with a new experimental technique. Two of the specimens were strengthened by wing walls with different strengthening details. The test results showed that the strengthened specimens exhibited ductile failure modes with beam yielding, whereas the unstrengthened control specimen failed at the oint in a brittle manner. We consider the effectiveness of seismic strengthening, focusing on strength and ductility, and discuss the test results in terms of the design calculations presented. Our seismic strengthening approach is valid for improving the seismic performance of substandard RC beam-column oints. 1 PhD Candidate, Dept. of Global Architecture, Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita, Osaka 565-0871, Japan 2 Associate Professor, Dept. of Global Architecture, Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita, Osaka 565-0871, Japan Yuebing Li, Yasushi Sanada. Strengthening with wing walls for seismically substandard R/C beam-column oints. Proceedings of the 10 th National Conference on Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014.
10NCEE Tenth U.S. National Conference on Earthquake Engineering Frontiers of Earthquake Engineering July 21-25, 2014 Anchorage, Alaska Strengthening With Wing Walls For Seismically Substandard R/C Beam-Column Joints Yuebing Li 1 and Yasushi Sanada 2 ABSTRACT A large number of reinforced concrete (RC) buildings designed for gravity loads according to older seismic design codes exist even in regions of moderate seismicity, and some of them contain no transverse reinforcement in the beam-column oints. In consideration of economic and technical conditions in developing countries, a seismic strengthening method using RC wing walls is proposed for this kind of seismically substandard beam-column oint. This paper presents a design procedure for applying the proposed method to exterior beam-column oints. In this study, three 3/4-scale substandard exterior beam-column oint specimens were constructed with the same structural details and tested with a new experimental technique. Two of the specimens were strengthened by wing walls with different strengthening details. The test results showed that the strengthened specimens exhibited ductile failure modes with beam yielding, whereas the unstrengthened control specimen failed at the oint in a brittle manner. We consider the effectiveness of seismic strengthening, focusing on strength and ductility, and discuss the test results in terms of the design calculations presented. Our seismic strengthening approach is valid for improving the seismic performance of substandard RC beam-column oints. Introduction Reinforced concrete (RC) buildings that were designed according to older seismic design codes, or without complying with current seismic codes, containing no transverse reinforcement in the beam-column oint regions (called unreinforced oint ), still widely exist, particularly in buildings designed before the 1970s in the western U.S. and in other seismically active regions worldwide [1]. Model tests have shown that unreinforced oints possess lower strength, ductility, and energy dissipation [2], and shear failure in oints may cause such buildings to collapse, as observed in recent severe earthquake disasters. Effective and economic strengthening methods for such buildings are urgently needed. Several studies have examined seismic strengthening of unreinforced oints using steel props [3], GFRP [4], etc. and have achieved significant strengthening effects, but particularly for developing countries, these methods are not easily implemented due to technical level, available 1 PhD Candidate, Dept. of Global Architecture, Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita, Osaka 565-0871, Japan 2 Associate Professor, Dept. of Global Architecture, Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita, Osaka 565-0871, Japan Yuebing Li, Yasushi Sanada. Strengthening with wing walls for seismically substandard R/C beam-column oints. Proceedings of the 10 th National Conference on Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014.
materials, and construction cost. Installing wing walls beside columns seems to be a realistic way of upgrading these seismically substandard buildings. This paper describes a series of tests conducted to clarify the strengthening effects of installing wing walls, and discusses the results. Investigated Building and Joint The study focused on an exterior beam-column oint of a three-story RC moment-resisting frame structure which collapsed in the 2009 Sumatra Earthquake, as shown in Fig. 1. It was built in 2005 with a floor height of 3,000 mm and span length of 7,000 mm. In this building, damage such as buckling of longitudinal bars and concrete spalling was concentrated in the oints. According to a field survey [5], there was no transverse reinforcement in the oints, as shown in Fig. 2. Figure 1. Earthquake-damaged building and beam-column oint [2]. Figure 2. Dimensions and reinforcing details of the structure [5]. Proposed Strengthening Design Seismic Behavior of Exterior Beam-Column Joint In a previous study [2], a small-scale (1/3) model test of the same structure was conducted and the deformation behavior was found to be as follows. One crack along the diagonal line of the oint panel (called diagonal crack ) and the other extending from the tensile corner of the beam and column to the center of the diagonal crack (called corner crack ) appeared under seismic loading in the positive/negative directions, as shown in Fig. 3. Each diagonal crack and corner crack widened with rotation at A, B, C/D, E, F under positive/negative loading, respectively. Concept of Seismic Strengthening Method Shiohara [6] proposed a resisting model for beam-column oints called a nine-parameter model and defined that the oint strength depends on its moment capacity (M u ), which is the total moment acting on the oint center (denoted by O in Fig. 4), mainly produced by tension of the reinforcements crossing the cracks of the oint and compression of concrete around the rotation points in Fig. 3. Kusuhara and Shiohara [7] developed this model and formulas for calculating
the moment resistance of exterior oints considering their details, and found good agreement with experimental results. Figure 3. Deformation behavior of exterior oint. Figure 4. Strengthening concept. Seismic bending moments around the oint panel (M b, M c ), as shown in Fig. 4, induce a diagonal crack and a corner crack at the oint, and rotations of the divided three parts (upper column, lower column and beam). It is conceivable that, when the seismic moment at the oint center (M b or 2M c, called oint moment ) exceeds the moment capacity, the oint will fail. It is expected that, by installing wing walls to the columns, the tension ( b T ay, c T ay ) of the anchor bolts that are used to connect the pulled wing wall to the existing structure, and the compression (N) of the strut in the pushed wall, will increase the moment capacity of the existing oint, thus preventing brittle failure of the oint and shifting it to a beam flexural hinging mechanism. For designing the wing walls, the following assumptions are made: 1) The moment capacity of the existing oint is known. This study refers to the previous test conducted by the authors [8]. The capacity of the strengthened oint is set to be higher than the oint moment when the beam yields at the end of the wing wall ( by M ). The difference between both is the required strengthening. 2) Firstly, the wing walls are individually designed for the pulled/pushed side because of the unclear contribution ratio, even though they seem to act together. The final strengthening scheme is adopted as the safer one for which more strengthening is needed, and is symmetrically applied to both sides. 3) The width of the strut along the edge of the beam (L c ) is half the width of the wing wall (b w ). The angle (α) between the strut and column is 45 o. 4) The amount of anchor bolt groups installed in the beam and column are equivalent, and their centroids are at the middle of the wing wall.
Based on the above assumptions, the design formulas are as follows: Strengthening design for the pulled wing wall: T b ay b L + T L M M (1) a c ay c Strengthening design for the pushed wing wall: P by a by ex ex N cos α L M M (2) Symbols in the formulas are stated above or shown in Fig. 4. Specimens Experimental Program Three 3/4-scale specimens of the exterior beam-column oint shown in Figs. 1 and 2 were prepared. They were modeled up to the inflection points of the upper/lower column and beam, however, pin supports attached to the column ends and a loading method described later were considered. Fig. 5 shows the dimensions and reinforcement details. One of them was the control specimen, J2, representing the existing frame. The other two were strengthened by installing wing walls; J2-W2 was strengthened at both sides, and J2-W1 only at one side. Figure 5. Dimensions and reinforcement details of the existing part of the specimens. The wing walls were designed as stated above. In this study, however, when determining the width of the wing wall, the effective width of the pulled wall for the partial specimen was assumed to be e b w in Fig. 4. This is equivalent to the length from the column interior surface to the intersection of the line extending from the corner crack and the upper edge of the specimen, and was 328mm. Using Eq. 1 for the pulled wall, the area of beam and column anchors should each be more than 330mm 2, when their yield strength is 295N/mm 2. Using Eq. 2 for the pushed wall, the thickness of the wing wall should be not less than 114mm when concrete compressive strength is 22N/mm 2. According to the Japanese guideline for retrofitting RC buildings [9],
however, the minimum thickness limit of the wing wall should be not less than 150mm, considering the scale of 3/4. Moreover, the width of wing walls was eventually increased to 340mm, because of the Japanese code restrictions on the minimum spacing of anchors and the minimum thickness of concrete cover. Fig. 6 shows the details of wing walls. Figure 6. Strengthened specimens. Figure 7. Validation of strengths of anchors. J2-W2 was strengthened at both sides of the beam, where wing walls were symmetrically installed to the upper and lower columns. To confirm the contribution of the pulled/pushed wing wall, only one wall was installed to the lower column for J2-W1. Two weeks after the existing part was constructed, anchor bolts were embedded and then concrete was cast for the wing walls. The tested cylindrical compressive strength of the concrete and the properties of the reinforcement and anchor bolts are shown in Table 1. Table 1. Properties of concrete, reinforcement, and anchor bolts (N/mm 2 ). J2 Region Ÿ E c 2.55 10 4 F c 20.2 f t 1.9 D16 E s 1.75 10 5 F y 373 T u 529 J2-W2 Existing 2.57 10 4 22.7 2.0 Φ9 1.78 10 5 344 455 wall 2.62 10 4 26.9 2.5 D10 1.68 10 5 380 554 J2-W1 Existing 2.80 10 4 22.6 2.1 D13 1.65 10 5 361 523 wall 2.62 10 4 27.7 2.4 where, E c : Young s modulus of concrete, F c : compressive strength of concrete, f t : tensile strength of concrete, E s : Young s modulus of steel, F y : yield strength of reinforcement, and T u : ultimate tensile strength of reinforcement. The properties of the concrete and reinforcement are similar to those of the studied building [2].
Strength Evaluation for Existing Frame, J2 The strengths of the members of J2 were calculated according to the Japanese standard: flexural strengths of the beam and column by Eqs. 3 and 4, respectively [10], and shear strength of the oint by Eq. 5 [11]. The shear strengths of the beam and column are considerably larger than their flexural strengths, and are not given here. b c M u = 0.9a σ d (3) t y N M u = 0.8atσ y D + 0.5ND 1 (4) bdfc V u = κ φ F b D (5) where, a t : area of tensile bars, σ y : yield strength of tensile bar, d: effective depth of beam, D: depth of column, N: axial force, b: width of the member, F c : compressive strength of concrete, κ: oint shape factor (0.7 for exterior oint), φ: factor depending on existence of orthogonal beams (0.85 for oint without it), F : nominal value of shear strength for oint (F = 0.8F c 0.7 ), b : effective width (= b, when beam and columns have the same width), and D : effective depth of oint. By calculations, the ultimate strengths are 134kN.m (142kN.m), 71kN.m (177kN.m), and 278kN (124kN.m) for the beam, column, and oint, respectively. Values in parentheses are the conversions to oint moment. The strength of oint is the minimum, meaning that oint failure will occur first for the existing frame. Validation of Strength of Anchor Bolt Groups Post-installed adhesive anchor bolts were used in this study. When an adhesive anchor bolt is pulled, it may fail in three patterns: fracture of anchor, bond failure or concrete cone failure. The minimum is taken as its tensile strength. However, when anchor bolts are close to one another, the existing concrete may fail as shown in Fig. 7. According to the Japanese guideline [9], the tensile strengths of anchor bolt groups, together with their conversions to oint moment and the required strengthening (referring to Eq. 1), are shown in Table 2. The shear strengths and their corresponding requirements are also given in the table. As shown in Fig. 7, the shear strengthening requirements are considered as the shear forces on the beam and column (V b, V c ) when the beam yields, and are resisted by the column and beam anchor groups, respectively. As shown in Table 2, strengths of the anchor groups were confirmed. Table 2. Strengths of anchor bolt groups and the corresponding design loads. Tensile Shear Anchor Place Strength Total moment on oint Requirement Strength Requirement bolts kn kn.m kn.m kn kn Beam 7-D10 135.9 102.9 54.0 135.4 76.0 Column 5-D13 190.3 131.0 63.1
Test Set-up, Instrumentation, and Loading Program Fig. 8 shows the test set-up, and Fig. 9 shows the strain gage arrangement. The specimens were installed on the loading facilities, rotated by 90 o. The left column (lower floor column) was supported by a pin hinge, and the right was supported by a roller. In order to measure the shear force of the column, a load cell was incorporated into the roller support. Horizontal reversed cyclic loading was applied to the beam tip controlled by the displacement of the beam tip. Meanwhile, an additional moment in proportion to the horizontal load was applied by two vertical hydraulic acks, to represent a realistic moment distribution of the full-scale beam. The loads of the vertical acks were controlled by Eq. 6. Axial load was not applied to the column. N ± ( 2.625 1.7) Q / 3.5 = ± 0. 264 Q = (6) Figure 8. Test set-up and loading method. Figure 9. Strain gage arrangement. Beam drift ratio is defined as R = δ/l, where δ is the horizontal displacement at the beam tip measured by a displacement transducer, and L is the distance between the transducer and oint center, as shown in Fig. 8. The loading program was one cycle for R = 1/800, 1/400, 1/200, 1/133, 1/100, 1/67, 1/50, 1/33, and 1/25 rad., followed by a pushover to R = +1/17. Experimental Results Fig. 10 shows damage to the specimens after the cycles to R = 1/200, and 1/67 rad., during which the oint diagonal crack appeared and the ultimate strength was recorded, respectively, for the control specimen, together with the final damage. The relationships between oint moment and beam drift ratio together with the maximum strength are also given in the same figure. The oint moments are the product of the distance between pin centers and the shear force of the column. Control specimen, J2 Diagonal cracks appeared at the oint panel during the cycle to R = ±1/200, as shown in Fig. 10, then extended along the external longitudinal column bars with increasing plastic deformation. The maximum strength was recorded at R = ±1/67. During the cycle to R = ±1/25, column
longitudinal bars became exposed and their buckling was observed. After that, the diagonal cracks pierced the outside of the column, and a substantial amount of concrete peeled off. Figure 10. Damage to specimens, oint moment beam deflection relationships. Specimen strengthened at two sides, J2-W2 During the cycle to R = ±1/200, the first beam anchor bolt (ABU/L4, referring to Figs. 6 and 9) yielded, and diagonal cracks appeared at the beam end where the walls were attached. During the cycle to R = ±1/133, the beam longitudinal reinforcement yielded at the location where the first anchor bolt was inserted (BU/L4), and diagonal cracks appeared at the existing oint panel. During the cycle to R = +1/17, the longitudinal reinforcement of the beam became exposed and buckling was observed. The damage concentrated at the beam where the wing walls ended. Obvious damage to the wing walls was not observed. Specimen only strengthened at one side, J2-W1 The main characteristic for J2-W1 was the asymmetry between positive and negative loadings. During the cycle to R = +1/200, a diagonal crack appeared from the existing oint to the part of the beam where the wing wall was attached. During the cycle to R = ±1/133, diagonal cracks appeared at the existing oint panel, and in the positive loading, the first beam anchor (ABL4) yielded. During the cycle to R = +1/100, the third beam anchor (ABL3) yielded, then the beam reinforcement (BL4) yielded at the location where the first anchor was placed, and a crack appeared along the beam depth where beam anchors were buried, as shown in Fig. 11. It seemed to be the start of cone failure of concrete, as shown in Fig. 7. During the cycle to R = -1/50, when the wing wall was pushed, shear cracks appeared at the wing wall, at an angle of about 45 o to the
column, as shown in Fig. 11. During the cycle to R = +1/17, the column reinforcement became exposed and buckling was observed. The maximum strengths between positive and negative loading, when the wall was pulled or pushed, varied widely. The damage was also asymmetric. Discussion The skeleton curves of the hysteresis loops, the strengths considering beam yielding and oint failure based on Eq. 3 and 5, and the ratios of the maximum strength of strengthened specimens to the control specimen are illustrated in Fig. 12. J2 behaved in brittle failure mode, showing low strength and deformability. Its ultimate strength was obviously below the value calculated according to the Japanese standard [10] (±124, depending on the oint strength). The failure mode of F2-W2 successfully shifted from brittle oint failure to ductile beam yielding, showing high strength, exceeding the beam yield strength (±164, assuming yield at wall ends), and good deformability. For J2-W1, when the wing wall was pushed, the strength was improved by 81%, about the same as J2-W2. When it was pulled, the strength was improved by 39% compared with J2, reaching the beam yield strength (±142, assuming yielding at column face) of the existing frame, but less than the yield strength at the wing wall end (±164). The reason seemed to be cone failure of concrete where beam anchors were buried, as shown in Fig. 11, although the strengthening capacity was validated as described above. Because the anchors further from the oint center carried a larger tension, the anchor bolts did not share the tension equally as had been assumed. Figure 11. Cone-style failure, and 45 o cracks at wing wall for J2-W1. Joint Moment (kn.m) 250 200 1.73MJ2(+) 164 150 1.39MJ2(+) 142 100 124 MJ2(+) 50 0-50 J2 J2-W2 J2-W1 MJ2(-) Ultimate strength -100 Calculations -150 Joint strength -200 1.85MJ2(-) 1.81MJ2(-) Beam yielding (at column face) Beam yielding (at wall end) -250-4 -3-2 -1 0 1 2 3 Beam Drift Angle (%rad.) 4 5 6 Figure 12. Skeletons, calculated strengths, and strengthening effect. Conclusions This paper proposed and verified a strengthening method using RC wing walls for exterior beamcolumn oints where shear reinforcement was not provided. Three partial frame specimens were tested. The maor findings were as follows:
1. In the case of the control specimen, damage concentrated at the oint, showing a brittle failure mode. The maximum strength was less than the calculated design value. 2. For the specimen strengthened by installing wing walls to both upper and lower columns, the failure mode shifted from brittle oint failure to ductile beam yielding. The strength and deformability were significantly improved. It was verified that the strengthening method effectively inhibited premature failure of the oint. 3. For the specimen strengthened on one side, the strengthening effect when the wing wall was pushed was greater than when the wall was pulled, although the strengthening capacity by the tension of the anchor bolt groups had been validated when designing. 4. The contribution of strengthening by pulling and pushing for the oint strengthened on two sides needs further examination. References 1. Park S, Mosalam KM, Shear strength models of exterior beam-column oints without transverse reinforcement. Pacific Earthquake Engineering Research Center Report 2009; 2009/106. 2. Sashima Y, Nitta Y, Tomonaga T, Sanada Y, Seismic Loading Test on an R/C Exterior Beam-Column Joint without Shear Reinforcements in Indonesia. Proceedings of the Thirteenth Taiwan-Japan-Korea Joint Seminar on Earthquake Engineering for Building Structures 2011; pp. 68-77. 3. Sharbatdar MK, Kheyroddin A, Emami E, Cyclic performance of retrofitted reinforced concrete beam column oints using steel prop. Construction and Building Materials 2012; Vol. 36, pp. 287-294. 4. Ghobarah A, Said A, Shear strengthening of beam-column oints. Engineering Structures 2002; Vol. 24, pp. 881-888. 5. Sanada Y, Kishimoto I, Kuroki M, Sakashita M, Choi H, Tani M, Hosono Y, Fauzan, Musalamah S, Farida F, Preliminary Report on Damage to Buildings due to the September 2 and 30, 2009 Earthquakes in Indonesia. Proceedings of the Eleventh Taiwan-Korea-Japan Joint Seminar on Earthquake Engineering for Building Structures 2009; pp. 297-306. 6. Shiohara H, New Model for Shear Failure of RC Interior Beam-column Connections. Journal of Structural Engineering 2001; Vol. 127, Issue 2, pp. 152-160. 7. Kusuhara F, Shiohara H, Ultimate moment of reinforced concrete exterior beam-column oint. Journal of Structural and Construction Engineering (Transactions of AIJ) 2013; Vol. 78(693), pp. 1949-1958. (in Japanese) 8. Sanada Y, Tomonaga T, Li Y, Watanabe Y, Behavior of an R/C Exterior Beam-Column Joint without Concrete Confinement under Seismic Loading. Proceedings of the Seventh International Conference on Concrete under Severe Conditions Environment and Loading 2013; Vol. 2, pp. 1598-1606. 9. The Japan Building Disaster Prevention Association (JBDPA), Guidelines for seismic retrofit of existing reinforced concrete buildings, (2001). 10. The Japan Building Disaster Prevention Association (JBDPA), Standard for seismic evaluation of existing reinforce concrete buildings, (2001). 11. Architectural Institute of Japan (AIJ), Design Guidelines for Earthquake Resistant Reinforce Concrete Buildings Based on Inelastic Displacement Concept, (1999). (in Japanese)