EFFECTS OF END REGION CONFINEMENT ON SEISMIC PERFORMANCE OF RC CANTILEVER WALLS

Transcription

1 10NCEE Tenth U.S. National Conference on Earthquake Engineering Frontiers of Earthquake Engineering July 21-25, 2014 Anchorage, Alaska EFFECTS OF END REGION CONFINEMENT ON SEISMIC PERFORMANCE OF RC CANTILEVER WALLS R. Taleb 1, S. Kono 2, M. Tani 3 and M. Sakashita 4 ABSTRACT The main objective of this study is to investigate the effects of end region detailing of reinforced concrete (RC) cantilever walls on their seismic performance. An experimental study was conducted on seven -scale cantilever type RC walls designed to fail in flexure under cyclic reversed loading. Considered walls included two walls with barbell shape section and five walls with rectangular cross-sections and having different transverse reinforcement ratio at their end regions. Primary test variables were sectional shape (rectangular and barbell shape), transverse reinforcement ratio in confined end regions, shear span ratio, and axial load ratio. For rectangular walls, concrete crushing spread widely over the plastic hinge region with buckling of longitudinal reinforcement at the final loading stage, while for walls with barbell shape, crushing of concrete was essentially limited within boundary columns but leading to a more brittle failure than that of rectangular walls. Walls with barbell-shape showed the efficiency of boundary columns in increasing deformation capacity and reducing damage level in a wall panel. Test results also made clear that end regions should be well confined when a structural wall, especially rectangular walls, is expected to sustain large drift capacity. It was noted that contribution of flexural deformation was clearly dominant and constantly as high as 7 of the total deformation and that more than 7 of the flexural deformation after yielding was concentrated at the lower wall portion. It was also shown that the damage region is limited in height and tends to spread in a horizontal direction toward wall center rather than a vertical direction. It was concluded that transverse reinforcement detailing in confined end region is a key parameter that controls the global buckling of compression zone and/or buckling of longitudinal reinforcement at ultimate drift. 1 Doctoral Student, Department of Environmental Sci. & Tech., Tokyo Institute of Technology, Yakohama, Japan 2 Professor, Structural Engineering Research Center, Tokyo Institute of Technology, Yakohama, Japan 3 Research Engineer, Building Research Institute, Tsukuba, Japan 4 Assistant Professor, Department of Architecture and Architectural Engineering, Kyoto University, Kyoto, Japan Taleb R, Kono S, Tani M, Sakashita M. Effects of end regions confinement on seismic performance of RC cantilever walls. Proceedings of the 10 th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014.

2 10NCEE Tenth U.S. National Conference on Earthquake Engineering Frontiers of Earthquake Engineering July 21-25, 2014 Anchorage, Alaska Effects of End Regions Confinement on Seismic Performance of RC Cantilever Walls R. Taleb 1, S. Kono 2, M. Tani 3 and M. Sakashita 4 ABSTRACT The main objective of this study is to investigate the effects of end region detailing of reinforced concrete (RC) cantilever walls on their seismic performance. An experimental study was conducted on seven -scale cantilever type RC walls designed to fail in flexure under cyclic reversed loading. Considered walls included two walls with barbell shape section and five walls with rectangular cross-sections and having different transverse reinforcement ratio at their end regions. Primary test variables were sectional shape (rectangular and barbell shape), transverse reinforcement ratio in confined end regions, shear span ratio, and axial load ratio. For rectangular walls, concrete crushing spread widely over the plastic hinge region with buckling of longitudinal reinforcement at the final loading stage, while for walls with barbell shape, crushing of concrete was essentially limited within boundary columns but leading to a more brittle failure than that of rectangular walls. Walls with barbell-shape showed the efficiency of boundary columns in increasing deformation capacity and reducing damage level in a wall panel. Test results also made clear that end regions should be well confined when a structural wall, especially rectangular walls, is expected to sustain large drift capacity. It was noted that contribution of flexural deformation was clearly dominant and constantly as high as 7 of the total deformation and that more than 7 of the flexural deformation after yielding was concentrated at the lower wall portion. It was also shown that the damage region is limited in height and tends to spread in a horizontal direction toward wall center rather than a vertical direction. It was concluded that transverse reinforcement detailing in confined end region is a key parameter that controls the global buckling of compression zone and/or buckling of longitudinal reinforcement at ultimate drift. Introduction Reinforced concrete structural walls are commonly used lateral-load resisting components in multi-story building structures. When well designed and detailed, walls are considered to perform well under earthquake loading by providing substantial lateral stiffness and strength to effectively resist lateral loads and reduce damage in non-structural elements by limiting lateral deformations. When large deformation capacity is required to withstand large earthquakes, the compression zone in the bottom part of the end regions should be well confined. The behavior of properly designed walls with aspect ratios more than 2.0 is generally dominated by flexure, while 1 Doctoral Student, Department of Environmental Sci. & Tech., Tokyo Institute of Technology, Yakohama, Japan 2 Professor, Structural Engineering Research Center, Tokyo Institute of Technology, Yakohama, Japan 3 Research Engineer, Building Research Institute, Tsukuba, Japan 4 Assistant Professor, Department of Architecture and Architectural Engineering, Kyoto University, Kyoto, Japan Taleb R, Kono S, Tani M, Sakashita M. Effects of end regions confinement on seismic performance of RC cantilever walls. Proceedings of the 10 th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014.

3 that of walls with aspect ratios less than 1.0 is dominated by shear. For walls with intermediate aspect ratio, the behavior is usually dominated by an interaction between shear and flexure. If shear behavior is significant, walls may have reduced strength, stiffness and deformability. Several experimental studies have been conducted to investigate the ultimate behavior of cantilever RC walls [ 1, 2, 3, 4]. However, modern architectural requirement pushed design engineers to produce slender walls with higher load and drift capacities than have been well verified in past studies. Observed damages of RC wall buildings in recent earthquakes in Chile and New Zealand raised concerns about the seismic performance of rectangular RC walls. In these earthquakes, severe damage happened to concrete walls in numerous walled buildings leading to partial or total collapse. Structural wall damage included spalling and crushing of concrete which often spread over more than half of the wall length, fracture under tension or buckling under compression of vertical reinforcement at end regions, and global wall buckling [ 5]. It was reported that lack of adequate confinement and detailing in end regions was one of the main causes of their damages, suggesting that more studies are needed to examine their seismic performance [ 6]. Structural walls in Japan usually have boundary columns and beams to provide confinement to wall panels and to allow large amount of axial force to be carried by boundary columns. However, the 2010 AIJ standard [ 7] allows the use of rectangular cross section walls with confined end regions. RC structural walls with boundary columns provide confinement to wall panels and are usually considered more effective than rectangular walls with confined end regions. A research program was undertaken in order to study the effects of end region confinement on the seismic performance of cantilever structural walls. Seven -scale cantilever type structural walls having different cross sectional configurations and transverse reinforcement at the end regions of the walls were constructed and tested under lateral cyclic reversed loading. The test specimens included two specimens with boundary columns and five specimens with rectangular shape section. Primary test variables included cross sectional shape (rectangular and barbell shape), transverse reinforcement ratio in confined end regions, shear span ratio, and axial load ratio. Description of Test Specimens Experimental Program Seven scale RC cantilever walls having different section configurations and amount of transverse reinforcement in confined end regions were constructed and tested under lateral cyclic reversed load. Test variables included sectional shape (barbell-shape and rectangular sections), shear span ratio (1.71 and 1.37), axial load level (0.10Agf'c and 0.20Agf'c), and transverse reinforcement ratio in confined end regions. Reinforcement details of the tested wall specimens are shown in Figure 1. Wall specimens BC40 and BC80 had boundary columns while other wall specimens had rectangular sectional shape with confined end regions. All specimens had same length (1,750mm). Wall thickness of rectangular wall were 120 mm or 128 mm for 1.71 and 1.37 shear span ratio walls, respectively. However, the barbell-shaped wall specimens (BC s specimens) had a wall panel with a thickness of 80 mm and boundary columns at its both ends with a cross-section of 250 mm 250 mm. Lateral load was applied at either 3000 mm or

4 2 mm above the wall-foundation interface. BC s and NC s wall specimens were designed to have nearly same total area (2,250 cm 2 for BC's and 2,240 cm 2 for NC's) and confined end region area (625 cm 2 for BC's and 666 cm 2 for NC's). The rectangular walls with 1.37 shear span ratio were tested to study the effect of confined area and axial load on the ultimate deformation. SC specimen had the smallest confined area (180 cm 2 ) with a small confining hoop spacing (40 mm) and MC specimen had larger confined area (300 cm 2 ) with 80 mm hoop spacing. HN specimen, which was tested under higher axial load ratio, had a large confined area (540 cm 2 ) and with 40 mm hoop spacing. Reinforcing bars D10 were used for longitudinal reinforcement in confined end regions while reinforcing bars D6 were used for transverse reinforcement and web horizontal and vertical reinforcement for BC s and NC s wall specimens, and D4 for MC, NC and HN walls. Figure 2 shows examples of the vertical reinforcement layouts of NC40, NC80 and MC wall specimens. The specimens were cast in two phases: first the foundation and then the wall and the loading beam as one part for the second phase. Table 1 provides the measured concrete and reinforcement properties. An axial load of approximately 0.10A g f' c was applied to all specimens, except HN specimen which sustained an axial load level of 0.20A g f' c. This axial load was applied to the wall specimens at the beginning of each test and kept constant throughout the test to represent the action of dead load. All specimens were designed to fail in flexure; the shear-to-flexural-capacity ratios vary from 1.1 to 1.5, where the flexural and shear capacities were calculated based on the AIJ standard. The detailing of hoops in end regions of rectangular walls satisfied the ACI [ 8] requirements. 1. BC80 2. BC40 D6@80 D6@80 D6@40 D6@ D D D D NC80 4. NC40 D6@80 D6@80 D6@40 D6@40 12D D10 12D D MC 6. SC D4@80 D4@80 D4@40 D4@ x x HN D4@40 D4@ D10 16x50 Figure 1. Reinforcement details of the tested wall specimens D

5 275 B NC40 NC80 B B-B Figure 2. Vertical reinforcement layout of NC40, NC80 and MC specimens. (Unit: mm, 1 in. = 25.4 mm.) Table 1. Properties of wall specimens. h w /l w (mm) a s (Shear span) End regions Wall panel No. Specimen A ch t (mm 2 p w p wh = ) l (mm) p wv N/A g f' c 1 BC80 2 BC % NC80 (2/) (3000 mm) 4 NC % % MC % SC (1700/) % 0.10 (2 mm) 7 HN % 0.20 Notes: h w /l w is the aspect ratio; a s is the shear span ratio; A ch is the area of confined end region; p l is the boundary longitudinal reinforcement ratio; t w is the wall panel thickness; p wh and p wv are the horizontal and vertical web reinforcement ratio, respectively; N/A g f' c is the axial load ratio. (1 in. = 25.4 mm.) Table 2. Measured mechanical properties of concrete and reinforcement. Concrete Reinforcement Specimen Compressive Young s Splitting Reinf. Yield Young s Tensile strength modulus strength bars strength modulus strength (MPa) (GPa) (MPa) (MPa) (GPa) (MPa) BC80/BC D NC80/NC SC 27.5 MC 29.6 HN D D D

6 Test Setup, Instrumentationn and Procedure Wall specimens were instrumented by LVDTs and strain gages to measure wall displacements and reinforcement strains. The shear-span ratio 1.71 walls were subdivided into four zones (Z1, Z2, Z3 and Z4) in the vertical direction for a separate measurement of the contributions of shear and flexure to the total displacement and the shear-span span ratio 1.37 walls were divided into three zones (Z1, Z2 and Z3). Reinforcement strains were measured in different locations on the longitudinal and transverse reinforcement of end regions as well as vertical and horizontal web reinforcement. Figure 3 shows the loading setups and LVDTs layouts. Lateral load was applied using a displacement-controlled led reverse cyclic load protocol. Each load increment was repeated two times at drift ratios (top horizontal displacement divided by height of lateral load application point) of 0.05%, 0.1%, 0.25% %, 0.5%, 0.75%, 1., 1.5%, 2% and 4%. 2 x 2 MN Actuators 1 MN Actuator 1 MN Actuator 1 MN Actuator 2 MN Actuator 1 MN Actuator Z4 Z3 Z2 Z Z3 Z2 Z (a) BC s and NC s walls (b) MC, SC and HN walls Figure 3. Test setup and LVDTs layouts. (Dimensions in mm,) Test Results and Discussions Damages, Hysteretic Behavior and failure modes At 0.05% drift ratio, flexural cracks started to appear in the lower part of the tensile region. The number of flexural cracks increased along the confined region and progressed into flexural-shear cracks at drift ratio of 0.5% with the yielding of tensile longitudinal reinforcement. For specimens with boundary columns, flexural cracks and shear cracks were not necessarily continuous at the column-wall panel interface. As drift ratio increased, these cracks increased and propagated to the upper part and to the center of the walls. Spacing between flexural cracks was larger for HN specimen under high axial load level compared to that of other specimens. Although the wall specimens behaved generally in a flexural manner by yielding of the longitudinal reinforcement and suppressing of a premature shear failure, these specimens failed finally by concrete crushing of the compression zone with buckling of longitudinal reinforcement in confined end regions after the drift ratio of 1.5%. Global buckling of damaged regions under

7 compression was more pronounced for HN specimen at final loading stage. Figure 4 shows the damage situation at 2% drift ratio of BCs and NCs walls and at 1.5% drift ratio of MC, SC and HN walls. The failure was brittle at final stage since the core concrete crushed in a brittle manner. Crushing was more brittle and happened mainly around the boundary columns for BC40 and BC80, while extended to the wall center for rectangular section walls. As expected, the performancee of wall with boundary columns was better than that of rectangular walls with similar shear-span ratio in terms of drift capacity. Boundary columns also showed the ability to reduce damage level in wall panel since they carry a large amount of axial force and reduce axial stress level in wall panels. In this manner, boundary columns can also contribute effectively in preventing failure mode due to global wall buckling when subjected to high axial load level. Damage observation for specimens with larger hoop spacing revealed that the two outer longitudinal reinforcement buckled first followed by simultaneous buckling of the other longitudinal reinforcement at the final stage. The damage in confined region under compression for all specimens seems to have concentrated within approximately 30 cm height above the wall base, whereas extendedd horizontally up to the center. MC and SC walls have almost same volumetric transverse reinforcement content in confined end regions. However, SC wall with narrow confined core and closer hoop spacing had similar or better performance than specimens with wider confined core with larger hoop spacing since core concrete crushing under compression. Failure might be accelerated by longitudinal reinforcement buckling, suggesting that amount and spacing of transverse reinforcement might be a key parameter when assessing detailing requirements for a large drift capacity. NC80 NC40 BC80 BC40 (a) 1.71 shear span ratio walls at 2% drift ratio (NC80, NC40, BC80 and BC40, respectively) MC SC HN (b) 1.37 shear span ratio walls at 1.5% drift ratio (MC, SC and HN, respectively) Figure 4. Damage patterns. Figure 5 shows lateral load-drift ratio relations for the tested wall specimens. All specimens yielded in flexure, reached the peak point, and deformed until the failure without too much degradation of lateral load carrying capacity. They showed ductile inelastic behavior after flexural yielding. and in the figure are calculated flexural capacity and ultimate flexural deformation based on AIJ guidelines [ 9], and exp is the experimental ultimate deformation.

8 Table 3 summarizes test results for lateral load and corresponding drift ratio at concrete cracking, yielding of longitudinal reinforcement in confined regions, peak point as well as ultimate deformation point for both positive and negative loading directions. The ultimate point was defined by either -drop of the peak load or the maximum drift reached during the loading process. Hysteresis curves of the shear-span ratio 1.71 walls showed pinching loops due to the high axial load, high concrete strength and the low longitudinal reinforcement content in confined end regions Lateral load (kn) Lateral load (kn) exp Q exp -600 m -600 BC80 NC80 BC40 NC Drift angle (%) Drift angle (%) R R -200 u -200 u - - exp exp SC MC Drift ratio (%) Drift ratio (%) Figure 5. Lateral load (kn) 600 Lateral load (kn) Lateral load-drift ratio hysteresis curves. Lateral load (kn) exp HN Drift ratio (%) Table 3. Experimental characteristic points. Cracking Yielding Peak Ultimate Specimen Q cr R cr Q y R y ax R Qmax (kn) (%) (kn) (%) (kn) (%) (%) NC80 Positive Negative NC40 Positive Negative BC80 Positive Negative BC40 Positive Negative MC Positive Negative SC Positive Negative HN Positive Negative

9 Flexural and Shear Deformations Although the shear force in a cantilever wall subjected to a horizontal top load was constant over the height of the wall, the shear deformation was not uniform after concrete cracking and reinforcement yielding. Figure 6 shows the contribution of flexure and shear deformations as percentage of the total lateral drift for four segments of the 1.71 shear-span ratio walls and three segments of the 1.37 shear-span ratio walls. Zone Z0 is the lower 50 mm region which had vertical and horizontal displacement gages to measure pullout of vertical reinforcement and sliding along the joint between the wall and foundation, respectively. The flexural contribution was clearly dominant and constantly as high as 7 for the 1.71 shear-span ratio rectangular walls while it slightly decreased from approximately 7 to 6 after yielding for 1.37 shearspan ratio rectangular walls. Furthermore, more than 7 of the flexural deformation after yielding concentrated at lower zone (Z1). Similarly, the shear contribution concentrated in the lower part, where the longitudinal reinforcement yielded, remained approximately constant for all peak drift ratio in the inelastic range for the 1.71 shear-span ratio walls, while it gradually increased for the lower shear-span ratio walls NC NC BC80 BC MC SC HN Figure 6. Variation of flexural and shear deformations with top drift ratio.

10 For walls with boundary column, it is noted that contribution of shear deformation was large before yielding since the flexural deformation changed from about % before yielding to approximately 7 for larger drift ratios. On the other hand, sliding at the wall-foundation interface was negligible in all specimens, whereas contribution of deformation due to pull-out of longitudinal reinforcement was more significant for rectangular walls especially for lower shearspan ratio walls. The contribution of deformation due to pull-out of longitudinal reinforcement was about 18% for larger shear-span ratio walls and about 25% for other rectangular walls while it was lower than 1 for walls with boundary columns. Strain distribution in confined regions Confined regions of MC, SC and HN walls were heavily instrumented by strain gages attached to longitudinal and transverse reinforcement over a height of about half of the clear wall height from the foundation-wall interface. Strain gages on transverse reinforcement were attached at 80 mm from the wall base and then at every 160 mm above. Only transverse reinforcement at 80mm and at 240 mm height yielded in end regions under compression. Figure 7 shows strain distribution in transverse reinforcements parallel to the wall thickness at 80 mm above the wall base. Figure 8 shows strain distribution in the outermost edge longitudinal reinforcement. The ordinate shows the distance from the center of the wall. Strain evolution in confined regions shows that yielding initiate earlier for HN wall under high axial load compared to MC and SC walls. It is also shown that the damage region was limited in height and tended to spread more horizontally than vertically. Takahashi et al. [ 10] proposed a plastic hinge zone length of 2.5 times the wall thickness, and that the height of confinement may be limited to 3 times the wall thickness if the compressive strain is not exceeding Figure 7. Strain distribution in transverse reinforcement parallel to wall thickness Figure 8. Yield strain Strain distribution in the outermost edge longitudinal reinforcement. Conclusions This paper presents experimental results of seven cantilever RC walls having different cross-

11 section shapes, shear-span ratios and the amount of confining transverse reinforcement at their end-regions. They were tested under quasi-static cyclic loading to see their ultimate deformation capability and to investigate the effects of end regions detailing of reinforced concrete cantilever walls on their seismic performance. The following conclusions can be drawn: Boundary columns can effectively enhance the wall performance by increasing its ultimate deformation capacity and reducing damage level in the wall panel. However, the final failure of walls with boundary columns was more brittle compared to that of rectangular section walls. Damaged regions due to concrete crushing in rectangular walls spread widely over the lower portion of the walls. The damage spread horizontally to the wall center and was limited in height. Flexure deformation was continuously dominant for rectangular walls while its contribution of flexural drift increased with the increase of drift ratios for walls with boundary columns. The amount and spacing of transverse reinforcement might be a key parameter for large drift capacity is desired. The flexural failures need more studies to deeply investigate the trends of ultimate drift as a function of transverse reinforcement and other key wall properties. Acknowledgments Financial support from the Japan Ministry of Land, Infrastructure, Transportation and Tourism (MLIT) is acknowledged. References 1. Thomsen, J. H., Wallace, J. W. (1995). Displacement-based design of reinforced concrete structural walls: Experimental studies of walls with rectangular and T-shaped cross sections. Rep. No. CU/CEE-95/06, Department of Civil and Environmental Engineering, Clarkson University, Potsdam, N.Y. 2. Zhang Y, Wang Z, Seismic Behavior of Reinforced Concrete Shear Walls Subjected to High Axial Loading, ACI Structural Journal 2000, 97(5): Oh YH, Han SW, Lee LH, Effect of boundary element details on the seismic deformation capacity of structural walls, Earthquake Engineering & Structural Dynamics 2002, 31(8): Dazio A, Beyer K, Bachmann H. Quasi-static cyclic tests and plastic hinge analysis of RC structural walls. Engineering Structures 2009, 31(7): Bonelli P, Restrepo JI, Boroschek R, Carvallo JF. The 2010 Great Chile Earthquake Changes to Design Codes. Proceedings of the International Symposium on Engineering Lessons Learned from the 2011 Great East Japan Earthquake, March 1-4, 2012, Tokyo, Japan 6. Wallace JW, Massone LM, Bonelli P, Dragovich J, Lagos R, Lüders C, Moehle J. Damage and Implications for Seismic Design of RC Structural Wall Buildings. Earthquake Spectra 2012, 28(S1): S281 S Architectural Institute of Japan. AIJ Standard for Structural Calculation of Reinforced Concrete Structures, Tokyo, Japan, ACI Committee 318, Building Code Requirements for Structural Concrete (ACI ) and Commentary, American Concrete Institute, Farmington Hills, MI, 2011, 473 pp. 9. Architectural Institute of Japan. AIJ Guidelines for Performance Evaluation of Earthquake Resistant Reinforced Concrete Buildings, Tokyo, Japan, S. Takahashi, K. Yoshida, T. Ichinose, Y. Sanada, K. Matsumoto, H. Fukuyama, and H. Suwada, Flexural Drift Capacity of Reinforced Concrete Wall with Limited Confinement, ACI Structural Journal 2013, 110(1):