SEISMIC TESTS ON REINFORCED CONCRETE BEAM-COLUMN JOINTS USING STEEL FIBER REINFORCED ULTRA-HIGH-STRENGTH CONCRETE

Transcription

1 SEISMIC TESTS ON REINFORCED CONCRETE BEAM-COLUMN JOINTS USING STEEL FIBER REINFORCED ULTRA-HIGH-STRENGTH CONCRETE Hiroto TAKATSU, and Hideki KIMURA Results of previous cyclic loading tests of columns with concrete compressive strength of 50MPa (Takatsu et al. 006) showed that adding steel fibers into concrete mix increased the flexural strength and the ductility of these columns. Moreover, by comparison with high-strength concrete columns without steel fibers, adding steel fibers to columns prevented early concrete cover spalling and reduced surface crack width. Herein, the bearing capacity of beam-column joints using steel fiber reinforced ultra-high-strength concrete was investigated experimentally. Cyclic loading tests were carried out on interior beam-column joint subassemblies containing different volumetric ratios of steel fibers for two concrete strength classes. The volumetric ratios added to the 0MPa-class concrete were 0 (no addition) and 0.5%, and to the 50MPa-class concrete were 0, 0.5 and.0%. Test results showed that the shear strength of the beam-column joints with steel fibers was higher than the shear strength of the beam-column joints without steel fibers. In addition, spalling of cover concrete was delayed and surface crack width was reduced in the specimens with steel fibers compared to those without steel fibers. The added 0.5% and.0% volumetric ratios of steel fibers induced an increase of % and % in shear strength, respectively. The introduction of steel fibers into concrete mix was suspected to have contributed to the increase of the effective joint width and effective compressive strength. Keywords: ultra-high-strength concrete; beam-column joint; steel fibers; shear strength. While research aiming to provide higher concrete strengths is still ongoing, the practiced concrete in Japan has already reached a maximum nominal compressive strength (Fc) of 00 MPa (Fc00). As known, increasing the strength of concrete affects its mechanical characteristics under compression where explosive spalling of concrete occurs. To allow reinforced concrete columns with concrete strength exceeding Fc00 sustain large earthquakes and ensure them go beyond their assumed ultimate limit, a large amount of transverse reinforcement should be inserted. However, such practice is not always possible due to limited spaces within columns. The authors solved these problems by mixing steel fibers (SF) into concrete. They constructed ultra-high-strength concrete columns and tested them successfully under combined axial-lateral loadings. Adding SF proved to be very effective. It resulted in an increase of strength, an improved ductility and a reduction of cover concrete spalling, as well as a uniform spreading of cracks (Takatsu et al. 006). Based on some investigations (Maruta et al. 004), the ultimate shear strength of beam-column subassemblies with concrete strength exceeding Fc00 was found not to comply with AIJ guidelines (AIJ 999), where it was shown that the evaluated strength could be on the unsafe side. Similarly to columns, the addition of SF to the concrete of beam-column joints was thought to be effective and improve the shear capacity of such subassemblies. For that purpose, beam-column joints with ultra-high-strength concrete containing SF were investigated through Chief Researcher, TAKENAKA Research and Development Institute, Japan, takatsu.hiroto@takenaka.co.jp Senior Chief Researcher, TAKENAKA Research and Development Institute, Japan, kimura.hideki@takenaka.co.jp - -

2 The 0th Taiwan-Korea-Japan Joint Seminar on Earthquake Engineering SHEAR BEHAVIOR OF ULTRA HIGH PERFORMANCE HYBRID FIBER REINFORCED CONCRETE BEAMS Manuel BERMUDEZ, Chung-Chan HUNG Ultra-high performance fiber reinforced concrete (UHPFRC) is a type of high performance fiber reinforced concrete that shows tensile strain hardening behavior accompanied by multiple hairline cracks. It also has an ultra-high compressive strength that is often reported to be greater than 00 MPa. The shear behavior of UHPFRC beams with hybrid fibers was investigated in this study. The major experimental parameters included the shear spans-to-depth ratio and types of mixing fibers. It was found that adequate combination of different types of fibers could enhance cracking and ultimate shear stresses as well as reduce the shear crack width for the beam specimens. Keywords: hybrid fibers; synergy; shear strength; shear reinforcement. Due to the growth of population density in urban areas, nowadays it is common that urban designers base the development of the cities in mid or high-rise buildings. This type of buildings are exposed to huge loads, and the structural elements need to deal with them. The cur-rent codes require a considerable amount of reinforcement for columns, beams, and column-beam joints. One of the main problems in this type of constructions is the congested reinforcement. This study aims to suggest a solution to this issue by utilizing Ultra High Performance Hybrid Fiber Reinforced Concrete in order to reduce the shear reinforcement in beam specimens. ACI Committee (04) has included the use of steel fiber as a minimum shear reinforcement, which opens the possibility to consider if hybrid fibers can also be included. Ultra-high performance fiber reinforced concrete (UHPFRC), which is a type of high performance fiber reinforced concrete (Hung et al., ), has an ultrahigh strength in both tension and compression while showing strain-hardening behavior under tension. Banthia et al. (04), Yoo et al. (06), and Fantilli et al. (08) have been trying to understand the characteristics of Ultra High Performance Fiber Reinforced Concrete (UHPFRC) as well as how to obtain synergy performance from hybrid fibers. Due to modern constructions, it is necessary to develop a material that can reach higher strength and ductility. Juárez et al. (007), Zakaria et al. (009), and Rafeeqi et al. (03) investigated that the hybrid fibers can increase the tensile strength of the matrix and therefore diffuse the crack pattern as well as delay the critical shear crack growth. The use of different type of fibers in the UHPFRC matrix can produce many benefits that can control the already known brittleness of concrete. It is important to observe the shear response on the specimens that have this type of material. In this research, the experimental parameters include the amount and type of fibers and the shear span ratio of the beam. In addition, two different material of fibers were studied to evaluate how the synergy of the mixture enhances the shear strength (see Figure ). In order to develop shear design provisions for UHPFRC beams it is necessary first to realize experimental research to generate broader criteria. The assumptions of many researchers that propose equations to explain the shear MSc., Department of Civil Engineering, National Cheng Kung University, Taiwan, mrbermudez7@gmail.com Professor, Department of Civil Engineering, National Cheng Kung University, Taiwan, cchung@mail.ncku.edu.tw -9-

3 DEVELOPMENT AND APPLICATION OF HIGHLY FLOWABLE STRAIN HARDENING FIBER REINFORCED CONCRETE IN NEW RC BUILDING SYSTEMS Wen-Cheng LIAO, Li-Wei TSENG, and Wei-Ru SU The purpose of New RC project was aimed to reduce the member sections and increase the available space of high rise buildings by using high strength concrete (f c > 70 MPa) and high strength rebars (f y > 685 MPa). Material consumptions and member sections can be further reduced owing to the upgrade of strength. However, the nature of brittleness of high strength concrete may also cause early cover spalling and other ductility issues. In ACI 38-4, higher transverse reinforcement and restrict detailing are required where Pu > 0.3A gf c or f c > 70 MPa in columns. Addition of steel fibers is an alternative as transverse reinforcement. Highly flowable strain hardening fiber reinforced concrete (HF-SHFRC) has excellent workability in the fresh state and exhibits the strain-hardening and multiple cracking characteristics of high performance fiber reinforced cementitious composites (HPFRCC) in their hardened state. The objective of this study is to investigate the feasibility of implementing HF-SHFRC in New RC building systems, such as beam-column joints and base columns as an alternative of transverse reinforcements. Test results show that the HF-SHFRC beam-column joint and base column specimens perform as well as specimens with intensive transverse reinforcements regarding failure mode, ductility, energy dissipation and crack width control. Integration of New RC Building systems and HF-SHFRC can assure construction qualities and further diminish labor work and give infrastructure longer service life, and eventually lower the life-cycle cost. Keywords: Fiber reinforced concrete; New RC; strain hardening; high strength concrete; base columns; beam-column joints. The concept of New RC, which is upgrade of concrete and reinforcement strength, was proposed in Japan in 989. The objective of New RC project was aimed to reduce the member section sizes and increase the available space of high rise buildings by using high strength concrete (f c > 70 MPa) and high strength rebars (f y > 685 MPa). However, the brittle behavior of high strength reinforced concrete during fracturing is still a primary concern despite wide application in the construction industry. Unlike normal strength concrete, as high strength concrete fractures, the fracture plane passes directly through the coarse aggregate and mortar causing an immediate drop in strength once the ultimate strength is reached and catastrophic failure. This failure behavior can be disastrous for societies like Taiwan, which is located in the Pacific Rim seismic belt. Earthquakes are frequent and unpredictable and consequently, structures must be constructed to meet minimum ductility requirements. Traditional manners to increase the ductility of reinforced concrete columns have relied on increasing the amount and number of lateral stirrups in columns. However, by increasing the amount of stirrups, constructability decreases. Resultantly, new materials are being applied to traditional reinforced concrete Associate Professor, Department of Civil Engineering, National Taiwan University, Taipei, Taiwan wcliao@ntu.edu.tw Graduate Student, Department of Civil Engineering, National Taiwan University, Taipei, Taiwan - 9 -

4 BUILDING DAMAGES IN POHANG EARTHQUAKE AND THE LESSONS Hong-Gun PARK, Chul-Goo KIM, Tae-Sung EOM 3, and Dong-Kwan KIM 4 Pohang EQ of Magnitude 5.4 occurred on November 5, 07, which caused heavy damages in earthquake -vulnerable buildings nearby, giving shock to the people in Korea which has been known as a relatively safe region. In this paper, the characteristics and amplitude of the Pohang EQ and the effects on various building types in the Pohang city are briefly described. Based on the field evaluations, the EQ- vulnerable building types that were subjected to heavy damages are classified, and the main causes of the damages are explained. Particularly, buildings with pilotis, which were heavily damaged, were investigated in details. The lessons and aspects for the improvement of design codes and practice are explained. Keywords: Pohang earthquake, earthquake damage, piloti, building, non-seismic On November 5, 07, at :9 P.M. (local time), an earthquake of magnitude 5.4 struck the southeastern region of Korea near the city of Pohang (Fig. ). Pohang, which has a population of 56,000 inhabitants, is about 9 km south from the estimated epicenter. The earthquake was located at 36. N and 9.3 E coordinates with a focal depth of 6.9 km, near the town of Heung-Hae. Fortunately, the earthquake caused no casualty, but left 9 injuries and approximately 754 damaged buildings. The total direct damage cost of the earthquake was estimated to be US $ 50 million (the actual cost is more). Seoul CHS PHA Pohang 9 km Epicenter (M5.4) 3 km Pusan DKJ Pohang HAK Fig. Epicenter of Pohang earthquake and locations of seismic stations Professor, Department of Architecture and Architectural Engineering, Seoul National University, Korea, parkhg@snu.ac.kr Head of KBC center in the Architectural Institute of Korea. Research fellow, Department of Architectural Engineering, Dankook University, Korea, these09@naver.com 3 Associate Professor, Department of Architectural Engineering, Dankook University, Korea, tseom@dankook.ac.kr 4 Assistant Professor, Department of Architectural Engineering, Cheongju University, Korea, dkkim7@cju.ac.kr - 7 -

5 COMPARATIVE ANALYSIS OF STRUCTURAL DAMAGE POTENTIALS OBSERVED IN THE 06 GYEONGJU AND 07 POHANG EARTHQUAKES IN KOREA Cheol-Ho LEE, Ji-Hun PARK, Sung-Yong KIM 3, Su-Chan JUN 4 This paper presents comparative analysis of the 06 Gyeongju and 07 Pohang earthquakes to provide probable explanations and reasons for the much more severe damage observed in the 07 Pohang earthquakes. The damage potentials of the Pohang earthquake such as magnitude, Arias intensity and effective peak ground acceleration (EPGA) were generally weaker than those of the 06 Gyeongju earthquake. However, building damage was much more severe in the Pohang earthquake and recovery cost was estimated to be 0 times the Gyeongju earthquake. In contrast to the high-frequency dominant nature of the Gyeongju earthquake, the Pohang earthquake record from soft soil often contained significant spectral power around Hz. The elastic base shear around Hz frequency was as high as 40% building weight. Unfortunately, this frequency band was shown to be very close to the fundamental frequency of the piloti-type residential buildings fatally damaged in the northern part of Pohang city. In addition to inherent vertical and plan irregularity in the piloti builidngs, most of the damaged piloti-type buildings were poorly designed and detailed. All these contributed to the fatal damage observed in the northern part of Pohang city. The 07 Pohang earthquake clearly showed that near-collapse damage to irregular and brittle piloti-type buildings is highly possible, when subsoil is soft and the epicenter is close, although the earthquake magnitude is just moderate (M 5+) and of short duration. Keywords: strong motion duration; Arias intensity; effective peak ground acceleration; constant ductility spectrum; seismic damage; piloti; irregularity On November 5, 07, a moment magnitude of M L 5.4 earthquake struck the northern part of Pohang, a city in the southeastern part of the Korean peninsula. According to the Korea Meteorological Administration (KMA), the depth of the epicenter was very shallow, 3.5 km deep. This quake was the nd destructive instrumental earthquake in Korea and drove Korean people into panic much more than the Gyeoungju earthquake that occurred one year ago. Compared to the 06 Gyeongju earthquake, however, building damage was much more severe in the Pohang earthquake and recovery cost was estimated to be 0 times the Gyeongju earthquake. Notably, 4 to 5 story-high piloti-type residential buildings located in the northern part of Pohang city were most severely damaged. Overall, the damage seemed very severe compared to the low to moderate magnitude of the earthquake M= 5.4 (see Figure ). This paper presents comparative analysis of the 06 Gyeongju and 07 Pohang earthquakes to provide probable explanations and reasons for the much more severe damage observed in the 07 Pohang earthquakes. This paper is based on the former studies conducted by the authors right after the two earthquakes (Lee 06; 08) and summarizes the time and frequency domain analysis results of the measured ground motions, the elastic and inelastic seismic responses of SDOF (single degree of freedom) structures subjected to the ground motions recorded in the two earthquakes. Based on these basic analysis results, damage potential contained in the measured ground motions and probable causes of the observed damages are discussed. Professor, Seoul National University, Korea, ceholee@snu.ac.kr Professor, Incheon National University, Korea, jhpark606@inu.ac.kr 3 Post-doctoral Researcher, Seoul National University, Korea, sungyong.kim7@gmail.com 4 Graduate student, Seoul National University, Korea, corrora90@snu.ac.kr

6 RETROFITTING BY INSTALLING WING WALLS FOR AN EXTERIOR RC BEAM-COLUMN JOINT WITH SUBSTANDARD STRAIGHT ANCHORAGE OF BEAM LONGITUDINAL REBAR Syafri WARDI, Yasushi SANADA, and Susumu TAKAHASHI 3 Several destructive earthquakes in developing countries have revealed that severe damage to RC buildings was often caused by substandard details of beam-column joints. This study focuses on seismic retrofitting of an exterior RC beam-column joint with substandard straight anchorage of beam longitudinal rebar, which is expected to fail in anchorage, by installing wing walls beside existing columns. The retrofitting method is proposed to extend the development length of beam longitudinal bars by considering the embedment length along the wing walls. Two 0.7 scale RC exterior beam-column joint specimens, one benchmark specimen and one retrofitted specimen with wing walls were fabricated and tested. The test results verified the effectiveness of the proposed retrofitting method to upgrade the exterior RC beam-column joint with vulnerability in beam rebar anchorage. Keywords: developing country; low strength concrete; anchorage failure; reinforced concrete; seismic strengthening; substandard structure. Reinforced concrete (RC) beam-column joints with poor detailing, where typically little/no shear reinforcement and/or insufficient anchorage of beam longitudinal rebar are provided in the beam-column joint region, exist in many buildings designed according to older design codes, such as those designed before the 970s in the USA or without complying with current seismic codes in developing countries. The poorly detailed joints have inadequate seismic performance and the joint failure often lead to the severe damage or collapse of buildings, as observed in recent earthquakes in developing countries (Sanada et al. 009, 0; Takahashi et al. 04). Also, such poorly detailed joints still exist in newly constructed buildings in an area which had experience of severe damage due to past earthquake, as observed in a field survey by two of the authors (Wardi et al. 08). The existing buildings with the substandard beam-column joints need to be retrofitted. In consideration of the economic situation and technical level of developing countries, a retrofit method by installing RC wing walls has been proposed (Li et al. 07). The effectiveness of the retrofit method has been validated to upgrade exterior beam-column joints without shear reinforcement. However, the retrofitting method has not been applied to exterior joints with substandard anchorage of beam longitudinal rebar yet. Figure shows a collapsed building due to anchorage failure of beam longitudinal bars, which had straight anchorage into the exterior beam-column joints. In this study, the retrofitting method by installing wing walls is proposed to upgrade exterior beam-column joints with substandard anchorage of beam longitudinal rebar. A series of static loading tests was conducted with two specimens: one existing joint with straight anchorage of beam longitudinal bars into joint and one retrofitted joint by installing wing walls. The length of wing walls is proposed to extend the development length of beam PhD Candidate, Graduate School of Engineering, Osaka University, Japan, wardi_safri@arch.eng.osaka-u.ac.jp Associate Professor, Graduate School of Engineering, Osaka University, Japan, sanada@arch.eng.osaka-u.ac.jp 3 Lecturer, Faculty of Engineering, Daido University, Japan, susumu-t@daido-it.ac.jp

7 INTEGRATION OF ARTIFICIAL NEURAL NETWORK AND STOCHASTIC SUBSPACE IDENTIFICATION FOR STRUCTURAL HEALTH MONITORING Chia-Ming Chang a, Tzu-Kang Lin *, Chih-Wei Chang b This study proposes a new artificial intelligence-based structural health monitoring strategy based on neural network modeling. A neural network model is developed in accordance with a numerical model which is derived from the identified modal properties under ambient vibrations. The stochastic subspace system identification is first implemented to derive the natural frequencies and mode shapes of a healthy structure. These natural frequencies and mode shapes are then employed to derive a simplified model of this structure, allowing changing stiffness terms to construct various damage patterns. A neural network model is trained and built by the modal properties of the structure with these damage patterns. After a critical event occurs, this neural network model can be employed to estimate the damage patterns in terms of stiffness reduction. Moreover, the proposed strategy is also applied to an experimental test of a scaled twin-tower building with weak braces in some floors. Partially modal properties of the structure are obtained from the stochastic subspace system identification, while a simplified model is developed in accordance to the identified modal properties of the healthy building. Then, a neural network model is established based on this simplified model. After seismic events, this neural network model is employed to carry damage detection of this building in terms of damage locations and levels. As a result, the proposed artificial intelligence-based strategy is quite effective to locate damage if the identified modal properties are relatively accurate. Keywords: artificial neural network; structural health monitoring; structural health monitoring; damage location. Performance of structures can be degraded by repetitiously natural hazards. After strong ground shaking or wind loading, these damaged members may be sometimes invisible due to non-structural components (e.g., partition walls and veneers). Thus, the damage should be assessed by alternative methods such as non-destructive testing methods or vibration-based structural health monitoring techniques. Deviations in dynamic characteristics of structures (i.e., natural frequencies, damping, and mode shapes) are one kind of these changes, while these characteristics can be obtained from the operational modal analysis. Model updating is a method that adjusts the parameters used in the structural model to be capable of presenting the current state of this structure. The parameters to be updated can be obtained by processing static or dynamic measurements of the structure. To form a representative model of a structure, system identification techniques play an important role. As mentioned previously, modal properties are often utilized to derive or update structural models. These modal properties can be obtained through the operational modal analysis along with system identification. In time domain, the stochastic subspace system identification proposed by Van Overschee and De Moor in 99 is one of the most acceptable methods. This method utilizes the extended observability to * Corresponding author, Associate Professor, tklin@nctu.edu.tw a Assistant Professor., changcm@ntu.edu.tw b Graduate Student., waynechang009@yahoo.com.tw

8 FUNDAMENTAL STUDY ON BOND PROPERTIES USING DISTRIBUTED OPTICAL FIBER SENSING TECHNOLOGY Atsushi SHIBAYAMA, This paper presents an experimental technique to investigate the bond behavior between the concrete and reinforcing bars in reinforced concrete specimens using optical fiber strain sensors installed in the specimen. In the experiment, the concrete specimen was tested with an axial tension load, and the reinforcement strain was measured along the length of the embedded bars using the strain sensors. This distributed strain measurement system allowed the determination of bond behavior and provided a high spatial resolution of mm. The experimental results are discussed and compared with calculations. It was found that the optical fiber sensor could measure not only large strain values due to crack formations but also slight changes in the reinforcement strain. The observed bond stress distribution was compared with calculations according to two bond stress slip models. The calculated mean bond stresses according to the fib Model Code 00 were in good agreement with the observed bond stresses on both sides of the formed cracks. Keywords: bond, distributed strain measurement technology, optical fiber sensor, Rayleigh backscatter, reinforced concrete The bond between concrete and reinforcing bars has considerable influence on the behavior of reinforced concrete. Crack spacing, crack width, and load displacement, among other behavior, are related to the bond slip relationship. Many experimental studies have been conducted to investigate the bond properties between concrete and reinforcing bars under tensile load. For example, Kokubu & Okamura (965) and Scott & Gill (987) created two half-moon-shaped bars that had grooves in the center of each bar. They installed a number of electrical resistance strain gauges (RSGs) into the milled grooves and measured the reinforcement strains. These studies clarified some of the bond properties between concrete and reinforcing bars. However, this technique is very expensive, time-consuming, and not suitable for parametric studies. Recently, distributed optical fiber strain sensing technology was developed, which enabled a simpler distributed strain measurement. Knenel & al. (05) and Davis & al. (07) experimentally clarified the effectiveness of this technology to investigate bond behavior and tension stiffening. The deformed bars used in Japan are slightly different in shape compared to those used in the US and Europe. There are very few studies that use distributed optical fiber sensors to investigate the bond behavior between concrete and reinforcing bars in Japan. In this study, longitudinal reinforcement strain distributions in reinforced concrete tension members were measured using a distributed optical fiber strain sensor to investigate bond behavior between the concrete and reinforcing bar. Research Scientist, Civil Engineering Research Laboratory, CRIEPI, Japan, atushi@criepi.denken.or.jp PhD Candidate, Dept. of Architecture and Architectural Engineering, Kyoto University, Japan, rc.shibayama@archi.kyoto-u.ac.jp

9 REAL-TIME SHAKE TABLE HYBRID TESTING OF POLYNOMIAL FRICTION PENDULUM ISOLATORS Shih-Yu Chu, Lyan-Ywan Lu, Shih-Wei Yeh 3 and K-Y Chu 4 The pseudo-dynamic testing of isolation devices that interacts with hydraulic actuator and reaction wall is usually a cost-effective alternative to the shaking table test. However, there are a few bottlenecks need to be overcame especially for the velocity-dependent or highly-nonlinear friction device, such as the polynomial friction pendulum isolators (PFPIs). A testing framework called real-time hybrid testing with a shake table (RTHT-ST) is adopted in this study as an alternative of shaking table test to investigate the isolation performance of a floor isolation system (FIS) equipped with PFPI. The PFPI-FIS is installed directly on a shake table which simulates the absolute displacement of the equipped floor of building by interacting with the measured total shear force between the isolation layers. In order to get more comparison cases, some important parameters of the numerical model of PFPI-FIS are identified through the component test. The shaking table test measurements are compared with the proposed RTHT-ST. From the good agreements on the responses and hysteresis loops, the feasibility of the proposed RTHT-ST framework is ensured. Keywords: Polynomial Friction Pendulum Isolator, Variable Curvature, Friction Bearing, Real-Time Hybrid Testing with a Shake Table, Floor Isolation System, Shaking Table Test Numerical analysis method has been developed for a long time and is able to simulate the dynamic behavior of a structural system subjected to seismic excitations. But the accuracy of the numerical analysis method for a nonlinear complex component still need to be verified through experimental tests. Shaking table test (STT) is a kind of experimental methodology which is wildly used in civil engineering. The concept of the STT is to simulate historic earthquakes by using a rigid table driven by a few hydraulic actuators and to observe the dynamic behavior of the tested structural system mounted on the table. Therefore, STT has been believed that is one of the most reliable experimental methodology for studying the dynamic responses of a structure subjected to seismic excitations. However, STT for a full-scale structure needs a large-scale hydraulic oil system and actuators. The cost of the STT for such a test should be very high. Moreover, the control technique for simulating full-scale specimen and the effect of specimen-table interaction are the major bottlenecks for this kind of STT. Associate Professor, Department of Civil Engineering, National Cheng Kung University, Taiwan, sychu@mail.ncku.edu.tw Professor, Department of Civil Engineering, National Cheng Kung University, Taiwan; lylu@mail.ncku.edu.tw 3 Associate Technologist, Southern Taiwan Experiment Division, National Center for Research on Earthquake Engineering, Taiwan; swyeh@narlabs.urg.tw 4 Master of Science, Department of Civil Engineering, National Cheng Kung University, Taiwan, quit03@gmail.com

10 The 0th Taiwan-Korea-Japan Joint Seminar on Earthquake Engineering An Experimental Study of Frictional-Pendulum Isolation System Subjected to PulseLike Ground Motions Ya-Heng Yang, Yu-Chen Lin, Yin-Nan Huang3 A series of shaking table tests were conducted using a friction-pendulum isolation system subjected to four sets of ground motions and 0 motions for each set. Set (Set ) includes pulse-like records with pulse periods ranged between 0.5 and ( and 6) seconds; Set 3 are non-pulse-like records with average spectral accelerations similar to that of Set ; and Set 4 are synthetic gourd motions spectrally matched to the response spectra of the records of Set. The displacement demand of the tested isolation system for Set is much greater than that for Set despite that both sets are all pulselike records. Also, similar mean spectral accelerations for two sets of ground motions over a wide period range do not guarantee similar responses of the isolation system. The commonly used index of peak acceleration ratio between super- and sub-structures may overestimate the effectiveness of the isolation system when subjected to pulse-like ground motions. Keywords: friction pendulum; isolation; near fault; pulse-like ground motion; shaking table test Near-fault (NF) ground motions are the ground motions that occur at sites close to a fault and they are different from far-field ones in often including large-amplitude long-period velocity pulses and possible permanent ground displacement which may be more destructive than what their peak ground accelerations indicate. This research studies the impact of velocity pulse on the performance of friction-pendulum bearings. Since not all the NF ground motions contain velocity pulses, the term pulse-like ground motion instead of NF ground motion is used in this paper to avoid possible misunderstanding. Although a few studies have recognized the higher demands imposed by pulse-like ground motions in the 950s (Bertero, et al. 999), the engineers refocused their attention back onto them following several destructive earthquakes, such as the 97 San Fernando earthquake, the 99 Landers earthquake, and the 994 Northridge earthquake. Bertero et al. (978) and Anderson and Bertero (987) indicated that the pulses in the NF ground motions are the cause for the large demands on multistory structures and also mentioned that obtaining inelastic design response spectrum of the pulse-like ground motion directly by modifying its linear-elastic response spectra may not be reliable, that is, records with similar elastic spectral accelerations may lead to different inelastic demands (MacRae, et al. 00). Different from the performance of isolation systems under non-pulse-like ground motions, the behavior of the systems subjected to pulse-like ground motions has the following features: () an excessive isolation displacement; () transmitting higher accelerations into the superstructure; (3) resonance-like behavior when the isolation period is close to the pulse period. (Hall et al., 995; Jangid and Kelly, 000; Rao and Jangid, 00; Shen et al., 004; Sharbatdar et al., 0; Lu et al., 03; Alhan and Öncü-Davas 06) In applications, in order to reduce the displacement demands on isolators, two strategies are commonly used: () using large isolators to accommodate 3 Graduate student, Department of Civil Engineering, National Taiwan University, Taiwan, r06506@ntu.edu.tw Former graduate student, Department of Civil Engineering, National Taiwan University, Taiwan, r0558@ntu.edu.tw Associate Professor, Department of Civil Engineering, National Taiwan University, Taiwan, ynhuang@ntu.edu.tw

11 Seismic Behavior of Rocking Reinforced Concrete Walls and Hybrid Walls Taku OBARA, Netrattana Chanipa,David MUKAI 3 and Susumu KONO 4 In the 995 Kobe Earthquake and the 0 Tohoku Earthquake, many structural and non-structural reinforced concrete walls had heavy damages and it was difficult to keep using buildings after earthquakes even if the buildings did not collapse. Economic loss due to these damage was sometimes more than repair cost. As a solution to low damage system, self-centering system emerged using unbonded post-tensioned precast concrete technology. This structural system has enhanced resilient performance of quick recovery after earthquakes because residual cracks are negligibly small and tensile yielding of mild reinforcement does not take place. However, unbonded post-tensioned precast concrete members have small energy dissipation and large seismic response. Hence, energy dissipating elements are used as a solution. An experiment was carried out on four unbonded post-tensioned precast concrete walls (hereafter, rocking wall) with and without energy dissipating elements (hereafter, dampers) to quantify damage level and energy dissipation. The experiment was conducted under cyclic static lateral loading with constant axial load. All rocking walls had few cracks with negligible crack width and small residual drift until.0% of drift angle. In terms of compressive damage of concrete, all specimens did not have heavy concrete s crushing until 0.5% of drift angle. In.0% of drift angle, all specimens had crushing of concrete at the edge of wall panel, but the region of concrete s crushing was small. In rocking walls with dampers, equivalent damping ratio was approximately 0% until.0% of drift angle because dampers were yield at small drift angle. This paper describes the superior recovery performance of rocking walls with and without dampers focusing on high damage controlling ability with better energy dissipation. (This paper was submitted to 6 th European Conference on Earthquake Engineering, 8- June 08.) Keywords: Rocking wall; Energy dissipating element; Damage evaluation; Limit states In the 995 Kobe Earthquake and the 0 Tohoku Earthquake, many non-structural Reinforced concrete (Hereafter, RC) walls had heavy damages and it was difficult to keep using buildings after earthquakes even if other structural members did not suffer damage. Buildings were demolished or repaired with high cost for a long period because non-structural RC walls around windows or doors had heavy cracks and spalling (NILIM and BRI (0)). Therefore, the society seeks for buildings with superior damage controlling ability like the base isolated buildings. One of other structural systems, which is proposed to realize damage reduction for non-structural RC walls, is unbonded post-tensioned precast concrete system (Hereafter, rocking system). It does not require higher initial construction cost nor extra space for an isolated story. Rocking system is able to reduce flexural cracks on walls (Priestley, M.J.N. (999), Jose I. Restrepo (007)). Priestley et al. and Restrepo et al. studied load deformation curve of rocking walls and compared energy dissipating performances of rocking walls with and without dampers (K. M. Twigden (04), Tony Holden Ph.D. Candidate, Dept. of Architecture, Tokyo Institute of Technology, Yokohama, Japan, obara.t.ac@titech.ac.jp Ph.D. Candidate, Dept. of Architecture, Tokyo Institute of Technology, Yokohama, Japan, chanipa.net@gmail.com 3 Associate Professor, Civil and Architectural Engineering, University of Wyoming, US, dmukai@uwyo.edu 4 Professor, Institute of Innovative Research, Tokyo Institute of Technology, Yokohama, Japan, kono.s.ae@m.titech.ac.jp

12 SEISMIC PERFORMANCE OF UNBONDED POST-TENSIONED PRECAST CONCRETE WALLS WITH INTERNAL AND EXTERNAL DAMPERS LIU Yuan, BEDRIÑANA Luis A., TANI Masanori 3, KONO Susumu 4, NISHIYAMA Minehiro 5 Despite the research efforts on Unbonded Post-Tensioned (UPT) Precast Concrete Walls over the last decades, the application of UPT walls in building structures is still insufficient, which can be attributed to the lack of experience with UPT walls featuring external replaceable dampers as well as design codes for UPT walls. This paper presents the outline and results of two half-scaled UPT walls, one with internal dampers and the other with external dampers. It was observed that UPT walls could sustain large lateral deformations within limited damage with drifts over 3%; damage in the specimens was limited to the base panel and no damage was found in the upper panel; core concrete and confinement hoops were sound through the test; the yielding of internal and external dampers provided the main source of energy dissipation in their corresponding specimen while PT tendons provided the main restoring force to the specimens and remained within the elastic range throughout the test; the gap opening at the bottom joint was extensive and consisted of the inelastic deformations of the specimens. It showed that the internal and external dampers were effective in dissipating energy; however, UPT-IA showed three times larger residual drifts than UPT-EA because of the buckling of internal dampers which leaded to the difficulty in closing bottom joint gaps. Therefore, replaceable external dampers can be considered more efficient for recovering after a severe earthquake. Keywords: post-tensioning; self-centering; precast concrete; structural wall; damper; unbonded. Over the last decades, researchers have been investigating the seismic performance of Unbonded Post-Tensioned (UPT) Precast Concrete Wall Structures (e.g., Kurama 999, Kurama 03, Kurama 05). The advantage of this system over conventional cast-in-situ reinforced concrete walls is that it has better seismic performance, to be specific, it is able to sustain large deformations while maintaining an adequate energy dissipation and smaller residual deformation (Restrepo 003, Restrepo 007). Besides, the damage on UPT walls can be restricted to cover concrete spalling at the critical joint. However, the application of UPT walls in building structures is still limited, which can be attributed to the insufficient experience with UPT walls featuring external replaceable dampers (Marriott 008).Although it has been reported that the ultimate performance of UPT walls relies on detailing at the bottom joint such as mechanical couplers for mild steel reinforcement and confining reinforcing details of the wall bottom (Kurama 0), there is still a lack of references for construction details of UPT walls (Kurama 08), comparing to common-seen precast concrete structures. In addition, previous experimental research featured UPT walls with low levels of axial load; then, further research is needed to evaluate the application of UPT walls as bearing structural walls with external replaceable dampers,. Graduate student, Department of Architecture and Architectural Engineering, Kyoto University, Japan, rc.liu@archi.kyoto-u.ac.jp Ph.D candidate, Department of Architecture and Architectural Engineering, Kyoto University, Japan, rc.luis@archi.kyoto-u.ac.jp 3 Associate professor, Department of Architecture and Architectural Engineering, Kyoto University, Japan, tani@archi.kyoto-u.ac.jp 4 Professor, Tokyo Institute of Technology, Japan, kono.s.ae@m.titech.ac.jp 5 Professor, Department of Architecture and Architectural Engineering, Kyoto University, Japan, mn@archi.kyoto-u.ac.jp

13 The 0th Taiwan-Korea-Japan Joint Seminar on Earthquake Engineering AN ANALYTICAL MODEL FOR PARTIALLY-CONFINED MASONRY PANELS Yi-Hsuan TU and Bor-Min CHANG A nonlinear backbone model for the lateral force-displacement relationship of partially-confined masonry panels surrounded by RC frames is presented in this paper. The model was developed mainly by assuming the partially-confined masonry panel a composite structure consists of an equivalent diagonal strut and a moment-resistant column. Multiple failure modes including diagonal tension, bed-joint sliding, and diagonal compression were considered for the panel. The failure modes govern the lateral strength as well as the form of the backbone curve. A rational way to estimate the share of lateral force between the panel and the column was also proposed. Test results of specimens from Taiwan and other countries were used to verify the analytical model. The experimental and analytical force-displacement relationships were compared. The comparison showed that the proposed model can adequately portray the behavior of partially-confined masonry panels and reasonably assess the lateral strength and deformation capacity. Keywords: Confined masonry; in-plane; opening; seismic assessment. Confined masonry (CM) is one of the common construction types used for the existing low-rise buildings in Taiwan. Unlike the typical CM in Southern Europe, Latin America, and other parts of Asia (EERI 0). The Taiwanese CM has moment-connected beams and columns with larger sections rather than pin-connected tiebeams and tie-columns. Also, there is usually no confining members around the openings, as required for the standard CM buildings (Meli and Brzev 0). It means that a CM panel contains a door or window can be regarded as a combination of partially-confined panel segments. As depicted in Figure, the combination varies with the opening type that is usually considered to affect the positions of the equivalent diagonal struts form in the panel. (b') (b) (b) (a) (a') (b') Figure. Confined masonry panels with openings as combinations of partially-confined segments. Evaluating the contribution of masonry panels with openings in the seismic assessment is a difficult issue. There have been several analytical methods proposed for CM panels with no opening. The famous research by Tomaževič and Klemenc (997) proposed a trilinear curve model defined by cracking, maximum resistance, and ultimate displacement. Riahi et al. (009) followed the trilinear model skeleton and derived the strength and displacement equations by regression analysis from a databank of 0 test specimens. The design guide by Meli and Brzev Associate Professor, National Cheng Kung University, Department of Architecture, Taiwan, yhtu@mail.ncku.edu.tw Master of Science, National Cheng Kung University, Department of Architecture, Taiwan, hsnu755@gmail.com - 3 -

14 STRUCTURAL BEHAVIOR AND RESTORATION OF MIREUKSAJI STONE PAGODA Fahimeh YAVARTANOO, Thomas H.-K. KANG and Sung-Gul HONG 4 Mireuksaji stone pagoda which is the largest and one of the oldest Korean pagodas was excavated in the 9 th century. More than half of the Mireuksaji stone pagoda had collapsed and remained damaged until 95 when repairs were made by applying concrete layers. Due to the historical, cultural and architectural value of the Mireuksaji stone pagoda, which reflects the style of wooden buildings of that era, it is necessary to evaluate and investigate the cause of collapse for reconstruction and restoration plans for this cultural heritage. In this study to evaluate the structural behavior of Mireuksaji stone pagoda, material properties of stones in the pagoda were investigated by taking samples from the collapsed part of the pagoda. Uniaxial and triaxial compressive strength tests, tensile strength tests, and nondestructive tests for the porosity and the absorption rate of the stone were carried out. Then three-dimensional numerical analysis for discontinuous modeling based on the distinct element method was reported to analyze the pagoda. Keywords: Mireuksaji stone pagoda; masonry; deterioration; collapse; material properties; interfaces. The Mireuksaji stone pagoda is the oldest and largest stone pagoda in Korea which was built in Baekje Muwangdae (660 ~ 640 years). It is an important cultural property showing the process of transition from wooden pagoda to the stone pagoda. However, the Mireuksaji stone pagoda largely collapsed before 95, and Japanese preserved the collapsed part of the pagoda by using concrete and stone in an emergency repair (see Figure ). The structural safety check carried out in 998 raised the risk of further collapse, and since 00 it has been repaired in order to investigate the cause of the collapse and restore it. Most of the previous studies on the Mireuksaji stone pagoda are concentrated on the architectural aspects and the aesthetic viewpoint such as its era time and style. This study is carried out to investigate its structural performance to find engineering reasons for the collapse of the Mireuksaji stone pagoda. It is assumed that the causes of collapse may be three-fold: () the collapse of the laminated structure due to the subsidence of the ground, () the collapse of the upper structure due to lightning, and (3) the collapse due to the excessive transverse displacement due to earthquake and/or wind load. The fundamental cause of the collapse is the structural instability of the stone pagoda. ) Ph.D. Student, Dept. of Architecture & Architectural Engineering, Seoul National University, Korea ) Professor, Dept. of Architecture & Architectural Engineering, Seoul National University, Korea 4) Professor, Dept. of Architecture & Architectural Engineering, Seoul National University, Korea - 3 -

15 SEISMIC SHEAR BEHAVIOR OF SLENDER HIGH-STRENGTH REINFORCED CONCRETE COLUMNS Yu-Chen OU and Bao-Nguyen NGUYEN-VAN To investigate the shear behavior of high-strength reinforced concrete columns, particularly columns for the first story of high-rise buildings, seven shear-critical high-strength columns with a shear span to effective depth ratio (a/d) of.5 were tested under double-curvature cyclic loading. Design parameters that were studied included the specified yield strength (80, 40 and 785 MPa) and the amount of shear reinforcement. Test results and analytical study showed that the concrete shear strength equations from the ACI 38 code are applicable to the slender high-strength columns and the use of an upper limit of 600 MPa for the steel stress is conservative for shear strength calculation by the ACI code equations. Keywords: high-strength reinforced concrete columns; shear span to effective depth; shear reinforcement; shear strength. Material strengths commonly used for reinforced concrete structures are from 8 to 35 MPa and from 8 to 4 MPa for concrete and steel, respectively. Nowadays, there is an increasing interest in the use of high-strength materials due to a lot of benefits they have provided such as relieving reinforcement congestion or decreasing consumption of construction materials, then advancing environmental sustainability. This research is a part of Taiwan New RC project which is aiming to advance reinforced concrete structures with high-strength materials and to make it more popular and useful in Taiwan. In this project, high strength concrete with specified compressive strength up to 70 MPa was used. The deformed reinforcement SD685 and SD785 with specified yield strength of 685 MPa and 785 MPa has been used for longitudinal and transverse reinforcement, respectively. There were 5 high-strength reinforced concrete columns in total (Ou 05a; Ou 05b) which have been tested with a shear span to depth (a/d) ratio of.88 under variant axial load ratio to examine the effect of axial load on behavior of the columns. Furthermore, the previous papers collected shear strength data of 6 high-strength columns with shear span to depth (a/d) ratio ranging from.5 to.88. Based on these all test data, the authors evaluated and proposed a new shear strength model for high-strength reinforced concrete columns. In this proposed model, the weakening effect of axial compression on diagonal cracking strength was considered. Additionally, the upper limits of concrete compressive strength, yield strength of steel and the minimum amount of transverse reinforcement to prevent sudden failure after formation of diagonal cracking were suggested. Note that all high-strength columns collected have a/d ratio equal to or smaller than.88. However, the a/d ratio might exceed this value in high-rise building, particularly for columns in the first story. Consequently, studying the shear behavior of high-strength columns with a/d ratio larger than.88 is necessary. Professor, Department of Civil Engineering, National Taiwan University, Taiwan, yuchenou@ntu.edu.tw PhD Student, Department of Civil Engineering, National Taiwan University, Taiwan - 9 -

16 Effects of lap splice details of column longitudinal reinforcement at plastic hinge region Chul-Goo KIM, Hong-Gun PARK, and Tae-Sung EOM 3 Longitudinal reinforcing bars in columns are often lap-spliced at the bottom of first story columns where plastic hinges can be formed. In this study, seismic resistance of columns with such lap splices was investigated by performing cyclic load tests. The effects of lap splice length and details (lower bar offset splice, upper bar offset splice, and side-by-side splice) of column longitudinal bars were evaluated. The test results showed that the flexural strength, deformation capacity, and energy dissipation capacity of columns were significantly affected by the splice details despite the same lap splice length. In columns with lower bar offset splice, the deformation and energy dissipation capacities were relatively large but the flexural strength was less. In the columns with upper bar offset splice or side-by-side splice, on the other hand, the flexural strength was high but ductility was less. Such differences in the seismic behaviors among splice details are mainly attributed to the location of lower bars from the pedestal. Keywords: SEEBUS; splice detail; splice length; lap splice; offset bar splice; seismic performance. In low and moderate seismic zones, for convenient bar placement, lap splices of column longitudinal bars are used at the bottom of columns where potential plastic hinges form. The seismic performance of columns with such lap splices can be degraded due to premature bond deterioration of spliced bars. Thus, in current seismic provisions of ACI 38-4 for special moment frames, lap splices are permitted in the center half of the column height where relatively small inelastic deformation is required. The flexural strength and deformation capacity of columns with lap splices are mainly influenced by the offset bar details as shown in Fig.. ACI specifies two types of offset bar details; lower bar offset splice and upper bar offset splice. In the lower bar offset splice, the lower bars from the lower story are offset inside and spliced with straight upper bars. In the upper bar offset splice, the upper bars are offset inside and spliced with the straight lower bars from the lower story. However, in small buildings, alternatively, lap splices of straight bars are used for convenience of construction (Fig. (c)). Lap splice length is also an important parameter for the seismic behavior of columns. Current design codes define tension splice length of deformed bars as Eqs. ()-(). l s l.3l.3 s l d f 0.9d f ' c b y c K d tr d f b yd 6 bd fbd b for class B in ACI 38-4 () in EC () Research fellow, Department of Architectural Engineering, Dankook University, Korea, these09@naver.com Professor, Department of Architecture and Architectural Engineering, Seoul National University, Korea, parkhg@snu.ac.kr 3 Associate Professor, Department of Architectural Engineering, Dankook University, Korea, tseom@dankook.ac.kr

17 The 0th Taiwan-Korea-Japan Joint Seminar on Earthquake Engineering EXPERIMENTAL STUDY ON BOND STRENGTH IN R/C BEAMS WITH DOUBLE LAYERED CUTOFF BARS Koshiro NISHIMURA, and Naoki ONISHI When a reinforced concrete flexural member is designed against shear loads according to the AIJ standard in japan, bond stress on the bar surface should be verified. In this standard, the bond strength in the second layer is reduced to 0.6 times due to influence of bond stress in the first layer. However, in previous study, the evaluation method of the bond strength was reviewed, and it has been indicated that evaluating the total bond capacity is logical. In this study, pull-out tests of reinforcing bars of different development lengths were carried out to evaluate bond capacity of double layered bars. The specimens were intended for the condition in R/C beams including cut-off bars. As a result, the following conclusions can be drawn: the effects of increasing transverse reinforcement in single and double layered specimens were almost the same; the calculations of the bond capacities by the proposed equation showed good agreement with the test results. Keywords: reinforced concrete; deformed bar; bond splitting; double layer; cut-off. When a reinforced concrete flexural member is designed against shear loads according to the AIJ standard for structural calculation of reinforced concrete structure published by Architectural Institute of Japan (00), bond stress between reinforcing bars and concrete should be verified. The bond strength is evaluated as shear stress on the surface of the bars. The bond splitting strength in the AIJ standard is based on pull-out test results of longitudinal deformed bars embedded in concrete (Fujii and Morita 983). Regarding double layered bars, there have been pull-out tests, and the development lengths of the bars were identical as shown in Fig. (a). These tests indicated that the bond strengths in the second layer weakened as the bond stress in the first layer became large. And the bond strength in the second layer was apparently lower than that of the bar in first layer because of influence of bond stress in the first layer (Tsuihiji et al. 993). Therefore, in the AIJ standard (00) and guidelines (999), the bond strength in the second layer is simply reduced to 0.6 times. In an R/C beam, some of longitudinal bars are terminated in the span if those are no longer required to resist flexural loads. In the previous experimental studies of R/C beams (Ito et al. 03), it was indicated that maximum bond stresses in cut off bars in the second layer were larger than those in bars passing through the span, as shown in Fig. (Ito et al. 03). That is because, when bars in the second layer were terminated, bond stress distribution in both the first and second layers is different from the case that all the double layered bars are arranged continuously through the span. The evaluation method of the bond strength was reviewed in the previous study, and it has been indicated that evaluating total bond capacity is logical (Nishimura et al. 06). In order to evaluate the total bond capacity, pull-out tests of reinforcing bars of different development lengths, as shown in Fig. (b), were carried out in this study. The specimens were intended for the condition in the beam including cut-off bars. Test variables were the number of layers, the number of bars in second layer, development length ld of bars in the first layer, transverse reinforcement ratio t and concrete strength. Three bars were placed in the first layer in all the twenty eight specimens, and each test variable was as follows: single or double layers; Associate Professor, Tokyo Institute of Technology, Japan, nishimura.k.ac@m.titech.ac.jp Assistant Professor, The University of Tokyo, Japan, onishi@arch.t.u-tokyo.ac.jp

18 SEISMIC PERFORMANCE OF BEARING WALL BUILDING WITH PILOTI CONFIGURATIONS DAMAGED BY 07 POHANG EARTHQUAKE Tae-Sung Eom, Seung-Jae Lee, and Hong-Gun Park 3 During Pohang Earthquake in 07, a number of residential buildings with piloti configurations at their first level were severely damaged. In this study, the results of an analytical investigation on the seismic performance of two bearing wall buildings with piloti configurations are presented. The vibration mode and lateral force-resisting mechanism of the piloti buildings with vertical and plan irregularity were investigated first. Then, based on the investigations, methods of nonlinear modeling of walls and columns at the piloti level were proposed. Finally, by performing nonlinear static and dynamic analyses, the seismic performance and vulnerability of the piloti buildings were discussed. The results have shown that the area and arrangement of walls at the piloti level were the most influential factors to the seismic performance of the overall buildings. During the initial behavior, the lateral resistance of the piloti story was contributed mainly by the walls exhibiting shear-dominated behavior. On the other hand, after premature shear cracking and yielding of the walls, the columns governed by flexure in double curvature contributed significantly to the residual strength and ductility of the piloti buildings. Keywords: Pohang Earthquake, Pilotis, Bearing wall, Seismic performance, Pushover analysis, Time history analysis, Shear wall, Reinforced concrete. During Pohang Earthquake on November 5 th in 07, a majority of low-rise residential buildings with pilotis, located at the area of Jangseong-dong 3 km distant from the epicenter, were severely damaged (see Fig. ). These buildings, most of which were built early in 000 s, have pilotis (ground-level supporting columns) to secure parking space. Forty seven buildings among 3 buildings located in Jangseong-dong were significantly damaged. Although the level of earthquake damages in these buildings were different according to their structural characteristics, such as the arrangement and area of walls, this is an example to show that buildings with piloti configurations is vulnerable to earthquake. Fig. Four-story residential buildings with pilotis damaged by 07 Pohang Earthquake Associate Professor, Department of Architectural Engineering, Dankook University, Korea, tseom@dankook.ac.kr Graduate student, Department of Architectural Engineering, Dankook University, Korea,, kambig@naver.com 3 Professor, Department of Architecture, Seoul National University, Korea,, Parkhg@snu.ac.kr