EFFICIENT ANALYSIS OF BUILDING STRUCTURES USING SUPER ELEMENTS

Transcription

1 EFFICIENT ANALYSIS OF BUILDING STRUCTURES USING SUPER ELEMENTS Dong-Guen LEE 1, and Hyun-Su KIM 2 Recently, building structures become larger and higher to satisfy the social and economical needs. In order to accurately predict structural behavior of a large-sized structure, a detailed and complicated finite element (FE) model is required. When the detailed FE model is used for structural analysis of a large-sized structure, significant computational time and efforts may be required. Therefore, many researches on efficient structural analysis methods have been conducted to date. In general, super elements and substructures are frequently used for an efficient finite element analysis. In this study, an efficient structural analysis method using an advanced super element technique has been investigated for various types of structures. Keywords: super element; fictitious beam; building structures; seismic analysis. When a large-sized shear wall building with openings or a flat slab structure having irregular shaped floors is analyzed, a refined mesh FE model is required to obtain accurate analysis results. In this case, a super element can be used for an efficient analysis. When developing a super element for a floor slab or a shear wall, the DOFs at inner area nodes of a super element are usually condensed by matrix condensation and the DOFs at boundary nodes are remained to combine the super element with adjacent super elements. If the size of a super element is too big, the remained DOFs of the analytical model may be still considerable. Therefore, the number of DOFs can be more reduced by eliminating most of DOFs except only a few DOFs which can represent the structural behavior of a large-sized structure. In order to avoid this inconvenience, a novel super element technique has been proposed in this study. In the proposed method, only the DOFs at corner nodes of a super element are remained and the developed super elements are combined to construct a full FE model. If these super elements are used for the numerical analysis, the compatibility condition will not be satisfied at interfaces of super elements because the nodes only at the corners of the super elements are shared by adjacent super elements. To solve this problem, stiff fictitious beams have been introduced in this study. In order to verify the efficiency and accuracy of the proposed method, static and dynamic analyses of various example structures including high-rise shear wall buildings with various types of openings and flat slab structures having irregular shaped floors with openings were performed. Based on analytical results, it has been confirmed that the proposed method can provide results with outstanding accuracy requiring significantly reduced computational time and memory. SUPER ELEMENTS Some of the high-rise apartment buildings adopt the box system that has repeated arrangement of residential units with the same architectural plan on each floor. Figure 1(a) shows a 3 story shear wall structure with window type openings and Fig. 1(b) shows a fine mesh model for this structure using many plane stress elements. This fine mesh model can be separated into 3 substructures having the same configuration in each story as shown in Fig. 1(c). Super elements illustrated in Fig. 1(d) could be generated if all of the DOF`s except those for the 1 Professor, Department of Architectural Engineering, Sungkyunkwan University, Korea, dglee@skku.ac.kr 2 Assistant Professor, Division of Architecture, Sunmoon University, Korea, hskim72@sunmoon.ac.kr - 1 -

2 The Eleventh Taiwan-Korea-Japan Joint Seminar on Earthquake Engineering for Seismic Upgrading of RC School Buildings in Taiwan Acknowledgement z z z Prof. Shyh-Jiann Hwang z Prof. of Department of Civil Eng., National Taiwan University Division Head of Building Structures, NCREE z z National Science Council Ministry of Education National Applied Research Laboratories National Center for Research on Earthquake Engineering (NCREE) County Governments of Hualien, Yunlin, Taoyuan and Tainan NTU, NTUST and NCKU SEEBUS 2009 Outline Damages of School Buildings z Damages of School Buildings z Seismic Evaluation z Seismic Retrofit z In-Situ Tests z Strategy for Seismic Upgrading z Conclusions Taiwan A Living Laboratory of Earthquake Geography Setting < M < < M 4.5 < M < < M < 6.5 Taiwan 25.0 Taipei North Latitude Area: 36,000 km2 Population: 23 million Taiwan Tectonic structure of Taiwan Seismicity (A.D ) East Longitude

3 SEISMIC FRAGILITY ASSESSMENT OF BUCKLING RESTRAINED BRACED FRAMES Heui-Yung CHANG 1 In comparison with a moment resisting frame (MRF), steel braced frames such as eccentrically braced frame (EBF) and buckling restrained braced frame (BRBF) have superior performance in reparability and costs. Since that, the use of EBF and BRBF has rapidly grown in seismically active zones, such as Taiwan, Japan and North America. On the other hand, towards the development of performance-based earthquake engineering (PBEE), there is a need of assessing the structural vulnerability and seismic risk on a reliability basis. Fragility assessment plays a key role in the assessment. In this paper, the reliability frame work used for MRFs has been extended and employed to assess the seismic performance and fragility of BRBFs. The analyzed 6-story steel office building uses chevron BRBFs in the central bays and has representative portions for a building found in Taiwan. This example also helps confirm the confidence with which BRBFs might achieve the seismic performance expected of new MRFs. Keywords: buckling restrained brace; steel frames; seismic performance; fragility analysis. The performance of steel framed buildings can vary significantly depending on the adapted lateral load resisting systems. For example, a traditional, concentrically braced frame (CBF) may have poor performance owing to brace buckling (Osteraas and Krawinkler1989; Kim and Goel 1992; Tremblay et al 1995 and 1996). During the past few decades, considerable efforts have been made to improve the performance of CBFs (e.g. Kim and Goel 1992). These efforts also have led to the development of eccentrically braced frame (EBF) (e.g. Malley and Popov 1984) and buckling restrained braced frame (BRBF) (e.g. Black et al 2004). In comparison with a moment resisting frame (MRF), the EBF and BRBF have superior performance in reparability and costs. Since then, the use of EBF and BRBF has rapidly grown in seismically active zones, such as Taiwan, Japan and North America. Towards the development of performance-based earthquake engineering (PBEE), there is a need of assessing the structural vulnerability and seismic risk on a reliability basis. The assessment requires characterization of earthquake hazard, estimation of structural demand and capacity, and identification of structural damages and losses associated specific damage states. A key ingredient of this assessment is the fragility, which describes the probability of failure to meet a performance objective as a function of seismic demand on the structural system (Yun et al 2002; Wen et al 2003; Celik and Bruce 2009). In this paper, the reliability framework used for MRFs has been extended and employed to assess the seismic performance and fragility of BRBFs. The analyzed 6-story steel office building uses chevron BRBFs in the central bays and has representative portions for a building found in Taiwan. This example also helps confirm the confidence with which BRBFs might achieve the seismic performance expected of new MRFs. 1 Assistant Professor, Department of Civil and Environmental Engineering, National University of Kaohsiung, Taiwan, hychang@nuk.edu.tw

4 CYCLIC TEST OF HIGH STRENGTH STEEL BEAM-TO-COLUMN CONNECTION COMPOSED WITH KNEE-BRACE DAMPER AND FRICTION DAMPER CONNECTED BY HIGH-STRENGTH BOLTS Atsushi SATO 1, Kei KIMURA 2, Keiichiro SUITA 3, and Kazuo INOUE 4 Recently, high strength steel H-SA700A for building structures was developed. This steel is used only under elastic and full-penetration is not considered. The purpose of this study is to develop weld-free beam-to-column connection for H-SA700A high strength steel members, and verify the structural behavior of the proposed connections. Knee-brace damper and Friction damper were proposed and used to connect beam and column by high-strength bolts. The trigger level of these dampers was designed to remain beam and column as elastic. Full-scale testing of knee-brace damper showed that damper yielded at the target level and all part of connections performed as expected. Friction damper, aluminum alloy is sprayed at the surface, obtained 0.7 as the slip coefficient and a stable elasto-plastic structural behavior was observed after the slippage. Consequently, validity of the proposed design procedure for knee-brace and friction damper beam-to-column connections were verified from the full scale testing. Keywords: high strength steel; beam-to-column connection; high-strength bolt connection; damage control structure; knee-brace damper; friction damper. Recently, high strength steel H-SA700A for building structures was developed to keep the members elastic under the huge earthquake. Specified yield stress is 700N/mm 2 and yield ratio of this steel is less than 98%; therefore, inelastic behavior is not considered. Additionally, H-SA700A is not considering full-penetration. The objective of this paper is to develop beam-to-column connection for these steel members, and verify the structural behavior of the proposed connections. The structural system considering in this study is damage control structures which beam and column remain elastic at the ultimate stage to prevent collapse. To achieve this objective, following requirements have to be satisfied. (1) H-SA700A is not considering welding; therefore, beam-to-column connection should be weld-free connection. (2) Structure fuse, i.e., damper, have to be composed in the connection to avoid beam and column overloaded and remain elastic. Two types of weld-free beam-to-column connections are proposed in this paper. One is knee-brace damper beam-to-column connection. Knee-brace is designed as buckling restrained brace and beam and column is connected through this damper (Suita et al. 2003). Elasto-plastic behavior of knee-brace damper will absorb energy and also control the maximum force occurs in beam and column. Another is friction damper beam-to-column connection. Aluminum alloy friction surface bolted connection will be used as a damper and also control the maximum force occurs in beam and column. For both connections, appropriate details are proposed and full scale testings are conducted to verify the structural performance. 1 Assistant Professor, Department of Architecture and Architectural Engineering, Kyoto University, Japan, asato@archi.kyoto-u.ac.jp 2 Structural Engineer, Nippon Steel Corporation, Japan, kimura.kei@nsc.co.jp 3 Associate Professor, Department of Architecture and Architectural Engineering, Kyoto University, Japan, suita@archi.kyoto-u.ac.jp 4 Professor, Department of Architecture, Okayama University of Science, Japan, inoue@archi.ous.ac.jp

5 SEISMIC PERFORMANCE OF ENHANCING LINK BEAM AND COLUMN CONNECTIOS IN ECCENTRICALLY BRACED FRAMES Chin-Tung CHENG 1 and Ten-Yu WANG 2 The moment connection of link beam to the column in eccentrically braced frame is still prohibited by current code regulations due to its high demands on the strength and deformation capacity. In this research, new designs of link beam to column connection are proposed. Five intermediate length of link beam to column connections are constructed and tested. There are three link beams with 1.4m length and two link beams with 1.6m, respectively. Investigated parameters also include methods for protecting the flange welds in the beam to column connection: tapered cover steel plate or reduced beam section. Connection of beam web to the column face is investigated by using either bolts or full penetration groove welds. In this research, seismic performance of newly designed link beam to column connections is evaluated. Keywords: seismic performance; link beam; beam-column connection; eccentrically braced frame. Eccentrically Braced Frame (EBF) has been investigated over 30 years and its design guidelines can be found in the codes (AISC, 1997). The researches on EBF started at Berkeley in 1977, where Roeder and Popov 1978 conducted a series of test on a one-third scale three-story EBF frame. Test results showed that EBF had excellent ductility and energy dissipation capacity. They suggested the link beam firstly yielded by the shear force rather than by the moment and the brace should sustain a maximum shear force 1.5Vp remaining elastic without buckling. In 1983, Hjilmstad and Popov continuously investigated seismic performance of EBF frame. Test results indicated that heat affected zone in the welds of flange-to-column face ruptured in a brittle manner when both end moments of the link beam reached its plastic strength. They suggested moment connection of the link beam to column face should be prohibited, except it is verified by the tests. Till now, this regulation is still accepted by most modern codes. In 1986, Kasai and Popov 1986a tested several link beam-to-column moment connections. Investigated parameters included the length of link beam, axial force, shear force and moment interaction, strain-hardening effect on the moment redistribution at both beam-ends and energy dissipation. Kasai and Popov, 1986b also proposed a method to calculate the spacing of the stiffener to prevent buckling of the beam web. Engelhardt and Popov, 1989 recommended the length of moment links should be less than 1.6Mp/Vp to prevent a sudden rupture of flange welds to column face. They also suggested the flange of moment links should be stiffened to prevent buckling within 1.5 times flange width from the column face. In 2005, Richards and Uang investigated the effect of flange width-thickness ratio on EBF link cyclic rotation capacity. Since rolled shapes of A36 steel have been replaced by those of A992 steel after Northridge earthquake, many efficient A992 wide flange sections are disqualified by the width-thickness ratio requirement. More than 100 finite element models, representing the spectrum of short, intermediate and long links were analyzed under cyclic loading. It was found that 1 Professor, Department of Construction Engineering, National Kaohsiung First University of Science and Technology, Taiwan, ctcheng@ccms.nkfust.edu.tw 2 Graduate Student, Department of Construction Engineering, National Kaohsiung First University of Science and Technology, Taiwan

6 APPLICATION OF ON-LINE RECURSIVE LEAST-SQUARES METHOD TO PERFORM STRUCTURAL DAMAGE ASSESSMENT OF BENCHMARK MODELS Shih-Yu CHU 1 and Shih-Chieh LO 2 The aim of system identification is to find a mathematical model and to determine its parameters using the input and output signals. Furthermore, the identification of structural damage is an important objective of health monitoring for civil infrastructures. The upmost goal is to provide the ability to assess damage in real-time or immediately after a catastrophic earthquake. Any methodology that satisfies the abovementioned criteria would be able to activate the early warning system and to determine which structures are safe by the building authorities. In order to validate the feasibility of fast assessment capability, both classical least-squares (LS) and on-line recursive least-squares (RLS) methods are implemented in this study to investigate the recorded strong-motion data of a three-floor shaking table benchmark model tested at NCREE in Taiwan. Computation of the least-squares estimation can be arranged recursively so that the estimated parameters at previous step can be used to predict the responses at current time. The one-step ahead predicted error between estimated response and measured response is calculated by the RLS method and the dynamic properties of system can be identified as well. By observing the variations of the identified time-varying modal properties of benchmark model, global damage behavior due to weak element or failure of components can be revealed. Keywords: structural health monitoring; damage assessment; system identification; adaptive identification; recursive least-squares identification technique. Generally speaking, system identification algorithms fall into two categories depending on whether they operate on the data in time domain, or in the frequency domain. Frequency-domain algorithms involve averaging temporal information, thus discarding most of their details. For structural systems whose parameters are expected to degrade with time, this trade-off of temporal information for frequency information is not always justifiable. Over the past decades, a few time domain methods have been transferred to civil engineering applications (Shinozuka et al. 1982; Ghanem et al. 1991). In order to facilitate a comparison of available system identification and damage detection techniques on a common basis, the ASCE Task Group on Structural Health Monitoring has established a benchmark problem by applying the simulated data (Johnson et al. 2004). A special issue of ASCE Journal of Engineering Mechanics has been published for this Phase-I benchmark problem (Bernal and Beck 2004), in which various time-domain approaches have been used. Bernal and Gunes (2004) proposed a flexibility based identification technique that involved sub-matrix inversions and full measurements. They used eigensystem realization algorithm (ERA) (Juang 1994) together with an Observer Kalman Filter Identification algorithm (OKID) (Juang 1994) to perform modeless identification when the input is known, and used a subspace identification algorithm when the input is stochastic. Caicedo et al. (2004) and Lus et al. (2004) also presented ERA based approaches in this special issue. While the former identified modal parameters before using least-squares optimization to locate and identify damage, the latter used an OKID estimator and a nonlinear optimization approach based on sequential quadratic programming techniques to identify a baseline model. The classical least-squares (LS) method has been used extensively to identify constant parameters; however, it 1 Assistant Professor, National Cheng Kung University, Taiwan, sychu@mail.ncku.edu.tw 2 Ph.D. Candidate, National Cheng Kung University, Taiwan, n @mail.ncku.edu.tw

7 LATERAL STIFFNESS AND NATURAL PERIOD EVALUATION OF FLAT PLATE TALL BUILDINGS FOR WIND DESIGN Je-Woo PARK, 1 Hongjin KIM 2, Young-Ju JANG 3, and Ji-Hoon LEE 4 Wind-induced vibration is one of the important structural design factors for serviceability of tall buildings. In order to evaluate the reliable wind-loads and wind induced-vibration, it is necessary to obtain the exact natural period of buildings. The discrepancy in the natural period estimation often results in the overestimation of wind loads. In this study, the effectiveness of lateral stiffness estimation method for tall buildings with flat plate system is evaluated. For this purposed, the results of finite element analysis of three recently constructed buildings are compared with those obtained from field measurement. For the analysis, factors affecting on the lateral resistance such as cracked stiffness of vertical members, elastic modulus of concrete, effective slab width, and cracked stiffness of link beam are considered. Form the results, it is found that the use of uncracked stiffness and application of dynamic modulus of elasticity rather than initial secant modulus yields closer analysis result to the as-built period. Keywords: tall building, flat plate system, natural period, Wind design, serviceablity. The necessity for measures against wind load is increasing as buildings becomes taller and the accurate calculation of design wind load and serviceability evaluation in relation with wind-induced vibration becomes more important structural design factors for such measures. The calculation of wind load and review of serviceability can be performed in design stage by calculating design load and corresponding response acceleration through design criteria or wind tunnel test with specific return period (Kim et al., 2005). At this time, the forecast of accurate natural periods has serious influence on enhancement of reliability of wind load and wind-induced vibration (Simiu et al., etc., 1996). Though natural periods can be generally calculated at design stage through eigenvalue analysis, there is serious discrepancy in calculating accurate natural period of an actual building because of difficulty in modeling non-structural members. According to study result, it is reported that measured natural frequency is larger than analysis result by 20 % in average (Yoon et al., 2003, Shinoya, 1996). Also, though formula for calculation of natural period are proposed by "Korean Building Code" (AIK, 2005) and Building Code Requirements for Structural Concrete (ACI, 2008), they are difficult to be applied to wind-resistant design as it is a natural period calculation formula for design of wind load but for seismic design. In case of flat plate system recently constructed in Korea, beam that is one of lateral resistance factors is omitted and accordingly it is different from the other structural system in designing lateral resistance for wind load. As flat plate system is a system where the lateral load is resisted by vertical members and slabs, there exist two-dimensional slab members and one-dimensional column model at the same time. Therefore dissimilarity arises in calculating lateral stiffness in comparison to the shear wall structures or moment frame structures. In calculating the lateral stiffness and natural period of flat plate system, the part which shows the biggest difference from general structural system is the effective stiffness of slabs. For the estimation of effective 1 Graduate student, School of Architecture & Civil Engineering, Kyungpook National Univ., Korea, unmetoo@knu.ac.kr 2 Assistant Professor, School of Architecture & Civil Engineering, Kyungpook National Univ., Korea, hjk@knu.ac.kr 3 Graduate student, School of Architecture & Civil Engineering, Kyungpook National Univ., Korea, mansky43@knu.ac.kr 4 Graduate student, School of Architecture & Civil Engineering, Kyungpook National Univ., Korea, lotus2918@knu.ac.kr

8 Computer-aided Tool for Seismic Evaluation of Existing RC Buildings * Shang-Hsien HSIEH 1, Ming-Der LU 2, and Kuang-Wu CHOU 3 This paper reports the development of a computer-aided tool that can effectively help engineers carry out seismic evaluation of existing RC Buildings. This tool takes advantage of the pushover analysis capability of commercial structural analysis software, currently either ETABS or MIDAS, and integrates all tasks needed to perform the seismic evaluation in a smooth and automatable way with a friendly user interface. In addition, this tool provides great flexibility and transparency for the user to manage and visualize all the tasks and their results in the seismic evaluation process. Keywords: Earthquake Engineering; Seismic Evaluation; Pushover Analysis; Computer-Aided Tool. Performance-based seismic evaluation methods have become popular in recent years. In general, these methods require pushover analysis that establishes a structure s capacity spectrum to estimate its nonlinear behavior. For the pushover analysis to bring out accurate structural behavior, the plastic hinges at the beam or column ends should have their nonlinear behavior defined as precisely as possible. Although both ATC-40 (1996) and FEMA-273 (1997) seismic evaluation methods propose approaches for definition of plastic hinges, these approaches are mainly for design purposes and often cannot accurately model the nonlinear behavior of plastic hinges of the existing buildings, especially those with poor design or construction. Therefore, engineers need to define their own plastic hinges that can better capture nonlinear behavior. In Taiwan, ETABS (Computers and Engineering, 2009) is popular structural analysis software for buildings and it allows users to define and configure the behavior of their plastic hinges. Engineers usually exploit spreadsheet software like EXCEL or write small programs like MATLAB scripts to generate the plastic hinge properties that ETABS requires. Without the help of any computer-aided tool, the engineers then need to go through a time-consuming and tedious manual process to input those properties into ETABS. Therefore, this research develops a computer-aided tool that automates not only plastic hinge definitions for the ETABS inputs but also the post-processing of the ETABS outputs for seismic evaluation of existing RC buildings. Although the tool has recently been extended to allow for the use of MIDAS (MIDAS/Civil, 2009) as structural analysis software, the discussions in this paper focus on the use of ETABS. * Extended from Hsieh, S. H., M. D. Lu, and K. W. Chou (2009). A Computer-aided Tool for Seismic Evaluation of RC Buildings, Proceedings of the 22th KKCNN Symposium on Civil Engineering, October 31 November 2, 2009, Chiang Mai, Thailand, Professor, Department of Civil Engineering, National Taiwan University, Taipei, Taiwan, shhsieh@ntu.edu.tw 2 Formerly Doctoral Research Associate, Department of Civil Engineering, National Taiwan University, Taiwan; Currently with Mighty Power Solutions Corp., mdlu@mightypower.com.tw 3 Associate Technologist, National Center for Research on Earthquake Engineering, Taipei, Taiwan, kwchou@ncree.org.tw

9 MODAL PROPERTIES OF TALL REINFORCED CONCRETE BUILDINGS BASED ON FIELD MEASUREMENT AND ANALYTICAL MODELS Ji-Young KIM 1, Eunjong YU 2, Seung-Nam KIM 3, and Hong Jin KIM 4 Modal properties are the key parameter to determine the seismic and wind loading of tall flexible structures. In this study, natural frequencies and associated mode shapes of three tall RC buildings were obtained from measured acceleration data using system identification technique. Subsequently, finite element (FE) models for tall reinforced concrete buildings were built using a popular PC-based finite element analysis program and calibrated to match their natural frequencies and mode shapes to measured values. The modification of the FE models for calibration included: i) consideration of the effect of beam-end-offsets, ii) modeling of floor slabs instead of using rigid diaphragm assumptions, iii) inclusion of major non-structural components such as plain concrete walls and cement brick walls, and iv) the increment of elastic modulus of concrete by using the mean compressive strength of mix design rather than the specified concrete strength. Estimates of modal properties from the calibrated FE models and the measured values showed remarkable agreement in all the buildings. Keywords: finite element modeling, structural analysis, natural frequency, full-scale field measurement Dynamic properties including the natural frequencies, mode shapes, and modal damping ratios are key parameters to determine the seismic and wind loading of tall flexible structures. Many experimental studies based on field measurements have been carried out to appreciate their dynamic properties. Through regression analyses of the collected data, a number of equations have been proposed to estimate the dynamic properties [Eurocode, ISO, AIJ, Jeary]. For natural frequencies, however, the actual values vary greatly depending on the size and shape of a structure and its configuration. Hence, it is more reasonable to use estimates from the FE model rather than the values from the proposed equations. Here, the use of appropriate assumptions is necessary because the estimates can differ greatly depending on the modeling assumptions such as the stiffness of the individual components and the extents of components included in the FE model. When inappropriate assumptions are made in the modeling, the predicted natural frequencies show significant differences with measured values. Table 1 is a list from previous studies comparing the natural frequencies obtained from field measurements with the analytical predictions based on various modeling assumptions. Larger differences are observed in reinforced concrete (RC) buildings. When the RC buildings were modeled only with the bare frame, the FE model considerably underestimated the natural frequencies, which was seemingly caused by simplified modeling techniques that may have been introduced for convenience of modeling and/or conservative design philosophy; for example, ignoring the stiffness of floor slabs and non-structural elements, the use of conservative material property values, and so forth. Generally, when the natural frequency decreases, the estimated wind-induced vibration increases. Hence, underestimated natural frequencies may result in excessive structural costs because an increase of lateral stiffness or damping by stiffening the structural members or using additional damping devices is required to mitigate the vibration level. 1 Researcher, Daewoo Institute of Construction Technology 3 Graduate Student, Hanyang University 2 Assistant Professor, Hanyang University 4 Assistant Professor, Kyungbook Univeristy

10 SHEAR STRENGTH OF PRECAST POST-TENSIONED CONCRETE BEAMS WITH HIGH STRENGTH SHEAR REINFORCEMENT Jaeman LEE 1, Minehiro NISHIYAMA 2, Masanori TANI 3 In Japan, a couple of design equations for shear strength of prestressed concrete members are specified in the Architectural Institute of Japan Standard for Structural Design and Construction of Prestressed Concrete Structures. The design equations have several drawbacks and uncertainties such as the upper limit of specified yield strength of shear reinforcement, which is low from the viewpoint of the current construction technology. The design equations should be improved based on experimental and analytical results. The authors constructed and tested seven precast concrete beams post-tensioned by multiple strands. The nominal yield strength of shear reinforcement is 785 MPa. Cyclic loading simulating earthquake load is applied. In this paper, the shear design equations along with the ACI318 shear design equations are scrutinized based on the loading test results and those found in past literature. Keywords : Prestressed concrete, post-tension, shear strength, high-strength reinforcement, truss model, cyclic loading tests In Japan, a couple of design equations for shear strength evaluation of prestressed concrete members are specified in the Architectural Institute of Japan (AIJ) Standard for Structural Design and Construction of Prestressed Concrete Structures 1 : one is empirical and based on the summation of contributions from concrete, transverse reinforcement and prestressing force (Eq.1-1). The other is based on truss and strut mechanisms analogous to a strut-and-tie model (Eq.1-2). Eq.1-2 assumes that the member should be subjected to earthquake loading shown in Fig.1. The beneficial effect of prestress on shear resistance is considered in a concrete strength factor, ν, which is defined as a ratio of concrete compressive strength in a strut to design concrete compressive strength, F c. { α ( f + 0.1σ ) f ( } b j Qu = s g wy ρw ) o (1-1) bo D Qu = bo joρw fwy + ( νfc 2ρw fwy ) tanθ (1-2) 2 where, α = 4/(a/d+1) (1 α 2), a = shear span, d = effective depth of the section, D = total depth of the section, L = member length, f s = concrete shear strength, σ g= average prestress (= P/b o D), f wy = yield strength of shear reinforcement (f wy 295 (N/mm 2 )), ρ w = shear reinforcement ratio (= A w /b o s), b o = width of the section at the centroid of the section, s = stirrup spacing, A w = sectional area of the stirrups, j = distance between tension and compression force resultants or (7/8)d, j o = distance between tension and compression non-prestressed mild-steel longitudinal bars, ν = concrete strength factor (= αl r (1+σ g /F c ), tanθ = (L / D) L / D, α = 60 / F c, L r = a/2 D, and F c =design concrete compressive strength. 1 Graduate Student, Department of Architectural Engineering, Kyoto University, Japan, jaeman.lee@ax2.ecs.kyoto-u.ac.jp 2 Professor, Department of Architectural Engineering, Kyoto University, Japan, mn@archi.kyoto-u.ac.jp 3 Assistant Professor, Department of Architectural Engineering, Kobe University, mtani@port.kobe-u.ac.jp

11 PLASTIC HINGE LENGTH OF CIRCULAR REINFORCED CONCRETE COLUMNS Yu-Chen OU 1, and Raditya Andy KURNIAWAN 2 This paper presents a parametric study to investigate the plastic hinge length of circular reinforced concrete columns using a three-dimensional finite element analysis method. The Taguchi robust design method is used to reduce experimental efforts. Parameters examined include the longitudinal reinforcement ratio, the column aspect ratio, the axial force ratio, and the concrete compressive strength. Longitudinal reinforcement with yield strengths of 414 MPa (60 ksi) and 689 MPa (100 ksi) are considered in the study. Based on the results of the study, simplified expressions for the plastic hinge length of circular reinforced concrete columns are proposed. Keywords: reinforced concrete columns; circular section; plastic hinge length; Taguchi method. When a column is subjected to large earthquake loading, the end region of the column may experience significant inelastic rotation, forming a plastic hinge. The inelastic rotation results from nonlinear distribution of inelastic curvature over the plastic hinge. For convenience in calculation, the inelastic curvature is assumed to be constant over a length referred to as plastic hinge length L p. The plastic hinge length is a critical parameter for seismic analysis and design of reinforced concrete structures. There have been many researchers suggesting various formulas to estimate the plastic hinge length. However, those expressions are significantly different from one another (Bae and Bayrak 2008). Based on the UW/PEER column database, Bae and Bayrak (2008) propose a new expression for plastic hinge legnth taking into account the effects of longitudinal reinforcement ratio, the column aspect ratio, and the axial force ratio. This paper examines the plastic hinge length of circular reinforced concrete columns using a three-dimensional finite element analysis method. The parameters investigated are the longitudinal reinforcement ratio, the column aspect ratio, the axial force ratio, and the concrete compressive strength. Two yield strengths, 414 MPa (60 ksi) and 689 MPa (100 ksi), are considered for the longitudinal reinforcement. Simplified expressions are proposed for plastic hinge length of circular reinforced concrete columns. General Description THREE-DIMENSIONAL FINITE ELEMENT METHOD A non-linear three-dimensional finite element analysis method employed using a general-purpose finite element software, ABAQUS (HKS 2006), is developed in this research for modeling the behavior of reinforced concrete columns under combined axial and lateral forces. Column specimen 415 examined by Lehman and Moehle (2000) was used to validate the finite element method. The column had an aspect ratio of 4 and 1.5% longitudinal 1 Assistant Professor, National Taiwan University of Science and Technology, Taipei, Taiwan, yuchenou@mail.ntust.edu.tw 2 Master student, National Taiwan University of Science and Technology, Taipei, Taiwan, m @mail.ntust.edu.tw

12 EARTHQUAKE SIMULATION TESTS ON A 1:5 SCALE PILOTI-TYPE LOW-RISE RC RESIDENTIAL BUILDING MODEL Han Seon LEE 1, Dong Wook JUNG 2, Kyung Bo LEE 3, Hee Cheul KIM 4, Young Hak LEE 5, and Ki Hak LEE 6 This paper presents the results of a series of uni- and bi-directional earthquake simulation tests conducted on a 1:5 scale five-story reinforced concrete building model, which represents a residential apartment building possessing the high irregularities of weak story, soft story, and torsion simultaneously in the ground story. Analysis of the test results leads to the following conclusions: (1) The torsional mode is the fundamental mode whereas the period of the second translational mode appears to be close to that of torsional mode; (2) The two orthogonal translational modes acted independently while the torsional mode frequently coincided with the next close translational mode, thereby leading to large inelastic responses; (3) The maximum torsional moment and deformation remained almost constant regardless of the excursion into inelastic behavior in two orthogonal translational motions; And, (4) the resistance and stiffness of the columns and wall increase or decrease greatly with the variation of acting axial forces. Keywords: earthquake simulation test; reinforced concrete; irregularity. Recently, many low-rise residential apartment buildings have been constructed in the densely populated areas of Korea. The lack of sites causes the ground floor be used as a parking lot and the piloti adopted. This type of buildings as shown in Fig. 1 commonly have high irregularities of soft story, weak story, and torsion simultaneously in the piloti story. Observations of the damages to the structures imposed by the severe earthquakes such as 1995 Kobe and 2008 Sichuan earthquake have drawn the conclusion that this type of building structures are vulnerable to severe damage or complete collapse in this piloti story. Many of these buildings have been constructed without considering earthquake-resistant design requirements in Korea. However, new Korean Building Code(KBC) 2005 (AIK 2005) and building law enforce the seismic design of these building structures. This research aims at the investigation of the realistic seismic responses of this type of building structures through earthquake simulation tests. 1 Professor, School of Civil, Environmental and Architectural Engineering, Korea University, Korea, hslee@korea.ac.kr 2 Graduate Student, School of Civil, Environmental and Architectural Engineering, Korea University, Korea, dwjung@korea.ac.kr 3 Graduate Student, School of Civil, Environmental and Architectural Engineering, Korea University, Korea, skyungbo@korea.ac.kr 4 Professor, Department of Architectural Engineering, Kyunghee University, Korea, kimhc@khu.ac.kr 5 Assistant Professor, Department of Architectural Engineering, Kyunghee University, Korea, leeyh@khu.ac.kr 6 Associate Professor, Department of Architectural Engineering, Sejong University, Korea, kihaklee@sejong.ac.kr

13 FEM SIMULATION FOR RC SHEAR WALLS WITH MULTI-OPENINGS Masato SAKURAI 1, Hiroshi KURAMOTO 2 and Tomoya MATSUI 3 This paper outlines two dimensional non-linear FEM analysis of RC shear walls with multi-openings executed to simulate the static loading test result such as hysteresis loops, failure mode and stress distributions of reinforcing bars, while the validity of the analytical method is verified by comparing experimental results with analytical ones. The analytical results showed good agreement with experimental results. It is also clarified that stress transferring mechanisms of RC shear walls of multi-openings were significantly affected by the difference of the number and layout of openings. Keywords: RC shear walls with openings, Static test, Seismic performance, Shear strength, FEM analysis In Japan, the shear strength of RC shear walls with openings are generally evaluated by using an equivalent perimeter ratio of openings in the AIJ design standard for RC structures (AIJ, 1999). Although the opening layouts are different, the shear strength of shear walls with the same value of the equivalent perimeter ratio is to be calculated by using the above method. On the other hand, looking at the existing experimental results and the earthquake damage to RC buildings in the past, the failure mechanisms of RC shear walls with openings are complicated. Especially, in case of those with multi-openings, the quantitative evaluation of the seismic performance is very difficult even in the present. The main objective of this study is to grasp the seismic performance of the walls, such as failure mode, hysteresis characteristic and deformability in order to improve the evaluation method of the shear strength of RC shear walls with multi-openings. Static loading tests of RC shear walls with openings carried out to investigate the influence of different number and layout of the openings in the past (e.g., Suzuki et al and Sakurai et al. 2008). The test results showed that the shear strength, failure mode and deformability of RC shear walls with openings were significantly affected by the difference of the number and layout of openings. Numerical simulation using FEM analysis is very useful to grasp internal stress transferring mechanisms of the RC shear walls, which can not be directly obtained from the test, while the accuracy of the analytical tools used is important. In this study, then, two-dimensional non-linear FEM analysis of RC shear walls with multi-openings was conducted to simulate the experimental results of the shear walls with openings. Through the analysis, the accuracy of the analytical method was confirmed. An outline of the analysis is described and the stress transferring mechanisms of RC shear walls are examined in this paper. 1 Graduate Student, of Dept. of Architectural Engineering. Osaka University, Suita, Japan, sakurai_masato@arch.eng.osaka-u.ac.jp 2 Professor, Dept. of Architectural Engineering. Osaka University, Suita, Japan, kuramoto@arch.eng.osaka-u.ac.jp 3 Assistant Professor, Dept. of Architectural and Civil Engeering, Toyohashi University of Technology, Toyohashi, Japan matsui@tutrp.tut.ac.jp

14 TORSIONAL STRENGTH OF REINFORCED CONCRETE BEAMS SUBJECTED TO PURE TORSION Jung-Yoon LEE 1 Jongwook Park 2 and Sang-Woo Kim 3 This paper presents the results of an analytical study on the performance of reinforced concrete beams subjected to pure torsion. In particular, the effect of the tension stiffening was discussed and included in the analytical study. Although the torsional strength of RC beams according to the existing design codes (ACI , EC2, and JSCE-02) depends on neither the average yield stress of steel bars nor tension stiffening effect, the test results indicated that the steel stress of the beams at peak load increased as the total percentage of reinforcement decreased due to tension stiffening effect. A new equation including the tension stiffening effect was proposed to predict the torsional moment capacities of RC beams. Comparisons between the tested and calculated torsional moments of the seventy-one beams showed reasonable agreement. Keywords: torsional strength; tension stiffening effect; space truss model; reinforced concrete members. The torsional behavior of reinforced concrete members is different from their behavior in flexure. Torsional failure of RC beams is abrupt and does not give sufficient advanced warning if not designed properly, and the diagonal cracks that develop are considerably wider than the flexural cracks. Because of the brittle nature of such failures, the designer has to design sections that are adequately strong to resist the external factored torsion loads without reaching their torsional strength capacity. In addition, the designer has to be able to consider the effect of bending moment while designing for torsion since the torsional moment seldom acts on its own but rather in combination with flexure. The current design codes (ACI (2005); EC2 1992); JSCE-02 (2002)) for torsional design of reinforced concrete (RC) members are developed based on the space truss analogy and the thin walled tube theory. Although these design codes are capable of predicting the torsional behavior of RC members with a reasonable accuracy, some test results indicated that the codes were not successful in encompassing the interactions between concrete and torsional reinforcements in torsional resistance of a RC member. The comparison between the experimental torsional strengths of the RC beams tested by Fang and Shiau (2004) with the torsional strengths calculated by the equations from ACI Code and EC2 Code is plotted in Figure 1. Since the codes determine the torsional strength at the yield strength of the reinforcements, the values of those specimens in which the yielding of both transverse and longitudinal steel bars took place were chosen for the comparison. In addition, no safety factors were considered in all torsional strength calculations by the codes. We note that all the beams tested by Fang and Shiau had a cross section of 350 x 500 mm, with the transverse and longitudinal reinforcements arranged according to the design provisions of ACI Associate professor, Sungkyunkwan University, Republic of Korea, jylee@skku.ac.kr 2 Graduate student, Sungkyunkwan University, Republic of Korea 3 Research assistant professor, Kongju National University, Republic of Korea

15 CONFINEMENT EFFECT OF CONFINED CONCRETE WITH HIGH-STRENGTH SPIRAL REINFORCEMENT Sang-Woo KIM 1, Young-Sik KIM 2, Jung-Yoon LEE 3, and Kil-Hee KIM 4 * This study investigated the influence of concrete compressive strength for the lateral confinement of high-strength spiral reinforcement. The main test parameters were the compressive strength of concrete and the yield strength of spiral reinforcement. A total of 36 cylindrical test specimens with a diameter of 150mm and a height of 300mm were cast and tested under monotonic concentric compression. To effectively evaluate the confinement effect of spiral reinforcement, the external diameter of spiral reinforcement was the same as the diameter of specimen. The experimental result indicated that the influence of the strength of spiral reinforcement on lateral confinement effect was reduced as the compressive strength of concrete increased. Keywords: confinement; confined concrete; high-strength; stress-strain relationship; spiral reinforcement. Modern structures are following the trend of higher, bigger, longer-span and specialized ones due to the industrial concentration, dense population, and the upgraded demand of users in modern society. In response to this new demand, it is necessary to develop and apply the high-strength materials, such as high-strength concrete and steel reinforcement. As the technique to satisfy the ductility as well as the strength of the steel bars has been improved, the interest in the application of high-strength reinforcements for reinforced concrete (RC) members has also intensified. The use of high-strength reinforcement is very effective, because it is able to not only avoid the reduction of construction performance due to the overcrowded arrangement of reinforcements but also obtain the diversity of structural design. Especially, because densely arranged steel bars in RC columns with seismic detailing has been pointed out as a factor to deteriorate the construction performance, research on the use of high-strength reinforcement for RC structures is very important as an alternative. Research on the behavior of confined concrete with lateral reinforcements has been actively investigated since the analytical study performed by Mander et al. (1988). As the high-strength concrete highlighted, many researchers (Ahmad and Shah 1982; El-Dash and Ahmad 1995; Issa and Tobaa 1994; Razvi and Saatcioglu 1999) have been studied to evaluate the confinement effect of transverse reinforcement in high-strength concrete. However, there are a few researches on the confinement effect of high-strength reinforcement in comparison to that of high-strength concrete. Especially, research on the interaction of the compressive strength of concrete and the yield strength of lateral reinforcement is rare. Accordingly, this research empirically evaluated the confinement effect of high-strength steel spirals according to the compressive strength of concrete. 1 Research Assistant Professor, Department of Architectural Engineering, Kongju National University, Korea, swkim91@kongju.ac.kr 2 Graduate Student, Department of Architectural Engineering, Kongju National University, Korea, nancy800@kongju.ac.kr 3 Associate Professor, Department of Architectural Engineering, Sungkyunkwan University, Korea, jylee@skku.ac.kr 4 Assistant Professor, Department of Architectural Engineering, Kongju National University, Korea, kimkh@kongju.ac.kr

16 BIDIRECTIONAL STRUCTURAL CONTROL BY A LIQUID DAMPER Sung-Kyung LEE 1, and Kyung-Won Min 2 Most researches relating to structural control by liquid dampers have concentrated on mitigating a unidirectional response of structures subjected to dynamic loads such as earthquake and wind. To reduce bidirectional responses of structures, two dampers should be installed along two principal axes of structures. A TLCD produces the control force by a reciprocal motion of liquid within a tank in the direction of longitudinal axis, while it generates the control force by the liquid sloshing in the direction of lateral axis. Accordingly, a TLCD has a similar feature with a rectangular tuned liquid damper (R-TLD) in the direction of lateral axis. This study presents bidirectional structural control using only single tuned liquid column damper (TLCD). Shaking table test is performed to experimentally investigate the dynamic characteristics of a TLCD under white noise excitation. Its dynamic characteristic is expressed by the transfer function from the shaking table acceleration to the control force generated from a TLCD. The analytical dynamics of TLCD is derived from the equivalent linearized TMD model. As a TLCD is excited with an inclined incident wave, its coupled dynamics is expressed in terms of those of both TLCD and R-TLD. Finally, the key parameters of both TLCD and TLD are identified based on the coupled dynamics proposed in this study, which include the mass ratio of horizontal liquid column to total liquid for a TLCD, and the participation factor of the fundamental liquid sloshing for a R-TLD and damping ratio for both cases. Keywords: liquid damper; tuned liquid column damper; tuned liquid damper; bidirectional structural control; shaking table test; coupled dynamics. During the last few decades, liquid-type vibration absorbers have been applied to wind-induced vibration control of many high-rise building structures. Especially, tuned liquid column dampers(tlcds) have been proven to be effective for reducing responses of building structures subjected to dynamic loads by many researchers (Soong and Dargush 1997). Hitchcock et al. have developed a special type of TLCD, called as liquid column vibration absorbers(lcvas), of which sectional areas of the horizontal and vertical columns are different with each other (Hitchcock et al. 1997). They have also developed a LCVA which can be applied to reduce bidirectional responses of building structures, and have verified its control performance through a series of dynamic test. In terms of simplicity in manufacturing and installation, a TLCD has dominant advantages over other passive types of energy-dissipating dampers. For example, its natural period is easily tuned to that of a structure by adjusting a liquid column length. Also, its inherent damping is simply determined by regulating the head loss coefficient that depends on the size of an orifice to resist to the flow of liquid in the horizontal column. Recent researches on TLCDs have been focused on their optimal designs (Chang and Hsu 1998), design guidelines (Wu et al. 2005) and practical application to real building structures (Hochrainer 2005; Hochrainer and Ziegler 2006). Most researches relating to structural control by liquid dampers have concentrated on mitigating a unidirectional response of structures subjected to dynamic loads such as earthquake and wind. To reduce bidirectional responses of structures, two dampers should be installed along two principal axes of structures. A TLCD 1 Research Professor, Department of Architectural Engineering, Dankook University, Korea, sungkyunglee@dankook.ac.kr 2 Professor, Department of Architectural Engineering, Dankook University, Korea, kwmin@dankook.ac.kr