Developments of the Current Performance-Based Seismic Design Code for Buildings in Taiwan

4th International Conference on Earthquake Engineering Taipei, Taiwan October 12-13, 2006

Paper No. 223

DEVELOPMENTS OF THE CURRENT PERFORMANCE-BASED SEISMIC DESIGN CODE FOR BUILDINGS IN TAIWAN

Qiang Xue1 Chia-Wei Wu2 Cheng-Chung Chen3 Kuo-Ching Chen4

ABSTRACT

Performance-based design has been widely recognized as an idealized method for future seismic design in practice. A draft seismic design code for buildings by introducing performance-based design methodology has been proposed in this research. First of all, the current code provisions have been examined according to the theoretical basis of PBSD. Several key issues have then been investigated to clarify why, what and how to incorporate the methodology of PBSD in domestic seismic design code currently. In the proposed draft code, transparent seismic design objectives and criteria for buildings of different use group have been established quantitatively. Drift limit for each considered earthquake level is clearly established. Site feasibility requirement, conceptual design scope and basic rule have been modified considering use group. A performance objective oriented preliminary design procedure is proposed. Direct displacement-based design procedures have been summarized in the appendix. Methods and procedures for seismic performance evaluation have been presented. Finally in this paper, what we have learned through the applied examples will be summarized. We believe that performance-based seismic design technology will bring a new era to engineering practice with increased confidence and reliability on seismic safety.

Keywords: Performance, Design, Code

INTRODUCTION

Taiwan is mainly located in an unstable tectonic zone between the Philippine Sea and Eurasian Plates. It has a very rich seismic history. Seismic safety has been our major concern. Similar as in other countries, strength and ductility are the most important factors in the local seismic design code. Serviceability under low earthquake level only is requested to check. After the severe damage and a large amount of life and economic loss caused by the Chichi earthquake in Sep. 21 1999, the need of performance-based seismic design (PBSD) (Xue, 2000) of structures has been recognized by professionals. Since then, in addition to keeping an eye on the international technology development in this area, researchers in local research organizations such as Sinotech Engineering Consultants Incorporation and National Center for Research on Earthquake Engineering and in the universities such as National Taiwan University, etc. have put a lot of effort on performance-based seismic evaluation of existing structures and performance-based seismic design of new structures. Seminars or workshops have also been held to introduce such concept to engineering practice. In 2003, more and more projects regarding seismic evaluation and retrofit design are requested to introduce performance-

1 Research Fellow and SE Section Chief, Civil Engineering Research Center, Sinotech Engineering Consultants Inc., Taiwan, [email protected] 2 Research Fellow, Civil Engineering Research Center, Sinotech Engineering Consultants Inc., Taiwan 3 Research Fellow and IT Section Chief, Civil Engineering Research Center, Sinotech Engineering Consultants Inc., Taiwan 4 Manager, Civil Engineering Research Center, Sinotech Engineering Consultants Inc., Taiwan based seismic design methodology. On recognizing an increasing need in performance-based seismic design guideline or provisions for local engineers to follow, in 2003, Architecture and Building Research Institute (ABRI), Ministry of Interior, sponsored Sinotech Engineering Consultants Incorporation to propose a tentative framework of performance-based seismic design code, which must be currently applicable to engineering practice. The results of this study were approved by an advisory committee, which was composed of professional specialists, before they are included in the final report. Following the framework, ABRI sponsored another project for the same research group to propose a tentative seismic design code for buildings by introducing performance-based design methodology. The research includes two parts, Part 1, Provisions, Commentary and References and Part 2, Examples. This paper focuses on part 1. Seismic performance of a building is associated with both structural and non-structural systems and elements. In this research, a non-structural system is designed to accommodate acceleration or drift after completion of the structural system design as in the current design code. No economic loss has been employed as a quantified design criterion. This research focused on the structural performance. Issues regarding non-structural systems, seismic evaluation of existing buildings, buildings including base isolation system and energy dissipative devices, design review, construction quality assurance and building management and maintenance are out of the scope of this paper. Due to paper length limit, examples are not given. Rather, some conclusions are drawn.

KEY ISSUES

The major difference of PBSD from the traditional design methodology is summarized as the following 6 issues. 1) Multi-level seismic hazard is considered with emphasis on the transparency of performance objectives. 2) Building performance is interpreted in terms of both structural and non- structural damage. Building performance is guaranteed through limited elastic or inelastic deformation in addition to strength and ductility. 3) Seismic design is oriented by performance objectives. 4) An analytical method, through which the structural behavior, particularly the nonlinear behavior is rationally obtained, is employed. 5) The designed building will meet the prescribed performance objectives reliably with accepted confidence. 6) The final design result must ensure the minimum life- cycle cost. Since performance-based design covers a wide range of fields, introducing performance- based design methodology into seismic design code for buildings must be done stage-by-stage. The current version focused on the preceding first 4 issues through which reliability of such designed buildings to meet the performance objectives is believed to increase.

PERFORMANCE OBJECTIVES AND CRITERIA

The purpose to establish individual performance objective for buildings of different seismic use group is to provide a platform for information exchange between the owners and the designers so that they can achieve a common view about the structural performance under different level of earthquake excitation. Therefore, transparency of the design objectives is very important. In the current seismic design code, each performance goal of an important building is adjusted by increasing the considered earthquake level through an importance factor while the expected performance level is maintained. This is the case even for a performance goal with respect to an earthquake level, which is equal to 2500 years of return period. Without seismic hazard analysis, designer may not know the return period of such amplified earthquake level. Even if the return period of such an earthquake level is available, no body can tell the corresponding intensity level exactly because 2500 years of return period corresponds to the maximum intensity level for most places in Taiwan (Loh and Hwang, 2002). Referred to the latest code or supporting reports in the USA (IBC2003, 2003, FEMA450, 2003), Japan (JSCA, 2000), New Zealand and Australia (AS/NZS 1170) (King and Shelton, 2004) and Canada (NBCC) (Heidebrecht, 2004), a good alternative to adjust the performance goal of an important building is to decrease the expected performance level while keeping the considered earthquake level. No importance factor is needed. Three seismic hazard levels (Fig. 1) are considered and can be distinguished by return period, probability of exceedence or corresponding site intensity. Performance has been classified as 5 levels, Operational (OP), Immediate Occupancy (IO), Damage Control (DC), Life Safety (LS) and Collapse Prevention (CP). For each building, a performance objective is composed of three performance goals, each of which is constructed by one considered seismic hazard level and an expected performance level. Performance objectives for seismic use group I、II、III are depicted in Fig. 1.

Figure 1. Performance Objectives

The purpose to establish performance criteria is for engineering design. Performance criteria are the quantified acceptance criteria to meet the performance objective interpreted by both quantified hazard levels and quantified performance levels. Each seismic hazard level is quantified by an elastic design spectrum. Each performance level is quantified by parameters associated with strength, stiffness and ductility. Regarding strength, OP corresponds to an elastic behavior. Overstrength must be ensured for other performance levels and no large strength degradation beyond ductility limit. No weak story exists and the structure has enough vertical capacity. Regarding ductility, the concept of inelastic displacement demand ratio (IDDR) (SEAOC, 1999) is employed. IDDR stands for the ratio of inelastic displacement demand over the ultimate inelastic displacement capacity. Acceptable IDDR values associated with structural system performance levels OP, IO, DC, LS, CP are 0, 0.2, 0.4, 0.6 and 0.8, respectively. For structural members, no less than 80% elements are needed to meet the same criteria. Regarding stiffness, the maximum inter-story drift ratio (IDR) is considered to limit building lateral displacement. It is not easy for everybody to agree with the same drift limit. In this research, based on the references such as SEAOC (1999), JSCA (2000), IBC2003 (2003), FEMA450 (2003), AS/NZS 1170 (King and Shelton, 2004) and NBCC 2005 (Heidebrecht, 2004) and according to the local seismic hazard analysis results and opinions from the advisory committee of this project, IDR limit as in Table 1 is suggested preliminarily. Structural system type in Taiwan is not discussed in this paper and can be referred to the local seismic design code (CPRMI, 2005) or Xue et al. (2005). It should be noted that limit of the maximum inter-story drift ratio is used as performance assessment criterion but may result in an unreasonable demand particularly for a less ductile structural system if a direct displacement based design method is used for preliminary design. Table 1. Allowable maximum inter-story drift ratio with respect to each performance level

Performance Level Structural system CP LS DC IO OP Systems with brick shear walls 0.009 0.007 0.007 0.007 0.005 Other systems 0.025 0.020 0.015 0.010 0.005 An obviously significant change arising from the current performance criteria is an increase in structural stiffness demand for buildings of higher seismic use group or in other words, for more important buildings. It is important to point out that the code suggested performance objective and criteria provides the minimum protection to the buildings. Increase of performance objective is allowed upon the owner’s acceptance.

SITE FEASIBILITY

Site feasibility study is to ensure that the established performance objective can be met on the construction site. Similar as the current code, this research focus on the determination of very weak soil layer and liquefaction potential of sandy soil so that the coefficient of sub-grade reaction can be reduced by a reduction factor DE . Under EQ1 level earthquake, liquefaction is not acceptable. For EQ2~EQ3 levels, soil liquefaction may occur as long as structural performance is maintained. The allowable degree of liquefaction is limited by the maximum allowable value of DE , which is set to be 0, 1/3 and 2/3 for buildings of seismic use group I, II and III, respectively. If this limit has been exceeded, soil improvement must be applied. For buildings of seismic use group II and III, back off distance from type 1 active fault and other hazard such as landslide must be taken into consideration.

CONCEPTUAL DESIGN

Basic conceptual design rules are given with emphasis on redundancy and uniform continuity of strength, stiffness and ductility, respectively. Structural systems have been classified as four types, namely, the load-bearing walls, frame systems, moment resisting frames and dual systems. Ductility capacity of each system is represented by the ultimate ductility ratio R according to their relative capacity to dissipate energy after yielding. Building height limit, horizontal and vertical irregularity, expected energy dissipation and yielding mechanism have been defined. Flexibility is given to allow design of special structural systems based on reliable technology.

PRELIMINARY DESIGN

Performance based design implies that the design procedure is oriented by the performance objective. It can be achieved through either an indirect or a direct displacement-based design method. Since approaches of the latter method are not mature enough to be applied directly to various structures, they are summarized in the appendix for reference only before detail parametric and case studies. Alternatively, the former method, which employs a force-based design procedure combined with a check on the inter-story drift ratio, is adopted.

Performance objective oriented design by force-based method Similar as traditional force-based design method, fundamental period of the structure system is given by empirical formula. Seismic design force is estimated by an elastic design force reduced by a reduction factor Fu associated with an allowable ductility ratio corresponding to each performance level Ra,PL . Various format of elastic design spectra are given for each refined seismic zone. Newmark-Hall (1982) reduction factors have been used with slight modification according to local seismic hazard analysis in the present design code and in this research. The allowable ductility ratio or the maximum usable ductility associated with each performance level is calculated by Eq. 1.

Ra,PL = 1+ IDDRPL (R −1) (1) where R is the prescribed ductility capacity for a given structural system and IDDR has been defined before. Subscripts a and PL stand for allowable value and considered performance level, respectively. Since R has been taken by considering some safety margin, it corresponds to 80-90% of the ultimate inelastic displacement capacity. Accordingly, during preliminary design, acceptable IDDR values associated with structural system performance levels OP, IO, DC, LS and CP are revised to be 0, 2/9, 4/9, 2/3 and 1, respectively. For sites within Taipei basin, they are revised to 0,1/6,1/3,1/2 and 1, respectively considering increasing number of earthquake cycles.

Analysis and Design The same as in the current design code, elastic analysis can be used at this stage. Static analysis or modal spectrum analysis can be selected according to building height and system irregularity. Element design can be conducted in the usual way according to the existing design specifications and tools.

Preliminary Check on The Inter-Story Drift Ratio Regarding concerns arising from complexity of inelastic analysis during detail seismic performance assessment in the next step, a simple method for preliminary check on the inter-story drift ratio is employed so that deficiency associated with stiffness can be noticed at an earlier stage. Concept of modal spectrum analysis (Goel and Chopra, 2004; FEMA440, 2004) has been employed. N modes with 90% modal mass participating are considered. Alternatively, N=3 is allowed. Under i earthquake level EQ2 or EQ3, for each mode i with period of Ti , inelastic spectral displacement S d associated with the corresponding performance level is calculated through Eq. 2.

i 2 i Ra,PL ⎛ Ti ⎞ i S d = ∗⎜ ⎟ ∗ S ae (2) Fu ⎝ 2π ⎠

i 1 where S ae is the elastic design spectral acceleration corresponding to Ti . For the first mode, Ra,PL is equal to Ra,PL estimated by Eq. (1) with revised IDDR. For other mode, the same value may be i≠1 i assumed or Ra,PL =1. S d also represents the inelastic target displacement of the equivalent single degree of freedom (ESDOF) system under mode i. Modal superposition is applied by either SRSS or CQC rule to obtain the displacement of ESDOF system. Finally, the maximum inter-story drift ratio is calculated by Eq. 3 (SEAOC, 1999).

IDR = x * (hn k1k2 ) (3) where hn, k1 and k2 are height at roof level, factor of effective height, factor of displacement shape function defined in SEAOC (1999). It should be noticed that this approximate method is very rough and is most suitable for a structure whose behavior is dominated by its first mode. For more flexibility, preliminary design according to the current force-based design method is allowed as long as the performance criteria are acceptable after detail performance assessment. It is suggested to use together with preliminary check on the inter story drift ratio.

DETAILED SEISMIC PERFORMANCE ASSESSMENT

Although the preliminary design is somewhat objective oriented, the assumed seismic ductility capacity may not be developed in the designed structure due to consideration of sizing standard and construction convenience. Besides, nonlinear behavior is not explicitly considered in the procedure. Real behavior of the structure may be different from the design model. Therefore, a detail seismic performance assessment is needed to ensure, although not with 100% reliability, that its seismic performance meet the prescribed performance criteria and objective.

Analysis Method for Performance Evaluation

In this research, the minimum allowable analysis method to be employed for performance evaluation is suggested in Table 2.

Table 2. The minimum allowable analysis method for performance evaluation

Seismic hazard Static analysis is Dynamic analysis must be used in the level allowed in the preliminary design preliminary design T ≤ 3.5Ts T > 3.5Ts EQ1 Linear Static Linear Dynamic Linear Dynamic EQ2 Nonlinear Static Nonlinear Static Nonlinear Dynamic EQ3 Nonlinear Static Nonlinear Static Nonlinear Dynamic Note: material nonlinearity is the only focus in this table.

Analysis and Performance Evaluation Linear static or linear dynamic analysis including modal spectral analysis and time-history analysis can be carried out as usual to evaluate seismic performance under EQ1 level earthquake. Nonlinear static pushover analysis is usually the choice for lots of buildings under EQ2 or EQ3 level earthquakes. In the case when nonlinear time-history analysis is needed, nonlinear dynamic pushover analysis may be carried out to evaluate structural system ductility. Concerning about complexity of nonlinear dynamic analysis, the advisory committee suggested that simplified modal be allowed to use in nonlinear dynamic analysis. This section focused on nonlinear static analysis. Structural Modeling Structural modeling is very important, sometimes may be even more important than selection of the analysis method. Plastic hinge properties of different elements can be adopted from reliable experiments or research reports. Cracking of RC elements are considered in the same way as suggested in FEMA356 (2000). Influence of non-structural walls such as the infill walls to the structural deformation must be considered. Nonlinear Static Pushover Analysis Based on a constant load pattern, nonlinear static pushover analysis is an incremental iterative method to obtain the base shear versus roof displacement relationship. In this research, two types of load pattern are suggested to employ. One of them is proportional to the fundamental mode. The other is a uniformly distributed load pattern proportional to the story mass. Other adaptive load patterns based on reliable studies are also allowed. The monitoring point is the mass center of the roof story including the penthouse. The ultimate or failure state along the capacity curve may corresponds to that when the structure lost stability, or 80% elements reach their deformation limit, or more than 20% (maximum or effective) base shear strength reduced after yielding, whichever is critical. Performance Evaluation Performance point is evaluated through seismic coefficient method (FEMA440, 2004) or capacity- spectrum method (Fajfar, 1999; Chopra and Goel, 2004; FEMA440, 2004; Xue, 2001). The structural is pushed again to the target displacement associated with the performance point to access the behavior of both the structural system and the elements. Performance Criteria Examination Performance criteria regarding all performance goals are examined regarding structural systems and elements. Structural systems z Regarding vertical load carrying capacity, the structure should not collapse due to loss of any element. z Regarding lateral strength, damage and weak story mechanism can be clearly understood through hinge location and formation consequence. Strength deterioration is controlled by the definition of failure state during nonlinear pushover analysis. z Regarding lateral deformation capacity, the maximum inter-story drift ratio IDR and inelastic displacement demand ratio IDDR are calculated and compared with the acceptable performance criteria. Extremely soft story and extreme torsion irregularity must be avoided. Building separation distance must be no less than either the maximum story displacement under EQ2 or 70% that under EQ3 to avoid pounding. Structural elements z Elastic behavior examination must be done for elements that need to remain elastic according to the damage mechanism or structural modeling. z Deformation capacity examination is conducted for deformation-controlled behavior based on force-displacement relationship employed in the nonlinear static analysis. According to the performance criteria, only no less than 80% elements need to meet the same allowable IDDR as that of the structural system. z Strength capacity examination is conducted for force-controlled behavior so that the lowest bound of strength will not be less than the estimated force induced by the considered loading. In this research, element performance examination is not applied to performance goal associated with CP performance level where global system stability is of the main concern.

CONCLUSIONS

The tentative performance-based code introduces a transparent platform for the owners and designers to exchange their view on the expected seismic performance of the buildings under difference level of earthquake excitation. For buildings of different seismic use group, different performance goals are specified instead of employing an importance factor. Performance levels are quantified through parameters associated with structural strength, stiffness and ductility. Requirement on site suitability with focus on controlling the degree of liquefaction potential is given. Conceptual design rules with focus on redundancy and uniform continuity of strength, stiffness and ductility, respectively are specified. A performance objective oriented preliminary design procedure is presented with consideration of flexibility. Instead of ensuring structural systems possessing the prescribed ductility capacity through ductile design, the expected behavior or performance of the preliminarily designed structures are analyzed and evaluated through analysis methods, which may capture the real behavior reasonably. The tentative performance-based design code is not to replace the current design code at present. Both codes are supposed to exist for a period of time, during which possible amendment may be introduced to the performance-based design code. Finally, performance-based design code will dominate after incrementally improvement over the present design code. According to the examples studied in this research, seismic performance of the design structures could be easily understood and clearly interpreted through the performance-based design procedures. The engineers may have more confidence in the seismic performance of the designed structure under different level of earthquake excitation. Preliminary check on the inter-storey drift limit may help finding the stiffness deficiency earlier in the preliminary design stage and save some computational effort. In engineering practice, structural detailing has been standardized considering construction convenience. It is found that such a structure usually has a lower ductility capacity than that specified in code. Although structural ductility is not uniformly distributed, structural strength and stiffness are increased. Therefore, the performance objectives are usually satisfied. If an idealized detailing has been used without considering construction convenience, the ductility capacity is in very good agreement with the code value.

ACKNOWLEDGMENTS

This work is the result of a research project sponsored by Architecture and Building Research Institute, Ministry of The Interior under grant No. 094301070000G1018. Special thanks are given to the reviewers and the advisory committee members of this project for their valuable comments.

REFERENCES CPRMI, (2005), Seismic Design Specifications and Commentary of Buildings, Construction and Planning Agency, Ministry of The Interior, CPRMI publish introduce, Taiwan. (in Chinese) Xue, Q. (2000). “Need of performance-based earthquake engineering in Taiwan: a lesson from the Chichi earthquake”, Earthquake Engineering and Structural Dynamics, 29, 1609-1627. Loh, C.H. and Y.R. Hwang, (2002). Seismic Hazard Analysis of Taiwan Area — considering multiple ground motion parameters, Report No. NCREE-02-032, National Center for Research on Earthquake Engineering, Taipei, Taiwan. IBC2003, (2003). International Building Code 2003, International Code Council, CA, USA. FEMA450, (2003). NEHRP Recommended Provisions for Seismic Regulations for New Buildings and Other Structures, Part 1: Provisions, BSSC, Washington, DC, USA. JSCA, (2000). Structural Design by Response Control Methods, Japan Structural Consultants Association, Shokokusha Publishing Co., Ltd. (in Japanese) King, A. and R. Shelton, (2004). “New Zealand Advances In Performance-Based Seismic Design”, 13WCEE, Vancouver, Canada, Aug.1~6. Heidebrecht, A. (2004). “Code Development Issues Arising From The Preparation Of The Seismic Provisions Of The National Building Code Of Canada”, 13WCEE, Vancouver, Canada, Aug.1~6. Newmark, N.M. and W.J. Hall, (1982). Earthquake Spectra and Design, Earthquake Engineering Research Institute, Berkeley. FEMA440, (2004). Improvement of Nonlinear Static Seismic Analysis Procedures, (camera ready draft), Prepared by Applied Technology Council (ATC-55 Project) for Federal Emergency Management Agency, Washington, DC. SEAOC, (1999). Recommended Lateral Force Requirements and Commentary, Structural Engineers Association of California-Seismology Committee, USA. Goel, R and A.K. Chopra, (2004). “Evaluation of Modal and FEMA Pushover Analysis: SAC Buildings”, Earthquake Spectra, 20(1), 225-254. Fajfar, P., (1999). “Capacity Spectrum Method Based on Inelastic Demand Spectra”, Earthquake Engineering and Structural Dynamics, 28, 979-993. Xue, Q., (2001). “A Direct Displacement-Based Seismic Design Procedure of Inelastic Structures”, Engineering Structures, 23(11), 1453-1460. FEMA356, (2000). Prestandard and Commentary for The Seismic Rehabilitation of Buildings, prepared by ASCE, published by the Federal Emergency Management Agency, Washington, DC. USA.