Development of the Design Codes Grounded on the Performance-Based Design Concept in Japan

SOILS AND FOUNDATIONS Vol. 50, No. 6, 983–1000, Dec. 2010 Japanese Geotechnical Society

DEVELOPMENT OF THE DESIGN CODES GROUNDED ON THE PERFORMANCE-BASED DESIGN CONCEPT IN JAPAN

YUSUKE HONJOi),YOSHIAKI KIKUCHIii) and MASAHIRO SHIRATOiii)

ABSTRACT This paper traces the recent introduction of performance-based design (PBD) into Japanese design codes (especially geotechnical design codes) and describes the distinguishing characteristics of PBD codes. An overview is provided for source documents that initiated the concept of performance-based design, and consideration is given to the impact of the WTO/TBT Agreement and Japanese government policies on this issue. Three design codes are introduced to ex- plain the development of PBD codes in Japan: JGS4001–2004 `Principles for Foundation Design Grounded on a Per- formance-Based Design Concept' (Geocode 21); Technical Standards for Port and Harbor Facilities (TSPHF); and Speciˆcations for Highway Bridges (SHB). The ˆrst one is a model code established by JGS, whereas the latter two are regulatory design codes used in practice. Performance-based design is now recognized as the most important concept in code drafting and the majority of design codes will introduce this concept at least partially. Thus, PBD is a key con- cept for understanding the current Japanese design codes.

Key words: design codes and standards, foundation design, limit state design, partial factors, performance based de- sign, reliability based design (IGC: A8/H0/H1)

based speciˆcations and the use of international stan- INTRODUCTION dards are requirements of this policy. This paper traces the recent introduction of perfor- On the other hand, several of Japan's professional en- mance-based design (PBD) into Japanese design codes gineering societies had started examining PBD in the (especially geotechnical design codes) and describes the mid–1990s as a concept re‰ecting the current necessity distinguishing characteristics of these codes. The PBD and demand for design codes. The Japanese Geotechnical concept was o‹cially introduced into the Japanese Society (JGS) was one of the organizations active in this regulatory design codes in the revised Technical Stan- area at a very early stage, and the results were delivered dards for Port and Harbor Facilities (TSPHF) enforced quickly. The major outcome was the establishment of the in April 2007 (JPHA, 2007). Revisions based on the PBD ``Principles for Foundation Design Grounded on a Per- concept are presently underway in other design codes in- formance-Based Design Concept (JGS4001–2004)'', cluding the Speciˆcations for Highway Bridges (SHB), known internationally by its nickname `Geocode 21' which is regarded as the most important and widely used (JGS, 2005, 2006). This code has had a considerable im- civil engineering structural design code in Japan (JRA, pact on the drafting of design codes based on the PBD 2002a, b). Performance-based design is now recognized concept in Japan. The contents and characteristics of as the most important concept in code drafting and the JGS4001–2004 are described in this paper so that the majority of design codes will be introducing this concept, reader can grasp the key elements of PBD-based geo- at least partially. Thus, PBD is a key concept for under- technical design codes. standing the current Japanese design codes. Lastly, the revision process of the two regulatory de- In the ˆrst part of this paper, two documents consi- sign codes, Technical Standards for Port and Harbor dered to be sources of the PBD concept are introduced Facilities (TSPHF) and Speciˆcations for Highway together with the ensuing developments. Emphasis is Bridges (SHB), is reported. These explanations will hope- placed on the fact that the introduction of performance- fully provide a detailed understanding of the features of based design was a requirement of Japan's trade policy PBD codes. It should be noted that since TSPHF has al- based on the WTO/TBT Agreement. In this context, the ready been revised and was enforced in April 2007, substance of the WTO/TBT Agreement is brie‰y summa- whereas SHB is still under revision, the two codes are not rized. In actuality, both the introduction of performance- described to the same level of detail in this paper.

i) Gifu University, Gifu, Japan (honjo＠cc.gifu-u.ac.jp). ii) Port and Airport Technical Research Institute, Japan. iii) Public Works Research Institute, Japan. The manuscript for this paper was received for review on August 17, 2010; approved on October 20, 2010.

983 984 HONJO ET AL.

The deˆnition of PBD (performance-based design) in uniform approach to structural safety. this paper is the `design practice of thinking and working PBD codes are regarded as having the following advan- in terms of performance requirements for structures tages compared to the traditional regulatory design codes rather than descriptive speciˆcations'. (CIB, 1997; JGS, 2000; Honjo and Kusakabe, 2002): (1) Easier to understand the intention of a regulation. (2) Higher transparency and accountability for evalua- DEVELOPMENT OF THE PERFORMANCE-BASED tion of alternative means. DESIGN CONCEPT (PBT) (3) Easier to maintain consistency of description in Origins of the PBD Concept drafting the codes. Technologies are often developed and proposed based (4) Provision of an outline helps the code-writer draft on current demands and requirements. In some cases, a new design code. they originate from a single source; in other cases, similar (5) Easier to arrange the existing material and insert ideas are simultaneously proposed in various parts of the new material in the code. world in order to fulˆll the public's requirements. We be- These advantages were achieved by adopting the follow- lieve that performance-based design (PBD) ˆts into the ing framework for PBD codes: latter case. (1) A top-down hierarchical structure from objectives To the best of our knowledge, the PBD concept stems to means. from at least two important sources that were developed (2) Consistency in the presentation structure, which independently based on their respective societal needs. makes it easer to understand the relationship be- One is the Nordic Five Level System (NKB, 1978), which tween the whole and its parts, and between the proposes the framework for a building regulation system main document and the sub-documents. in Scandinavian countries, and the other is Vision 2000 (3) Consistency of description, e.g., to clarify a state- (SEAOC, 1995), which presents a seismic performance ment as `required', `optional', or `informative'. matrix. The minimum top-down structure includes at least two If we extend our scope to government policies on elements, namely the `objectives' and the `allowable economy and trade, one of the central regimes that con- means to accomplish the objectives'. There are some vari- trol world trade is that of the WTO, which was clearly ations in the structure, but it essentially follows the Nor- adopted for one of Japan's primary economic and trade dic Five Level System, as presented in Table 1. policies. The 1995 WTO/TBT Agreement requires per- In the Nordic Five Level System, the ˆrst three levels formance-based speciˆcations for industrial products and deˆne the required objectives of a structure in simpler to the use of international standards. Without doubt, the more detailed descriptions in a stepwise manner. The Nordic Five Level System was introduced in anticipation other two levels specify more concrete conditions that of the TBT agreement on a building regulation system. satisfy the required performance. The veriˆcation methods speciˆed in PBD codes for Nordic Five Level System conˆrming structural compliance with the performance Performance-based speciˆcations were ˆrst introduced requirements can be classiˆed into three categories (CIB, in the building regulation system known as the Nordic 1997): Five Level System (NKB, 1978). In the preface to the (1) Veriˆcation by approved procedures: This type of report, the objective of proposing this new system is ex- veriˆcation includes design by calculation, nondes- plained as follows: tructive in situ testing, destructive testing of speci- mens at a laboratory, item count and dimensional The system of rules which now governs building in the check Nordic countries is made up of legislation, regulations (2) Veriˆcation by comparison with standards or relat- and other building rules. In the action program of the ed documents known to comply with a perfor- Nordic Council of Ministers for the Nordic co-opera- mance requirement: This type of veriˆcation is tion (NU 1977:32) which the building sector is stated based on various national and international `type that the system of rules should in the ˆrst place be structures into a limited number of levels characteriz- ing the purpose of the regulations from the comprehen- Table 1. Nordic Five Levels System sive objective of the statute down to the technical solu- Level 1—GOALS: essential interests of the community at large with tion. In this way co-operation would be facilitated even respect to the built environment. if the administrative system varies from country to Level 2—FUNCTIONAL REQUIREMENTS: building or building element speciˆc qualitative requirements. country. (NKB, 1978; p. 23) Level 3—OPERATIONAL REQUIREMENTS: actual requirements, in terms of performance criteria or expanded functional It is clear from this explanation that from the very be- description. Level 4—VERIFICATION: institutions or guidelines for veriˆcation ginning, the intention of developing this framework was of compliance. to harmonize the regulatory speciˆcations and technical Level 5—EXAMPLES OF ACCEPTABLE SOLUTIONS: supple- ments to the regulations with examples of solutions deemed criteria as well as the administrative systems that diŠered to satisfy the requirements. from one country to another in order to achieve a PERFOMANCE-BASED DESIGN 985

approvals', deemed to satisfy solutions, acceptable ic damage caused by these earthquakes was US$7 billion solutions or accreditations for the former and US$30 billion for the latter, far higher (3) Veriˆcation by accreditation agencies, peer review than would have been expected considering the intensity or other experts of the earthquakes (Hamburger, 1997). Furthermore, it PBD codes have the potential to contribute in the follow- was discovered that communication gaps existed between ing aspects, and have already been introduced in various the owners and the designers of buildings regarding seis- countries: mic performance requirements. (1) Elimination of technical trade barriers One of the reasons for proposing the performance (2) Introduction of new technologies and exploration matrix was to close this communication gap. The desig- of creative design and construction methods ner designs the building based on the performance re- (3) Cost reduction in design and construction quirements speciˆed in the performance matrix selected On the other hand, PBD codes have the following prob- by the owner. This is based on the concept that the desig- lems to be overcome: ner and the owner should have a clear understanding and (1) Frequent gaps between qualitative objectives and agreement on the required performance of the building, performance requirements, and performance which was not always possible with the traditional criteria and veriˆcation methods prescriptive speciˆcations. (2) Prescriptive solutions are still employed in the The seismic PBD of buildings in the US has progressed veriˆcation, and the technical level is not yet ma- along the lines indicated by Vision 2000. This progress ture enough to quantitatively evaluate the degree of can be seen in reports by SEAOC (1998), FEMA (1997a, compliance between the solution and the required b), etc. One of the extensions of the PBD concept is the performance. SAC steel moment-resisting frame project (Wen, 2000). (3) Liability issues are not yet fully resolved. The project integrated the most advanced experimental These issues were anticipated from the beginning and are and theoretical results including knowledge in seismic de- considered unavoidable in the introduction of PBD sign and reliability assessment methodologies, and em- codes. EŠorts are continuously made to solve these prob- ployed the LCC concept to propose the design criteria for lems. steel frame buildings. Further progress can also be seen in In North America, the Nordic Five Level System is im- a report by Hamburger et al. (2004). plemented with the addition of the Vision 2000 perfor- Regarding highway bridges, the ATC-32 report (ATC, mance matrix, which is described below, and the further 1996) covers a project sponsored by Caltrans (California addition of a risk evaluation (e.g., ICC, 1998, 2000; Department of Transportation). It is also based on the Meacham et al., 2002). experience gained from the Loma Prieta and Northridge earthquakes, and includes the PBD concept. Other Performance Matrix and PBD in North America progress in PBD that has been made by AASHTO The Vision 2000 performance matrix was proposed in a (American Association of State Highway and Transpor- report by SEAOC (Structural Engineers Association of tation O‹cials) can be seen in the report by Buckle California) as a tool to ensure eŠective communication (2002). between the owner and the designer regarding their buil- The ICC (International Code Council), who published ding's seismic performance requirements (SEAOC, the harmonized International Building Code, has devel- 1995). The matrix indicates the seismic performance re- oped a PBD code that combines the Nordic Five Levels quirements with the impact of earthquakes on the y-axis, and the Vision 2000 performance matrix (ICC, 1998, the post-quake condition of the building on the x-axis, 2000). This code covers not only the structural aspects of and the signiˆcance of the building as a parameter (Fig. buildings but also ˆre safety, a superior concept in the 1). area of building design. It is important to understand the background of the performance matrix. The Vision 2000 report was pre- WTO/TBT Agreement: Performance-based Speciˆca- pared based on the damage incurred during two Califor- tions and Use of International Standards nia earthquakes, namely the 1989 Loma Prieta earth- There is no question that the rapid progress made in quake and the 1994 Northridge earthquake. The econom- adopting performance-based speciˆcations in Japan was triggered by the conclusion of the WTO/TBT (World Trade Organization/Technical Barriers for Trade) Agree- ment (Hirano et al., 2002). Thus, the contents of the agreement are brie‰y reviewed here. The preface states the following:

Members, Having regard to the Uruguay Round of Multilateral Trade Negotiations; Desiring to further the objectives of GATT 1994; Recognizing the important contribution that interna- Fig. 1. Performance matrix for buildings by SEAOC (SEAOC, 1995) tional standards and conformity assessment systems 986 HONJO ET AL.

can make in this regard by improving e‹ciency of WTO/TBT Agreement scheme, which we use for the pur- production and facilitating the conduct of internation- pose of explanation, is presented in Fig. 2 (Honjo et al., al trade; 2005). The structural performance requirements are de- Desiring therefore to encourage the development of ˆned by the performance-based speciˆcations, whereas such international standards and conformity assess- the design veriˆcation is based on limit state design ment systems; (LSD), which is speciˆed in international standards such Desiring however to ensure that technical regulations as ISO2394 (ISO, 1998). Note that in this paper, LSD and and standards, including packaging, marking and RBD (reliability based design) are used exactly in the labelling requirements, and procedures for assessment same sense. of conformity with technical regulations and standards do not create unnecessary obstacles to international Development of PBD Concept in Japan trade; (omission) Japanese Government Trade policy Hereby agree as follows: Since 1995, the year in which the WTO/TBT Agree- ment came into eŠect, the Japanese government has The preface is followed by the main body of the Agree- adoptedapolicyofderegulationwithregardtovarious ment, which consists of 15 articles and 3 annexes. Article laws and rules related to economic activities and trade. In 1.3 states `All products, including industrial and agricul- March 1998, the Three-Year Program for Promoting tural products, shall be subject to the provisions of this Deregulation was determined by Cabinet decision, and Agreement'. Concerning technical regulations, Articles the following tasks were delineated: 2.4 and 2.8 may be closely related to our interests: 1 All economic regulations should be eliminated in prin- ciple, and social regulations should be minimized. 2.4 Where technical regulations are required and 2 Rationalization of regulation methods. For example, relevant international standards exist or their comple- testing can be outsourced to the private sector. tion is imminent, Members shall use them, or the 3 Simpliˆcation and clariˆcation of the contents of regu- relevant parts of them, as a basis for their technical lations. regulations (omitted). 4 International harmonization of regulations. 2.8 Wherever appropriate, Members shall specify tech- 5 Speeding up of regulation-related procedures. nical regulations based on product requirements in 6 Transparency in regulation-related procedures. terms of performance rather than design or descriptive Following the above program, the Three-Year Plan for characteristics. the Promotion of Regulatory Reform was determined by Cabinet decision in March 2001. The objectives of this In these articles, `the use of relevant international stan- plan are shown below: dards' as the basis of technical regulations and `regula- 1 Realization of sustainable economic development tions based on product requirements in terms of perfor- through the promotion of economic activities. mance' are clearly stated. 2 Realization of a transparent, fair and reliable econom- It is understood that the WTO/TBT Agreement is ic society. aimed at overcoming the defects of the traditional design 3 Assurance of diversiˆed alternative lifestyles. codes and production standards based on descriptive 4 Realization of an economic society that is open to the characteristics, shortening the innovation time of new world. technologies, and unifying regional standards, thereby In order to realize such objectives, the promotion of es- expanding world trade markets, encouraging free compe- sential and active deregulation in various administrative tition and obtaining higher cost performance of the mar- services was planned. In the ˆeld of standards and ac- ket. This agreement has had a signiˆcant impact on the creditations, the following basic policies were implement- construction industry. ed: The framework that the design codes should follow in – Essential reviews of standards and accreditations in order to check the necessity of government involve- ment. – In cases where administrative involvement was still re- quired, administrative roles should be minimized, and self-accreditation or self-maintenance of standards and accreditations by the private sector should be promot- ed. – The international harmonization of standards and per- formance-based speciˆcations and the elimination of multiple examination procedures in accreditation proc- esses should be promoted. The third item had a very strong impact on the revision of design standards and codes for civil structures. In March Fig. 2. Basic framework of design codes: PBD and RBD 2003, the Ministry of Land, Infrastructure and Transpor- PERFOMANCE-BASED DESIGN 987 tation (MLIT) initiated a program entitled Restructuring performance-based design (PBD) is the Technical Stan- of Public Works Costs, which includes the following dard for Port and Harbor Facilities (TSPHF) revised and tasks: enforced in April 2007 (JPHA, 2007; Nagao et al., 2009). – Revision of common speciˆcations for civil works. The Speciˆcations for Highway Bridges (SHB) (JRA, – Review of the Highway Bridge Speciˆcations. 2002a, b), which is regarded as the most widely used civil – Revision of the Technical Standards for Port and Har- engineering design code in Japan because of its large mar- bor Facilities (TSPHF) to performance-based speciˆca- ket size and variety of code users, is under revision with tions. the focus on incorporating the PBD concept. The back- ground for the revision of these codes is brie‰y explained Revision of the TSPHF had started around this time with in this section. the goal of achieving harmonization between the stan- (1) Technical Standard for Port and Harbor Facilities dards and the international agreement. (TSPHF) Response of the Professional Societies Japan's ˆrst guideline for port and harbor facilities, Professional societies, on the other hand, recognized which was more or less a collection of design case histo- the impact of the WTO/TBT Agreement at a relatively ries, was published by the government in 1930. Engineers early stage, and several activities were initiated. The at that time designed port facilities by themselves with the Japanese Geotechnical Society (JGS) was one of several aid of such design examples. In 1967, the Port and Har- organizations that took action very early on. The JGS es- bor Bureau published the Standards for Port and Harbor tablished a technical committee to examine `Prospects of Facility Design, which were the basis for standards for present and future foundation design: international har- approximately 40 years. However, at that time, such monization of the design method and soil investigation', standards were not legally enforceable. In 1973, the Port chaired by Osamu Kusakabe. This committee, which and Harbor Law was revised to make these standards operated from 1997 to 2000, started the activities that led mandatory. In 1979, standards and commentaries were to establishing the ``Design Principles for Foundation revised to suit the law. The Port and Harbor Bureau Structures Grounded on a Performance-Based Design revised the standards two more times during the period Concept' (JGS4001–2004) in 2005 (JGS, 2005, 2006). The up until 1999. In these revisions, the concept of the stan- ˆrst draft of this code in Japanese was prepared as early dards was the same as in 1967. In 2007, new technical as 2000 (JGS, 2000), and the English version in 2002 standards were presented. The concept of these standards (JGS, 2002), which is known by the nickname `Geocode was diŠerent from the former standards. These standards 21'. The concept proposed by Geocode 21 was further ex- were formulated to coincide with the WTO/TBT Agree- tended to all civil structures, and was published as Code ment. PLATFORM ver.1 by JSCE (JSCE, 2003; Honjo and Based on this background, the Port and Harbor Law Kusakabe, 2004). This was one of the basic documents was revised in the Parliament, which made a proclama- used to draft the revised TSPHF, which became eŠective tion in September 2006, and the revised law was im- in 2007. plemented on April 1, 2007. The item that in‰uenced the Similar activities took place simultaneously in various revision of the standards, Article 56 Item 2–2, is shown in civil structure ˆelds. Published documents include JSSC Table 2. In the revision, rather than prescribing the (2001) for steel structures, JSCE (2002) and ICCMC speciˆcations for design details, the performance of facil- (2001) for concrete structures. Among these, ICCMC ities is regulated. (2001) was drafted to harmonize the design codes for con- Based on the revised Port and Harbor Law, the crete structures in Asia and the Paciˆc region based on TSPHF was fully revised. The main points of the revision the PBD concept. This activity continued and ICCMC cover two aspects, namely, the system for the performan- (2006) was published. In the ˆeld of geotechnical seismic ce-based speciˆcations, and the performance veriˆcation. design, ISO23469 (ISO, 2005) was developed based on the Only the performance requirements are mandatory, concept of PBD, with Japan taking the lead in its estab- lishment. The basis of design for all civil structures, Code PLATFORM ver.1 (JSCE, 2003; Honjo and Kusakabe, Table 2. Sequence of events on PBD codes in Japan 2004) was completed in 2003. JSCE also published a 1995 WTO/TBT agreement enforced guideline in 2008 for actions on structures based on the 1998 ISO2394 ``General principles on reliability for structures'' PBD concept (JSCE, 2008). These documents are not published. based on any mandatory regulations, but are mainly pub- 1998 3 years program for promoting deregulation (cabinet decision) 2000 Completion of 1st draft of Geo-code 21. lished as model codes or guidelines for introducing PBD 2001 3 year program for promotion of regulatory reform (cabinet into design codes. However, these documents were decision) referred to when regulatory design codes, such as TSPHF 2002 Completion of ``Principles on design of civil and building structures'' andSHB,werebeingrevisedinordertointroducethe Restructuring of Public Works Costs (MLIT) PBD concept. 2003 Review of SHB, and Revision of TSPHF. 2005 JSCE, ``code PLATFORM ver.1'' published. 2007 JGS, JGS4001–2004 published. Revision of TSPHF and SHB Enforcement of revised TSPHF. One regulatory design code that has already introduced 988 HONJO ET AL. whereas the veriˆcation can be based on the attached smooth introduction of new technologies and the ex- commentaries or on any methods considered appropri- perience of engineers into new projects and to accelerate ate. technological innovation and improvement, while main- Previously established comprehensive design codes, taining existing standards for quality and safety in de- MLIT (2002), JSCE (2003) and JGS (2005) provided the sign. foundation for revising these technical standards. The As performance-based speciˆcations, the conventional framework of the standards was the one proposed in speciˆcations in SHB were restructured into mandatory JSCE (2003), which was considered the most appropriate performance requirements allocated to levels in a hierar- for introducing PBD into the design codes in Japan. chical manner. In doing so, it was also necessary to draft (2) Speciˆcations for Highway Bridges (SHB) some new requirements. Furthermore, standard design The Speciˆcations for Highway Bridges (SHB) is an- methods and solutions are presented as possible veriˆca- nounced by the Ministry of Land, Infrastructure, Trans- tion methods. The revised SHB allows designers to select port and Tourism (MLIT) as the code of practice for Sec- design methods, new materials, new types of structure, tion 30 of the Road Law and Section 35 of the Road etc. diŠerent from those presented in SHB on the condi- Structure Ordinance. The SHB was drafted by the Bridge tion that the speciˆed performance requirements would Technical Committee convened by the Japan Road As- be satisˆed. sociation (JRA) based on the request of MLIT. The In addition, the importance of seismic performance clauses ˆnalized and announced by the MLIT are also accountability was widely recognized after the disastrous published in book form by JRA with the Committee's damage to infrastructure during the 1996 Hyogo-ken commentaries and further recommendations (JRA, Nanbu (Kobe) earthquake. Seismic performance of high- 2002a, b). The SHB is composed of the following ˆve way bridges has been expressed with what is referred to as parts: a performance matrix. The basic principles have already Part I: Common design principles been tabulated in the 1996 version of the SHB and there Part II: Steel structures wasnorevisioninthe2002version.However,theterms Part III: Concrete structures ``Seismic Performance Level'', ``Level 1 earthquake'' Part IV: Substructures and ``Level 2 earthquake'' are newly deˆned for the per- Part V: Seismic design formance matrix in the 2002 Speciˆcations. In order to Part I, Common design principles, prescribes the prin- achieve the performance requirements for a bridge, a ciples for the basis of design, loads, and structural combination of relevant limit states for individual struc- materials common for all structures. Part V, Seismic de- tural components of the bridge would be assigned, and sign, prescribes the basis for seismic design, earthquake veriˆcation would be made to ensure that individual eŠects including seismic loads, seismic instability of soils structural components do not exceed the speciˆed limit such as liquefaction and liquefaction-induced lateral state. Actually, the introduction of performance-based spreading, and prevention of unseating of superstruc- design and limit-state design has signiˆcantly improved tures. Part IV, Substructures, speciˆes the basis of design the transparency and accountability of the seismic design for piers, abutments, and foundations. With respect to of highway bridges. seismic design, Part IV also describes speciˆc modeling Durability is an essential performance requirement of a for ground resistance; for example, formulas on the ulti- bridge. In the conventional SHB, durability issues were mate bearing capacity of a spread foundation and the covered by specifying the structural details and safety shaft and bottom resistance of a pile, as well as structural margin embedded in the design. Nevertheless, the ac- resistance and earthquake-resistant structural details of cumulation of highway bridges and their maintenance foundation members. Parts II and III deal with the de- records raised concerns about deterioration damage sign of superstructures. caused by heavy tra‹c loads and severe environmental The SHB is traditionally considered the national de conditions. At the same time, research on durability de- facto standard composed of established design methods sign was making progress. Based on these eŠorts, basic and structural details. It is widely recognized that, as the design principles for durability were set out in the 2002 de facto standard on civil structures, the SHB has con- SHB, where a newly speciˆed fatigue design procedure tributed considerably to the eŠective and economical con- for steel members and the chloride ingress design for con- struction of highway networks that supported Japan's crete members were introduced. rapid economic growth from the 1960s to 80s. Innovation It should be noted that the introduction of PBD into in design and construction were occasionally re‰ected in the SHB is being done in two steps: The performance- earlier revisions of the SHB, and technologies were ac- based speciˆcation format was introduced in the 2002 re- cumulated. vision of the SHB, but LSD was not introduced at that The latest revision of the SHB was implemented in time. The introduction of LSD together with a more 2002 (JRA, 2002a, b). The target of the revision was to thorough introduction of PBD are scheduled at present introduce the concepts of performance-based speciˆca- (Tamakoshi, 2006). The introduction of the reliability- tions and durability design. The motivation for introduc- based design (RBD) concept is presently at the middle ing performance-based design was to accommodate working stage. It is essential to clarify speciˆc safety mar- diversifying procurement rules, that is, to encourage the gins based on various quantiˆed uncertainties so that de- PERFOMANCE-BASED DESIGN 989 signs that are diŠerent from the standard solutions can be Table 3. Table of contents of Geo-code 21 ver.3 accommodated under the same umbrella, thus achieving 0. BASES OF STRUCTURAL DESIGN full PBD. The related government research agencies are 0.1 Scope of application drafting the load and resistance factor design (LRFD) 0.2 Objective version of the SHB based on RBD together with the im- 0.3 Functional statements proved performance-based speciˆcations. 0.4 Performance requirements 0.5 Acceptable veriˆcation methods JGS4001–2004 has developed since 1997 and PBD, 0.6 Veriˆcation by Approach A LSD, RBD and the characteristic values of geotechnical 0.7 Veriˆcation by Approach B parameters, etc., have been studied in the process of this 0.8 Documents related to design and construction development. The core drafting members of SHB cooper- 0.9 Revision of the present code ated with other members in this activity. As mentioned 0.10 Deˆnitions of terms and notations 1. BASES OF FOUNDATION DESIGN earlier, the JGS4001–2004 is published as a model code 1.1. Scope of the design code for code writers so that it is referred to during the revision 1.2. Objectives of foundations process for all Japanese design codes. Accordingly, the 1.3. Functional statements foundation design in the SHB borrowed some ideas from 1.4. Performance requirements JGS4001–2004 such as the LSD concept in seismic design, 1.5. Design of foundations 1.6. Veriˆcation by Approach A the LRFD format and the deˆnition of characteristic 1.7. Veriˆcation by Approach B values for geotechnical parameters. 1.8. Seismic design of foundations 1.9. Foundation design report 2. GEOTECHNICAL INFORMATION DEVELOPMENT OF DESIGN CODES BASED ON 2.1. Scope PBD CONCEPT IN JAPAN 2.2. Objective 2.3. Interpretation of geotechnical information The previous chapter described the background and 2.4. Relationship between geotechnical investigation and structural development process of design codes based on the PBD design concept. In the chapter, the contents of each code are ex- 2.5. Procedure of geotechnical investigation 2.6. Other matters plained. We ˆrst look at JGS4001–2004, published by 3. DESIGN OF SHALLOW FOUNDATION JGS (2005, 2006). This code is for code writers, and it is 3.1. Scope conceptual and ideal seeking. TSPHF (JPHA, 2007) and 3.2. Objective SHB (JRA, 2002a, b) are explained next. Both of these 3.3. Functional statements codes are regulatory design codes in Japan. 3.4. Performance requirements 3.5. Investigation of ground and surrounding conditions 3.6. Matters to be considered in design Principles for Foundation Design Grounded on a Perfor- 3.7. Analysis of shallow foundation mance-based Design Concept (JGS4001–2004) 3.8. Veriˆcation Contents 3.9. Execution In 1997, the JGS established a committee started in 4. DESING OF PILE FOUNDATION (Sections are omitted here) 1997, and decided to draft a new code based on PBD con- 5. DESIGN OF COLUMN TYPE FOUNDATION cept. The code, nicknamed Geocode 21, was an attempt (Sections are omitted here) to harmonize all the major foundation design codes in 6. DESIGN OF RETAINING STRUCTURE Japan that had been developed in a rather independent (Sections are omitted here) manner due to various reasons. The purpose was to pub- 7. DESIGN OF TEMPORARY STRUCTURE (Sections are omitted here) licize the Japanese foundation design technology Annex: worldwide in a consistent manner. This attempt was A An example of comprehensive design code: somewhat earlier than similar attempts in other areas, B Comments on seismic design of foundations and thus has had a considerable social impact. C Comments on geotechnical information for foundation design D Determination of characteristic values from a small number of The ˆrst draft of Geocode 21 was completed in 2000 samples (JGS, 2000; Honjo, 2000), with the English version com- E Comments on shallow foundation design pleted in 2002 (JGS, 2002; Honjo and Kusakabe, 2002). F Comments on pile foundation design The code was subsequently published as the `Principles G Comments on column type foundation design for Foundation Design Grounded on a Performance- H Comments on earth retaining structures design I Comments on temporary structures Based Design Concept (JGS4001–2004)', which is an o‹- cial document of JSG. The English version has since been published (JGS, 2006), and a paper to the ICSMGE has been published for reference (Honjo et al., 2005). (JSCE, 2003; Honjo and Kusakabe, 2004). Presented in Table 3 is the table of contents of Chapter 0 and 1 describes the basis for foundation de- JGS4001–2004. Chapter 0 was drafted to propose a com- sign. Chapter 2 `Geotechnical information' describes the prehensive design code for all civil and building struc- problem of determining the characteristic values of geo- tures as there was no such code in Japan at that time. This technical parameters, which is one of the central issues in chapter was developed and extended to all civil struc- a geotechnical design code. Chapters 3 to 7 deal with vari- tures, and was published as Code PLATFORM ver.1 ous types of foundation structures. 990 HONJO ET AL.

The design of earth structures is not included in JGS4001–2004. However, several additional drafts based on the PBD concept were prepared by a committee on earth structures such as embankments and cut slopes (Honjo and Honda, 2007; Honda et al., 2009), although this work is not covered here.

Special Features (1) Seeking the ideal design code JGS4001–2004 was drafted pursuing the ideal founda- tion design code in Japan. That is to say, the code strives to systematize and harmonize the major foundation de- sign codes in Japan that had been developed in a rather independent manner. In proposing such a code, it is neither meaningful nor feasible to try to develop a code at Fig. 3. Hierarchy of requirements, veriˆcation and codes the same level as the existing major design codes; an ad- vanced concept is required in proposing such a code. The PBD concept is the backbone of this code. International standards such as ISO2394 (ISO, 1998) and regional standards such as Structural Eurocodes were carefully reviewed, and their essence was adopted. ForallmajordesigncodesinJapan,itisofprimary importance that the design changes immediately a revised code is enforced for the category of structures governed by that code because of the legal background. This is too strong a constraint for a code to introduce a new concept. For this reason, it is our experience that all new concepts introduced into the codes are slowly deformed, stripped of the essential contents in the drafting process, and ˆnally enforced with no substance. It is not expected that JGS4001–2004 would be used in Fig. 4. Performance matrix the actual design process from the day it is issued; rather it is the pursuit of an ideal code in which all codes ˆnally merge together in the near future. It is expected that vari- criteria of a structure should be determined based on the ous foundation design codes in Japan accept the concepts magnitude and frequency of load that the structure is ex- and formats, etc., proposed in this code, and eventually posed to during its service life, and the importance of the calmly harmonize with this code within a certain time in- structure (Fig. 4). terval. In this sense, JGS4001–2004 is regarded as a code (3) Diversiˆcation and standardization of design veriˆ- for code writers. cation methods (2) A code based on performance-based speciˆcations There seems to be two major global trends in the de- One of the distinguishing features of JGS4001–2004 is velopment of structural design codes. One is the diversiˆ- the introduction of the PBD concept. The performance cation or increase of freedom in design, which has gained requirements of foundations are hierarchized in order to momentum since the conclusion of the WTO/TBT increase transparency and accountability of the code Agreement where a consensus was reached on the use of (Fig. 3). The framework of the code consists of three lay- performance-based speciˆcations for all industrial ers: Objectives, Performance requirements and Perfor- products. mance criteria. The other trend is standardization or uniˆcation as Objectives: Objectives are the ˆnal social requirements of seen with ISO and Eurocodes that attempt to regionally a structure for one of its speciˆc performances (e.g., or globally standardize and unify all design veriˆcation structural performance) described in general terminolo- methods. It is necessary to account for these two trends gy. (i.e., diversiˆcation and standardization) simultaneously Performance requirements: Performance requirements in developing a new code, even though they sometimes describe the functions of a structure that should be pro- appear to be contradictory. vided to achieve the objective stated in general terminolo- In order to account for these two trends at the same gy. time, two diŠerent approaches to the veriˆcation of struc- Performance criteria: Performance criteria specify the tural performance, namely Veriˆcation approach A and details that are necessary to fulˆll the performance re- B, are proposed in JGS4001–2004 (Fig. 4). Veriˆcation quirements. In principle, they should be quantitatively approach A is the fully performance-based design ap- veriˆable in the structural design. The performance proach where designers are only given the performance PERFOMANCE-BASED DESIGN 991 requirements of the structures; the designers are request- value is deˆned as a mean value of a geotechnical ed to verify their design, and the results are checked by an parameter. By doing so, it is preventing designers from authorized organization. arbitrarily including a safety margin in the determination On the other hand, Veriˆcation approach B is a veriˆ- of a characteristic value by taking a conservative value. cation procedure based on design codes: these codes may On the other hand, it is encouraging the introduction of be established for each category of structures (e.g., high- engineering judgments that are a most important element way bridges, buildings, etc.) by an authority who is either in geotechnical engineering by certifying the goal (i.e., es- the owner or the one responsible for the administration timating the mean value of a geotechnical parameter). and safety of the category of structures. In Veriˆcation The other important reason for applying the mean approach B, JGS4001–2004 is to be used as acodefor values to the design is that it helps designers get a ``feel'' code writers. for the most likely behavior of their design up to the last (4) Limit state design based code stage of their design calculation. This is important for en- JGS4001–2004 is based on ISO2394, General principles gineering judgments in geotechnical design. on reliability for structures (ISO, 1998), which is based on (6) A checklist for design the LSD and RBD concepts. The notations and terminol- JGS4001–2004 is a comprehensive foundation design ogy are deˆed in accordance with ISO2394 as much as code; thus we made the following points in our policies possible. It is presumed in JGS4001–2004 that LSD is one while drafting the chapters for a particular type of foun- of the most suitable methods for realizing PBD. dation: (5) Characteristic values of soil parameters It was the aim of such chapters to create a checklist for The most important role of design codes is to deter- designing foundations based on state-of-the-art mine the safety margin (or elements) in design by balanc- knowledge. ing the uncertainties involved in actions, resistances and In this checklist, we tried to avoid quantitative descrip- calculation models in order to su‹ciently satisfy the vari- tions and to use only the qualitative descriptions. This ous performance requirements of a structure during its is to secure su‹cient room for the code writers to in- service life (Ovesen, 1989). troduce their own performance requirements in draft- In geotechnical design, the geotechnical parameter ing the code. values diŠer from one site to another, and they are esti- This code can be used as a table of contents for draft- mated based on in situ investigations, laboratory tests or inganewcode. past experiences. It is very diŠerent from the design of We are including some typical concrete foundation de- concrete or steel structures where the material parameter sign methods in the appendices of the code. These values are speciˆed based on industrial standards and are methods are simpliˆed versions of the actual design controlled in the manufacturing process. Therefore, in methods used in major Japanese foundation design order to introduce an equal margin of safety to all codes. designed structures in geotechnical design, it is necessary and inevitable for all designers to understand in what Revision of the Technical Standards for Port and Harbor sense a soil parameter value (i.e., a characteristic value) is Facilities (TSPHF) a representative value of the ground. If there is no com- The TSPHF was revised and enforced in 2007 (JPHA, mon understanding among the designers, the safety mar- 2007). The revised TSPHF is a fully PBD-based design gin introduced in the design may diŠer among structures. code, and its contents are introduced here (Nagao et al., In JGS4001–2004, the deˆnition of the characteristic 2009). value of a soil parameter is given as follows: Performance-based Speciˆcations System (1) REQ The characteristic values of geotechnical The basic system for the TSPHF is that the required parameters are the representative values carefully esti- performances of the structures are given as mandatory i- mated as the most appropriate ones for the foun- tems in three levels, i.e., objectives, performance require- dation-ground models for design calculations taking ments and performance criteria, whereas the perfor- into account variations of various sources. mance veriˆcation methods are not mandatory but are (2) REQ These representative values of geotechnical given in the commentary, the annexes or the reference parameters are principally the averages of the meas- documents as examples of possible methods (Fig. 5). ured values. These averages are not mere mathematical The performance requirements are classiˆed into basic averages, but taking into account estimation errors as- requirements and other requirements (Table 4), where sociated with statistical averaging. Moreover, these the former speciˆes structural performances against vari- values must be determined as careful estimations of ous actions and their combinations, and the latter speci- averages exercising due consideration on geologic/geo- ˆes structural dimensional requirements arising from technical as well as experiences in similar past projects, usage and convenience. The basic requirements are fur- and based on comprehensive interpretation of diŠerent ther classiˆed into serviceability, reparability and safety kinds investigation techniques and testing methods. requirements, as deˆned in Table 4. The basic requirements should be combined with the The most signiˆcant point here is that the characteristic actions considered in the design, which are summarized 992 HONJO ET AL.

Fig. 5. Performance based speciˆcations for the TSPHF

Table 4. Performance requirements in the TSPHF

Classiˆcation Deˆnition Performance of structural response (deforma- Basic requirement tion, stress, etc.) against actions.

The function of the facility would be recovered Serviceability with minor repairs.

The function of the facility would be recovered Reparability in a relatively short period of time after some repairs. Signiˆcant damage would take place. However, Safety the damage would not cause any loss of life or serious economic damage to the hinterland.

Performance requirements for structural dimen- Fig. 6. Classiˆcations of performances, actions and frequency Other requirements sions concerning usage and convenience of the facilities

Table 6. Summary of the design situations

Table 5. Classiˆcation of actions in TSPHF Design situation Deˆnition

Permanent Any actions that continuously act on harbor facilities dur- Persistent Permanent actions (self weight, earth pressures) are actions ing their service life, e.g. Self weight, hydro and earth pres- Situation major actions. sures, environmental actions (temperature, corrosion etc. Transient Variable actions (waves, level 1 earthquakes) are any physical, chemical or biological actions to deteriorate Situation major actions. the structures and members) etc. Variable Any actions that likely to act on harbor facilities during Accidental Accidental actions (tsunamis, level 2 earthquakes) are actions their service life. e.g. Wave forces, live load, impacts to Situation major actions. quays by ships, Level I earthquake force (earthquake forces that are likely to be expected during the service life) etc. Accidental Any actions that are not likely to act, but have some im- actions pacts on harbor facilities during their service life, e.g. colli- sion of ships, Tsunami, Level 2 earthquakes (the maximum possible earthquake at a site) etc.

in Table 5. The combinations of performance require- ments and actions are referred to as design situations, where performance veriˆcation of the structure should be carried out for each design situation. The actions are clas- siˆed into accidental and permanent/variable actions em- Fig. 7. Typical caisson type breakwater ploying an annual occurrence rate of approximately 0.01 (i.e., a return period of 100 years) as a threshold value. For both permanent and transient design situations, trated in Fig. 6. It should be noted that the performance serviceability must be satisˆed, whereas in accidental situ- of a structure may not always be veriˆable in accidental ations, one of the three performance requirements must situations. be satisˆed depending on the importance and function of The objectives and performance requirements are the structure under design. This concept is further illus- prescribed in the MLIT Ministerial Ordinance part of the PERFOMANCE-BASED DESIGN 993

TSPHF, whereas the performance criteria are speciˆed in The stability for sliding can be checked by considering the the MLIT declaration part of the TSPHF that deˆnes the equilibrium between the horizontal external force action details of the TSPHF. In this way, the hierarchy of the on the caisson and the friction resistance between the performance speciˆcations is maintained. caisson and the mound. The conventional veriˆcation Table 6 shows an example of the provisions in the new equation based on the safety factor is given below: TSPHF. This example is for a breakwater, a representa- m･(W0－U) tive protective facility. Figure 7 shows the cross section Fsà (1) of a caisson-type breakwater. In the new TSPHF, the ob- P jectives, performance requirements, and performance where Fs is the safety factor, m is the friction coe‹cient criteria are clearly written in accordance with the hierar- between the caisson and the mound, W0 is the total chy shown in Fig. 5. However, these were not clearly de- weight of the caisson at the time of still water, U is the up- scribed in the former TSPHF. With regard to veriˆca- lift water pressure on the caisson, and P is the horizontal tion, this was mandatory in the former TSPHF, but it is wave force on the caisson. Conventionally, the safety fac- not mandatory in the new TSPHF. Since performance tor for this equation is set to 1.2. veriˆcation in accordance with TSPHF is Approach B In the new TSPHF, the veriˆcation formula is given as veriˆcation shown in Figure 5, the recommended veriˆca- follows: tion method is presented in the guidelines, but it is not mandatory. g f g W －P －g P Æg P (2) f kØS Wi ik Bd PU UK » PH Hk i Performance Veriˆcation where g are partial factors, su‹x k is the characteristic Due to the introduction of PBD, it became necessary to value, su‹x d is the design value, f is the friction specify amounts of deformation and displacement in coe‹cient between the caisson and the mound, Wi is the some structures. It is especially important to determine total weight of the caisson, PB is the buoyancy force act- the residual displacement after a seismic event for deˆn- ing on the caisson at the time of still water, PU is the uplift ing reparability and serviceability. For this reason, force acting on the caisson, and PH is the horizontal wave response analyses are introduced for important structures force on the caisson. The failure probability used to set in addition to the conventional pseudo-static methods the partial factors in this case was 8.7×10－3 or below. such as the seismic coe‹cient method. It is also empha- This probability is a system failure probability, and not sized that for the prediction of deformation, it is im- only for the sliding failure mode. The system failure portant to use model tests and real-size experiments on probability is estimated by superposing the failure site to support the analytical methods. probabilities of the three failure modes without double RBD methods are introduced to ensure that the counting the failure probabilities. The partial factors are designed structure satisˆes the conditions speciˆed in the determined based on the average failure probabilities esti- performance matrix. By doing so, the failure probability mated from the sections designed by the conventional of a structure becomes apparent, which was not the case safety factor method for each failure mode. About 40 ac- in the conventional safety factor method. It goes without tual sections of the breakwater are used in this code saying that structures designed by the same safety factor calibration. may not have the same failure probability. The design value method is employed in determining RBD is usually classiˆed into 3 levels, namely Level 1, 2 the partial factors, where the recommended partial fac- and 3. The TSPHF employed Level 1 RBD as adopted in tors are presented in Table 7. the majority of design codes in the world. The format for veriˆcation has been changed due to the Characteristic Values of Geotechnical Parameters introduction of Level 1 RBD, and is described here taking The issue of how to determine the characteristic values the design of a gravity-type breakwater (caisson) as an ex- of geotechnical parameters is of vital importance in geo- ample. Three modes of failure should be considered in technical design. The method recommended in designing this structure as indicated in Fig. 8. JGS4001–2004 is illustrated in Fig. 9. The essential point Let us consider here the sliding failure of the caisson. of this recommendation is that a safety margin is not in-

Fig. 8. Failure modes of gravity type break water 994 HONJO ET AL.

Table 7. Partial factors for veriˆcation for sliding failure of gravity type break water

Recommended partial factors (variable action by wave force)

Target system reliability bT 2.38

－3 Target system failure probability PfT 8.7×10

Target reliability index for partial factors bT? 2.40

gam/xk s/m

gf Friction coe‹cient 0.79 0.689 1.060 0.150 slow water depth change 1.04 0.740 0.239 gPH, gPU －0.704 Rapid water depth change 1.17 0.825 0.251 Fig. 10. Flowchart for the determination of soil parameters in TSPHF

rwl＝1.5 1.03 1.000 0.200

Table 8. Values for correction factor b1 sliding gwl rwl＝2.0, 2.5 1.06－0.059 1.000 0.400

H.H.W.L. 1.00 — — Correction factor b1 Coe‹cient of gWRC Unit weight of RC 0.98 0.030 0.980 0.020 variation COV Parameter for Parameter for safe side unsafe side gWNC Unit weight of NC 1.02 0.025 1.020 0.020 COVº0.1 1.00 1.00 gWSAND Unit weight of sand 1.01 0.150 1.020 0.040 0.1ºCOVº0.15 0.95 1.05 (Note) P is calculated from b assuming b follows a normal distribu- fT T 0.15ºCOVº0.25 0.90 1.10 tion. g is determined partial factors, a is a sensitivity factor obtained in the reliability analyses, m/xk is the ratio between the mean value and the 0.25ºCOVº0.4 0.85 1.15 characteristic value and s/m is COV. The partial factors are calculated 0.4ºCOVº0.6 0.75 1.25 by the formula g＝m/xks1－abT(s/m)tby the design value method. Reexamination of the 0.6ºCOV data/Reexamination of the soil test

acteristic value is proposed (Watabe et al., 2009). Ac- cording to Ovesen (1995), the lower bound of the 95z conˆdence interval of the parameter x, for which the arithmetic mean value is given as x*, is obtained by a sim- ple formula: 1.645 x ＝ 1－ COV x* (3) k Ø n » Fig. 9. Flow chart for determining the characteristic values of the geo- technical parameters in JGS4001–2004 where xk is the estimated value of x, COV is the coe‹cient of variation of observed x,andn isthesamplesize.The

TSPHF proposes the following formula to obtain xk: troduced in the determination of characteristic values; it COV 0.5 * * is only introduced in setting the design values of geo- xk＝b1×b2×x ＝Ø1－ »Ø1－ »x (4) technical parameters. There has been considerable discus- 2 n sion on this point, which can be seen in the commentary where and the annexes of JGS4001–2004. COV 0.5 The determination procedure in the TSPHF is present- b ＝ 1－ , b ＝ 1－ 1 Ø 2 » 2 Ø n » ed in Fig. 10 where some modiˆcation has been made to the JGS4001–2004 procedure. The modiˆcation consists It is understood from this equation that the reduction of the introduction of the estimated value between the parameters b1 and b2 are introduced, where the former derived value and the characteristic value. The reason for concerns the degree of variability and the latter concerns this modiˆcation is that in practical situations, mechani- the sample size. When the scattering of data is large, i.e., cal or physical geotechnical parameters are obtained as a large COV, b1 is reduced, and when n is small, b2 is function of depth, and it is easier to specify the procedure reduced. In practice, b1 in Eq. (4) can be selected from taking into account (or limited to) the modeling of soil Table 8 based on the range of COV.Itisalsoarulethatif layers in an explicit fashion. Therefore, the procedure n is larger than 10, b2 can be set to 1.0. It is also suggested shown in Fig. 10 separates the modeling of parameters as in Table 8 that if COV exceeds 0.6. reexamination of the a function of depth from obtaining the estimated values. obtained data should be carried out. b1 and b2 in Eq. (4) is In addition, a simpliˆed procedure to determine a char- for cases where a smaller characteristic value of x is less PERFOMANCE-BASED DESIGN 995

Fig. 11. Depth proˆle of the undrained shear strength favorable. The diŠerence between Eqs. (3) and (4) is ones, it is necessary to conˆrm that the alternative minimal when n exceeds 10. method oŠers equivalent or greater performance com- The procedure for determining the characteristic value pared to the standard ones by experimental, theoretical from the measured values is explained here taking the un- or other means. drained shear strength, cu, data shown in Fig. 11 as an For example, the provision ``6.2 Examination of Chlo- example. Usually, observation logs, physical properties ride Ingress'' (Chapter 6 Examination of Durability in such as natural water content, unit weight of soil, etc., Part I) states the following: are attached with cu values, and the characteristic values are determined taking this information into account. (1) The durability of RC members of substructures However, for the sake of simplicity, we just use un- shall be maintained against the eŠects of salt. drained shear strength vs. depth information to deter- (2) If a measure to ensure the minimum depth of con- mine the characteristic value. crete cover stipulated in Table 6.2.1 (Omitted in the Three alternative models can be considered in this case, present paper) is taken, the RC members of substruc- namely (a) constant depth model, (b) one-layer linear in- tures located in the regions shown in Table 6.2.2 crease model, and (c) three-layer model. The results for (Omitted in the present paper) are deemed to satisfy (1) the three diŠerent models are presented in Fig. 11. It is above. understood from the results that the characteristic value cannot be determined using model (a) (i.e., Fig. 11(a)) On the one hand, Term (1) in Clause 6.2 demonstrates the because COV exceeds 0.6 in this case. Either model (b) or mandatory requirement and leaves room for the desig- (c) (i.e., Fig. 11 (b) and (c)) can be used. However, model ners to decide on the design methods and veriˆcation (c) gives higher characteristic values in most of the depth procedures for the prevention of chloride ingress. On the range and the characteristic values ˆt better to the meas- other hand, Term (2) allows for the regulatory speciˆca- ured values. tion for conventional structures. Another example is the veriˆcation requirement for a Revision of the Speciˆcations for Highway Bridges new piling method. New piling methods have been en- (SHB) thusiastically proposed by contractors to reduce costs and Performance-based Design solve environmental problems such as excessive noise and Performance-based design (PBD) in the SHB (JRA, disposal of excavated soil. The SHB (JRA, 2002b) pro- 2002a, 2002b) implies the combination of clearly deˆned vides a guideline for introducing such methods by ex- performance requirements of the bridges and several al- plaining the procedure for modifying the veriˆcation for- ternative solutions that could satisfy the deˆned perfor- mula for conventional piles in the commentary. mance requirements. This is diŠerent from the conven- tional regulatory speciˆcations where detailed descrip- Some believe that only the mandatory performance re- tions are given for the design procedures, material used, quirements should be mentioned in PBD codes and that conˆgurations and dimensions. However, standard de- veriˆcation methods should be the responsibility of the sign veriˆcation methods and structural details are still designers. Nevertheless, the performance requirements, speciˆed for structural safety and durability. When the the standard veriˆcation methods and the structural de- employed veriˆcation method diŠers from the standard tails coexist in the SHB. There are two reasons for this. 996 HONJO ET AL.

Fig. 12. Code structure of the Speciˆcations, Part V: Seismic Design

Table 9. Seismic performance matrix

Type of design ground motion Standard bridges (Type-A) Important bridges (Type-B)

Level 1 earthquake: high-probability ground motion SPL 1: Prevent damage

Level 2 earthquake: Interplate earthquakes (Type-I) SPL 2: Limited damage for function low-probability ground motions SPL 3: Prevent critical damage recovery Inland earthquakes (Type-II)

First, the SHB should provide a fundamental back- tails were applied, a su‹cient safety margin for larger ground by showing the widely accepted standards in prac- earthquakes was considered to be ensured. However, tice, because alternative methods are not necessarily used based on the catastrophic damage to highway bridges in in all projects. Second, designating quantitatively as- the 1995 Kobe earthquake, the latter was introduced in sessed requirements in the SHB is extremely di‹cult the 1996 version to employ a design methodology involv- without speciˆc values, shapes, materials and so on. Ac- ing a realistic description that directly represents the non- cordingly, the standard methods must be presented so linear behavior of bridges during a large earthquake. that engineers can conˆrm the performance of other ac- Table 9 shows the basic principles of seismic design in the ceptable solutions by comparison with the conventional SHB through a performance matrix of design earthquake methods of the SHB. ground motion and seismic performance level (SPL). The seismic performance level (SPL) depends on the Seismic Design importance of the bridge. Bridge importance is classiˆed The code structure for the seismic design speciˆcations into two groups: standard bridges (Type A) and im- is illustrated in Fig. 12. The principal requirements and portant bridges (Type B). Both Types A and B bridges the existing detailed veriˆcation methods are clearly sepa- shall resist Level 1 earthquakes to achieve SPL 1, where rated into two parts: requirements for which compliance SPL 1 stipulates that the bridge shall perform elastically, is mandatory and detailed standard veriˆcation methods without signiˆcant damage during an earthquake. Type- that oŠer acceptable solutions. Mandatory principal re- A bridges shall resist Level 2 earthquakes to achieve SPL quirements on the seismic performance of highway 3, where SPL 3 stipulates that the bridge shall perform so bridges, design earthquake ground motions, and princi- as to prevent critical failure that would lead to its collapse ples for verifying seismic performance are speciˆed as the during an earthquake. Type-B bridges shall resist Level 2 upper level in the code structure. earthquakes to satisfy SPL 2, where SPL 2 stipulates that A two-level design method has been employed in the the damage to the bridge shall be within a limited degree SHB since the 1996 version. The ˆrst level is the seismic so that the bridge quickly recovers its function. design against small-to-medium earthquakes, which have Table 10 summarizes the perspectives for characteriz- traditionally been considered. The second level is the seis- ing the seismic performance of bridges. SPL 1 to 3 are de- mic design against large earthquakes based on the disas- scribed from the viewpoint of ``Safety,'' ``Functionality trous damage from the 1995 Hyogo-ken Nanbu earth- (i.e., function recovery)'', and ``Reparability (i.e. repair quake (Kobe earthquake). Conventionally, when the work di‹culty)'' levels during and after an earthquake. former design was satisˆed and the relevant structural de- The relationship between the SPLs and the corresponding PERFOMANCE-BASED DESIGN 997

Table 10. Key issues of seismic performance

Repairability SPL Safety Functionality Short-term Long-term

SPL 1: Safety against unseating of Same function as that Repair not necessary for Prevent damage superstructure before earthquake function recovery Simple repairs

SPL 2: Safety against unseating of Early function recovery Function recovery possible Full recovery can be Limited damage for superstructure possible by temporary repair completed via relatively function recovery easy repair works SPL 3: Prevent critical Safety against unseating of ——— damage superstructure

Table 11. Limit state of foundation

Related seismic performance of bridge Limit state of foundation Veriˆcation items

The local plasticization of structural members of the foundation and ground resistances does not lead to clear inelastic behavior of the Repairability, foundation system. Functionality The state remains in a limited nonlinear region which seems to be linear when viewed as a (Veriˆcation of residual displacement of the foundation can be system. negligible for this degree of damage.) (Naturally satisˆed in the case of satisfying the above veriˆcation Safety items)

The foundation does not lose strength as a system. The damage to foundation members can be repaired or reuse of the Repairability, damaged foundation is possible by reinforcement. Functionality The state remains in a region in which the foun- dation retains su‹cient strength and the The residual displacement of the foundation does not cause an exces- damage can be repaired. sive inclination to cause the replacement of the piers. Safety (Naturally satisˆed in the case of the satisfaction for the above veriˆ- cation items)

Note: The response displacement of the foundation is also considered in the veriˆcation of unseating of the girders. levels of Safety, Functionality, and Reparability is de- system and their plastic deformation shall not exceed the scribed in Table 10, which is also given in the Speciˆca- onset of strength capacity loss. Therefore, designers must tions as a commentary. Note that Reparability comprises perform the following tasks at the beginning of the veriˆ- short-term reparability and long-term reparability. Short- cation of the seismic performance of a bridge for SPL 2 term reparability represents the di‹culty of repair work and SPL 3: required in order to recover bridge function. For exam- Select expected sections for plastic hinges and devices ple, su‹cient short-term reparability is considered to ex- for energy dissipation ist when emergency access to a bridge is possible by car- Determine the limit state of each member such that it is rying out temporary repair work shortly after the earth- assured that the bridge shall not reach its limit state quake, and some damaged members or devices may when a member reaches its limit remain unrepaired. Long-term reparability is a measure Design shall be conducted such that the bridge shall not of the di‹culty involved in quickly repairing damaged be damaged beyond the speciˆc limit states that are de- members or devices; however, repair work to be able to scribed from the perspective of mechanics, depending on use damaged members and devices is possible. Repair the performance levels for checking the performance re- work may begin after full resumption of bridge service. quirements. A bridge comprises structural components The distinction between short-term and long-term such as girders, piers, bearings, foundations, etc., which reparability is important in foundation design. serve as a system. For achieving the required total bridge The limit state for SPL 1 is speciˆed such that the be- system performance, the SHB also requests that the limit havior of the bridge as a system shall be within the elastic states of individual structural components be determined. limit. The limit state for SPL 2 is speciˆed such that pos- The idea is that integrating the components for which the sible plastic hinges (in other words, inelastic sections) limit state criteria are veriˆed can achieve the required shall be developed only at expected sections and devices total bridge system performance. Accordingly, the SHB of the bridge system and their plastic deformation shall regards limit state design as a concept for the design of in- be within the reparable limit. The limit state for SPL 3 is dividual structural components, not for the design of the speciˆed such that possible plastic hinges shall be devel- bridge structure as a system. oped only at expected sections and devices of the bridge The seismic performance of bridges, limit states of 998 HONJO ET AL.

Fig. 13. Ductility design of pier foundations for veriˆcation in the case where foundations have su‹cient ductility capacity foundations, and veriˆcation items are summarized as istic values of geotechnical parameters for calculating Table 11. In Table 11, the veriˆcation items are speciˆed foundation responses should target the values that will corresponding to the stability of the foundation, the predict the most likely behavior of the foundation under reparability of foundation members, and the in‰uence of the speciˆed load combinations. Namely, expected or foundation displacement on the bridge system. Since average values are recommended in principle. foundations are placed underground, it is time-consum- For example, Structural Eurocode 7, Geotechnical De- ing to examine and repair damaged portions before sign (CEN, 2004), states that the characteristic values restarting the service of a bridge after an earthquake. Ac- should be determined as a cautious estimate of the value cordingly, the plasticizing behavior of the foundation aŠecting the occurrence of the limit state without a system should be avoided even in a severe earthquake. speciˆc detailed guideline for the determination process. However, it is unavoidable to consider inelastic behavior ``A cautious estimate'' may work when considering the in rare scale earthquakes, especially, when liquefaction of foundation stability check based on a comparison of subsoil layers occurs or when a pier ends up possessing a strength or bearing capacity with applied force. large capacity due to factors outside the seismic design However, a cautious estimation does not work well espe- process. In these cases, it is essential to restrict the cially for the ductility design of foundations. When con- damage to foundations to satisfy with the short-term sidering several failure modes, reduced geotechnical repairability of bridge function and the long-term parameter values lead to design results on the safety side repairability of foundation damage. Accordingly, as for a particular failure mode but could have the adverse shown in Fig. 13, the nonlinear response of foundation result for some other failure modes. For example, given subjected to design seismic horizontal coe‹cients, khe and reduced geotechnical strength parameter values, the non- khg, respectively, is estimated by the energy conservation linear displacement response of foundations tends to be method. The veriˆcation is performed by use of the estimated on the safety side (i.e., larger displacement), response and allowable ductility factors, mFR and mFL re- and the peak sectional force values in foundation struc- spectively, where the ductility factor, m, is deˆned in m＝ tural members such as piles can be underestimated. Fur- d/dy by dividing the displacement d by the yield displace- thermore, very unrealistic failure modes of pile founda- ment dy at the point of seismic lateral load in the upper tions may occur in the design when piles yield or reach the structure. The yield point of the system behavior of foun- ultimate bearing capacity with intentionally reduced geo- dation is obtained by the so-called `log kh-log d analysis' technical parameter values. with the relationship between seismic coe‹cient, kh,and Since nonlinear design calculations are used in the SHB horizontal displacement, d, at the point of seismic lateral simultaneously considering several failure modes, the load in the upper structure. Finally, the foundation shall SHB deˆnition of the characteristic values for geo- be designed not to reach the deˆned limit state point. The technical parameters should be relevant. In addition, the allowable ductility factor, mFL, is recommended based on fact that the use of the average value should be supported previous large-scale experimental studies such as grouped by geotechnical communities based on a questionnaire piles subjected to lateral and overturning moment cyclic survey conducted by Shirato et al. (2002) also supported loads (Nakatani and Shirato, 2006), assuming that the the present deˆnition. damage to foundations should not prevent the bridge service for emergency vehicles while the damage to foun- dations may be repaired after the restart of the bridge CONCLUSION service. Performance-based design (PBD) is the main concept in the revisions of not only geotechnical design codes but Characteristic Value of Geotechnical Parameters also the major civil engineering design codes in Japan The commentary of the SHB states that the character- today. The background and the process of development PERFOMANCE-BASED DESIGN 999 are explained with special emphasis on the relationship 8) Hamburger, R. O., Rojaha, C., Moehle, J., Bachman, R., Comar- between the government trade policy and the design codes tin, C. and Whittaker, A. (2004): The ATC–58 project-develop- as well as the role of professional societies. The contents ment of next generation performance based earthquake engineering design criteria for buildings, 13 WCEE, Vancouver, Canada, 2004, of JGS4001–2004, TSPHF and SHB are described in the 1819. context of the PBD concept. It is expected that the PBD 9) Hirano, Y., Yamamoto, S. and Nishikawa, K. (2002): Present and concept will play a major role in the development of de- future of technical standards of civil and building structures, Tech- sign codes in Japan for some time. nical Note of National Institute for Land and Infrastructure The development described based on the PBD concept Management (58), 83–120. (in Japanese). 10) Honda, M., Kikuchi, Y. and Honjo, Y. (2009): Application of con- has taken place in the past ten to ˆfteen years, which is a cept in `Geo-code21' to earth structures, Geotechnical Risk and relatively short period. This could not have been achieved Safety, Proc. of 2nd ISGSR and IS Gifu, 155–158. without Japan's trade policy based on the WTO regime. 11) Honjo, Y. (2000): Proposal of a comprehensive foundation design This is especially true in the PBD codes for the introduc- code `Geo-code 21', Journal of JGS (Tsuchi to Kiso), 48(9), 17–20 tion of performance-based speciˆcations prescribed in (in Japanese). 12) Honjo, Y. and Kusakabe, O. (2002): A keynote lecture `Proposal of the WTO/TBT Agreement. The design codes developed a comprehensive foundation design code: Geo-code 21 ver.2', along this line should contribute to the removal of techni- Foundation design codes and soil investigation in view of interna- cal trade barriers, promotion of construction markets, in- tional harmonization and performance based design, Proc. of IWS troduction of new technologies and improvement of cost Kamakura, Balkema, 95–103. performance of design and construction. 13) Honjo, Y. and Kusakabe, O. (2004): Some movements toward es- tablishing comprehensive structural design codes in Japan: `Geo- The other important aspect of PBD codes is improve- code 21' and `code PLATFORM ver.1', Proc. of the 3rd Civil En- ment of communications between owners and designers, gineering conference in the Asian region (CECAR),Seoul,Korea, which was the motivation for developing the perfor- 217–220. mance matrix for seismic design of buildings. The PBD 14) Honjo, Y., Kikuchi, Y., Suzuki, M., Tani, K. and Shirato, M. codes developing in Japan also contain this aspect of (2005): JGS Comprehensive Foundation Design Code: Geo-code 21, Proc. 16th ICSMGE, Osaka, 2813–2816. PBD, and have the potential to more directly accommo- 15) Honjo, Y and Honda, M. (2007): Perspectives and issues concern- date the requirements of the general public for the design ing performance based speciˆcations of earth structures: a technical of infrastructures in the long run. committee report, Journal of JSCE: Division C—geotechnical en- PBD codes are drafted employing RBD (or LSD) as the gineering, 63(4), 993–1000 (in Japanese). design veriˆcation method. Since PBD and RBD com- 16) ICC (1998): ICC Building Performance Committee: Preliminary Committee Report. pensate each other quite well, this line should be followed 17) ICC (2000): Final draft ICC performance code for buildings and in the present development of the new design codes. facilities, 197. It should not be forgotten that in order to fulˆll the 18) ICCMC (2001): Asian Concrete Model Code 2001, International true PBD concept in design and construction, the de- Committee on Concrete Model Code for Asia, March 2001. velopment of design codes is just one of many necessary 19) ICCMC (2006): Asian Concrete Model Code 2006, International Committee on Concrete Model Code for Asia, June 2006. changes. These changes include modiˆcations to the tech- 20) ISO (1998): International Standard ISO/FDIN 2394, General prin- nical approval system, bidding system, insurance system ciples on reliability for structures. and contract system of the construction industry, some of 21) ISO (2005): ISO 23469 Bases for design of structures—Seismic ac- which have already been implemented. tions for designing geotechnical works, 2005 22) JGS (2000): Report on international harmonization of design and soil investigation method, Technical Committee on present and fu- REFERENCES ture of foundation design method in Japan, 45th Geotechnical En- gineering Symposium (in Japanese). 1) ATC (1996): ATC: Improved seismic design criteria for California 23) JGS (2002): The JGS foundation design guideline (Draft ver.2), bridges: provisional recommendations (ATC-32), Jun. 1996. Foundation design codes and soil investigation in view of interna- 2) Buckle, I. G. (2002): AASHTO LRFD limit state design of bridges tional harmonization and performance based design, Proc. of IWS with emphasis on seismic performance, Foundation design codes Kamakura, Balkema Publishers, 399–457. and soil investigation in view of international harmonization and 24) JGS (2005): Principles for foundation designs grounded on a per- performance based design, Proc. of IWS Kamakura, 17–30, Balke- formance based design concept (Japanese version) (JGS4001– ma Publishers. 2004). 3) CEN (2004): EN 1997–1 Eurocode 7 Geotechnical Design—Part 1: 25) JGS (2006): Principles for foundation designs grounded on a per- General Rules. formance based design concept (English version) (JGS4001–2004). 4) CIB (1997): Final report of CIB Task Group 11 - Performance 26) JPHA (2007): Technical Standards for Port and Harbor Facilities: based building codes, CIB report Publication 206. Speciˆcations and Commentary (vol. 1 and 2), Japanese Port and 5) FEMA (1997a): FEMA-302: NEHRP Recommended provisions for Harbor Association. seismic regulations for new buildings and other structures, Part 27) JRA (2002a): Speciˆcations for Highway Bridges and Commenta- 1—Provisions, Feb. 1997. ry: I. Common Design Principles, IV. Substructures, Maruzen, 556 6) FEMA (1997b): FEMA–303:NEHRP Recommended provisions for (in Japanese). seismic regulations for new buildings and other structures, Part 28) JRA (2002b): Speciˆcations for Highway Bridges and Commenta- 2—Commentary, Feb. 1997. ry: V. Seismic Design, Maruzen, 406 (in Japanese). 7) Hamburger, R. O. (1997): The development of performance based 29) JSCE (2002): Standard Speciˆcations for Concrete: Structural per- building structural design in the United States of America, Proceed- formance veriˆcation (2002 version), Japanese Society of Civil En- ings of International Workshop on Harmonization in Performance gineering (in Japanese). based Building Structural Design in Countries Surrounding Paciˆc 30) JSCE (2003): A comprehensive design code: code PLARFORM Ocean. ver.1.0 (Principles, guidelines and terminologies for structural de- 1000 HONJO ET AL.

sign code drafting founded on the performance based design con- 38) Ovesen, N. K. (1989): General Report, Session 30, Codes and Stan- cept), 29 (in Japanese). dards, 14th ICSMFE, Riodejaneiro. 31) JSCE (2008): A guideline for actions on civil structures based on 39) Ovesen, N. K. (1995): Eurocode 7 for geotechnical design, Proc. PBD concept, Structural Engineering Series (18), Japanese Society Bengt B. Broms Symposium on Geotechnical Engineering, Singa- of Civil Engineering (in Japanese). pore, 333–360. 32) JSSC (2001): A guideline for performance based design of civil steel 40) SEAOC (1995): Vision 2000—Performance based seismic engineer- structures, Technical Report (49), Japanese Society of Steel Con- ing of buildings, Vision 2000 Committee, Final Report, 1995. struction (in Japanese). 41) SEAOC (1998): SEAOC seismology PBE ad hoc committee, Per- 33) Meacham, B., Tubbs, B., Bergeron, D. and Szigeti, A. F. (2002): formance based Seismic Engineering Guidelines. Performance system model—a framework for describing the totali- 42) Shirato, M., Fukui, J. and Matsui, K. (2002): Present situation re- ty of building performance, presented at 4th International Confer- garding Geotechnical investigations and the determination of geo- ence on Performance Based code and ˆre safety design method, technical parameters in Japan, Proc. of International Workshop on March 2002, Melbourne, Australia. Foundation Design Codes and Soil Investigation in view of Interna- 34) MLIT (2002): Principles of design on civil and building structures, tional Harmonization and Performance Based Design, eds by Ministry of Land, Infrastructures and Transportation. Honjo et al., IWS Kamakura, Tokyo, Japan, 237–243. 35) Nagao, T., Watabe, Y., Kikuchi, Y. and Honjo, Y. (2009): Recent 43) Tamakoshi, T. (2006): Performance-based design speciˆcation and revision of Japanese technical standard for port and harbor facili- quality ssurance of highway bridges, Kyouryo-to-Kiso (Bridges and ties based on performance based design concept, Geotechnical Risk Foundations), August 2006, 19–22 (in Japanese). and Safety, Proc. of 2nd ISGSR and IS Gifu, 39–47. 44) Watabe, Y., Tanaka, M. and Kikuchi, Y. (2009): Practical determi- 36) Nakatani, S. and Shirato, M. (2006): Engineering Philosophy for nation method for soil parameters adopted in the new performance the Threshold Ductility Factor of Deep Foundations. Technical based design code for port facilities in Japan, Soils and Founda- Note of Public Works Research Institute (4030). Public Works tions, 49(6), 827–839. Research Institute, Tsukuba. 45) Wen, Y. K. (2000): Reliability and performance based design, 37) NKB (1978): Structure for building regulation, NKB report (34), Proceedings 8th ASCE Special Conference on Probabilistic 1978. Mechanics and Structural Reliability (CD-ROM), 21.