Application of Accelerated Bridge Construction Connections in Moderate-to-High Seismic Regions (2011)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

43 This report has reviewed recent advancements in ABC tech- niques that are either being used currently or show promise for use in regions of the United States that are subject to mod- erate-to-high seismic hazards. ABC techniques have been applied on many projects, primarily in regions of low seismic activity. However, their use in moderate-to-high seismic regions of the country has been limited, because the conven- tional, linear, precast elements used with ABC cause the con- nections to be located at the intersections of framing elements, and those locations are typically the regions expected to expe- rience the highest demands under earthquake loading. Accord- ingly, significant work is under way and more is needed to ensure that ABC connections can meet the required seismic performance, in addition to having the necessary non-seismic properties of constructability, cost effectiveness, durability, and inspectability. Conclusions The use of precast or prefabricated elements in bridges located in seismic areas can be characterized into two cate- gories, energy-dissipating and capacity-protected, and this sep- aration is useful in focusing the development of SABC. Bridge systems are designed such that the inelastic response that is unavoidably induced by the ground motion is concentrated in a few predetermined components. These components are typ- ically the columns, which act like fuses. Other components are thereby protected from the heavy loading demands and do not need to be designed for the more rigorous conditions experi- enced by the columns. Many of the connections that were reviewed participate in the dissipation of earthquake-induced energy. These elements are termed energy dissipating and they significantly influence the overall seismic performance of the bridge. The components and connections of a bridge that are protected by the fuse-like behavior of the columns are designed with capacities that are large enough to prevent damage from occurring in them. Such elements are denoted as capacity- protected connections between components and comprise the integral connection type that has been considered here. In some cases, emerging nontraditional concepts seek not to dissipate energy so much as act compliantly and accommodate seismically induced displacements with minimal damage. Such connections are called deformable, and are used in lieu of energy-dissipating connections at locations where inelastic deformations are expected. They may allow the bridge designer to improve the seismic response of the system by selecting an optimal distribution of moments within it. Significant knowledge gaps remain to be closed for energy- dissipating connections, so a focus of additional research should be on energy-dissipating and deformable connections. The reason for this is that capacity-protected connections may be designed largely with data that supports the use of these same elements in non-seismic areas. Typically, such design data exists. Gaps in the experience and knowledge base for these capacity-protected components must eventually be closed, but the energy-dissipating work is a more pressing impediment to implementation. Long-Range Needs The status of the existing state-of-knowledge and practice for SABC, coupled with the wide range of construction prefer- ences by owners and engineers around the country, suggests that a broad and extensive testing program will ultimately be necessary to fully support the use of SABC in the United States. Such a program should eventually include large-scale sub- assemblage (full pier) tests, as well as field demonstration proj- ects to build confidence in the use of SABC. Experience suggests that a single technology will not fit the needs of all the states with moderate-to-high seismic areas, especially in view of the fact that SABC technologies are a significant departure from conventional CIP systems. A realistic estimate of the total time to develop the required knowledge could be in the range of 20 years. Therefore, this section summarizes the broader, C H A P T E R 4 Conclusions and Suggested Research

long-term needs and proposes the first, but significant, step of a journey that will likely last a number of years. This assessment of duration is supported by recalling the time taken for the les- sons learned in the 1971 San Fernando (California) earthquake to be formally adopted into bridge design practice in 1990. Compared with conventional bridge technology (e.g., CIP), SABC elements are somewhat more complex and include a wide range of possible behaviors, which suggests that time will be required to fully develop the technology. Nevertheless, such a research undertaking is crucial for the implementation of ABC in seismic areas and is an important step for the states, the traveling public, and work-zone safety. Many ABC systems have been proposed for seismic use and limited testing has vastly improved the knowledge base and reduced the gaps in it. However, for owners and designers to have the confidence to deploy SABC technology, it is necessary to develop definitive design and construction specifications, design examples, demonstration projects and field experience. Future efforts should fill the remaining gaps in this knowledge in a systematic way. The objective is to provide the user with a palette of connections that can be constructed easily and that will accommodate inelastic deformations expected at inter- mediate piers—an essential “toolbox” for bridge designers in moderate-to-high seismic regions. An ancillary benefit of development work on SABC systems is that such systems may address other extreme events, such as vessel impact, blast, or other loadings that may load a bridge beyond its elastic limits in ways not addressed by design for gravity load alone. The design principles used for seismic load- ing require continuity of load path, reserve inelastic strength, and a high level of structural integrity, and such attributes directly benefit the structure for other extreme events with strong lateral effects. Previous NCHRP and state DOT-funded research to develop workable solutions to meet seismic performance requirements for ABC applications has produced a good start, and the sug- gested additional effort represents the next logical step on a longer journey. Leveraging existing knowledge and experience is necessary to prevent redundant research. An example of such leverage is the use of existing data for the design of capacity-protected connections. Deferring consideration of capacity-protected connections focuses the next stage of SABC work on a smaller universe of connections that are affected by the seis- mic demands placed on energy-dissipating and deformable- element connections. The overall strategy for the implementation of SABC should also account for owner preference in deployment of technol- ogy. There is a distinct preference apparent in the survey results for technologies that emulate CIP construction performance, and this is a manifestation of comfort level with the known. It is also apparent that there are emerging technologies that may provide seismic performance that is superior to that available from present designs, but the effort required to get those tech- nologies both to a level of maturity that will instill high confi- dence in owners and designers and to imbue them with characteristics of rapid constructability will take longer than that needed for CIP-emulative types. Thus, the recommended strategy is to give higher priority to development of the tech- nologies that align with current preferences and to take them to a deployable level. The review of existing technologies undertaken in this proj- ect led to the generalization of connections into seven types, which are listed in Table 34. The use of connection types in a bridge follows a building-block approach, where the overall bridge system is built of SABC connections, along with more conventional connections in non-seismic critical locations. For each of the seven connection types, the available informa- tion is insufficient to justify implementation as an SABC sys- tem in the field. The concept of TRL has been adapted for the use of SABC, providing rankings within nine categories that range from initial concept development to the system having successfully performed in the intended environment (in this case, a design earthquake). In addition to the ranking of TRLs, a judgment of the level of completeness at each TRL has been used and this helps identify three ways in which connection or system development may be deficient. The deficiencies fall into one of the three broad classifications defined above by the degree of completion of the steps in the TRL. The generalizations of the deficiencies are meant to represent a composite status for the type of connection under consideration. • The catch-up classification indicates that a step along the connection’s development is missing altogether and must be provided to justify the connection’s use at the highest level at which other information is available. In some cases, the connection may be in use today, despite the lack of information in one previous step. • The infill classification indicates that in one or more steps, the needed information has been partially, but not fully, developed. In some cases, the partial knowledge may jus- tify use of the technology in moderate seismic areas where 44 Connection Type Catch-up Infill Advancement Bar couplers Grouted ducts Pocket Socket Hybrid Integral Emerging Table 34. Work remaining by SABC connection type.

demands are somewhat lower, but not in higher seismic regions. • The advancement classification indicates that all the infor- mation up to and including a given step is available and the remaining developments are those needed to push the technology to a higher level of readiness. In an ideal world, all TRLs would be completed before moving on to the next level. In the real world of bridge engi- neering, this has not been done. The reasons for the foregoing classifications are as follows. Bar Couplers refer to devices that connect reinforcing bars for tension or compression using grouted sleeves or various types of mechanical connections. A wide range of tension capacities are available, but only a limited set of such couplers are potentially suitable for seismic applications. Of these, the grouted sleeves have been used in a number of applications, and several versions are commercially available. A primary shortcoming is that a comprehensive test series on grouted- types of bar couplers is lacking. Partial information is avail- able (tests on couplers in air under high strain rates; isolated tests on members connected using particular couplers, [e.g., Splice Sleeve Japan Ltd., undated, Riva 2006]), but tests covering the full range of behaviors under seismic loading have not been conducted. Open questions include not only the cyclic response of the couplers themselves, but also sys- tem effects, such as the influence of the coupler stiffness on the strain distribution in the plastic hinge zone and its effects on the strain penetration in the opposing connected element and on the overall deformation capacity of a coupled-bar system. Because grouted sleeve bar couplers have already been deployed in seismic regions, the paucity of cyclic performance data for that type of bar coupler represents a serious short- coming, so they are placed in the “catch-up” category. Grouted Ducts refer to the anchorage of bars from one ele- ment into another by means of grouting the projecting bars into ducts. The load is then transferred from the duct into the surrounding concrete by bond. A number of test programs have demonstrated the high anchorage capacity of grouted ducts under monotonic loading, but only a few tests have been conducted using inelastic cyclic loading. Other areas where more information is needed include the effects of the size of duct, the type of duct (particularly the nature and roughness of the corrugations), the location of the bar in the duct (eccentric or otherwise), group pullout failure, and transfer of the load from the duct wall through the concrete to neigh- boring reinforcement. Grouted ducts have been deployed in non-seismic applications, on the basis of the available static strength data. The connection is therefore placed in the “infill” category. In a Pocket Connection, bars projecting from the top of a column are fitted into a single void, or pocket, in the cap beam, which is subsequently filled with concrete. The primary information was developed in NCHRP Project 12-74. The sys- tem provides considerable promise, but mechanics-based design procedures are needed for the joint region and more extensive testing is needed to advance their development. The joint region includes not only the confined pocket itself, but also the surrounding region in the cap beam. The dimensions of that region are limited by the width of the cap beam and the size of the pocket, so it may be quite small and, therefore, highly stressed. In particular, the required quantity of tie and other confining reinforcement and procedures for computing it need to be established. The connection is placed in the “infill” category. In a Socket Connection, the footing or cap beam is CIP around a precast column, from which no reinforcement proj- ects. These are simple to fabricate and transport and offer excellent onsite constructability characteristics. Some cyclic testing has been conducted on both precast concrete and steel columns embedded in footings that are typical of bridge con- struction. Other studies have investigated footings suitable for buildings, but those results appear to translate poorly to bridge construction. A more extensive study is needed to define the relationships between the embedded length of the column, the column diameter, surface roughness, and confining reinforce- ment. Clear, mechanics-based design guidelines are needed for the design of the critical connection region. The connection is placed in the “infill” category. Hybrid Systems typically contain unbonded prestressing tendons that remain elastic at all times during an earth- quake and re-center the bridge system when the lateral load is removed—a highly advantageous characteristic for post- earthquake use. Hybrid systems differ from many of the others discussed here in that their primary purpose is to provide supe- rior seismic performance, with rapid erection seen as a desir- able, but not essential, additional feature. This ranking of priorities is the opposite of most of the other systems presented here. The principles of hybrid structures have now been well established in the vertical building industry and a number of such structures have been built. This level of development in a parallel industry, coupled with their demonstrated potential for improved performance, justifies their being treated differ- ently from other emerging technologies. However, details suit- able for bridge construction have not yet been fully worked out. This is the case partly because many variants on the con- figuration are possible. The primary questions include the choice between pre- and post-tensioning, use of bars or strand, corrosion protection, anchorage details in the footing and pier cap, confinement needed at the rocking interface, and so forth. Hybrid systems are, thus, classified as “advancement” in terms of needed work to advance to deployment. Integral Connections are taken here to mean the connec- tion between girders, cap beam, and columns that resists 45

longitudinal load at a pier. The girders may be steel or concrete, and the cap beam may be completely or partially precast. Such connections are likely to be designed as capacity-protected, so that any inelastic deformations are forced to occur in the top of the column, below the connection. Their expected capacity- protected behavior renders them different from most of the other connections described here. Many configurations are possible, and they tend to follow the dictates of the local bridge-building culture. The internal forces must be trans- ferred from bending in the girders to torsion in the cap beam and back to bending and shear in the columns. That load path is complex and its integrity can only be investigated with a large-scale test set-up. Consequently, very few such tests have been completed and those that have were on specialized sys- tems. Integral connections have been used in many bridges, with the connection details determined using conventional principles of structural design. However, such principles are least reliable when load is transferred between many elements and the geometry of those elements is complex. The system performance must be understood clearly to ensure that such connections really can behave as capacity-protected elements. The shortage of system information, combined with extensive field deployment, place the connection type in the “catch-up” category. Emerging Technologies are a number of technologies that offer promise of excellent seismic performance, but most are not particularly suited to ABC. Examples include specialized materials, such as SMAs and engineered cementitious compos- ites (ECCs) to improve toughness and or damping, elastomeric bearings to increase deformability, and so forth. Most of the technologies have not yet been sufficiently developed to permit evaluation of their promise for ABC, and only a few prelimi- nary tests have been conducted to investigate their inelastic response to cyclic loading. Therefore, they are classified here in the “advancement” category. The emerging technologies show considerable promise for excellent seismic performance, but the concepts will require significant additional effort to bring them to a high TRL and to deployment in the field. This cate- gory of technology will likely affect bridge seismic design prac- tice some years in the future. Such systems should be nurtured by continued development effort, but immediate SABC deployment will come from the other technologies that have been developed more completely and, in many cases, already tried in the field either in non-seismic regions or in specialized innovative projects. Suggested Research The objectives of future research for SABC are first, to address immediate needs for use of ABC in moderate-to-high seismic regions, then second, to address more promising areas. This project has identified and prioritized these needs, and those that will provide the quickest and most widespread value are suggested for near-term efforts. In the spirit of addressing the most urgent next steps first, the suggested work for immediate research should provide substantiation of seismic performance and further develop design and construction guidance for the following: • Bar coupler systems that have already been deployed in high seismic regions (i.e., grouted sleeves) • Connections for a complete pier or bent system, inclusive of top, bottom, and splice column connections using either grouted ducts or pocket-type connections at the top and socket-type connections for the bottom. By addressing the infill effort judged to be remaining for grouted duct, pocket, and socket connections, the suggested work will enable a complete pier system to be deployed with confidence in moderate-to-high seismic areas. Of these, the grouted duct and pocket connections are particularly well suited for the column to cap beam connection, and the grouted duct and socket connection types are suited to the foundation connection. Bar couplers, once verified, could be applied anywhere in the structure. Quasi-static, statically determinate tests are preferred for most of the testing because these permit, without ambiguity, the relationships of internal force and displacement to be quantified. Such data is necessary to support the development of design procedures compatible with the AASHTO Guide Specifications, which use displacement-based methodologies. Beyond such simple testing, eventual proof-of-concept tests should be performed on large-scale subassemblages using shake tables. This provides additional confidence in the tech- nologies under near-actual dynamic conditions. Because we cannot control the occurrence of large damaging earthquakes, which would provide actual field proof-of-concept, such dynamic testing is the next best thing to boost confidence in TRL toward the highest level. The next phase of work that is suggested, potentially sev- eral years in the future, is comprehensive evaluation and development of integral connections that form part of the load path for longitudinal seismic loading in common with prestressed girder bridges including the following: • Two-stage cap beams with a precast lower drop cap with- out prestress in the connection region • Flush-soffit cap beam types where longitudinal post- tensioning may or may not be used • Innovative connecting approaches beyond those currently in use for cap beams 46

Overall, testing and development of such integral systems is no less important than development of pier systems. Pier system development was prioritized ahead of integral con- nection development, in part, because such systems have been deployed, they are common throughout the country, and testing is somewhat less expensive due to specimen size. However, ultimately, both pier and integral testing should be undertaken. It is apparent that there is other infill work and much remain- ing advancement work that falls beyond the scope of the near- term suggestions. The reasons for not suggesting any of that work for immediate priority are as follows: • Adequate development work for even one emerging sys- tem would consume the entire likely budgets available in the near term. • There is a higher likelihood of achieving the highest readi- ness level in the shortest time for conventional bridges with the catch-up and infill work that has been recommended. Accordingly, an underlying assumption to the recommen- dations made herein is that the best approach in the near term is to focus on bringing the technology with the highest poten- tial to benefit the most users to a deployment-ready stage as fast as possible. This, in the research team’s judgment, is to bring a bent system to market that can be used with widely used precast girder conventional bridges. Beyond these near-term goals and to the extent possible, development should continue on technologies that will pro- vide enhanced seismic performance, in addition to enhanc- ing ABC, and that will benefit more specialized bridge types that are used in smaller numbers. 47