Innovative Bridge Designs for Rapid Renewal (2014)

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.

70 C h a p t e r 3 Overview Sixteen ABC design concepts developed by the R04 team in Phase I have been described with concept sketches and photographs in Chapter 2. They include new concepts, or adaptations of existing concepts, that are proposed as solu- tions to various ABC problems. Some concepts such as the modular bridge systems are truly complete systems and generally focus on the use of large prefabricated compo- nents in order to expedite construction. Many of the other concepts are at the elemental level, such as superstructure systems, deck concepts, various innovative column con- struction ideas, and so forth. They are not complete bridge systems, but they can be used together to form complete bridge systems. The work of Task 6 in Phase II was to incrementally win- now the collected findings and ABC concepts from Phase I through screening, more detailed engineering and con- structability evaluations, and assessment of implementa- tion challenges. The R04 team further developed and refined the more highly rated concepts in Phase II so that they can be readily implemented by state DOTs and other bridge owners with minimal additional design effort. The most promising technologies from the Task 6 evaluations have been recommended for standardization and use in field demonstrations in Phase III. Any technology recom- mended must meet minimum standards of readiness for execution, suitability for ABC, a promise of durability, and value to the owner. A key objective of SHRP 2 R04 is to produce pre-engineered ABC design standards for the recommended substructure and superstructure systems. By standardizing designs the opportunities for local or regional fabricators will be greatly increased. In many cases, standardizing designs also promotes the ability for local contractors to self-perform their pre- casting, thus encouraging broader acceptance of ABC and reducing costs. To perform thorough and consistent evaluations on the recommended concepts from Phase I, all Phase II evaluations were required to follow a series of steps, as outlined below: Step 1: Compile and summarize published and unpublished materials pertinent to the technology. Step 2: Perform an engineering evaluation of the concept that focused on the soundness of the underlying engineering design behind any particular concept. Step 3: Perform a constructability evaluation of the concept to evaluate issues specific to transportation of components, erection methods, equipment needs, and the suitability of the system to rapid construction. Step 4: Discuss implementation challenges and barriers to more widespread use of the recommended technologies and any specific obstacles that may inhibit the use of a technology. Step 5: Develop recommendations for testing for concept development and implementation. The Phase II evaluations provided a short list of concepts recommended for standardizing and Phase III implementation. Results of the evaluations are presented in three parts in this chapter: Part 1: Evaluation of Precast Decks and Complete Super- structure Systems; Part 2: Evaluation of Precast Substructure Systems; and Part 3: Evaluation of ABC Construction Technologies. In Part 1, the results of the evaluations for precast decks and complete prefabricated superstructure systems have been presented under the following headings: • Modular superstructure systems 44 Deck bulb tees; 44 Double tees; Findings and Applications

71 44 Decked stringer system; and 44 Decked trapezoidal box girders. • Segmental superstructure systems 44 Box girders; 44 Segmental slabs; and 44 Segmental voided slabs. • Precast decks. In Part 2, the evaluations for precast modular abutments and complete piers are presented under the following headings: • Precast modular abutment systems 44 Body; 44 Wings; and 44 Support options, piles, shafts, spread footing. • Precast complete pier systems 44 Whole pieces, footing, shaft, cap; and 44 Support options, piles, shafts, spread footing. • Segmental columns and piers 44 Segmental columns; 44 Pier caps; and 44 Footings. In Part 3, the construction concepts introduced in Phase I and subsequently carried forward for further evaluation in Phase II are presented. part 1: evaluation of precast Decks and Complete Superstructure Systems Summary This review of accelerated bridge construction (ABC) super- structures system is part of the SHRP 2 R04 research project Innovative Bridge Designs for Rapid Renewal, administered by the Transportation Research Board (TRB) of the National Academies. This TRB project intends to develop standardized approaches to designing, constructing, and reusing complete bridge systems that address rapid renewal needs and make efficient use of modern construction equipment. The work of Task 6 is to incrementally winnow down the collected findings from Phase I through screening and fur- ther evaluations. The purpose of these evaluations is to pro- vide recommended ABC concepts and techniques that can be advanced to standard plans. In this chapter, the results of the evaluations for precast decks and complete prefabri- cated superstructure systems have been presented under the following headings: • Modular superstructure systems 44 Deck bulb tees; 44 Double tees; 44 Decked stringer system; and 44 Decked trapezoidal box girders. • Segmental superstructure systems 44 Box girders; 44 Segmental slabs; and 44 Segmental voided slabs. • Precast decks. This review of superstructure design concepts documents each of the concepts, provides a review of the associated research literature, and provides a review of the engineering and constructability evaluations. The review also pinpoints implementation challenges and provides suggestions to over- come those challenges. In addition, testing needs and future research are also discussed. The engineering evaluation focuses on the soundness of the underlying engineering design behind any particular concept. This type of assessment is a more detailed review of a concept than that which was conducted during Phase I. The engineer- ing evaluation is undertaken to evaluate the recommended concept critically and to assess the quality of the underlying research, the suitability of any proposed design approaches, the quality of proposed specifications, and so forth. The constructability evaluation is aimed at assessing issues specific to the transportation of components, erection methods, equipment needs, and the suitability of the system to rapid construction. Part of the constructability evaluation is to document the potential time savings of an ABC technique as compared with conventional construction. An evaluation matrix for each option is included in the report. Each option is evaluated based on criteria such as initial cost, durability, system simplicity, market readiness for rapid construction, ease of evaluation for overload permits, and other factors. A score for each criterion is assigned on a scale of 1 to 5 with 1 being poor, 3 being average, and 5 being very good. Design Concept Descriptions Modular Superstructure Systems The intent is to develop pre-engineered standards for modular deck segments for concrete and steel bridge superstructures with spans of up to 140 ft that can be transported and erected in one piece. Longer spans, up to 200 ft, can be transported in sec- tions and spliced on site and erected using special techniques such as girder launching. Standardizing the designs to not more than five sections (for each of the three deck segments) that will cover the span range from 40 ft to 140 ft will increase their availability through local or regional fabricators, reduce lead times, lower costs, and increase familiarity among local con- tractors. For short spans, these segments can be purchased and erected in a few days by county crews using conventional

72 equipment. The deck segment concepts incorporate proven elements and details used by several states. Four options for standardized modular superstructure systems are presented: • Modular steel superstructure systems 44 Decked steel stringer system (two-beam steel sections with slab); and 44 Decked steel trapezoidal boxed girders. • Modular concrete superstructure systems 44 Concrete deck bulb tees with integral deck; and 44 Concrete double Tees with integral deck. DeckeD Steel Stringer SyStem The construction of the superstructure is a time-consuming part of cast-in-place (CIP) bridges; therefore, its prefabrication, in part or total, can significantly reduce construction time and traffic disruption. Increasingly, innovative bridge designers and builders are finding ways to prefabricate entire segments of the superstructure. Preconstructed composite units may include steel or concrete girders prefabricated with a composite deck that are cast off the project site and then lifted into place in one operation. One method of prefabrication involves constructing conventional composite stringer bridges off site and installing them rapidly on site. DeckeD Steel trapezoiDal Box girDerS Steel tub girder use is becoming more commonplace in modern infrastructure design. It offers advantages over other superstructure types in terms of span range, stiffness, and durability—particularly in curved bridges. Torsional rigid- ity is an advantage for spans with tight horizontal curvature. In addition, steel tub girders have distinct aesthetic advantages, due to their clean, simple appearance. Bracing, stiffeners, utili- ties, and other components are typically hidden within the box, resulting in a smooth, uncluttered form. The steel surface area exposed to the environment is greatly reduced, since half of the web and flange surfaces are enclosed. Welded steel tub girders are more costly to fabricate than plate I-girders. It takes highly skilled workers to fabricate and erect steel tub girders, which results in a premium on labor costs even if material costs are competitive with plate girder alternatives. An example of steel trapezoidal box girders is given in Figure 3.1. A folded plate bridge system offers an economical solution over welded box girders for short-span bridges. The system consists of a series of standard shapes such as trapezoidal tubs or boxes that are built by bending flat plates using a brake press. Bending plates to specified shapes is rapid and very economical when compared with welded construction. A folded plate bridge system can be constructed using light construction equipment and can provide long service life with minimal maintenance. Almost 45% of the bridges in the U.S. bridge inventory are less than 60 ft in length. Most are simple spans located on county roads. Span length for this sys- tem without a splice is currently limited to about 60 ft, reflect- ing the longest press brakes that are available in the industry. Folded plate girders suitable for different span lengths differ only by their cross-sectional dimensions. Span lengths longer than 60 ft can be developed using spliced sections. More spe- cifically, varying the width of the top and bottom flanges and the depth of the web while keeping the plate thicknesses to typically ½ in. can accommodate span length requirements. The different top and bottom flange widths and web depth can easily be accommodated by changing the bend locations, so fabricators can build folded girders quickly while stocking only one or two plate thicknesses. That will assure delivery of steel bridge girders without long lead times. Plate sizes are available in 10-ft widths from four U.S. manufacturers. Sec- tions that can be fabricated within these width limitations will aid the standardization process. Sections can also be galva- nized for a small additional cost. Initially, only simple spans are planned but the design can be economically extended to continuous girders while also preserving the rapid renewal advantages. The sections can be fabricated with a precast deck in the yard prior to shipping to the site. A 60-ft folded plate girder with precast deck should weigh less than 30 tons and can be easily erected without the need for heavy cranes. In some cases, even county crews will have the capability to erect such bridges. concrete Deck BulB teeS anD DouBle teeS Precast concrete deck girders are becoming increasingly pop- ular as an economical solution for short-span and medium- span bridges. For local bridges, segments for short spans can be purchased and erected in a few days by county crews with conventional equipment. The top flange is designed to func- tion as an integral deck, making them an attractive option for Figure 3.1. Steel folded plate trapezoidal box girder.

73 rapid replacement applications. Omitting the concrete top- ping and the transverse posttensioning will significantly reduce the on-site time for construction. The top ¼ in. can be ground after installation to achieve a smooth riding surface. Use of an overlay system is optional. The added protection from an asphalt overlay and membrane can increase service life. Cast-in-place longitudinal joints between girders can be designed to allow full moment transfer or only shear transfer. Recent advances in joint design have greatly enhanced dura- bility and performance. The type of joint depends on the traffic exposure or the functional classification of the route. Control of camber is a key consideration for constructability and rideability. Several states, such as Washington and Idaho, have standardized deck girder sections using the bulb tee and double tee configurations to increase their availability. These girders can also be spliced in the field by posttensioning to allow longer spans than is feasible using a single transport- able section. Segmental Superstructure Systems The segmental superstructure system consists of short sec- tions that can be connected to each other to form the entire superstructure for a bridge. Segmental bridge sections are either precast or cast-in-place sections. The segmental super- structure system provides a number of advantages. Long spans are possible. The finished structure is durable and aesthetically pleasing. Erection methods can be easily adapted for safe and rapid construction over existing roadways, rivers, and other obstructions. Highly skewed supports are easily accommodated by this system. Segmental bridges are generally very economical for lon- ger spans. They are the design of choice for spanning deep valleys and wide water crossings, and across highways and existing facilities without the use of costly false work. Precast segmental construction is also typically used with long via- ducts. Precast segmental technology can be readily adapted to typical highway bridges with very little effort. Sections can easily be adapted from traditional boxes to solid or voided slabs sections, or other less common sections such as chan- nel bridge sections. Technologically, these adaptations are relatively simple. Three types of precast segment sections are suitable for ABC applications: • Conventional single cell box girders; • Solid slabs and voided slabs; and • Channel sections. These typical cross sections can be used in simple grade separations, span-by-span viaducts, or continuous, balanced cantilever-type construction. conventional Single cell Box girDerS Conventional single cell box girders have been used in dozens of applications across the United States in the past 30 years. They are typically cost-effective for longer spans (continuous versus simple). They are commonly used for span-by-span highway or tran- sit construction in the 125-ft to 150-ft span range. They can be extended up to 180 ft by using continuity posttensioning. The United States is beginning to see the application of single cell box girders more and more on typical highway construction for interchanges, overpasses, and grade separations. In Minne- sota, single cell box girders were used on various interstate reconstruction projects, and in Indiana, they have been employed along I-80 and I-90. SoliD SlaBS anD voiDeD SlaBS This technology represents the greatest opportunity for man- ufactured bridges and ABC. These solutions have been used widely in the United States and Canada over the past half century, but mostly for CIP applications. They are typically the most cost-effective and durable bridge types available for spans up to 130 ft. Examples for segmental solid and voided slab bridges are shown in Figures 3.2 and 3.3, respectively. The use of solid deck slab panels is a standard approach for the composite deck of cable-stayed bridges with spans of 12 ft to 25 ft. Posttensioned solid slab spans are economical up to about 80 ft. Voided slab spans are economical up to about 135 ft. This technology is readily adaptable to solid slab or voided slab segments and can be accomplished economi- cally. It has been done elsewhere. A real market exists in the United States for a manufactured ABC bridge replacement technology. The aging U.S. interstate infrastructure drives the need for a durable replacement technology that can be applied on a large scale for “bread-and-butter” grade sepa- ration bridges. Figure 3.2. Segmental solid slab bridge.

74 Segmental channel SectionS These sections have not been widely used in the United States, or other markets. They are limited in width (typically two lanes) without introducing transverse ribs or floor beams, include some casting issues due to their flexibility, and are mentioned only to demonstrate the feasibility of using sim- plified erection girders and small picks. An example is shown in Figure 3.4. Precast Concrete Bridge Decks A variety of precast bridge deck technologies have been used for the past 50 years to facilitate rapid construction and ease traffic congestion. Precast bridge deck panels can be used in place of a CIP con- crete deck to reduce bridge closure times for deck replacements or new bridge construction. The precast panels are most com- monly prefabricated at a casting yard, which provides optimal casting and curing conditions and which normally results in durable, long-lasting decks. For smaller projects, or those located beyond reasonable access to a PCI-certified precasting plant, a number of precast bridge deck panels have been site- cast by the bridge contractor. Some of the challenges inherent with any bridge deck tech- nology include: performance of deck panel joints, accommo- dation of a normal crown, composite action with girders, and attachment of bridge railings. In general, precast concrete bridge decks can be classified as either partial-depth or full-depth. Partial-depth precast concrete deck panels (PCDP) are often used to provide a durable, stay-in-place concrete form system that is later supplemented by 3 to 5 in. of CIP concrete. Although these partial-depth panel systems are valuable in providing a durable bridge and have been standardized for use on both conventional and long-span bridges in a number of states, the need for a cast-in-place concrete structural topping does not make them suitable for use in an ABC environment. Therefore, the remainder of this section will focus only on full-depth concrete deck panels. To further narrow the focus of these recommendations, the PCDP described here include those panels that span trans- versely between conventional prestressed concrete beams or steel girders. These pieces, typically limited in width to a size transportable without specialized permits, are usually 8 to 10 ft wide and 8 to 9 in. thick. The vast majority of bridges in the United States are two lanes wide and include moderate shoul- der width of 8 to 10 ft. To eliminate the often troublesome lon- gitudinal joint, PCDP are typically designed for the full width of the roadway, including barriers. Approximate lifting weights for full-depth concrete deck panels range from 15 to 20 tons. Figure 3.5 schematically depicts a prestressed concrete beam bridge with full-depth precast deck panels. The construction process consists of the following: 1. Installing the panels on top of the beams. The self-weight of the panels is transferred to the beams through leveling bolts protruding through the bottom of the panels. The haunch depth is adjusted to provide the desired top-of- deck elevation. 2. The haunch areas are filled with grout or concrete between the top of the beam and the bottom of the deck. The trans- verse joints are filled next. Figure 3.3. Segmental voided slab bridge. Figure 3.4. Segmental channel section.

75 3. If the deck is to be posttensioned, this operation is then performed. After the posttensioning operation is com- plete, the posttensioning ducts are pressure-injected with a corrosion-resistant grout. 4. The haunch is placed after the posttensioning operation. 5. To obtain composite action between the deck and under- lying beams or girders, a mechanism must be used to pro- vide shear transfer between the panels and beams. This is normally accomplished through the use of shear connec- tor blockouts arranged to match with either shear rein- forcing extended up from prestressed beams or welded studs on steel girders. 6. Once the grout in the haunch has cured, the leveling bolts are removed and the panels and beams act as a composite system. 7. Barrier rails are then cast. 8. A wearing surface may be placed if desired, but the time involved for this operation could significantly reduce the ABC benefits of the system. The shear pockets provided in each precast panel are designed to fit over shear studs (in the case of steel girders) or shear rein- forcing (in the case of prestressed beams) and the pockets are grouted to ensure fully composite action. Figure 3.6 depicts a typical shear pocket for use with a steel girder. ReinfoRced concRete deck Panels Precast reinforced concrete panels are only reinforced using conventional mild steel. These panels are designed per AASHTO requirements for service, strength, and serviceability. The Figure 3.5. Anatomy of a precast concrete deck panel system. Figure 3.6. Precast deck panel with shear pocket.

76 panels can be designed as either continuous over a series of par- allel girders or simply supported over the longitudinal girder. Figure 3.7 depicts an example of precast reinforced concrete panels used by the Utah DOT on the C-437 bridge rehabilita- tion of the county road over I-80 to Wanship. The precast pan- els are conventionally reinforced with two layers of epoxy-coated steel bars in each direction. The longitudinal reinforcement is designed to resist the negative moment over the piers resulting from the superimposed dead and live loads applied when the deck is made composite with the superstructure. PRestRessed concRete deck Panels Perhaps the most common type of panel in use today, pre- stressed concrete panels are typically transversely pretensioned (perpendicular to the driving direction) and conventionally reinforced in the longitudinal direction. The panels are reinforced for temperature and shrinkage on both directions. Prestressed panels must be longer to meet the transfer and development length criteria of strands. The main advantage of prestressed panels over reinforced panels is that the permanent compressive force provides a crack-free deck. Posttensioned concRete deck Panels Posttensioned concrete deck panels are very similar to the prestressed panels described above, except that the main reinforcement in the panels is provided by a posttensioning system rather than prestressing. The posttensioning force is typically provided in the longitudinal direction, but can also be applied in the transverse direction for especially wide bridges or those that require staged construction. The panels are generally lightly reinforced to resist self-weight, tem- perature, shrinkage, and creep. Figure 3.8 depicts typical post- tensioned concrete deck panels used for the recently constructed Bill Emerson Memorial Bridge, spanning the Mississippi River in Cape Girardeau, Missouri. The precast concrete deck panels are conventionally reinforced with top and bottom meshes of epoxy-coated bars. The posttensioning is provided in the longitudinal direction. The thickness of the precast panels is 10 in. Figure 3.9 illustrates an example of deck posttensioning in both longitudinal and transverse directions for the Door Creek project on Interstate 39/90 in Wisconsin. The precast system consists of full-depth precast concrete deck panels, which were constructed off site and delivered to the site ready for placement. Because the panels were posttensioned in both longitudinal and transverse directions, ducts were placed in both directions so that they would not interfere with each other. The longitudinal posttensioning duct is located at mid- depth of the slab, while the transverse posttensioning ducts are placed above and below the longitudinal ducts. The panel thickness for the Door Creek Bridge is 8¾ in., which is similar to a typical cast-in-place deck. One significant advantage of using posttensioned deck panels to improve an existing bridge is that they are often Figure 3.7. Reinforced concrete deck panels for the Utah DOT. Source: Missouri DOT. Figure 3.8. Posttensioned deck panels for the Bill Emerson Memorial Bridge, Missouri.

77 thinner than a CIP deck due to the existence of internal ten- dons in one or both directions. Thinner panels translate into a lighter-weight deck, which could improve the live-load rat- ing of older bridges. In addition, the use of high-performance, high-strength materials allow precast decks to be significantly lighter than a conventional cast-in-place system. Although the use of exotic fiber composite material can reduce the panel loads even further, AASHTO LRFD service- ability requirement limit how far this envelope can be pushed. UltRa-HigH-PeRfoRmance concRete Panels (Waffle slab) An innovative precast concrete bridge deck system that is cur- rently being developed uses an ultra-high-performance con- crete (UHPC) waffle slab system that is designed to provide the equivalent stiffness of a solid concrete panel while remov- ing much of the normal dead load. This system is currently undergoing testing at both the FHWA Turner-Fairbank lab- oratory and Iowa State University. A demonstration bridge project, funded through the FHWA Highways for LIFE pro- gram, was constructed in Wapello County, Iowa, in 2011. The use of UHPC materials, which offer much higher strength and significantly improved resistance to intrusion of chlorides when compared to conventional concrete, offers future potential as a lightweight, durable bridge deck system. However, the material costs remain rather expensive and the technology is still being improved. UltRa-HigH-PeRfoRmance concRete foR deck Panel Joints The New York State DOT is currently investigating the use of full-depth precast deck panels with field-cast UHPC joints to develop the continuity in the deck panels. This solution had been previously attempted by the Ontario Ministry of Trans- portation on an experimental basis, but the New York State DOT has been using this solution for rapid replacement of bridge decks in high-traffic areas. The New York State DOT found that the strength and ductility properties of UHPC functioned well as a joint fill material when combined with precast deck panels. The New York State DOT specifies that full-depth precast concrete deck panels are designed with HPC and use epoxy- coated or galvanized reinforcing steel. The agency further spec- ifies a minimum of 40 MPa (4,800 psi) compressive strength. The panel reinforcement design was based on continuity through the joints. UHPC joint material was assumed to pro- vide sufficient bond development to develop full continuity of the rebar just as if it were continuous through the joint. The strength and low permeability of UHPC provides excel- lent protection of the rebar against corrosion and improved bond with the rebar, thereby providing short bond develop- ment lengths. Testing has shown that the bond development length of a #4 bar in UHPC is less than 3 in. UHPC also offers excellent bond development length, resistance to freeze and thaw cycles, and high flexural strength and toughness, which provides the critical resistance to flex- ural loads generated by truck loads passing across the joint. Results of the New York State DOT indicate that the UHPC/ HPC deck interface is bonded with no potential for leaking. The New York State DOT conducted a demonstration proj- ect selected using field-cast UHPC joint fill on a 127-ft long, single bridge with full-depth precast deck panels supported on steel beams near Oneonta, New York, as shown in Figures 3.10 and 3.11. The UHPC joints were 6 in. wide and 8 in. deep. Figure 3.9. Posttensioned deck panels for the I-39/90 bridge in Wisconsin. Figure 3.10. The New York State DOT Demonstration Bridge in Oneonta.

78 Following installation of the full-depth precast deck panels, the panels were adjusted and leveled for grade and a smooth, flush riding surface. UHPC joint material was transported to the joints by power buggy and then dumped directly into the joints without any vibration, which is an acceptable prac- tice for this material. The joints were covered with form grade plywood strips and allowed to cure until reaching 100 MPa (14,500 psi) before being opened to traffic. This cure time required approximately 3 days, but could be reduced though the use of an accelerator and heat. The field mixing of UHPC joint fill proves that this material can be batched on site and provide adequate strengths during typical field curing conditions and that local contractors can easily adapt to using UHPC in bridge projects. DUCTAL is the only supplier for UHPC. The steel fibers are from a European supplier subject to Buy America provisions. It should also be noted that match cast and posttensioned joints are well established. They are acceptable alternatives for ABC for which designers can find information from other sources. Engineering Evaluation The engineering evaluation focuses on the soundness of the underlying engineering design behind any particular concept. This type of assessment is a more detailed review of concepts than that conducted during Phase I. The engineering evaluation is undertaken to critically evaluate the recommended concept and assess the quality of the underlying research, the suitability of any proposed design approaches, the quality of proposed specifications, and so forth. The engineering evaluation covers the two important aspects of modular construction: the pre- fabricated elements and systems, and the connections. The evaluation begins with a discussion of prefabricated elements and systems. Modular Superstructure Systems The intent of this project is to develop pre-engineered stan- dards for modular deck segments for concrete and steel bridge superstructures. It is critical that only concepts that have been thoroughly vetted be advanced to the following tasks. Stan- dardized sections need to be versatile so that they can be used in varied situations, provided they meet the decision cri- teria for selecting prefabricated construction. Design consid- erations for standardized modular superstructure systems should include the following: • Pre-engineered standards for modular construction. Designs that can be used for most sites with minimal bridge specific adjustments. • Optimized designs for ABC and use of high-performance materials. Simplicity and efficiency of design, availability of sections, and short lead times are key considerations. • Segments that can be used in simple spans and in continu- ous spans (simple for dead loads and continuous for live loads). Details to eliminate deck joints at piers. Details for live load continuity at piers to be included for use as required. • Use of high-performance materials: HPC/UHPC concrete, HPS or A588 weathering steel. Consider lightweight concrete for longer spans to reduce weights of deck segments. • Deck tees and double tees with minimum 8-in. flange to function as decks with integral wearing surface so that an overlay is not required. Use of overlay is optional as part of a long-term preservation strategy. • Limit the number of standardized designs for each deck type to five, which should cover span ranges from 40 ft to 140 ft. Consider steel rolling cycles and sections widely available. • Segments designed to be used with either full moment con- nection between flanges or with shear-only connections. Each flange edge needs to be designed as a cantilever deck overhang. • Skewed bridges: The complexity of the geometry makes pre- fabrication for these types of bridges a challenge. This does not preclude the possibility of using prefabrication for these structures. Prefabricated bridge replacement projects have been completed on bridges with significant skews. Attention to tolerances and field fit-up is essential for these complex structures. • Prefabricated components can be the most cost-effective solution for any alignment. However, straight alignments allow multiple identical components, which tend to be the most economical. The alignment will affect super- structure member types. Curved alignments also typically require shorter segments in order to be transported over city streets. Initial construction costs and long-term main- tenance costs are typically less for bridges on straight align- ments due to their simpler construction and load paths. Preference should be given, if possible, to straightening the roadway alignment along the bridge length for lower life-cycle costs. Figure 3.11. The New York State DOT precast panel joint detail for UHPC.

79 Modular superstructure systems are particularly suited to be used as Tier 1 concepts for weekend bridge superstructure replacements or as Tier 2 concepts, when the entire bridge may be scheduled to be replaced within a month using a detour to maintain traffic. DeckeD Stringer SyStemS Prefabricated decked steel stringer systems have been a very popular option for accelerated construction of bridges in this country. Their light weight, easy constructability, low cost, and availability were seen as advantages over other systems. The length and weight of each module can be designed to suit transportation of components and erection methods. Erection can be made using conventional equipment. Cast-in-place clo- sure pours or grouted or welded joints are typically used to connect adjacent units in the field. The modules can be made to different widths to fit site and transportation requirements. They can be fabricated with square or skewed ends. Steel stringers/girders with precast decks have become increasingly common in steel bridge construction. Advantages include the following: • Improved efficiency with lighter steel beams (shored construction). • Uses standard rolled shapes and welded plate girders. • Economical or average construction costs. • Can be fabricated with exact camber and skew to meet existing site requirements. • Top of deck can be textured for riding surface. • Easy and rapid erection and construction. • Suitable for use as continuous spans. • Durable, since deck is cast in controlled conditions. Inverset-type concrete deck and steel composite systems have had a successful track record as an economical alterna- tive for rapid superstructure replacement. Many such bridges have been built over a weekend, which demonstrates their constructability even in congested urban locations. Some examples summarized here are provided to illustrate their suitability of the system to rapid construction: • In Virginia, the I-95 bridge over James River had 102 super- structure spans replaced in just 137 nights, with no impact to rush-hour traffic. Full-span-length prefabricated super- structure segments, up to 114 ft long, were fabricated at a nearby casting yard and transported using conventional flat- bed trailers. Each prefabricated segment consisted of three steel plate girders with an 8¾-in. deck. During the night, the old segments were removed, and cranes then installed the new prefabricated superstructure segments. • Also in Virginia, the US-15/29 bridge over Broad Run was completed with road closures on three weekends to replace the 12 superstructure segments, one span per weekend. Each segment consisted of two rolled steel beams made composite with a concrete deck. • In New Jersey, each of three bridges along Route 1 was replaced during a weekend closure. Each superstructure span consists of five full-length segments with two steel girders and a 9-in.-thick composite concrete deck (Inverset) system. It is common to cast these units in an inverted position at a prefabrication yard so that the deck is in compression in the final condition. Such a casting method may not always be fea- sible if the contractor is self-performing the precasting at the bridge site. Even if the deck is precast under conventional shored conditions, this modular system will provide the ben- efit of shored construction where the dead load is carried by the composite section. It should be noted that the beams are designed for non-composite dead loads in consideration of future deck replacement. One advantage of decked steel mod- ular systems is that they allow the replacement of the deck, while the steel stringers can be reused. Use of preconstructed composite units is relatively new. Accordingly, the performance of these systems is not well documented. There is less experience with and less literature discussing decked stringer systems than exist for other con- ventionally built systems. Although no literature was found on problems specifically associated with decked stringer sys- tems, it is assumed that they could experience problems simi- lar to those suffered by full-depth deck panels and prestressed concrete multi-beam superstructures. One key issue could be the longitudinal and transverse joints between the units and the durability of these joints. Durability of these joints can be achieved through proper detailing, as discussed in this section. Different top slab elevations of adjacent units could also be another key issue, which could be adjusted through a leveling procedure. DeckeD Bent plate Steel Box girDer SyStem A bent plate bridge system requires steel plates to be cold formed or bent in a press brake. Cold forming involves plastic deforma- tion of the metal surface on the outside of the bend. Cold form- ing results in strain hardening of the material and this in turn affects mechanical properties. The extent to which the plastic deformation can take place without exceeding the limits of material ductility controls the minimum radius of bend. As the ductility and fracture toughness decrease in the areas subject to plastic deformation, it will be necessary to determine the radius of bend and the material properties and steel types suitable for bent plate girders. In the Japan bridge specifications, an allow- able bending radius for cold bending is set at 5t, where t = plate thickness, to ensure proper fracture toughness after cold bend- ing. Bent plate connections have been successfully used on highly skewed bridges in the United States over a long period.

80 Generally, low carbon content is a prerequisite to good formability in bent plate applications. Steel plates are more readily formed with the bend axis transverse to the rolling direction of the plate. Though not warranted for bridge girder applications discussed here, it should be noted that heat treat- ment can eliminate all traces of cold working in steel plates. Engineering considerations pertinent to cold-formed steel applications are the effects of strain hardening and consequent increase in strength with reduction in fracture toughness and ductility. In the past, fracture and cracking concerns have impeded the adoption of cold bending for bridge structures. The recent availability of high-performance steel (HPS) with high fracture toughness has largely eliminated this issue and has provided a significant impetus for cold-formed applications. Though it is important to recognize and account for the effects of cold working on girder behavior, it should not be a limiting factor anymore for the use of bent plate girder systems as an economical and rapidly constructible bridge alternative. ASTM A709 Grade HPS 50W is contained in A709-01 and is produced by using conventional hot-rolling up to 4 in. thick in lengths similar to Grade 50W steel. HPS is produced to a lower carbon content than grade 50W. The fracture toughness of high-performance steel is much higher than conventional bridge steels. The brittle-ductile transition of HPS occurs at a much lower temperature than conventional Grade 50W steel. This means that HPS 50W remains fully ductile at lower tem- peratures where conventional Grade 50W steel begins to show brittle behavior. The fatigue resistance of high-performance steel is controlled by the welded details of the connections and is not a concern with bent plate structures. HPS has the ability to perform without painting under normal atmospheric conditions. HPS steel has enhanced atmospheric corrosion resistance that is even better than the conventional grade 50W steel did. The same guidelines and detailing practice for con- ventional weathering grade steel should be followed to assure successful applications of HPS steel in the unpainted conditions. The cost-effectiveness of HPS has been demonstrated by the design and construction of HPS bridges in many states. A steel/concrete composite bridge had been proposed in Japan with cold-formed steel U-girders, which are filled with concrete and partially prestressed near the intermediate sup- ports of a continuous bridge. This U-girder is cold formed from a single steel sheet. Laboratory tests were performed on cold-formed U-girder models to investigate bending behav- ior (Nakamura, 2002). These models were about one-fourth of the preliminary designed bridge, with a span of 60 m. Bending tests were carried out to investigate the static bend- ing behavior of the girder models in the positive and negative bending moment areas. The girder model at the span center behaved as a composite beam. In all the cases, the bending strength of the girder models was higher than the calculated yield moment, and nearly reached or exceeded the plastic moment. The U-girder section is therefore regarded as the compact section, and the plastic design can be applied for the proposed bridge. The proposed bridge system with cold- formed steel U-girders has sufficient bending strength and good deformation and rotation capacity, and it is feasible and economical. aDvantageS of Bent plate Steel Box girDer SyStem • Economy of steel section compared with I-girder system: The elimination of cross frames reduces steel weight. The box girders can be designed as continuous for live loads. Composite action is achieved with the precast deck. • Economy of construction: Precasting of the deck and elim- ination of cross frames eliminates time consuming and expensive components of the structure. • Reduced weight and increased underclearance: Total super- structure weight and depth are reduced when compared with concrete superstructures or I-girder steel systems. Ver- tical underclearance is increased. This avoids overloading substructures. • Allows future widening: Superstructure widening can be done quickly and effectively without disrupting traffic. Fascia girders can be produced to accept future widening. • Cost competitive: Cost competitive with concrete or other steel superstructure systems. • Torsional rigidity: Allows the use on spans with tight horizontal curvature. • Aesthetics: Clean appearance for very visible structures. DiSaDvantageS of Bent plate Steel Box girDer SyStem • Specification support: AASHTO LRFD Specifications do not address the design of cold-formed structures. How- ever, the AISC Steel Construction Manual does address cold bending. • Fatigue resistance: There are potential problems of fatigue resistance at the longitudinal bend locations. This can be alleviated through the use of HPS. Direction of plate rolling needs to be considered in cold bending applications. • Inspection access: The optimum box depth is structurally less than ideal for physical access for maintenance crews. Because the section does not include welded details, internal access would not be necessary. Weep holes at the bottom can be provided for drainage as necessary. • Press brake limitation: Maximum length and thickness of plates that can be cold bent are limited by press brake capabilities. Inquiries made by HNTB show that current manufacturing capabilities include cold bending a 5⁄8-in. steel plate up to 54 ft long. This could allow bent plate box girder spans up to 100 ft with a single splice at midspan. Press brakes can bend plates over 1 in. thick, but in smaller lengths.

81 conceptual DeSign for compoSite Bent plate Box girDer SyStem A preliminary load and resistance factor design (LRFD) was performed for a bent plate box girder system. The concrete deck width and thickness were taken as 8 ft and 8 in., respec- tively, and the concrete compressive strength was assumed as 4,000 psi. A 2-in.-thick asphalt wearing surface was assumed to be incorporated in the dead weight calculations. The material for the steel plate was 0.5-in.-thick ASTM A709 Grade HPS 50W. In addition to its more ductile performance at low tem- peratures compared with conventional steel, the enhanced fracture toughness of high-performance steel results in less microcracking during the cold forming process, thus improv- ing fatigue performance. The preliminary design of the bent plate system is based on the current AASHTO LRFD Bridge Design Specifications, 4th ed., box-section flexural members section (AASHTO LRFD, Section 6.11). Although reasonable results were achieved, it should be noted that current AASHTO LRFD specifications do not specifically cover cold-formed steel members. The conceptual design was performed for a simply sup- ported bridge with a span length of 60 ft. The box girder was dimensioned to carry the factored weight of components, the wearing surface, and the HL93 design live load; including a 33% dynamic load allowance for the Strength I limit state. The factored loads that were used in the evaluation are given in Table 3.1. To perform a parametric analysis, the available sheet plate width was taken as 10 ft. This also corresponds to the maximum width that can be bent at the shop. Since the deck thickness and width were known, by keeping the top flange of the box girder constant at 2 in. by 6 in., it was possible to directly calcu- late the exact location of the plastic neutral axis for different values of the box girder depth (d) and the inclination angle. Trial analyses were performed for three different depth values: 24 in., 36 in., and 48 in. The lower and upper boundaries for the inclination angle were chosen as 30° and 90°, respectively. The conceptual cross section is shown in Figure 3.12. Results from the trial analyses are summarized in Table 3.2. The relationship between the ratio of flexural capacity to flexural demand and the inclination angle is also shown in Figure 3.13. Table 3.1. Load Effects on 60-Ft Trapezoidal Box DC DW LL 1.25DC  1.50DW  1.75LL Shear (kips) 30.3 3.9 100.0 219 Moment (kips-ft) 454.5 58.5 1,356.0 3,029 Note: DC = dead load of structural components and nonstructural attachments; DW = dead load of wearing surfaces and utilities; and LL = vehicular live load. Figure 3.12. Decked steel bent plate box section— conceptual design. Table 3.2. Parametric Analysis Results d (in.) Inclination Angle (degree) Capacity-to- Demand Ratio 30.00 1.14 40.00 1.30 50.00 1.39 24 60.00 1.45 70.00 1.48 80.00 1.50 90.00 1.51 50.00 1.69 60.00 1.82 36 70.00 1.90 80.00 1.94 90.00 1.95 70.00 2.10 48 80.00 2.17 90.00 2.20 Trial analyses showed that the ultimate capacity of the sys- tem is always governed by flexure rather than shear. However, even with a depth of 24 in. and a relatively shallow inclination angle of 30°, the bent plate box girder system is capable of carrying the factored dead and live loads. Depending on the deck geometry and the material availability (the maximum width of the sheet plate that can be bent at the shop is 10 ft), it is possible to improve the flexural performance by increas- ing the depth of the box girder, although an increase in depth from 24 in. to 36 in. resulted in a greater gain in strength, when compared to an increase in depth from 36 in. to 48 in. In addition, the rate of strength gain also seems to diminish at higher inclination angles.

82 AASHTO LRFD Specifications also include limits for the cross-sectional proportions, most of which are based on the thickness of the steel plate. Since the steel plate thickness is fixed at 0.5 in., it is possible to calculate upper-bound values for the depth of the box girder, and also for the upper flange width: • Per AASHTO LRFD 6.11.2.1, maximum depth for the box girder is 75 in. • Per AASHTO LRFD 6.11.2.2, maximum width for a single upper flange is 12 in. Although the specifications require the thickness of the flanges to be at least 10% more than the web thickness (AASHTO LRFD, 6.10.2.2–3), it is not possible to satisfy this clause, since the box girder has an overall uniform thickness. Per AASHTO LRFD, 6.11.2.1.1, the inclination of the web plates to a plane normal to the bottom flange should not exceed 1 to 4. This requirement is met in the conceptual design trials. concrete Deck BulB teeS anD DouBle teeS With integral Deck These superstructure systems have been used by various DOTs including those of Washington State, Idaho, and Utah. These states have developed standards for these systems, which is indicative of the use in each state. The Utah DOT has developed the Precast Bulb Tee Girder Manual. The purpose of this manual is to provide guidance with the design and detailing of precast prestressed concrete bulb tee girders. The manual discusses the design, detailing, fabrication, and handling of precast prestressed girder bridges. The girders may be pre- tensioned or posttensioned. Three families of bulb tee girders have been developed by the Utah DOT. The girders are based on the Washington State DOT WF Series girders. The girder depths range from 42 in. to 98 in. in 8-in. increments. The Washington State DOT deck bulb tees have widths up to 6 ft and 6 in. minimum thickness. The Washington State DOT standard details cover square ends and skewed ends, up to a maximum 30° skew. Welded and grouted joint details are shown between the units and an asphalt overlay is required. The typical unit widths for the double tee units vary between 8 ft and 11 ft. The Washington State DOT and the Precast/ Prestressed Concrete Institute Northeast have developed standards for these types of beams. As indicated in these standards, double tee with integral deck units are currently feasible for spans up to approximately 90 ft using concrete strengths of 10 ksi. This modular superstructure concept involves deck bulb tee beams and double tees with a welded plate connection between the deck units for bridges subject to light-to-moderate traffic, and with full moment deck closure pours for bridges subject to heavy truck traffic. Both options are acceptable from an engi- neering viewpoint; however, the welded plate connection option should be limited to bridges that carry light truck traf- fic and are located in the regions of low seismicity, where seis- mic design of bridges is not required. The option with full moment deck closure is recommended for bridges that are expected to carry moderate-to-heavy truck traffic and are located in zones that require bridges to be designed for seis- mic design loading. The option with full moment connection is market ready, but the option with welded plate connection 1 1.25 1.5 1.75 2 2.25 2.5 30.00 40.00 50.00 60.00 70.00 80.00 90.00 100.00 Fl ex ur al Ca pa ci ty / De m an d Inclinaon Angle (Degrees) Bent Plate Box Girder Assembly (Sheet Plate Length = 10 ) 24" Deep Girder 36" Deep Girder 48" Deep Girder Figure 3.13. Capacity-to-demand ratio vs. inclination angle.

83 required further research, testing, and code development. Headed reinforcing bars, as indicated in NEXT D Beams Stan- dards by the Precast/Prestressed Concrete Institute Northeast, also could be used with closure pours. However, this type of connection requires further testing and the design require- ments for this type of connection need to be codified. The concept that uses welded plate connections between adjacent deck bulb tee units requires further research and testing for suitability and the required minimum capacity of the connection. Because this type of a connection is a shear- only connection and does not provide moment transfer between adjacent units, the ability of one beam to transfer its load to the adjacent beam is limited, thus reducing super- structure redundancy, especially when one beam is damaged or deteriorated. With this type of connection between the adjacent units, as the beam loaded under traffic would have limited ability to transfer the load to adjacent beam, AASHTO live-load distribution factors would not be applicable because the deck is not fully continuous in the bridge cross section. The welded plate connection also reduces the redundancy of the structure if a beam failure occurs and does not provide an efficient alternate load path to transfer the force to adjacent beams or an adjacent portion of the deck. Deck slabs on each beam would behave as cantilever beams and would need to be designed as such for the applicable design loadings. The weight of the traffic barrier would also not get distributed to multiple beams, but would rather be supported primarily by the fascia beam, which could require additional prestressing strands in the beam that supports the barrier. With increasing spans, as the deck tee beam depth increases, the steel channel diaphragms should be replaced with truss-type, steel cross frames. As the load distribution to beams would be signifi- cantly different than those used for conventionally con- structed bridges, design criteria to be included in the design codes for this type of construction need to be developed based on testing and further research. Because the deck slab does not transfer moment with the welded plate connection, this option reflects reduced struc- tural capacity under seismic loading compared to convention- ally constructed bridges that have a fully continuous deck slab in cross section. Additionally, the welded plate connection with grouting detail is not covered by AASHTO LRFD Speci- fications for seismic loading and therefore, seismic details and design criteria for the connectivity between adjacent deck tee elements need to be developed, tested under seismic loading, and codified. When the adjacent deck tee units are connected to provide full deck continuity in cross section, the design is more suit- able for bridges subject to moderate and heavy truck traffic, and the design covered by the current AASHTO LRFD Speci- fications is applicable. Deck continuity in cross section also offers improved load distribution under traffic loading and superimposed dead loads and redundancy. As the deck slab is continuous to transfer moment and shear and axial loads, the structure capacity under seismic loading would be compa- rable to that of a structure constructed using conventional techniques. With the increase in span length, as the beam depth increases, steel channel diaphragms should be replaced with truss-type, steel cross frames. Both of these concepts are also suitable for multi-span bridges made continuous for live loads. For deck tees, con- tinuity for live loads could be achieved by splicing the deck longitudinal rebars over the piers and extending the beam prestressing strands into the cast-in-place concrete dia- phragms by bending them up 90° into the diaphragms. Continuity diaphragms and the deck slab in the area of con- tinuity diaphragms are placed by using a closure pour. This detail would be similar to that provided in the NEXT D Beam Sample details developed by the Precast/Prestressed Concrete Institute Northeast. Live load continuity reduces the number of deck joints, which in turn reduces initial con- struction cost and eliminates the maintenance costs associ- ated with these joints. overlayS for moDular SuperStructure SyStemS The combination of high-performance concrete and high- quality construction will provide a long service life for these systems. Some owners may, however, have concerns about the long-term durability of bare decks. Use of an asphalt overlay with a membrane could be a desirable option in such situations to provide enhanced durability. In most cases, the overlay can be installed in a day prior to opening the bridge to traffic, or the overlay can be done during night lane clo- sures at a later point. If the bridge is constructed and opened during cold-weather months, the asphalt overlay can be installed when warm weather returns and the asphalt plants open. Asphalt overlay will also provide an improved ride quality. Use of an asphalt overlay may be required in bridge widening and for multiple simple spans, to even out the roadway profiles. European practice is to always use an asphalt overlay with a membrane as a protective system for bridge decks. Their experience indicates that keeping water away from bridge decks significantly improves service life. The preservation strategy for bridges with an asphalt overlay would be to replace the overlay on an as-needed basis. For bridges without an overlay, a new bonded concrete overlay or topping slab may be added to compensate for any deck deterioration. This project should investigate the substitution of steel rebars in the deck slab/top flange with fiber reinforced polymer (FRP) bars to achieve a longer deck life. The FRP bars may cost two or three times more than steel, but the overall cost impact would not be much. Several FRP reinforced bridges are in service and have performed well.

84 Connections for Modular Superstructure Systems The ease and speed of construction of a prefabricated bridge system is paramount to its acceptance as a viable system for rapid renewal. Additionally, as discussed, connections between the modular segments can affect the live-load distribution characteristics, seismic performance of the superstructure sys- tem, and also the superstructure redundancy. The designers need to develop a structure type and prefabrication approach that can be executed within the time constraints of the project site and also achieve the desired structural performance. Connections play a critical role in this approach. Connections of the modular units are important elements for accelerated bridge construction, as they determine how easily the elements can be assembled and connected together to form the bridge system. Often the time to develop a structural connection is a function of cure times for grouted connections, and the time it takes to make a welded or bolted steel connection. The connection details between the modular segments and their load transfer capabilities are critical with respect to sev- eral design issues: • The amount of wheel load transferred from a loaded beam to an adjacent unloaded beam must be considered to deter- mine appropriate design loads for the girders. The amount of load that can be transferred depends on the ability of the joints between the girders to transfer forces, as well as the sectional properties (e.g., torsional stiffness) of individual girders and the system as a whole. • The development of precast concrete components for bridges located in seismic areas is complicated by increased require- ments on structural continuity, increased ductility, and increased development length for the reinforcement. These requirements make the design of connections between the precast components more difficult than the connections used in low and moderate seismic regions. Full moment con- nections are preferred in seismic areas. Recommendations for joint details are provided in this section. • Questions can arise about the redundancy of these modu- lar superstructure systems when the connections between the units, such as a grouted joint, do not provide full moment transfer capabilities. This needs to be considered and accounted for in the design and detailing procedures. Full moment connections will provide the same level of redundancy as cast-in-place construction. Joints are prone to deterioration and are considered the weakest link in any structure, thereby reducing a structure’s effectiveness and long-term performance. The number of joints and the type of joint detail is crucial to both the speed of construction and to the overall durability and long-term maintenance of the final structure. The use of cast-in-place concrete closure joints should be kept to a minimum for accel- erated construction methods due to placement, finishing, and curing time. Durability of the joint should be achieved through proper design, detailing, joint material selection, and construction procedures. The following are design considerations for connections between deck segments: • Achieve durability at least equal to that of the deck. • Joint designs should consider truck traffic severity to achieve durability. • Joint details suitable for heavy/moderate/light truck traffic sites. • Achieve acceptable ride quality (similar to CIP decks). • Does not require overlays (overlay use is optional). • Does not require posttensioning. • Details can accommodate slight differential camber. • Can be opened to traffic in a matter of hours or days. • Preferably avoids the need for placement and removal of formwork, requiring access from below. Connection Details for Prefabricated Bridge Elements and Systems (Culmo, 2009) has introduced three classifications for connection details: • Level 1: This is the highest classification level. It is assigned to connections that have been used on multiple projects or that have become standard practice by at least one owner agency. Level 1 details are typically practical to build and will perform adequately. • Level 2: This classification is for details that have been used only once and were found to be practical to build and have performed adequately. • Level 3: This classification is for details that are either experimental or conceptual. Some Level 3 details have been researched in laboratories, but to the knowledge of the authors, have not been put into practical use on a bridge. Also included in the Level 3 classification are con- ceptual details that have not been studied in the laboratory, but are thought to be practical and useful. These standardized designs will use primarily Level 1 details. Level 2 details will be considered only where they are appropri- ate and can be justified through a critical evaluation. Joint typeS • Match cast and posttensioned joints are well established and are acceptable alternatives for ABC. Designers can find information on these from other sources. • Passively reinforced joints (full moment connections suit- able for heavy truck traffic sites). • Welded and grouted joints (shear-only connections suit- able for moderate to light truck traffic sites).

85 Two alternates may be considered for passively reinforced joints for modular construction at heavy truck traffic sites: • Full moment connection using ultra-high-performance concrete (UHPC) joints. • Full moment connection using high-performance con- crete (HPC) joints. In addition to these two full moment connections, a welded and grouted joint option is available for modular systems for sites with light truck traffic, such as local roads. These three options are evaluated from the standpoint of rapid renewal requirements and structural behavior and durability considerations. In precast construction, continuous connections exist when both moment and shear are transferred through the joint. Connections that just transfer shear work as hinges. For ductility and redundancy, AASHTO LRFD Article 1.3.3 notes that the requirements for ductility are satisfied for a concrete structure when the resistance of a connection is not less than 1.3 times the maximum force effect imposed on the connection by the inelastic action of adjacent components. For nonductile connections the ductility factor shall be at least 1.05. Systems with nonductile connections should be classified as nonredundant. For system redundancy and ductility, this project will recommend the use of continuous connections for ABC as the preferred approach. The hinge-type connection is an available option for low-traffic sites. heavy truck traffic SiteS: full moment connection uSing ultra-high-performance concrete (uhpc) JointS The term “ultra-high-performance concrete” (UHPC) refers to a class of advanced cementitious materials. When imple- mented in precast construction, these concretes tend to exhibit properties including compressive strength above 21.7 ksi, sustained tensile strength through internal fiber reinforce- ment, and exceptional durability as compared to conventional concretes (Rosignoli, 1998a). The specific UHPC investigated in this study is a product of a major worldwide construction materials manufacturer and supplier. It is currently the only product of this type that is widely available in the United States in the quantities necessary for large-scale infrastructure applications. European and Asian markets currently have multiple suppliers, and a similar situation will likely occur in the United States as the market for this type of advanced cementitious product develops. the compoSition During the summer of 2009, the New York State DOT com- pleted two bridge projects using the UHPC closure pour con- cept. The New York State DOT was interested in full-depth precast deck panels and deck bulb tee prestressed girders for use in constructing and reconstructing bridges. In both bridge types, the precast concrete elements needed to be connected together at the deck level via a permanent, durable connec- tion. This connection is heavily stressed both structurally and environmentally, meaning that the long-term performance of the bridge is dependent on acceptable performance of the connection. Conventional construction practices for con- nection details can result in reduced long-term connection performance as compared to the joined components. UHPC presents new opportunities for the design of modular compo- nent connections due to its exceptional durability, bonding performance, and strength. The properties of UHPC may make it possible to create small-width, full-depth closure pour connections between modular components. These connec- tions may be significantly reduced in size compared with con- ventional concrete construction practice, and could likely include greatly simplified reinforcement designs. Use of advanced cementitious composite materials such as UHPC in connection design presents new opportunities to advance the use of modular components with the following advantages: • Passively reinforced joint only 6 in. long. No posttension- ing needed. Figure 3.11 shows UHPC joint detail used by the New York State DOT. • Full moment connection suitable for heavy truck traffic sites, but can also be used under less-severe traffic situations. • The placement and curing of UHPC can be performed by using procedures similar to those already established for use with some HPCs. The fluid mix is virtually self-placing and requires no internal vibration. • UHPC can provide significant durability improvements to bridge decks due to the high strength, extremely low per- meability, high resistance to freeze thaw, and improved connection details inherent in the system. Research dem- onstrates that UHPC exhibited almost no permeability and was not susceptible to chloride ingress. • In the New York State DOT detail, the shorter development length of reinforcing bar in UHPC allowed a narrower joint, which reduced the total shrinkage. Tests done by the New York State DOT show that a 5-in. development length was sufficient for #6 rebars. This allowed a full moment connection to be made using a 6-in. closure pour and straight rebars. The agency successfully completed a proj- ect in 2008 using a UHPC joint. • In tests made at Michigan Tech Transportation Institute, the UHPC showed compression strength of 28,000 psi, compared with 4,000 psi for normal concrete. Tensile crack- ing strength was above 1,000 psi, compared with 400 psi for normal concrete. In testing for resistance to road salts and chlorides, UHPC withstood these chemicals at a rate 100 times greater than that of normal concrete.

86 • The compressive strength gain behavior of UHPC is an important characteristic of the concrete. UHPC does not have any compressive strength for nearly 1 day after casting. Then, once initial set occurs, UHPC rapidly gains strength over the course of the next few days until over 10 ksi of strength is achieved in about 3 days. No special curing is needed for the joint material (though steam curing is ben- eficial when applied). Regardless of the curing treatment applied, UHPC exhibits significantly enhanced properties compared with standard normal strength and HPCs. • Level any differential camber between adjacent beams before placing the joint. Slight differences in camber (<¼ in. can be tolerated). • Installation time of about 3 days includes erecting, placing, closure pours, and curing. • FHWA is conducting additional testing on UHPC joints. The results will provide improved guidance for design, as discussed in this chapter. • DUCTAL is the only supplier for UHPC, and the steel fibers are from a European supplier subject to Buy America provisions. teSting of uhpc connectionS In conjunction with the New York State DOT, researchers at the FHWA Turner-Fairbank Highway Research Center (TFHRC) are investigating whether the exceptional durability, high strengths, and superior bonding characteristics of UHPC lend themselves to the development of a new generation of connec- tion details applicable to modular bridge components. The TFHRC’s ongoing research program into the use of UHPC in highway bridges has recently begun focusing on deck-level connections between modular precast components. A physical testing program has been initiated (Transportation Pooled Fund Project TPF-5[217], titled Ultra-High-Performance Concrete Connections Between Precast Bridge Deck) in which subassemblages of full-scale precast bridge deck panels are connected via UHPC closure pours and then cycled under repeated truck wheel loadings. The test program has six speci- mens, with variables including joint orientation, slab thickness, reinforcement configuration, and reinforcement type. None of the specimens include any pre- or posttensioning. Four of the six test specimens will simulate 8-in.-thick precast deck panels, while the remaining two will simulate 6-in.-thick top flanges on deck bulb tee girders. Cyclic testing has been completed. All specimens performed well through the more than 2 million cycles of 2-to-16-kip loading and the more than 5 million cycles of 2-to-21.3-kip loading. No specimens showed any evidence of leakage along the joint interface. No joint inter- face debonding was observed. Also, the cracking behavior of the specimens demonstrated that individual structural ten- sile cracks in conventional concrete were interrupted and replaced by multiple tight-width cracks in UHPC. Static testing will begin next month. The final report is scheduled to be complete by the end of June 2010. The performance of the specimens tested to date has met all benchmarks. Test results to date, along with two New York State DOT bridges constructed in 2009, demonstrate the potential viability of using UHPC as a closure pour material. UHPC early age behavior and its compressive strength gain behavior are important material characteristics for ABC applications. Results indicated that UHPC does not have any compressive strength for nearly 1 day after casting. In tests, UHPC didn’t begin setting for approximately 22 hours. Once initial set occurs, UHPC rapidly gains strength over the course of the next few days until over 10 ksi of strength is achieved 2 days later. At that point, the rate of strength gain decreases, but the strength gain continues until over 18 ksi of compressive strength is achieved by 28 days. Cure time for UHPC connections will limit their suitability for weekend replacement projects. heavy truck traffic SiteS: full moment connection uSing high-performance concrete (hpc) JointS HPC denotes high-strength concrete that must have other characteristics specified to ensure durability, including per- meability, deicer scaling resistance, freeze-thaw resistance, and abrasion resistance. These characteristics are particularly suited for connections with the following advantages: • Alternate passively reinforced joint with HPC, no post- tensioning. • Full moment connection suitable for heavy truck traffic sites, but can be used under less-severe traffic situations. • The lapping of steel may be achieved with overlapping looped bars or short straight bars whose development is improved by the geometry of the joints or by external means such as confining spirals. • The greater widths (up to 3 ft) that are typical for these types of joints, relative to UHPC or welded/bolted joints, may increase the likelihood of shrinkage cracking and may require erecting and removing formwork from below. • The interface between the precast deck and the cast-in- place closure is of particular concern since cracks can develop due to shrinkage. A penetrating sealant should be applied to the top surface of grouted joints after curing to enhance durability. • Level any differential camber between adjacent beams before placing the joint. Slight differences in camber (<¼ in. can be tolerated). • Installation time is about 3 days including erecting, placing closure pours, and curing. • Researchers are investigating the use of a small closure pour with headed reinforcing bars.

87 moDerate-to-light truck traffic SiteS: Shear-only connection By WelDing anD grouting Connections that transfer shear only may be adequate for local roads. This detail has the advantage of reduced con- struction time, lower cost, and easy adaptability to all types of modular systems. Concerns about joint performance and durability have limited their use in ABC applications. • Use a welded tie connection combined with a grouted key. The 6-in.-long by ½-in.-thick steel ties are normally spaced 5 ft on center along the edge of the beam. They are welded to angles embedded in the beams and anchored with studs. • This connection is primarily designed as a shear-only con- nection. There is no intent to make this connection a deck moment connection. Each flange edge needs to be designed as a cantilever deck overhang. • The Texas DOT researched transverse welded connections for adjacent precast members and found that when com- bined with a grouted shear key, the connection is sound and durable. The Washington State and Idaho DOTs have also used a welded joint detail for precast members. • Any differential camber should be leveled before welding. Connection can be made even if there is slight camber dif- ferential between the beams. • Installation time is about 2 days, including erecting, weld- ing, and grouting. Multiple spans can be built in the same time frame with larger construction crews. • Use of fiber reinforced grouts can enhance joint perfor- mance. Some welded joints have not worked too well under certain applications. The Utah DOT has had some issues with leakage. More study of welded joints is recommended. The states use a welded tie connection combined with a grouted key. The ties are normally spaced 5 ft on center along the edge of the beam. This connection is primarily designed as a shear-only connection. There is no intent to make this connection a deck moment connection; therefore each flange edge needs to be designed as a cantilever deck overhang. Some designers have concerns with the long-term fatigue behavior of the welded tie connections. Therefore, at this time, it is recommended that these welded tie details be used for bridges with lower truck volumes. Several issues need to be considered when using welded plate connections. First, the bottom portion of the welded plate connector is not protected from the weather on the underside of the connection. If a bridge is to be constructed in a corrosive environment, the designers may want to con- sider the use of stainless steel plates and rods. Second, it is important that the grout placed in the keys between the beams completely fills the void. The most common keyway failure results from inadequate filling of the keys. If voids are present, the mechanical interlock of the key is lost. Posttensioning is a well-established and acceptable alterna- tive for ABC for which designers can find information from other sources. This toolkit focuses on innovative materials such as UHPC and advances their use for ABC connections. Use of high-performance lightweight concrete is a viable option to reduce the weight of prefabricated elements and systems. Segmental Superstructure Systems Segmental precasting for highway bridges is a well-established design and construction technology that offers many benefits, including: • Manufactured solution that is highly adaptable to many demands; • Speed of manufacture; • Quality control; • Speed of erection; • Small segments that are easily handled, shipped and erected; and • Repetitive erection technology. This concept includes a combination of both design and construction concepts for ABC. Both the design of greatly simplified segmental superstructures and the erection equip- ment for smaller spans is addressed. The adaptation of segmental technology to manufactured small-scale bridges is technically and economically feasible. It will require a paradigm shift on the part of owners and contractors. The net result will be a greatly simplified tech- nology from the perspective of both design and construc- tion. Much simpler segments, much lighter lifts, and greatly simplified equipment and erection technology will result in enhanced production and erection and lower costs. The adaptation of this technology will also result in high-quality, durable, and low-maintenance bridges with a longer life expectancy. Segmental technology has been thoroughly proven in the United States, and current standards cover most aspects of both the design and construction technology. AASHTO and American Segmental Bridge Institute (ASBI) have published standards on precast segmental construction. Several publications are also available that further discuss segmental construction in detail. References to these stan- dards and publications can be found in the References. channel SectionS anD SlaBS The engineering underlying the concept of precast segmental bridges with channel sections and solid or voided slabs is well developed. Channel sections are frequently used for channel bridges, railway bridges, and light rail transit (LRT) systems.

88 In-place voided slabs were commonly used during the early period of prestressed concrete bridge development within the span range of 80 to 115 ft and with span-to-depth ratios of up to 20 to 25. Adapting the precasting plant to different deck widths would also be simpler and less expensive. Extruding wet concrete along longitudinal reinforcement with a movable form is a typical construction method for voided slabs of buildings. Slab depths up to 2 ft with a total length of about 70 ft are typically achieved; the maximum length of the precast slabs in this case is dictated by transportation requirements. In the case of pre- cast segmental bridges, the extrusion lines may be lengthened to optimal values for the casting process. Rectangular or circu- lar voids would be easily achievable in the slab segments. If embedded forms with a few large shear keys are used to divide the segments during the extrusion process, after seg- ment separation and removal of the forms, the contact sur- faces would require only sandblasting and application of bond enhancer. The webs and slabs of the cross section may be particularly thin because of the industrial casting process, the reduced surface exposed to the atmosphere, and the smooth exterior surface that avoids gathering water. In addition to pleasant aesthetics, the span-to-depth ratio of voided slab bridges can be particularly low due to a better transverse distribution of live loads and the presence of a wide bottom slab. The bridge would be composed of two types of segments: the standard voided slab span segments and solid segments at the piers and abutments. The two types of segments may be cast in two separate extrusion lines or alternatively in the same line. External prestressing tendons may be deviated with steel saddles bolted to the internal surface of the cells; the end anchorages may be embedded into the abutment segments. The longitudinal prestressing tendons would possibly extend the full length of the bridge, anchored on the end faces. Transverse prestressing and the use of high-performance, lightweight concrete for the precast segments may further increase durability and diminish the weight of segments or increase their length if transported vertically. Transverse pre- tensioning and posttensioning are both possible, although pre- tensioning would increase forming costs. The use of unbonded monostrand tendons (also frequent in building construction) would add durability and further lighten the section. Dead anchorages could be alternatively used on the opposite sides of the bridge to diminish the cost of transverse posttensioning. The following aspects of the construction process require additional research and investigation. • Treatment of construction joints and reliability of stress transfer; • Details and technology of concrete stitches between pre- cast segments; • Optimization of cross-sectional design; • Optimization of longitudinal and transverse prestressing; • 3D solid modeling and analysis of stress dispersal; • Modular support girders (discussed later in this chapter); and • Incremental launching erection of precast segmental bridges. Precast Concrete Deck Panels As noted for modular systems, the intent of this project is to develop pre-engineered standards for ABC systems that offer the greatest potential for future advancements. In fact, there are already a number of standardized precast concrete deck panel systems that have been developed by various state DOTs. The issues involved with advancing precast concrete deck panels to greater acceptance in the industry lie not in devel- oping the details, but rather taking the details that already exist and have been used with some success and addressing the most critical deficiencies found to date. Design considerations for standardized precast concrete deck panel systems should include the following: • Pre-engineered standards for precast deck panels should address the most common bridge sites without consider- ing site-specific geometry, and so forth. • Advanced materials, including high-performance concrete and UHPC, when cost-effective. • Optimize designs for ABC and use of high-performance materials. • Designs should be established for a range of common bridges widths (36, 40, and 44 ft) and girder spacings (8 to 11 ft). • Designs should prioritize panels that span the entire roadway width and can thus be installed without a centerline joint. • Designs should simplify posttensioning (PT) details and should include only longitudinal PT. • Eliminate skewed and curved bridges from consideration for standard precast deck systems. • Lightweight concrete should be evaluated for use in post tensioned systems to ensure that creep behavior is acceptable. challengeS to BroaDer uSe of precaSt Deck panel SyStemS Although precast concrete deck panels have been used for more than 40 years, a number of challenges to the wider use of this technology remain. The most critical challenges are presented in the following paragraphs. BriDge Deck JointS. The transverse panel joint is a funda- mental part of virtually all precast bridge deck systems. How- ever, longitudinal panel joints are typically used only for

89 certain projects, such as those in which the bridge is so wide that the individual panels are simply too large to be econom- ically transported to the site and installed using moderately sized equipment, or those that are constructed in stages and require the shifting of traffic from one lane to another dur- ing replacement of an existing deck. In many heavily traveled bridge replacement scenarios, a structure is closed during overnight or weekend hours for bridge deck replacement and then reopened to traffic in the morning. Essentially, if a single panel can be used to cover the entire width of the bridge, it is strongly recommended that the longitudinal joint be avoided. A bridge deck is subjected to considerable exposure to deicing salts and thus any cracks that permit the intrusion of chloride-laden water will cause the bridge deck accelerated deterioration. When compared to a cast-in-place bridge deck, which is often placed continuously from end to end of a bridge, a precast concrete deck panel system by its very nature pro- vides a multitude of opportunities for water intrusion. There are several components of a successful precast con- crete deck panel system, including • A smooth riding surface and effective load transfer between panels; • Effective filler material for panel joints; and • A durable, reliable posttensioning system. It should be noted that a precast concrete bridge deck that provides a smooth wearing surface is desirable not simply as a comfortable ride for the traveling public. In fact, one of the fundamental causes of deterioration of a precast concrete bridge deck occurs if adjacent panels do not provide a com- pletely flush-fitting and effective load transfer system at a transverse joint. Repeated wheel loads passing across these joints, and amplified by the dynamic impact effect, will even- tually cause one or more reflective cracks to develop either immediately at the bond line between a panel and the inter- stitial grout or in the panel concrete itself. These reflective cracks permit the ingress of water, and the cyclic freeze-thaw effect will eventually widen these cracks. This self-propagating crack development pattern continues and continual maintenance of the deck will be required to slow the deterioration of concrete panels. Another component of successful joint performance is the application of the appropriate filler material. In most cases, a cementitious grout product is used; however, there has been limited use of neoprene materials to date. Ongoing research has been working toward the improvement of these joints by evalu- ating different shape joints and filler material. As yet, only lim- ited success has been achieved, and considerable work remains. A variety of systems that use posttensioning in the longitudi- nal direction have been developed and tested by bridge owners. These systems are designed to provide a crack-free, durable deck that remains in compression throughout its life cycle. poSttenSioneD connectionS BetWeen panelS. Posttension- ing is a well-established and acceptable alternative for ABC for which designers can find information from other sources. SuggeSteD aaShto lrfD coDe improvementS. In Phase III of the ongoing project, the research team will develop and propose AASHTO–formatted design specifications to assist owners and designers with wider implementation of acceler- ated construction. One specific issue that could significantly improve the cli- mate for ABC using precast deck panels is related to maximum stud spacing requirements. National bridge design standards, including both the AASHTO standard and AASHTO LRFD Specifications, pro- vide a requirement for maximum stud spacing of 24 in. for steel girders. These shear studs are placed in clusters to accommodate shear pockets in the precast panels. In case of concrete beams, most bridge owners prefer that the stirrups placed for vertical shear are extended into the bridge deck to provide composite action. As mentioned before, it is very cumbersome to clear concrete around the shear studs or shear reinforcement. It is recommended that the AASHTO provisions for shear connector spacing be evaluated for suit- ability with accelerated construction. At least in the case of prestressed concrete beams, the roughed surface on the tops of the beam flange transfers substantial horizontal shear though friction alone. It is evident from several studies that this shear friction theory can be used to satisfy the horizontal shear requirement, which would allow the precast deck panel and beam to contribute to the composite action. AASHTO currently requires that these stud clusters and their corresponding panel blockouts be provided at a maxi- mum 24-in. spacing. This requirement frequently leads to congestion and significant discontinuity of the mild reinforc- ing in the panels. An effort is under way through the PCI Bridge Committee, and supported by ongoing and future research into the 24-in. requirement, to modify this require- ment to allow shear pockets at a 48-in. maximum spacing. If this amendment is incorporated into AASHTO LRFD, the fabrication and placement of precast deck panels will be much simpler and thus less expensive. In addition, the future removal of precast deck panels will be considerably improved; approximately half of the shear studs to be worked around would be eliminated during the removal of each panel for bridge widening or replacement. AASHTO Chapter 9, which deals with precast concrete deck panels, should be carefully revisited. The research team envisions code language that is more definitive and less open to varying interpretations than the current code. The following

90 recommendations should be considered for inclusion in future interim specifications: • Guidelines for stress requirement on the panels for stage construction or full construction. It is anticipated that stress requirement for stage construction is higher than that for full construction. • Example detailed content to address both transverse and longitudinal joints. • Guidelines for the use of reinforced concrete or prestressed concrete for a variety of span configurations and beam/ girder spacings. • Reconsideration of resistance phi factors to account for the higher level of quality control available in the closely moni- tored fabrication environment when compared with cast-in- place concrete design. As noted previously, the recent NCHRP 12-65 study devel- oped an extensive list of proposed AASHTO LRFD Section 9 modifications. These recommendations were presented to the appropriate AASHTO committee. Currently, there are ongoing discussions among the AASHTO committee mem- bers as to whether to provide the proposed modifications in the AASHTO construction specifications or AASHTO LRFD Bridge Design Specifications. Members of the current R04 research team believe that the inclusion of the proposed modifications in AASHTO LRFD Specifications are valid and should be strongly considered for inclusion in the AASHTO code. A more detailed evaluation of these code provisions will be made during Phase III of the project. Constructability Evaluation Constructability evaluation is aimed at assessing issues spe- cific to transportation of components, erection methods, equipment needs, and the suitability of the system to rapid construction. ABC Designs need to be optimized to meet the transportation and erection requirements for prefabricated construction. Modular Superstructure Systems conStructaBility iSSueS anD conSiDerationS • Usually length ≤140 ft, weight ≤100 tons, width ≤8 ft for transportation and erection using conventional construc- tion equipment. • Design for sections that can be transported and erected in one piece, for lengths up to 140 ft, may be feasible in cer- tain cases. Provide one method of erection. (Spans longer than 140 ft may be erected by shipping the segments in pieces, splicing on site, and using a temporary launching truss for erection.) • Segments designed for transportation and erection stresses, including lifting inserts. Sweep of longer beams should not be an issue for erection because there is an opening between the beams. • Able to accommodate moderate skews. For rapid renewal, it would be more beneficial to eliminate skews altogether by making the bridge spans slightly longer and square. • Provide standard details for durable connections between deck segments that can also be rapidly constructed. • Segments that can be installed without the need for cross frames or diaphragms between adjacent segments. Improves speed of construction and reduces costs. Use of diaphragms is optional and based on owner preference. • Deck segments when connected in the field should provide acceptable ride quality without the need for an overlay. Deck segments to have ¼-in. concrete overfill that can be diamond ground in the field to obtain desired surface profile. • Control of camber for longer spans will be important for modular superstructures. Control fabrication of concrete sec- tions, time to erection and curing procedures so that camber differences between adjacent deck sections are minimized. Leveling procedure to be specified to equalize cambers in the field during erection. • Edge sections of deck with curb piece ready to allow bolt- ing of precast barriers. tranSportation anD erection iSSueS Transportation and erection (crane capacity) limitations could present challenges that need to be considered in the standardization process. Every state has requirements for shipping of oversize and overweight loads that can limit the permissible size of the elements. These limitations could influence the maximum span lengths of the new bridge sys- tems suitable for ground transportation. The width that can be transported without special permit is generally lim- ited to 8 ft. Optimizing the weight of these bridge systems through design or the use of new lighter and durable materi- als will allow the transportation and erection of larger and longer bridges. High-performance materials that are light- weight and durable are most suited for prefabrication in large sizes. camBer anD riDing Surface iSSueS One of the greatest construction difficulties is eliminating the differential camber between the girders. It is important to develop an adequate means of removing the differential cam- ber between the girders on site. Differential camber in pre- fabricated elements could lead to fit-up problems and riding surface issues. If the differential camber is excessive, the con- tractors in some states will apply dead load to the high beam to bring it within the connection tolerance.

91 To the traveling public, the smoothness of the riding surface is a significant riding comfort issue. This is also an important factor for durability and maintenance, as vibrations from an irregular surface can affect the structural steel components of a bridge. Due to irregularities in the riding surface that can occur at longitudinal and transverse joint locations between modular components, it may be necessary to plane the deck surface through diamond grinding. LRFD Article 2.5.2.4, Rideability, requires that [t]he deck of the bridge shall be designed to permit the smooth movement of traffic. . . . Construction tolerances, with regard to the profile of the finished deck, shall be indicated on the plans or in the specifications or special provisions. The number of deck joints shall be kept to a practical minimum. . . . Where concrete decks without an initial overlay are used, consideration should be given to providing an addi- tional thickness of 0.5 in. to permit correction of the deck pro- file by grinding, and to compensate for thickness loss due to abrasion. achieving riDe Quality With prefaBricateD SuperStructure SegmentS • While the application of an overlay helps overcome finite geometric tolerances, it also requires another significant critical path activity prior to opening a structure to traffic. • Today’s availability of low-permeability concretes and corrosion-resistant reinforcing steels allows owners to forgo the use of overlays on bridge decks. • With prefabricated superstructure construction, the chal- lenge is to develop methods that achieve the final ride sur- face without the use of overlays. Control of cambers during fabrication and equalizing cambers or leveling in the field is intended to achieve the required ride quality. • An attractive option is diamond grinding decks with sacrificial cover to obtain the desired surface profile. Such a method can be faster and more cost-effective. • For continuous or multiple simple spans, beam cambers may affect ride quality to a point where an asphalt overlay system may be recommended (see the discussion below). control of camBer During faBrication anD eQualizing camBerS in the fielD • Differential camber of beams can lead to dimensional problems with connections. • Schedule fabrication so that camber differences between adjacent deck sections are minimized. Measure camber on each deck section immediately after transfer of prestress forces. (The Washington State DOT requires that at trans- fer of prestress, the difference in camber between adjacent deck sections of the same design must not exceed ¼ in. per 10 ft of span length or a maximum difference of ¾ in., whichever is less.) • Equip all deck sections with leveling inserts for field adjust- ment or equalizing of differential camber. The inserts with threaded ferrules are cast in the deck, centered over the beam’s web. The Washington State DOT specifies a mini- mum tension capacity of 5,500 lb for the inserts. After all adjustments are complete and the deck sections are in their final position, fill all leveling insert holes with a nonshrink epoxy grout. • The welded joint details can accommodate minor differen- tial camber. If the differential camber is excessive, the con- tractors in some states will apply a dead load to the high beam to bring it within the connection tolerance. A level- ing beam also can be used to equalize camber. • Have available a leveling beam and suitable jacking assem- blies for attachment to the leveling inserts of adjacent beams. Adjust the deck sections to the tolerances required. More than one leveling beam may be necessary. • If the prescribed adjustment tolerance between deck sec- tions cannot be attained by use of the approved leveling system, shimming the bearings of the deck sections may be necessary. • See Figure 2.21 in the previous chapter for an image depict- ing a New York State DOT bridge leveling procedure for adjacent beams. aSSemBly planS It is common for designers to require the submission of erec- tion plans for conventional construction projects. This is nor- mally limited to the erection of beams and girders. Bridges built with prefabricated elements require special erection and assembly procedures due to the larger number of elements that need to be erected. The New Hampshire DOT required the contractor to submit an assembly plan for its first fully pre fabricated bridge project. The assembly plan is similar to an erection plan; however, it also includes information such as grouting and grout curing procedures, timing and sequence of construction, and temporary shoring of substructure elements during each phase of construction. It is recommended that projects built with prefabricated elements contain specifica- tions requiring the submission of a detailed assembly plan. Deck BulB teeS anD DouBle teeS With integral Deck The girders may be pretensioned or posttensioned. Post- tensioned girders are often used for long spans in which ship- ping limitations preclude the use of pretensioned girders. Posttensioned girders are often cast in two or more pieces that are connected in the field by splicing. Posttensioning can also be used to simplify girder shipping. Spliced girder technology can be used to create multi-span bridges. The girders are spliced with reinforced concrete closure pours. Accurate predictions of the deflections and camber are dif- ficult to determine since modulus of elasticity of concrete (EC),

92 varies with stress and the age of concrete. The effects of creep on deflections are difficult to estimate. An accuracy of 10% to 20% is often sufficient. Leveling of beams to offset camber differences can be carried out in the field, as discussed in this report. The durability of grout also should be tested, as the loss of grout would accelerate corrosion of the welded plate con- nections and could result in a potential loss of connectivity between the beams. The loss of grout and connectivity would also accelerate the deterioration of beams and the sub- structure units. The current AASHTO LRFD Specifications do not address the welded connectivity between the adjacent deck tee beam units and this type of construction. Construc- tion specifications for this type of construction, including tolerances, should be separately developed. This type of con- struction is suitable for bridges with no skews or small (10° to 15° AASHTO-permitted) skews. For larger skews, this type of construction may lead to fit-up issues with welded connec- tion plates due to differential deflections between beams. Leveling beams could be used to achieve the intended top of deck elevations and to improve the connectivity of the welded plates. The beam cambers must be strictly maintained in the fabrication shop to limit the differential between the theoreti- cal camber and actual camber to approximately ¼ in. This type of construction is suitable for bridges with constant cross slopes and on tangent alignments. It is not suitable for bridges on a curved alignment, on significant skews, or for bridges supporting a flared roadway. The grout and closure pour concrete properties should be carefully selected to avoid cracking and provide durability. It is estimated that the addition of concrete diaphragms for live-load continuity would add approximately 3 days to con- struction time. Given several weeks of construction duration using ABC, this additional time is acceptable. Precast deck tee sections with weights up to 100 tons and widths up to 8 ft could be transported using conventional equipment and erected using conventional cranes. With the use of precast sections, the superstructure construction time could be reduced to weeks or days depending on proj- ect needs. Lifting locations for deck tees should be located such that cracking during transportation is minimized. Criteria for lift- ing locations and details are not presently addressed in the AASHTO LRFD construction specifications and need to be developed. Segmental Superstructure Systems Bridge length, the span length, and the nature of the obstruc- tion to overpass significantly influence the construction method. Most bridges for ABC applications will be short or medium-length bridges. Erection methods commonly used for long viaducts require some adaptation and simplification. Simple erection girders and light crane picks are the pre- ferred technology for erection of short and medium-length precast segmental bridges with solid or voided slabs or with channel sections. If the access below for cranes is limited, or the traffic effects are significant, the segments can be erected from either end with light cranes and simply rolled into place on erection girders. This technology can be implemented by smaller local con- tractors on a few overpass locations, or even more economi- cally, on a larger number of ABC sites. Manufacture and erection of segments are processes that can be self-performed by small to intermediate-size contractors when the technol- ogy is simple and repetitive. Precast segmental highway bridges cover span lengths ranging from 120 ft to 160 ft at the lower limit and 390 ft to 460 ft at the upper limit. The longest precast balanced canti- levers constructed in Europe are now in the 600- to 700-ft range. Below 120 ft, the use of precast girders and in-place deck slabs is generally more economical, although the quality of in- place deck slabs is lower than that of precast segments and the construction duration is generally longer. Achieving longitudi- nal continuity is also complicated and time-consuming. For ABC applications, segment length and weight need to be manageable for available cranes, weight limits, and the typical height and width of handling and transportation requirements. Lengths of up to 12 ft are often transportable on public roads without excessive restrictions. Application of ABC techniques to modular short-or medium-span precast segmental bridges may pose new technological challenges in relation to the bridge length. When the bridge length permits amortization of the invest- ments and of the mobilization and demobilization costs of a launching gantry, the viaduct can be rapidly built with an under-slung or overhead gantry without new technological challenges. When the bridge is just a few spans, however, sim- pler erection means should be used to avoid the costs of an erection gantry and to simplify and accelerate mobilization and demobilization of the erection site. The simplest erection method for a precast segmental bridge is supporting the spans with ground falsework. In addition to the high labor cost and the long construction duration, however, the area under the bridge must be acces- sible for the entire bridge length, which is often incompatible with crossing highways, railroads, rivers, and environmentally sensitive sites. Support girders may be used to support the segments and diminish the impact of construction on the area under the bridge. The support girders would be simplified erection gan- tries, with some of the standard features of the latter removed

93 to diminish the investment and the mobilization and demo- bilization costs. A support girder for short bridges would have the following: • Simplified self-launching capability. The girder can be pulled along support rollers with a small winch anchored to the abutment. The truck used for transportation of the support girder may also be equipped with a special winch. • Overhead or under-slung configuration. Both configura- tions should be compatible with channel bridge sections and solid or voided slabs. In the under-slung configura- tion, the support girder may pose clearance problems when overpassing railroads or highways. This may require lifting the vertical profile of the bridge to avoid conflicts or erecting the span in a raised configuration and lowering it onto the bearings when the support girder is removed. • Modular composition to fit different span lengths. For the scale of projects targeted by this research for ABC implementation, it is expected that the size, complexity, and cost of erection equipment, scaled down as it is for the size of the project, will not present a large initial cost, or a disincen- tive to cost-effective and competitive solutions. In fact, it may represent a reduction in cost due to the small size and simplic- ity of the segments and corresponding equipment necessary for erection. As an alternative, a precast segmental bridge with channel sections or a voided slab bridge can be assembled behind an abutment and positioned onto the piers by incremental launch- ing. This construction method would offer many advantages, including the following: • Safety for traffic: no work adjacent to traffic, no erection equipment between the piers, no construction clearances for support girders, no drop of materials; • No detours of traffic when overpassing highways; • No speed limitation on vehicles and trains; • Safety for workers: bridge built entirely on the ground; • High quality and easy inspection: bridge erected behind the abutment; • Context-sensitive solution: minimal disturbance to envi- ronmentally sensitive sites; • Easy crossing of rivers: only interference is pier erection within tight work windows, no reduction of the hydraulic section, no consequences from floods; • Compatible with hard-to-access sites: rivers, channels, wetlands, highways, railroads, deep valleys, steep slopes, piers of any height; • Compatible with transverse shifting and ABC replacement of bridges in service; • Small erection yard with no additional right-of-way; • Low labor costs: labor used entirely on production, mini- mized access problems, minimized crew transportation, minimized use of ground cranes; • Industrialization of the erection process easily adaptable to bridge length; and • Inexpensive and adaptable erection equipment with rapid mobilization and demobilization. Incremental launching construction of prestressed con- crete bridges has seen hundreds of applications worldwide. Its application to short and medium-length precast segmen- tal bridges with channel sections or voided slabs would be an innovative evolution of a time-tested construction method. The risks of innovation would be mitigated by the positive 50-year history of the launch techniques and the absence of serious accidents. Ample research is available on construction of precast segmental bridges. The American Segmental Bridge Institute specifically addresses this type of construction. Books and manuals are also available. Joints In modern segmental construction, segments are typically match cast in a precasting plant and glued on site with epoxy joints. Dry joint technology is not applicable to ABC con- struction in which slabs and channels might be considered, as external posttensioning is not feasible. With epoxy joints, the segments to be erected are first guided into position to ensure that no damage occurs to the concrete. The segment is offered up to the previously erected segment on a dry run, with the joint prestressing bars already in place to ensure that everything fits together and matches. The segment is then moved back and the epoxy applied to the joint surface, after which the segment is pulled into position and the joint prestress fully installed. The temporary joint prestress normally consists of bars. These are quick to install and hold the segments in place until the permanent prestressing tendons are installed and tensioned. To keep the epoxy thickness uniform over the joint, the temporary prestress is applied with an average compressive stress of 30 to 45 psi. The joint prestressing bars are either internal or are arranged outside the concrete. Internal bars are usually left in place and grouted to become part of the permanent prestress for the deck. External bars are positioned above the two slabs and anchored on temporary steel or concrete blocks stressed down to the segment. Permanent anchor blisters within the box cell are often used to avoid holes in the slabs and to save labor during erection. The advantage of using external bars is that they are easy to destress and reuse. When the segments are erected, they join up to the adjacent segments with the same horizontal and vertical angle deviation

94 that existed when they were match cast against each other. This is facilitated by the presence of shear keys on the webs and alignment keys on the slabs, which guide the segment into position. Small inaccuracies in individual segments or misalignment across joints accumulate when long sections of deck are constructed, and construction and surveying tolerances may require corrections to the segment alignment during erection. In superstructures erected by balanced cantilever assembly, small misalignments are corrected within the cast-in-place joints at mid-span. With the span-by-span method and con- tinuous span structures, misalignments are corrected with short mortar or concrete stitches typically located at the joints of the pier diaphragms. Simple spans do not pose any problem for small geometric discrepancies. When it is necessary to adjust the alignment of segments during erection, fiber or polyethylene (PE) shims are inserted into the joints to increase the thickness of the epoxy joint along one edge. Although this technique achieves only a small adjustment to the deviation at every joint, the effect is magnified as subsequent segments are built on. Segmental Construction for ABC Using precast segmental construction technology with solid or voided slabs and channel sections for short and medium- length bridges for ABC applications should not pose major technical challenges. However, several aspects of standard precast segmental construction should be revised in relation to the specific requirements of these types of bridges and cross sections. The relatively short length of several bridges for ABC applications, in particular, suggests the use of spe- cialized joints between segments to facilitate amortization of the investment associated with precasting facilities. For ABC construction, the development of standard, sim- ple sections that can be mass produced in relatively small and inexpensive fabricating plants should be readily achievable. Construction of solid slab, or even voided slab, segments that are posttensioned is even less costly than the construction of long casting beds with anchor bulkheads and deviation anchors for typical prestressed girder fabrication. The use of external tendons avoids problems of durability from leaking joints. However, external tendons are hardly compatible with solid slabs and channel bridge sections because both types of cross sections are devoid of cells for cable containment. The costs of these precast segmental solutions would be gov- erned by the amortization of investments (i.e., by the entity of the investment and by the number of elements to be built in the precasting facility). In turn, the entity of the investment would be governed by the construction deadlines and the opti- mal level of industrialization of the casting process. Segmental precasting typically requires a precasting or fabrication facility, transportation, and erection equipment. The following sub- sections present the requirements in more detail. precaSting or faBrication facility This includes storage areas for loose materials, cage prefabri- cation templates, casting cells for short- or long-line match casting of segments, gantry cranes and straddle carriers for handling of segments, coverings of the working areas, and storage areas for segments. Production requirements and the time available for bridge construction dictate the number of casting cells and the dimensions of the storage area. Figure 3.14 shows a long-line match cast segmental setup for solid deck slabs. A very limited amount of special equipment and technology is required. The process involves the casting of every other segment, the removal of the bulkhead forms, and the infill match-casting of the intermediate segments. tranSportation The cost of transportation depends on the weight of segments. The number of transportation units depends on the bridge construction schedule and the distance between the precasting plant and the erection site. In the case of long distances, storing segments close to the erection site may diminish the number of transportation units but requires double handling of segments. Figure 3.15 shows the handling and transportation of a seg- mental solid slab. erection eQuipment Three different techniques are typically used for erecting a precast segmental bridge: the span-by-span assembly, the bal- anced cantilever assembly, and the progressive placement of segments with the help of temporary stays or props. An Figure 3.14. Long-line match casting of a solid slab.

95 example illustrating the erection of a segmental solid slab is shown in Figure 3.16. With the span-by-span method, all the segments for a span are positioned before the prestressing tendons are installed and the complete span is lowered onto the bearings. This method is used for both simply supported spans and con- tinuous superstructures. The adjacent spans of continuous bridges are joined together with concrete stitches to avoid propagation of the geometry tolerances of segmental pre- casting. After the stitch concrete has hardened, prestressing tendons are tensioned to make the deck continuous. With span-by-span erection and epoxy joints, a typical 130-ft span is usually erected every 2 or 3 days. With an under-slung gantry and dry joints, an erection rate of up to a span a day is achievable. Overhead or under-slung gantries support a complete span of segments during erection; after application of longitudinal prestressing, the gantry releases the span onto the bearings and launches itself forward to erect the next span. Span-by-span erection with launching gantries is typically used for long bridges and spans shorter than 160 ft; for longer spans it tends to be more expensive than balanced cantilever erection because the erection gantries become very heavy. The cost of launching gantries, including investment and mobilization and demobilization, requires long viaducts for amortization; when the area under the bridge is accessible, therefore, short precast segmental bridges are typically erected onto ground falsework. The balanced cantilever method involves erecting the seg- ments as a pair of cantilevers about each pier; the pairs of opposite segments are prestressed with tendons that cross the entire hammer. This method is primarily suited to long spans; long viaducts with shorter spans are better processed using the span-by-span method. Segments can be positioned with a launching gantry or a lifting frame supported by the deck itself. Standard cranes can sometimes work on the deck although this causes significant load unbalance on the piers. Ground cranes are used only when the area under the bridge is accessible and the piers are short. With the progressive placement, a lifting frame or ground crane raises and places the segments in one direction from the starting point, passing over the piers in the process. Pro- gressive placement is usually the most time-consuming erec- tion technique because of the single work location; however, the specialty equipment can be particularly inexpensive, especially when ground cranes can erect the segments along the entire length of the bridge. Only one bibliographic reference has been found on the incremental launching erection of precast segmental bridges. The reason may be that the high investments of segmental pre- casting require long viaducts for amortization while the incre- mental launching construction is typically addressed to shorter bridges. ABC applications of precast segmental construction would be addressed to numerous short or medium-length bridges rather than a long viaduct, so the financial break-even point would be different. A combination of precast segmental construction and incremental launching erection might merge the advantages of two construction methods that have amply demonstrated their capabilities in the respective typical fields of application. The precast segments are typically fabricated in a precasting facility by short-line match casting. With this technique, the new segment is moved to the end of the casting cell and the next segment is match cast against it. The geometric relation- ship between the two segments is achieved by rotating the front segment in 3D space before match casting the new segment. This permits the application of cambers and plan and vertical curvatures. The segments can also be fabricated close to the bridge; seg- ment transportation is thus minimized but on-site processing of loose materials is necessary. Long-line match casting is Figure 3.15. Segmental solid slab handling and transportation. Figure 3.16. Segmental solid slab erection.

96 often used in this case because of the less stringent geometry tolerances. With long-line match casting, two foundation beams support the entire segmental span during construction. On completion of construction, the span is dismantled and reassembled at the final erection site with any of the above construction methods. Precast segmental construction of box girder bridges is a well-established technique that does not require additional research within the purposes of this study. Precast Concrete Deck Panels Many of the challenges of constructing a bridge deck using precast concrete panels are not unlike those described for the other modular superstructure systems presented earlier in this section. However, the flexibility to use precast con- crete deck panels (PCDP) for both new construction and deck replacement projects offers some unique challenges as well. These issues will be briefly discussed in the following paragraphs. conStructaBility iSSueS anD conSiDerationS • For a typical bridge, the panel is usually ≤50 ft in length, ≤50 tons in weight, and ≤8 ft wide for transportation and erection with conventional construction equipment. • Panels typically span entire bridge width with transverse joints. • Panels will be transported and erected in one piece. A piece typically weighs less than other controlling items on the project. • Panels, including lifting inserts, must be designed to with- stand both transportation and erection stresses. • Panel flatness must be considered. Proper handling and storage of panels prior to arrival on site is critical. • Precast panels for skewed bridges are problematic. Rectangu- lar panels are typically much easier to form in a production environment. Small skews (less than 10°) may be accommo- dated in precasting. • Large skews may require an additional step in construction to cast a closure pour at the end of the bridge. For acceler- ated construction, it would be more beneficial to eliminate skews altogether by making the bridge spans slightly longer and square. • Provide standard details for durable connections between deck panel segments that can also be rapidly constructed. • Deck panels absolutely must provide a smooth riding sur- face. A joint that is not flush will be subjected to significant impact loads and will suffer premature deterioration. • Deck panels are often designed with an additional ¼-to- ½-in. cover that can be ground after installation to provide smooth ride. • Edge sections of deck with curb pieces must be fabricated for bolting or posttensioning of precast barriers. • Panels are typically cast flat to simplify casting and instal- lation in the field. • Shear connectors must be installed to coordinate with pockets cast into the deck panels. tranSportation iSSueS Limitations on the transportation, including crane capacity, of precast concrete deck panels could present challenges to the standardization process. Every state has established a maximum size and weight for pieces to be shipped without the need for a permit. Given their general shape, precast deck panels will often times be difficult to ship because they must lay flat. These limitations could influence the maximum bridge width that could be transported by ground. In most states, the width that can be transported without special per- mit is generally limited to 8 ft. High-performance materials that are lightweight and durable are most suited for prefabri- cation in large sizes. erection iSSueS: enD pour for SkeWeD BriDgeS Another challenge during the construction of precast con- crete deck panels is the end closure pour. This detail occurs in two specific situations: • At the end of skewed bridge; and • For bridges that incorporate an integral abutment. In normal (square) bridges, the panels can be laid end-to- end along the bridge and posttensioned parallel to the beams or girders with no additional complications because the panels are all rectangular with square corners. The precast operation can be very economical, as all panels use the same forms and essentially the same details. However, if the bridge contains one or more skewed ends, the end panel must either be specialty formed and precast (for small skews only) or be cast-in-place with an end closure pour. In the case of a special precast end panel, the cost of forming this piece can be significantly higher than for the normal production panel. One particular complication can arise when replacing an existing bridge deck with even minimal skew and attempting to salvage the existing abutment backwall. In this situation, the skewed end panel must fit precisely between the final normal panel and the backwall without creating a tapered expansion joint opening. The contractor must accurately measure the skew angle of the existing backwall prior to casting the skewed end panel. A deviation of even a small amount from the correct skew angle may make it impossible to place the final precast deck panel and working under an overnight closure is not the preferred time to learn that the final panel will not fit.

97 In the case of an integral abutment bridge with a longitu- dinally posttensioning deck, a cast-in-place closure pour is required to create a monolithic connection to the abutment, which provides for moment transfer. In either of these cases, the need for appropriate curing time of the closure pour can significantly affect the time savings expected in accelerated construction and somewhat retards the purpose of rapid construction. ErEction issuEs: crown from flat PanEls A review of the literature and surveys of bridge owners dem- onstrates that the vast majority of projects completed to date using precast deck panels have not had a crown built into the deck, but rather are constructed flat. For a situation in which a single panel can be used to construct the full width of the bridge, a moderate crown can be built in by thickening the slab along the centerline. However, this method of positive deck drainage is not very practical for wider bridges or where the cross slope for the bridge exceeds the normal 2% value. In addition, where panels are thickened at the center, a significant additional dead load is imposed on the supporting superstructure. In most cases, wider precast decks that cannot be con- structed as a single panel are designed as a two-panel system with a longitudinal joint supported on a center girder. In new construction, the center girder can be elevated to provide the proper support geometry. However, in the case of a redecking in which the existing girders are to remain in place, a possible solution is the inclusion of an exceptionally tall beamline haunch that is constructed of CIP concrete. A challenge for future development is the need to construct a durable longi- tudinal joint for this scenario that can be quickly constructed without the need for a CIP closure pour. PanEl flatnEss and riding surfacE issuEs As discussed in the section on constructability issues for modular superstructure systems it is important to develop an adequate means of removing the differential camber between the girders on site to ensure comfortable rideability for users. Where concrete decks without an initial overlay are used, consideration should be given to providing an additional thickness of 0.5 in. to permit correction of the deck profile by grinding, and to compensate for thickness loss due to abra- sion. AASHTO LRFD Article 2.5.2.4 requires that the deck of the bridge be designed to permit the smooth movement of traffic. This section is primarily concerned with the service- ability of the deck and not necessarily with the structural per- formance. Construction tolerances, with regard to the profile of the finished deck, should be indicated on the plans or in the specifications or special provisions. The number of deck joints should be kept to a practical mini- mum; however, the maximum panel width of 8 ft will likely govern this criterion. constructability issuEs and considErations To develop the largest possible load carrying capacity, precast concrete deck panels are made composite with supporting beams or girders using shear connectors. In the case of steel girders, this composite connection is achieved through shear studs welded to the top flange. With concrete beams, the com- posite connection is achieved by reinforcement protruding from the girders. In either case, the physical connection between panels is made through a variety of shear pockets, as shown in Figure 3.17. Typically, rectangular openings in the deck panels correspond with clusters of studs or reinforcing steel. These pockets are grouted after any deck posttensioning is completed to minimize any force transfer to the girders and subsequent loss of effective posttensioning force in the process. Installation and Removal of Concrete and Shear Connectors Two different concepts to streamline the installation of shear studs for use with precast concrete deck panels exist. Shear studs used in composite steel bridge construction are typically ¾-in. or 7⁄8-in. diameter. Research has been con- ducted within the past 10 years to development a much larger super stud with a 1¼-in.-diameter. This new stud offers approximately twice the strength and a much higher fatigue capacity than a conventional 7⁄8-in.-diameter stud. The use of these larger-diameter studs would require far fewer studs to be installed along the length of the girder with cast-in-place concrete and provide far more room within the shear pocket when used with precast deck panels. Not only would this system increase bridge construction speed and future deck replacement, but it would also reduce the potential for damage to the studs and girder top flange during future deck removal. The number of studs or stirrups in each pocket is usually based on AASHTO design requirements. The maximum shear stud spacing (or distance between shear stud block- outs) permitted by AASHTO LRFD is 24 in. There is some opinion in the industry that this limit is an arbitrarily safe “rule of thumb” limit imposed by AASHTO to assure com- plete composite action and avoid fatigue conditions. In most Figure 3.17. Shear pocket connection for steel girder bridge.

98 circumstances, this spacing is based on the fatigue capacity of the studs, and not the ultimate capacity. With precast panels it is beneficial, however, to place the shear connector block- outs at the largest spacing possible. This allows for fewer blockouts in the panels, which in turn increases panel strength for shipping and decreases manufacturing time and cost. Research and tests using both static and cyclic loads have been used to support the claim that 48-in. shear pocket spac- ing is adequate. An effort is under way through the Precast/ Prestressed Concrete Institute (PCI) to gather support for a revision to the LRFD code that would permit the use of 48-in. stud spacing under certain conditions. The replacement of an existing composite concrete deck, whether consisting of cast-in-place concrete or not, is compli- cated by the need to work around existing shear connectors. Past studies have shown that one of the most time-consuming parts of rapid deck replacement is the clearing of concrete around the existing studs. The bridge construction specifica- tions used by bridge owners state that it is necessary to protect the shear connectors during deck removal and replacement. This is especially true for prestressed beam bridges in which a particularly desirable method for replacing any shear reinforc- ing that is damaged during deck removal does not exist. In addition, the potential for extensive damage to the beam con- crete must be carefully monitored by contractor and inspector alike. Figure 3.18 shows an example of exposed shear studs following concrete removal. In the case of a steel girder superstructure, it is frequently acceptable to simply cut the existing shear connectors off within a ½ in. of the top flange to accommodate precast panel installation. Shear studs are installed in clusters to match the location of stud pockets in the precast panels to be installed. The recent development of ever-larger bulb tee beams has allowed for the efficient use of precast concrete spans well in excess of those available with AASHTO sections of years past. However, with these advantages come significant complica- tions. With their wider and thinner top flanges, an effective and time-sensitive deck removal procedure that can be performed without damaging the underlying beams will present an even greater challenge for future generations of bridge engineers. Uniform Bearing of Panels on Beams and Girders Fabrication variations and differential camber among bridge beams and girders can cause the bearing of the full-depth deck panels on the girders to be uneven. Full-depth deck panels should be leveled and bear evenly on the girders to ensure opti- mum performance and long life. If the panels are not leveled, extensive spalling of the transverse joints and a poor riding sur- face may result, while placing the panels directly on the girders leaves voids that can cause the panel and joint to crack and spall. To alleviate this problem, the void between the girders and panels must be filled completely with grout to provide a solid, uniform bearing surface. Leveling bolts or shims are commonly used to level the panels and temporarily support them above the girders. Bridge girders with a large amount of camber, such as long-span prestressed concrete beams, require that the full- depth panels be leveled to produce the proper bridge profile, as illustrated in Figure 3.19. There are two common ways of achieving this level condition: • Leveling the full-depth panels on the girders so that the pan- els align with the bridge profile. This allows for a thin wear- ing course to be used, which promotes rapid construction. Figure 3.18. Clustered shear studs following concrete removal. Figure 3.19. Two methods for leveling precast deck panels.

99 When this method is used, the clear distance between the bottom of the panels and the top of the girders can become very large near the ends of the bridge, making it difficult to level the panels and form the haunch region. • Leveling the full-depth panels with the girder profile a short depth above the girder. This minimizes the distance between the bottom of the panels and the top of the gird- ers, making the leveling easier, but the difference between the girder profile and bridge profile must be made with a varying depth wearing course, which requires a longer cure and adds significant additional dead load to the bridge. Posttensioned Connections Posttensioning is a well-established and acceptable alterna- tive for ABC for which designers can find information from other sources. Wearing Surface A bridge deck constructed from full-depth concrete deck pan- els often exhibits an extremely rough surface because of the grouted joints and shear pockets. A typical full-depth panel deck is shown in Figure 3.20. This type of wearing surface is not acceptable on many bridges, especially those with large volumes of high-speed traffic, so a high-performance concrete wearing surface is often added to provide additional safety, rider satisfaction, and improved durability. A wearing surface offers another advantage in that it eliminates potential impact effects on any deck joints that are not absolutely flush. Placing a wearing surface will have a significant effect on the time required to replace a bridge deck. The alternative to a placing a wearing surface would be to diamond grind the surface immediately prior to opening the bridge to traffic. Splicing of Tendons During Overnight Closure Operations For heavily traveled bridges or for bridges without a reason- able detour that would permit full closure of the bridge, the replacement of an existing deck with precast concrete deck panels is normally performed during overnight and weekend bridge closures. In these cases, the existing deck is removed and replaced to the extent permitted by closure duration each night. There are examples of many successful projects that have been completed this way, including the SR-433 Lewis and Clark Bridge and the Route 64 bridge over Lake Pomme du Terre in Missouri. For precast concrete deck panels that incorporate longitu- dinal posttensioning, the need to splice onto in place tendons and continue adding panels each night complicates the prob- lem. The posttensioning has to end at a specific location, which depends on the number of panels that can be com- pleted within the closure duration. Thus, the posttensioning bulkhead is established at a particular panel location corre- sponding to the end of an overnight of work. If the post- tensioning needs to be continuous for subsequent panels, the panel and stressing design will need to incorporate splicing and resumption of stressing in order to maintain a uniform posttensioning force in all panels. In addition, the contractor may need to make provisions for an unplanned panel instal- lation endpoint in the event of a mechanical breakdown or unexpected weather delay. Potential Time Savings During Construction The time required to construct a typical cast-in-place concrete deck is considerable and normally lies on the critical path to opening the bridge to traffic. In addition to the actual time to construct the deck and place the concrete, most owners require a 7-day wet curing period for all bridge decks. Potential time-related considerations for precast deck sys- tems include the following: • Panels are precast and delivery is scheduled when needed on site. • Panels are erected directly from the delivery truck. • Panel systems may require grinding or overlay in order to achieve a smooth ride and to eliminate the potential for salt intrusion. • Construction staging (and replacing half of the deck at a time) can be achieved with a longitudinal joint. • Precast barrier sections have been developed. They can be installed using bolts or posttensioning. • CIP concrete placement may be delayed by hot weather or rain. • Removal of existing concrete decks is very time-consuming and requires working around existing shear connectors without damaging the supporting members, which is challenging. To illustrate part of the constructability evaluation, the team compared the time required to construct a bridge deck Figure 3.20. Rough surface texture of full-depth precast panel deck.

100 for a typical DOT grade-separation bridge deck, as summa- rized in Table 3.3. The contractor is assumed to have consid- erable experience with DOT bridges, but not necessarily with posttensioned deck panels. The following basic data were assumed: • Bridge consists of four spans: 70 ft, 100 ft, 100 ft, and 70 ft, for total length of 340 ft. • Bridge width = 44 ft (two 12-ft lanes plus two 8-ft shoul- ders and barrier rail). • Bridge superstructure consists of steel beams with shear studs installed in the field. • Shear studs are assumed to require the same amount of time for each alternative. Although the CIP deck alterna- tive may have a greater number of studs to be installed, it is assumed that the greater precision required to match shear pockets in the panels is roughly equivalent. • CIP deck alternative consists of single 8-in.-thick high- performance concrete deck with epoxy-coated reinforcing steel. Barrier rail will consist of slip-formed, jersey-style barrier rail. During the 7-day curing period, a number of operations can take place that do not require the place- ment of heavy loads on the bridge deck (screed rail removal, curb form stripping). Given the length and width of the bridge, the contractor will be allowed to place the deck concrete in a single, continuous operation. • Precast deck alternative consists of 8-in.-thick precast pan- els, with longitudinal posttensioning. Transverse joints are spaced at 8 ft centers. In order to increase durability, the precast deck panels will be overlaid with a 2-in.-thick high- density concrete overlay. The barrier rail will consist of precast barrier units installed after the deck panels are installed, but prior to the placement of the overlay. • Bridge deck will be constructed in a single operation. No staging of traffic is required and a detour is available. • Abutment is non-integral, so both bridge deck alternatives will have a strip seal joint. This simplified example illustrates the potential time sav- ings for a routine bridge project. A total time savings of 23 days, or nearly 40%, is available through the use of a precast concrete deck panel system. Even given some fairly conservative assump- tions for construction durations, and notwithstanding that a contractor can proceed much faster simply by applying more labor for those items that are not simply a waiting period, the potential time savings are significant. There are a number of variations for bridge deck replace- ments that cannot be performed as simply as closing the bridge to traffic and pouring a CIP deck in one operation. Examples include the following: • Staged replacement of a two-lane bridge deck where no reasonable detour exists. The existing bridge is removed in a series of transverse strips. In this case, as many deck sec- tions of the existing bridge are removed as can be replaced each night. • Staged replacement of a four-lane deck where traffic can be shifted to other lanes during overnight operations. With a Table 3.3. Time Durations for CIP Decks and Precast Decks Cast-in-Place Deck Precast Concrete Deck System Task Duration (Days) Task Duration (Days) Install walers and plywood deck forms between girders 12 Install neoprene haunch strips for each beam line 2 Install overhang jacks, brackets, forms, walkway, and handrail 8 Erect precast deck panels (8 per day) 7 Install curb forms and screed rail 4 Assemble ducts and place closure pour concrete 4 Tie main reinforcing steel, including vertical barrier rail reinforcing 7 Thread posttensioning tendons 2 Assemble and adjust screed rails 2 Perform posttensioning 2 Assemble and test deck finish machine 1 Grout posttensioning duct 1 Place deck concrete and curing system 1 Grout shear pockets and girder haunch 2 Wet curing period 7 Install precast concrete barrier rail sections 5 Strip deck forming 7 Place high-density concrete overlay 3 Underdeck patching 2 Wet curing period 3 Slipform barrier 3 Install strip seal expansion joint gland 3 Install strip seal expansion joint 3 Total Time 57 days Total Time 34 days

101 longitudinal joint along the centerline, it may be possible to remove half the deck, replace it with precast panels, switch traffic to the new deck, replace the other half the deck with precast panels posttensioned longitudinally, and finally connect the halves with transverse posttensioning. • Long viaduct deck replacement that cannot be completed in a single, continuous concrete placement. In these types of situations, the contractor is required to place deck con- crete in a particular sequence to avoid damaging previously placed sections. Typically, each positive moment region in the bridge is placed in a separate operation (with 7 days of cure time at each step) and the same process is repeated for the negative moment regions. • Horizontally curved bridges would require precast panels that are tapered in order to accommodate the varying radii from the interior side of the curve to the exterior side of the curve. • Bridges with a superelevation transition. Any time of pre- cast operation is best suited for repetitive operations where the mass production of the precast panels can be leveraged to obtain the best possible bid price from the precaster. Implementation Challenges Most of the systems described in this report, though proven, are only occasionally used for bridge replacement projects. Even relatively simple-to-deploy solutions such as precast deck panels, prefabricated modular bridges, or movement solutions such as sliding, rolling, launching, movement using SPMTs, and so forth, are used for a very small fraction of ongoing bridge work. This section addresses issues pertaining to implementa- tion and what should be done in the development stages to overcome the impediments to ABC implementation. Modular Superstructure Systems Precast concrete girders have seen widespread acceptance as an economical construction alternative over the last 50 years. In recent times, deck girder systems and prefabricated com- posite stringer systems have gained some traction in rapid construction situations. More needs to be done to achieve greater penetration of prefabricated systems and components to minimize on-site construction and change the current cast-in-place construction culture. While most agencies are aware of ABC technologies, very few practice it on a large scale. According to the surveys in Phase I, many ABC tech- niques are ready for implementation, yet DOTs are hesitant about using ABC techniques because of certain concerns, including higher initial costs and longevity of connections. The objective of the R04 project is to develop standardized approaches to designing, constructing, and reusing (includ- ing future widening) complete bridge systems that address rapid renewal needs and efficiently integrate modern con- struction equipment. Pre-engineered standards for proven systems will make available a toolbox of ABC systems to designers who may be new to this method of construction to help overcome the initial resistance to trying a new approach that they may perceive as being risky or complex. Advancing the state of the art to overcome obstacles to ABC implementation and achieve more widespread use of ABC is a goal of this research. Key findings from the Phase I outreach efforts of owner and contractor concerns and impediments to ABC implementation pertinent to superstructures are as follows: 1. There is a cast-in-place (CIP) construction culture among contractors. Contractors like to keep as much work for themselves as possible to keep crews employed and max- imize profits. Precast options may require work to be subcontracted out and reduces the control of the prime contractor. 2. The largest impediment to increased use of ABC appears to be the higher initial costs. Reducing cost was a priority with most owners. 3. ABC is perceived as raising the level of risk associated with a project. It is also perceived by some contractors as being too complex. Proven superstructure and sub- structure systems that reduce overall risks would be quite attractive to owners and contractors. 4. There are concerns about the durability of joints and connections in precast elements. 5. There are concerns about seismic performance of precast elements and connections in seismic regions. 6. Lack of familiarity with ABC methods is a concern. States are looking for design manuals and other aids that could help them to design and implement ABC. Training could be beneficial. 7. Standardizing components is good but also offers chal- lenges in getting the industry and the states to come together in a regional approach to ABC. Developing ABC standards that could be adopted regionally is one goal. 8. There is a need for design considerations for structures to be moved, for acceptable deformation limits during movement, and for better specifications. 9. ABC designs should be adaptable to a number of place- ment options to be cost competitive. A majority of con- tractors are not receptive to owners requiring that a specific method of construction be used in ABC contracts. 10. Lack of access for equipment and the need for large stag- ing areas unavailable in urban locations are hindrances to large-scale prefabrication. Use of smaller elements that can be assembled on site for superstructures and sub- structures will overcome mobility issues. The modular concept of building bridges could overcome this concern.

102 11. Contractors would be more willing to make equipment purchases if bridge construction became more standard- ized or industrialized, and was based on certain methods of erection to speed the assembly. Standardization increases the prospects for repeated use of the same equipment. Any obstacles to implementing these modular super- structure systems will depend on how effective these systems will be in addressing these owner and contractor concerns about ABC. Item 1 There is a cast-in-place (CIP) construction culture among con- tractors. Contractors like to keep as much work for themselves as possible to keep crews employed and maximize profits. Precast options may require work to be subcontracted out and reduces the control of the prime contractor. One of the main reasons for the lack of support for ABC among contractors is the reluctance to subcontract out much of the work to precasters and other specialty subcontractors for transportation and erection, which is seen as reducing the gen- eral contractors’ profit and control of project operations. Con- tractors want to keep their own crews busy. With precast decks, there have been problems with unions; the ironworkers’ union has complained because there was no deck to reinforce. This is a valid concern that will not go away with innovative designs or standardization of ABC concepts. Ways to work with the con- tracting industry need to be found to make this transition to ABC happen. The solution is to introduce the CIP industry to precast technology and demonstrate its profitability. Self-performing of precaSting By general contractor This is an approach to ABC in which the designs allow maxi- mum opportunities for the general contractor to do its own precasting at a staging area adjacent to the project site or in the contractor’s yard with its own crews. The more the con- tractor performs using its own crews, the more profits it will realize. There is no reason to go to a precaster, unless the pre- cast is posttensioned or prestressed, which would need to be fabricated in a certified plant. In this regard, the prestressed deck girders would need to be fabricated by a precaster. How- ever, contractors have always obtained their precast pre- stressed beams from precasters. Using ABC would not signal a significant change except for the integral deck that is cast with the beam. With regard to the decked steel girder systems, contractors can perform the fabrication themselves because the deck is made of conventionally reinforced concrete. The casting of the deck can be done off-line by contractor crews at a staging area and then transported for erection. The casting of the deck can be done under fully shored conditions in which the beams are ground supported, which is advantageous for ease of construction, worker safety, and enhanced struc- tural resistance of the system, since it avoids buildup of non-composite stresses. The weight of modular systems should be kept under 100 tons to allow for erection by con- ventional cranes. Lightweight concrete mixes can be used to lighten sections. Self-performing is perhaps even more signifi- cant for substructure components and is discussed further in the substructure evaluation report. Items 2 and 3 The largest impediment to increased use of ABC appears to be the higher initial costs. Reducing cost was a priority with most owners. ABC is perceived as raising the level of risk associated with a proj- ect. It is also perceived by some contractors as being too complex. Proven superstructure and substructure systems that reduce over- all risks would be quite attractive to owners and contractors. Standardizing modular superstructure systems for ABC is aimed at increasing their availability through local or regional fabricators. Doing so will greatly increase their availability to owners and contractors and reduce lead times, which should result in more widespread use and thus reduced costs. Stan- dardizing also makes these systems more familiar to engi- neers, owners, and contractors, thereby reducing complexity and the level of risk associated with the project. Contractors will be more willing to offer competitive bids on a system they have experience with and that they perceive to be proven and easily constructible. Repeated use of a standardized ABC design also allows the owner or engineer to iron out kinks in the system through continuous improvement, leading more to a better design than to a onetime, customized solution. Repeated use of these systems will also encourage contractors to provide suggestions about how the constructability could be further improved, which will lead to further reductions in overall risks and cost. The modular superstructure systems will be developed to achieve cost and risk reductions through the adoption of the guiding philosophy for all ABC concepts advanced in this project. This philosophy is stated as follows: As light as possible 44 Simplify transportation and erection of bridge components. As simple as possible 44 Fewer girders, splices, or bracings. As simple to erect as possible 44 Fewer workers on site; 44 Fewer fresh concrete operations; 44 No falsework structures required; and 44 Simpler geometry.

103 Items 4 and 5 There are concerns about the durability of joints and connec- tions in precast elements. There are concerns about seismic performance of precast ele- ments and connections in seismic regions. The quality and durability of joints and connections between prefabricated components has been a significant concern that has impeded the greater use of ABC. Western states have also had concerns about shear-only grouted joints and their performance during a seismic event. With this in mind, modular superstructure systems have placed maxi- mum emphasis on developing durable connection details between prefabricated elements. Full moment connection using ultra-high-performance concrete (UHPC) or high- performance concrete (HPC) is being recommended as the preferred connection for details that are strong, durable, and seismically sound. Standardizing these connections with proven, easy-to-construct details will go a long way in over- coming the past concerns with performance of joints and connections. Items 6 and 7 Lack of familiarity with ABC methods is a concern. States are looking for design manuals and other aids that could help them to design and implement ABC. Training could be beneficial. Standardizing components is good but also offers challenges in getting the industry and the states to come together in a regional approach to ABC. Developing ABC standards that could be adopted regionally is one goal. Standardizing ABC systems will bring about greater famil- iarity with ABC technologies and concepts and will foster greater regional cooperation, which will help achieve region- specific customization that accommodates regional practices and industry needs. Pre-engineered standards to be devel- oped in this project will emulate cast-in-place construction but will be optimized for modular construction and ABC. These standards can be inserted into project plans with min- imal additional design effort to adapt to project needs. Using these standardized designs will serve as a training tool to increase familiarity about ABC among engineers. Formal training courses should also be developed to provide back- ground information on ABC and the application of design specifications. Items 8 and 9 There is a need for design considerations for structures to be moved, for acceptable deformation limits during movement, and for better specifications. ABC Designs should be adaptable to a number of placement options to be cost competitive. A majority of contractors are not receptive to owners requiring that a specific method of construc- tion be used in ABC contracts. Modular systems can be erected using conventional equip- ment for most span ranges. Longer spans may be erected using specialized erection methods or movement technologies. The intent of ABC construction technologies being developed in this report is to develop standard concepts for erecting high- way structures by using adaptations of proven long-span tech- nology that can also be easily adapted from project to project. This project will develop conceptual design of equipment suit- able for ABC use. Another important class of ABC projects includes those that require the movement of large components or bridges completed using various movement techniques. These movement techniques include self-propelled modular transporters (SPMTs); bridge sliding, skidding, and rolling using various sliding surface movement methods; and incre- mental launching. These ABC construction concepts will pro- vide the contractor a range of construction options, from conventional erection to specialized erection and movement techniques, to make ABC projects cost competitive. Design considerations for structures or components to be moved and acceptable deformation limits during move- ment are topics that will need further clarification during the development of the design standards. Additional speci- fication language may also need to be developed to guide practitioners. Items 10 and 11 Lack of access for equipment and the need for large staging areas unavailable in urban locations are hindrances to large-scale prefabrication. Use of smaller elements that can be assembled on site for superstructures and substructures will overcome mobility issues. The modular concept of building bridges could overcome this concern. Contractors would be more willing to make equipment pur- chases if bridge construction became more standardized or industrialized, and was based on certain methods of erection to speed the assembly. Standardization increases the prospects for repeated use of the same equipment. Moving complete bridges by using wheeled carriers requires large staging areas, which may be in short supply in congested urban areas. Modular systems allow the superstructure to be built in place with smaller components, thus overcoming the mobility issue. In short, modular systems allow a more versatile option to ABC not limited by space availability at the bridge site. Standardized designs will allow for the repeated use of modular superstructure systems, which will make contrac- tors more willing to invest in equipment on the basis of

104 certain methods of erection to speed assembly. Repetitive use will allow contractors to amortize equipment costs over sev- eral projects, which is an important component to bring overall costs in line with conventional construction. Segmental Superstructure Systems There is an unfortunate perception in many parts of the country that voided slabs have been plagued by performance and durability problems. This is not the case, considering that the United States has not used voided slabs as they are defined elsewhere in the world. True voided slab systems are regularly and successfully designed and constructed as a least initial cost solution in Canada and Europe, where they are known for superlative strength, redundancy, and durability in highly aggressive environments. Voided slabs, as defined in U.S. practice, generally mean box girders that have been used for short and intermediate spans. Almost always, they have been implemented (unfortunately) with inadequate detailing, poor workmanship, and a lack of proper posttensioning. All of these factors have combined to create a negative perception of this technology in many quarters. Short and medium-length precast segmental bridges for ABC applications do not pose particular implementation challenges. Segmental precasting is a well-known construc- tion method, the use of concrete stitches and long-line match casting have been amply tested, and the engineering under- lying channel or ribbed sections and solid or voided slabs is also proven technology. Short and medium-length precast segmental bridges may be erected with support girders or by incremental launching, as shown in Figure 3.21. The support girders would be sim- plified versions of self-launching gantries, which are amply tested machines. Incremental launching is also a time-tested construction method, applied for 50 years all over the world. The first launched bridge was composed of precast segments joined with concrete stitches. Multi-cellular box girders and ribbed slabs have also been launched. In both cases, therefore, the technical obstacle is not the risk of innovation. Simplifying existing machines and learn- ing how to use a construction method described in tens of publications and several domestic and international codes involve minimal risk. The technical obstacle to the adoption of these technologies is the poor knowledge of the developments achieved by the international bridge industry and the inertia of the U.S. bridge industry in exploring construction techniques that are amply consolidated elsewhere. Inertia can be won with incentives for contractors to erect short and medium-length segmental bridges with support girders and incremental launching. These incentives include the following: • Merging several bridges into one contract to facilitate amortization of the initial investment for designing and implementing a modular system of segmental bridges. • Relaxing the design requirements where possible, includ- ing simple geometry, constant radii of plan and vertical curvature, and constant width. Move all transitions out of the bridges. • Educating DOTs, contractors, and designers with itinerant courses given by specialists in bridge erection machines and incremental launching of bridges. The courses should provide continuous education credits. • Financing experimental projects to spread information on costs, quality, and technology. Testing Needs UHPC Joints for Modular Superstructures Continue to monitor the Turner–Fairbank Highway Research Center’s (TFHRC’s) ongoing research program into the use of UHPC in highway bridges, specifically the research into deck-level connections between modular precast compo- nents. In conjunction with the New York State DOT, research- ers at the TFHRC are investigating the use of UHPC for a new generation of connection details applicable to modular bridge components [Transportation Pooled Fund Project TPF-5(217), UHPC Connections Between Precast Bridge Deck]. The findings from this research will be pertinent to the UHPC connection details to be developed for this project. Precast Decks To address the challenges described in the preceding sections on engineering and constructability evaluations, a number of research and testing needs have been identified. To promote the wider use of full-depth precast concrete panels, these challenges will need to be addressed to produce a durable and easily constructible deck for a wide variety of situations.Figure 3.21. Segmental channel beam erection.

105 A number of these important topics are currently under investigation by researchers around the country. These needs can be addressed through a combination of theoretical analy- sis, numerical analysis, or laboratory testing. Identified research needs include the following: • Evaluation of transverse and longitudinal joints; • Guidelines for a closure pour for skewed and integral abut- ment bridges; • Practical, durable solutions for providing crown in precast decks; • Shear stud or shear steel configuration to make girders composite with decks; • Guidelines for intermediate closure pour in the case of stage construction; and • Specific modification to AASHTO LRFD Chapter 9. Evaluation of Transverse and Longitudinal Joints Additional research is needed to evaluate joint details. This research should focus on eliminating posttensioning to the greatest extent possible through the use of advanced materials such as ultra-high-performance concrete. The New York State DOT and FHWA are currently involved in research into UHPC joints for modular superstructures and precast deck panels. There is considerable evidence that the incorporation of longi- tudinal posttensioning in precast deck systems can result in a crack-free and leakproof deck. As noted previously, many bridge owners and contractors are inexperienced with post- tensioning. As with any unknown technology, contractors will increase their bid prices to help them mitigate uncertainties. The research focus should be to fully understand the com- plex forces initiated in a joint by the applied dynamic wheel loads. This research can take two forms: either finite element analysis of a joint or by testing small specimens in a laboratory. Two primary forces exist in a typical transverse bridge deck joint: shear and bearing. The application of force is fatigue in nature. In a longitudinal joint, the stress field becomes even more complicated through the addition of flexural forces. If a longitudinal joint is over a girder, negative flexural moments in the joint are predominant. If a joint is not properly detailed to accommodate these flexural loads, there is considerable chance that a crack may occur in the joint, which eventually permits the intrusion of water through the deck and pro- motes the rapid deterioration of the joint. In the event that this research reliably concludes that post- tensioning cannot be eliminated completely, additional research should focus on developing a system in which the pan- els are lightly reinforced for self-weight to overcome handling and installation stresses, while the main reinforcement in both the transverse and longitudinal directions is provided through posttensioning in nature. Guidelines for a Closure Pour for Skewed or Integral Abutment Bridges Additional guidelines are required for the design and construc- tion of skewed PCDP bridges. As mentioned previously, if the bridge skew is severe, it is difficult to design and cast panels to match the skew. It is far simpler to design panels in a skewed fashion when only reinforced-concrete design concepts are used. Due to the skewed alignment at one end of a particular panel, the prestressing or posttensioning will introduce addi- tional eccentric forces into the panels. In situations that are not time-critical, it may be desirable to design the panels as tangents and provide cast-in-place closure pours. In addition, further investigation is needed for rapidly constructed closure pours. Practical, Durable Solutions for Forming Crowns in Precast Decks Reliable design details are needed to provide a practical solution for crown formation using two flat panels. Current solutions either provide additional dead loads to the structure or provide a longitudinal joint that is susceptible to water leakage. Addi- tional hardware can be built into the panels to provide rota- tional capability and eventually a crown to the deck panels. Recommendations Recommendation of concepts considered suitable for stan- dardization is based on a critical evaluation of factors that were uniformly applied to each concept via an evaluation matrix included with each evaluation report. Twenty-one evaluation criteria, as given below, were used to rank these concepts. Each criterion was given a score from 1 to 5, where 5 = very good and 1 = poor. The maximum score possible was 105. The score for each superstructure concept is as follows: • Concrete deck bulb tees = 84 • Concrete double tees = 83 • Decked steel stringer system = 88 • Decked steel trapezoidal box girders = 72 • Segmental concrete superstructure systems = 78 Certain important criteria about the appropriate super- structure system should be considered before a decision is made to choose prefabrication as the best course for bridge construction. A matrix of criteria for selecting modular and segmental superstructure systems is given in Table 3.4. Modular Superstructure Systems Modular superstructure systems composed of both steel and concrete girders are recommended for advancement to the subsequent tasks in which pre-engineered standards will be

106 prepared for these systems. Deck bulb tee, deck double tee, and decked steel stringer systems received the highest scores, as these are proven systems for rapid renewal. Even two lower- scoring alternatives have specific advantages for certain sites. Each modular system is expected to see a 75- to 100-year ser- vice life due to the quality of its prefabricated superstructure, the use of high-performance concrete, and the attention given to connection details. precaSt concrete Deck girDerS (BulB tee/DouBle tee) Conventional precast concrete girders have been well estab- lished for bridge construction in the United States for more than 50 years. There is wide acceptance for them among owners and contractors because they are easy and economi- cal to build and to maintain. In most cases, the girders are used with a cast-in-place (CIP) deck built on site. For ABC applications, the key difference is that the girders will have an integral deck, which eliminates the need for a CIP deck. The precast deck girders combine all the positive attributes of con- ventional precast girder construction with the added advantage of eliminating the time-consuming step of CIP deck construc- tion. Contractors familiar with conventional precast girder construction should have no difficulty in adapting to these newer deck girders installed using an ABC approach. Deck girders are a proven system, having been standardized for use by several Western states. The team expects that the deck girder bids will be very competitive when compared with the girder and CIP deck systems and may come in even lower for sites in which constraints to deck casting operations may exist. DeckeD Steel Stringer SyStem Similar to the concrete deck girder system, the decked steel stringer system is also a proven concept shown to be quite economical and rapidly constructed. Many states are familiar with the Inverset system or some variations of it. Standard- izing generic designs for commonly encountered spans and skews will provide a big boost to this modular concept, gain- ing quick acceptance and more widespread use. Length limi- tations resulting from press brake capacities will mean that the splicing of bent plate boxes will be required to reach spans in the range of 60 ft to 100 ft or more. As for the precast deck girders, the recommended connection will be the full moment connection for all the same reasons previously discussed. DeckeD Bent plate Box girDer SyStem The bent plate trapezoidal box is another innovative alternate design concept for steel girder systems that will be a good complement to the stringer systems in certain situations, such as curved ramps and bridges, bridges with limited under- clearance, and bridges where aesthetic considerations become deciding issues. At the present time, this system is recom- mended only for simple spans. A single sheet can be used for cold bending trapezoidal boxes up to about 60 ft. Spans up to 100 ft can be accommodated by splicing sections. This system Table 3.4. Selection Matrix for Modular and Segmental Superstructure Systems Criteria Decked Steel Stringer Decked Bent Plate Steel Box Concrete Deck Bulb Tee Concrete Double Tee Segmental Systems Spans <140 ft X X X X X 140 ft < Span < 250 ft (with field splicing) X X X Spans >250 ft X Tier 1 ABC: Can be completed over a weekend X X X X Tier 2 ABC: Can be completed in a few weeks X X X X Tier 3 ABC: Accelerate larger projects, saving weeks or months X X X X X Light weight is a priority due to site access X X Able to be constructed by local contractors X X X X Bridge has multiple similar spans/Long viaducts X X X X Bridge has continuous spans X X X X Should allow future widening X X X X Bridge has curvature X X Bridge has underclearance issues X X Spans with limits on falsework or ground access for construction X

107 is currently envisioned for short spans with no end continuity. Further research and development will be needed to validate a design approach for the negative bending regions. connectionS The option with full moment deck continuity is recom- mended to provide a more durable structure with redun- dancy as compared with the option with a welded plate connection combined with a grouted key. For bridges subject to moderate-to-heavy traffic and located in zones in which design is required for seismic loading, it is recommended that a full moment deck slab closure pour be used between adja- cent modules or precast sections. For multi-span bridges, additional efficiency could be achieved by making the beams continuous for live loads by installing cast-in-place concrete diaphragms at the piers using closure pours. The welded plate connection between the deck girders, though not recommended as a standard option, may be used by agencies as a lower-cost detail for bridges subject to light traffic, bridges carrying local roads, or bridges located in low seismicity zones in which seismic design is not required. As the design of prefabricated sections with connection plates and grout is not covered by AASHTO LRFD specifications, structure evaluation for permit loading would be difficult at the present time. Evaluation requirements for permit loading need to be developed in addition to the other requirements indicated in this report. overlay An overlay is not considered necessary when the girder and connections will be constructed with high-performance materials, which should provide good durability without an overlay or concrete topping. The deck segments will have ¼-in. overfill, which will be diamond-ground in the field to provide adequate rideability. All modular systems also allow for future widening. The design of modular systems and pre- cast girders could be standardized for various span lengths and commonly used beam spacings for efficiency. Spans in the 200-ft to 250-ft range could be constructed by splicing two separate girders of transportable lengths in the field. Segmental Superstructure Systems Segmental precasting of box girder bridges is a well-established construction method that offers many benefits on suitable projects. The transfer of precast segmental technology to chan- nel sections and solid or voided slabs should not pose particu- lar technical challenges and would result in new structural solutions for ABC applications. Posttensioned slab spans can provide economical, low-maintenance ABC systems for spans up to about 150 ft. aDapting Segmental SlaB SyStemS for routine BriDgeS Segmental slab systems, based on years of experience and the application of proven technology in the United States, can be scaled back and applied on typical grade separation projects for ABC with significant cost and schedule benefits. The technology can be adapted to a variety of widths and span arrangements as follows: • Solid slabs are economical up to 80 ft in length (for simple spans) and can be extended to 100-ft spans with continuity in the longitudinal direction. • Voided slabs are cost-effective between span lengths of 90 ft and 150 ft with continuity. • Voided slabs must be transversely prestressed prior to ship- ping. Solid slabs do not require transverse posttensioning unless the deck width and pier configuration require it. • Longitudinal posttensioning is applied in the field. • Joints are match cast in simple long-line forms with epoxy applied in situ. • Manufacture of segments does not require expensive forms, beds, or exotic detailing and can be readily self-performed by any competent contractor. • Erection equipment consists of simple erection girders that can be delivered on flat beds and erected or dismantled overnight. The critical issues for the success of this technology are as follows: • Simple and effective detailing; • Quality workmanship; and • Proper design of prestressing and posttensioning. The proper execution of the above parameters should result in competitive solutions on a bridge-by-bridge basis, and does not necessarily require a minimum number of segments to be cost competitive. The increasing number of smaller projects using conventional segmental technology across the country supports this point. Larger projects and multiple bridges will result in only a more competitive solution. Finally, significant life-cycle and performance benefits will also accrue, including the virtual elimination of cracking, watertight deck systems, reduced maintenance, and enhanced life expectancy. Overlays can be applied to any deck system to further extend durability and life expectancy. Precast Concrete Deck Panels Precast concrete deck panels (PCDPs) are a proven technology awaiting an engineering solution for several of the bridge renewal challenges. PCDP is somewhat unique from two

108 different perspectives when compared with the other super- structure elements presented in this report. First, full-depth precast concrete deck panels, whether posttensioned or used with a variety of cast-in-place joint materials, are the only ele- ments presented that are truly applicable to both the complete replacement of existing bridges and also in the rapid replace- ment of a deteriorated bridge deck where the underlying superstructure remains in serviceable condition. Second, precast concrete deck panels have been used for projects across the country for at least 40 years and have a record of proven performance. The advancement of precast concrete deck panel technology is not as much about the development of a new system as it is about promoting an exist- ing system as worthy of consideration in Phase III of the cur- rent project. This system will not reach its full potential until owners, designers, researchers, and contractors can use this technology for development of a precast deck system that addresses the need for the following: • Durable, long-lasting details that provide composite action with beams or girders. • Rapid construction for present-day applications. • Reliable removal for future deck replacement projects. • Ride quality, which translates into a longer-lasting deck through the elimination of impact loading from wheels passing across the joints. A number of systems associated with precast concrete deck panels have been developed, tested, and implemented with a wide range of success. These systems include reinforced con- crete, pretensioned concrete, and posttensioned concrete panels with a variety of joint materials. The research team has evaluated a variety of these systems and identified key fea- tures to be incorporated into a system for future standardiza- tion on a national or regional basis. Recommendations for implementation include panel and connections to beams and girders, panel joints, and posttensioning, all of which are cov- ered in the rest of this section. panelS anD connectionS to BeamS anD girDerS Full-depth precast concrete deck panels offer significant advan- tages in construction over conventional cast-in-place concrete. Advantages and recommendations include the following: • Reinforced concrete panels are most suitable for mass pro- duction and minimize the need for special hardware or casting yards. • Panels up to 8 ft wide and up to approximately 50 ft long will permit the largest possible pieces without the need for special transportation permits and large cranes for installation. • A minimum panel thickness of 7½ in. for girders spaced up to 11 ft, and 8 in. for girders spaced up to 12 ft. • Fully composite panel connection to the superstructures using shear pockets. Shear pockets should be grouted after any posttensioning (if used) is applied. • The use of larger-diameter shear studs should be further studied with an eye toward future implementation. • The current effort to modify the 24-in. maximum spacing requirement for shear connections up to 48 in. should be strongly considered. Research under NCHRP 12-65 pro- vides supporting lab testing results. This modification would simplify both new construction, as well as the future removal of precast deck panels when a deck replacement is required. • Flowable, self-leveling, freeze–thaw durable, nonshrink grout mix should be used. The grout material, if stored on the construction site, should be kept protected from envi- ronmental factors such as humidity and rain. panel JointS To create a durable maintenance-free bridge deck system, precast concrete panels must use the most-reliable joints pos- sible. Given that most precast deck panel installation projects will likely include full-depth panels that span the entire road- way width, only transverse joints will be discussed in these recommendations. • Ultra-high-performance concrete joint filler. Although UHPC has not been widely used for precast concrete deck panel joints, the advantages offered by its very high strength, low permeability, and bonding capacity with precast concrete panels make this material highly suitable for this application. • Transverse, shear key joints shall be used to connect adja- cent precast deck panels. It is critical that all joints be designed, detailed, and constructed to be completely flush and provide full shear transfer across the joint. Flush joints are essential to eliminate impact loading due to truck wheel applications and the consequent problems with water intru- sion and long-term deterioration. • For staged construction in which only a partial bridge deck is constructed overnight, and in which longitudinal post- tensioning is provided with staged construction, the end panel of every night’s work shall be provided with tendon splicing devices (“dog bones”) that allow anchoring of the longitudinal posttensioning at the end of one work period and the resumption of posttensioning at a later stage. • Nonposttensioned connections that use bulged structural tubes for spliced reinforcing connections, as developed during NCHRP 12-65, should be considered for further development and implementation. These details offer the advantage of eliminating the need for posttensioned joints while creating a durable load transfer connection. • Posttensioning is an acceptable alternative to UHPC for ABC construction.

109 Posttensioning If posttensioning (PT) is used, the following additional rec- ommendations should be considered: • PT force should be applied to only the precast deck to obtain the greatest effective prestressing force in the con- crete deck. Therefore, the panel-to-girder connection should not be constructed until the PT is tensioned and anchored. • The minimum average effective stress on concrete due to PT should be at least 250 psi. • PT tendons should be uniformly distributed across the slab width and spaced no further than four times the structural composite slab thickness. • Maximum jacking stress in PT reinforcement should not exceed 80% of the specified minimum of the guaranteed ultimate tensile strength (GUTS) of the posttensioning steel. exhibits Different types of precast deck panels are shown in Fig- ures 3.22 through 3.27. Part 2: Evaluation of Precast Substructure Systems Overview In this section, the results of the evaluations for precast mod- ular abutments and complete piers are presented under the following headings: • Precast modular abutment systems 44 Body; 44 Wings; and 44 Support options, piles, shafts, spread footing. Figure 3.22. Deck bulb tee girders.

110 • Precast complete pier systems 44 Whole pieces, footing, shaft, cap; and 44 Support options, piles, shafts, spread footing. • Segmental columns and piers 44 Segmental columns; 44 Pier caps; and 44 Footings. This review documents each of the concepts, provides a review of the associated research literature, and provides a review of the engineering and constructability evaluations, as well as pinpoints implementation challenges and provides suggestions to overcome those challenges. In addition, testing needs and future research are also discussed. The review shows that the precast modular abutment and precast pier design concepts described in this report are worthy of promotion to Phase III implementation. With the results from this research project, and those in the future, the codification (development of code specifications) process of precast substructures will begin. Codification will lead to stan- dardization, and standardization will give the design engi- neers a measure of comfort and liability protection. These tools will lead to greater designer acceptance. With designer acceptance, precast substructures will be proposed as solu- tions more frequently and owner acceptance will increase. With DOTs committed to the concept, contractors will begin to embrace the concepts as well. As is proven with all new technologies, as acceptance and use increases, the costs will decrease and the product will evolve. The reduced costs will provide owners the tangible financial incentive necessary to pursue precast substructures for projects on a greater scale. This research project, as well as future projects, will provide the necessary stepping stones for this evolution. The precast substructure systems are design concepts on their way to codification, standardization, and implementation. Figure 3.23. Deck double tee girders.

111 Design Concept Descriptions Precast Modular Abutments Precast modular abutments are composed of separate com- ponents fabricated off site, shipped, and then assembled in the field into a complete bridge abutment. Precast modular abutments have been constructed in several states. The cur- rent schematic details employed by the Utah Department of Transportation (DOT), have been used by the Utah DOT and are proven and complete. The Utah DOT precast modular abutment details include a stub-type abutment on drilled shafts or piles and a cantilever abutment on spread footings. The Utah DOT makes use of an integral connection of the superstructure to the substructure. Since not all states use integral abutments, standards should be created for both integral and non-integral abutments. Also, non-integral abutments would be easier to reuse. In signifi- cant seismic zones, easily detachable seismic restraint devices may be used to connect the abutment to the superstructure to prevent the superstructure from losing vertical support during an earthquake. The abutment consists of standard- length cap sections, or wall sections for cantilever abutments, with a precast backwall attached by grouted splice sleeves. An example of a precast modular integral abutment can be found in Figure 3.36 and alternate connection ideas can be found in Figure 3.42. The individual precast components should be designed to be shipped over roadways and erected using typical construc- tion equipment. The precast components are made as light as is practical. Voids are used vertically in the cap section, or wall section, to reduce shipping weights and to allow for larger elements to be used. These voids are also used to attach drilled shafts or piles to the cap for stub-type abutments. Once the components are erected, the voids and shear keys are filled with high early strength concrete. Wingwalls are also precast, with a formed pocket to slide over wingwall drilled shaft Figure 3.24. Modular decked stringers.

112 reinforcing. Once in place over the wingwall drilled shaft, the wingwall pocket is filled with high early strength concrete. Precast Complete Piers Precast complete piers are composed of separate components fabricated off site, and then shipped and assembled in the field into a complete bridge pier. While the Utah DOT includes standard schematic details for complete piers, they have not been used in their current state. Similar concepts are envisioned by the PCI Northeast Bridge Technical Commit- tee and have been studied by the University of Alabama at Birmingham. Thus, the Utah DOT precast complete pier standard schematic details is a good place to begin. Piers with single-column and multiple-column configura- tions are available. Foundations can be drilled shafts with precast footings or precast spread footings. Attached to the foundation by grouted splice sleeve connectors is a precast column. Precast columns are an octagonal shape, the top of which is connected by grouted splice sleeves to the precast cap. The precast cap is a standard rectangular shape. Many states employ the use of integral piers, so it is recommended that standards for both non-integral and integral piers be created. Also, non-integral piers would be easier to reuse. In significant seismic zones, easily detachable seismic restraint devices may be used to connect the pier to the superstructure to prevent the superstructure from losing vertical support during an earthquake. An example of a precast complete pier can be found in Figure 3.37, and alternate pier ideas and con- nections can be found in Figures 3.40 through 3.44. Like the precast modular abutment, the components of the precast complete pier should be designed to be shipped over roadways and erected using typical construction equipment. The precast components are made as light as practical. Precast spread footings can be partial precast or complete precast com- ponents. A grout-filled void beneath the footing is used to transfer the load to the soil, avoiding unexpected localized point loads. Column heights and cap lengths will be limited by Figure 3.25. Modular decked steel trapezoidal box girders.

113 transportation regulations and erection equipment. Alterna- tively, the cap length limitation can be avoided by using multi- ple short caps combined to function as a single pier cap. Precast bearing seats can also be used. Segmental Columns and Piers While it is preferable in precast concrete construction to have the columns fabricated in full-height segments, many times the need arises to fabricate them in several pieces for ease of transportation to the site and placement in the bridge sub- structure. Usually weight controls the maximum size of each segmental component for the columns. Segmental columns consist of components of varying length, based on the demands of the design, that are stacked vertically until the designed height for the columns has been reached. Once they are in place, these column segments may be vertically posttensioned together and to the foundation for stability. For large-size piers, it is much easier to handle and erect these discrete components as compared with whole col- umns of equal heights. Concrete segmental column components can be thin-walled hollow segments to reduce weight. They can be match cast to ensure proper alignment, as well as full contact of the concrete at the joints. The segments can also be mass-produced with a thin layer of mortar bed between segments. The mortar bed should be designed to resist the actual loads from design, to provide a thorough closure of the joint, and to be designed with proper creep and shrinkage characteristics. Shims can be used to maintain vertical alignment. Engineering Evaluation Precast Modular Abutments Many pilot projects have been completed with precast modu- lar abutments. The initial success of these projects has proved the viability of the design and construction procedures of pre- cast substructures. Although these systems have been used, Figure 3.26. Deck segments.

114 additional testing such as seismic response and strength tests of connections are recommended to gain confidence from designers nationwide. Over the next several years, these structures will undergo the scrutiny of maintenance inspections. The evolution of these systems depends greatly on these inspections and on how durable these structures prove to be. Research into the use of modern materials for the purpose of increasing ease of construction and improving durability is also recommended. Use of high-performance concrete, early strength concrete, and self-consolidating concrete may reduce section size and weight, improve strength, and ease construction. The Utah DOT Precast Substructure Elements Manual (Utah DOT, 2010c) and the PCI Northeast Bridge Technical Committee’s Guidelines for Accelerated Bridge Construction (PCINE, 2006) provide adequate guidance for the design of precast substructures, and the Utah DOT’s standard sche- matic details with guidance on their use provides adequate direction on the development of design details for precast substructures. In addition, NCHRP Project 12-74 and its report, NCHRP 681: Development of a Precast Bent Cap System for Seismic Regions (Restrepo et al., 2011), and FHWA’s Con- nection Details for Prefabricated Bridge Elements and Systems (Culmo, 2009) provide guidance on available connections. This report provides a short summary as well. Currently, a design can be developed by using existing codes and design guides. However, additional information and commentary is suggested for the LRFD specifications concerning emulation design, mechanical couplers, and cur- rent seismic criteria. Any additional information in the LRFD specifications would contribute to a designer’s confidence in the design process. Precast Complete Piers As with precast modular abutments, pilot projects have been completed using precast complete piers. Most precast complete Figure 3.27. Joint details.

115 piers are constructed by posttensioning the precast cap to the precast column. Additionally, precast caps on cast-in-place columns and complete segmental piers have been used and have proven to be successful on many projects. The initial suc- cess of these projects has proved the viability of the design and construction procedures of precast substructures. Although these systems have been used, additional testing for seismic response and strength of connections is recommended to gain confidence from designers nationwide. The future of these systems also will be tested through their ability to prove durable and to withstand the scrutiny of maintenance inspections. As in precast modular abutments, research into the use of modern materials to ease construction and improve durabil- ity is recommended. Using high-performance concrete, early strength concrete, and self-consolidating concrete all may reduce section size and weight, improve strength, and ease construction. The Utah DOT precast substructure details presented in Figure 3.37 make use of the grouted splice sleeve connectors for mechanical connections in which 100% of the splice is at one location. Currently, this requirement would require special detailing for high seismic regions. The grouted splice sleeve connector does not depend on the surrounding concrete cover to develop its strength as with typical lap splices. The Utah DOT and other state DOTs are currently pursuing funding for additional research into the seismic behavior of these mechanical connections and funding to develop AASHTO code specifications. While these details include the grouted splice sleeves, other connection methods are available. The designer should refer to the FHWA’s Connection Details for Prefabricated Bridge Ele- ments and Systems (Culmo, 2009) and NCHRP Project 12-74, and its report, NCHRP 681: Development of a Precast Bent Cap System for Seismic Regions (Restrepo et al., 2011), for additional information. Here, too, designs can be developed using existing codes and design guides, but they would be greatly enhanced through additional information and commentary in the LRFD specifi- cations concerning emulation design, mechanical couplers, and current seismic criteria. Segmental Columns and Piers Many projects—whether pilot projects or those completed for convenience and ABC purposes—have been successfully completed with segmental columns and piers. Successful completion of such projects has proven the viability and advantage of such applications. Although the application of segmental columns has been successful in the past, a few issues have arisen during the experience. Some issues such as tendon corrosion are similar and apply to all types of segmental construction, whether horizontal or vertical, that took place prior to 2000. These issues are related to the location of the region within the columns where the posttensioned tendons were applied, the types of grout used for filling out the duct, poor workmanship, some inefficient detailing of the anchorages, and inadequate levels of protec- tion that were required to protect the tendons and anchorages from moisture. Match casting the segments is the preferred way of fabri- cating column segments because it avoids most of the issues discussed. Match casting results in an almost-perfect joint fit between the two adjacent components. It provides for tighter joints between components, better distribution of stresses across the joint, and easier erection on site, which accelerates on-site construction. Segments can be match casted vertically or horizontally. While vertical casting of columns segments provides better quality of components, it usually results in a more costly prod- uct. It requires several handling and movement operations for the components. In addition, formwork depth is limited for long column segments. Horizontal match casting has also been used efficiently. It avoids the repetitive handling and movement operations by casting the segments in line next to each other for the available length of the bed. Also, the length of the formwork does not impose any limitation to the length of the segments to be cast. But casting the segments horizontally has disadvantages. Some of the most popular cross-section shapes, such as the circular shape, can be difficult to cast horizontally and are therefore more costly. In addition, horizontally cast components can result in finishes that vary from rough to smooth within the same segment, which could be aesthetically unacceptable. Segments can be produced in mass with a thin layer of mortar between segments. The number of segments that could be produced at one time with this procedure is limited only by the length of the bed and the number of formworks available. Although it is usually cheaper to fabricate the seg- ments this way, many contractors prefer the match cast proce- dure of segment fabrication. This is mostly because there is a need to include connection hardware in the ends of the seg- ments. In addition, erection of such segments requires form- work at each joint before the grout can be placed, which adds to the labor and time to erect these segments. The latter, com- bined with the additional time needed for the grout to cure before the placement of additional segments, reduces the advantage of the accelerated intent of the project. Grouted joints may unintentionally be built with a non- uniform bearing surface. In such instances the column com- ponents may suffer stress concentration resulting in edge crushing, cracking, gaps, chlorite penetration, and ultimately corrosion of the mild steel and of the posttensioning.

116 Shear along the joint surface of the column is adequately resisted in most instances by shear friction. Shear keys can be used to enhance the shear resistance of the column at the joint location. In segmental columns, posttensioning is designed to resist flexure in the column. Adequate posttensioning is usually required to avoid the opening of joints due to service load flexure. This requirement may impose excessive demand for initial compressive stress in the segments, thus diminishing available ductility, which is very important in seismic applica- tions. Special care should therefore be applied when design- ing these systems in seismic regions to avoid crushing of the concrete from the imposed reversed cyclic loading. Due to their integrity during transportation and erection, solid shapes would be preferred, where possible, to hollow thin-walled shapes that could be fragile and more easily dam- aged during handling. In addition, in instances in which seg- ments are located within the limits of potential plastic hinges, solid shapes would be more suitable to handle the deforma- tion demand, as well as to accommodate the confinement reinforcement that is required in such regions. The segmental nature of this type of construction makes these applications very suitable for the use of innovative modern materials to make hybrid columns. For example, engineered cementitious composites (ECC) could be used to fabricate certain segments to be included within the overall column at potential hinge regions, thus maintaining the integ- rity of the component and the system considerably better than regular reinforced concrete can. Small to moderate-size columns frequently used in stan- dard applications around the country would be more suitable for standardized segmental construction than are larger-size or unique columns and pylons that are used sporadically or only in special projects. The benefits of standardization and precasting are obtained through multiple uses of the standards. Segmental column applications have been used in many projects to date. These projects were intended to test the advantages of the systems in the field. Other applications were made to accelerate construction time in congested areas or were used where casting the concrete in place would have been difficult. Currently, there is a knowledge database of such system applications, including use in the field as well as testing in the lab. While no specifications currently address segmental col- umn design, such design could be made with existing appli- cable codes and guides such as the AASHTO LRFD Bridge Design Specifications and pertinent ASBI publications. It is recommended that AASHTO publications include appropri- ate language that would directly address the design of such systems as well as provide references of previous applications, connection details, and research performed to date. Constructability Evaluation Precast Modular Abutments While specialized equipment and innovative erection meth- ods can be effectively used with prefabricated complete super- structures, currently it is best to design the substructures so that their components can be shipped and erected using typi- cal equipment. The benefits of using specialized equipment and innovative erection methods would have to be great to justify the additional cost. For precast modular abutments, the benefits would not justify the cost. The designer should always consider state and local ship- ping size regulations, as well as erection load limits, for typi- cal cranes. Consideration should also be given to the pavement capacity of local streets, if necessary. The following can be used as a guideline for sizing precast concrete substructure elements on design–bid–build projects: • The width of the precast component and any projecting reinforcing should be kept below 12 to 14 ft. • The height of the precast component and any projecting reinforcing should be kept to 12 ft or less for vertical clear- ance at existing bridges. • The weight of each precast component should be kept below 100,000 lb to keep the size of cranes needed reason- able. Weights of 60,000 lb should be anticipated. For ABC projects, the designer should work with both the fabricator and contractor to size the elements on the basis of the fabricator’s equipment, the contractor’s equipment, and the load limitations of local shipping routes. An assembly plan showing all elements and connections should be a requirement for all prefabricated bridge projects. The R04 team estimates that an abutment with drilled shafts or spread footings can be built in about 6 days, from the beginning of the first concrete pour to the completion of an appropriate cure time. Alternatively, the team estimates that a precast abutment can be assembled on the site in about 3 days, from the beginning of the foundation construction to the completion of an appropriate cure time for closure pours and grouted connections. This results in a 50% reduction in construction time for the abutments. Precast Complete Piers Although specialized equipment and innovative erection methods can be effectively used with prefabricated complete superstructures, currently it is best to design precast complete piers so that their components can be shipped and erected using typical equipment. The benefits of using specialized equipment and innovative erection methods would have to be great to justify the additional cost. For precast complete piers, the benefits would not justify the cost.

117 As is the case with precast modular abutments, the designer should always consider state and local shipping size regula- tions, as well as erection load limits for typical cranes when designing precast complete pier components. If necessary, con- sideration should also be given to the pavement capacity of local streets. The following can be used as a guideline for sizing precast concrete substructure elements on design–bid–build projects: • The width of the precast component and any projecting reinforcing should be kept below 12 to 14 ft. • The height of the precast component and any projecting reinforcing should be kept to 12 ft or less for vertical clear- ance at existing bridges. • The weight of each precast component should be kept below 100,000 lb to keep the size of cranes needed reason- able. Weights of 60,000 lb should be anticipated. For design–build projects, the designer can work with both the fabricator and contractor to size the elements on the basis of the fabricator’s equipment, the contractor’s equipment, and local shipping routes. The R04 team estimates that a cast-in-place pier with drilled shafts or spread footings can be built in about 3 weeks, from the beginning of the first concrete pour to the completion of an appropriate cure time. Alternatively, the team estimates that a precast complete pier can be assembled on site in about 1½ weeks, from the beginning of the foundation construction to the completion of an appropriate cure time for grouted con- nections. This results in a 50% reduction in construction time for bridge piers. The total substructure construction time sav- ings increases with each additional pier. Since most bridges have multiple piers, the benefits of complete precast piers become clear. Segmental Columns and Piers Segmental columns and piers should be designed with as few components as possible. Fewer segments would allow for more accelerated construction and would improve the dura- bility of the system, with fewer joints. However, the size of the columns combined with transportation and erection equip- ment limitations often dictate the number of segments that are needed for a particular application. Highway and water- way transportation limits must also be considered. Navigable waterways that could allow barging have fewer limitations than highways. The design should carefully evaluate at least one possible route that would potentially accommodate the transportation of components. Sizes of each component should be designed with consideration for all limitations, as well as what erection equipment is readily available to local contractors. However, if larger equipment would be needed for particular applications, the design could suggest the type of equipment to be used for that particular application. Construction of the substructure may expend 60% to 70% of the total construction time required for a project. If prop- erly designed, segmental columns could be very versatile and easy to erect, thus drastically reducing the time of construc- tion. It is estimated that such systems could take a few days for smaller projects to about 2 weeks for larger projects. This would mean a time reduction of about 30% to 50% of the time needed to erect the same column as a cast-in-place application. The use of segmental columns instead of single-piece pre- cast column segments needs to be considered during the design phase so that the project will run smoothly during the con- struction phase. Smaller multi-column piers can be efficiently constructed with single-piece columns, whereas taller single- column and wall-type piers lend themselves to the segmental approach. Most routine bridges with column heights under 25 ft should allow for the fabrication and erection of the col- umn as a single piece. It is not expected that segmental col- umns will see widespread use in ABC applications. The size and weight of the precast element has much to do with this decision. When the column sections would start to exceed 100,000 lb, the availability of cranes and transporters needs to be considered. Segmental columns can keep the crane sizes and precast element weights reasonable and economical. Implementation Challenges Designer Perspective A designer must provide a safe design. Safety is at the fore- front of each and every design. It is natural for a designer to be comfortable with a design, or design process, that has been in use for many years and to be equally uncomfortable with the opposite. A typical, or standard, design will help alleviate a designer’s reluctance to use precast substructures. The precast modular abutment and precast complete pier can be standardized in a manner similar to how cast-in-place standards were created. Standardization should start with typical roadway widths and common skews, with limits on abutment and column heights that take the emulated design to their strength limits. The Utah DOT precast substructure details presented in Figure 3.36 make use of the grouted splice sleeve connectors for mechanical connections in which 100% of the splice is at one location. The AASHTO LRFD specifications currently allow this only for structures in regions of the country with low seismicity (Zones 1 and 2). This requirement would require special detailing for highly seismic regions.

118 These details include the grouted splice sleeves, but other connection methods are available. The designer should refer to FHWA’s Connection Details for Prefabricated Bridge Ele- ments and Systems (Culmo, 2009) and NCHRP Project 12-74, and its report, NCHRP 681: Development of a Precast Bent Cap System for Seismic Regions (Restrepo et al., 2011), for additional information. The Utah DOT and other state DOTs are currently pursu- ing funding for additional research into the seismic behavior of these mechanical connections and funding to develop AASHTO code specifications. This research project, and those in the future, will begin the codification (development of code specifications) process. Codification will lead to stan- dardization, and standardization will give the design engi- neers a measure of liability protection. The designer must also provide an economical design. One might think that given the size of the proposed project, the eco- nomic benefit of a precast substructure system can be readily determined. However, there is an unknown factor that the designer may not take into account when determining the cost of a precast substructure system. This unknown is the risk pre- mium a contractor will factor into its bid. Currently, a design can be performed using existing codes and design guides. However, clarification is needed in the LRFD specifications for how precast substructures would be handled. Additional information and commentary would contribute to a designers comfort in the design process. Many designers are reluctant to propose or implement segmental column applications, even in projects in which the benefits of their use are readily obvious. There are no published documents that provide design standards for such applications. While engineers could combine creativity and their engineering judgment to implement such applica- tions, doing so could open their work to unnecessary liabil- ity. In addition, as a result of the lack of design standards, such applications may require additional design time and adequate time to perform an independent peer review of the design. It is usually very difficult for the designer to be awarded additional time from the owner to design such applications. All of the above are even more applicable for special designs such as seismic applications. While several studies have shown the adequacy of different segmental column applica- tions, there is still some uncertainty as to the behavior of these columns with opening and closing of joints, behavior and serviceability of the posttensioning tendons, and avail- able ductility when subjected to seismic reversible loading. Current testing has demonstrated that emulation design of precast substructures is sufficient for use of precast sub- structures for most regions. Testing of emulation design for use in high seismic zones will be required for complete accep- tance on a nationwide scale. NCHRP Project 12-74 has taken the first steps toward this goal. Refer to the section on connec- tions for additional information. Owner Perspective Despite the success of prior projects, owners are still reluctant to pursue prefabricated substructures for bridge projects without a financial incentive to do so. This financial incentive will need to be actual dollars added to the DOT’s budget, not just savings of user costs, which although substantial, do not provide adequate incentive in most cases. To encourage the use of ABC, such as precast substructures, perhaps the federal government could pay the states a substantial percentage of the additional cost of the ABC approach to remove the ini- tial financial disincentives. Currently, there are two critical financial disincentives that contribute to owner reluctance. These are the initial cost increase and the unknown future life-cycle costs. The initial cost increase can be rationalized if the user cost is included in the analysis, but in many cases this rationale is not adequate enough, since user cost cannot be recovered. The life-cycle costs depend on the long-term performance of the precast substructure. Thus, durability becomes a major concern. While precast elements by themselves are viewed as durable, it is the durability of the connections of these ele- ments that needs to be confirmed. Although many of these connections have been used in commercial buildings for many years, an owner will want to see long-term, time-tested performance of the connections in bridge applications before confidence is gained. Outside of these costs, other peripheral costs, such as the cost for future widening, the cost for future demolition or removal, and the cost for load ratings for overload permits, are also present. The precast substructure is designed to emu- late cast-in-place construction. Thus, an analysis for over- loads should not be more difficult than that of cast-in-place construction. Additionally, widening and removal of precast structures should not be more difficult than that of cast-in- place structures. While segmental columns have successfully been imple- mented in several projects, there are still concerns about the durability of the components. The latter, combined with some findings of corroded tendons of segmental pylons in prime bridges such as the Sunshine Skyway Bridge in Florida, make owners more uneasy about these applications. Using actively applied preloading on the components and counting on that load to ensure the integrity of the overall system poses some uncertainty. Single columns especially are by default nonredundant elements, and the possibility of tendons cor- roding without being detected is a dreadful thought to many owners.

119 This research project, as well as those in the future, will begin the standardization process, and standardization will lead to repeatable, cost-effective, and constructible projects. The cost-effectiveness will give the owners the unassisted financial incentive they need to pursue precast substructures for their projects. Contractor Perspective The time savings of precast substructures over cast-in-place substructures could be as much as 50%, which is significant. The more piers involved, the greater the total time savings. Contractors make money when their forces are busy and gainfully employed. Some contractors see an issue with the use of a subcontractor to supply the precast elements because it cuts into work contractor forces normally perform. Being able to do more projects per season may be the answer to keeping contractor forces working and turning a profit for the contractor. Keeping the details as simple as possible and allowing contractors to self-perform the precasting instead of subcontracting will also be attractive features. There is no reason to go to a precaster, because the more the contractor does itself, the more money it makes. Self-performing SuBStructure precaSting By general contractorS This is an approach to ABC in which the designs allow maxi- mum opportunities for the general contractor to do its own precasting at a staging area adjacent to the project site or in the contractor’s yard with its own crews. Unless a precast is posttensioned or prestressed, there is no need for a precaster. Substructure components are made of conventional rein- forced concrete that can be precast by the general contractor without the need to apply for any plant certification. Con- tractors have precast pier caps to the side of the project site in the past. Several states, including Nevada and West Virginia, do not have an established precast industry nor certified pre- casters, so precast girders come from elsewhere and incur an extra shipping expense. These states have contractors who are very good with CIP. Contractors self-performing the precast elements will overcome such impediments. The solution is to introduce the industry to precast technology and demonstrate its profitability. Substructure components need to be designed to allow contractors to self-perform the precasting, giving special con- sideration to the following: • Components that are simple enough to fabricate. • Components that allow some tolerance for erection. • Maximum repetition of components to reduce form- work cost. • Component weights preferably not exceeding 50 tons, to allow easy transportation and erection. Heavier compo- nents can be used depending on contractor capabilities and site conditions. • Substructure components that do not need prestressing or posttensioning in the field. • Connection details that are easy to construct. The use of precast elements for substructures has been impeded by the weight of components and by hauling. In addi- tion, precast substructure elements do not provide much toler- ance for field installation. Limitations on the cranes that can be used at a site are a major concern with pre fabricated elements. A reasonable upper limit for such elements is 100 tons. Site access is a large problem, and the use of larger cranes is also more expensive. Trucking and lifting can be an issue with larger precast elements. Lightweight concrete mixes can be used to lighten sections, enabling geometrically larger section placements by the same cranes. Designing for self-performing of precasting by the contrac- tor requires closer collaboration between designer and con- tractor. States are often restricted in what they can do to engage the contractor during the design phase, which makes ABC more difficult. Some states have used construction manager/ general contractor (CM/GC) contracting to improve this early communication without going to design–build. ABC is per- ceived as being more risky by contractors. Closer collaboration between designer and contractor will greatly reduce the risk premium that is often built into ABC bids. Rapid construction requires the designer to spend a lot more time on site and to be instantly available in the design office to work with the contractor and DOT construction personnel to provide speedy responses. A close partnership between designer, owner, and contractor is a key to success for ABC projects. Such a collaborative approach is particu- larly important when precast substructures are involved and when field adjustments are far more likely to be required to accommodate site conditions and foundation issues. Since precast elements are designed for transport over the road and erected with typical construction equipment, trans- portation and erection should not pose a problem. Particular attention needs to be paid to weight limits that may govern the constructability of large substructure components. An increase in transportation costs will occur if precasting is done off site, and an increase in time and cost for crane rental could occur. Most precast substructure elements will require a crane for placement. Cranes require a level surface capable of with- standing the anticipated ground loads. The project site should lend itself for easy crane placement. If not, increased costs for crane platforms will add to the erection costs. A contractor’s reluctance to participate in this type of proj- ect stems from not knowing the risks involved and the ability

120 to price that risk. The contractor is in business to make a profit. Take profit away and the contractor will follow. Utiliz- ing past experience, a contractor will bid on a typical cast-in- place project with a good understanding of the tasks involved, how much those tasks cost, and how fast those tasks can be performed. When presented with a new process or procedure, the unknown risks become a factor to be included in the bid items and the actual profit margin will remain unknown until the end of the project. Another reason for contractor reluctance is the potential requirement for investment in new equipment. This reluctance can be greatly reduced if the state departments of transporta- tion were fully committed to the new construction techniques. New technologies, such as grouting, posttensioning, and so forth, may be necessary to erect the segmental columns. The contractor may not be fully familiar with these technologies and they may not be available locally. Investment in training for new skills and technologies may be required, or it may be necessary to outsource these activities to outside subcontrac- tors that may become future competitors. Precasters may need to make initial investments in buying new beds for casting segmental components. They have to be trained in the new technologies. However, the sporadic nature of such projects at the current time does not justify the initial investments for acquiring the new equipment and skills if future similar projects are not a certainty. Despite all of the issues enumerated above, contractors have shown interest in such future projects. During the focus group meetings, representatives from contractors and owners showed keen interest in this type of construction. Some com- ments indicated there is a clear advantage for using such applications in locations with seasonal work limitations and that the reduction of curing times would be very favorable. Preference was shown for the use of posttensioning bars instead of strands and tendons. Some concerns were expressed about the limitations of knowledge in posttensioning opera- tions from some contractors, which could limit the number of contractors that could do the work. Large design–build projects could potentially be a perfect opportunity to apply precast elements, mostly because the designer and contractor will work side by side to design and erect the precast systems. This research project, as well as those in the future, will begin the standardization process; and for the contractor, standardization will lead to repeatable structures that are bid- dable, constructible, and profitable. Connections Background Cast-in-place concrete structures are built with construction joints that usually involve lapped reinforcing bars. The principle of emulation design is to substitute an alternate connection that mimics or emulates the standard lap splice. For emulation design in seismic regions, the goal is for a prefabricated system to be comparable to a cast-in-place system in performance such as energy dissipation, ductility, stiffness, strength, and similar reliable failure modes. During Phase II of this study and prior to the evaluation of technical information related to ABC connection details, pre- vious Phase I information was updated and additional infor- mation was identified. Specifically, the sources of information for this report section include the following: • Phase I report; • Connection Details for Prefabricated Bridge Elements and Systems (Culmo, 2009) (hereafter referred to as “FHWA Connections Manual”); • NCHRP 12-74 project report, Development of a Precast Bent Cap System for Seismic Regions (Restrepo et al., 2011); • Guidelines for Accelerated Bridge Construction, (PCINE, 2006); and • Additional contacts with academia, state DOTs, and other pertinent federal agencies. Most states believe additional testing of connection details is important. In particular, they are concerned with connec- tion effectiveness (strength and stiffness) and durability. States in seismic regions noted that there is inadequate precast con- nection performance information. Many state DOTs are just starting to consider using ABC concepts and are only beginning to investigate experimental testing (laboratory or field). A popular beginning point for states has been to investigate the use of full-depth deck panels and their connection to each other and to bridge girders. A general theme noted is that the testing research for connec- tions is just scratching the surface and there are still testing needs. A few states indicated that while they would like better information, particularly with regard to long-term perfor- mance (durability), they will still proceed to implement an ABC concept when a situation calls for it. Figure 3.28 shows a pier substructure with 10 connection types labeled (AP, A, B, C, D, E, F, G, H, and I). Detailed dis- cussions for substructure connection types I, G, F, and E are provided in this section. The connections discussed were previously chosen from the FHWA Connections Manual and determined by the SHRP 2 R04 team to be the most useful for implementation. Subsequently, these connections (or similar types) were studied to identify previous laboratory test information. If such information was available, a brief description of the information is provided and a summary of additional laboratory testing needs is also provided. A dis- cussion on the need for construction specifications is also included.

121 Connection Type I An important aspect of project acceleration is expediting the construction of bridge foundations and substructures. In tra- ditional foundation construction, driven piles are frequently embedded in a cast-in-place footing. However, in an ABC environment, the use of precast footings could greatly speed construction. A reliable method of connecting driven piles to precast footings needs to be established. Additionally, the technology could be extended to not only driven piles (in groups) connected to a pile cap but more generically driven piles connected to integral abutments or bent caps (such as for pile bent construction). The recently published FHWA Connections Manual includes numerous concepts for the connection of concrete columns to cap beams, prestressed piles to footings, and details to con- nect steel H-piles and pipe piles to concrete footings or bent caps. Some details are provided and generally consist of forming a void to confine the piled head, which is then back- filled with concrete. The size and configuration of these blockouts tends to vary widely. The majority of details proposed at this point for con- necting steel piles to pile caps involve leaving an individual void for each pile or a larger void to capture the entire pile group. Some suggested details are provided for both piles subjected to uplift and those without uplift. It is undeter- mined to what level testing has progressed to provide the current recommendations. To date, very few prefabricated footing systems have been constructed. The Northeast PCI Bridge Technical Committee has developed conceptual details for the connection of a spread footing to steel piles. The major issues with a pile-to- footing connection are whether or not there is anticipated uplift on the piles or if there is a need to provide moment capacity in the pile connection. Uplift capacity can be achieved by welding reinforcing steel to the pile end and embedding the reinforcement in a closure pour (note that weldable rein- forcing steel is required for this connection). Moment capac- ity is achieved by embedding the pile top at least 12 in. into the footing. An example for the pile supported precast foot- ing is shown in Figure 3.29. Figure 3.28. Precast component connection types for pier substructure. Figure 3.29. Pile-supported precast footing with uplift.

122 No literature could be identified that documented previ- ously completed testing of these details. It is anticipated, however, that these details will perform just as well as cast-in- place footings. Connection Type G A review of the FHWA Connections Manual for footing to bottom-of-column precast connections was completed. The noteworthy Type G detail in the manual is Detail 3.1.4.2 B (Utah, Precast Column to Precast Spread Footing, FHWA Connections Manual, page 3-59), as shown in Figure 3.30. It should be noted that the specific Detail 3.1.4.2 B had not been implemented per information contained in the FHWA Connections Manual, and specific literature could not be identified that documented previously completed testing of these connection details. However, literature related to other connection details (precast column to precast spread footing) was identified and evaluated. Details for connection of pre- cast column to cast-in-place footings or pile caps follow one of two general approaches. • The first approach involves temporarily supporting the column base segment with mild steel reinforcement pro- jecting from the bottom and casting the footing around the column. This approach has been applied in practice in Washington, but no research was found in the literature regarding laboratory or field testing of such a connection. There is some literature related to analytical modeling, which did contain suggested specific experimental research needs. The requirement for temporary support introduces obvious construction and alignment issues. While strength would not appear to be at issue with this detail, durability concerns, especially in terms of moisture penetration, and seismic concerns may be raised. • The second general approach involves the use of propri- etary grouted coupler systems to splice reinforcement (either mild steel or posttensioning bars) at the joint between the column and footing. While tensile strength testing has been performed for specific coupler systems, and strength performance of similar details applied at column-to-pier cap tests are encouraging, little research was found in the literature to quantify behavior of the joint detail in terms of below-grade durability. This detail has been applied in low seismic regions, but durability in gen- eral, as well as development issues for seismic applications. should be investigated. In terms of constructability, careful tolerance control is required to ensure adequate alignment of spliced reinforcement. A third approach that involves casting a socket in the foot- ing to receive the column base has been tested in the labora- tory and has been applied widely in building construction in Europe. The annular cavity between the socket and column base is filled with a flowable epoxy grout once the column is appropriately aligned. This detail has proven robust in labo- ratory testing under cyclic loading and would likely provide a high degree of protection against moisture ingress. Source: Culmo, 2009. Figure 3.30. Column-to-footing connection using grouted splice coupler.

123 Seismic performance of grouted sleeve couplers is a research area that needs to be further explored before the couplers’ use in high seismic zones can gain acceptance. The use of these connectors is common in the precast building market and is being increasingly promoted for bridge construction using precast elements as well. Notable projects such as the Mill Street Bridge replacement in New Hampshire, the Edison Bridge in Fort Meyers, Florida, and many others have made use of this simple connection made in the field by grouting a coupler connecting two bars in lieu of a lap splice or other mechanical system. The use of these couplers is commonly noted in the FHWA Connections Manual. Additionally, the Utah DOT depicts such a mechanical joining concept for its development of statewide ABC standards. There are no design issues with the use of these couplers in nonseismic or low seismic regions. The underlying problem is the current AASHTO prohibition on splicing 100% of the reinforcing steel in a single plane for Seismic Zones 3 and 4. The requirement is that no more than 50% of the bars be spliced in a single plane and the remaining bars are at least 24 in. away. This would require deeply embedded couplers, and the likelihood of constructability problems would poten- tially be greater. The question is whether such a prohibition is warranted. These coupler systems have been extensively used in Japan and other high seismic locations around the world. The field performance has been reported to be excellent. Some states are still reluctant to accept the concept of splicing all the bars in a single plane. Their additional concerns are that the stiff- ness of the coupler assembly will produce short plastic hinge lengths and inadequate ductility. Recommended testing pro- gram of various axial-flexural tests would need to be con- ducted with multiple bars joined in a single plane to assess the ductility and overall seismic performance of precast elements connected with grouted splice sleeves. Connection Type F Connection types listed in the FHWA Connections Manual were reviewed. The most mature Type F connection identi- fied is Detail 3.1.2.1 B (Utah DOT, Precast Pier Column Sec- tion to Precast Pier Column Section, FHWA Connections Manual, page 3-45), as shown in Figure 3.31. Detail 3.1.2.1 B is listed in the FHWA Connections Manual as “Under Development.” It appears, however, that this con- nection has been used since publication. Published literature that documents experimental evaluations of Detail 3.01.2.1 B does exist. Generally, the experimental descriptions indicate that the evaluations were more of a constructability nature. In some cases, however, the constructability was at least par- tially evaluated using strength type tests. Although very little information on the specific application of Detail 3.1.2.1 B was identified for connection Type F, similar details have been applied in other connection types and experimental results do exist for these other connection usages. Concerns about future inspection of this type of connection exist. Connection Type E: Pier Column–to–Pier Cap Connection Considerable research has been conducted to develop reliable, constructible precast connection details. However, implemen- tation in seismic regions has been limited, primarily due to uncertainty in the seismic performance of these connections. New research has been conducted under NCHRP 12-74, Development of Precast Bent Cap Systems for Seismic Regions, to validate the expected performance of various precast sub- structure column-to-cap connection types for use in all seis- mic regions in the United States. The research used a CIP control specimen designed in accordance with the AASHTO LRFD Bridge Design Specifications, 3rd ed., 2004 with 2006 Interims, and the 2006 Recommended LRFD Guidelines for the Seismic Design of Highway Bridges. The control specimen was classified as seismic design category (SDC) D and designed, detailed, and tested as such. The precast specimens used the same design and testing basis as for the CIP control specimen. The performance of the precast specimens was then compared Source: Culmo, 2009. Figure 3.31. Column-to-column connection using grouted splice coupler.

124 with that of the CIP control specimen. Many connection types were tested, including the grouted duct connection and the cap pocket connection. The grouted duct connec- tion and the cap pocket connection were tested for seismic performance; however, the grouted splice sleeve was not tested. Descriptions of the grouted splice sleeve connection, the grouted duct connection, and the cap pocket connection are discussed below. Grouted Splice Sleeve (or Grouted reinforcinG Splice couplerS) The Utah DOT precast complete pier details make use of the grouted splice sleeve connectors. According to the FHWA Connections Manual, grouted splice sleeves are produced by several manufacturers. They are hollow-cast steel sleeves sim- ilar to a pipe. The sleeve, preferably used in the vertical direc- tion, is cast into the end of one element and a protruding reinforcing bar is cast in the end of the adjacent element. The elements are connected by inserting the protruding bars from one element into the hollow end of the coupler in the other element. The joint between the pieces is then grouted, and grout is pumped into the couplers to make the connection. An illustration of a grouted splice sleeve connection is shown in Figure 3.32. illuStration These connections have been thoroughly tested and can develop as much as 125%, 150%, and even 160% of the speci- fied yield strength of the reinforcing bars. However, only lim- ited testing has been performed to determine their behavior under seismic conditions. The Utah DOT precast complete pier details make use of grouted splice sleeve connectors in which 100% of the splice is at one location. The AASHTO specifications allow for splicing 100% of the longitudinal bars with mechanical splices at one location for low to moderate seismic zones. Additionally, NCHRP 12-74 has recommended that the grouted splice sleeve connection be used for limited- ductility connection applications only. The FHWA Connections Manual suggests the splicing lim- itation of this connection may be overcome with special detailing. The Utah DOT and other state DOTs are currently pursuing funding for additional research into the seismic behavior of these mechanical connections and to develop AASHTO code specifications relating to their use. Grouted duct (or Grouted poSttenSioninG ductS) Several states have experimented with the use of grouted duct, also known as grouted posttensioning ducts, for connections between precast concrete elements. These connections are similar to grouted reinforcing splice couplers (grouted splice sleeves) in that reinforcing bars or threaded rods are inserted into a sleeve made up of standard posttensioning duct. The difference is that the duct is nonstructural; therefore, addi- tional confinement reinforcing is required around the pipe to develop a significant connection. The posttensioning duct is much larger than a grouted coupler; therefore, tolerances are not as strict. A photograph of a grouted duct connection is shown in Figure 3.33. While the FHWA Connections Manual indicates that this connection was not recommended for high seismic areas at the time of publication, it does concede that research is ongoing (NCHRP 12-74) about precast connections for use in all seis- mic regions. More recently, NCHRP 12-74 has concluded that a grouted duct connection can be used for all seismic regions. The findings of NCHRP 12-74 concluded that the grouted duct specimen satisfied the performance goal of the seismic design, achieving an extensive drift without appreciable Source: Culmo, 2009. Figure 3.32. Grouted splice sleeve. Source: NCHRP 12-74 (Restrepo et al., 2011). Figure 3.33. Grouted duct connection.

125 strength degradation and exhibiting extensive plastic hinging of the column, limited joint distress, and essentially elastic behavior of the bent cap. Additionally, the report found that the emulative performance is concluded on the basis of closely matching overall behavior to the CIP control speci- men, including lateral force–displacement response; plastic hinging; joint shear stiffness; level of joint distress; pattern of joint cracking; strain patterns of bent cap and joint reinforce- ment; integral behavior between the bedding layer, column, ducts, and bent cap; and minor bar slip. cap pocket The cap pocket connection uses a medium-diameter steel pipe to create a void in the bent cap to house the column bars and serve as a stay-in-place form and equivalent joint hoop reinforcement. Normal-weight concrete is placed in the bent cap void and bedding layer to anchor the column bars. The cap pocket connection studied under NCHRP 12-74 used a single 18-in. nominal-diameter 16-gauge steel pipe. Photo- graphs of a cap pocket connection are shown in Figures 3.34 and 3.35. The report concluded that the cap pocket connec- tion can be used for all seismic regions. The findings of NCHRP 12-74 concluded that the cap pocket (full ductility) specimen satisfied the performance goal of the seismic design, achieving an extensive drift with- out appreciable strength degradation and exhibiting exten- sive plastic hinging of the column, limited joint distress, and essentially elastic behavior of the bent cap. Additionally, the report found that the emulative performance is concluded for the cap pocket (full ductility) specimen on the basis of closely matching overall behavior to the CIP control specimen, including lateral force–displacement response; plastic hing- ing; joint shear stiffness; strain patterns of bent cap longitu- dinal reinforcement; integral behavior between the bedding layer, column, pipe, and bent cap; and minor bar slip. connection uSe On the basis of the results of NCHRP 12-74 and the informa- tion provided in the FHWA Connections Manual, the team believes the grouted splice sleeve, the grouted duct, and the cap pocket connections emulate cast-in-place construction and can be used for precast substructure column-to-cap connec- tions for various seismic regions in the United States. Table 3.5 is a summary for precast substructure column-to-cap connec- tion use expected on implementation of the recommendations of NCHRP 12-74. Note that NCHRP 12-74 has recommended the grouted splice sleeve for limited-ductility applications only. While the grouted duct and the cap pocket can be used for both limited- ductility and full-ductility applications, the detailing require- ments between limited and full ductility may differ. Refer to NCHRP 12-74 for design and detailing examples of the grouted duct and cap pocket connections, as well as for design flowcharts. Also note that NCHRP 12-74 tested a limited- ductility cap pocket connection in addition to the full-ductility cap pocket connection. However, the performance of the limited-ductility specimen did not match the expressed intent of Article 4.7.1 of the 2009 AASHTO Guide Specifications for LRFD Seismic Bridge Design. While the results from NCHRP 12-74 tested precast sub structure column-to-cap connection types, the team Source: NCHRP 12-74 (Restrepo et al., 2011). Figure 3.35. Cap pocket connection close-up. Source: NCHRP 12-74 (Restrepo et al., 2011). Figure 3.34. Cap pocket connection. Table 3.5. Connection Types for U.S. Seismic Regions Column-to-Cap Connection Type Seismic Design Category Grouted splice sleevea A, B, C Grouted duct A, B, C, D Cap pocketb A, B, C, D a NCHRP 12-74 has recommended use for limited-ductility applications only. b NCHRP 12-74 tested both a limited-ductility and a full-ductility cap pocket connection.

126 recommends that additional testing be conducted for foundation-to-column connections. It also recommends that the grouted splice sleeve connection be further studied and tested for use in high seismic regions. Grouted splice sleeves should be subjected to characteristic cyclic loading (both seismic and non-seismic). Although more expensive, some specimens used in the cyclic and durability testing pro- grams should be full-size specimens. In scaled specimens, it is frequently difficult to get the reinforcement in the desired loca- tions, obtain the desired cover, and so forth. Proposed construction specifications for precast systems and connections could be developed specifically for projects in which accelerated bridge construction is one of the proj- ect goals. The construction specifications could be based on the testing results as well as on field observations of precast assembly erection from the demonstration project. The con- struction specifications could describe both materials (e.g., hydraulic cement, grout, connection hardware) and methods (grouting, placement, testing). Girder Connections at Piers for Seismic Regions The superstructure of a bridge is intended to remain elastic during an earthquake, and seismic superstructure designs are similar to non-seismic designs. Therefore, the precast con- crete superstructure designs developed for non-seismic areas can be implemented in the seismically active western United States. Unlike the superstructure, the substructure can expe- rience large inelastic deformations during an earthquake. Often, the frame action in the transverse and longitudinal directions is the primary mechanism for resisting seismic loads. Engaging the superstructure and substructure in the longitudinal direction is necessary to mobilize the frame action. Proper seismic design of precast piers entails a detailed evaluation of the connections between precast components (as discussed previously), as well as the connection between superstructure and the supporting substructure system. Monolithic action between the superstructure and sub- structure components is the preferred approach for the design of seismic-resistant precast concrete bridge systems. Plastic hinges are expected to form at the ends of the columns during an earthquake. Ductile action of the bridge during an earthquake event is achieved through plastic hinging in the columns. Plastic hinges may be formed at one or both ends of a reinforced concrete column. After a plastic hinge is formed, the seismic loads are redistributed until the second plastic hinge is formed. If there is no monolithic action between the superstructure and the bent cap in precast construction, either the girder seats or the column tops will act as pinned connections. This requires that for stability in the longitudinal direction, the column bases will need to be fixed to the foun- dation supports, which places substantial force demands on the base of the columns and the foundations, particularly in moderate to high seismic zones. Developing a moment con- nection between the superstructure and substructure makes it possible to reduce moments at the base of the column. An experimental research program at the University of Washington has developed and evaluated details for a precast concrete bridge bent substructure system with satisfactory seismic performance and suitability for rapid construction. The proposed system uses a small number of large bars grouted into ducts much larger in diameter to achieve the connection between components so that the substructure can be rapidly constructed. This product innovation, developed through Highways for LIFE programs, is intended to create a design methodology and details and includes laboratory test- ing to ensure that the detail can be deployed in varied applica- tions of ABC in seismic regions. Testing Needs Several different areas of testing needs have been identified during this research project. Some of these needs are cur- rently being addressed by researchers around the country; a few other areas will need to be examined in the future. The use of self-consolidating concrete has great potential when producing precast substructure elements. Several states have done research on the use and properties of this material. Regional differences in aggregate and other concrete elements will likely require some localized research to develop strength, shrinkage, and creep characteristics for the local application of self-consolidating concrete. The strength component is critical for substructure applications. The shrinkage and creep issues are less of a concern than they would be for a superstructure beam made from self-consolidating con- crete. Local precast suppliers will need to work with their DOTs to establish mix designs that meet the structures’ strength requirements. Joint systems have been tested for strength-related needs on grouted splices and grouted metal sleeves. The results of this test have shown good strength performance for these joints. Extending this research into the performance of these joints under seismic loads is needed. To use the ABC approach in seismically active regions, an extensive testing program is needed. This testing should show how the different joints react to cyclic loads and determine if the needed joint plastic- ity can be achieved for the precast substructure elements to be prescribed in seismically active regions of the county. NCHRP Project 12-74, Development of Precast Bent Cap Systems for Seismic Regions, has taken the first steps toward this goal. Refer to the earlier section on connections for addi- tional information. However, the team believes that testing the grouted splice sleeve for high seismic applications would be beneficial to expand the connection toolbox. In addition,

127 testing various foundation-to-column connections for high seismic regions is also needed. The acceptance of precast elements, such as prestressed concrete girders, by state DOTs has been well established for more than 50 years. Most state DOTs report that they are comfortable with precast concrete in individual elements. Issues arise when the precast elements are assembled, as prob- lems have been documented in joints between elements and in posttensioning systems, when they have been employed. Further proof testing in an accelerated environment may be needed to show the longevity of the currently proposed joint systems. One recurring query from the DOTs was, Will the precast system last as long as the tried-and-true cast-in-place alternate? Proving the longevity of the precast joint systems will do much in moving the precast system toward imple- mentation in many states. Summary and Recommendations Precast modular abutments, precast complete piers, and pre- cast segmental piers are design concepts worthy of promotion to Phase III implementation. Sufficient design guidance and details exist and can be used to form the basis of nationwide standard designs. Standard drawings should be created for both integral and non-integral abutments and piers. Stan- dardization should start with typical roadway widths and common skews. Standard details for alternate connections should also be included. The precast substructure elements should be designed so that they can be shipped and erected using typical equipment. Over the last 10 years, many successful segmental pier proj- ects have been completed. Sufficient design guidance and details now exist and can be used for ABC Designs. The use of segmen- tal columns over the possible use of single-piece precast col- umn segments needs to be considered during the design phase. Smaller multi-column piers can be efficiently constructed with single-piece columns, whereas taller single-column and wall- type piers lend themselves to the segmental approach. Most routine bridges with column heights under 25 ft should allow for the fabrication and erection of the column as a single piece. It is not expected that segmental columns will see widespread use in ABC applications. Standardizing their designs for ABC use is not recommended. This project takes the approach that for ABC to be success- ful, ABC Designs should allow maximum opportunities for the general contractor to do its own precasting at a staging area adjacent to the project site or in the contractor’s yard with its own crews. This is particularly true for substructure com- ponents that have traditionally been constructed by contrac- tor crews. Substructure components are made of conventional reinforced concrete and can be precast by the general contrac- tor. Components will be designed to allow the contractor to self-perform the precasting by paying special consideration to the following: • Components that are simple enough to fabricate. • Components that allow some tolerance for erection. • Maximum repetition of components to reduce formwork cost. • Component weights preferably not exceeding 50 tons. • Substructure components that do not need prestressing or posttensioning. The R04 team recommends testing and research be pro- moted for developing specifications and commentary con- cerning emulation design of current cast-in-place elements by using mechanical couplers and required seismic criteria. Codification will provide designers with needed knowledge and liability protection, freeing them to propose precast substructure solutions in greater numbers. The benefits of precast substructures can be emphasized to owners, and with increased exposure to solutions from designers, own- ers can be better informed, leading to a technical acceptance of the concept. This research project begins the standardization process, and standardization will lead to repeatable, cost-effective, and constructible projects. The eventual cost-effectiveness will give the owners the incentive they need to pursue precast sub- structures for their projects. As long as the owners are com- mitted to precast substructures, contractors will be committed as well. When the tipping point is reached, the risks will be known and contractors can bid on precast substructure proj- ects with the knowledge and experience needed to reduce or eliminate the risk premium associated with current precast substructure projects. exhiBitS Different types of substructure configurations and details are shown in Figures 3.36 through 3.45. part 3: evaluation of aBC Construction technologies Overview The objective for the SHRP R04 project, Innovative Bridge Designs for Rapid Renewal, is “to develop standardized approaches to designing, constructing, and reusing complete bridge systems that address rapid renewal needs and effi- ciently integrate modern construction equipment.” While other parts of this Phase II report have focused on prefabricated substructure and superstructure components, the primary focus of this part will be on the application of proven and new construction technologies used for the rapid completion of bridge projects that employ either conventional

128 designs or ABC Designs similar to those presented in this report. Costs associated with investment in these technologies will not be addressed nor will they be evaluated at this time, as the intrinsic soft costs driving the need for accelerated or special- ized construction equipment and techniques may outweigh the additional project hard costs. Therefore, the opinion of the research team is that the needs of an individual project should be used to evaluate and select the type of construction equipment and techniques used and that the additional hard costs should be used as a final evaluation tool to decide whether to recommend a renewal project as an ABC project. With this said, the ABC type designs presented in this report have been developed with the intent that the prefabricated modular components could be erected using conventional equipment when adequate site access above, below, and sur- rounding the bridge project are available. The ABC technolo- gies reviewed in this part offer an alternative means to erect conventional or ABC bridge substructures and superstruc- tures when site access for conventional equipment is limited or restricted, or in situations in which the prefabricated lengths and weights of modularized components exceed the capacities or limits of conventional equipment. The construction concepts introduced in Phase I and sub- sequently carried forward for further evaluation in Phase II are as follows: C-1: Above-deck driven carriers (ADDCs); C-2: Launched temporary truss bridges (LTTBs); C-3: SPMTs and other wheeled carriers; and C-4: Launching, sliding, and lateral shifting. These ABC Construction Technologies can be grouped into two categories for use: • Bridge erection systems: Concepts C-1 and C-2 are tech- nologies in which the erection equipment is designed to deliver individual components of a proposed structure in a span-by-span process. The C-1 and C-2 technologies are intended to be easily transportable, lightweight, and mod- ular systems. The use of this type of equipment to deliver Figure 3.36. Precast modular integral abutment.

129 fully preassembled structures is not practical (although it is possible on a very small scale). • Bridge movement systems: Concepts C-3 and C-4 are tech- nologies in which the erection equipment is designed specifically to lift and transport large complete or partial segments of preassembled structures. The ultimate goal for this report would be to demonstrate that ABC technologies in both categories provide a means for a more rapid construction time period; create a safer working environment for both the contractors and the traveling pub- lic; minimize disruption to the surrounding natural environ- ment, residential neighborhoods, and business districts; and generate a method for better quality control over the final constructed product. Before setting the goals and expectations for the ABC technology vision too high, the team recognizes that not all bridge projects requiring rapid renewal may fit into one of the two ABC categories and thus may not justify or allow for the use of specialized equipment for a more rapid removal of existing structures or installation of new structures. A case could even be made that given the proper project criteria, use of conventional equipment would be the first choice for constructing a bridge designed with ABC modu- larized components. With this perspective, all bridge renewal projects can be categorized into one of the four design and construction project types, as follows: 1. Bridge Designs built with ABC Construction Technologies a. Designed assuming the use of precast or modularized substructure and superstructure components. b. Detailed using the ABC standardized component details presented in previous parts of this report. c. Constructed using ABC Construction Technologies. Figure 3.37. Precast complete pier (see Figure 2.6 for illustration). (text continues on page 135)

130 Figure 3.38. Alternate drilled shaft–to–footing connection.

131 Figure 3.39. Alternate column-to-foundation connection—Option 1.

132 Figure 3.40. Alternate column-to-foundation connection—Option 2.

133 Figure 3.41. Alternate column-to-foundation connection—Option 3.

134 Figure 3.42. Alternate drilled shaft to integrated precast abutment cap and backwall connection—Option 1.

135 2. Bridge Designs built with conventional construction a. Designed assuming the use of precast or modularized substructure and superstructure components. b. Detailed using the ABC standardized component details presented in previous parts of this report. c. Constructed using traditional equipment and pro- cesses currently accepted by the vast majority in the contracting community. 3. Conventional bridge designs built with ABC Construction Technologies a. Designed assuming the use of traditional structural systems. b. Designed assuming the use of well-proven standard DOT details. c. Constructed using ABC Construction Technologies. 4. Conventional bridge designs built with conventional construction a. Designed assuming the use of traditional structural systems. b. Designed assuming the use of well-proven standard DOT details. c. Constructed using traditional equipment and pro- cesses currently accepted by the vast majority in the contracting community. To properly evaluate the correct use of ABC Construction Technologies, the era in which the bridge construction indus- try originally exploded must be considered. This was a time when many smaller cranes dominated project sites, when significant detours, lane reductions, lane shifts and disrup- tion to the surrounding environment were standard prac- tice, and when months to years were acceptable periods for project completions. Today, and looking beyond, traffic vol- umes have ballooned beyond original estimates to a point in which detours, lane reductions, and shifts are typically unsafe; in which disruption to the surrounding natural envi- ronment, residential neighborhoods, and business districts may have heavy social and economic impacts; and in which Figure 3.43. Alternate drilled shaft to integrated precast abutment cap and backwall connection—Option 2. (continued from page 129)

136 more innovative construction equipment and structural sys- tems are available to provide more rapid renewal and less disruptive construction. The key for owners and their engineering consultants is to define the goals of a renewal project, survey the limits and constraints that could affect the design and construction of the project, evaluate the impact of those limits and con- straints, and finally develop a list of design criteria that will be used to prepare plans and specifications for construction. Goals The Phase I goals for the Task 6 evaluations of ABC Construc- tion Technologies are as follows: • Develop standard concepts for erecting highway structures using adaptations of proven long-span technologies that can also be adapted from project to project. • Document the potential time savings of an ABC Construc- tion Technology as compared with more conventional con- struction techniques and equipment. • Continue to examine the various innovative methods of erection to see if there are better ways to employ these in accelerated bridge construction. After reviewing the Phase I goals, the Phase II evaluation approach for ABC design concepts was modified to better suit the evaluation criteria for ABC Construction Technologies. The evaluations included • The development of a matrix of questions for owners and their consultants that would guide them in selecting a construction technology that best fits a project’s needs and limits. 44 The matrix would consider all four design and construc- tion project types. Figure 3.44. Alternate precast cap (two-column module).

137 44 The matrix would factor in the site and traffic variables and constraints. 44 The matrix would factor in the surrounding environment. 44 The matrix would be independent of the structure type. 44 The matrix would consider the project construction time period. • The development of a checklist of items that must be addressed during the design and construction phases of a project. a. These lists would be tailored to each of the four specified ABC Construction Technologies evaluated in this part. b. These lists would address design considerations and expectations for both the engineer of record and the contractor’s construction engineer. • The development of a set of standard conceptual details that define terminology or demonstrate the possibilities and lim- its of the four specified ABC Construction Technologies. Review of ABC Construction Technologies Above-Deck Driven Carriers Above-deck driven carriers (ADDCs) are designed to deliver individual components of a proposed structure in a span-by- span process with minimal disruption to the activities and the environment below the structure. Current ADDCs exist in two forms, and both perform a similar function. An ADDC rides over an existing bridge struc- tures and then delivers components of the new bridge spans by using hoists mounted to overhead gantries with traveling bogies. As shown in the examples below, the ADDC equipment can be quite specialized as in the case of the RCrane Truss sys- tem used by railroads to replace existing short bridge spans. Some systems, like the Mi-Jack Travelift overhead gantry, require specific site adaptations to align their wheel set with the centerlines of the existing girders that support the heavy Figure 3.45. Alternate integrated bent (two columns with cap).

138 moving loads. An example of ADDC equipment is shown in Figure 3.46. A modified ADDC concept would be a combination of the RCrane Truss and the Mi-Jack Travelift to create pairs of lightweight steel trusses supporting an overhead gantry system. This lightweight equipment could then be used on structures in which the existing bridge deck or girders are insufficient to support the heavier wheel loads of current ADDC equipment. This construction technology would be multifunctional, would be easily transportable both on urban and rural road systems, and would be mobilized with mini- mal erection and de-erection time. The trusses of the modified ADDC would be modularized into lengths that are easily trucked over both primary and sec- ondary roads (either shipped on flatbed trucks, or towed with the mountable rubber-tired bogies). Once assembled at the project site, the system would be equipped with several rubber- tired bogies that would be spaced to reduce and more evenly distribute the localized equipment dead load. Once the modi- fied ADDC is rolled out over across the bridge span(s), tempo- rary jack stands would be lowered at the piers and abutments and would bear on the deck where blocking had been added below, from the pier up to underside of the bridge deck. By bearing at the piers and abutments, the modified ADDC prevents overloading of the existing bridge structure dur- ing the delivery of the bridge components, as illustrated in Figure 3.47. This ABC Construction Technology would be applicable when an existing bridge or set of twin bridges is planned to be widened and when portions of the existing bridge are to be replaced, as shown in Figures 3.48 and 3.49. In an extreme case, with several movements, the technology could be used to replace an entire bridge. Although the modified ADDC concept is intended for use on more typical ABC bridge components, a heavier, more special- ized version of the concept could also be used for those cases in which the component lengths or weights exceed the limits and capacities of conventional equipment. Advantages of ADDCs include the following: • Minimizes disruption to traffic and the environment at lower level of bridge project. • Can be used where conventional crane access is limited by site constraints. • Allows for faster rates of erection due to simplified delivery approach of components. • Component delivery occurs at the end of the existing bridge, which minimizes disruptions at the lower level of the project site. • Reduces the need to work around existing traffic and decreases the need to reduce lanes, shift lanes, and detour lanes, which in turn improves safety for both the workers and the traveling public. • Can be used to deliver prefabricated modularized compo- nents of ABC substructures and superstructures. Additional advantages of modified ADDCs include: • The dual truss concept will allow either overhead gantries or under-slung hoists. • The lightweight trusses can be driven over existing bridge decks. • The temporary loads generated while delivering the bridge components are introduced directly into piers, eliminating potential overloading of the existing bridge structure. Figure 3.46. RCrane Truss. Figure 3.47. Modified ADDC rolled across span and delivering a bridge component.

139 Figure 3.48. Modified ADDC concept used for widening the inside edges of a set of existing twin bridge structures. Launched Temporary Truss Bridges Launched temporary truss bridges (LTTBs) are designed to deliver individual components of a proposed structure in a span-by-span process with minimal disruption to activities and environment below the structure. Examples of LTTBs are given in the next two figures. Currently, LTTBs exist in many forms; however, the basic technology is the same for each. LTTBs are launched across or lifted over a span or set of spans and, while acting as tempo- rary bridges, are used to deliver the heavier components of a span without inducing large temporary stresses into those components. As shown in the examples below, the pieces of LTTB equipment are designed and modified on the basis of varying sets of criteria from project to project. The equip- ment can be quite specialized, depending on the needs of the project, and can require extensive modifications from project to project based on changes in span lengths and component weights. With the industry moving forward with more-standardized bridge designs of different span lengths and component sizes and weights, the idea behind a modified LTTB is to create a set of standardized lightweight steel trusses that could be assem- bled to a specific length for a given project. The truss design and details would follow the quick connect concepts used in crane boom technology and would allow site modifications with relatively minimal effort. The lightweight equipment could then be used to bridge new spans to deliver components for a new bridge structure. This construction technology would be multifunctional, would be easily transportable both on urban and rural road systems, and would be mobilized with minimal erection and de-erection time. The trusses of the modified LTTB would be modularized into lengths that are easily trucked over both primary and sec- ondary roads (either shipped on flatbed trucks, or towed with mountable wheel-tired bogies). Once assembled at the project site, the lightweight equipment would then be launched from span to span or could be lifted into position with cranes, as Figure 3.49. Modified ADDC concept used for widening the outside edge of an existing bridge structure.

140 shown in Figure 3.50. Once the modified LTTB has bridged the new span, it would be stabilized and supported at each pier or abutment substructure unit. This ABC Construction Technology would be applicable whenever new bridge structures are to be erected; it could also be applicable when an existing bridge or set of twin bridges are planned to be widened, as illustrated in Figure 3.51. Although the modified LTTB concept is intended for use on more-typical ABC bridge components, a heavier, more spe- cialized version of the concept could be used for those cases in which the component lengths or weights exceed the limits and capacities of conventional equipment. Advantages of the LTTBs include the following: • Provides construction options when launching demands on the permanent structure add extra cost to the steel, con- crete, or posttensioning. • Minimizes disruption to traffic and the environment at lower level of a bridge project. • Can be used where conventional crane access is limited by site constraints (launched option only). • Component delivery occurs at the end of the existing bridge, which minimizes disruptions at the lower level of the project site. • Reduces need to work around existing traffic and reduces need to reduce lanes, shift lanes, and detour lanes, which in turn improves safety for both the workers and the traveling public. • Increases the possibility of erecting longer spans without significantly increasing the cost of bridge spans because the components of the spans can be delivered without addi- tional temporary erection stresses. • Where access to the bridge site is limited due to approach roadways that are difficult to traverse, shorter girder Figure 3.50. Modified LTTB launched across span and delivering a bridge component. Figure 3.51. Modified LTTB concept used for erecting a new bridge structure or possibly for widening an existing bridge structure.

141 segments can be delivered and preassembled at the site and then delivered as one unit over the span. • Allows work to proceed on multiple fronts (i.e., where multiple-span LTTBs are used, girders can be set while the next girder is delivered). • Temporary loads are introduced directly into piers, mini- mizing the need for falsework bents. • Can be used to deliver prefabricated, modularized compo- nents of ABC substructures and superstructures. SPMTs or Other Wheeled Carriers Self-propelled modular transporters (SPMTs) and other wheeled carriers are used to remove entire spans or full- length strips of existing bridges and to replace them with new preassembled structures in a quick and efficient manner. SPMTs are multi-axle mobile lifting devices that have been standardized by the heavy lift industry, as shown in Fig- ure 3.52. They are easily transportable and can be mobilized with minimal assembly time. With the use of computer syn- chronization, SPMTs can “crab crawl” in numerous direc- tions, perform 360° carousel-type turns, and traverse uneven terrain. With the steel cribbing and timber dunnage mounted to its lift table, an SPMT unit can be used to raise, transport, and set any type of heavy preassembled structure. Other types of wheeled carriers have been developed on a more specific project-by-project basis, but their purpose is similar to that of the SPMTs—to raise, transport, and set preassembled structures. The U.S. bridge industry is growing to accept SPMTs and other wheeled carriers as plausible ABC Construction Tech- nologies. Success stories involving the use of SPMTs are well documented: • The Utah DOT: 4500S over I-215E (Figures 3.53 and 3.54). • The Florida DOT: Graves Avenue over I-4. • The Louisiana Department of Transportation and Devel- opment: I-10 over LA-35. • The Chicago Transit Authority, Illinois: Main Street Via- duct Replacement. • The City of Toledo, Ohio: MLK Replacement Bridge. AASHTO has produced a Manual on Use of Self-Propelled Modular Transporters to Remove and Replace Bridges (FHWA, 2007c), which was developed through cooperation of the FHWA, AASHTO, NCHRP, and the Florida DOT. The man- ual provides information about the use of SPMTs as well as guidance toward working with and specifying the use of SPMTs for bridge projects including cost benefits, planning criteria, design considerations, contracting considerations, and case studies. The Utah DOT has developed its own Manual for the Mov- ing of Utah Bridges Using Self Propelled Modular Transporters Figure 3.52. Example of an SPMT. Source: Utah DOT. Figure 3.54. 4500S over I-215E—installing a new bridge. Source: Utah DOT. Figure 3.53. 4500S over I-215E—removing the existing bridge.

142 (SPMTs) (Utah DOT, 2008e) which includes a general over- view of the responsibilities of design, owner, and contractor teams and a comprehensive checklist for the engineer(s) of record, the contractor, the heavy lift team, the resident engi- neer, and the specialty bridge engineer. Heavy lift contractors in the United States that provide ser- vices that use SPMTs include the following: • Mammoet in Houston, Texas; • Fagioli in Pearland, Texas; • Berard Transportation in New Iberia, Louisiana; • Barnhart Crane and Rigging in Memphis, Tennessee; and • Bigge Crane and Rigging in San Leandro, California. This ABC Construction Technology has proven to be very applicable when removal of complete existing bridges and installation of preassembled new bridge structures must be achieved in a very short time. SPMTs may also be applicable to bridge widening when the new widened portion of the struc- ture is to be installed as a preassembled unit. Advantages of the SPMTs and other wheeled carriers include the following: • The standard SPMT unit is available in four- and six-axle units and can be grouped in any number of longitudinal and transverse combinations. • The standard SPMT units and power packs have been devel- oped with quick-connect pin and hydraulic connection points, making assembly and reconfiguration relatively easy. • By pre-erecting the bridge in its entirety on a support system mimicking the permanent condition, the deflections and reactions can be monitored to confirm design assumptions. • Traffic disruptions are minimized and safety for workers and the traveling public is improved by moving critical demolition and erection work tasks off-line from the traffic. • The quick removal of existing structures and quick installa- tion of new preassembled bridge structures are provided for. • Safety of the traveling public is improved by minimizing exposure to construction zones. • By pre-erecting the new structures in a controlled environ- ment, improved quality control can be achieved. Launching, Sliding, and Lateral Shifting The incremental launching construction method was devel- oped in Europe in the 1960s and is now typically used for con- struction of prestressed concrete, steel, and steel/composite bridges. The method involves building a bridge at a single con- struction location in sections and launching the bridge incre- mentally as each section is completed. Examples are shown in Figure 3.55 and Figure 3.56. Similarly, lateral shifting is a method in which a structure is constructed adjacent to an active bridge that is to be replaced. On a given closure period, the existing bridge is lifted and slid out of the way of the new bridge. The new bridge is then slid into position on the existing alignment. Figure 3.55. Example of bridge structure that has been launched or installed by sliding or lateral shifting technologies. Figure 3.56. Examples of bridge structures that have been launched or installed by sliding or lateral shifting technologies.

143 As illustrated by the examples below, both launching and lateral shifting technologies involve the use of specialized equipment. To launch a bridge structure, a contractor would potentially require a launching nose, a method of movement (jacks, tuggers, push/pull rams), guides, and rollers. To shift a bridge laterally, a contractor would require falsework to sup- port the new bridge while it is under construction, falsework to support the track system, a method of movement (jacks, tuggers, push/pull rams), guides, and rollers. To launch a bridge requires the use of a staging area behind one or both of the abutments where a partial or complete structure is constructed. For concrete structures, the entire bridge section is launched integrally. For steel structures, the entire steel system (girders, diaphragms, and floorbeams) is launched and then on completing the launch, the deck is cast in place and made composite with the steel girders. Typi- cally a launching nose is provided to reduce the weight of the lead cantilevered section. Depending on the available length in the staging area behind the abutment, the structure may be constructed full length and launched in one sequence or can be erected and launched in a series of partially completed stages. Depending on the available support mechanisms, the launched structure can either be pushed from a deadman at the back end of the staging area, be pushed or pulled by using rams that anchor to a travel track, or be pulled with winches anchored to the abutment or pier ahead of the launched struc- ture. Launching uphill against gravity is preferred to provide control of the launch movement; however, it is not manda- tory as long as the braking and locking mechanisms are prop- erly engaged. To slide or laterally shift a bridge requires the use of an available staging area adjacent to the existing bridge structure to be replaced. The falsework, track, and movement mecha- nisms are similar whether the structures are cast-in-place concrete slabs or boxes, precast girders with concrete decks, or steel girders with concrete decks. The size of the falsework, track, and movement mechanisms will be based on the weights of the different bridge types. By using the sliding or lateral shift technology, an existing structure can moved inde- pendently of a new structure or coincidently with the new structure. Examples are shown in Figures 3.57 and 3.58. A number of reference manuals and research papers dis- cussing incremental launching techniques exist; however, most of these focus on the launching of concrete structures. Although many aspects of the technology are similar, there are no specific reference manuals that focus specifically on the launch of steel structures. Neither are there specific guides addressing minimum design criteria for lateral shifting tech- nology. For reference purposes, readers may consult the follow- ing publications in this report’s reference list: AASHTO, 2003; American Segmental Bridge Institute, 2008; Podolny and Muller, 1982; Rosignoli, 2002, and Hewson, 2003. This ABC Construction Technology would be applicable when the construction of a new structure is challenged with difficult terrain below the bridge, when crane access below the bridge is limited or prohibited, or when traffic volumes require minimal disruption. Although not common, this ABC Construction Technology may also be applicable to bridge widening. Advantages of launching, sliding, or lateral shifting include the following: • Minimizes traffic disruptions and improves safety for work- ers and the traveling public by reducing or eliminating detours, lane reductions, or traffic shifts. • Improves quality control by pre-erecting the new structures in a controlled environment. • Minimizes the need for temporary works between spans (specific to launching). • Allows substructure and superstructure construction activities to occur concurrently. Figure 3.57. Example of a horizontally curved launched steel girder. Figure 3.58. Example of skewed bridge installed by lateral shifting technology.

144 • Minimizes disruption of environmentally sensitive areas under the bridge. • Reduces crane size requirements by pre-erecting the new structure behind the abutment. • Reduces overall construction time. Development of the ABC Construction Technology Selection Matrix The role of owners and engineers is to define the goals of a bridge construction project, survey the limits and constraints that could affect the design and construction of the bridge project, evaluate the impact of those limits and constraints, and develop a set list of criteria that can be used to prepare plans and specifications for construction. To assist owners and engineers in this process, the first goal of this part of the Phase II report on ABC Construction Tech- nologies was to “develop a matrix of questions for owners and designers that will guide them toward the proper selection of the construction technology that best fits a project’s needs.” The proposed tool for their assistance is the ABC Con- struction Technology Selection Matrix. This matrix con- tains a logical progression of questions; after answering these questions, the user will be able to ascertain whether a bridge renewal project warrants further consideration of the use of specialized ABC Construction Technologies or whether the site and project limits are more suitable for the use of conven- tional equipment and technologies. Although this SHRP report is focused on ABC Designs for bridge renewal and widening, we have included new construc- tion as part of the matrix criteria to give a complete picture of the total range of the selection process. The ABC Construction Technology Selection Matrix compels owners and consultants to consider the following variables: • Bridge project type a. New construction; b. Bridge widening; and c. Bridge replacement. • Site and traffic constraints a. Volume of traffic on the mainline; b. Volume of traffic on the secondary road; c. Detour possibilities or restrictions; and d. Feasibility or possibility of lane reductions or traffic shifts. • Available space (if any, where and condition) for construc- tion staging areas a. Off site, away from project site; b. Adjacent to existing bridge; c. Behind one or both abutments; and d. Physical limitations or potentials of the surrounding grades and structures. • Environment surrounding the project site a. Effects on residential neighborhoods; b. Effects on business districts; c. Degradation of sensitive areas of the natural environ- ment; and d. Critical access to existing roads. • Project construction time period a. Total project time; b. Limited closure times; and c. Early milestones. The questions and evaluation criteria within the ABC Con- struction Technology Selection Matrix have been selected to allow the evaluation of the proper construction technology to be independent of the structural system, potential costs, span configurations, and span lengths. It is the research team’s opinion that including these variables in the evaluation crite- ria at this stage would distract from critical design issues behind the reasoning for classifying the renewal project as a potential ABC project. The structural system, cost, and span data may eventually play a role in the final selection of the construction technology; however, a similar selection matrix has already been developed in this report for the independent evaluation of the proper ABC substructure and superstructure designs that best fit a renewal project. Once the independent evaluations of the two matrices are complete, the results can be reviewed and the best design and construction technologies can be selected. The ABC Construction Technology Selection Matrix recog- nizes that all bridge projects fall into one of four basic categories of design and construction types: • ABC bridge designs built with ABC Construction Tech- nologies; • ABC bridge designs built with conventional construction; • Conventional bridge designs built with ABC Construction Technologies; or • Conventional bridge designs built with conventional construction. Furthermore, the ABC Construction Technology Selection Matrix recognizes that all bridge sites can be categorized as one of the following four typical cross sections (see Fig- ure 3.59 through Figure 3.62). As a result of completing the matrix, owners and engineers will have a list of possible construction technologies that are applicable to their renewal project. These can be summarized into one of the following groupings: • ABC Construction Technologies are not practical given project restraints. • Launching or use of LTTBs.

145 Figure 3.59. Secondary road over mainline–rural (single or twin). Figure 3.60. Secondary road over mainline—urban (single or twin). Figure 3.61. Mainline road over secondary road (single or twin). Figure 3.62. Mainline crossing road (single or twin). • Use of only ADDCs. • Use of only SPMTs. • Launching or use of LTTBs or ADDCs. • Launching or use of LTTBs or SPMTs. • Use of ADDCs or SPMTs. • Launching or use of LTTBs, ADDCs, or SPMTs. • Sliding and lateral shifting. • Launching, sliding, and lateral shifting or use of LTTBs, ADDCs, or SPMTs. Step 1. Matrix Questionnaire A: Bridge Project Type The first of the selection matrices was developed to sort proj- ects quickly by project type. Once the bridge project type is selected, three subsequent questions sort and establish the worthiness of the project to warrant further consideration for ABC Construction Technologies. • Can service be disrupted with full closures during con- struction of the bridge project? a. If full closures are allowed, the matrix directs the user to a follow-up question to test the worthiness for ABC consideration. b. If full closures are not possible, this becomes a key cri- terion to establish possible worthiness for ABC Con- struction Technologies and the matrix directs the user to a new matrix. • Is a short construction duration critical to the project completion? a. If a shortened construction period is required, this becomes a key criterion to establish possible worthiness

146 for ABC Construction Technologies and the matrix directs the user to a new matrix. b. If a shortened construction period is not mandatory, the matrix directs the user to a follow-up question to test the worthiness for ABC consideration. • Is there full equipment access to the lower level of the proj- ect site? a. If equipment can be used on the lower level, the project at this point in the matrix has not demonstrated a manda- tory reason to consider ABC Construction Technologies. Therefore, the matrix suggests that the user consider more conventional construction equipment and techniques. b. If equipment cannot access the lower level, this becomes a key criterion to establish possible worthiness for ABC Construction Technologies and the matrix directs the user to a new matrix. Step 2a. Matrix Questionnaire AB: New Construction Once the user has been directed to the New Construction matrix questionnaire, the probability of the project being worthy for consideration of an ABC Construction Technology is very good. A second series of questions specific to the proj- ect site will help the user further evaluate whether ABC Con- struction Technologies might be applicable and, if so, which technologies best suits the project criteria. Note, however, that the answers may also demonstrate that due to physical site constraints, ABC Construction Technologies may not be applicable for the project. The New Construction matrix focuses on the physical aspects of the project site. • Is there room or access nearby for a staging area on the lower elevation where complete or partial assembly of the new bridge can occur? a. The use of SPMTs requires the ability to preassemble the new bridge structure in an off-line staging area, away from the project site. If there are no open areas nearby, SPMTs may not be applicable; however, the matrix directs the user to other questions that may open the possibility for other ABC Construction Technologies. • Is the staging area accessible to the bridge site without overhead, width, or grade restrictions? a. If an off-line staging area is available, the SPMTs and their cargo must be able to travel without interference from the staging area to the project site. b. If there are physical restraints preventing a clear travel path between the staging area and the project site, the matrix will direct the user to other questions that may open the possibility for other ABC Construction Technologies. • Is there room behind one or both abutments for a staging area? a. If there is room behind one or both of the abutments for a staging area, launching or use of the LTTBs may be applicable. b. If there is no room behind the abutments for a staging area, at this point in the matrix, the physical constraints of the project site have severely limited the possibility of using ABC Construction Technologies. Therefore, although some of the design criteria indicate the need for ABC consideration, the matrix suggests the user consider more conventional construction equipment and techniques. • Is a short duration closure or detour (1 to 2 days) allowed? a. If short closures are allowed, then SPMTs may be applicable. b. If short closures are not allowed, the work may have to be performed from above with launching techniques or by use of the LTTBs. • Is there space adjacent to the proposed bridge location to preassemble the new structure? a. If there is a staging area adjacent to the proposed bridge location, the new bridge superstructure could be erected as a parallel activity to the substructure construction and the bridge eventually moved into position by using sliding or lateral shifting technologies. Depending on how the questions are answered, the New Construction matrix questionnaire either will lead the user to the conclusion that the site constraints restrict the applicabil- ity of ABC Construction Technologies or will lead the user to one of the following set of ABC Construction Technology opportunities: • Launching, sliding, or lateral shifting, or the use of SPMTs or LTTBs. • Launching or use of SPMTs or LTTBs. • Sliding or lateral shifting, or the use of SPMTs. • Use of only SPMTs. • Launching or the use of LTTBs. Step 2b. Matrix Questionnaire AC: Bridge Widening Once the user has been directed to the Bridge Widening matrix questionnaire, a series of questions focusing on criteria developed for traffic control, the number of bridges, and the type of widening proposed will help the user further evaluate the applicability of ABC Construction Technologies and, if applicable, which technologies best suit the renewal project cri- teria. Note, however, that the answers may also demonstrate

147 that due to the ability to adjust the traffic flows, ABC Con- struction Technologies may not be required for the project. The Bridge Widening matrix focuses on the traffic con- trol, the number of bridges, and the type of widening pro- posed for the project. • Can traffic on the lower level be partially disrupted by short-duration closures, shifting, or detours, or is the lower level devoid of traffic? a. If the traffic below the bridge has high volumes and has been determined to be unsafe for traffic flow adjust- ment, this becomes a key criterion to establish possible worthiness for ABC Construction Technologies. b. If the traffic below the bridge can be disrupted or adjusted or if there is no traffic below the bridge, the matrix points out to the user that ABC Construction Technologies may not be required for the project. This does not preclude ABC Construction Technologies from consideration but does demonstrate that conventional construction equip- ment and technologies can be considered. • Is access for crane equipment limited on the project site? a. If crane access will be limited, this becomes a key crite- rion to establish possible worthiness for ABC Construc- tion Technologies. b. If the site allows for easy crane access, the matrix points out to the user that ABC Construction Technologies may not be required for the project. This does not preclude ABC Construction Technologies from consideration but does demonstrate that conventional construction equip- ment and technologies can be considered. • Is the project a twinning of a parallel structure or a widen- ing of an existing structure? If the project is a twinning, the matrix directs the user back to the New Construction matrix questionnaire; otherwise, the user is directed to additional questions related to the widening of an existing structure. Note that if the user had declared that the twin bridge was new construction in the opening matrix, the user would have been directed to the same New Construction matrix. • Is the existing structure a single bridge or twin bridges? a. If the bridge project consists of widening a single struc- ture, the matrix directs the user to a new matrix specific to the widening of a single bridge. b. If the bridge project involves widening two parallel bridges, the matrix will ask additional questions related to traffic control design criteria. • Can traffic from one direction be shifted to the other bridge? a. If the project involves twin structures and traffic from one direction cannot be shifted to the other bridge with a crossover, this becomes a key criterion to establish pos- sible worthiness for ABC Construction Technologies. b. If the project involves twin structures and a temporary crossover can be constructed to move traffic completely off one bridge, the matrix points out to the user that ABC Construction Technologies may not be required for the project. This does not preclude ABC Construc- tion Technologies from consideration but does dem- onstrate that conventional construction equipment and technologies can be considered. • What type of widening is proposed? a. The user will be directed to the next matrix level based on the type of widening proposed for the twin set of bridges. 1. Widen each bridge to the outside. 2. Widen each bridge to the inside. 3. Widen each bridge along both edges. Depending on how the questions are answered, the Bridge Widening matrix questionnaire either will lead the user to the conclusion that the flexibilities related to adjusting traffic patterns do not mandate that ABC Construction Technolo- gies be considered or will lead the user to one of the following matrix questionnaires: • Widen Exterior Edges of a Single Bridge; • Widen Both Interior and Exterior Edges of Twin Bridges; or • Widen Interior Edges of Twin Bridges. Step 2c. Matrix Questionnaire AD: Bridge Replacement Once the user has been directed to the Bridge Replacement matrix questionnaire, a series of progressive questions will help the user evaluate the applicability of ABC Construction Tech- nologies and, if applicable, which technologies best suit the renewal project criteria. Note, however, that the answers may also demonstrate that due to physical constraints of the project site, ABC Construction Technologies may not be applicable for the project. The Bridge Replacement matrix will focus on availability of staging areas at the project site. • Is a short-duration closure or detour (1 to 2 days) allowed? a. If short closures are allowed, SPMTs may be applicable. b. If short closures are not allowed, the work may have to be performed from above by using launching techniques or by use of the LTTBs. • Is there room or access nearby for a staging area on the lower level, where complete or partial assembly of the new bridge can occur? a. The use of SPMTs requires the ability to preassemble the new bridge structure in an off-line staging area, away from the project site. If there are no open areas nearby, SPMTs may not be applicable; however, the matrix directs the user to other questions that may open the possibility for other ABC Construction Technologies.

148 • Is there space adjacent to the proposed bridge location to preassemble the new structure? a. If there is a staging area adjacent to the proposed bridge location, the new bridge superstructure could be erected as a parallel activity to the substructure construction, and the bridge eventually moved into position using sliding or lateral shifting technologies. • Is the staging area accessible to the bridge site without overhead, width, or grade restrictions? a. If an off-line staging area is available, the SPMTs and their cargo must be able to travel (without interference) from the staging area to the project site. b. If there are physical restraints preventing a clear travel path between the staging area and the project site, the matrix will direct the user to other questions that may open the possibility for other ABC Construction Technologies. • Is there room behind one or both abutments for a staging area? a. If there is room behind one or both of the abutments for a staging area, launching or use of the LTTBs may be applicable. b. If there is no room behind the abutments for a staging area, at this point in the matrix, the physical constraints of the project site have severely limited the possibility of using ABC Construction Technologies. Therefore, although some of the design criteria indicate the need for ABC consideration, the matrix suggests that the user consider more conventional construction equipment and techniques. Step 3a. Matrix Questionnaire AC3: Widen Exterior Edges of a Single Bridge If the user has been directed to the Widen Exterior Edges of a Single Bridge matrix questionnaire, the series of question will further focus on criteria developed for traffic control, avail- ability of staging areas, and the proposed width of the widen- ing, and will help the user evaluate the applicability of ABC Construction Technologies and, if applicable, which tech- nologies best suit the renewal project criteria. Note, however, that the answers may also demonstrate that due to the ability to adjust the traffic flows, ABC Construction Technologies may not be required for the project. The Widen Exterior Edges of a Single Bridge matrix focuses on the traffic control, the availability of staging areas, and the width of the widening proposed for the project. • Can traffic on the bridge be reduced to allow partial closures along each edge of the bridge? a. This question will help establish the boundaries for possible erection equipment. b. If the ability to safely constrict the traffic lanes to create space along the bridge edges for the contractor to run equipment exists, ADDCs may be applicable. c. If the space along the bridge edges must be kept to a min- imum, there may not be sufficient room for construction equipment to be staged on the bridge. • Is there room behind one or both abutments for staging area(s)? a. If there is room behind one or both of the abutments for a staging area, launching or the use of LTTBs may be applicable. b. If there is no room behind the abutments for a staging area, at this point in the matrix, the physical constraints of the project site have severely limited the possibility of using ABC Construction Technologies. Therefore, although some of the design criteria indicate the need for ABC consideration, the matrix suggests that the user consider more conventional construction equipment and techniques. • Is the widening a single girder or multiple girders? a. If the widening involves multiple girders, there may be an option to preassemble major, if not all, components of the widened bridge structure. This would open opportu- nities for launching or the use of SPMTs. As a result of completing the matrix, owners and engi- neers will know whether the site and traffic constraints minimize, restrict, or promote the need for ABC Construc- tion Technologies. If ABC Construction Technologies are promoted, they will be categorized into the following set of opportunities: • Launching or the use of ADDCs or LTTBs, or explore the use of SPMTs. • Launching or the use of ADDCs or LTTBs. • Use of ADDCs or explore the use of SPMTs. • Use of only ADDCs. • Launching or the use of LTTBs, or explore the use of SPMTs. • Launching or the use of LTTBs. Step 3b. Matrix Questionnaire AC4: Widen Both Interior and Exterior Edges of Twin Bridges The logic behind the questions in this matrix match those discussed for Matrix Questionnaire AC3. If the widening on twin bridges is to only exterior edges, the results of the ques- tion set will exactly match the results of the question set of the AC3 matrix. Note, however, that due to the physical restric- tions of widening the interior edges of twin bridges, some of the potential ABC Construction Technologies available for widening exterior edges will not be applicable to widening the inside edges.

149 As a result of completing the matrix, owners and engineers will know whether the site and traffic constraints minimize, restrict, or promote the need for ABC Construction Technolo- gies. If ABC Construction Technologies are promoted, they will be categorized in the following set of opportunities: exterior eDgeS • Launching or the use of ADDCs or LTTBs, or explore the use of SPMTs. • Launching or the use of ADDCs or LTTBs. • Use of ADDCs or explore the use of SPMTs. • Use of only ADDCs. • Launching or the use of LTTBs, or explore the use of SPMTs. • Launching or the use of LTTBs. interior eDgeS • Launching or the use of ADDCs or LTTBs. • Use of only ADDCs. • Launching or the use of LTTBs. Step 3c. Matrix Questionnaire AC5: Widen Interior Edges of Twin Bridges The logic behind the questions in this matrix match those discussed for Matrix Questionnaire AC3 with the exception that, due to the physical restrictions of widening interior edges, some of the potential ABC Construction Technologies available for widening exterior edges will not be applicable to widening the interior edges. As a result of completing the matrix, owners and engineers will know whether the site and traffic constraints minimize, restrict, or promote the need for ABC Construction Tech- nologies. If ABC Construction Technologies are promoted, they will be categorized in the following set of opportunities for interior edges. • Launching or the use of ADDCs or LTTBs. • Use of only ADDCs. • Launching or the use of LTTBs. Depending on how the questions are answered, the matrix either will lead the user to the conclusion that the site con- straints restrict the applicability of ABC Construction Tech- nologies or will lead the user to one of the following set of ABC Construction Technology opportunities: Demolition of the exiSting BriDge • Demolition with short-duration closures. • Complete removal with the use of SPMTs, followed by demolition off-site. • Removal by sliding or lateral shifting, followed by demoli- tion with short-duration closures. inStallation of the neW BriDge • Launching, sliding, or lateral shifting, or the use of SPMTs or LTTBs. • Launching or the use of SPMTs or LTTBs. • Sliding or lateral shifting, or the use of SPMTs. • Use of only SPMTs. • Launching or the use of LTTBs. To test the logic of the series of ABC Construction Tech- nology Selection Matrix questionnaires, a random sample of past projects was evaluated by using the progression of questions presented in the various matrix question- naires. The results of the evaluations were then compared with the actual construction technology used to complete the project. ABC Construction Technology Design Checklists It is assumed that once owners or engineers have completed the series of ABC Construction Technology matrix question- naires, they will have been directed to a variety of nonconven- tional ABC Construction Technology options for their renewal project. It is also assumed that owners and engineers used the evaluation matrices presented in previous parts of this report to perform a concurrent, yet independent, evaluation of the ABC substructure and superstructure systems to determine structure types for their renewal project. Once the evaluations of the two matrices are complete, the results can be reviewed and the best ABC Design and ABC Construction Technology can be selected. Once an ABC Con- struction Technology has been selected, owners and engi- neers must integrate this technology into the bridge design and must consider a new set of technical issues and challenges specific to that technology. To assist owners and engineers in this design coordination process, the second goal of this part of the Phase II report on ABC Construction Technologies was to develop checklists of items that must be addressed during the design phase of a project. To be useful to owners and engineers, these lists were tailored to the specific needs of each ABC Construction Technology. Development of Standard Concept Details for ABC Construction Technologies To assist owners and engineers with the selection of an ABC Construction Technology, the third goal of this part of the Phase II report on ABC Construction Technologies was to develop a set of standard conceptual details defining termi- nology and demonstrating the possibilities and limits of each ABC Construction Technology.

150 Recommendations for Further Development of ABC Construction Technologies This section focused on the following: • Developing a matrix of questions for owners and their con- sultants to assist with the proper selection of a construction technology; • Developing a checklist of items that owners and their con- sultants must address during the design and construction phases of a project; and • Developing of a set of standard conceptual details that own- ers and their consultants could use as a guide for visualizing the possibilities and defining the limits of the four specified ABC Construction Technologies. Future development of ABC Construction Technologies could evolve around the demonstration of which technolo- gies work best with the ABC Designs (both substructure and superstructure) proposed in this report. Additional exhibits could be developed and attached with the specific ABC Designs. Automation of the selection matrix would also be a viable future development for the ABC Construction Technologies. Automation could evolve with the use of a Microsoft Visual Basic program embedded within an Excel file. The program would include a page of questions matching those listed in the matrix questionnaires and would use the selected responses to make recommendations for a construction tech- nology (whether it be conventional or one or more of the ABC Technologies). recommended Design and Construction Concepts The role of Phase II was to provide a critical evaluation of the recommended concepts and technologies from Phase I so that an incremental shortlist of technologies could be created and standardization of details and field demonstra- tions could be recommended. Standardizing ABC systems will bring about greater familiarity about ABC technologies and concepts and will also foster greater regional cooperation to achieve region-specific customization that will accommo- date regional practices and industry needs. Pre-engineered standards to be developed in this project will emulate cast- in-place construction but will be optimized for modular construction and ABC. These standards can be inserted into project plans with minimal additional design effort to adapt to project needs. Using these standardized designs will serve as a training tool to increase familiarity about ABC among engineers. Modular Bridge Systems for ABC The modular superstructure and substructure systems will be developed in this project to achieve cost and risk reduc- tions through the adoption of the guiding philosophy for all ABC concepts advanced in this project. This philosophy is stated as follows: • As light as possible 44 Simplify transportation and erection of bridge components. • As simple as possible 44 Fewer girders, splices, or bracings. • As simple to erect as possible 44 Fewer workers on site; 44 Fewer fresh concrete operations; 44 No falsework structures required; and 44 Simpler geometry. The successful use of prefabricated elements to accelerate construction requires a careful evaluation of the requirements for the bridge and an unbiased review of the total costs and benefits. This part provides a review of the engineering and constructability evaluations, pinpoints implementation chal- lenges, and provides suggestions to overcome those challenges. The recommended concepts from this evaluation have been further screened to provide a shortlist of the most promising technologies for standardization in Phase III and for use in the field demonstrations. The recommended technologies meet minimum standards of readiness for execution, suitability for ABC, and the promise of durability, economy, and value to the owner. Minimizing road closures and traffic disruptions is a key objective of ABC. For ABC systems to be viable and see greater acceptance, the savings in construction time should be clearly demonstrated. As outlined in Phase I, ABC Design concepts can be classified into three categories: • Tier 1: ABC concepts that can be completed over a weekend closure; • Tier 2: ABC concepts that can be completed in a few weeks; or • Tier 3: ABC concepts that accelerate larger projects and save weeks or months on the overall schedule. Modular bridge systems are particularly suited to be used as Tier 1 concept for weekend bridge replacements or as Tier 2 concept when the entire bridge may be scheduled to be replaced within a month with a detour to maintain traffic. Complete prefabricated modular systems and construction technologies recommended to be advanced to standard plans

151 and conceptual details in Phase III are presented under the following headings: • Precast modular abutment systems 44 Integral abutments; 44 Semi-integral abutments; and 44 Precast approach slabs. • Precast complete pier systems 44 Conventional pier bents; and 44 Straddle pier bents. • Modular superstructure systems 44 Concrete deck bulb tees; 44 Concrete deck double tees; and 44 Decked steel stringer system. • ABC bridge erection systems 44 Above-deck driven carriers; and 44 Launched temporary truss bridges. These design standards aim to overcome known obstacles to widespread adoption of ABC as a preferred method of bridge replacement. Standardizing ABC systems has many benefits that could encourage greater use of these systems in bridge renewals. Overcoming obstacles to implementing the modular superstructure and substructure systems will depend on how effective these systems are in addressing owner and contractor concerns on ABC expressed in the focus group meetings and enumerated in the Task 6 report. ABC and precast construction are viewed by local contrac- tors as projects requiring large amounts of outsourcing of work. Standardized structural systems that general contrac- tors can fabricate as much as possible themselves without having to subcontract out to specialty precasters or fabrica- tors will allow contractors to keep their crews employed and maximize profits. Perhaps the greatest impediment to increased use of ABC appears to be the higher initial costs. ABC is perceived as rais- ing the level of risk associated with a project. Standardized modular superstructure systems for ABC are aimed at increas- ing their availability through local or regional fabricators, thereby reducing costs and lead times. Standardizing also makes these systems more familiar to engineers, owners, and contractors, which will reduce complexity and the level of risk associated with the project. Contractors will be more willing to offer competitive bids on a system they have experi- ence with and that they perceive to be proven and easily con- structible. Repeated use of a standardized ABC Design also allows owners and engineers to iron out the kinks in the sys- tem through continuous improvement, leading to a more superior design than a onetime customized solution. Repeated use of these systems will also encourage contractors to provide their own suggestions about how constructability could be further improved, which will lead to further reductions in overall risks and cost. Concerns have also persisted about the durability of joints and connections in precast elements. Many earlier ABC sys- tems used grouted joints that did not prove to be very durable, especially under heavy traffic conditions. Seismic performance of precast elements and connections in seismic regions has also been an issue. With these concerns in mind, the modular superstructure systems advanced in this project have placed maximum emphasis on developing durable connection details between the prefabricated elements. Full moment connec- tions using ultra-high-performance concrete (UHPC) or high-performance concrete (HPC) are being recommended as the preferred connection details because they are strong, durable, and seismically sound. Standardizing these connec- tions with proven, easy-to-construct details will go a long way in overcoming the past concerns with performance of joints and connections. Lack of access for equipment and the need for large stag- ing areas unavailable in urban locations have been a hin- drance to large scale prefabrication. Moving complete bridges using specialized wheeled carriers requires large staging areas, which may be in short supply in congested urban areas. Modular systems allow the superstructure to be built in place using smaller components, thus overcoming mobility issues. In short, modular systems allow a more ver- satile option to ABC not limited by the space available at the bridge site. Standardized designs geared for erection using conventional erection equipment will allow repetitive use of modular superstructure systems, which will make contrac- tors more willing to invest in equipment based on certain methods of erection to speed assembly. Repetitive use will allow contractors to amortize equipment costs over several projects, which is an important component in bringing overall costs in line with conventional construction. The use of pre-engineered systems in bridge engineering is commonplace. Many states have decided to make best use of their program dollars by greatly standardizing design through the development of pre-engineered systems, plans, and so forth that encompass entire bridge systems including the quantity take-off for various standard configurations. In this environ- ment, engineering need be done only once, to create the stan- dard, so that quality assurance and quality are satisfied in the creation of an accurate standard. The greater portion of pro- gram funds can be used for construction instead of design. The use of pre-engineered bridge systems has led to low in-place constructed costs. The research team believes that a transition from pre- engineered but stick-built systems to pre-engineered and prefabricated systems is a worthy objective of this project. A number of state DOTs, such as those in Utah, Idaho, and

152 Washington State, have already developed extensive sets of pre-engineered bridges for precast construction or rapid renewal. These serve as the team’s starting point. It is not the intent of the proposed work to redo all of this work and publish many additional sets of pre-engineered bridge plans like those mentioned. This is far beyond the resources this contract provides. Rather, to the extent possible, the research team will collect, synthesize, and build upon the most useful sets of existing pre-engineered bridge plans from a variety of states and recommend modifications of their details (connections, erection methods, etc.) to make them sufficiently suitable for accelerated construction and national use. The research findings from this project, including the survey and focus group findings, lab test results, and lessons learned from the demonstration project, will provide the basis for making significant enhancements to existing ABC systems and for developing new ABC Designs that address these impediments to greater ABC use. Lab testing of UhpC Joints The durability of full-depth deck joints between prefabri- cated panels has been a major concern for many years. Unless posttensioned, these joints may allow penetration of water and chemicals leading to corrosion. Posttensioning of bridge decks to induce compression in joints, however, has tradi- tionally been a time-consuming field operation not compat- ible with ABC. UHPC was investigated as a possible solution to this problem because its high bond strength to reinforcing bars allows narrow joints, its relatively high bond strength to precast concrete may negate the need for posttensioning, and its low permeability enhances long-term durability. UHPC has been tested and implemented as a deck joint material in several joint details, including two instances in the United States—in the positive bending moment region trans- verse deck joints of Route 23 Bridge in Oneonta, New York, and in the longitudinal deck joints of the Route 31 Bridge in Lyons, New York. This work focuses on a UHPC joint appli- cation previously untested, namely in a transverse joint located in the primary negative bending region over the piers of a continuous bridge. This detail was developed for use in a three-span demonstration bridge over Keg Creek on U.S. Hwy 6 in Pottawattamie County, Iowa. The joint inves- tigated in this study is of high significance because if success- ful, it would enable this prefabricated deck construction strategy to be applied not only to single-span bridges but also to multiple-span, continuous-under-live-load bridges, which are very common throughout the United States and the world. Unlike other UHPC deck joints implemented in the past, this joint will be subjected to significantly higher levels of tensile stresses that will be oriented to be perpendicular to the joint itself. Three suites of laboratory tests were conducted to evaluate the UHPC deck joints used in the demonstration bridge. The lab tests conducted for this study include abrasion testing of the UHPC closure joint material, constructability testing of the intersecting deck joints, and strength and serviceability testing of the transverse deck joint at the pier. Posttensioning details were added to the Keg Creek transverse joints on the basis of findings of these lab tests. The testing of UHPC joints was performed by Iowa State University (ISU) and covered three primary areas of interest that were considered critical to the use of UHPC in ABC appli- cations and in the Keg Creek demonstration project. The lab tests that were conducted and their objectives follow. • Abrasion Testing (Test 1): Grinding of the UHPC closure joint material for the longitudinal and transverse joints 44 Evaluate the grindability of the cast-in-place UHPC in relation to the accelerated construction schedule. • Constructability Testing (Test 2): Placement, handling, and quality of the UHPC material at the intersecting clo- sure joints 44 Evaluate the constructability of intersecting cast-in-place UHPC joints with respect to the flow characteristics and properties of the material. 44 Qualitatively assess the feasibility of the UHPC joint placement procedure. • Serviceability and Strength Testing (Test 3): Evaluation of the strength and serviceability of the transverse bridge deck joint at the pier 44 Evaluate the negative bending performance of the module- to-module transverse connection detail at the piers. 44 Determine the cracking moment at this location. 44 Determine the ultimate moment capacity at this location. The ISU report on the UHPC testing is included in Appen- dix C and provides a detailed explanation on the testing pro- gram and findings. A summary is provided in this chapter. Abrasion Testing Abrasion testing of cast-in-place UHPC was conducted to determine the early age grindability of the material when used in the demonstration bridge. Specifically, identifying a period of time in which the contractor is able to grind the joint material without causing damage to the joints or equipment had to be identified. For the demonstration bridge it was specified that the UHPC closure joint attain 10,000 psi of compressive strength before it should be ground. This test helped determine the relative ease of grinding for this material after the 10,000 psi threshold has been reached. Experimental variables for this test included the maturity of the UHPC and the specimen surface

153 finish. Testing of the UHPC material for abrasion resistance was completed at Iowa State University in February and March 2011. Three surface finish types were tested for grindability dur- ing the abrasion testing: a rough top surface, a diamond cut surface, and a smooth formed surface, as shown in Fig- ure 3.63. The UHPC specimens were cured at 40°F, 70°F, and 100°F, and tested 2, 4, 7, and 28 days after casting. To evaluate the UHPC material for grindability, testing was completed following ASTM C944, Standard Test Method for Abrasion Resistance of Concrete or Mortar Surfaces by the Rotating- Cutter Method. Figure 3.63, a plot of the percentage of mass loss versus compressive strength for the three surface finish conditions, presents the results of the abrasion testing. Based on the compressive strength test results for the dem- onstration bridge UHPC mix design, the UHPC will reach the 10,000 psi compressive strength required for grinding in the project specifications for the demonstration bridge at approximately 2 days if cured at 70°F. The 14,000 psi com- pressive strength threshold, required in the demonstration bridge project specifications for opening the bridge to traffic, will likely be reached 4 days after placement. Thus, the con- tractor will have roughly 2 days to perform grinding of the joints: from the time the 10,000 psi threshold is reached prior to the opening of the bridge to traffic at 14,000 psi compres- sive strength. The percentage mass loss for both formed and top finishes at the 10,000 psi compressive strength threshold is approximately 0.12%. At 14,000 psi compressive strength of the UHPC mix, the percentage mass loss is approximately 0.07%. Over that 2-day duration of time, the UHPC’s resis- tance to abrasion increases by approximately 40%. That would be a significant factor for the contractor in terms of grinding time and accelerated scheduling. Figure 3.63 illustrates that the formed surface and rough surface finishes displayed the lowest abrasion resistance. Specimens with formed surface finishes exhibited lower Figure 3.63. Abrasion testing: Percentage of mass loss versus compressive strength.

154 abrasion resistance than did cut surfaces because of the steel fibers present in the UHPC. At the formed surface, the steel fibers were aligned preferentially, parallel with the surface. Thus, the fibers tended to pull off easily. The fibers lay parallel with the form surface because as the UHPC flowed along the bottom of the form, the fibers tended to align and lie flat. The rough surface finish generally also included small entrapped air bubbles that allowed for easier removal of the UHPC material. As was expected, the cut surface finish had the high- est abrasion resistance. Because the cast-in-place UHPC joints in the Project R04 demonstration bridge are to have a formed surface, the abrasion resistance in the field is expected to most nearly resemble that of the formed surface finish observed in the abrasion tests. Constructability Testing Joint constructability testing was completed to qualitatively evaluate the intersecting, cast-in-place UHPC deck joints to be used in the demonstration bridge. Specifically, a full-scale mock-up of the intersection between one longitudinal and one transverse UHPC deck joint was constructed to investi- gate issues relating to casting sequence, material mixing and placement rates, effects of ambient temperature on construc- tion, flow characteristics of the UHPC, and consolidation of material at congested locations. Testing of the UHPC joints for constructability was completed at Iowa State University in April 2011. Casting Sequence The original proposal for the construction sequence of the demonstration bridge outlined continuous placement of the entire grid of UHPC deck joints (longitudinal and trans- verse). Through discussions with the engineer, contractor, and material supplier, several logistical issues arose that chal- lenged the feasibility of full deck continuous placement. Typ- ical mixers used by Lafarge Canada for UHPC placement mix 5.11 ft3 per batch. On the job site, the mixers are used in pairs in order to provide a continuous supply of UHPC. Each batch is then discharged into buggies and transported onto the bridge to the placement location. With the large volume of UHPC necessary in the bridge deck joints, continuous placement could be achieved only by using a large number of mixers and laborers. Without employing many mixers and laborers, cold joints could potentially form in the UHPC deck joints. As an alternative, stay-in-place acrylic vertical bulkheads were proposed by Lafarge to control the location of potential cold joints. A prototype of the stay-in-place acrylic vertical bulkheads was fabricated and used during the joint constructability testing, so its performance could be evaluated. Ambient Temperature Effects on UHPC The extent of the susceptibility to variations in temperature for the workability and flow characteristics of the UHPC mix design was observed during batching of the joint con- structability test specimen and the transverse joint strength and serviceability test specimen that followed. Ambient air temperatures were steady at around 65°F at the time of batching for the intersecting joint specimen. However, dur- ing the batching for the transverse joint strength and ser- viceability specimen, ambient temperatures were 75.5°F. Without compensating for the change in ambient air tem- perature, the flow characteristics of the mixes were much different. When ambient temperatures were 65°F, the temperature of the UHPC on discharge from the mixer ranged from 82°F to 85°F for the intersecting joint specimen’s three batches. Within this range, the UHPC had acceptable flow characteristics for placement. The temperature of the UHPC on discharge from the mixer for ambient temperatures around 75.5°F was over 100°F. At this ambient temperature, the UHPC never reached its anticipated flow characteristics in the mixer, thus the batch was rejected. To correct the issue, water in the mix design was replaced by mass with ice and the UHPC material tempera- ture was reduced. This modification, the replacement of water by mass with ice, enabled extended working time and improved the flow relative to the previous batch. Flow Characteristics and Consolidation of UHPC Evaluating the flow of the UHPC around the corners at the intersection of the longitudinal and transverse deck joints was a critical aspect for this test. Adequate consolidation of the UHPC in the joint cross section around steel reinforcement is important to the deck joint performance. During UHPC placement, when the final mix temperature was limited to a maximum of 85°F, the UHPC material appeared to have ade- quate flow characteristics to achieve good consolidation and flow around corners at the intersections of longitudinal and transverse joints. After the specimen was cured and removed from the forms, it was cut into several sections to examine consolidation and potential cold joints. Upon investigation of the cut specimen, no significant voids around steel reinforcing bars were observed. The test also validated the use of top forms and chimneys at the high end of the 2% cross slope at transverse joints. The top forms were applied sequentially as the joint was filled from the lowest elevation to the highest. The chim- neys provide additional hydrostatic head in the freshly placed UHPC to aid in consolidation within the joint. It was sug- gested that top forms and chimneys be used in the demonstra- tion bridge.

155 Joint Intersection Detail Recommendations Final inspection of the specimen upon removal from the forms allowed for additional observations and recommenda- tions. The proposed stay-in-place acrylic bulkhead success- fully allowed for sequential placement of the UHPC, but also created a possible infiltration plane where water and chemi- cals could access the embedded steel joint reinforcement. The reinforcement cage and the finished product are shown in Figure 3.64. To maintain sequential placement of UHPC in the deck joint grid and avoid possible infiltration planes, a detail for a partial-height, removable acrylic bulkhead was developed and suggested for use in the demonstration bridge, as shown in Figure 3.65. Removable acrylic bulkheads should be used in the longi- tudinal joint in compression zones where possible. Placing the bulkheads at these locations will provide better continuity at the interface between the hardened and freshly placed UHPC, which will help prevent the ingress of water and other chemicals. Transverse Joint Strength and Serviceability Testing The module-to-module transverse connection used in the SHRP 2 ABC demonstration bridge was a unique and critical detail that had never been implemented in a bridge nor tested to quantify structural performance, as shown in Figure 3.66. Strength and serviceability testing of the module-to-module transverse connection was performed to evaluate the negative bending performance of this detail over the piers, determine its cracking moment, and verify the ultimate moment capacity. Testing of the module-to-module transverse connection was completed at Iowa State University from July to October 2011. Service-Level Static Test Load testing through live-load Service Level II moment was completed on the specimen. Loading was completed at 5,000-lb increments in order to complete visual inspection of the speci- men and to check for the appearance of cracks and accrual of damage. Strain levels were monitored with the embedded and surface-mounted strain gauges located throughout the speci- men. Strain levels for surface-mounted strain gauges at loca- tions that spanned the HPC/UHPC interface exceeded 110µe, the HPC cracking strain, at approximately halfway to Service Level I moment, as shown in Figure 3.67. The disparity between immediately adjacent gauges and the strains register- ing in excess of the HPC cracking strain across the interface suggested debonding and an opening at the interface between the precast HPC deck and the UHPC joint. This raised con- cern and called the bond strength of the UHPC to the precast concrete into question. Visual inspection of the joint interface at Service Level II confirmed the debonding and substantial opening of the Figure 3.64. Joint intersection form (top) and joint intersection specimen (bottom). Figure 3.65. Removable acrylic bulkhead.

156 interface that was suggested in the strain gauge data, as shown in Figure 3.68. Later, inspection during fatigue testing further confirmed the interfacial debonding and opening occurring below service-level conditions. In addition to joint interface debonding and substantial opening, strain levels in the embedded strain gauges also registered above the expected HPC cracking strain prior to reaching Service Level I moment conditions, as shown in Fig- ure 3.69. In the top-of-deck reinforcement, maximum strains of 540µe, 550µe, and 475µe were recorded in gauges S1-1-1T, S2-2-2T, and S2-3-2T, respectively. Strains in the UHPC joint were relatively lower in the top-of-deck, not exceeding 160µe, which is below the expected UHPC cracking strain level of 250µe. Nearly all gauges located at the termination of the joint hairpin bar registered strain levels exceeding 110µe prior to the Service Level II conditions. These data suggested cracking in the prefabricated HPC deck modules under service-level loading. Cracking was not visually confirmed near the joint in the HPC deck during the incremental static loading, but opening and closing of the cracks during cyclic loading made cracking in the HPC clearly visible. The prevalence of the high strains at the termination of the hairpin reinforcement in the top-of-deck suggests cracking of the HPC would be expected for the demonstration bridge, indicating that the transverse connection detail might not satisfy the original project aim to avoid cracking in the deck over the pier. After the static tests were completed, fatigue testing com- menced. Fatigue tests consisted of loading the specimen through the full service-level moment range for 1,000,000 cycles. The embedded strain results for the fatigue testing generally resembled those from the static testing. Similarly, the gauges near the interface consistently exhibited the highest strains while the gauges within the UHPC registered the lowest in each of the instrumentation rows. Some higher strain levels at 1,000,000 cycles when compared to the static testing results suggested propagation of cracking and damage accrual within the specimen. On inspection at 250,000 cycles, cracks were identified in the precast deck. At 500,000, 750,000, and 1,000,000 cycles, further visual inspection confirmed prop- agation of the existing cracks and formation of new full- depth cracks in the precast deck panels up to 10 ft away from the joint. To mitigate the serious durability concerns that arose with respect to the early debonding of the UHPC to precast inter- face and cracking in the HPC deck panels, a modified detail, Figure 3.66. Module-to-module transverse connection detail.

157 as shown in Figure 3.70, was devised to posttension the deck in this region and minimize tensile stresses in the concrete through Service Level II moments without compromising the accelerated construction aspect of the project. The transverse module-to-module connection detail was modified to include high-strength steel rods mounted just under the deck surface to posttension the entire joint region. The retrofit detail was tested through the full range of service- level moments with a 60-kip posttensioning force per rod and again with a 70-kip posttensioning force per rod. The 60-kip posttensioning force in each of the rods reduced tensile strain across the joint interface such that the HPC cracking strain was not reached until Service Level I conditions. However, strains did exceed the HPC cracking strain in the top-of-deck embedded gauges before reaching Service Level II. Figure 3.67. Selected surface-mounted strain gauges adjacent to the joint interface. Figure 3.68. Joint interface opening.

158 Figure 3.69. Row 1, top-of-deck embedded strain gauges (static). Figure 3.70. Connection retrofit detail.

159 By contrast, applying 70 kips posttensioning force in each of the rods minimized or negated the tensile strain across the interface entirely when loaded to Service Level I. All surface- mounted strain gauges spanning the interface registered below the HPC cracking strain until after the Service Level I condi- tions were exceeded, as shown in Figure 3.71. Tensile strain data across the interface revealed a maximum 29µe at Service Level I moment. All embedded strain gauges, top- and bottom- of-deck, did not exceed 110µe until Service Level II condi- tions were applied. The 70 kip per rod posttensioning force was recommended for application in the SHRP 2 Project R04 demonstration bridge. Ultimate Capacity Test Upon completion of static testing for the modified detail, the posttensioning rods were removed and the transverse module- to-module connection detail was tested to ultimate moment capacity. Strain data for the embedded gauges were analyzed in combination with qualitative observations to determine the failure mechanism for the transverse module-to-module connection detail. As loading incrementally increased, the opening at the interface between the HPC deck and the UHPC joint wid- ened. In addition, cracks from service-level testing propagated and widened throughout the precast deck, as shown in Fig- ure 3.72. As the specimen was pushed well beyond service-level moments, reinforcement in the HPC deck near the UHPC interface began to yield. Eventually, the moment-displacement curve entered into the nonlinear region, and reinforcement near the joint began to deform plastically. The two large cracks parallel to the joint interface continued to widen, and eventu- ally the UHPC suffered tensile rupture near the shear studs located in the joint, as shown in Figure 3.73. Spalling at the edges of the precast deck exposed the outer- most reinforcement hairpins that entered into the joint, allow- ing for pullout. Load application continued, and the specimen reached a peak moment of 2,239 kip-ft before successive Figure 3.71. Top-of-deck surface-mounted strain gauges over interface (70-kip retrofit).

160 fractures of multiple hairpin reinforcement bars acted as the ultimate mode of failure for the transverse connection. UHPC Testing Conclusions The durability of full-depth deck joints between prefabricated panels has been a major concern for many years. Unless post- tensioned, these joints may allow penetration of water and chemicals, leading to corrosion. Posttensioning of bridge decks, however, has traditionally been an additional field operation requiring a specialty subcontractor. UHPC was investigated as possible solution to this problem because its high bond strength to reinforcing bars allows narrow joints, its relatively high bond strength to precast concrete may negate the need for posttensioning, and its low permeability enhances long- term durability. Figure 3.72. Interface opening and crack propagation. Figure 3.73. UHPC rupture (top- and bottom-of-deck). Through the project’s three suites of laboratory tests, the UHPC deck joints were evaluated for use in the ABC dem- onstration bridge. Abrasion testing was completed to assess the abrasion resistance of the cast-in-place deck joints with respect to anticipated grinding operations, a constructability test was carried out to assess the placement procedure and feasibility of the longitudinal and transverse UHPC joint intersection detail, and strength and serviceability testing were completed to quantify the cracking moment and ulti- mate moment capacity of the transverse module-to-module connection detail over the bridge pier. Abrasion and maturity testing of the UHPC material indi- cated that when cured at 70°F, the compressive strength thresh- olds required for grinding (10 ksi) and opening the bridge for traffic (14 ksi) were reached at 2 and 4 days after placement of the UHPC, respectively. Thus, the contractor would have 2 days before grinding could commence and 2 days to com- plete grinding prior to reopening the bridge to traffic. Abrasion resistance increased by roughly 40% for the UHPC material from 2 to 4 days, emphasizing the advantages in time and equipment to grinding the joints as early as possible. A UHPC placement procedure for the demonstration bridge based on the findings of the constructability testing for the longitudinal and transverse joint intersection detail was discussed with the contractor and material supplier. Partial- height removable bulkheads were recommended in order to control the placement of the UHPC in the deck joints. In addi- tion, the sensitivity of the UHPC mix design to ambient air temperatures was identified while batching for the laboratory tests. Provided that the UHPC’s sensitivity to ambient tem- perature effects were accounted for, the UHPC exhibited excellent flow characteristics and consolidation during place- ment of the intersecting deck joints. In addition, the acceler- ated rate of compressive strength gain and higher cracking

161 strain level relative to regular concrete proved useful for application in this ABC project. While the UHPC displayed several superior material char- acteristics with respect to the strength of the deck joints them- selves, the direct tensile bond strength between the UHPC and the precast HPC deck observed during the strength and serviceability testing raised a durability concern. Testing revealed that the interface between the transverse UHPC joint and the HPC deck underwent early debonding and sig- nificant opening well below service-level moment conditions. This raised concerns as to the durability of the module-to- module transverse joint connection for the demonstration bridge. Consequently, a posttensioned retrofit detail was devel- oped and tested in an effort to eliminate opening at the inter- face and cracking in the HPC deck immediately adjacent to the transverse joint over the pier. With adequate posttension- ing force, the retrofit detail successfully limited strains levels to below the cracking threshold of the HPC. For this design using steel girders, the posttensioning retrofit was a successful solution even within the constraints of the accelerated con- struction schedule. However, if prestressed concrete girders had been used, posttensioning might be more difficult to apply within an accelerated schedule. The strength of the transverse module-to-module connection detail was found to be more than adequate. However, due to the interfacial bond issues observed over the course of this testing, further investigation into the direct tensile bond strength between the UHPC and HPC is recommended. This testing would help to better evaluate the durability of the longitudinal and transverse UHPC deck joints present in the ABC demon- stration bridge and help to determine the long-term viability of this UHPC deck joint detail as a solution in future ABC projects. (See Appendix C for the UHPC Testing Report.) Field Demonstration project Phase III of the project required the construction of a dem- onstration bridge that used the most-promising bridge details identified earlier in the research and modular systems being incorporated into ABC standards. The US-6 bridge, which crosses Keg Creek near Council Bluffs, Iowa, is representative in size and length to a large majority of bridges across the United States. This bridge was replaced as a demonstration bridge that incorporates proven ABC bridge construction details with the innovative use of ultra-high-performance concrete (UHPC) to shorten the normal bridge replacement period of 6 months to only 2 weeks of traffic disruption. The improvements consist of replacing the bridge located on US-6 over Keg Creek in Pottawattamie County, Iowa. The existing 180-ft by 28-ft continuous concrete girder bridge (with spans of 81 ft, 48 ft, and 81 ft) was constructed in 1953 and was classified as structurally deficient with a sufficiency rating of 33. The replacement structure is a three-span (67 ft, 3 in.; 70 ft, 0 in.; 67 ft, 3 in.) composite steel modular bridge, 210 ft, 2 in. by 47 ft, 2 in., with precast substructures and pre- cast bridge approaches. The bridge replacement is intended to increase the structural capacity of the bridge, improve roadway conditions, and enhance safety by providing a wider roadway. The original and new Keg Creek bridges are shown in Figures 3.74 and 3.75, respectively. This application provided a unique opportunity to effec- tively promote ABC for rapid renewal of the bridge infrastruc- ture and to demonstrate various ABC technologies being advanced in the R04 project. The steel modular option was chosen as the most cost-effective on the basis of early discus- sions with local contractors and fabricators. Although it will not be fully detailed on the design plans, the contractor was allowed to propose a precast concrete modular alternative under a value engineering option if it could be constructed within the same ABC schedule and at a lower cost—none was Figure 3.74. Original Keg Creek Bridge. Figure 3.75. New Keg Creek Bridge.

162 proposed. The bridge was originally designed in-house to be constructed with a planned 13-mi detour (ADT = 4,000) with an estimated construction duration of 6 months. HNTB rede- signed this bridge by using ABC techniques and standard designs so that the replacement could be completed in a 14-day period. The ABC period of 14 days pertains only to the time that traffic was disrupted. Although the total duration for the project, including time for prefabrication, was about 7 months, the traveling public was affected for only just over 2 weeks. A daylong workshop, including a site visit, provided an ideal opportunity to disseminate information to bridge owners from around the country. The demonstration bridge features precast concrete semi- integral abutments, precast columns and pier caps connected with high-strength grouted couplers, and an innovative mod- ular superstructure constructed using prefabricated concrete decked steel stringer units and field-cast UHPC joints. The enhanced durability provided by the elimination of all open deck joints is seen as a major advance in long-life ABC projects and the assembly of precast units without the need for any posttensioned connections avoids the need for specialized contractors. The project was the first in the United States to use ultra-high- performance concrete (UHPC) to provide a full, moment- resisting transverse joint at the piers. This detail allowed the prefabricated superstructure elements to be erected as a sim- ple span and, once the UHPC joints were constructed, to per- form as continuous joints. The project team performed full-scale laboratory testing of the critical field-cast UHPC continuity joints to ensure their long-term reliability and ultimate load capacity. These UHPC joints provided simple construction, additional load-carrying capacity, and a dura- ble joint that prevents moisture intrusion and long-term maintenance problems. Demonstration Project Innovative Features This demonstration project implemented a series of innova- tions. It incorporated details drawn from diverse locations and applied them in a single demonstration project that was visited by DOT and FHWA personnel from numerous states. Project innovations included the following: • Overall, a complete bridge system was designed and con- structed using superstructure and substructure systems composed of prefabricated elements. The bridge approach slab also consisted of precast elements. • Superstructure units that incorporate precast suspended backwall elements created a semi-integral abutment. • Ultra-high-performance concrete was used in the joints between the modular superstructure units and between the approach slab panels. UHPC was used for longitudinal joints and transverse joints over the piers. This project was the first in the United States to use UHPC to provide a full, moment-resisting transverse joint at the piers. The elimi- nation of open deck joints provides for a more durable, low-maintenance structure in the final condition. • Self-consolidating concrete (SCC) was used to improve consolidation and increase the speed of construction for abutment piles (fill pockets) and abutment-to-wingwall connections. Abutments consisted of prismatic precast concrete elements that feature a series of open holes that accommodated driven steel H-piles. • Use of fully contained flooded backfill at abutments: This proven construction method ideally suited for ABC involves placing a granular wedge behind the abutment backwall that is flooded to achieve early consolidation and signifi- cantly reduce the potential for formation of voids beneath the approach pavement. • A structural health monitoring system (HMS): A monitor- ing plan was implemented to evaluate and document the innovative aspects of accelerated construction. The moni- toring plan included health monitoring instruments to assess the integrity of the structure and deck panel system during and after construction. • ABC entails prefabricating as many of the bridge compo- nents as feasible considering site and transportation con- straints. This project takes the approach that for ABC to be successful, ABC Designs should allow maximum opportu- nities for general contractors to do their own precasting at a staging area adjacent to the project site or in their yards with their own crews. The components were designed such that a local contractor could self-perform all the precasting work without outsourcing much work to precasters. The winning bidder chose to do that by leasing a temporary casting yard next to the bridge site. • The technologies incorporated into this bridge project have been successfully used in constructed projects drawn from around the United States. The fact that several diverse structural systems have been assembled and incorporated into a single project reinforces the concept that innovation does not necessarily mean creating something completely new, but rather facilitating incremental improvements in a number of specific bridge details to fully leverage previ- ously successful work. This demonstration project can affect the future practices of the industry and the DOTs. New technologies that are implemented successfully on this project will accelerate the adoption of the innovations in the United States. This will be accomplished by educating and creating awareness related to the innovative features, which will increase confidence in rec- ommending their use on other projects.

163 Demonstration Project Construction The construction letting for the project was held on Febru- ary 15, 2011. A total of seven fully responsive bids were received on the Keg Creek Bridge project, with a low bid of $2.65 mil- lion submitted by Godbersen–Smith Construction (G-S) of Ida Grove, Iowa. During the 14-day ABC period, the contractor would be subject to liquidated damages at a rate of $22,000 per day. The project was re-opened to traffic on November 1, 2011. The 14-day ABC period occurred between October 17, 2011, and November 1, 2011. A detailed account of the construction process for the bridge, including the prebid meeting, contractor bids, the site preparation and prefabrication, the construction process for the bridge including work prior to the ABC period, and the post-construction review meeting, is included in Appendix D. Demonstration Project Lessons Learned Overall, the Keg Creek Bridge project was a tremendous suc- cess. The bridge was completely replaced in 16 days by using only conventional equipment and labor and without signifi- cant problems. All parties (owner, designer, and contractor) worked closely together to resolve challenges as they arose during the ABC period. A SHRP 2 R04 representative was on site during the ABC period to make immediate decisions when questions arose. This was a critical component to the overall success. Following the post-construction review meet- ing, a number of lessons learned were identified. They include the following: • On-site prefabrication of bridge components can be per- formed by contractors and result in a high-quality prod- uct. On-site inspection staff should be prepared for work that is not exactly like their normal projects. • On-site mixing and placement of trial batches of UHPC should be considered to help eliminate fiber balling issues. Early and proactive communication with the UHPC provider is critical to the success of on-site placement operations. • Special provisions for projects should be carefully written to provide for both on-site and more-traditional precast concrete operations. The special provisions should describe casting, quality assurance, and inspection. • The bond between UHPC and conventional precast concrete is critical. Surface preparation prior to placement of UHPC should be performed per the manufacturer’s recommenda- tions. Future direct-tension testing of bond specimens at ISU will be beneficial in understanding this condition. • Field placement of UHPC in large quantities can be chal- lenging to manage. For future projects, cold joint bulkheads should be strategically placed to manage the UHPC pours efficiently. It might also be beneficial to separate the UHPC pour in the suspended backwall from the slab joint pour. • Joint reinforcement using hairpin bars should be carefully evaluated for future projects. It may be possible to simplify this joint construction with reinforcement details that would allow these joints to be more easily constructed. Bars should be staggered, and projecting bars potentially shortened if possible. • Joint reinforcement congestion should be carefully evalu- ated for future projects. It may be possible to reduce the number of longitudinal bars that would allow these joints to be more easily constructed. Bars crossing at the joint intersections create congestion and time-consuming place- ment methods. • Surveying is a critical element of fast-track bridge replace- ment projects. To avoid critical and time-consuming errors, two sets of independent surveys should be used to verify accurate pile driving and foundation placement during the ABC period. • The importance of following project special provisions and supplier-directed UHPC on-site placement protocols to ensure good performance and long-term durability should be stressed in contract documents and contractor communications. All concrete surfaces in contact with the UHPC shall be cleaned and coated with an approved epoxy bonding agent. Both longitudinal and transverse joints are to be coated. • Consider providing additional isometric views to plans to allow contractor and inspection personnel to better under- stand how the bridge components fit together. • Although one was not needed on this project, the contrac- tor should have a backup plan in the event that a bridge component is damaged during the ABC period. At the very least, a repair plan should be agreed on in advance. • Ideally, the designer should be present on site during the ABC period for quick decision making. On the basis of lessons learned from the Keg Creek Bridge demonstration project, several adjustments were incorporated into the proposed ABC standards presented elsewhere in this report. These adjustments were intended to improve constructability and reliability, and to provide improved opportunities for future plant-cast and site-cast ABC projects. Highways for LIFE Workshop To more widely disseminate the construction process and les- sons learned from the Keg Creek Bridge demonstration project, a national Highways for LIFE showcase was held in Council Bluffs, Iowa, on October 28, 2011. The showcase was attended by nearly 80 people from 14 states. These participants

164 represented state DOTs, the FHWA, designers, and contractors who shared an interest in accelerated bridge construction. The showcase agenda included presentations from a vari- ety of viewpoints. Topics discussed included an overview of the Highways for LIFE program, both national and Iowa per- spectives on accelerated bridge construction, and a detailed presentation on the design and construction of the Keg Creek Bridge. Showcase attendees were encouraged to visit the proj- ect site during the afternoon. Bus transportation was pro- vided to the project site, and attendees were allowed to freely observe all aspects of the construction progress. Weather was cool and windy, with temperatures in the low 50s in the after- noon. On the day of the showcase, the contractor was placing the UHPC material for all of the superstructure deck joints, an ideal presentation for the attendees. Visitors were able to observe the mixing, transporting, placing, finishing, and cur- ing operations. Video A time-lapse video of the ABC period for Keg Creek Bridge and a longer, narrated video that captures each stage of the project and tells the story of how innovative technologies can be inte- grated into the bridge replacement process can be viewed at http://www.trb.org/StrategicHighwayResearchProgram2 SHRP2/Pages/Keg_Creek_Bridge_Project_619.aspx.