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.

195 ABC Case Studies Task 1 of this project included a comprehensive review of published and unpublished literature. Using various online databases to access published sources and relying on various outreach efforts to solicit unpublished information, the team gathered and collected the literature data throughout Phase I. Representative examples of various types of projects and ABC solutions are presented in this appendix to illustrate the broad range of solutions available to owners, designers, and contractors. This appendix focuses heavily on case studies of prior proj- ects, with specific focus on the accelerated bridge construc- tion (ABC) aspects of the projects; provides information on ongoing research that has not yet resulted in field applica- tions; and provides some information on contractual and planning aspects of ABC projects. For the case studies, exam- ples of different types of projects are presented to highlight the variety of ABC solutions available. They are generally grouped into projects that executed ABC through prefabrica- tion and those that used innovative construction techniques or structure movement as the core of the solution. Many projects increasingly involve both of these components, so a strict division was not always possible. The following discussion provides a general overview of the material to be presented and then, by using a series of illustrative projects, describes various innovations. Although this section presents the material as a literature review, the team has in many instances written a short synopsis on how a particular concept, technology, method, etc., can be applied to future ABC projects. Many of these reviews should thus be considered as concepts for implementation. The literature review clearly demonstrates that through repetition of and improvement on existing ideas, dramatic progress can be made. Owners and designers should consider the following material a starting point for future improvements and greater implementation. Standard Bridge Systems This section addresses the development of standard bridge sys- tems. By standard bridge systems, the report refers to work that attempts to look at the problem or opportunity of ABC from the standpoint of integrating various components into a cohe- sive model. Several agencies have completed work in this area, and their reports are summarized here. However, one synthesis study (Shahawy, 2003) is particularly noteworthy; it provides a wealth of pertinent references and information related to pre- fabricated bridge elements and systems for ABC applications. Notably, one effort is not specifically discussed because it is a work in progress, and that is the development of a series of ABC standards by the State of Utah. (Note that only the cre- ation of standards in the State of Utah is a work in progress; the overall ABC initiative in Utah is significant, as evidenced by the number of case studies from Utah contained throughout this appendix.) Utah DOT, through a series of workshops involving local and national experts, contractors, consultants, and state and FHWA personnel, has embarked on a substantial effort to make ABC standard practice. That goal will partly be accom- plished through the development of standards and compo- nents that work together in a systems approach to the problem. Users and owners should look to the future completion of this work as one of the starting points for ABC deployment. Other states can and likely will transition their standard practices to ABC (or at least allow it as an option) through modifications to their standard drawings. States such as Iowa, Texas, Washington, and others have extensive sets of standard drawings in presentation or near presentation quality, but these sheets are typically detailed for conventional construc- tion. Re-detailing a number of these standards for ABC imple- mentation is a fairly simple matter. For instance, the Iowa DOT has completely designed and detailed piers and abutments for various standard roadway widths and span lengths as part of its standard bridge plans. Piers consisting of hammerhead and A p p e n d i x A

196 multi-column bents are fully detailed and designed—but for cast-in-place construction. Simple alternative detailing of the column/footing joint and column/cap joint would make these plans immediately useful for precast ABC substructure con- struction. Similarly, the Iowa DOT has fully detailed slab designs based on the standard cross sections for common roadway widths. Simply replacing the cast-in-place deck detailing with a standard full-depth precast deck and adding the required connection pocket and posttensioning details (if used) could convert these plans for ABC superstructure use. Certainly, other states could do the same. Several illustrative project summaries are presented on the following pages. Project Title: Guidelines for Accelerated Bridge Construc- tion Using Precast/Prestressed Concrete Components Citation: Precast/Prestressed Concrete Institute Northeast Technical Committee, 2006 ABC Design Features: Complete construction of bridges by using precast concrete components ABC Construction Features: Rapid construction of typical small and medium-size bridges that use common materials and equipment Project Description The PCI Northeast Bridge Technical Committee worked col- lectively with engineering, producer, and DOT members to develop a design and construction guidance document for providing completely precast/prestressed/prefabricated bridges for rapid erection. This appears to be the first and, as yet, only comprehensive document that addresses bridges as a complete product. As Figure A.1. indicates, the manual provides guidance on all elements of traditional bridge construction, including details for prefabricated superstructure elements such as full- depth deck panels, various details for common bridge piers such as multi-column bents, and more uniquely, details for prefabricated abutments. Connections between the various subsystem components are provided. Most innovative are the abutment details including various grouted void details for attaching the abutment shells to a pile group. A portion of the manual is related to connection design, specifically the promoted use of grouted sleeve splice connec- tors to join abutting elements. The use of grouted splice sleeves allows for what is known as “emulative design,” in which the member is designed as if the connection is not present. These connections are easily executed in the field and develop, as a minimum, 125% of the yield strength of the rebar. Some commercial products can resist a force of 150% of the strength of the rebar. ABC Opportunity Agencies looking to develop ABC standards for typical bridges should examine the PCINE recommendations in detail. They provide simple and tested connections to bond various pre- cast concrete elements. They also provide case studies that demonstrate the use of these products in several bridge recon- struction projects. Project Title: Alabama Precast Bridge System Citation: Fouad, Rizk, and Stafford, 2006 ABC Design Features: Creation of complete integrated pre- cast bridge system ABC Construction Features: Rapid construction of com- plete bridge systems by using precast elements and con- ventional equipment Project Description In a study by the University of Alabama, Fouad et al. (2006) discuss the development of an entirely precast bridge system for the State of Alabama. The study focused on all elements of a traditional bridge, specifically deck panels, superstruc- ture elements, bent caps, columns, solid piers, and abutment configurations. The intent was to create an integrated precast bridge system that could be constructed rapidly, presumably by local forces and contractors. The final result consists of bulb tee beams, a rectangular bent cap, hollow column sec- tions, and precast abutment sections. In developing the final products, the team was careful to use only commonly available materials and design strengths and to specify that all connections between precast ele- ments be made with the commonly available NMB splice sleeve grouted coupler. [Note that this same coupler has been used for decades in the precast building and bridge market and is currently being advocated by others in the transportation industry as a standard method for connect- ing precast elements through a process known as emulative design.] The following limits were adopted for design of the standard bridge system: • Girder weight, not to exceed 160 kips nor a length of 130 ft; • Bent cap weight, not to exceed 150 kips; for example, a cap 5 ft (h) by 4.5 ft (w) by 45 ft (l); • Maximum column weight, 100 kips, and maximum slen- derness, kl/r = 100 (to preclude second order analysis); and • Standard roadway widths of 40 ft and 44 ft. The girder section chosen is a deck bulb tee beam, pat- terned after the section in the PCI Bridge Design Manual. The top flange varies; either 85- or 93-in.-wide flanges are used. The section depth varies; 36-, 50-, or 64-in.-deep sections are

197 proposed. This gives a span range capability of 40 ft to 130 ft with three sections. The beams are connected in the field with welded flange connections and covered with a concrete topping to obtain the required ride profile and additional strength. For the bridge columns, rectangular sections with an inner void were chosen to limit weight. The end sections near the connections are solid. This design is similar in many regards to pretensioned box beams with solid end sections. Based on both the weight and kl/r requirements for the column, limit- ing maximum lengths were established for columns, includ- ing 36-, 42-, 48-, and 54-in.-square sections. These columns could typically have heights of 45 ft to 55 ft, generally tall enough for a large variety of bridges. Figure A.1. PCINE modular bridge system.

198 The precast caps were designed to be used with a two- column bent and the roadway widths described previously. The caps are connected to the pier columns with the grouted sleeve couplers also described previously. They are made of conventionally reinforced concrete. Details are provided to modify the bent cap for attachment to multiple piles as would commonly be used for a pile bent. Abutment details allowing for the use of spread footings, driven pile, or drilled shaft foundations were also developed in concept. In addition to the design details, shipping and handling details were provided for the various elements. Borrowing from the traditional pick point diagrams com- monly used in the prestressed concrete industry and found in sources such as PCI design manuals, stripping, lifting, and bracing details for various elements were provided. Sev- eral example structures were designed and partly detailed in the Alabama study to illustrate application of the design and detailing recommendations. ABC Opportunity This project and its findings can be immediately implemented at the state level or modified by others to suit their local prac- tices. The project illustrates the development of a cohesive system concept of entirely precast elements with weights, dimensions, and levels of innovation that will not inhibit their use by local contractors. Some of the components can be assembled ahead of time for rapid movement; but even in the basic configuration, erection of a completed bridge is possible in days once the foundation construction is completed or if portions of the existing foundations can be reused. Project Title: Washington State DOT Prestressed Bridges for High Seismic Regions Citation: Hieber et al., 2005a; Wacker et al., 2005; Hieber et al., 2005b; Khaleghi, 2005; Khaleghi, 2008 ABC Design Features: Systematic development and evaluation of precast systems for high-seismic ABC environments ABC Construction Features: Use of fully precast pier and superstructure systems erected with conventional equipment Project Description The Washington State DOT has been sponsoring research in the area of ABC applications for precast/prestressed structural con- crete for more than a decade. With a large amount of research in the area of precast/prestressed elements for ABC application, Washington State DOT was looking to evaluate the viability of these systems for use in the high-seismic regions of western Washington State. A series of studies evaluated existing structural systems, commented on their merits and disadvan- tages, and then developed new concepts suitable for high-seis- mic use and rapid erection for ABC applications. In Hieber et al. (2005a), which appears to be the first of a series of incremental reports in the use of precast elements for ABC construction, the authors evaluate various ABC concepts, including the use of full-depth deck panels, partial-depth stay- in-place form panels, prestressed concrete multi-beam super- structures (i.e., double-T systems), prefabricated composite units, and precast concrete pier units. For each type of system evaluated, the authors identified various key issues that repre- sented long-term durability concern or possible constructabil- ity issues, or certain aspects that were troubling or unique from a seismic design perspective. A brief description of their find- ings follows for some of the systems evaluated. Full-depth deck panels. The full-depth deck panel was studied and determined to be an attractive system for rapid deck instal- lation (or redecking of existing bridges) even though the authors acknowledge that such systems have an inconsistent performance history. The study examined connection details between the panels and girders, the joints between panels, required prestressing, and wearing surface issues. The panels were recommended for seismic area use despite the lack of complete confidence in the water tightness and durability of the transverse joints and some question about the diaphragm action capabilities of the deck in resisting seismic loading. Multi-beam superstructure. These structures are composed of systems such as double-T, box beam, inverted channel, and deck bulb tee sections placed adjacent to one another and attached via grouted joints, shear keys, mechanical fas- teners, or transverse posttensioning—or sometimes a combi- nation of methods. The evaluations concluded that all such systems are viable and have performed well. Concern cen- tered on the durability of the longitudinal joints (grouted or mechanically connected) and establishing a durable connec- tion for higher average daily traffic locations. Other concerns pertained to constructability, such as camber match issues between adjacent beams and attaining a smooth riding sur- face as a consequence. None of these is a substantial impedi- ment to the use of the system. Again, adequate diaphragm resistance must be achieved in spite of the many longitudinal joints in the finished bridge. Precast pier systems. The most extensive portion of the study by Hieber et al. (2005a) focuses on the potential use of precast substructure units for bridge pier construction in high-seismic regions. Although bridge superstructures tradi- tionally have some degree of vulnerability, the nature of seis- mic loading is that the controlling demands are concentrated in the substructures, more specifically in the bridge piers. This study focuses on the development by others of various precast pier system concepts and the implications of seismic loading on the various concepts.

199 Given the possibility of high-demand seismic loading, the authors identified various critical issues that would need special attention in the use of prefabricated columns. These issues include the development of a connection between the footing and precast segmental column segments, connection between segments, connection between segments and the cap beam, connection of cap beams together for a wide frame, and weight/handling concerns. The report discusses details for connecting the base segment to the footing, including the options to use multiple staged pours to anchor the footing and base segment and variations of dowel bar connections between the two elements. Regarding joints between column segments, the authors advocate the use of match cast segments to provide the proper dimensional tol- erance and fit for the columns. These segments must be pre- stressed enough that the transverse joints do not open under lateral loading. A high degree of prestress introduces a high compression in the concrete and a high initial strain in the tendons, both of which lead to limited ductility. The possible use of large tendons at a lower stress is discussed as a solu- tion. For the joint between the column and cap, details such as individual dowels in preformed holes, posttensioned connec- tions, and the use of headed bars to develop the connection strength in a shallow cap depth are all considered and dis- cussed briefly. Weight limitations were discussed with local fabricators and contractors in Washington State, resulting in a range of acceptable weights for precast elements listed as 120 kips to 180 kips. On the basis of only those weights, the authors computed the length and diameter of columns that would be achievable in these conditions. When kl/r consider- ations are also used to control column slenderness, the range in heights for a pre fabricated piece meeting a limit k/r = 100 and the 120-kip or 180-kip limit is such that a 3-ft-diameter column could have a maximum unsupported length of 41 ft, and a 6-ft column could have an unsupported length of as much as 75 ft. The study recommends that precast segmental column con- cepts be considered for column design in high-seismic regions. The authors suggest that unbonded posttensioning in con- junction with mild steel bridging the joints between segments would be the most effective total solution. The unbonded post- tensioning allows for a ductile behavior during seismic load- ing and creates a restoring force to recenter the column. In summary, precast/prestressed elements could have many uses in a high-seismic region such as Washington State. Many other such areas exist around the United States, the obvious being California; but locations such as St. Louis, Missouri; Memphis, Tennessee; Charleston, South Carolina; Salt Lake City, Nevada, and others are challenged to provide a valid seismic design for all projects as well. The precast solution has the promise of being standardized. Hieber et al. (2005b) conducted a detailed analytical study. The researchers used nonlinear pushover analyses and sim- ulated earthquake loading to evaluate the performance of Figure A.2. Washington State DOT precast bridge system.

200 precast pier systems constructed with reinforced concrete connections designed by using an emulation design concept and a second pier system connecting the pier cap to the col- umn by using a posttensioning system. The results of the para- metric study indicated that acceptable performance could be expected for both a 10% in 50 years and a 2% in 50 years earthquake event. No experimental work was conducted— only a series of analyses was done. The researchers recom- mend that additional physical testing be conducted, that further analysis of the joint regions be done, that construc- tability again be evaluated, and that some consideration of post-earthquake repair be part of further studies. Wacker et al. (2005) focused on developing displacement and force-based design procedures that do not require nonlinear analysis and that could be used in a design office. The focus again is on the emulation design and hybrid posttensioned con- nections studied first in the Hieber report (2005b). The study first looked at developing an equivalent lateral force design approach to pier design—analogous to the design approaches currently used in the AASHTO specifica- tions and based on the use of “R” factors to scale the design forces. The downside of the R factor approach is that no description of expected behavior is possible. In a second effort, a displacement target design method was developed that allows the user to choose acceptable levels of pier displacement during a design event and, using elastic dynamic analysis methods, determine the pier forces. Both methods led to similar and acceptable results. Both the reinforced emulation details and the hybrid posttensioning connections were found to provide acceptable designs. Khaleghi (2005) presents the Washington State DOT implementation of this research as well as other work in the area of precast elements used for structures in high-seismic demand regions. The reports details the state DOT philoso- phy on connection design between superstructure elements in prestressed concrete bridges made continuous for live load, and also details the design criteria for the connection between superstructures and substructures. Khaleghi also discusses the use of precast/prestressed concrete elements for rapid pier construction. He briefly details the design requirements for the main member design, including geometric consider- ations for ease of constructability. And he provides a sug- gested design procedure for successfully using precast pier elements in high-seismic regions with a specific application to ABC projects. Similar information is presented in Khaleghi (2008), including the precast seismic resistant bridge with decked girders presented in the Figure A.2. ABC Opportunity The use of precast pier elements has obvious advantages over the slower process associated with cast-in-place construction. A hesitancy to adopt a structural design concept with discrete elements connected mechanically at the joints has led to slow adoption of precast substructure systems in areas of high- seismic demand. The studies by the Washington State DOT, with support from the University of Washington, have evalu- ated most critical aspects of precast element use in ABC appli- cations in high-seismic demand areas and provided adequate design and detailing guidance. Bridge Superstructure Concepts This section focuses on components that can be used to acceler- ate bridge superstructure construction. Obvious solutions such as conventional prestressed concrete beams are not explored. Those are common practice, and no additional review is required. The projects that follow involve various newly researched systems and details that are unique to U.S. practice. The concept of developing a bridge system from precast concrete components for accelerated construction is not a new idea. A pair of bridges along I-87 between the Westchester Expressway and Armonk, New York, was constructed com- pletely of precast and epoxy connected concrete components in 1965 (Engineering News Record, 1965). The bridge super- structures consist of precast concrete box girders supported on precast concrete piles. The substructure consists of precast pile cap beams with cast-in-place abutment backwalls. Bridge superstructure erection can be advanced in many additional ways other than those presented here. Concepts such as the Nebraska Inverted T beam system, the Washing- ton State DOT’s use of deck bulb tee girders, and others have been used to accelerate bridge superstructure erection. There are numerous references for these projects, and they are rea- sonably well known. The use of deck bulb tees is the focus of ongoing research as part of the NCHRP 12-69 project, and recommendations on improved joint details and design con- siderations should be available as the project progresses. Data will be incorporated as they become available in subsequent phases. Similarly, some useful and unique information related to precast segmental construction of major bridges inter- nationally is available, including a report on the Sutong Bridge in China which is the world’s longest cable-stayed bridge (Liu et al., 2007). The experimental Roize Bridge, located near Grenoble, France, uses an innovative composite space truss and unique modular construction methods specifically developed for efficiency and economy. The Roize bridge, completed in 1990, was the first structure to use a composite space truss combin- ing concrete, structural steel, and prestressing to provide a stiff, yet lightweight structural system (Mueller, 1993). The Texas DOT (Cox, 2008) has developed a prestressed concrete decked slab beam (with accompanying standard

201 drawings). These decked slab beams are cited by William Cox as being “a good selection when rapid construction is desir- able and when minimal superstructure section depth is nec- essary for bridges on low-volume roadways.” With a standard width of only 24 ft, these decked slab beams are not likely to be viable for most locations where rapid construction is desired. However, the promotion of these types of standard sections represents a significant step forward. The I-95 James River Bridge in Richmond, Virginia, com- pleted in 1958, is six lanes wide and 4,185 ft long (Kozel, 2009). In 1979, a latex concrete overlay was used to rehabilitate the bridge deck. Approximately 20 years later at a cost of $49 mil- lion, the superstructure on the bridge was replaced. By using an innovative reconstruction technique, rehabilitation of the bridge was managed with minimal disruptions to the traveling public. Before work began on the superstructure of the bridge, dozens of preconstructed composite units consisting of an 8¾-in. deck and steel plate girders were fabricated approxi- mately 1 mi away; those units varied in size and weight, with the heaviest weighing close to 120,000 lb. Minimal disruption was achieved by working on the bridge only Sunday through Thursday between 7:00 p.m. and 6:00 a.m. During that time period, the six lanes were reduced to one lane in each direction. Using concrete saws, construction crews cut free a segment of the existing superstructure. That segment was then removed by using two high-capacity cranes and hauled away. Next, a new preconstructed segment was transported to the bridge and set in place with the same cranes. Each segment carried three lanes of traffic; two segments were needed to form one span of a three-lane roadway. At the end of the 11-hr construction period, the six lanes of the bridge were reopened for rush-hour traffic with the new superstructure segments in place. This procedure was repeated until the entire superstructure of the bridge was replaced. By using precast bridge abutments, wingwalls, and deck girders, the replacement bridge at Mitchell Gulch was com- pleted in 37 hr of actual construction time over a 46-hr period over one weekend in August 2002 (Culmo, 2009). Before this period, the contractor had driven H-piles to support the pre- cast substructure; and those H-piles were positioned to avoid the existing bridge structure, which was removed in just over 5 hr. Each abutment (consisting of lower and upper back wall units) was 44 ft wide; wingwalls were 23 ft long. Superstructure elements, which also provide the deck of the bridge, were 5 ft, 4 in. wide, 18 in. deep, and 38 ft, 4 in. long; eight of those units were required to obtain the desired bridge width. To save time, the outside deck girders were constructed with integrated bridge railings. Since precast concrete elements made up 90% of the structure, a significant amount of welding was required to connect them. More than 1,200 linear feet of field welding was required in this short construction time, which led to one observation that other methods are needed for connecting pre- cast units in ABC projects. By replacing the deck on the Lewis and Clark Bridge on SR-433 over the Columbia River between the states of Washing- ton and Oregon, the two state DOTs extended the life of the bridge by an estimated 25 years (FHWA, 2009d). More than 70% of the deck (3,900 ft) was replaced using 103 prefabricated full-width deck panels that were 36 ft wide and varied in length from 20 ft to 45 ft. The panels, which were 7 in. thick (6 in. of lightweight concrete plus a 1-in. modified concrete overlay), comprised two longitudinal steel stringers with several interme- diate transverse stringers. In addition to the full closures, which were allowed for only 120 nights between 9:30 p.m. and 5:30 a.m., four weekend closures were allowed. Rather than use the Washington State DOT–designed placement procedure, the contractor developed a placement procedure by using self-pro- pelled modular transporters (SPMTs) with specially designed steel truss frames for lifting, transporting, and installing the new panels. The modified SPMTs were able to transport a new panel onto the bridge, remove the old panel that had just been cut out, and then lower the new panel into place before taking the old panel off the bridge. One panel was replaced each night in approximately 6 hr. One indication of the success of this system is that the contractor received a bonus for early completion. In all likelihood many additional ideas have been used suc- cessfully by owners, both domestic and international. (That an idea is not discussed in this report should not be interpreted as having any connotations.) This report is not intended to cata- log every possible idea but to highlight and draw attention to the concepts this team reviewed that appear to have promise and have already performed successfully in limited or broader application. Both steel and concrete bridge solutions are pro- vided in the following illustrative projects. Project Title: I-95 James River Bridge Replacement, Virginia Citation: Kozel, 2009 ABC Design Features: Nighttime replacement of modular sections of existing superstructure ABC Construction Features: Preassembled segments lifted into position during overnight closures Project Description The Interstate 95 James River Bridge in Richmond, Virginia, is approximately 4,200 ft in total length and carries six lanes of traffic with a total volume of 110,000 vehicles per day. The original structure was built in 1958. The renewal project consisted of substructure rehabilitation and superstructure and deck replacement. The renewal project was started in 1999 and completed in late summer 2002. The Virginia DOT decided to use an innovative construction method allowing partial closure (leaving one lane open in each direction),

202 Sunday through Thursday nights, with the bridge returned to nominal full capacity each morning by 6:00 a.m. to main- tain traffic during peak travel times. While substructure rehabilitation was under way, large sections of the replace- ment superstructure were constructed at an assembly site within a mile of the bridge. At closing each night, existing deck and superstructure spans were cut at the bents and removed. Bearing pockets were prepared, and preassembled replacement superstructure and deck segments were set in place by using high-capacity flatbeds for transport and two ground-mounted cranes for hoisting. Preassembled seg- ments were typically less than 100-ft long and weighed an average of 60 tons. Each segment carried three lanes of traf- fic, and the contractor was able to set one segment per night. A full span (six lanes) required two preassembled segments. A total of 102 preassembled segments were replaced. For the main span of 800 ft, the existing steel truss superstructure was left in place. The existing concrete deck was removed and replaced with precast deck panels, also during nighttime partial closures. The contractor was allowed 179 nights of closure for segment removal and replacement but needed less than that. ABC Opportunity The James River Bridge in Virginia was a good candidate for accelerated construction for a variety of reasons. The bridge carries a high volume of rush-hour traffic through a major metropolitan area. Also, the Virginia DOT and the design/ consulting team carried out a thorough community dialogue before deciding on a final design and construction method. That process, along with a thorough public relations and communication plan during construction, created greater acceptance of the use of innovation and accelerated construc- tion. Removal and replacement of the existing bridge spans over the James River was accomplished with minimal dis- ruption of traffic through the use of full-span, partial-width, steel-girder-with-concrete-slab prefabricated segments erected during off-peak hours with high-capacity cranes and conven- tional flatbed trailers. Each span was replaced before reopen- ing the bridge at 6:00 a.m. The contractor was able to set up a temporary fabrication yard close to the bridge site, and the span weights and geometries (with the exception of the main span) were such that they could be designed to accommodate off-site assembly and transported into their final position. Project Title: Lewis and Clark Bridge Replacement over Columbia River, Washington and Oregon Citation: FHWA, 2009d ABC Design Features: Full-depth precast deck replacement ABC Construction Features: Preassembled segments lifted into position during overnight closures Project Description The Lewis and Clark Bridge across the Columbia River carries SR-433 between Oregon and Washington State near the Port of Longview, Washington. The bridge is 5,478 ft in total length, with 34 spans carrying 21,000 vehicles per day (13% trucks). The main center span is 1,200 ft long. The original structure was built in 1929 and appears on the National Register of Historic Places because, at the time of construction, it was the longest and highest cantilever steel truss bridge in the United States. After considering complete replacement with a new cable stayed bridge, the Washington DOT opted to extend the life of the existing bridge by 25 years by using full-depth precast deck replacement design. An innovative approach to deck panel replacement involved the contractor’s use of large spe- cialized transport equipment that could both remove the old deck panels and transport new panels to the site. Full-width, full-depth, precast replacement deck panels were hoisted into location with a specially designed steel truss framed gantry system. This use of innovative construction equipment and prefabricated design elements allowed the contractor to meet scheduling constraints. The precast concrete deck panels were constructed of light- weight concrete and received a concrete overlay after place- ment. The panels and overlay were supported by a frame of longitudinal and transverse steel stringers. The bridge was also widened by incorporating prefabricated sections supported by a single longitudinal steel girder. Construction was further accelerated by using precast approach slabs. The project was completed in August 2004. ABC Opportunity The Washington DOT was committed to accelerated construc- tion to minimize the impact on the traveling public and main- tain efficient operations within the Port of Longview adjacent to the bridge. To accomplish the accelerated schedule, the con- tract allowed for full closure nightly, from 9:30 p.m. to 5:30 a.m., for 120 nights (124 were actually used), plus four weekend clo- sures (three were actually used). Traditional designs and con- struction methods would have required either a 4-year construction schedule if performed under traffic (single-lane closures) or full closure for several months. In addition to accel- erating the construction schedule, use of prefabricated elements allowed for material inspection before installation, which elimi- nates the need for specialized testing equipment and promotes worker safety by reducing exposure to traffic during construc- tion. Furthermore, the use of full-depth, full-width precast replacement deck panels, precast widening sections, and precast approach slabs improved the constructability of the bridge. Project Title: Steel Box Girder Bridges Made Continuous for Live Load

203 Citation: Azizinamini, Yakel, and Niroumand, 2008 ABC Design Features: Design of a pre-topped steel box for accelerated construction ABC Construction Features: Use of prefabricated steel box girder modular superstructure units Project Description Azizinamini et al. (2008) focuses on the use of pre-topped steel box girder units as an ABC technology. The team of researchers had previously conducted research in the area of steel bridges constructed as simple span for dead loads and continuous for live loads. This is analogous to the traditional approach used to construct prestressed concrete beam bridges. In the conventional approach for prestressed concrete, beams are erected in a span-by-span situation, simply supported and not connected to each other in any way at the substructures. Because of their own weight and the slab weight, the girders resist loads in a simple-span condition. Only after the deck hardens can the structure behave in a continuous fashion through special detailing of the closure pour at the pier loca- tions. This technology has traditionally allowed for rapid erec- tion of concrete bridges. Steel bridges, on the other hand, have been designed and constructed as continuous structures, largely because we can connect them in the field. That requires a timely operation to complete bolted field splices in the air, many times over active traffic lanes. Azizinamini provides a design approach to use the simple-made-continuous approach for steel bridges as well. The slight loss of economy in the positive moment region—due to lack of continuity for dead loads—is counter- acted by a reduction in the negative moment at the piers—since continuity is present for only a portion of the loads. These two tend to cancel each other in terms of steel required to resist the loads. Additionally, substantial benefits are attained in stability and safety by using simple-made-continuous systems. Earlier development involved the use of simple-span steel I-beams and a cast-in-place concrete deck. The new proposal by the University of Nebraska researchers is to construct steel box girders in lieu of I-beams and to add a concrete deck to the box, with the completed section lifted into place on a bridge. A sim- ple longitudinal closure pour is required to complete the bridge. The researchers propose that, in lieu of traditional lap splices, the closure pour contain headed reinforcing bars, shown to result in shorter lap lengths and thus a smaller closure pour. In Azizinamini, various end details are proposed and tested for the steel beams. To achieve continuity, a concrete pier dia- phragm is cast that envelops the ends of the beams. In negative bending, a compression block forms near the bottom flange and can be transferred by the flanges bearing against each other or—at greater expense and with greater difficulty—by welding in the field. That was deemed too expensive and time- consuming. A modified detail was developed that eliminates the contact condition for the flange and uses end plates to provide the bearing surface. For the tension component, the connection again is as in a concrete bridge; supplemental lon- gitudinal slab reinforcing is used to make the connection. The report provides several details. Figures A.3 and A.4 illustrate a typical section of a bridge; dimensions are given in inches. The slab width for a typical prefabricated section is 118 in. (about 10 ft), and the slab mea- sures 6.5 in. deep with a subsequent overlay. A hybrid girder with Grade 50 steel for the top flange and webs and Grade HPS70 steel for the bottom flange is proposed. A typical pre- fabricated section is shown. Source: A. Azizinamini. Figure A.3. Typical section of modular steel box beam bridge.

204 Source: A. Azizinamini. Figure A.4. Modular steel box beam bridge, typical unit. Source: A. Azizinamini. Figure A.5. Pier continuity details. Project Title: Pre-Topped U-Beam Structures Citation: Griffin, Wolf, Freeby, Hyzak, Hohmann, and Cox, 2008 ABC Design Features: ABC Construction Features: Rapid construction of high- way overpass structures with traditional materials Project Description In response to the development of a pre-topped steel tub girder system by the Texas DOT, the Precast/Prestressed Concrete Manufacturers Association of Texas hired Project R04 team Once the composite beams are prefabricated, they are erected in simple spans and adjacent to one another. The longitudinal closure pours are completed, as is the pier diaphragm connec- tion. Completed pre-topped sections can be constructed with either lightweight concrete (120 pcf) or normal-weight con- crete and again with either the 6.5-inch subdeck or a conven- tional 8-in.-deep complete deck. Azizinamini provided data to the study team as a supplement to the published paper, indicat- ing that for a series of study bridges, the completed box sections would range in weight from roughly 70 tons to 100 tons. That is comparable to the weight of heavy prestressed concrete bulb tee or tub girder bridges, so erection with cranes is possible with this solution. It is promoted specifically as an ABC solution involving the use of conventional technology. However, it is conducive to ABC construction techniques as well. The study cites the case of a planned bridge replacement of 262nd Street over I-80 near Lincoln, Nebraska. Because a convenient detour is available, the bridge can be closed and demolished. However, the new bridge will be constructed on the existing approach paving, at grade, and then launched forward by using an SPMT at the leading end to move the bridge. ABC Opportunity The ABC opportunity provided by this technology is the pos- sibility of eliminating the costly, time-consuming, and some- times dangerous operation of field splicing steel girders while in the air. Analogous to concrete bridges, the girders are erected in simple-span conditions under noncomposite dead load; following the pier joint closure pour, the bridge is made con- tinuous for composite loads such as barrier rails, live load, and any future wearing surface or overlay. The girders can be erected in a pre-topped condition with a full-depth deck or subdeck, and yet the pick weights still remain within the capac- ity of commonly available cranes. Thus, an opportunity exists to erect a superstructure in not much more time than it could be installed with movement techniques such as SPMT or lat- eral skidding. The only extra time is the required field closure pour and curing time.

205 member Structural Engineering Associates (SEA) to develop a similar pre-topped U-beam concept using precast girders for the structural system. Following review of the system, the Texas DOT elected to institute a pilot project of four bridges in the Waco District. The pre-topped U beams developed for these bridges have an overall structure depth of 3 ft, 2 in. and a span length of 115 ft. The beams’ construction process is geared toward minimizing the time of on-site construction. To minimize fabrication cost, the pre-topped U beam has the same soffit width, strand layout, and web face slope as standard Texas DOT U beams. The pre-topped U beam is composed of a 10.5-in.-thick bottom flange, 6-in.-thick webs, and a 7-in.- thick top flange. The top right and left corners of the top flange are dapped (9 in. wide by 4.5 in. deep). After fabrication, the beams are transported to the contrac- tor’s yard (near the job site), where they are erected on tempo- rary supports that simulate the permanent supports—providing the same cross-slope, longitudinal slope, and relative elevation between beams. By using a screed machine and blockouts for the closure pour strips, a 4-in.-thick topping is placed on each beam. The metal portion of the exterior rail is then installed on the exterior beams. Once the 4-in. topping has cured, the beams are placed side by side onto the permanent substructure. The required reinforcement is then placed in the longitudinal and transverse closure areas, and the closure strips are then poured. After the strips are cured, the bridge is ready for traffic. The erected weight can be as much as 150 tons per beam line. Even so, production rates of two spans of beams per night have been achieved. An additional seven bridges have since been designed. These last seven bridges use a modified pre-topped U-beam system. The modified system uses the same pre-topped U beams. The difference is that, in this modified system, the beams are trans- ported from the fabricator directly to the site and erected onto the permanent substructure. The entire topping and beam joints together are then cast. The beams themselves provide a majority of the required formwork. Some steel plates are needed to span the gaps between the beams. Although this method may increase on-site construction time slightly, the quality of the riding surface can be better controlled. To reduce the on-site construction time, the slab and beam joint re inforcement may be pre-tied near the construction site, so that immediately after beam erection, the reinforcement can be placed. Welded wire fabric can also be used. ABC Opportunity The proposed system offers a precast concrete beam option with span–depth ratio and total span capabilities typically reserved for shallow steel tub or I-beam bridges. The use of butted flanges and a pre-topped section minimizes the Figure A.6. Typical construction details.

206 amount of field forming and thus expedites slab construc- tion. Providing a completely cast-in-place topping provides the optimum ride quality and control over final deck geom- etry. Large-capacity cranes are able to erect several spans of girders in a single night given the simple procedure of span- by-span and butted construction. Project Title: Blackhawk County Prefabricated Bridge Construction Citation: Wipf, Klaiber, Phares, and Fagen, 2000 ABC Design Features: ABC Construction Features: Rapid construction and erec- tion of prefabricated bridges by county personnel Project Description A steel-beam precast unit bridge was developed through lab- oratory testing at Iowa State University. The steel-beam pre- cast (S-BP) units consist of two steel beams connected by a reinforced-concrete deck of limited thickness. The thin deck reduces the weight of the units and makes it possible to con- struct the units off site and to transport them to the bridge site for rapid assembly. Once the S-BP units are connected, a cast-in-place reinforced-concrete topping is cast over the S-BP units, and the bridge railing is attached. This system can be economically used in bridges in the 30-ft to 80-ft range. A demonstration bridge consisting of four S-BP units (each 7 ft, 6 in. wide and 65 ft long, weighing 23 tons) was con- structed. The units were fabricated in the Black Hawk County Maintenance Facility in Waterloo, Iowa, after which they were transported to the bridge site and assembled. A cast-in-place reinforced-concrete deck was placed over the S-BP units to provide the required total deck thickness. The demonstration bridge was instrumented and load tested at various states of construction. On the basis of those tests, the S-BP unit bridge was shown to be a viable and economical alternative that can be rapidly constructed. The same county completed an additional ABC project. The substructure portion of the project is highlighted. To expedite foundation construction, a cantilevered wall was built for the abutment and wings. Interlocking sheet piles were used for the wing walls. The abutment front face consists of vertical steel H piles with a concrete panel infill. Once the piles were driven, forms were attached, and the space between the piles was filled. It is possible to use precast wall panels with this concept as well, though tolerance of the pile driving must be controlled. ABC Opportunity A simple and cost-effective abutment and superstructure con- cept were developed and tested at Iowa State University and constructed by local agency forces for use in ABC projects. The project demonstrates that simple and durable systems can be deployed with limited equipment and resources and with con- struction times measured in days. Project Title: Minnesota DOT/Poutre Dalle Inverted T-Slab System Citation: Hagen, Ellis, Fishbein, Molnau, Wolhowe, and Dorgan, 2005; Bell, French, and Shield, 2006a and b ABC Design Features: Development of an inverted T-beam precast deck system ABC Construction Features: Rapid construction of short- span bridges with conventional equipment Project Description One of the objectives of the 2004 prefabricated bridge elements (PFBE) scan tour was to identify technologies that could be used for rapid bridge construction in the United States. The Minnesota DOT sponsored research and a series of research projects to develop a local variation of the French Poutre Dalle system, a system of inverted precast slabs used for short- and medium-span bridge replacements. The specific application of the Minnesota DOT precast slab system (PSS) is to replace bridges that were commonly built as cast-in-place slabs requir- ing extensive falsework and long construction times. The bridge uses a series of precast/pretensioned sections, with deep reinforced closure pours between sections and a cast topping providing composite action and the riding sur- face. The section was designed with input from fabricators and contractors. It makes use of existing stressing beds and forming systems for maximum economy. The section is designed with the AASHTO load and resistance factor design (LRFD) specifications and can be used for simple or con- tinuous spans. Secondary effects from creep and shrinkage restraint were a significant part of the testing and analysis of the system. The system has a span–depth ratio of as much as 30 and can achieve maximum spans of 65 ft. In the first of two demonstration projects, a three-span bridge with equal spans of 45 ft over the Tamarac River was reconstructed. The existing substructures were retained, and the bridge superstructure was replaced with PSS components with typical segment weights of approximately 25 tons. The beam costs for the project were $43 per ft2, and the total proj- ect cost for the staged reconstruction was $83 per ft2. In a second project, a three-span bridge with spans of 22 ft, 27 ft, and 22 ft was completely replaced. Piece weights on the order of 10 tons were installed with a light crane. Total con- struction time was 20 days to construct abutments, pile bents, and the PSS units. The PSS units cost $48 per ft2, and the total bridge cost was $157 per ft2. This is much higher than traditional cast-in-place concrete slab projects, which in Minnesota range from $80 to $85 per ft2.

207 Subsequent to the demonstration projects, the Minnesota DOT has continued to advertise projects that use the PSS sys- tem, and the cast-in-place prices have continued to decline. The latest data indicate total installed costs of $95 to $100 per ft2, with other projects in the pipeline. The department con- tinues to work with researchers and fabricators to economize the system further by reducing the amount of reinforcing and through wider use of the products. The Minnesota DOT has observed some longitudinal and transverse cracking on the deck surface on several of the PSS bridges and has begun an investigation to determine the cause. Several changes have been included in the latest generation of PSS designs and will be closely monitored after initial construction. ABC Opportunity The proposed system offers a rapidly constructed, shallow, mostly prefabricated solution for structures that convention- ally would be constructed as cast-in-place slab systems. With light elements easily erected by local contractors and even by county or department personnel with access to light-duty equipment, the system provides another option to owners in the short- to medium-span precast bridge market. Project Title: PCINE NEXT Beam Bridge Citation: Precast/Prestressed Concrete Institute Northeast, 2008 ABC Design Features: Development of new double-T pre- cast bridge superstructure modules ABC Construction Features: Rapid construction of com- pleted superstructure by using standard modules Project Description The Precast/Prestressed Concrete Institute Northeast regional association has made multiple advancements in the area of accelerated construction and in standardizing the practice among the New England states and New York State. The asso- ciation’s latest effort is to develop a short- and medium-span Source: Minnesota DOT. Figure A.7. Inverted T-slab installation. Source: Minnesota DOT. Figure A.8. The Minnesota DOT inverted T-slab bridge.

208 shallow replacement structure specifically for standard appli- cation and for use in accelerated bridge construction (PCINE, 2008). The use of precast double-Ts is not a new advancement. They have been popular bridge choices in many states throughout the country. The northeast extreme tee beam (NEXT beam) does offer some structural and construction advantages over the other systems, as will be described. The proposed beam section is presented, and as indicated by standard notes, it comes in an 8-ft or 12-ft nominal width. When stacked side by side, the top flange is considered to be nonstructural; it simply replaces stay-in-place formwork. A full-depth cast-in-place concrete slab is then applied to create a fully composite section. The section is also designed for an additional 3-in. bituminous overlay. The standard 8-ft section has a span capability of approximately 85 ft, and the 12-ft sec- tion has a span capability of 75 ft. The section depth is 36 in. Charts and graphs are included to provide the span capability of each variation of the NEXT beam, listing section width, number of strands, and concrete strength. The sections do not need to be connected along the longitudinal joints with any welded connection as is common with other double-T sys- tems, nor do they require transverse posttensioning—another cost and complication avoided with this refinement. ABC Opportunity This simple bridge shape—requiring a cast-in-place topping as the only field-cast concrete—accommodates a wide range of spans consistent with bridge sizes in the existing inventory. The sections are shallow (maximum depth of 36 in.) and are easily erected; the heaviest section weighs less than 60 tons and erected easily with commonly available cranes. These sections could just as easily be combined ahead of time (including the slab) and moved into place with one of various heavy move techniques. Project Title: CBDG Concrete Modular Bridges Citation: Benaim, 2006 ABC Design Features: Innovative design of box girder sys- tem for short- and medium-span bridge replacement ABC Construction Features: Innovative box system built by crane, gantry, or incremental launching Project Description The Concrete Bridge Development Group (CBDG) is a con- crete bridge industry association in the United Kingdom. The group sponsored development of standardized modular con- crete bridges for the construction of routine overpass bridges as a marketing initiative to establish prestressed concrete as the preferred solution for spans up to 50 m (Benaim, 2006). The result of those efforts was development of a shell system that is precast off site and shipped to the project; permanent prestressing is applied and then the shell is filled, forming a solid section. Details are provided. The development and capital costs of new equipment can be capitalized over an estimated 5 to 10 bridges, indicating that an initial invest- ment would likely be paid off within several years for any contractor with a steady stream of work. What makes the system unique is, first, the lack of a top slab. That makes forming more cost-effective than traditional inner-void box girder forming. Additionally, unlike U.S. prac- tice, even for external prestressing the entire box is filled with concrete as protection for the prestressing. That step adds substantial weight to the final section and limits its efficiency in some ways. However, it is included, at least in part, because of U.K. limitations and past problems with hollow box sec- tions and external prestressing. The advantages of the system include the use of light sec- tions, adjustability in length and in width, usability for vari- ous bridge widths and lengths, and suitability for multiple erection methods including gantries, cranes, incremental launching, and full-span erection. Once the shells are erected, they are prestressed together, the infill concrete is placed, and the full span prestressing is applied. Shell depths of 1.0 m to 1.5 m can be used for spans up to 25 m; shell depths of 1.5 m to 2.0 m can span up to 35 m; and shell depths of 2.0 m to 3.0 m deep are required for spans of 35 m to 50 m. The depth– span ratio is 1:17 for a typical installation. ABC Opportunity The issue of standardizing designs is again proposed as an ABC solution for bridges of the routine overpass type. The CBDG system advocates the use of segmental shell systems to create bridges of various lengths and widths by simple variations in segment geometry. The bridge is designed to be erected by both conventional and innovative methods, mostly innovative due to the segmental nature of the structure. Like most segmental bridges, one method of erection involves the suspension/ support by a truss/gantry until a span is complete. Unlike other segmental bridges, the shells are light enough to be post- tensioned together on the ground and lifted like a single-span box girder into place, at which point the infill and final pre- stressing are applied. Other installation techniques include the use of incremental launching or SPMT installation. Project Title: Flexi-Arch Construction Citation: Macrete, 2008 ABC Design Features: ABC Construction Features: Rapid installation of arch bridge structures for short and medium spans

209 Project Description The construction of an innovative short- to medium-span bridge can be possible through the use of an innovative pre- cast arch system. In a system developed in the United King- dom, precast arch bridges are constructed by using an articulated block system that ships flat but rotates into the desired arch shape when lifted and placed against simple footings. Once the footings are in place, typical superstruc- ture erection is completed in a day for a typical bridge. Construction consists of flat stacking the various arch seg- ments and shipping them to the site on a flatbed truck. This system has some advantages over shipping precast arch seg- ments in that fewer trips are likely needed and the rise of the arch is no longer a shipping consideration. Assembly of the bridge consists of several simple steps: placing the footings (which could be precast), erecting the adjacent Flexi-Arch segments, erecting spandrel walls, and typically filling the section with lightweight flowable fill. The lightweight flowable fill ties the arch sections together, ties the spandrel walls to the arch, and provides a nonsettling and waterproofed fill to the arch system. Additionally, given the reduced weight of the flowable fill, the arches can be designed for lesser load than if conventional granular backfill was used. Once the flowable fill is allowed to cure, the final riding sur- face (asphalt or concrete) can be applied. ABC Opportunity This system offers an alternative structural system for short- to medium-span bridges. It competes against other structural systems such as precast arches, large culvert sections, arched metal pipe systems, and short-span prestressed beams such as planks or voided slabs. It is inherently faster than all of those systems except the precast arch or culvert, which probably have similar construction durations and simplicity. Project Title: FEHRL New Road Construction Concepts Citation: Forum of European Highway Research Laborato- ries, 2008/Tanis, Nicolas, Cardin, Keller, Schaumann, Toutlemonde, and Godart, 2007 ABC Design Features: ABC Construction Features: Development of multiple bridge replacement concepts that use high-performance materials Project Description The Forum of European National Highway Research Labo- ratories (FEHRL) is sponsoring a program analogous to the SHRP 2 initiatives to investigate changes in the highway design, construction, and maintenance practices that are likely to influence the profession for many decades. The challenge for European road owners is the same as in the United States: a multiplicity of road owners, slow innovation, lack of consistent design guidance, and incomplete cost-benefit ratio informa- tion. The New Road Construction Concepts (NR2C) program aims to address these deficiencies by investigating new con- cepts for the road of the future and specifically investigating the enabling technologies and concepts. A series of reports and working documents have been published describing the evolving process of bridge concept development in Europe. These reports are described here to reflect the progress of a similar research effort and the findings. The State of the Art Review—A New Vision for Bridges (Tanis et al., 2007) provides an overview of the needs and expecta- tions for future European bridges, describes various new structural materials that might be used in future bridge con- cepts, and provides a vision for what the bridge of the future might be. Though not the same as the U.S. bridge population, Euro- pean owners face a similar problem of aging infrastructure, deficiencies on ever-larger portions of the road network, and inefficient programs for rehabilitation and replacement. The expectation of owners moving forward is for bridges with robustness, long life, safety, economy, little maintenance, and some interest in improving construction methods. They are not particularly concerned about the ability to widen or lengthen a bridge in the future. That is understood to be a necessity of a future road improvement project, and the pub- lic is expected to understand the need for such changes. Considering new materials, various applications for new materials are proposed. High-performance steel is advocated as a way of making composite girders (steel-and-concrete) more effective in low-clearance environments (shallower structures are possible), for use in moveable bridges, and for situations where lessening the structure’s weight can result in savings in shipping, erection, or foundation costs. This can be useful in many bridge types and situations. Very-high-performance concrete and ultra-high-performance fiber-reinforced concrete (UHPFRC) are both discussed for their ability to provide high strength and durable solutions in future bridges. Among the uses of UHPFRC, in particular, are highly efficient beam shapes, reinforcing-free concrete decks, shallow or thin arch sections, box beam sections with perforated webs, and also in bridges where shock loading (blast, earthquake) are of concern. In all of these applications, the materials’ high strength, easy workability, long-term durability, and reinforcing-free nature (except for internal fibers and prestressing) are the reasons the proposed innovation is possible or contemplated. The use of fiber reinforced polymer (FRP) materials con- tinues to receive attention as a material of the bridge of the future though we are well into our second decade of research

210 and development of trying to use these materials in civil engi- neering applications. Because of their light weight, high strength, and long durability, these materials are proposed for various structural applications. In addition to use as external reinforcing and retrofitting (a relatively mature and estab- lished market already), their use is proposed for all-FRP bridges as well as in composite applications that blend other materials (e.g., UHPFRC) with FRP in a hybrid structure. An interesting aspect of the NR2C project is its focus on timber for short-span bridges. Viewed positively as a renew- able resource, timber has environmental attributes that, from a holistic project viewpoint, make it an attractive solution when deployed appropriately. Other materials such as alumi- num and titanium are also discussed as having desirable properties. The report also presents the prospect of enhancing existing structural concepts for use in future bridges. The greatest pos- sibility for innovation seems to be in using new materials to provide new opportunities for economy and efficiency built on existing design concepts. The maximum opportunity is in the area of beams and slabs. For bridge slabs, the authors advocate the use of FRP reinforcing in lieu of steel reinforcing as the first step toward improving slab durability. As a second level of enhancement, they propose integrating an FRP formwork sys- tem that serves to reinforce the bridge deck at the same time, thus providing two functions in one construction operation. Steel-free slabs of UHPFRC are also suggested as a concept for the future. With no reinforcing in the slab, the structural strength derives from the strength of the fiber reinforced con- crete alone. Intermittent tension ties are required between beams to anchor the shallow arching behavior that forms at failure. A commercially available precast product known as ArchPanel is also recommended for use in steel-free decks. For new beam concepts with innovative materials, several FRP superstructure concepts are proposed. In the hybrid tube system, an FRP box girder is constructed. At the bearing loca- tions, the full depth of the box is filled with concrete (for strength, stiffness, and stability) while otherwise remaining hollow. A concrete deck is cast on top. The concrete-filled carbon-filled shell system was a spin-off of filament-wound pipes used for noncivil applications or for the fabrication of shells for column jacketing. Once filled with concrete, the shell is restrained from buckling and the strength of the confined concrete is enhanced. Given the challenges of using round sec- tions for bridge girders, advancements have focused on using concrete-filled square and rectangular sections instead. The NR2C bridge report also focuses on new bridge con- cepts. It states the following: Developing a completely new concept is the most difficult way to innovate, but also the most challenging and rewarding of the methods discussed. Therefore, combining existing concepts may be useful for coming up with unusual structural solu- tions. The construction process may become more compli- cated and thus more expensive. Aesthetics are of prime importance in bridge design and a mixture of structural ele- ments will not necessarily satisfy the observer, since the struc- tural simplicity and visual attractiveness of the bridge may be interfered with. In its study of new bridge concepts, even these are not so new. The recommended new bridge types to be studied in the future include extradosed structures (a hybrid between post- tensioned internal tendons and a cable-stayed bridge); fin- back bridges, which employ a solid concrete variable depth fin or rib extending above the deck in negative moment regions; and the stress ribbon bridge, possibly one of the oldest bridge types in existence, using draped tensile elements and decking to span two supports. NR2C’s general perspective on innovation is that it is expected to be a slow and continuing process of sustained development. Rarely do remarkable leaps of technology appear in the market (prestressed concrete is cited as one recent example of a dramatic change in materials technology). The researchers comment that radical changes in technology and capability (using bridge spanning capability as an example) have only come about in the presence of new materials. Materials drive innovation. New concepts must continue to meet current functional, safety, schedule, and cost demands. Future innovation should start with a better understanding of existing bridges and existing innovative technologies. Those include making use of better surveying tools, materials testing procedures, management systems, structural monitoring, and innovative repair techniques. Once these existing enhance- ments are mastered and common in practice, the industry can look to future concepts. The greatest opportunity for innova- tion is perceived to be in the area of “ordinary” bridges. Following the preliminary study, the NR2C team went about developing several new bridge concepts for the short- to medium-span traditional bridge market. Several solutions are proposed for a 10-m-span bridge and for a 25-m-span bridge, using various materials. The first solution proposed using glulam timber beams with a combination of high-strength materials for the top and bottom flanges, UHPFRC for the top slab, and FRP for the bottom flange. The section is hollow. In the simple-span con- dition, the FRP bottom flange will always be in tension. The UHPFRC slab will be used to increase the bending strength of the section and to form a durable wearing surface. In a second scenario, the glulam timbers are replaced, and the FRP and UHPFRC plates separated by a low-strength material, with foam as the spacer. Available in standard widths and easily cut to size, this solution can be easily customized for various depths, and lengths. For both of these solutions, the proposed

211 module width is 1.25 m and the height is expected to be about 0.8 m for the total depth. A third solution that uses both tim- ber webs and foam infill was also proposed but not shown because it simply married the first two concepts. The pres- ence of the internal foam stiffens the system and also elimi- nates the need for any internal diaphragms between the vertical glulam beams. Testing of these three systems indicated that the best solu- tion is the first approach. In testing the second, it was found that the system had much greater deflection due to the low stiffness of the foam infill. That resulted in high tensile stresses in the slab due to local longitudinal shearing distortion of the foam in high-shear regions. The authors compared the total weight of the first solution with other types of bridges that could be used for a 10-m span. Comparisons were made between a conventional reinforced-concrete slab, a composite- steel bridge, prefabricated concrete girders, and Solution 1. The usual weights of these structures are as follows: con- crete slab, 125 kN/m; prefabricated girders and composite girders, 90 kN/m; Solution 1, 31 kN/m. Thus, weight savings are substantial. A UHPFRC waffle slab system was also studied. The system consists of 0.6-m by 0.6-m panels, with a top slab 50 mm thick. The system has a total depth of 475 mm, the longitudi- nal ribs being somewhat deeper than the transverse ribs. The system is prestressed in the longitudinal and transverse direc- tions for strength and stiffness and to join the segments in the field. When shipped full width, with transverse joints, each section is the width of the bridge by 2.5 m long. When shipped full length with longitudinal joints, segments 1.8 m or 2.3 m wide are available. The volume of material in the section is equivalent to assuming a uniform slab thickness of 164 mm for the entire width and length. A traditional concrete slab bridge would have a thickness of 500 mm for this same span. The weight is thus about one-third of a traditional slab bridge. One nuance of this design may change the results for U.S. application: The service limit state deflection allowed for this system is L/350. Traditional U.S. practice has limited live- load deflections to L/800 or L/1,000, depending on whether the bridge is designed for vehicular loads only or also carries sidewalks. Finally, the authors provide solutions for a 25-m span. The first system, not depicted, is a waffle slab concept similar to those described above but made deeper for the additional span length. This solution increases the rib spacing to 1 m on center both ways, and the top slab is thus thickened to 80 mm. The total structural depth is 1.05 m for a span–depth ratio of about 25. A standard module is full length, and the sections are 2 m wide. They are prestressed longitudinally, with exter- nal tendons passing through deviator diaphragms and trans- versely adjacent to the ribs. The resulting bridge has an equivalent quantity of material analogous to a 230-mm-thick slab, a dramatic reduction in materials from that used in a conventional bridge. Other more exotic solutions are also proposed, including a notched slab system. In the notched slab system, used for a 25-m span, portions of the longitudinal ribs are removed to save weight. In additional modifications, the longitudinal posttensioning is eliminated, and additional tensile capacity is provided to the system by bonding FRP plates to the bot- tom of the notched ribs. The plates measure a total of 50 mm thick. This system is about 50% heavier than the traditional waffle slab, mainly because it requires much wider longitudi- nal webs to bond with the FRP plates. The high volume of FRP added to this system brings into question whether this is truly an innovation above and beyond the traditional post- tensioned slab system. A hybrid deck system is also proposed, meeting the objec- tives of combining a stay-in-place forming system with the advantages of FRP reinforcing. The total slab depth of 200 mm matches the conventional 8-in. bridge decks in use for common beam spacings. However, the comparison ends there. The proposed system uses an FRP grid to span in the transverse direction and provide forming and strength to the system. Lightweight concrete is used for a majority of the deck depth to reduce weight, and a normal-weight UHPFRC topping is added. Schaumann and Keller (2007) provide experimental test results of short- and long-span specimens fabricated as indicated. Their results indicate that full com- posite action between the grid and concrete is possible. Dif- ferent types of lightweight concretes were studied as fill. Only simple-span conditions were studied, so the behavior of this system in negative moment regions (such as over beam lines) is untested. ABC Opportunity Many possible innovations are discussed. The basic waffle slab system appears to have the most immediate promise. In the 10-m-span configuration, small bridges are easily replaced with a very lightweight system that is quickly erected. The team also envisions the 10-m-span waffle slab system as a potential concept for a deck system with widely spaced gird- ers. Traditional concrete slabs become excessively thick and heavy with wide beam spacing (e.g., two girder systems). The waffle slab can serve as an effective spanning element in that scenario. Another option for the thin waffle slab is to use it to span through-girder edge beams, a possible concept for future bridges in low-clearance environments. The deeper waffle slab for the 25-m-span condition is the same or lesser depth than bridges used today for such a span, yet is much lighter and easier to install. Finally, the concept of stay-in-place FRP decking that also reinforces the tension face of the slab is a time and money saving opportunity.

212 The performance of this deck in negative bending requires further validation. Bridge deck Concepts This section focuses primarily on components that can be used to accelerate bridge deck construction. By far the most widely available information for ABC is related to precast bridge decks, and the information dates back to the 1970s and earlier. The main focus is on concrete because, in the short term, that will likely continue to be the predominant material used for bridge deck construction. This expectation reflects concrete’s long-standing position as the “familiar” material, its ready supply, ability to be widened and repaired when needed, and other characteristics. A variety of other deck types exist. These include glued-laminated and stress-laminated timber decks; open-grid steel decking; partially filled systems such as half- filled grid decks and exodermic bridge decks; and lastly, vari- ous forms of fiber reinforced polymer (FRP) bridge decks. FRP decks deserve special, but brief, mention because of FHWA’s level of research investment in the recent past and because many state DOTs surveyed during this project consider them still to be in a demonstration phase of implementation. More than a decade of published history exists concerning FRP bridge deck systems, and hundreds of installations in the United States and around the world have been performed (many of the installations were done with demonstration project funding). In spite of the generally good performance of these bridge decks (save several overlay failures), the num- ber of commercial manufacturers is down to one or two firms in the United States, and fabricated prices continue to hover in the range of $75 to $100 per ft2, a high cost for a deck sys- tem (higher in fact than some states’ total in-place bridge costs on a square-foot basis). The promise of FRP is its light weight and long life. From an innovation standpoint, FRP bridge decks may still hold promise over the long term as a viable deck system for rapid construction. However, recent innovations in concrete decks may begin to chip away at some of the inherent advantages of FRP. Several useful sources of information related to FRP deck construction are available on FHWA websites, including lists of completed FRP deck projects (FHWA, 2009b) and a library containing pertinent FRP-related publications (FHWA, 2009c). Another noteworthy reference describing innovative and pre- fabricated bridge deck types for minimizing traffic disruptions during construction is the National Cooperative Highway Research Program (NCHRP) Synthesis study (Shahawy, 2003). In addition, Bakis et al. (2002) is a state-of-the-art review that contains useful bridge deck information. From an inter- national standpoint, FRP bridge deck applications began in the early 1980s and continue to be constructed. Two recent studies of FRP bridge decks implemented in Asia (Prachasaree and Shekar, 2008; Kim et al., 2009) validate the continued interest in better understanding the potential application of FRP decks. Several representative summaries related to both partial- and full-depth deck systems are presented below. A partial-depth prefabricated deck is a deck that combines a layer of precast concrete and a layer of cast-in-place concrete. This system can be developed with either steel or concrete girders. The deck is connected to the steel I-beam through shear studs on the beam. The deck is placed in the field in panel segments; once all of the segments are placed, a cast-in- place concrete deck is poured on the precast deck. The precast deck acts as a stay-in-place form, eliminating the need for temporary formwork. This system can also be installed by using a precast girder and partial deck system. In that case, concrete girders are used instead of steel I-beams. The instal- lation procedure is identical regardless of the superstructure material (Russell et al., 2005). The inverted-tee system is another partial-depth pre- fabricated deck developed to accelerate the construction of bridges. This system is designed to accommodate longer span lengths. The inverted-tee portion of the deck is precast and placed at the construction site. A cast-in-place concrete deck is constructed later. Researchers have shown construction of this system to be faster and less expensive than conventional methods of construction. The inverted-tee deck system can be customized for different bridge projects and used in spans up to 26 m (Kamel and Derrick, 1997). A full-depth prefabricated deck, which has a few features that make it unlike other deck systems, is one that includes a completely precast concrete deck. Deck panels are designed with shear keys that are filled with a high-strength grout for a strong connection between panels. Composite action between the deck and the girders is created by using shear studs or high-strength bolts with steel beams and precast (PC) beams. Posttensioning strands are generally used in the longitudinal direction to keep the deck panels in compression. Full-depth prefabricated decks are effective because they can be installed quickly and constructed in stages to allow traffic to flow dur- ing the process (Issa, Yousif, and Issa, 1995). Durability, ease of construction, and reduced maintenance are only a few of the advantages of using a precast deck system for the rehabilitation of deteriorated bridge decks. The sys- tems can be adapted to any bridge, even those with horizontal and vertical curves. Segmental bridges were first constructed in Europe in the late 1940s as a response to the demolition of many bridges during World War II and the need to replace those bridges and decks quickly. Segmental construction came to the United States in the early 1970s and has improved dramatically since then. One application that has gained pop- ularity is the precast deck system. Within this category are two types: stay-in-place precast deck panels and integrally

213 constructed full-depth decks. The stay-in-place panels are used as a working platform for workers and machinery and as the forms for the final cast-in-place deck pour. The full- depth precast concrete decks are just that, no cast-in-place concrete is needed to finish the deck. With these two deck applications, as with other precast concrete applications, careful attention must be paid to the joints between each panel, as rapid deterioration may occur. Biswas (1986) provides more than 20 case studies of applica- tions of precast bridge deck systems. Each case study empha- sizes the system construction and the economy of using precast concrete. The case studies also include detailed descriptions and drawings of joints and joint materials used. The author discusses the advantages and difficulties during the design and construction of the precast deck system in each of the case studies. Case studies of concrete deck options are highlighted here along with an innovative steel plate deck system. Project Title: Lake Pomme de Terre Precast Deck Replacement Citation: Desai and Blakemore, 2008; Wenzlick, 2005 ABC Design Features: ABC Construction Features: Precast full-depth deck pan- els, match cast adjacent to the site, night work Project Description The Missouri DOT bridge carrying Route 64 over Lake Pomme de Terre is a 1,684-ft-long steel stringer bridge that, because of deck deterioration and a desire for a slightly wider deck, was reconstructed with full-depth deck panels. With one lane of traffic in each direction, half-width construction with traffic signals at each end to control the flow was not feasible due to long cycle times and large queues at each end. Full closure was also not an option because of prohibitive detour lengths. The only viable option was nighttime closures. This decision dictated the use of prefabricated deck panels. A typi- cal panel is shown in the figure. The contractor was able to arrange for a strip of local farmland to be used for an at-grade long line, match casting bed for the fabrication of the deck panels. Thus, all construc- tion was completed near the final bridge location and the bed could be graded and panels match cast for proper fit in the field. In the first night of construction, the contractor was able to demolish several panels of existing bridge deck. Time was not sufficient to install the new concrete panels, so a set of steel, open-grid deck panels (already prepared) was used to span the gap. In subsequent nights, the steel panels were removed, sev- eral more sections of existing deck were removed, precast pan- els were placed and grouted, and the steel panels set again to cover the remaining gaps. This process occurred every evening from 7:00 p.m. to 7:00 a.m., with a daily production rate of several 10-ft-long panels. Steel posts were attached to the tem- porary and permanent deck panels for a temporary railing to be attached as construction progressed. At the completion of the project, the temporary steel rails were removed, the per- manent concrete railings were cast, and the deck was overlaid. Placing and curing the overlay required two 1-week patterns where only a single lane was in operation. Figure A.9. Lake Pomme de Terre bridge deck panels, typical section.

214 An innovative screw jack system was used to level the panels and set the haunch. With clamps that would hold the bottom flange of the steel stringers and mating steel tubes for the ver- tical with a threaded rod screw jack installed in the center, the panel heights could be easily adjusted in the field. That allevi- ated any complications arising from final girder geometry once the existing deck was removed. The screw jacks raised and lowered steel angles that supported the panels and also provided the haunch forming. The cost of the deck replacement portion of the project was about $56 per ft2; the Missouri DOT’s statewide average for deck replacements is generally in the range of $40 per ft2. The agency estimates that for this bridge—a long bridge over water—the statewide average would not be a realistic estimate anyway. According to Wenzlick (2005), the costs of a cast-in- place solution would be close to the precast solution but would have required an additional construction season, would have required half-width operations at all times, and would have incurred substantial additional traffic control costs. ABC Opportunity The challenges of bridge deck replacement on long, narrow structures is particularly troublesome. Replacing a short bridge, by using a complete closure, might take only a week of construction to remove an existing deck and install new deck panels. That might be tolerable. Alternately, for wider bridges, staging might be undesirable but tolerable and at least possible. For this bridge, neither closure nor staging was realistic. An innovative approach that uses match cast precast deck panels, infill open-grid steel deck panels, and nighttime closures effec- tively demonstrated a low-impact reconstruction project com- pleted with conventional forces and equipment. An innovative haunch forming/panel leveling system is presented. Project Title: Rapid Precast Deck Construction Citation: Ronald and Theobald, 2008 ABC Design Features: ABC Construction Features: Innovative construction method to place precast deck panels Project Description Ronald and Theobald (2008) describe the development of a new semiautomated method for placing full-depth deck pan- els. By using a series of small carts placed between the bottom flanges of prestressed concrete girders, full-depth precast deck panels are incrementally placed and launched down the length of a span. Beginning at the supply end of the span, the first set of carts—carrying the first deck panel—is placed between the beams. The carts are winched toward the far end, and the next panel is placed in the same way. The second set is tied to the first so that the train of panels is moved forward incrementally. The carts can be equipped with jacking plates so that screw jacks installed in the panels have a surface to bear on while the panel grade is adjusted. By using a portable computer control system, jacks can be adjusted individually or in a group to set Figure A.10. Deck panel fabrication, installation, and haunch-forming method.

215 the proper geometry. Posttensioning is completed before- hand, then panels are lowered to reduce any frictional effects between the panels and girder tops. ABC Opportunity For long viaduct-type structures in particular, the system has various advantages. Most important, it allows cranes to con- tinue to be dedicated to superstructure beam erection and freed from erecting deck panels. The latter can be done sepa- rately with the mobile cart system. However, even for long spans where a long boom or heavy crane would be required to place heavy panels at the far end, this cart system can greatly simplify deck panel placement. Project Title: UHPFRC Bridge Deck Studies Citation: Garcia, 2007; Toutlemonde et al., 2005, 2007a and b ABC Design Features: ABC Construction Features: Development of lightweight UHPFRC deck panels for composite bridge construction Project Description Ultra-high-performance fiber-reinforced concrete (UHPFRC) bridge decks appear to offer promise in many areas, including replacement of traditional bridge decks in multi-beam bridges and as an interesting long-span slab alternative used in two- girder systems (in lieu of more-traditional orthotropic decks). In deeper configurations the material may be used in “slab bridge” types of applications. The use of UHPFRC as deck slab material for girder bridges has been investigated and has potential to radically change concrete bridge deck design in the future. Beginning with a discussion on UHPFRC bridge decks for traditional deck slab replacement, Garcia (2007) describes the design of a precast deck system for traditional multi-girder bridge construction. The system is depicted in plan and typi- cal section. The slab itself is 2.5 in. deep and is bounded by ribs of the waffle that are an additional 5.5 in. deep, for the total depth of 8 in. The ribs are spaced 24 in. on center. The weight of this system is approximately 55 psf, compared with 100 psf for a traditional 8-in.-thick solid slab. This has significant implications for reducing structural weight (and thus beam section requirements) as well as foundation loadings and struc- tural mass (for reduced-seismic demand). Additionally, for long spans, as dead loads become dominant, the system pro- vides the opportunity to use lighter beams, wider spacings, or combinations of these to gain extra efficiencies. Garcia compares the design and resulting product with a series of examples produced by Michael Baker and Modjeski and Masters as part of LFRD design examples and transition design examples. Both of these consultants use a traditional I-beam bridge with conventional beam spacing as the basis of their sample designs. Garcia includes the prior designs to provide a baseline for the predicted strength of conventional bridge decks. The UHPFRC system should be designed with these capacities as a minimum requirement. Because of the unique strength and stress–strain charac- teristics of UHPFRC prestressed concrete, the traditional AASHTO strength models cannot be used directly. An alter- nate method is shown, using strain compatibility to establish the equilibrium conditions and resulting strength. The fun- damental difference between the prestressed UHPFRC and conventional prestressed concrete is the concrete tensile strength contribution to the force balance equation; this con- tribution is negligible for normal concrete and is traditionally Figure A.11. UHPFRC slab-on-girder concept.

216 neglected. The sustained tensile strength of the UHPFRC assumed for this design was 1.125 ksi, a significant strength. Compression strength is 23.8 ksi. Design of the deck was computed in accordance with AASHTO LRFD specifications for vertical loading and colli- sion loading effects on the cantilever slab. In the transverse direction, T-beam behavior was assumed based on the regu- larly spaced ribs in the waffle slab (24-in. centers). Results of the design development exercise, which is completely detailed and instructive for future designers using UHPFRC, indicate that the prestressed UHPFRC system has somewhat higher strength than conventionally reinforced concrete decks. This is considered prudent until additional validation is performed. The author recommends that additional physical testing be conducted to validate the expected performance. Waffle slab systems have been tested in Europe, and the use of waffle slabs for slab on girder construction is promoted as a viable new technology for bridge systems. In a study by Toutlemonde et al., the flexural and punching shear capacity of UHPFRC slabs constructed with different commercially available cements is assessed. Both Ductal and Ceracem cements were used in the study. The motivation for the deck testing was the development of a two-girder bridge concept that uses a waffle slab for long-span bridge applica- tions as proposed in recent French research. The typical section is provided along with section details and several photographs of the section during fabrication. Though not shown, the pro- totype bridge has two 11.5-ft-wide lanes, two 3-ft shoulders, and two 3-ft-wide sidewalks. The main girder spacing is 19.7 ft, and the prototype bridge has spans of 295 ft, 426 ft, and 295 ft. In this span range, weight reduction has dramatic benefits in overall economy. This slab system weighs roughly 80 psf. As a frame of reference, a conventional 8-in. slab, used for beam spacings up to 10 ft (half of the provided spacing) weighs 20 psf more. This is a highly efficient system from a weight standpoint. The study describes various details of the testing. The objec- tive was to determine the flexural strength and punching shear strength. Load was applied to the slab through contact patches set to simulate the Eurocode design loading of 16 in.2 on top of an overlaid slab; the effective bearing area is 2 ft2 at the interface of the overlay and deck. Other tests were conducted on a slab without overlay (direct loading of the slab) and with a reduced contact areas as small as 8 in. by 10 in. Only in the deck tests with direct loading on the base deck was failure achieved. Before the punching shear tests, the waffle slabs had already been subjected to substantial static flexural and fatigue load testing. Those tests were not able to degrade the performance of the slab in any discernable way. The punch- ing shear tests on the reduced contact area had the worst per- formance factor: 2.5 times the level required by the Eurocode for design. When more realistic contact areas and strength along the four-sided shear perimeter are considered, the load factors are 4.6 to 6.0 times that required by the Eurocode. As mentioned, the punching shear tests were conducted on specimens with prior fatigue testing. These slabs were subjected to about 20 weeks of constant cycling in the lab under loads simulating the Eurocode axle spacing but with magnitudes that for some cycles greatly exceeded normal fatigue-loading magnitudes. The initial loading consisted of 10,000 cycles of 2-kip to 34-kip wheel loads per axle. Those are described as very high and represent the extremes of loads at the service limit state. An additional 2 million cycles were applied with 2-kip to 22-kip loads; that represents a damage equivalence of 100 years of routine fatigue loading. That test was followed by 100,000 cycles of 27-kip and 34-kip axle loads. The deflection and strain behavior was stable throughout this testing. The researchers examined the fatigue behavior of the sys- tem as well as the local performance of the slab and ribs in the most heavily loaded waffle cells. They concluded that no fatigue damage is expected in the slab with unlimited applica- tions of 22-kip wheel loads and periodic overloading with Figure A.12. Fabrication details and completed waffle slab.

217 34-kip wheel loads. The results of the testing indicate that the safety factor for fatigue with 100 years of standard loads is about 1.25 and for elevated loads is about 1.15. Thus, the sys- tem is adequate for fatigue for a long design life. ABC Opportunity Several research projects demonstrate the viability of UHPFRC bridge decks. With this new material, significant weight sav- ings and a promise of long-term durability are shown to be possible because of the excellent environmental perfor- mance of the UHPFRC material. The long spanning capabil- ity of the material opens the possibility of wide girder spacing (including the viability of two-girder systems), lighter beams, longer spans, etc., as byproducts of the weight savings. These decks can easily be made composite with steel or concrete beams simply by filling one of the cells over the beam lines. Project Title: FRP Grid Reinforced Concrete Decks Citation: Bank et al., 2006; Matta et al., 2006; Matta et al., 2008 ABC Design Features: Design of a steel-free concrete bridge deck ABC Construction Features: Rapid construction through use of FRP grids as internal reinforcing Project Description Multiple research projects focus on the use of bidirectional, double-mat FRP grids to replace the internal reinforcing of reinforced concrete deck slabs. The work of two such projects is described here. The first project, using bidirectional grids coupled with a stay-in-place forming system, is described in two articles: Matta et al. (2006) is a summary article high- lighting the construction process; Matta et al. (2008) is the full research report containing recommended construction specifications and detailed design calculations. The FRP grid consists of I-beam-shaped bars running in the direction perpendicular to traffic (for main bending strength), while the I-sections are drilled and round rods are passed through in the longitudinal direction. Spacers are used to keep the mats 100 mm apart. Attached to the bottom mat of the grid is a continuous FRP faceplate that also serves as the stay-in-place forming system. A four-span bridge in Green County, Missouri, with a total length of 144 ft and a width of 24 ft, was decked by using this system of prefabri- cated panels. The panels are 24 ft wide and 8 ft long yet weigh only 900 lb—or less than 5 psf for a system that includes the formwork and all required reinforcing. The bridge deck construction time for a typical cast-in- place operation is estimated as 3 weeks for this structure; that includes the time required to place forms, place and tie the four mats of steel rebar, and pour the deck. The Greene County project was completed in 5 days. Placement of all 18 FRP pan- els was completed in 1 day. On Day 2, prefabricated FRP rail- ing cages were placed and tied, and the finishing machine was set up. Deck casting was done on Day 3, and the railings were formed and cast on Days 4 and 5. The cost estimate for this system, in-place, is about $45 per ft2, including $26 per ft2 for the fabrication and delivery of the prototype FRP panels. Similar research was conducted by Bank et al. (2006) for a project in Wisconsin. This project also included laboratory testing and field validation of the concept. A bridge similar in length to the Greene County project was constructed with the bidirectional grid system. However, the De Neveu Creek proj- ect in Wisconsin was a much wider structure. The total deck size for the De Neveu Creek project was 130 ft by 45 ft. For the De Neveu project, the deck panels were again fab- ricated in the full width of the bridge and in panels 8 ft long. Construction involved placement of all 18 deck panels in a single day. Panels were typically lifted to the deck in stacks of four and then maneuvered manually by four workers. Time for placement of a panel averaged 11 min per piece. The bridge was built next to a companion bridge that was also redecked but with conventional steel-reinforced deck. The data indicate the following: • Construction time: FRP—111 work-hours, steel deck— 239 work-hours; • Equipment hours: FRP—21 hours, steel—32 hours; • Bridge deck cost: FRP—$31.33 per ft2, steel—$11.54 per ft2; and • Reinforcing cost: FRP—$22.50 per ft2, steel—$3.05 per sq ft2. The authors comment that the bid cost is not a fair reflec- tion of the true costs of the system. The contractor’s bid is believed to have substantially overestimated the construction difficulty and did not reflect the rapid installation of the grids. Also, a substantial part of the grid cost was in develop- ing a spacer system to separate the decks in this prototype. The connectors were a small material component but com- prised 30% of the installed grid costs. That cost will come down for future projects with improved details and given that the R&D costs have already been incurred. ABC Opportunity Both of these projects demonstrate a rapid method of recon- structing bridge decks in the field with schedule savings of two- thirds compared with a traditional project. Although costs appear high, the same is true of all demonstration projects, and capitalization of some of those costs plus familiarity in the

218 construction community will bring the price down. Whether this grid system will equal steel-reinforced decks in cost is unknown. However, it involves a simpler and much faster installation and provides a higher degree of durability. Most likely, these grid systems could also be used as internal rein- forcing for precast deck panels—if such a system were desired. That would likely be the most rapid method of installation but would involve lifting much heavier pieces as well. A com- bination of FRP grids, precast panels, and high-performance concrete (or regular concrete with HPC/UHPFRC overlays) would result in a dramatic advancement in bridge deck tech- nology yet not require any real changes in construction prac- tices or equipment. Project Title: FRP Reinforced Concrete UHPFRC Decks Citation: Perry, Scalzo, and Weiss, 2007 ABC Design Features: Design of an FRP reinforced con- crete bridge deck with UHPFRC closure pours ABC Construction Features: Rapid construction through use of FRP grids as internal reinforcing Project Description In an innovative use of ABC systems (precast decks) and new materials (UHPFRC and FRP rebar), Perry et al. (2007) describe the staged reconstruction of a bridge deck carrying Ontario Highway 11 over the Canadian National Railway at Rainy Lake, Ontario. The existing bridge, roughly 24 m long and 11 m wide; construction was staged in half widths. To accelerate construction, precast full-depth concrete panels were selected for this project. Several innovations were implemented for this project. First, the use of precast full-depth deck panels, though grow- ing in popularity, is still considered an innovative method of construction. Second, because of concerns about durability in the deck panels, the owner, Ontario Ministry of Transpor- tation (MTO), chose to provide a top mat reinforced with bidirectional glass fiber reinforced polymer (GFRP) reinforc- ing bars; the bottom mat was reinforced with steel reinforcing bars. The panels were constructed of conventional 35-MPa concrete (7 ksi). An additional feature, which will be described in greater detail, is the use of UHPFRC materials for the shear stud blockouts, transverse joints, and new longitudinal joint configuration. The nuances of the innovative details are best understood through a comparison with traditional details. The Utah DOT recently prepared standard precast deck panel plans, and those plans represent current thinking on the design and detailing of closure pour details. The closure pour must provide flexural and shear continuity of the deck. The issue is the required development length of deck reinforcing, traditionally epoxy reinforced, 100% of it spliced at a single plane, and typically designed to near its maximum capacity. That combination of factors causes long lap slice lengths; and that requires a correspondingly large closure pour in the field. A similar situation exists for the transverse joints between panels. Except for single-span bridges where states have allowed simple unreinforced female-to-female joints to exist between panels, development of continuous reinforcing steel in these joints has also required wide joints to lap-splice the longitudinal reinforcement. This is particularly critical for continuous structures in negative bending, although pre- stressed concrete structures under long-term creep and shrinkage camber growth can place a deck slab in tension even in simple-span conditions. At its best, the traditional system requires the use of female-to-female joints without reinforcement (simple spans only) or, more likely, with wide transverse joints (conventionally reinforced) or tight joints with the need for longitudinal posttensioning to control cracking and provide durability. The work of the NCHRP 12-65 project, resulting in the publication of NCHRP Report 584 (Badie and Tadros, 2008), addresses some of these details with new “non-prestressed” continuity details between panels, which are quite elabo- rate. The Rainy Lake project takes advantage of the unique material qualities of the GFRP rebar and UHPFRC joint to create a simple detail for both the traditional transverse joints between panels and the sometimes required longitudinal closure pour for staged construction. On the combined basis of Canadian and Swedish research, it was determined that the full strength of the reinforcing bars could be developed with as little as 190 mm (about 8 in.) embedded in the UHPFRC material used for these joints. The joint material, Ductal made by LaFarge, has a 48-hour com- pressive strength of 100 MPa and 28-day strength of 140 MPa. Equally impressive is the flexural strength of 30 MPa. This material is highly durable and has equally impressive perme- ability and carbonation characteristics. Thus, it is well suited for critical applications such as closure pours, long known to be a weak link in structural performance. The longitudinal joint was placed at a critical stress loca- tion, a maximum negative moment region on top of a beam flange. Yet the small size of the joint is large enough to fully develop the strength of the lapped connection, and the flex- ural strength of the material is far greater than any traditional reinforced concrete that would be used in a similar applica- tion. Use of the top-mat GFRP rebar was simply another level of corrosion protection and long-term durability desired by the owner. It does not appear to be a fundamental require- ment of the system, though pull-out strength tests of steel bars would need to be conducted to substantiate any changes in the joint design.

219 Construction of the deck was otherwise similar to other pre- cast concrete full-depth deck panel projects. The panels were leveled, and the Evazote-filled joints were tested for watertight- ness. Following that inspection, the joints were grouted, shear pockets were filled, and the haunch constructed—all with the Ductal material. The material was completely batched on site with a 0.2-m3 portable mixer and delivered to the various loca- tions along the bridge with a small powered buggy. The entire operation took approximately 8 hr to complete. The deck was left to cure for 48 hr, any high spots were removed by grind- ing, and the bridge traffic moved to the deck for completion of the other half of the bridge. ABC Opportunity The opportunity to use this system of deck construction in ABC applications is fairly obvious. The system builds incre- mentally on past successful projects and research in the area of full-depth deck panel design and construction. In tradi- tional applications, full-depth deck panels either were pro- vided with large cast-in-place joints (filled with traditional concretes) or were limited to simple-span applications or posttensioned decks for proper durability. The Canadian deck project clearly demonstrates that existing concepts can be enhanced with the addition of new materials. The specific enhancement in this project is the use of durable and high- strength UHPFRC materials for all closure pours and shear pocket grouting for the deck system. The addition of the GFRP rebar as an upper mat provides an even greater overall protection to the deck system. Project Title: Sandwich Plate Steel Decks Citation: Kennedy, Dorton, and Alexander, 2002; Kennedy, Ferro, and Dorton, 2005; Knoblauch, 2004 ABC Design Features: ABC Construction Features: Rapidly installed lightweight deck system Project Description The sandwich plate system is a novel deck system concept comprising a sandwich of steel faceplates and an elastomer core. The deck system was developed for more traditional, stiffened plate applications such as in the shipbuilding indus- try. It is a product of collaboration between a subsidiary of the BASF Group and Intelligent Engineering. Kennedy et al. (2005) detail the design of a sandwich plate system to meet the intent of the Canadian Highway Bridge Design Code. Design rules are proposed to cover missing aspects of the code that are unique to this configuration. The elastomer core’s role is to separate the steel faceplates and thus provide a high bending stiffness to the system. It provides no strength since it is a weak material. But given its intimate contact with the plates, it prevents local buckling of the plates and provides damping. The panels are simply fabricated. Two steel plates are spaced the required distance apart with containment plates on all four sides; the void is injected with an expanding elastomer. The panels attain composite action with the main stringers by bolting longitudinal angles installed under the panels to the main stringers. This process can take some time, but the time is not much different from what is needed to grout shear pockets and wait for those to cure for several days. Another field process is to groove weld the seams between adjacent panels. The decks are thin. For the Shenley Bridge, each face- plate measured 6 mm (<.25 in.) and the core was 38 mm (1.5 in.) for a total thickness of 50 mm (about 2 in.). The deck plates are attached to cold-formed angles that are 250 mm tall (10 in.). The weight of the system is 35 psf, or about 35% of a typical concrete deck and comparable to other innovative deck solutions such as FRP, grid decks, exodermic decks, etc. for this span range. ABC Opportunity This product adds another solution to the list of deck options for lightweight deck construction. Its simple construction of spaced faceplates should lead to economical fabrication. The expense of the system appears to be in the connection to the beams. In a deck replacement in particular, extensive field drilling would be required. It may be possible to modify this system to use shear pocket blockouts and conventional shear stud technology for retrofit projects. Bridge Substructure Concepts This section focuses on components that can be used to accel- erate bridge substructure construction. Bridge substructure and foundation construction are always on the critical path in bridge construction and reconstruction project. Whereas superstructure components can be made in advance and read- ied for installation when needed, the construction of founda- tions and substructure units is a linear process. All time savings in this area immediately translate into total project time sav- ings. A number of techniques will be presented that allow combinations of foundation construction to occur without hindering existing operations and for bridge piers and abut- ments to be built quickly once the site becomes available. The most-common solutions involve the use of pre- fabricated substructure elements. Cast-in-place construction must be eliminated to the maximum extent possible to expe- dite project delivery. Several concepts have been proven in the

220 past and should be encouraged. One is the continued use of pile bents for bridge piers or spill-through abutments. Con- structed rapidly of driven piles, pile bents represent a rapid construction technique that no longer requires cast-in-place construction. Traditionally built of driven piles and cast-in- place cap beams, advancements in precasting and connection design now allow for precast caps with grouted connections to the piles. The pockets are made somewhat larger than the piles to allow for some driving tolerance. Another solution (not specifically described here but effective for short abut- ments) is a soldier-pile or sheet-pile abutment. With short projecting lengths from the ground, the soldier or sheet piles are outfitted with a cap for beam support. In the case of the soldier piles, a concrete panel lagging system is used to retain the earth; for an interlocking sheet-pile wall, no additional elements are required. For taller walls, tiebacks can be used, but that affects the economics of the system, so other systems (e.g., short cantilevered concrete abutments) may be more cost competitive. Additional representative substructure sys- tems are described below. One substructure design used on accelerated bridge proj- ects is a precast abutment and pier cap component supported by driven piles. This design worked well for a bridge built in Boone County, Iowa. On that project, steel H-piles were driven within a given positional tolerance. Precast abutment and pier cap elements were then placed on the piles. Indi- vidual blockouts were formed for each pile, and each pier cap section was installed in less than 30 min. A grouted connec- tion was then used to bond the cap segments to the piers. A high early strength concrete was used in the blockouts (Bowers et al., 2007). Another precast pier cap was used on a bridge in Black Hawk County, Iowa, combining concrete and steel. In that design, a W-section was laid on its side with the flanges set vertically. Concrete, along with reinforcing steel, was then cast around the steel from the web up. That created a concrete pier cap with the properties of steel in the bottom half. Before casting, holes were torched into the steel, creating composite action between the steel and concrete. The flanges protruding from the bottom of the pier cap can be used to set it onto an abutment wall (Wineland et al., 2007). Pier and abutment caps are not the only precast compo- nents used in a substructure. Abutment walls can also be pre- fabricated and brought to the site to reduce construction time. In this arrangement, a precast, posttensioned abutment wall is placed on a concrete footing. The footing can be installed by using either cast-in-place or precast methods. The precast abutment segment is then lowered onto the foot- ing. Dowels protruding from the footing are typically used as the connections between the footing and abutment wall. In some cases the dowels are secured to posttensioning strand through the abutment wall. On one project, cast-in-place footings were chosen because of the flat bearing surface they provide (Scanlon et al., 2002). All of the substructure designs previously described can be combined to produce a completely precast substructure sys- tem. One such design was developed by the New Hampshire DOT. The foundation of that design is composed of a precast footing. A flowable grout bed surface was placed below the footing to provide a flat bearing surface. Leveling screws were installed at the corners to ensure the footing was level. A pre- cast cantilevered abutment was then placed on top of the footing. All vertical joints in the system were sealed with grouted shear keys. Grouted splicers connected the abutment wall to the footing. This system has decreased substructure construction time from 1 month to less than 2 days (Stamnas and Whittemore, 2005). An innovative pier construction method used in Japan has also been shown to accelerate substructure construction. This method is called the Sumitomo precast form for resisting earthquakes and for rapid construction, or SPER for short. The SPER system uses stay-in-place concrete forms to con- struct piers. The forms are stacked on top of each other at the site and sealed with epoxy joints. Cast-in-place concrete is used to fill the forms. Shorter piers are composed of solid forms, and taller piers are made with hollow forms. The SPER system has proven to be 60% to 70% faster than conventional cast-in-place substructure systems (Russell et al., 2005). For shorter piers, the segments are stacked on top of each other, epoxied together, and then filled with cast-in-place concrete, creating a solid pier. For taller piers, inner and outer panels are used to create a hollow pier. For both types of piers, cross ties and couplers are used to provide transverse reinforce- ment. High-strength bars are typically used for the transverse reinforcement to reduce congestion between the panels. Cast- in-place concrete is typically used to connect the piers to the superstructure (Russell et al., 2005). Precast bridge substructures can be made earthquake resis- tant with a variety of innovative designs. The University of Nevada at Reno successfully tested an unbonded segmental precast column under high-seismic conditions. The column consisted of a footing, three hollow column segments, and a column head. The segments were constructed separately and joined together with an epoxy adhesive. Prestressing forces were then applied throughout the column cross section by using 12 prestressing strands. The column was tested by using motion comparable to the Kobe earthquake and performed successfully, exhibiting very small residual displacements and limited damage. Throughout testing, the joints between col- umn segments remained secure, but the joint at the footing experienced some disconnecting action. It was determined that any repairs necessary for the column could be completed in a short amount of time (Sanders et al., 2006).

221 Additional work was done at the University of California to determine the effects of various materials to improve the self-centering capability of columns. Residual displacements experienced during seismic loading can cause significant damage to or even failure of concrete columns. Researchers analyzed several design elements, including conventional reinforced concrete, partially prestressed concrete, longitudi- nal posttensioning, bonded and unbonded longitudinal mild reinforcing, and the use of a steel jacket. Experimentally, partially prestressed concrete performed much better than conventional reinforced concrete. Reinforced concrete and prestressed concrete experienced drift indexes of 1.0% and 0.1%, respectively. The introduction of longitudinal post- tensioning strand greatly increased the self-centering capa- bility of the column. Mahin et al. (2006) recommend using debonded longitudinal mild reinforcing in columns. Debonded reinforcing carries less strain because it is debonded from the concrete, thus decreasing the chance of fracture. Unfortu- nately, using debonded reinforcement slightly increases resid- ual displacement in the columns. The use of a steel jacket was also advantageous because it prevented spalling of concrete at the base of the column. Additional research was conducted by Amjad Aref to find the optimum steel ratio in bridge columns in seismic areas. Aref used analytical, finite element, and experimental methods of testing. Different ratios of mild steel reinforcement were added at the segment connections. It was found that, as the mild steel reinforcement ratio increases, the energy dissipation ability increased. Unfortunately, the residual displacement of the column also increased with increasing ratios of steel. Aref concluded that a steel ratio between 0.38% and 0.7% was most favorable range for seismic conditions (Aref, 2006). Several pertinent prefabricated bridge projects are sum- marized in an article by Ralls et al. (2004). For example, dur- ing construction of the Newark, New Jersey, monorail system, a unique prefabricated steel bent cap and column system was used. In that application the unique construction/fabrication methods were used to meet the rather significant design constraints at the site. The authors also describe the Loop 340 bridges over I-35 near Waco, Texas, which were con- structed with precast, pre-topped U beams. The U beams were cast near the site with the top slab and outside curbs cast in place. To speed construction of the substructure, precast column shells were cast nearby and quickly erected on site. The combined system (which cost approximately 40% more than conventional construction) minimized the impacts to I-35 while also minimizing the associated environmental impact of construction. Ralls et al. also reports that the precast construction improved the aesthetics of the project as well. The possibility of seismic forces or terrorist attacks has increased the need for blast-resistant bridges. The Multi- disciplinary Center for Earthquake Engineering Research completed a study on an innovative column design sub- jected to blast-type loading. The column consisted of a con- crete column poured within a steel tube. The concrete-filled steel tubes (CFSTs) were framed by fiber-reinforced con- crete pier caps at the top and bottom. The CFST columns proved to be resistant to a range of experimental blast forces. They were both resistant and ductile, preventing any dam- age to the pier caps. In addition to being a seismic-resistant design, CFST columns can be constructed in an accelerated manner (Bruneau et al., 2006). A review of current practice related to precast bent cap systems for seismic regions was conducted as part of NCHRP Project 12-74. Although the overall project is expected to be completed at the end of 2009, an interim publication (Toboloski et al., 2006) summarized the state- of-the-practice and the work needed to provide designers with proper tools. The investigators found that precast bent caps have been used by almost half of U.S. states and in other countries and continents. Further, more than 60 unique details could be classified as being used as part of either an integral or non-integral connection, with the greatest num- ber of details being of the non-integral type. Most use of pre- cast bent caps has occurred in low- to nonseismic regions. Most non-integral bent caps use some type of cap pocket. Generally, the investigators found that these types of connec- tions are not suitable for seismic activity resistance. Of prin- cipal interest, the investigators noted the inherent difficulty in actually achieving an integral connection when using precast construction. Realizing that the ultimate goal of Project 12-74 is to develop tools that are useful to designers, the investiga- tors will use a combination of analytical and experimental work. The project deliverables are planned to include design and construction specifications (LRFD format), design exam- ples, standard details, and an implementation plan. Yen and Aref (2007) provide a summary of ongoing work related to the development of construction details for seis- mic zones. Their paper specifically discusses the development of a model for predicting the behavior of segmental piers. The authors describe the developed model as a two-stage approach. The first stage consists of “pre-decompression” behavior during which a bridge column behaves like a con- ventional column with a fixed based as there is no opening between segments. During the post-decompression stage, the joints begin to open and are no longer in full contact. This behavior creates analysis difficulties because of the lack of strain compatibility. Further, a finite element model was developed to investigate the seismic behavior in detail. The researchers found that the simplified model is consistent with the one used in conventional bridge columns in that the model can be used with conventional techniques. The finite element model was found to help the engineer understand the behavior at a very detailed level.

222 A unique technique which employs a steel–concrete com- posite socket joint has been used to connect the foundation steel piling directly to the steel pier elements quickly without the need of a concrete foundation (Yoshida and Horiguchi, 2005; Takashima et al., 2005). To the authors’ knowledge, this procedure has not been used in highway bridges. However, it has been used in railroad bridges and has been extensively tested in the laboratory. The procedure has also been ana- lyzed by using finite element method (FEM) analyses to develop structural design data. The connection has flexibility to account for misaligned piling. Through testing of this con- nection, it was determined that shear connectors were not required between the steel shell and the core concrete. Various substructure and foundation ABC solutions are presented in the following case studies. For a list of additional details and concepts, consult FHWA Connection Details for Prefabricated Bridge Elements and Systems (Culmo, 2009). Project Title: Integral Steel Box-Beam Pier Caps Citation: Wassef, Davis, Sritharan, Vander Werff, Abendroth, Redmond, and Greimann, 2004 ABC Design Features: ABC Construction Features: Possible adaptation of inte- gral steel boxes for ABC purposes Project Description NCHRP 527 focuses on developing design recommendations and approaches and validating the performance of integral steel box girder/concrete column connections. Integral steel box connections are an attractive solution for low-clearance environments in particular—the beam and cap lie in the same vertical plane as opposed to conventional construction where the beams pass over the cap. A schematic of a bridge concept with an integral steel box pier cap from the NCHRP 527 report is provided. In this typical scheme, a multi-span plate girder bridge is spliced directly to an equal depth box girder. A full-moment connection between the box and col- umns is required for stability. This type of structure promotes beam continuity and—because of the elimination of joints— promotes ride quality, structural efficiency, and redundancy. Several connection details are provided in the report. The uppermost portion of Figure A.14 provides the beam conti- nuity detail. Given the need to transfer shear and control box distortion under torsion, the box has a rigid interior dia- phragm. Top and bottom tie plates transfer the flange forces between spans and partly into the box. Since compatibility between the stringers and box is enforced in this situation, deflection and twist of the girders generate forces in the box girder cap beam. Figures A.13 and A.14 depict the connection between the concrete column and steel cap. The test connections and pro- totype bridge designed in the NCHRP 527 report depict a box with a perforated bottom flange, allowing the column steel to penetrate the center void of the pier cap. Bounded by the four walls of the box and two internal diaphragms, this chamber will be filled with concrete to establish the final connection. To promote shear transfer between the column and cap, the top of the column is voided. A group of shear studs welded to the underside of the box eventually rest in this void and, by using a small hole in the bottom of the box, the void is grouted. This provides the horizontal shear connection with- out relying on the shear friction capacity of the vertical steel. ABC Opportunity Though not envisioned as an ABC innovation, this connection– pier concept has many ABC possibilities. Coupled, for instance, with a precast pier column, it is possible to quickly erect the column using grouted splice sleeve couplers and techniques such as SPMT or lateral sliding, to bring a two- or more-span bridge to the site, and to lower it directly onto the columns. Once the grouted connections are made, the structure is nearly complete. Whether or not the deck is installed before or after movement should be left to the conditions of the project and how much weight can be supported reliably dur- ing moving. Variations on the column-to-cap connection could also be proposed, the most apparent being the use of headed reinforcing in shallow-cap situations where develop- ment length is a potential problem. High-strength self- consolidating concrete would effectively fill this joint with little problem. Another possibility is to eliminate the studs penetrating the top of the column and simply rely on the shear Source: Wassef et al., 2004. Figure A.13. Integral bent cap concept.

223 (continued on next page) Figure A.14. Bent cap connection details.

224 friction of the longitudinal steel. For geometry control, it may make sense to stop the column just short of the required elevation and, by using friction collars and an edge band, form the last inch or two of the section after the cap is low- ered and the proper grade and cross slope are established. At that point the short void between the column and cap would be pressure grouted with a high-strength grout. Project Title: Concrete-Filled Steel Tubes for Bridge Columns Citation: Roeder and Lehman, 2008 ABC Design Features: ABC Construction Features: Rapid pier construction with concrete-filled steel tubes for bridge columns Project Description Roeder and Lehman (2008) present the results of experimen- tal testing of concrete-filled steel tubes to be used for bridge column construction. Large-diameter, thin-walled steel tubes both stay in place while forming but also serve as the verti- cal (primary) and lateral (confining/shear) steel simultane- ously. The net column diameter is less than that of a traditional reinforced concrete column. Their work also includes devel- opment of a simple and strong connection between the concrete-filled tubes (CFT) and a proposed precast concrete pier cap or footing. The proposed system allows for complete piers to be built in a matter of days instead of the weeks or months needed for typical cast-in-place construction. As part of this study, alternate connection details capable of sustaining large inelastic rotation demands of seismic loading were developed and tested—with excellent results. A simple connection between the CFT and a concrete footing was developed. In that connection, a concrete footing is con- structed in two lifts. The first lift is poured to support the steel column. This first lift has a series of hooked anchor bolts in a circle that will attach to an annular steel plate welded to the underside of the steel column. Once the column is erected and bolted to the first lift, the second lift is poured, completely enveloping the bolted connection and base of the steel column in concrete. No variations to the footing are required for the two-stage construction; the footing is conventionally rein- forced for shear and flexure. This connection performed well, but further economizing was thought to be possible, specifi- cally in the need for two-stage construction and the expensive full-penetration welded annular plate at the base of the tube. In the modified connection, a steel annular ring is still provided but only as an erection aid. The footing is cast in a single lift with a corrugated metal pipe blockout to receive the column at some future time. When the column is installed, light bolts are used to stabilize the column; the annular space Source: Wassef et al., 2004. Figure A.14. Bent cap connection details (continued).

225 between the outside of the steel column and the corrugated metal pipe is filled with a high-strength, nonshrink, fiber- reinforced cementitious grout. Simultaneously, the inside of the tube is filled with self-consolidating low-shrinkage con- crete. Completion of an entire column is likely to take just hours from start to finish, and smaller columns could be com- pleted in under an hour. To connect a pier cap, a cast-in-place pier cap can be con- structed, but for maximum speed of construction precast cap elements should be used. To connect a precast cap, a friction collar or similar device is installed at the top of the column to support the weight of the cap. The cap contains a void similar to the footing that receives the upper portion of the column. The connection between the column and cap is established by pressure grouting; small vents through the top of the cap ensure complete filling of the grout pocket. The authors cite test results that indicate the spiral-welded tubes used for the columns were more slender than typically permitted by code, with diameter-over-thickness (D/t) ratios as high as 80 and strengths of 70 ksi. This slenderness did not have a detrimental affect on the CFT performance. Test results of the bolted detail indicate drift ratios as high as 8% were attained with stable hysteretic behavior. This result was attained with a column penetration into the footing of 0.9 D, thus indi- cating that large footings are not required to anchor this detail. No significant damage of the footing was noted. The authors comment that for similar demands, concrete columns begin to spall at 6% drift, and their performance was not as good as the CFT option. Embedment of as little as 0.6 D is possible and can fully develop the yield strength of the tube. However, signifi- cant damage occurred at large displacements. The authors provide some narrative indicating that corro- sion and fatigue do not appear to be issues of any significance for these structures. Details to shed water away from the con- nection are proposed; and despite the high degree of com- pression expected in the bridge columns, no evidence suggests that fatigue of the spiral welds will be of any concern. ABC Opportunity Roeder and Lehman present an alternative to concrete col- umns for ABC construction. Because the section is com- posed of a spiral welded steel shell, erected first and then filled with concrete, the construction can be accomplished with a light crane. The external steel shell serves three pur- poses at once: external formwork, longitudinal reinforcing, and confinement/shear strength. Several proven connection alternatives are provided that develop the plastic strength of the column section while preventing damage penetration into the footing. Of these, the voided footing with fiber rein- forced concrete infill offers the simplest and most cost-effective detail. Project Title: Concrete Filled FRP Shells for Pier Con struction Citation: Li, Shi, and Mirmiran, 2008 ABC Design Features: ABC Construction Features: Rapid pier construction with concrete-filled tubes for bridge columns and caps Project Description Li et al. (2008) provide details of a concrete-filled FRP forming system to be used for pier construction (columns and caps), with the FRP serving both as formwork and as part of the required reinforcing. As with the steel shell system, external confinement is desired to enhance the ductility and perfor- mance of such frames under seismic loading. Four 1⁄6-scale two-column bridge piers were fabricated and tested. In the test frames, the columns included internal reinforc- ing. The reinforcing included longitudinal reinforcing but with a greatly reduced amount of transverse steel—only enough to control the geometry of the bar cage. The columns were con- structed with FRP tubes (glass, carbon, and a hybrid), and the FRP formwork for the pier was composed of multiple layers of fabric. Traditional push–pull cyclic load tests were conducted on the model piers. Whereas a reference reinforced concrete pier attained a drift ratio of about 3% at a lateral load of 20 kips, the specimen using external glass fiber reinforcing attained a lateral strength of nearly 40 kips, doubling the reinforced con- crete (RC) frame strength and achieving a drift ratio of 5.4%. The series of tests was conducted on a frame with mono- lithic pier-to-column joints; all concrete was cast in place. The external FRP forms were used for all elements. Follow-up was conducted with the principal investigator (PI) for this project to gather additional details that were ABC relevant. The ques- tion was whether the intent was to develop this into a system that uses externally FRP-reinforced precast sections that could be assembled in the field. The PI indicated that the intent is to make these piers of precast elements, externally confined. Testing has already been done on frames with precast caps and on pier frames with posttensioning used to connect the cap, column, and footing. Work continues on this project. The results from the precast pier system testing of these FRP-reinforced pier elements are presented by Zhu et al., (2004). This paper describes the precast variant of the system as well as the details of testing various types of connections, including simple dowel bars, embedment of the FRP tube into the footing, male and female insertion joints, and post- tensioning to join precast elements. A schematic of the pier and the various details tested is provided, and the results are briefly summarized. The posttensioned connection used to join the cap beams was the most robust and ductile connec- tion. The male and female joints did not provide adequate

226 performance due to geometric imperfections in the mating of the joints. Embedment of the tube into the footing provided performance benefits (as observed by Roeder and Lehman for steel filled tubes as well), and the embedment is recom- mended. The study concluded that internal reinforcing could be eliminated outside the connection regions. Further work on this system was suggested. We were unable to find any follow-up testing of this system. ABC Opportunity The ABC opportunity for this system is similar to the concrete- filled steel shell system described by Roeder and Lehman (2008) but uses FRP materials as the shell system. The differ- ence with this concept is that the shell forming system is used for the pier cap as well as the columns. The system has been tested with both monolithic joints and precast joints. Only reduced-scale tests appear to have been conducted. Addi- tional testing should be performed to validate this system. Availability of FRP shell materials may affect the use of this concept unless the FRP shells are laid up or spun for the spe- cific application. Project Title: UHPFRC Bridge Column Shells Citation: Brühwiler and Denarié, 2008 ABC Design Features: ABC Construction Features: Rapid pier construction with UHPFRC stay-in-place forms for bridge columns Project Description A related application is the use of prefabricated ultra-high- performance fiber-reinforced concrete (UHPFRC) shells for the rehabilitation of deteriorated bridge columns. Brühwiler and Denarié (2008) discuss the use of UHPFRC in various applications as a rehabilitation and restoration material for deteriorated structures or bridges requiring partial restora- tion. The idea is to use UHPFRC to “harden” particularly vul- nerable portions of structures, taking advantage of its high strength but, more important, low permeability and enhanced durability to protect existing structures. The concepts explored include remedial repairs such as thin bonded overlays or bar- rier rail overlay that uses UHPFRC to provide prolonged life. In an application that has potential technology transfer to new ABC bridge construction, the authors discuss the retro- fit of existing reinforced concrete columns with a UHPFRC overlay/jacket. The existing bridge columns in the case study were 40 years old and had substantial chloride contamina- tion. The rehabilitation strategy was to hydro-demolish the contaminated concrete and then install new UHPFRC shells around the existing column. The shells were 4 cm thick, cast as high as 4 m, and shipped in split halves to the job site. Once the shells were erected around the existing column, the inter- stitial void was grouted under pressure to both fill the void and bond the shell to the existing column. ABC Opportunity This concept was explored as a method of new column construction by the R04 team, including discussions with LaFarge, the North American producer of Ductal, a commer- cially available form of UHPFRC. The concept explored by the R04 team is to build on the “shell” concept used for retro- fit, or explored by others using steel or FRP shells, but for new column construction. In the proposed new column applica- tion, extruded or spun cast shells would be used as “lost forms” to construct a traditional reinforced concrete column. The UHPFRC would serve as formwork but also an extremely effective environmental barrier with low permeability and high tensile strength. Those qualities make it desirable to use as forming. The column would be built in traditional fashion; but instead of erecting steel or wood forms, the concrete shells would be used. Alternative concepts include the use of thicker walls for hollow column applications. These precast elements could then be stacked and posttensioned together for an ABC column application. At this point, solid columns composed of UHPFRC do not appear to be a cost-effective use of the material, thus the shell-type applications seem most suited for future development. MSE Abutments Though mechanically stabilized earth (MSE) abutments are not a new technology, general discussions with geotechnical engineers over time and more recent conversations with engineers specifically related to the R04 project reveal that the use of MSE abutments is an underused existing technol- ogy, proven through many decades, which should be used to accelerate bridge construction. Brabant et al. (2008) present some data to substantiate this statement. Currently, about 300 bridges annually are constructed with MSE abutments, but the opportunity is much larger. When selected, MSE abutments are chosen because of speed of construction, cost savings, and ability to be built over poor native soils. The true MSE abutment consists of facing panels and anchors placed near the beam seat level. A lift of compacted gravel is placed at this location; on top of that is set a grade beam with pedestals to support the superstructure beams. A backwall is used to restrain the backfill. In this system, the reinforced fill is used to support a spread footing (the grade beam), and the loads are carried thusly. It has been the authors’ experience, however, that many owners do not prefer this sys- tem; the concern is generally settlement. The second type of

227 MSE abutment is one in which the MSE facing is simply used to restrain the embankment fill, while the bridge has a deep foundation of either drilled shafts or more traditionally driven piles. In this mixed system, owners expect to get zero settle- ment in the completed foundation, which thereby supports the bridge in a fashion that is familiar and similar to other types of bridge foundations. MSE wall abutments offer various advantages to ABC proj- ects. They can be used in phased construction to effectively widen and replace existing bridges. Because of the way they are built from the bottom up with small equipment, they have been used in many instances for construction underneath an existing bridge to within several panels (5 ft to 10 ft) of the existing bridge envelope. Only at that point does demolition need to proceed to advance the completion of the abutments. Production rates for typical MSE walls vary but can equal 2,000 ft2 per day. With average abutments for small- to medium-span bridges ranging as high as 10,000 ft2, the total abutment construction time can be measured in just days, dramatically different from the speed of construction of a traditional cantilevered abutment on deep foundations. Once the abutment is completed, the sill and beam seat (precast in advance) can simply be installed. Geosynthetic Reinforced Soil Abutments Geosynthetic reinforced soil (GRS) abutments have been constructed in various research and field deployment proj- ects over the past two decades. Documented in various reports—including test results; analytical studies; and design, construction, and specifications information—these rein- forced soil substructures have the opportunity to change the way bridge substructures are constructed (Wu et al., 2006; Adams, 2008). In the GRS substructure concept, what appears to be simi- lar to an MSE structure is in fact a very different structural system. In an MSE substructure, the facing panels serve to anchor the straps and contain the fill and are thus an integral part of the structural system required to support the rein- forced soil mass. In the GRS system, the role of the discrete facing elements is more decorative than functional. The pri- mary stability of the system lies in the use of shallow lifts (approximately 8 in.) with geosynthetic reinforcing installed between each lift for strength and stability. The facing blocks are typical segmental block retaining wall units that are used in various forms of residential and commercial construction or split-face cinder blocks. The GRS need not be positively attached to the blocks; often, it is simply laid between courses as a friction anchor. Additionally, no pins or mechanical anchors between block courses are required since the lateral pressures exerted on the facing are minimal. Wu et al. detail various installations of such systems around the world, pro- vide the results of various physical tests of such systems, and present a comprehensive design methodology specifically to be used for GRS abutments with flexible facings (such as dry- stacked block facing units). Adams (2008) provides a detailed description of 11 bridges built in Defiance County, Ohio. Each bridge used local forces and conventional equipment and had bid cost savings of approximately 25% over a conventional project. The schedule compression was substantial, with a typical bridge requiring only 2 to 3 weeks of normal construction by a five-man team, the fastest bridge being competed in 10 days. Unlike some ABC solutions that we accept as costing a premium to justify their speed, GRS construction has proven to result in a prod- uct constructed both faster and cheaper. The in-service expe- rience also indicates excellent ride quality after opening and stable performance even in the event of a flood as each of the new bridges has been nearly completely inundated by recent flooding events. Each of the 11 bridges constructed by Defiance County was constructed without the use of any cast-in-place concrete, a major advantage to rapid construction. In addition to the use of the GRS abutment technology, the bridges were con- structed by using traditional and readily available adjacent concrete box beams. On Bowman Road bridge, the beams sit directly on top of the compacted abutment backfill. Addi- tional GRS fill is placed behind the ends of the beam in the bridge approach area, and continuous asphalt concrete pav- ing is used to bridge the approach roadway and bridge pave- ment. In-service inspection of the bridges reveals minimal settlement and excellent performance of the bridge deck and approach paving. Performance is considered better than for conventionally constructed bridges with regular abutments and approach pavement sections. ABC Opportunity The ABC opportunity for this project is to replace the time- consuming process of cast-in-place abutment construction with a rapidly constructed integrated bridge superstructure, substructure, and approach paving system entirely completed in several weeks with small construction equipment and local forces or small local contractors. It demonstrates a complete systems approach to bridge replacement that could be applied to literally thousands of bridges. The system could just as well be used for abutments for bridges crossing land features as for stream crossings. The system could also be used in bridge widening projects as well as to extend the width of existing abutments in a low-impact way. Potentially, bridge piers could be constructed by using this system. This type of foundation can be used in marginal soils since a relatively low

228 bearing pressure is applied. Additionally, the strength of the confined soils is sufficient to support most common bridges. The bearing capacity of the reinforced soils varies from 1.9 tsf to 2.9 tsf for a 5-ft-wide footing/sill to as high as 4.5 tsf to 6.9 tsf for a 2-ft-wide sill/abutment beam seat. These numbers convert to span capabilities for usual structure self-weights and design live loads that range from a low of 120 ft to nearly 200 ft depending on backfill quality and sill width. Thus, this abutment type is readily adaptable to the vast majority of potential bridge replacements in the United States. Bridge Movement for ABC projects The previous discussion focused primarily on prefabrication solutions for ABC projects. Those concepts make use of exist- ing contractor equipment and are already in deployment around the country to some degree. Acceptance of prefabri- cated solutions will likely grow substantially, and the team expects that future ABC projects will make significant use of some of those concepts. Another class of ABC projects consists of those that involve transporting large components or completed bridges by using various movement techniques. These techniques include self- propelled modular transporters (SPMT); bridge sliding, skid- ding, and rolling with the use of various sliding surface movement methods; and incremental launching. The follow- ing discussion focuses on these methods since they are dem- onstrated techniques with broad applicability. Other methods of erection are presented as well, including the use of gantry systems and several unique structure erection concepts. Notably, some topics are not covered. These include bridge rotation, which is the construction of a bridge on a rotating foundation, usually over a navigable waterway, and then rota- tion of the bridge to a closed position when completed. First, many international articles discuss this method, which is not a new technique. Double-leaf swing bridges have been con- structed this way since the concept was invented centuries ago. Second, bridge rotation is perceived as a niche applica- tion that does not have the wider range of applicability of some other techniques. Another technique that this report omits is the floating of bridges into place using barges. Again, the concept is neither new nor broadly applied. The team dis- covered one interesting concept worth mentioning briefly: segment floating. A Chinese bridge was recently constructed such that the center-span steel box girder, 100 m long, was fabricated more than 1,000 miles from the bridge site. It was hermetically sealed and floated like a barge itself down river to the bridge site. At that point, strand jacks were used to raise the section into place. Though innovative and unique, poten- tial applications are few and far between. Finally, one recent and unique article worth mentioning describes arch bridge construction that uses an accelerated construction method with chorded arches (Zhou et al., 2008). Each highlighted movement technology is first presented, then followed by sample cases of its use. Design Practices One of the concerns about bridge movement by various end users is overloading or in some way damaging the bridge dur- ing the process. The R04 team is aware of two European doc- uments on this subject, CUR 68 and CUR 81 authored by de Boer, but to date they have not been made available for our review and use. (English versions of these documents do not exist.) Having discussed design criteria with multiple agencies, the team has concluded that no consistent set of standards is applied. Contacts at the French agency SETRA (Service d’Etudes Techniques des Routes et Autoroutes) did provide some clarity on this issue—at least from the perspec- tive of their agency’s practices. Generally, the rules for strength and stability of structures being moved are the same as for the bridge in its final position. The Eurocode does provide some guidance. Some stability checks are specific (e.g., web buck- ling due to a patch load between two vertical stiffeners in EN 1993-1-5, or overall stability in the absence of the concrete slab in the case of a composite bridge). Checks are applied at the ultimate and service load conditions, and the checks are required to take into account the erection history. The loads during erection are defined in EN 1991-1-6. Some loads could be different due to their short application period (wind, for example, in the case of a 2-day launching). Special care is taken for geometrical tolerances during erection, uncertainties on some loads (e.g., counterweights), and any special equipment supported off the partial structure. SPMT Bridge Construction As an outgrowth of the FHWA scanning tour of prefabricated bridge elements and systems (PBES), subsequent AASHTO Technology Implementation Group (TIG) efforts, FHWA promotion, and the success of several high-profile emergency replacement and demonstration projects, SPMTs continue to gain popularity and use in the United States for bridge move- ment (Ralls et al., 2005b). The findings from the FHWA PBES scanning tour can likely be considered the first concerted step in accelerating bridge construction through the use of mobile bridge technologies. The most prominently mentioned tech- nology in the PBES scanning tour report is the SPMT. The PBES scanning tour report documents many interesting proj- ects (although the report lacks details because of its overview approach). The report identifies several bridge construction projects, including two multi-span bridge projects that were placed with SPMTs. The findings are significant because these types of moves have not yet been tried in the United States.

229 SPMTs, though somewhat new to the U.S. bridge building market, are becoming increasingly familiar to owners, engi- neers, and contractors. Numerous conferences, presentations, and publications in the past few years have drawn consider- able attention to unique projects whose construction was made possible (or caused less impact) through use of SPMTs. For example, self-propelled trailers have the ability to lift very heavy loads, turn them to respond to site constraints, and vertically and laterally position them with precision. Several projects are highlighted here, and excellent references provide substantial guidance to end users (owners, engineers, con- tractors) on the use of SPMTs for bridge construction proj- ects. Additionally, since demand is still relatively low, the SPMT heavy-lift firms continue to provide specialized sup- port and guidance for specific applications. In 2007, in response to an increasing interest in SPMT use for bridge construction, the FHWA published the Manual on Use of Self-Propelled Modular Transporters to Remove and Replace Bridges (FHWA, 2007d). An excellent primer that pro- vides general characteristics of SPMTs (which vary by manu- facturer), this document is available in electronic form from the FHWA website. The manual presents important charac- teristics, such as length (varying from 20 ft to 30 ft for a 4- to 6-axle unit), width (8 ft to 10 ft), height (platform height of 4 ft), vertical travel or adjustment capability (as much as 2 ft), axle capacities of 26 to 33 tons (useful capacity is less because of the weight of shoring, etc.). It provides an excellent defini- tion of the “usual parameters” so that a designer or owner can assess how an SPMT lift might be configured or proposed for a specific project. Given the geometric characteristics of the trailers and their load capacities, an “as-designed” lift could be designed by the engineer of record, understanding that the contractor might still choose another approach. The manual also presents the benefits and costs of SPMT projects. The benefits include items such as reduced traffic disruption and enhanced safety and quality due to off-site fabrication; the costs include those of equipment rental as well as cost savings measured by a reduction in user costs. In a section on design, the manual highlights several key issues. First among the considerations for SPMT use is the issue of adequate space and conditions in which to make the move. Following those are issues related to soils capacity. Given the high loads that are transported on pneumatic tires, high contact pressures are generated. These can be in the range of 1,500 psf to 2,000 psf. Adequate ground capacity must be available. Otherwise, ground improvements such as the addi- tion of a gravel haul road or steel plating of the traveled way might be considered for softer soils. A significant portion of the manual relates to contracting considerations and guidance for the owner. Included as a starting point are sample user cost models to assess the worthiness of SPMT use, sample specifications for SPMT projects, and sample project acceleration specifications— such as incentives and disincentives, bonuses, lane rental, value engineering, and partnering. Other sample documents, such as traffic control and structure movement path plans, are provided as well. Under a contract with Corven Engineers, the Utah DOT became the first state to develop a specific manual for use on state-sponsored SPMT projects. The objective of the manual is to assist the department in its rapid transformation to ABC deployment. Given that this is a new technology, with multiple layers of contracting, the manual was developed to describe the various elements of an SPMT project, the roles and responsibilities of various parties, engineering require- ments for certain critical operations, and to provide guidance on geometry control during various movement phases (Utah Department of Transportation, 2008e). The title is Manual for the Moving of Utah Bridges Using Self Propelled Modular Transporters (SPMTs), and a copy is available on the Utah DOT website. The front matter of the Utah DOT manual primarily pro- vides guidance to agencies; although it focuses on the Utah DOT, it could be considered standard guidance for any agency considering an SPMT project. The type of information pro- vided relates to the interaction of various parties, including the state agency, bridge contractor, engineer of record, spe- cialty moving firm, moving firm’s specialty engineer, and other parties that might be involved. The process by which these individuals interact and in what contractual forms are presented. Structural design considerations are provided primarily for the use of the engineer of record. A challenge to date has been providing analysis and stress and stability requirements for bridges to be moved in place with SPMTs. This section recommends critical stages that should be evaluated. The analysis is based on an assumed lifting location. A method is proposed for evaluating stresses and deflections in the struc- ture as it is lifted from its supports. Relevant sections of the LRFD specifications are referenced. These include specific recommendations for controlling stresses in already-cast concrete—such as might be the case if a bridge with a deck already cast on top is moved. Typical lifting locations are some distance in from the permanent bearings; that places portions of an already-cast deck and railing in tension during the lifting location since they cantilever out past the tempo- rary supports. Guidance is also provided to the contractor’s specialty engineer in the area of load recommendations for the design of strong and stable falsework on which to con- struct the initial bridge. Some guidance is also provided for the specialty mover. Specialty movers and bridge contractors may need to per- form certain activities to support the operations of the other party. Generally, the specialty mover relies on the bridge

230 contractor to provide various support functions. The manual presents information on some of these interactions. For instance, a specialty mover may require the bridge contractor and his specialty engineer to design the temporary support structures. Sometimes that work is done by the SPMT firm. This step should be clarified between the two parties. Addi- tionally, the bridge contractor should ensure that a proper traffic control plan is in place for the time of the move; that a movement path of proper grades, turning radii, and soils bearing capacity is provided; and finally, that a large enough staging area is available to build the bridge, stage the moving equipment, and complete the move. Geometry control of the bridge during moving is critical in that excessive twisting of the structure (e.g., one of the cor- ners departing from the plane of the deck) can result in deck cracking, excessive girder stresses, cross frame or diaphragm distress, or other undesirable behaviors. A finite element study of structure twist was performed to determine what stresses are induced by varying amounts of settlement or change in planarity of one of the bridge corners. Recommen- dations for structural twist of W/200 to W/300 are provided, where W is the width between edge beams. The maximum permissible twist should be limited to 2 in. to 3 in. (Note that in conversations between the team and SPMT manufacturers about their ability to meet these tolerances, the manufactur- ers all said this standard was generous and easy to meet.) Several appendices to the Corven–Utah DOT report pro- vide useful information, such as contact information for heavy lift ers and a sample set of plans for an SPMT move- ment project—including staging area and movement path drawings. Project Title: Graves Avenue Bridge, Florida Citation: Trimbath, 2006 ABC Design Features: Precast concrete girder and deck system ABC Construction Features: SPMTs used to move out old bridge and move in new bridge Project Description The Graves Avenue Bridge is two-span bridge composed of a precast concrete beam and deck system. The two-span bridge extended 286 ft across Interstate 4. Each span is 143-ft long and 59-ft wide. The total weight of each span is 1,300 tons. The existing bridge needed to be replaced with a longer span to accommodate a wider road below. The Florida DOT decided to experiment with the use of self-propelled modular transporters (SPMT) for the removal and installation stages of the project. SPMTs are computer-controlled multi-axle trailers that can lift precast elements into place. They have been used extensively in Europe and have proven to be a reli- able method of accelerated construction in the United States as well, with successful implementation in Washington State, Oregon, Utah, and elsewhere. Completed in 2006, the Graves Avenue Bridge replacement project was one of the first inter- state overpass projects to employ SPMT. By using SPMTs, traffic disruption was reduced from 32 days for a conven- tional bridge replacement project to just under 4 days. ABC Opportunity The Graves Avenue Bridge in Florida was a good candidate for accelerated construction for a variety of reasons. The bridge is located very near a high school, and the project had to be finished before the start of the school year. Also, the Florida DOT wanted to reduce the effect on traffic as much as possible, thereby reducing road-user costs. Removal and replacement of the existing bridge spans over the interstate was accomplished quickly and with little interruption to traf- fic. The spans were removed one at a time by using two SPMT trailers. Each span was raised 6 in. from its supports and then removed from the site. The removal process was completed on two separate nights without blocking traffic. Once the deck was fully removed, eight trailers were used to raise the new spans. The contractor constructed a temporary fabrica- tion yard close to the bridge site. The spans were then brought to the site and placed at their final resting locations. The erec- tion stage was completed in less than 4 days. Project Title: I-10 Louisiana Twin Span Bridges, New Orleans, Louisiana Citation: Fossier, 2005 ABC Design Features: Original precast units replaced; where original units were lost, Acrow steel truss was used ABC Construction Features: Staged construction and SPMTs used Project Description The infrastructure of New Orleans and the surrounding areas took a significant blow during the Hurricane Katrina disaster of 2005. The I-10 Bridge over Lake Pontchartrain was hit particularly hard by the storm. During the surge, a total of 473 spans were either moved or pushed into the lake. Because this bridge was part of one of the few major routes in and out of New Orleans, it had to be repaired immediately. The Louisiana DOTD used a staged construction schedule and SPMTs to restore traffic across the bridge. The I-10 Bridge over Lake Pontchartrain is a twin-span bridge crossing a total distance of 5.4 miles in 436 spans. Most of the spans are low-level and are composed of precast,

231 prestressed concrete girders. Three spans at the north side of the lake are high-level spans and provide access for marine traffic. Those spans are longer and are composed of compos- ite-steel plate girders. The substructure is a cast-in-place rein- forced concrete cap supported by precast, prestressed cylinder piles. Each bridge carries two lanes of traffic. Following the storm, 170 eastbound spans and 303 west- bound spans were displaced. Of those, 38 eastbound spans and 26 westbound spans were completely submerged in the lake. In addition to the span displacements, nine cylinder piles were damaged by the collapsing spans. The department was unable to reuse the spans that were completely sub- merged in Lake Pontchartrain. To replace those spans the contractor used temporary “Acrow” steel truss bridges. The Louisiana DOTD had to move quickly to restore traffic through the area. They made use of accelerated contracting and construction techniques to facilitate a rapid repair. The bridge was partially opened only 47 days after the storm and completely open to traffic a few months later. ABC Opportunity Immediately after investigation of the damage, the Louisiana DOTD went to work on the contract. All plans and specifica- tions were completed in 1 week, and the bids were submitted 12 days after the storm. To minimize traffic disruption, the decision was made to replace the eastbound bridge first because it had fewer spans. That bridge would then be open to traffic, and the westbound bridge would be repaired. The contractor was given 45 days to complete the first phase of construction but finished in 28 days. The second phase of construction was scheduled for 120 days. The spans were repositioned in their original loca- tions by using a combination of a barge and SPMT. Project Title: Yokogawa Bridge Corporation—YS Quick Bridge Citation: Yokogawa Bridge Corporation, 2007 ABC Design Features: ABC Construction Features: Movement of completed bridge systems to create grade-separated intersections at busy road crossings Project Description Continued traffic congestion problems in Japan demand innovative construction approaches. As one solution to this problem, at-grade intersections are being partially replaced with elevated through lanes; surface construction is reserved for local lanes and turning movements. The traditional approach has resulted in traffic restrictions for more than 1 to 2 years, causing new traffic jams. Thus, public support for such projects has been difficult to attain. Furthermore, secur- ing temporary access rights to the land needed for construc- tion has also been difficult. The YS Quick Bridge Method was proposed to solve these problems. In this method, the superstructure is assembled in the median of the approach to the road to be elevated, near the proposed intersection. Construction is completed in an ele- vated position on falsework, including the new bridge piers. As soon as the structure components are completed, they are assembled and moved into place with SPMT or heavy haul trailer technology. A time lapse image of construction is pre- sented. Numerous innovative aspects to this project include the use of existing lanes on the approach road as a staging area. Though imposing some restrictions, this method has less impact than erecting the bridge over an active intersection. Additionally, the connection between the bridge and foun- dation involves a unique “well” connection. By using the same work zone restrictions as for the superstructure, drilled shaft foundations are constructed with the upper portions left open, protected by steel casing. The bridge is then driven to the final position and lowered into the wells—where the final connection is made between the above- and below-grade portions via a cast-in-place closure pour collar. This method of construction is called the precast reinforced concrete (PRC) well method. It is suitable to urban environments since drilled foundations create less noise, are low vibration, and are installed in more-compact situations than is a multi-element driven foundation. The shaft is drilled to within 3 m of the finished ground line. A steel casing is used throughout and extends to the ground line. The superstructure with inte- grated bridge column is delivered and then lowered into the void provided by the steel casing. In the Quick Bridge sys- tem, the bridge pier is also a large-diameter steel column. This prototype joint type has been in use for a number of years, primarily for railroad bridge construction. None- theless, physical testing of the joint was conducted. During reversed loading push–pull lateral load tests, the system attained a behavior in excess of six times the yield displace- ment of the column with significant energy absorption as evidenced by stable hysteretic response. Displacements as large as 250 mm were recorded. ABC Opportunity The opportunity afforded by this system is to move com- pleted integrated bridge superstructures and substructures into place and to rapidly form a connection to the foundation element. Though construction of the bridge will take some time, and traffic restrictions are required for this method, it may prove to be one of few viable ABC techniques for crowded urban environments.

232 Project Title: Innovate 80 Bridge Replacement Projects Citation: Utah Department of Transportation, 2008d ABC Design Features: ABC Construction Features: Use of SPMTs and skid shoe systems to deliver and launch bridges Project Description The use of SPMTs and lateral skid shoes to install bridges was combined for multiple bridges installed by the Utah DOT in the summer of 2008. A staging area was created several miles from some of the bridges. Dubbed the “bridge farm,” the area formed a com- mon location to build multiple bridges. The bridge farm was located adjacent to an interchange. As individual structures were demolished and new substructures readied for the new bridge, the completed structures were driven up and onto the Interstate, sometimes crossing old bridges, sometimes newly installed ones, until the desired location was reached. When the bridge reached the site, it was transferred from SPMTs to the skid shoe systems. Figure A.15 shows a view looking back at the new structure still supported on SPMTs and being driven toward the abutment. The span cantilevers out; a series of skid shoes with jacks were used to incremen- tally receive the load from the SPMTs and drive the bridge across the opening. Once specified temporary support loca- tions were reached, the bridge was jacked down and lowered to the permanent bearing condition. ABC Opportunity The combination of SPMT movement from a remote location and use of skidding as a variation on incremental launching was effectively used for multiple bridge replacements in Salt Lake City, Utah. The combination of multiple techniques in a project demonstrates that no one particular technology will always work or work alone. Flexibility in the planning process, coordination with contractors, and the use of heavy movers were all essential for project success. The engineer of record would not likely have contemplated this method of installa- tion or been able to develop the correct details. A project such as this can be accomplished only with early and effective con- tractor interaction that employs contracting methods such as design–build or payment-in-advance for planning studies. Structural Skidding and Sliding The use of structural skidding and sliding is not a new con- cept. Bland and Miller (1915) discussed the lateral sliding of new truss spans to replace a series of temporary spans installed after a flood washed away a large portion of a railroad bridge in Ohio. The paper details the movement of a three-span truss weighing 3,500 tons in 10 min, 17 sec—between consecutive trains. That demonstrates the capacity of low-technology equipment to move large loads quickly and stably from a sec- tion parallel to the existing bridge into the final alignment location. Several movement techniques are described in this section. Lateral sliding requires that adjacent space parallel to the existing bridge be available. Sometimes space is required not just to stage the new bridge before rolling it in but also to roll the old bridge out—unless it can be demolished in place quickly and without impact to the roll-in operations. Two jack-based systems are described. The first uses a Teflon- and-steel sliding surface lubricated with ordinary dish washing Figure A.15. SPMT span delivery. Figure A.16. Skid shoe span launching.

233 detergent for maximum efficiency of the move. The skid shoes have capacities of as much as 600 tonnes (660 tons) each, with vertical stroke capacity of as much as 2 ft. Thus, the vertical alignment and horizontal sliding are all provided in one unit. The vertical adjustability of the jacks allows skid shoe systems to be used even on less than ideal soils. The jack can be extended if any soft spots in the foundation are located under the skid shoe. With a fully automated control system, the jacks self- compensate for any geometry imperfections. Horizontal move- ment is managed through the use of standard modular steel guide channels, which can be quickly placed at a site and removed afterward. The jack responsible for moving the struc- ture locks into this channel and pushes the bridge a set amount. At the limit of the strokes capacity, the jack retracts, advances itself forward along the track and resets for the next sequence. In the second jack-based system, sliding occurs via an air pad. Best described as a hovercraft or “air hockey” game, the jack rests on a cushion of pressurized air resulting in very low frictional resistance. A launching jack pushes the structure forward, and this too results in the jack resetting itself for the next series of moves. Finally, moving structures by using the French Autoripage system is discussed. In this concept, structures slide on a con- crete sleeper mat on soils that range from excellent to poor. The new bridge is built on top of this mat. When the bridge has to be moved, a pressure grout layer of bentonite is injected between the bridge foundation system and sleeper system. Project Title: Lateral Sliding of Oakland Bay Bridge Approach Spans Citation: Warta, 2007 ABC Design Features: ABC Construction Features: Lateral skid shoe system for rapid bridge installation Project Description Following demolition of the existing approach spans, a new bridge was jacked laterally into place with the Mammoet skid shoe system over the course of a Labor Day weekend closure in 2007. The new bridge measures roughly 350 ft long by 92 ft wide and weighs more than 7,100 tons. Once the road clo- sure was established, demolition of the bridge deck was com- pleted in several hours, followed by demolition of the existing bridge columns and site clean up. Once the site was turned over to the mover, a series of skid rails were installed to extend those already in place under the new bridge section. The new bridge was jacked up off its temporary supports by skid shoes and integrated jacks, with a capacity of 660 tons each. A total of 16 jacks/shoes were used. The skid shoe system incremen- tally locks itself into the track, extends the jacking cylinder, advances the locking mechanism, and repeats the process. The lateral shift of the span was completed in a little over 2 hr, and the bridge work was completed 11 hr ahead of schedule. ABC Opportunity This project demonstrates the unique capabilities of high- capacity skid shoe systems. A bridge of considerable length, width, and height was moved laterally as a complete unit. The skid shoes allow for synchronized hydraulic elevation control of the entire structure during the move, and the incremental ratcheting movement of the jacks provides steady and consis- tent movement of the bridge. This technique is ideally suited for an installation like the Oakland Bay project where work- ing room is readily available alongside the existing bridge. In that case an entire bridge can be built and, when the road closure is established, simply lifted from its temporary bents and moved laterally. The skid shoe system does not allow for turning structures, so its use is limited to linear moves. Project Title: Lateral Sliding of Bridge on Cantilever Brackets Citation: Hebetec Engineering, Ltd., 2008 ABC Design Features: ABC Construction Features: Lateral movement system for rapid bridge installation Project Description This project involves an innovative method for replacing small to medium structures. In this example, a small bridge weighing 240 tonnes is replaced with a similar structure weighing 360 tonnes. The structure is over a small river or canal. A self-reacting framework is built around the existing wall pier. In the usual situation, the grillage would be erected beneath the pier cap(s) without any interruption to traffic above and minor or no disruptions to traffic likely below the bridge. With nighttime or off-peak closures, materials for the new bridge would be delivered by using the old structure as the delivery platform. Various forms of prefabricated elements could be employed to speed construction of the new bridge. One issue is the asymmetric loading of the pier. For large-wall piers, this type of loading may not pose a problem since their string axis strength is usually many times the demand. For narrower piers, such as typical hammerheads or single round- or square-column narrow piers, an assessment of the pier should be done with greater care. Very likely, the pier could not accommodate the highly eccentric loading. In that case, the tip of the cantilever brackets could be supported directly with a vertical prop or shoring tower. That would relieve

234 any bending moments from the system during erection and lateral sliding. The outboard cantilever should also be rein- forced in anticipation of movement of the old bridge onto the falsework. This system would likely be constructed with common materials that contractors typically have on hand. The effi- ciency of the materials being used in direct tension or com- pression allows for shapes such as H-pile (HP) sections and salvaged steel sections to be used, so a cost-effective falsework system is likely available. If needed, commercially available adjustable shoring elements could also be supplied. For addi- tional safety and load carrying capacity, the top tie element could be easily prestressed with high-strength steel rods (Dywidag or similar) to relieve some tension force from the system. However, this would generally not be required and would likely add cost without much benefit. Once ready for moving, the concept is easily implemented. The existing bridge is lifted several inches on jacks resting on the transverse girders. It is moved laterally with center hole jacks connected to high-strength bars or by strand jacks attached to a temporary multistrand tendon. Movement of the existing bridge and new bridge should be synchronized by linking the two systems, which is easily done. ABC Opportunity This system can be used in unique circumstances where removal of the old bridge by SPMT or skidding system is not possible yet a rapid removal is desired. One instance is a proj- ect analogous to the example bridge: over water but not deep enough or otherwise conducive for barge access. In the exam- ple described here, an internal reaction frame is used to carry the structural load. The existing piers must be checked for their capacity to serve as part of the removal and sliding sys- tem. Once the old bridge is relocated, it can be disassembled and removed from the project site in numerous ways. Project Title: Forges Les Eaux, French Bridge Displacement Using the Air Pad Sliding System Citation: Freyssinet, 2006 ABC Design Features: ABC Construction Features: Rapid lateral movement of a three-span bridge by using air pad sliding Project Description The Forges Les Eaux bridge in France was moved laterally with the Hebetec Air Pad Sliding (APS) system. Best described as analogous to hovercraft technology, the APS system uses a cushion of trapped air under heavy lift jacks to provide a very low coefficient of friction for moving heavy structures. Originally conceived and prototyped for use in shipyards and heavy load moving for industrial uses, the APS system also has many advantages in civil construction. With a weight of 1,300 tons, this bridge was a significant load to be moved. The project required that the bridge be shifted laterally a distance of 45 m. Because of the low coeffi- cient of friction of the movement system and quick cycle times of the small hydraulic jacks, it was moved at a rate of 20 m per hour, or a total movement time of just over 2 hours. The sliding system has modular components that are arranged as necessary to execute the required move. A total of eight APS units were provided for this lift, each with a ver- tical capacity of 3,850 kN and a jack stroke range of 280 mm. Coupled with these APS sliding jacks are standard push–pull units, each with the ability to extend the jack and then retract, ratcheting themselves forward down the track. A total of four 320-kN capacity jacks were used with a stroke capacity of 1,250 mm. These small jacks are sufficient to move this heavy structure because the low moving friction rate is less than 1%. ABC Opportunity This project, using instead an air pad “frictionless” system to move the completed bridge, demonstrates an alternative to conventional skid shoe systems. Hydraulic jacks lift and con- trol the vertical elevation of the bridge, and small jacks push the bridge along the slide tracks. Thus, the system is relatively inexpensive to mobilize and requires only a firm footing for the slide tracks themselves. Project Title: Autoripage Bridge Erection Methods Citation: Freyssinet, 2005 ABC Design Features: ABC Construction Features: Lateral sliding of fully com- pleted bridges into place Project Description The use of structural sliding can be applied to complete structural systems including bridges with complete founda- tions and pier elements. The Freyssinet Autoripage system is one such method of rapidly moving completed structures over large distances and over marginal soils. An example of this technique includes the replacement of a high-speed railway bridge in Saint Cheron, France. Given the traffic density on the rail line, bridge replacement was not viable unless completed in an accelerated manner. The deci- sion was made to cast the new bridge next to the old one. In the Autoripage system, a concrete sliding slab is first cast under the entire project site. This concrete is not structural; it is there to level the site and to provide a sliding surface. Once the slab

235 has cured, construction of the new bridge can commence. The innovation lies in the method of movement. When the bridge is to be moved, a layer of bentonite is pressure injected between the sleeper slab and the slab that connects the vari- ous foundation elements. The two concrete surfaces are lubricated to facilitate sliding; and a reaction block, located at the final bridge location, and strand jacks are used to pull the bridge into place. This bridge weighed 2,400 tons with a total length of 35 m. Three 1,000-ton jacks were used along with a strand jack to move the bridge into place. Speeds up to 13 mph have been attained with this method. For the Saint Cheron bridge, the total time for movement was 5 hours, including a substantial amount of time for excavators to remove the required embankment materials to allow the new bridge to fit in the prior opening. Because of the large sliding surfaces between the bridge and sleeper slab, very low contract pres- sures are applied. For this bridge the pressures were less than 10 psi or ¾ tsf. This makes the Autoripage system well suited for construction of bridges where poor soils are present. ABC Opportunity The Autoripage system is a unique method of sliding heavy loads particularly in instances where the soil conditions are poor and might limit the use of other removal methods with higher contact pressures. The system is unique in its use of a large “sleeper slab” cast under the entire construction site. Strand jacks are synchronized and used to move the load across the slickened surfaces. The process is similar to the use of skid shoes and other lateral sliding techniques but appears best suited for marginal soils locations Incremental Launching Bridges have been constructed with the incremental launching method (ILM) for more than 100 years. In this method of con- struction, the bridge superstructure is assembled on one side of the obstacle to be crossed, then pushed longitudinally—or launched—into its final position. The launching is typically performed in a series of increments so that additional sections can be added to the rear of the superstructure unit before sub- sequent launches. The launching method has also been applied to tied-arch or truss spans, although those are fully assembled prior to launching. The incremental launching method will never become the most economical procedure for constructing all bridges, yet it has unique advantages for complex site conditions. The ILM requires considerable analysis and design expertise but not very sophisticated construction equipment. However, the ILM may often be the most reasonable way to construct a bridge over an inaccessible or environmentally protected obstacle. When used for the appropriate project, the ILM offers sig- nificant advantages to both the owner and the contractor. These advantages include the following: • Minimal disturbance to surroundings, including environ- mentally sensitive areas; • Smaller but more concentrated area required for super- structure assembly; and • Increased worker safety, since all erection work is per- formed at a lower elevation. The ILM can be used to construct a bridge over a wide range of challenging sites that feature limited or restricted access, including those with the following characteristics: • Deep valleys or water crossings; • Complex urban environments (complex road and railway facilities); • Steep slopes or poor soil conditions, making equipment access difficult; and • Environmentally protected species or cultural resources beneath the bridge. Estimates indicate that more than 1,000 bridges have been constructed with the incremental launching method world- wide over the last 50 years. Despite its advantages, however, this method of construction has seen limited application in the United States. The Po Bridge in Italy was an early pre- cast structural bridge system that used incremental launching (Gentilini and Gentilini, 1974). As noted, the list of bridges that have used this construction technology is too long to provide here, so some selected literature is provided for mod- erate- and relatively long-span bridges, including multiple precast concrete bridges in Hong Kong (Tung et al., 1988), the Kap Shui Mun Bridge in Hong Kong (Lau and Wong, 1998), the Donghai Bridge in mainland China (Ren et al., 2005), and the Pingsheng Bridge in mainland China (Cheng, 2006). Applicability and Limitations of Incremental Launching During the launching of a bridge, the superstructure acts as a continuous beam supported on roller or sliding bearings and is transversely restrained by lateral guides that prevent drift- ing movement. Any constraint eccentricity (vertical mis- placement of launching bearings or transverse misalignment of lateral guides) will cause unintended secondary stresses and may cause launching problems such as excessive wear of bearing devices. The case studies presented herein highlight the fact that incremental launching is applicable to a wide variety of challenging bridge sites. The recent FHWA PBES scanning

236 tour of Europe and Japan identified a number of bridge launching projects for which launching was considered the most efficient solution to a difficult bridge construction problem. Ideally, a bridge intended for incremental launching would be designed along a tangent alignment in both horizontal and vertical planes to simplify fabrication and construction. How- ever, bridge sites rarely fit those ideal conditions. Although somewhat more challenging, it is possible to construct a bridge by incremental launching while maintaining a curved align- ment in either or both planes. To eliminate the relocation and adjustment of lateral bearings, however, those surfaces must remain perfectly aligned with the superstructure during launching operations, and that can only be guaranteed in the case of a common geometry. Rosignoli (1998b) states that a bridge constructed by launching must be designed with one of the following alignments: • Tangent in plan and tangent or circular in profile; • Circular in plan and horizontal in profile (no launch gradient); • Circular in plan and inclined with respect to the horizontal plane; or • Curvilinear both in plan and in profile. The geometry of curved structures and the desire for uni- form distribution of launch stresses strongly favor the use of constant-depth superstructures such as a parallel flange I-girder. A variable-depth steel superstructure can be used, with temporary steel plate or trussed extensions of the bot- tom flange. A variable-depth superstructure is greatly com- plicated by the higher dead load present during launching operations. Structural Monitoring During Construction The use of structural monitoring during construction of an incrementally launched bridge has received considerable atten- tion from both owners and university researchers. Structural performance information through monitoring can supple- ment visual observations and may provide critical alerts during the launch stages at structure locations during the launch pro- cess. It can also provide validation of the design and construc- tion process, which is useful for implementation of subsequent ILM projects. Project Title: US-20 Iowa River Bridge Citation: LaViolette and McDonald, 2003; LaViolette, Wipf, Lee, Bigelow, and Phares, 2007 ABC Design Features: Innovative use of incremental launching with steel I-girders ABC Construction Features: Erection by incremental launching to satisfy multiple environmental and geo- technical constraints Project Description The bridge consists of two parallel deck superstructures, each with five equal spans of 302 ft. A 62-ft prestressed concrete jump span is provided on each end of the steel unit. The bridge was constructed by using the incremental launching method because of a number of stringent environmental restrictions near the project. The environmental issues included endan- gered mussel species residing in the Iowa River, endangered plant species near the site, and Native American artifacts near the site. In addition, a bald eagle roosting area was identified near the site. An extensive environmental monitoring pro- gram was established and maintained during construction. To make the I-girder superstructure act as much like a tor- sionally rigid box girder as possible during launching, a stiff system of diaphragms and lateral bracing was used. A dia- phragm spacing of 23 ft was used for Spans 2 through 5, but that was reduced to 11 ft, 6 in. in the leading span that would be cantilevered during launching. The I-girders were fabri- cated from ASTM A709 Grade 50W steel; they are 11 ft deep and spaced at 12-ft centers, and the constant 7⁄8-in. web thick- ness was designed as unstiffened for steel dead load. The bridge superstructure was completely erected on steel falsework and custom-made 18-in.-diameter rollers behind the east abutment. A 146-ft-long, tapered steel launching nose was erected at the leading end of the girders and used to reduce the cantilever deflection during each launching opera- tion. After each span was launched forward, additional steel girder sections, including diaphragms and bracing, were pushed forward to land on the subsequent pier. The process Figure A.17. Incremental launching of US-20 bridge; span at rest pier location.

237 was repeated five times for each steel superstructure. After the complete launching of the eastbound girders, the falsework was removed and reinstalled to perform an identical launch- ing of the westbound superstructure. ABC Opportunity This project illustrates the ability of a determined owner, designer, and contractor to construct a unique bridge to solve challenging environmental concerns. Although the project did not specifically intend to advance the ABC agenda, it provides evidence that incremental launching can be suc- cessfully applied to larger, more complex projects. Project Title: Tiziano Bridge Launch and Shift Citation: Rosignoli and Rosignoli, 2007 ABC Design Features: Uses proven bridge system of pre- cast, posttensioned box girders ABC Construction Features: Rapid construction of complete bridge systems by using innovative combination of proven bridge movement concepts of incremental launching and transverse shifting Project Description This project involved the longitudinal launching and transverse shift of a new bridge across the Tanaro River in Alessandria, Italy. Initial plans called for a 656-ft-long, 55.4-ft-wide bridge comprising cast-in-place concrete constructed by using a movable formwork system. The Tanaro River is prone to sudden floods, however; and the girders supporting a mov- able formwork system would have significantly diminished the available hydraulic section during construction. A value engineering approach was presented that included the following: • Construction of a 663-ft-long, 26.7-ft-wide box girder by incremental launching; • Transverse shifting of the first box girder by 28.7 ft until it reached its final location, thus clearing the launch align- ment for the construction of a second box girder; • Incremental-launching construction of the second box girder; and • Joining of the twin box girders with a cast-in-place central closure curb. This proposal was accepted, and the bridge was constructed in 2001 using this innovative technique. Although launching the bridge required some specialized construction equipment, the components were inexpensive and were limited to the following: • A 184-ft-long foundation beam for the casting yard; • A 39.4-ft-long steel launching nose; • Four temporary piers constructed on single foundation shafts; and • Steel launch saddles for the transverse shifting of the deck. Notably, although this project was constructed on reinforced- concrete temporary piers, construction of future bridges on smaller, reusable steel falsework towers may be possible. Compared with the cost of a movable formwork system, the cost savings were significant. Additional savings resulted from avoiding the transportation, assembly, use, and final dismantling of the moveable formwork. Avoiding disruption of the hydraulic section of the river also proved especially important, as two floods reached the 200-year return level during construction. Casting the bridge in a highly repetitive series of 48 segments produced highly efficient construction by allowing iron workers to work independently of the carpenters. Casting the superstructure on a rigid foundation made set- ting the form geometry easy because precamber was not required. The rigid foundation also minimized formwork deflections during concrete placement and therefore helped avoid formation of cracks in the setting concrete. Both pre- camber and cracking would have been expected for a bridge constructed by using a movable formwork system because casting an entire span takes several hours and the progressive deflection of the formwork affects the setting concrete. The use of temporary launch piers reduced the number of launch tendons and increased the number of continuous ser- vice tendons for a more efficient final prestressing scheme. In turn, transverse deck shifting allowed the use of a single set of temporary launch piers and a single foundation beam for Figure A.18. Tail section and tugger beam.

238 Figure A.19. Tiziano Bridge elevation and plan views. Figure A.20. Sequence of launching and sliding operations.

239 both box girders. Because of the numerous advantages, the owner quickly approved the value engineering proposal. ABC Opportunity This project illustrates how two proven bridge movement techniques—namely, incremental launching and transverse shifting—can be combined in an innovative way to construct a complete bridge superstructure to suit complex site challenges. This concept could be utilized in a much smaller context and constructed with readily available jacking and roller equip- ment to construct a pair of bridge superstructures on a com- bined substructure with minimal effect on the traffic below. Project Title: Reggiolo Bridge Launching Citation: Rosignoli, 2007 ABC Design Features: Uses bidirectionally prestressed and posttensioned concrete waffle slab ABC Construction Features: Construction of a very short bridge by using monolithic launching to avoid interfer- ence with obstacles below bridge Project Description This project involved the launching of a small prestressed con- crete deck plate bridge that spans the Verona–Mantua Railway in Reggiolo, Italy. The railway had to remain fully operational during construction, which would have made a more tradi- tional bridge difficult to construct because of limited locations for falsework. The superstructure consists of a multicellular prestressed concrete plate spanning 85.3 ft. The bottom surface is hori- zontal in the transverse direction and slightly inclined longi- tudinally; the top surface is inclined in both directions to shed water. Thus, the total thickness of the structure varies slightly. In the transverse direction, the plate is stiffened by three 15.7-in.-thick internal diaphragms spaced at 21.3 ft and two 3.3-ft-thick end beams at the abutments. The prestress- ing level in the superstructure was quite high because of the railway authority’s request for a fully prestressed section in both the longitudinal and transverse directions under full design loads and also during launching. The prestressing included longitudinal launch tendons in the internal webs and edge beams, longitudinal tendons in the bottom slab, Figure A.21. Plan view, Reggiolo Bridge.

240 transverse tendons in the end beams, and internal diaphragms and transverse tendons in the deck slab. Three disposable prestressed concrete launching nose gird- ers were installed at the front of the deck to control negative moments by shortening the cantilevered span and to limit the risk of the deck overturning. High negative moments were alleviated by cutting the launch span in half by using a tem- porary steel pier placed near the railway. Because of its varying width, the deck alignment could not be maintained with conventional lateral guides during launch- ing. Instead, guide pins at the rear of the deck and at the launch abutment were used. The pins guided the bridge through a steel-lined recess in the underside of the deck. At the tempo- rary pier, the central guide was used to laterally stabilize the pier through the deck. The results of launching and service load analyses showed that positive moments in the longitudinal webs were gov- erned by service conditions, while launching governed nega- tive moments (which would not have existed with conventional deck construction on shoring). The use of hydraulic launch supports in the casting yard—instead of conventional continu- ous low-friction extraction rails—generated substantial sav- ings. And the launch equipment was particularly inexpensive. Because of the bridge’s modest weight—about 1,000 tons—a pair of small long-stroke launch pistons generated adequate operational speed at minimal cost. Constructing the bridge in this way—compared with erecting and dismantling the bridge on shoring, plus the higher embankments required for work- ing clearance above the railways—generated about 12% overall savings. ABC Opportunity This project illustrates how two proven bridge movement techniques—incremental launching and transverse shifting— can be combined in an innovative way to construct a complete bridge superstructure to suit complex site challenges. This concept could be used in a much smaller context and con- structed with readily available jacking and roller equipment to construct a pair of bridge superstructures on a combined sub- structure with minimal effect on the traffic below. Project Title: Bridge Launching in Milan, Italy Citation: Rosignoli, 2008 ABC Design Features: Uses proven bridge system of pre- cast, posttensioned box girders ABC Construction Features: Incremental launching over heavily traveled rail line without suspending traffic Figure A.22. Reggiolo deck section. Figure A.23. Small casting yard serviced by conventional equipment.

241 Project Description This project involved the longitudinal launching of three prestressed concrete bridges in downtown Milan, Italy. The location required that the bridges be constructed over six electrified railroads, and the project represented one of the most complex applications of ILM construction in Italy. Site constraints, including road and tramway junctions on either side of the bridge and difficult maintenance condi- tions, and the architectural requirements of urban bridges complicated the project. Reduced clearances and the opera- tional requirements of six electrified railroads suggested construction by incremental launching, as this method is particularly suitable for crossing sensitive obstacles. The road bridges comprised 18.0-m, 49.0-m, and 26.5-m continuous spans and their three-cellular section is 13.5-m wide. The base of the eastern approach embankment was erected, an off-shelf formwork was placed onto it, and the first road bridge was cast and incrementally launched over the railroads (Figure A.24). After diverting traffic onto the first road bridge and demolishing the old Palizzi Bridge, the eastern embankment was widened southward, and the sec- ond road bridge was built. The entire project—including utility relocation, the construction sequence for the bridges, and the demolition of the old bridge—was designed to reduce interference with train traffic (>300 trains a day) and heavy road traffic. Parallel activities were strictly coordinated, and the completed structures were immediately used to open new sites for the next activities. Launching the road bridges over the long central span required the use of a temporary pier in conjunction with a launching nose. The temporary pier was cable-stayed to the rear abutment to diminish the bending moment generated by launch friction in the narrow foundation block. The launch operations for the first road bridge were initially limited to a 2-hr work window each night. Initial launching operations proceeded so smoothly and safely that eventually the railroad authority allowed launching to proceed without any restric- tions on the activities and the train traffic. After the first launches, the electric catenaries were kept energized during launching operations. ABC Opportunity The project demonstrates the value of incremental launching for crossing areas that severely limit construction access. In this case study, the railway authority allowed construction to continue without any delay to rail traffic. The applicability of incremental launching to short bridges is also validated. Other Movement and Construction Techniques Other techniques are described here. Some are proven tech- niques used in the past for larger structures (e.g., overhead gantry systems), while others, such as beam launching, are not frequently used in the United States. Several other meth- ods of bridge erection are also discussed, including the com- bination of gantry and SPMT technology and a few innovative techniques. Several international articles worth noting include the use of land beam transporters to deliver very large precast components for the construction of major bridge crossings (Liu, 2009; Lu et al., 2006). A significant amount of published literature exists on the use of heavy-lift cranes for moving prefabricated bridges (or components) into place. One example is the replacement of a high-speed rail bridge in Scotland. The project required ABC techniques because the owner specifications required that the existing bridge be removed and a new structure placed during a single 5-day closure. Several sliding options were consid- ered, and the contractor and designer investigated the fea- sibility of using large mobile cranes (primarily to reduce temporary construction in the river). The replacement struc- ture is a half through plate girder bridge. The main girders were approximately 6 ft deep and spaced at approximately 13 ft. Once on site, the superstructure was assembled into four sections varying in weight from 120 tons to 170 tons. The individual units were placed by using a 1,000-ton mobile crane (Hackney et al., 2002). Project Title: Beam Erection Using Mobile Overhead Truss Citation: VSL Singapore, 2007 ABC Design Features: ABC Construction Features: Rapid erection of superstruc- ture components Figure A.24. Incremental launching and use of temporary piers.

242 Project Description Rapid erection of superstructure components is made possi- ble with the use of a small gantry system. This type of system is ideal for crowded urban environments where occupying existing ground with construction equipment such as cranes would have a severe impact on traffic flow. This design feature was developed in the 1980s and involved the construction of an elevated freeway in Jakarta, Indonesia. Beams were erected over live traffic; the road below was occupied for only the sev- eral minutes needed to deliver the beams. The gantry rests on two piers at any one time and has a crane bridge with winch lifting system attached to it. The gan- try travels to the lateral extents, picks a new beam from the delivery truck, lifts it to the proper setting height, and then traverses laterally to the correct beam line position. Setting rates in excess of one span per day are possible. The cycle time of the gantry controls the speed of erection. Once a span is set, the truss launches forward to the next span, rolling on top of the existing legs. This process repeats. ABC Opportunity This construction method could become viable and popular for routine viaduct construction. In a recent project in Tampa, Florida, a reversible high-occupancy-vehicle bridge was built with similar site constraints, down the middle of an existing roadway. However, that bridge was made of box girders. This gantry system would allow for conventional steel and concrete multi-beam bridges to be built with industrialization. The truss would not be nearly as expensive as those used for seg- mental bridges since the level of sophistication is less. How- ever, a significant investment, likely several million dollars, would be required for such a system. This system could be broken down and reused many times for similar urban via- duct construction or for situations in any environment where work restrictions below the bridge present a complication. Project Title: Full-Span Launching of Concrete Box Girders Citation: Beijing Wowjoint Machinery Company, 2008 ABC Design Features: ABC Construction Features: Highly automated construc- tion of long viaduct structures Project Description The use of full-span launching has been applied successfully on a number of recent projects in Asia. Among these are vari- ous highway and rail expansion projects in Korea and China. One example is the erection of full-span concrete boxes for the Beijing–Tianjin High Speed Railway. For this project, approximately 75 miles of new track work, largely elevated viaduct, was constructed to support trains operating in excess of 215 mph. The solution was to establish multiple casting yards along the line and, by using multiple gantries, begin construction simultaneously. The system includes precast/ posttensioned box girders (35-m spans, 900 tons) that are cast in a yard and then delivered via SPMT down portions of the already completed viaduct to the new span location. The spans are then received by the gantry, lifted forward and down to their final bearing locations, and then the gantry rolls across the newly completed span and establishes its new position. On average, two to three spans were completed in a given work shift for this project. ABC Opportunity This equipment affords the opportunity to radically change the construction of longer viaduct structures. According to the manufacturer, these gantry systems cost on the order of US$3 million in 2008, so the project or opportunity must be large to capitalize that cost. However, the system is not lim- ited to a single use and can be easily collapsed and used for other projects. Large contractors might invest in these types of systems for projects such as urban viaduct reconstruction where the equipment could just as easily be used for large- scale demolition as for new-span installation. A similar oppor- tunity would exist for over-water crossings. The Florida DOT has conceptualized a similar concept to use half-width redecking as a reconstruction solution for urban interstate projects. The department communicated its interest in such a system to the R04 team and has made con- ceptual estimates of two to three spans per day, or an aver- age of 1 mile a month of completed viaduct. The hypothetical project is the construction of a “managed lanes” facility, such as a tolled high-occupancy-vehicle facility in the median of an existing facility. Using top-down construction would allow for SPMT delivery of new spans and the gantry to place them in halves. A similar new bridge construction project is under way in North Carolina (the Washington Bypass Project) where top-down construction is being used to drive prestressed pile bents, erect bent caps, and set the beam lines; the decks are cast in place. An average of one completed span per week is being attained. This concept is not relegated to heavy element delivery. For years, individual beam lines have been erected with this type of delivery system in much smaller systems. Beam launchers are used to set individual beam lines or possibly smaller decked beam sections, and capacities of these systems are up to 200 tons. This concept could easily be deployed for rapid erection of new multi-span bridges; in some cases it might even be useful for shorter spans to aid bridge removal by lift- ing large sections from above and moving them out of the

243 way. A short or medium bridge gantry system is considered viable as a reusable and flexible ABC delivery system. Project Title: Overhead Gantry Lift Method Citation: Mi-Jack Products, Inc., 2007 ABC Design Features: ABC Construction Features: Low-impact construction with highly adaptable and customizable equipment Project Description An underutilized technology is the use of travel lift cranes commonly found in fabrication yards but not often used for on-site bridge construction. A common situation is infill con- struction in the median of an existing roadway. This work can be complicated from below because of crane placement issues or challenges in lifting and placing materials, or because of access constraints such as waterways, environmental con- straints, etc. The travel lift can run on rubber tires or rails and be supported on a narrow strip of existing bridge deck, such as the shoulder area. Several example projects are cited to demonstrate the use of this system. Route 46 over the Overpeck Creek is a busy thoroughfare in northern New Jersey and is a major feeder route to the busy George Washington Bridge. The superstructure required com- plete reconstruction. The estimated cost of a conventional replacement was $20 million with a 3-year schedule. That was deemed unacceptable. The New Jersey DOT estimated $19,000 per day in user costs for this construction project (Keith, 2006). One of the first solutions that helped compress the sched- ule was to insist on prefabricated superstructure units in lieu of cast-in-place concrete. Complete new beams and decking were furnished for the project in sections measuring 9 ft, 4 in. wide by 93 ft long and weighing 70 tons. The contractor determined that a crane large enough to erect those sections would take up too much room and was unsure if the existing bridge, in its deteriorated condition, could support the weight of the cranes plus the lifted loads. The solution proposed by the contractor, Railroad Construction, was to erect a longitu- dinal runway system and use an overhead gantry crane for erection. In that configuration the gantry could run the entire length of the bridge, and the contractor could work in two closed lanes 24 hr a day while traffic flowed in the two lanes remaining open. The project cost $18 million and was com- pleted in less than a year. A similar positive example is the reconstruction of I-77 in Virginia over the New River. The project involved building a new superstructure in the median between two existing bridges. Each bridge measures 1,800 ft long and rises as high as 140 ft in the air. With the presence of existing bridges, con- struction from below was difficult. Even if possible, a 140-ft vertical lift of heavy sections would have required a large crane mounted on barges that would have to have been con- stantly repositioned to erect the various spans. As a solution, the contractor elected to use a 70-ton capacity Mi-Jack mobile lift crane. The crane was used to erect all the structural steel in the bridge. ABC Opportunity The use of overhead gantry cranes provides substantial flex- ibility for erecting and dismantling structures that would be more difficult to erect with traditional cranes because of operating constraints and space limitations. Using an existing bridge as a work platform (or constructing a runway system independent of the bridge) allows the gantry to run freely throughout long stretches of the job site, delivering materials and completed portions of a bridge. Conversations with gan- try suppliers indicate that only a small fraction of bridge con- tractors use these methods compared with the perceived available opportunity. Gantries of various capacities and sizes are available for lease or purchase, and availability does not appear to be a limiting factor on use.