Bridges for Service Life Beyond 100 Years: Innovative Systems, Subsystems, and Components (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.

136 A p p e n d i x A As a result of the activities within Phase 1, several concepts were identified that merit research by others. This appendix summarizes the topics classified as Category 3 research topics and provides brief introductory comments for some of them. During the development of the Phase 1 report, limited research was conducted on some of these topics; however, these investi- gations are not documented. Concrete durability: Using Smart Materials and nanotechnology to enhance the Service Life of Concrete elements This section provides two innovative approaches to improv- ing concrete durability through the use of nanotechnolo- gies: the use of self-healing systems for closing cracks and the use of nanoparticles for reducing the permeability of existing concrete. Background Nanotechnology is a branch of engineering that deals with the design and manufacture of extremely small electronic cir- cuits and mechanical devices built at the molecular level of matter and the use of very small particles to improve the prop- erties of materials (Chong and Garboczi 2002; Ratner and Ratner 2003). Nano indicates one billionth of a unit of mea- sure. Nanotechnology can be used successfully in a variety of construction materials, including concrete. Nanoparticulate additives are now widely used as fillers in protective paints, coatings, and clean-up systems for buildings. Cement hydration products are porous with a pore-size distribution that ranges from nanometers to millimeters. The pores are a pathway for water and other aggressive solutions to penetrate into concrete, causing cracking and deteriora- tion. Thus, a better understanding of cement hydrates at the nanoscale would provide information for reducing and mini- mizing the pores. The addition of nanoparticles such as nano- silica and nanoclays to concrete can reduce the occurrence and size of pores that can improve concrete strength and durability through physical and chemical interactions. The microstruc- ture of the binder is expected to be uniform without apparent lime crystals or large pores. Nanomaterials for use in concrete include nanosilica (nano- SiO2), nanoclays, nanotubes, nanocomposites, and nano- titanium dioxide (nano-TiO2). Nanotubes are very thin sheets (one atom thick) that are rolled into a cylinder (de Ibarra 2006). Carbon nanotubes have the potential to enhance strength, effectively hinder crack propagation in cement composites, and act as nucleating agents. Reinforcing concrete with nano- fibers may produce tougher concretes by interrupting crack formation as soon as it is initiated. A very small amount of nano-SiO2 reduces cement perme- ability and makes cement paste denser than the control paste (He and Shi 2008). Nanosilica is more effective than silica fume in increasing the strength of mortar (Li et al. 2004). Also, the strength of the interfacial transition zone is increased, and there are more pozzolanic reactions in nanosilica cement paste than in silica fume cement paste (Qing et al. 2007). The addition of polymer-modified nanoclay is also able to increase compressive strength in cement paste (Birgisson 2006). Past attempts to improve concrete life have focused on pro- ducing denser, less porous concretes; unfortunately, these for- mulations have a greater tendency to crack. Another approach to improve the permeability of concrete has been to reduce dif- fusion in concrete by using viscosity-enhancing admixtures (Bentz et al. 2009). Larger molecules, such as cellulose ether and xanthum gum, increased viscosity but did not reduce diffusion rates. Smaller molecules, less than 100 nm (nano- meters), slowed ion diffusion. Smart materials in the simplest case would detect and warn of problems. In a more advanced form, they could not only detect but correct problems. One idea is the application of Brief Description of Category 3 Research Topics

137 electrorheological fluids for vibration control of structures. According to Chong and Garboczi (2002), “Electrorheologi- cal fluids stiffen very rapidly in an electric field, changing their elastic and damping properties. [. . .] An important research challenge in smart structures and materials is to achieve optimal performance of the total system rather than just in the individual components. Among the topics requir- ing study are materials with energy-absorbing and variable damping properties, as well as materials having a stiffness that varies with changes in stress, temperature, or acceleration. [. . .] There is [also] the associated problem of simply being able to detect (predict) when repair is needed and when it has been satisfactorily accomplished.” Self-Healing Systems: Using Nanotechnologies for Closing Cracks One approach for incorporating smart materials is to use conventional concrete and add nanomaterials to fresh con- crete that could detect cracks and start the self-healing process by filling those cracks. The concept of self-healing concrete was inspired by biological systems in which damage triggers an automatic healing response. The new self-healing concrete uses the same idea, in which the addition of self- healing properties would be activated when cracking occurs. Based on recent investigation, certain microcapsules can be added to the concrete mixture to change the behavior of concrete after cracking. The new technology consists of three components: • A composite material, in this case concrete; • A microencapsulated healing agent; and • A catalyst. When cracks form in such concrete, the crack will tear the microcapsule and thus release the healing agent (see Fig- ure A.1). The agent then flows into the crack through cap- illary action and the pressure from osmosis. The agent will then come into contact with the catalyst, which will initiate the polymerization process. This whole process results in the formation of insoluble crystals that will fill the crack up to a width of 400 mm (microns). The polymerization process also bonds the crack closed. The specifications of self-healing con- crete are as follows: • Quantity: 100 to 200 capsules per cubic inch; • Capsule size: ~100 mm; and • Regained strength: as much as 75% of the original strength. One approach for achieving self-healing concrete is to embed hollow fibers (which act as microcapsules) in the repair matrix before it is subjected to damage. When cracking occurs in the concrete, all fibers in the crack path are broken, and the repair materials are released from inside the hollow fibers, infiltrating the matrix. The repair material penetrates and rebonds the crack. Consequently, this mechanism occurs when needed, repairing the concrete automatically without manual intervention. Neither fibers nor the released material sacrifices strength for durability. Figure A.2 shows the self- repairing mechanism. Because the healing component of this newly developed concrete is not released during the mixing of the fresh con- crete, the fresh concrete properties remain unchanged. Fibers with high aspect ratios can improve the ductile behavior of concrete and also increase strength and stiffness. Generally speaking, fibers can improve concrete behavior. When self- healing concrete is activated, the repaired region is more duc- tile and stronger in tension than the original concrete. Using self-healing concrete will address many of the issues associ- ated with multiple failure modes. Fibers can be used to increase reinforcing bars’ resistance to corrosion. The fibers are coated with a substance that may be dissolved in salt water. These fibers are filled with an anti- corrosion chemical component. When the fibers come in contact with salt water, the coating substance is released and attaches to the rebar, delaying or even preventing further corrosion. Purpose and Scope Smart materials and nanomaterials have the potential to enhance the properties of concrete and provide long-lasting service with minimal maintenance. This study could explore available smart materials and nanomaterials that can be used in concrete. The potential candidates should be evaluated for their effectiveness in enhancing the properties of concrete and sealing the cracks or performing self-healing. Suggested Research Methodology for Others This section provides a guideline for others to continue research in this field. It is recommended this research be car- ried out in two phases. In the first phase, literature will be surveyed to determine smart materials and nanomaterials that can be used in concrete for improving the paste and the interfacial transition zone, and materials with potential for improvement will be selected. In the second phase, the potential smart materials and nanomaterials will be added to concrete mixtures prepared in the laboratory. The concretes will be tested for strength, permeability, and shrinkage to determine improvements against controls without smart materials and nanomaterials. Concrete samples will be evaluated under a microscope to

138 (b) (a) (c) Figure A.1. Self-healing concrete process: (a) incipient crack forms (lower left); (b) crack ruptures microcapsule, releasing healing agent; and (c) polymerized healing agent fills crack. structural fibers repair fiber matrix micro-crack sealed crack Figure A.2. Self-repairing mechanism.

139 determine improvements in the microstructure that can lead to enhancements. Using Nanoparticles to Reduce the Permeability of Existing Concrete The second part of the study would involve the use of charged nanoparticles to reduce concrete permeability. Treatment of concrete using charged nanoparticles has the potential to improve the concrete properties of existing structures. If suc- cessful, this technique could be used to extend the service life of existing structures or to eliminate the need to replace cer- tain structures with new structures. Charged nanoparticles can be moved into open pores and moved through the pore structure of the cover concrete toward the surface of the steel, decreasing the porosity and forming a much less permeable concrete cover, thus restricting further migration of chlorides to the steel. The improved permeabil- ity of the concrete will improve the structure’s resistance to corrosion and can be used to extend the service life of existing reinforced concrete structures. Use of this technique on concrete bridge decks could elimi- nate the need to apply overlays, in many situations reducing dead load, maintaining load rating, and maintaining clear- ances on overhead structures. Purpose and Scope Charged nanoparticles could be evaluated on chloride- contaminated and, possibly, new reinforced concrete elements. Treatment with charged nanoparticles would enhance the properties of the concrete cover up to the steel interface by reducing porosity and defect size, which act as initiation sites for the corrosion of the steel. The scope of the research would be to identify the correct type and size of nanoparticles that would allow their penetration into the concrete pore struc- ture and to evaluate the increased corrosion resistance pro- vided to the steel. Suggested Research Methodology for Others The purpose of this section is to provide a guideline for others to continue research in this field. It is recommended this research be carried out in the laboratory using samples of rein- forced concrete elements subjected to chloride contamination. Chloride-contaminated samples would be subjected to treat- ment using charged nanomaterials. The ability to improve concrete properties would be determined by testing sam- ples cored from the elements for strength and permeability. The improvements in corrosion resistance will be determined by exposing the treated and untreated reinforced concrete samples to high levels of chloride solution and measuring chlo- ride profiles and corrosion activity. If corrosion is initiated after a sample has been treated, the level of chloride at the steel depth would be determined and compared with thresh- old levels for untreated concrete elements. development of innovative Hybrid Orthotropic deck System for Short-Span Bridges Several concepts have been identified that are related to steel bridge systems. One concept that was studied by the principal investigator of the R19A project is related to an innovative orthotropic deck system for short-span steel bridges. Accord- ing to John Fisher of Lehigh University (phone communica- tion, May 2008), an orthotropic deck system is the only deck system that has a chance of providing 100-plus years of service life. However, the available systems are suited for long-span bridges and are prone to fatigue and delamination of mem- branes that are used for the riding surface. Both of these prob- lems are related to the flexibility of the top steel plates used in orthotropic deck systems. The use of a steel-only option is to produce a lightweight deck system for long-span bridges. For long-span bridges, a steel orthotropic deck system provides a lightweight alternative. However, for short-span bridges, some increase in deck weight is tolerable. The innovative ortho- tropic deck system described in the following paragraphs takes advantage of steel and concrete to develop a system that should have well over 100 years of service life. The high cost of orthotropic decks, which is primarily due to fabrication, prohibits their use for short- and medium-span bridges. They become an affordable option for long-span bridges because of their light weight. However, with modifi- cations, the concept of an orthotropic bridge deck may be well suited for short- to medium-span bridges. The proposed concept is to develop a system in which the deck consists of top and bottom steel plates with a series of vertical plates. Such a system could also act as the top flange of girders and presents various options for accelerated bridge con- struction. Once the system is transported to the field, the hollow spaces between the top and bottom plates could be filled with concrete. The proposed system is shown in Figure A.3. Advances in welding technology make this system possible by allow- ing welding of the steel elements from the “blind” side or outer surface. General Description of Proposed Hybrid Orthotropic Deck System The proposed hybrid orthotropic deck system consists of a top and bottom plate with a cavity. The cavity is divided into smaller cells by using vertical ribs. These smaller cavities are filled with self-consolidating concrete. The total thickness of the deck is about 4 in. (for 10-ft spacing between girders),

140 and the thicknesses of the top and bottom plates are less than 0.25 in. The system offers the following advantages: • There is a significant increase of the load-resistance capacity of the deck. • The concrete sandwiched between the top and bottom plates is protected from wear and weathering elements and could last 100-plus years without any maintenance or special design provision. • The concrete between the top and bottom plates stiffens the deck system to a point that could eliminate fatigue cracking in the deck components. • The system lends itself to modular construction and could significantly decrease on-site construction activities. The system could be used for both new and existing bridges. • The concrete between the top and bottom plates stiffens the system and can therefore eliminate the delamination of wearing surfaces seen in purely steel orthotropic deck systems. • The addition of concrete prevents the orthotropic deck com- ponents from buckling, which makes it possible to decrease the thickness of the steel plates. • The concrete inside the cells could be added after the deck is attached to girders, thereby reducing the weight of pieces that have to be transported. • The deck system is so stiff that it could eliminate the need for cross frames in the positive bending regions of steel bridges and, therefore, eliminate a detail that is a source of many fatigue-related issues with steel bridges. • The weight of the system is about 30% lower than an all- concrete option. The disadvantages of the proposed system include a poten- tially higher initial cost and the addition of weight (as com- pared with a purely steel orthotropic deck system). In addition, inspecting a large portion of this deck system for fatigue crack- ing that is due to the presence of the concrete would be impos- sible. However, past research indicates that presence of the concrete can effectively prevent crack propagation (Hanson et al. 1987). Preliminary Study A preliminary finite element study was performed using the general-purpose finite element program ABAQUS to investi- gate the general behavior of the system. Figures A.4 through A.6 show the dimensions of the structure modeled. In con- ventional orthotropic bridge decks, the ribs are parallel to girder direction, as shown in Figure A.4. However, in this pre- liminary study, the possibility of a transverse orthotropic deck was also considered, in which the direction of the ribs is perpendicular to the girder direction, as shown in Figure A.5. A parametric study was performed to investigate bridge behavior with different orthotropic deck geometries. The Ribs Steel Girder Top Cover plate Concrete Figure A.3. Proposed hybrid orthotropic deck system. Figure A.4. Longitudinal ribs.

141 geometric specifications that can be changed in the orthotropic deck are total deck thickness, thickness of the plates (top, bottom, and ribs), and spacing between the ribs. The rib orien- tation can be either longitudinal or transverse. The spacing of the rib plates greatly affects the cost of the deck because weld- ing of these plates comprises a majority of the fabrication cost. Table A.1 presents the deck dimensions that were modeled. Figure A.7 shows a typical model structure with a longitu- dinal orthotropic deck. To investigate the behavior of the structure, five analyses were performed on each combination of parameters: 1. Analysis of the steel part of the structure only (ortho- tropic deck and steel girders without concrete) under its self-weight. 2. Buckling analysis of the steel part of the structure only (orthotropic deck and steel girders without concrete). This analysis helps to find the potential buckling modes of the structure when no concrete exists. 3. Analysis of the structure under the weight of wet concrete filled in the orthotropic deck hollow spaces. 4. Buckling analysis of the structure with hollow spaces of the orthotropic deck filled with hardened concrete. The buckling modes that resulted from this analysis will also be used in the ultimate load analysis of the structure as the implied imperfections. 5. Ultimate load analysis of the structure. This analysis resulted in the load–deflection curves of the structure. Two major loading conditions were considered: two HS-20 trucks side by side and one HS-20 truck on the bridge such that one wheel line was centered between two adjacent girders. Dead Load Results Table A.2 summarizes the maximum deflections and maxi- mum von Mises stresses in the orthotropic deck and girders due to the weight of steel and wet concrete. The maximum stresses occurred midway between the two adjacent girders. The deflection of the deck and girder were measured relative to the undeformed shape of the structure. Ultimate Loading of Structure with Hardened Concrete The buckling mode shape of the structure resulting from a buckling analysis of the whole structure with hardened con- crete was incorporated into the ultimate loading of the struc- ture as an imperfection. A concrete damage plasticity model was included in the concrete model of the structure to address the concrete crushing effect in the analysis (CEB-FIP 1993). Two Trucks Side by Side Figure A.8 shows an example of the von Mises stresses in the bridge at the ultimate load with two HS-20 trucks side by side. Figure A.5. Transverse ribs. Figure A.6. Supporting girders. Table A.1. Dimensions of Hybrid Orthotropic Deck Bridge Modeled Deck Direction Total Deck Thickness (in.) Deck Plate Thickness (in.) Longitudinal 4.0 0.250 0.125 3.0 0.250 0.125 Transverse 4.0 0.250 0.125 3.0 0.250 0.125

142 Table A.2. Modeled Structure’s Responses Under Construction Maximum Deflection Maximum von Mises Stress Deck Direction Total Deck Thickness (in.) Deck Plate Thickness (in.) Construction Stage Orthotropic Deck (in.) Girder (in.) Orthotropic Deck (psi) Girder (psi) Longitudinal 4.0 0.250 Steel self-weight 0.18 0.15 1,250 2,204 Wet concrete and steel weight 0.42 0.34 3,485 5,015 0.125 Steel self-weight 0.24 0.13 2,151 1,626 Wet concrete and steel weight 0.9 0.33 9,836 4,318 3.0 0.250 Steel self-weight 0.18 0.15 1,609 2,223 Wet concrete and steel weight 0.37 0.3 3,872 4,343 0.125 Steel self-weight 0.25 0.13 2,422 1,657 Wet concrete and steel weight 0.81 0.31 9,420 3,894 Transverse 4.0 0.250 Steel self-weight 0.16 0.16 1,860 2,200 Wet concrete and steel weight 0.37 0.37 4,420 5,034 0.125 Steel self-weight 0.16 0.15 2,111 1,752 Wet concrete and steel weight 0.43 0.43 5,898 4,897 3.0 0.250 Steel self-weight 0.16 0.16 2,233 2,218 Wet concrete and steel weight 0.32 0.32 4,457 4,374 0.125 Steel self-weight 0.15 0.15 2,907 1,728 Wet concrete and steel weight 0.36 0.35 6,554 4,084 Truck tires’ positions (two trucks side by side) Figure A.7. Structure modeled with ABAQUS with a longitudinal orthotropic deck.

143 Figure A.9 shows an example of the deflections in the bridge due to the ultimate loading of the structure with two HS-20 trucks side by side. This is the final stage of the loading. One Truck with One Side on the Middle of the Girders’ Spacing Figure A.10 shows an example of von Mises stresses in the bridge at the ultimate load with one truck on the bridge. Figure A.11 shows an example of the deflections in the bridge due to the ultimate loading of the structure with one truck on the bridge. This is the final stage of the loading. Load–Deflection Results The following two subsections describe the results obtained from the ultimate loading of the structure. This discussion provides an understanding of how changes in the dimensions Figure A.8. Von Mises stresses in the bridge; gray represents yielded regions. Figure A.9. An example of the deflections in the bridge at the ultimate load.

144 Figure A.11. An example of the deflections in the bridge at the ultimate load. Figure A.10. Von Mises stresses in the bridge; gray represents yielded regions. of the structure affect load resistance and the ultimate ductil- ity of the structure. Two Trucks side by side Figure A.12 shows the load–deflection curve of the structure associated with the ultimate load analysis of the structure using two trucks on the bridge side by side. The deflection is associated with girder deflection at midspan. Figure A.13 shows the load–deflection curve of the struc- ture associated with the ultimate load analysis of the struc- ture in terms of the number of bridge weights that can be applied to the bridge. As these curves show, decreasing the thickness of the ortho- tropic deck plates will reduce the ultimate load capacity. How- ever, the system still has significant reserve capacity. Figure A.13 may also be used to find the most efficient dimensions for the structure if it is assumed that the most efficient structure is the one that carries larger loads compared with its self-weight. one Truck wiTh one side on The Middle of Girders spacinG Figure A.14 shows the load–deflection curve associated with the response of the orthotropic deck. Conclusions from Finite Element Study The following are the conclusions of the preliminary finite element analysis performed on a hybrid orthotropic deck system: • Buckling in the deck of the completed structure is unlikely. Adequately stiffened girders prevent the buckling failure mode from governing.

145 Figure A.12. Load–deflection curves of the structure associated with two trucks side by side. Number of HS-20 Trucks-Deflection Curve; Behavior of Entire Bridge 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 125 130 135 140 Deflection of the midspan of the girder (in.) N um be r o f H S- 20 T ru ck s Longitudinal, Total Deck Thickness = 4", Deck Plates' Thickness= 0.250" Longitudinal, Total Deck Thickness = 4", Deck Plates' Thickness= 0.125" Longitudinal, Total Deck Thickness = 3", Deck Plates' Thickness= 0.250" Longitudinal, Total Deck Thickness = 3", Deck Plates' Thickness= 0.125" Transverse, Total Deck Thickness = 4", Deck Plates' Thickness= 0.250" Transverse, Total Deck Thickness = 4", Deck Plates' Thickness= 0.125" Transverse, Total Deck Thickness = 3", Deck Plates' Thickness= 0.250" Transverse, Total Deck Thickness = 3", Deck Plates' Thickness= 0.125" Figure A.13. Load–deflection curves in terms of bridge weight using two trucks side by side. 0 2 4 6 8 10 12 14 16 18 20 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 125 130 135 140 Deflection of the midspan of the girder (in.) Se lf- W ei gh t M ul tip le s Longitudinal, Total Deck Thickness = 4", Deck Plates' Thickness= 0.250" Longitudinal, Total Deck Thickness = 4", Deck Plates' Thickness= 0.125" Longitudinal, Total Deck Thickness = 3", Deck Plates' Thickness= 0.250" Longitudinal, Total Deck Thickness = 3", Deck Plates' Thickness= 0.125" Transverse, Total Deck Thickness = 4", Deck Plates' Thickness= 0.250" Transverse, Total Deck Thickness = 4", Deck Plates' Thickness= 0.125" Transverse, Total Deck Thickness = 3", Deck Plates' Thickness= 0.250" Transverse, Total Deck Thickness = 3", Deck Plates' Thickness= 0.125" Bridge Self-Weight Multiples-Deflection Curve; Behavior of Entire Bridge

146 • An efficient spacing between the ribs of an orthotropic deck can be obtained from cost analysis. • Increasing the orthotropic deck plate thickness will increase the ultimate ductility of the bridge and, to a lesser extent, the ultimate load-carrying capacity of the bridge. Possible Research Plan for Others The issues needing investigation are listed below: Welding process—So-called blind welding technology using laser welding is one available technology that can be used to weld steel plates from the “blind” side. Optimal orthotropic deck dimensions—The behavior, effi- ciency, constructibility, and economy of the proposed hybrid orthotropic deck system depend on a number of factors, including its dimensions. The thickness of the deck plates, the thickness of the deck, and the dis- tance between the ribs are some of the dimensions to be investigated. Economical bridge span—The proposed hybrid orthotropic deck system would not be economical for all span lengths. This range can be a subject of research. Connection to supporting girders—There is a need to develop details that could be used to attach prefabricated sections to each other and to steel girders. Corrosion protection of the orthotropic steel plates—Corrosion protection philosophy needs to be developed for exposed steel. Use of A1010 steel is an alternative. A1010 steel costs about twice as much as regular steel; however, it has excel- lent corrosion properties. The following steps are suggested for investigating the merits of the system: Step 1. Conduct a preliminary cost analysis and finite element study to determine the most efficient configuration of the orthotropic deck system. Step 2. Determine the best available welding technology through discussion with industry. Step 3. Perform laboratory testing of a small-scale, one-girder hybrid orthotropic deck system under fatigue and then ulti- mate loading. The test specimen should include the riding surface, and membrane delamination from the riding sur- face should be evaluated. Step 4. Develop detailed design, fabrication, and construction guidelines. 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 10.5 11 11.5 12 12.5 13 13.5 Relative Deflection of the Deck with respect to the Girder (in.) N um be r o f H S- 20 T ru ck s Longitudinal, Total Deck Thickness = 4", Deck Plates' Thickness= 0.250" Longitudinal, Total Deck Thickness = 4", Deck Plates' Thickness= 0.125" Longitudinal, Total Deck Thickness = 3", Deck Plates' Thickness= 0.250" Longitudinal, Total Deck Thickness = 3", Deck Plates' Thickness= 0.125" Transverse, Total Deck Thickness = 4", Deck Plates' Thickness= 0.250" Transverse, Total Deck Thickness = 4", Deck Plates' Thickness= 0.125" Transverse, Total Deck Thickness = 3", Deck Plates' Thickness= 0.250" Transverse, Total Deck Thickness = 3", Deck Plates' Thickness= 0.125" Number of HS-20 Trucks-Deflection Curve; Deck Behavior Figure A.14. Load–deflection curves in terms of bridge weight using one truck with one side on the middle of girders spacing.

147 Development of Seismic Details for Simple for Dead and Continuous for Live Load Steel Bridge System Recently a new steel bridge system, referred to as simple for dead load and continuous for live load, has gained popularity in nonseismic areas of the country. Research conducted over the past 10 years (Azizinamini 2013) has resulted in the develop- ment of complete design and detailing provisions for this new system for nonseismic applications. As there are no bolts or expansion joints in this system, it should have a long service life. This system is also best suited for accelerating the construction process. It is ideal for building individual spans off site, trans- porting them to the final location, and then joining the spans over the middle piers to create continuity for live load. Another version of the system that uses adjacent beam technology can significantly reduce the on-site construction activities. Background To date no research studies have been conducted to extend the applicability of this system to seismic regions. The main objective of this proposed concept as a research area is to use the available research data (for nonseismic application) and conduct mainly numerical work to develop details and design provisions for extending the application of the simple for dead and continuous for live steel bridge system to highly seismic areas. The focus of the study should be on using this system in conjunction with accelerated bridge construction philosophy. This work should provide highly seismic areas with an excellent steel bridge system alternative that does not have expansion joints and potentially produces long service life while meeting seismic design requirements. Brief Overview of Seismic Design Requirements Steel bridges are about 40% lighter than concrete alternatives and, at first glance, are expected to provide better perfor- mance in a major earthquake. However, observations from a past earthquake demonstrate that the use of wrong details or systems could result in steel bridges sustaining major damage (Astaneh-Asl et al. 1994). In the 1995 Hyogoken-Nanbu earth- quake in Kobe, Japan (Bruneau et al. 1996; Chung 1996; Shinozuka et al. 1995; Azizinamini and Ghosh 1997), steel bridges suffered damage to superstructure elements (inadequate cross-frame detailing leading to lateral bending of the girder webs near girder ends) resulting in major retrofit activities and closing of major highways, such as the Hanshin Expressway, for more than a year. The Kobe experience demonstrated that even minor damage to steel bridges in seismic events can result in types of damage that could be very difficult to repair. Among the lessons to be learned is that critical elements of the bridge that are difficult to inspect and repair must be protected from any level of damage and remain elastic during the entire seismic excitation. Seismic input is largely unknown; therefore, the design phi- losophy for buildings and bridges is to work on behavior of the structure under known conditions. Specifically, the design objective is to predefine the damage locations and design them accordingly by providing adequate levels of ductility. In the case of bridges, the preferred damage locations are at the ends of pier columns (formation of plastic hinges). In the direction of traffic, it is preferred to put columns in double curvature, as shown in Figure A.15. This arrangement allows larger portions of the pier column (two plastic hinges versus one for single curvature) to participate in energy dissipation. In the transverse direction, pier columns are usually designed to act in single curvature, as shown in Figure A.16. Figure A.15. Deflected shape of a three-span bridge under the longitudinal (along traffic) direction. (a) (b) Figure A.16. Deflected shape of pier column in (a) longitudinal and (b) transverse directions.

148 Under longitudinal excitation, plastic hinges are located near the top and bottom of the columns; under transverse excitation, the plastic hinge is located near the bottom of the pier column. Seismic design of steel bridges is still a developing subject. One of the most comprehensive design provisions for design of bridges for seismic events is given by Caltrans Seismic Design Criteria (Caltrans 2008). The main design requirement in the seismic design of bridges is to keep the superstructure elements, called pro- tected elements, completely elastic during an entire seismic event. The inelasticity is then forced to take place at pre- defined locations within the substructure. The predefined damage locations are the weak links or fuses that control the level of forces to be transmitted to superstructure elements. This design approach, which is referred to as the capacity design approach, is used for designing bridges in seismic regions. In the capacity design approach, protected elements are designed for the largest possible force effects they might expe- rience, considering overstrength that may exist because of higher actual material strength than that specified in design. The capacities of the bridge elements in the desired damage location (plastic hinge locations) are controlled through design. The plastic hinge regions are also detailed so that they can provide the desired capacities while deforming in an inelastic manner during a seismic event (ductility through adequate detailing). Brief Description of Simple for Dead and Continuous for Live Load Steel Bridge System Continuous steel bridges are usually constructed so that the system provides continuity for noncomposite dead loads in addition to superimposed dead load and live loads. Figure A.17 shows a conventional two-span continuous steel bridge girder. For a large number of bridges, the construction sequence con- sists of first placing a middle segment over the interior support and then connecting the two end pieces by using a bolted or welded field splice. In the simple for dead load and continuous for live load system, the girders are spliced over the pier. Girders are placed spanning directly from abutment to pier within each span. The individual spans are simply supported when the deck is cast. Once the deck is in place, reinforcing steel cast into the deck provides continuity of the tensile forces for live load and superimposed dead loads (the weight of barrier and future wearing surfaces) only. The compressive component is trans- ferred through direct bearing of the bottom flanges. An exam- ple of this detail is shown in Figure A.18. Falsework Restricted Traffic Pier Field Splice Field Splice (a) (b) Figure A.17. (a) Conventional two-span continuous bridge girder and (b) typical splice detail. Figure A.18. Simple for dead and continuous for live detail.

149 The main challenge is to use the correct detail over the pier to connect the girder ends. Bottom flanges of steel embedded in the concrete diaphragm transmit large stresses to the con- crete and, if not detailed correctly, can crush the concrete. Fig- ure A.19 shows the results of tests on three details that could be used to connect girder ends over the pier in non seismic regions. The main lesson learned is that the bottom flanges of the steel girders embedded in the concrete diaphragm should be con- nected somehow. This arrangement will allow the transfer of compressive force (generated by live loads) from one girder to another girder without passing through the concrete dia- phragm. The three test specimens shown in Figure A.19 were identical, except the details used at the ends of the girders, embedded in the concrete diaphragm. In one specimen, labeled Test 2 in Figure A.19, the ends of the girders did not incorporate any detail and were simply embedded in the concrete dia- phragm. As noted, the capacity of that specimen was signifi- cantly less than the other two specimens, in which the ends of the girders incorporated details that allowed a smooth path for transferring the compressive force from one girder to another. In nonseismic applications, the forces to be transmitted from the bottom flange of one girder to another girder in the concrete diaphragm are predominantly compressive, and the detail to be used is not required to handle tension force. Fig- ure A.20 shows a detail that is suitable for nonseismic applica- tion and has been used in several bridges in service. However, in seismic application, the types of details shown in Figure A.20 will not work. In seismic areas, the detail at the ends of the girder in the concrete diaphragm region needs to be able to resist cyclic loading, which will involve both tension and compression. In nonseismic application it only needs to resist compressive-type force. Further, in seismic applications the entire concrete dia- phragm region, including the girders ends and details, needs to remain completely elastic (protected elements) during the entire seismic event. An additional requirement is that the pre- defined damage areas (plastic hinge locations) must be forced to be at some distance away from the concrete diaphragm region, allowing repair and inspection after a major seismic event. Test 1 Test 2 Test 3 Figure A.19. Comparing performances of various details for connecting girder ends over the pier in nonseismic regions. Figure A.20. Detail used in nonseismic regions.

150 Suggested Scope of Work An extensive amount of research data for the proposed sys- tem in nonseismic regions exists. However, there are spe- cific requirements for seismic applications that need to be addressed. For seismic applications, the detail over the pier should have the following characteristics: 1. It should be adequate for cyclic loading (designed for ten- sion and compression). 2. The plastic hinge locations should be at some distance away from the girder ends and column ends near bearings. 3. Design should follow a capacity design philosophy, allow- ing different performance levels as desired by the owner. To achieve these objectives, the following five tasks should be carried out: Task 1. Develop preliminary ideas on types of details that could be used to join the girder ends over the pier. During a seismic event, the column moment becomes a torsional force for the concrete diaphragm, as shown in Figure A.27. The concrete diaphragm will be the protected element and must remain elastic. Therefore, the concrete diaphragm Figure A.21. Single (interior) box girder and deck unit. Figure A.22. Three adjacent box girder units. Figure A.23. Closure region detail. Figure A.24. Headed bar detail. Simple for Dead and Continuous for Live Load System Using Adjacent Beam Technology The adjacent box concept uses prefabricated units consisting of an individual steel box girder topped by a portion of deck slab, as shown in Figure A.21. These units are prefabricated and then shipped to the job site. The portion of the deck shown in Fig- ure A.21 is cast at the fabrication shop or temporary staging location. Once on site, the individual units are set in place on the supports adjacent to one another, as shown in Figure A.22. A longitudinal deck closure strip between the individual units is then cast to join them. At the same time, the turndown over the interior support is cast. The interior turndown connects the spans and provides continuity between them for sub- sequent loading (live load). The individual units (steel box and pretop deck) are rela- tively lightweight, typically less than 100 tons for spans of 140 ft. Once the pretop girders are placed side by side, they are then connected through casting closure pour regions. Figure A.23 shows the closure region detail. One option is to use headed bars, as shown in Figure A.24, in the closure pour regions. Figure A.25 shows a rendering of the closure splice region over the interior support. The hooked bar ends from each span anchor the longitudinal reinforcement into the turndown and provide continuity over the support. A similar detail is shown in Figure A.26. Objective The objective of the study must be to develop recommen- dations for extending the application of the simple for dead and continuous for live load steel bridge system to highly seismic areas.

151 must be designed for maximum credible torsion and remain elastic. Several approaches can be used to ensure that the concrete diaphragm remains elastic under maximum cred- ible torsion. Posttensioning the concrete diaphragm is one possibility (Patty et al. 2002). Further, the plastic hinge at the column end must form at some distance away from the column ends. The detail to be selected will be such that it will create a frame action between super structure and sub- structure (integral pier cap). The integral pier cap detail is the preferred choice in some parts of the United States, such as the West Coast, because of aesthetics considerations. Task 2. Modify the existing detailed nonlinear finite element models used in a previous research study (Farimani et al. 2013) on applicable nonseismic details to determine the force transfer mechanisms of the details identified under Task 1. Figure A.28 shows an example of models used in previ- ous research studies. Task 3. Conduct small-scale tests to verify specific aspects of the detail selected under Task 1. This test can also be used as a means to calibrate the analysis model in Task 2. Task 4. Based on the results of Tasks 1, 2, and 3, develop pre- liminary design provisions using a capacity design approach. Task 5. Develop a plan of action for research studies by others to develop test-verified details suitable for seismic regions. List and Description of Other Category 3 Research Concepts Table A.3 lists other concepts for Category 3 research areas that have been identified and are proposed for research by others. Girder Cope Top Flange Bearing Blocks Longitudinal Reinforcement Deck Figure A.25. Interior support continuity detail. Figure A.26. Interior support detail showing multiple girders. Figure A.27. Transfer column moment to concrete diaphragm.

152 Figure A.28. Typical finite element models used to investigate force transfer mechanism for nonseismic details. Table A.3. Additional List of Category 3 Research Items Research Area Topic Concrete durability Develop tension limits and mix designs defining low-cracking concrete. Investigate use of nanomaterials to improve the paste and the interface between the aggregate and the paste, leading to reduced permeability. Determine lightweight concrete benefits. Investigate fatigue in concrete (predicting microcrushing of deck concrete). Investigate use of inorganic nanomaterials and electric field for filling cracks, even when they are very tight. Investigate using smart materials to respond to distress in concrete. For example, if cracks occur, the smart material would close the crack; if the pH decreases, the smart material would increase the pH. Bridge deck Develop systems for protecting the underside of the bridge deck from seawater intrusion, which may include innovative means and methods of forming for the deck. Substructure Investigate high-performance concrete applications in splash zones needed to reduce concrete permeability and enhance chemical resistance to salt water. Investigate jacketing systems for piles and shafts. Consider incorporating with new construction to protect critical parts before damage and chloride intrusion can occur. Investigate pile splicing systems for improvements in long-term durability. Develop substructure components using innovative (composite) construction materials that have strong resistance to saltwater environments. Bearings Develop life prediction model using fatigue testing for steel-reinforced elastomeric bearings. Develop life prediction model using fatigue testing for cotton duck pads. Develop life prediction model using fatigue testing for high-load multirotational disc bearings. Develop life prediction model using fatigue testing for high-load multirotational pot bearings. Determine the proper orientation of guided bearings for curved and skewed bridges by analytical study. Joints Develop rapid replaceable expansion joint assembly. (continued on next page)

153 Table A.3. Additional List of Category 3 Research Items Research Area Topic Fatigue and fracture Develop guide for fatigue retrofitting. Investigate potential cracking problem with galvanized steel members. Develop toughness standards for tubular and hollow-section members. Develop criteria that tie inspection frequency of fracture critical members to level of damage accumulation, that is, to average daily truck traffic. Investigate the feasibility of using a coating system for crack detection. Investigate the use of real-time instrumentation and monitoring for crack detection and monitoring. Corrosion protection Develop a best practices manual. Steel bridge systems None for Category 1 Concrete bridge systems Develop nonposttensioned systems that will ensure transverse continuity and nonleaking joints for adjacent member bridges. Add threaded-rod continuity to create full continuity for deck weight and to eliminate potential creep restraint cracking at the piers (Nebraska; Alberta, Canada; Illinois; Florida). Add posttensioning to keep superstructure sections under compression that are prone to cracking. (continued)