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.

160 A p p e n d i x C This appendix presents details of the experimental testing per- formed on a precast adjacent box beam system with transverse posttensioning. Additional information describing the develop- ment of the system can be found in the following PCI journal articles: Simplified Transverse Post-Tensioning Construction and Maintenance of Adjacent Box Girders (Hansen et al. 2012) and Transverse Post-Tensioning Design and Detailing of Pre- cast, Prestressed Concrete Adjacent-Box-Girder Bridges (Hanna et al. 2009). A source of general information is the PCI Bridges Committee publication titled The State of the Practice of Precast/Prestressed Adjacent Box Beam Bridges (PCI 2011). Background and problem Statement Precast adjacent box beam bridges are one of the prevalent box girder systems for short- and medium-span bridges (typically 20 to 127 ft), especially on secondary roadways. These bridges consist of multiple precast box beams that are butted against each other to form the bridge deck and superstructure. The system also has applicability to accelerated bridge construction for rapid delivery. Adjacent box beams are generally connected using partial- or full-depth grouted shear keys along the sides of each box. Transverse ties are usually used in addition to the grouted shear keys; these may vary from a limited number of threaded rods to several posttensioned tendons. In some cases, no topping is applied to the structure, but in other cases a noncomposite topping or a composite structural slab is added. Bridges constructed using box beams have been in service for many years and have generally performed well. However, a recurring problem is cracking in the grouted joints between adjacent units (see Figure C.1), which results in reflective cracks in the wearing surface and deterioration and cracking on the bottoms of beams (see Figure C.2 through Figure C.6). The development of these longitudinal cracks over the shear keys jeopardizes the durability and structural behavior of adjacent box girder bridges (Kahl 2005; Huckelbridge et al. 1995). In most cases, the cracking leads to leakage, which allows chloride-laden water to penetrate the sides and bottom of the beams and cause corrosion of the nonprestressed and prestressed reinforcement. In addition, the load distribution among the beams is adversely affected, so that the loaded beams are required to carry more load than originally intended (Huckelbridge et al. 1995). Objectives of the Research The general objective of this research topic is to improve the performance of adjacent box beam bridges, specifically with regard to the transverse connection. The overall objectives are (1) to achieve the structural capacity of this system in the transverse direction and (2) to prevent longitudinal joint leakage. These objectives are accomplished by developing a detail that has the ability to transfer moment and shear in the transverse direction. Previous Research Previous research by the University of Nebraska–Lincoln (UNL) investigated the benefits of nonposttensioned sys tems and developed two new connection details that are both effi- cient and economically practical. Advantages of these systems include • Improved production; • Easier inspection of voids; • Better drainage of moisture; • Lighter weight for handling and shipping; and • Continuous rather than discrete (quarter-point) connec- tions. Below is a brief description of the two systems; additional information can be found in (Hanna et al. 2011). Joints in Modular Systems of Adjacent Box Girders

161 Wide Joint System The wide joint system developed by UNL does not use a top- ping or diaphragms. Top and bottom transverse reinforce- ment is used in a wide shear key filled with concrete to connect adjacent boxes instead of the reinforced concrete composite topping. This monolithic joint with top and bottom reinforce- ment provides continuous connection that transfers shear and moment between boxes and eliminates the need for inter- mediate or end diaphragms. The elimination of topping and diaphragms will significantly speed the production and con- struction operations and reduce the material, labor, and erection costs. Shear keys used in the wide joint system are wide full-length and full-depth shear keys that require a slight modification to the AASHTO standard box cross section and consequently the forms. Figure C.3 shows a modified 33-in.- deep and 48-in.-wide AASHTO PCI box section as an exam- ple. Modifications include (1) a 5-in.-wide shear key (2.5 in. at each box side); (2) two block-outs every 4 ft at the top flange, which are 4 in. deep, 5.5 in. long, and 5.5 in. wide; and (3) two block-outs every 4 ft at the bottom flange, which are 8 in. long and 4 in. wide. Two reinforced concrete connections are used at the block-out locations (every 4 ft) in the top and bottom flanges to connect adjacent boxes. Reinforcing steel bars extending from the top and bottom flanges of each box are lap spliced using short bars confined by 3-in.-diameter, 1-in.-pitch, 0.125-in.-thick spirals, as shown in Figure C.4. This confinement reinforcement is necessary to provide ade- quate development length for such short lap splices. It is also recommended that the top surface of the box and the shear keys have 0.5 in. of extra thickness to be grinded to provide a roughened surface for skid resistance and shear transfer. The use of self-consolidating concrete to fill the wide shear key is also recommended to eliminate the need for grouting, which is a costly and time-consuming operation. Narrow Joint System The narrow joint system developed by UNL does not use dia- phragms and incorporates top and bottom transverse ties rather than a single middepth transverse found in other details. Grade 75 threaded rods are used every 8 ft to connect each pair of adjacent boxes at the top and bottom flanges. These rods provide continuous connection that transfers shear and moment between adjacent boxes more efficiently than the middepth transverse ties at discrete diaphragm loca- tions. A slight modification is made to the standard box cross section by developing full-length horizontal and full-depth vertical shear keys, as shown in Figure C.5. The boxes are Figure C.1. Crack developed in joint between adjacent box beams. Figure C.2. Crack developed in adjacent box beam bridges. Figure C.3. Wide joint system box dimensions.

162 fabricated with a plastic duct at the top and bottom flanges to create openings for the threaded rods, as shown in Figure C.6. The bottom duct is inserted between the two layers of pre- stressing strands, and the top plastic duct is located 3 in. from the top surface to provide adequate concrete cover. Vertical vents are provided at one side of each box to allow the air to escape while the ducts are grouted. Posttensioned System Development A modified version of the narrow joint system has been devel- oped by UNL to allow for posttensioning of the transverse high-strength steel rods. Details of the development of the system including literature review, numerical studies, and recommended design provisions can be found in Hansen et al. 2012 and Hanna et al. 2009. The system continues to eliminate the typical use of dia- phragms and a concrete topping for load distribution, a detail common to most adjacent box girder applications. The system maintains the use of top and bottom reinforcement; however, ducts have been moved inward so that posttensioning extends between girder voids. This arrangement simplifies inspection of the posttensioning as it is now unbonded with the box sec- tion and shear keys. Construction time is reduced due to the elimination of grouted ducts, and venting is no longer an issue of girder production. A duct-within-a-duct setup is being proposed so that posttensioning can be applied after grouting of the shear key. The interior duct is necessary; without it, the duct would close during grouting, and posttensioning could not be threaded between adjacent members. Posttensioning should be applied after the shear key is placed for the joints to be placed under initial compression. Figure C.7 is a diagram of the proposed modifications to the current AASHTO PCI box section. The exterior ducts have a diameter of 3.5 in. to provide enough space for an interior duct, high-strength steel rods, and couplers. This duct can be formed by plastic tubing cut Figure C.4. Wide joint connection details. Figure C.5. Narrow joint system box dimensions.

163 to the appropriate dimensions. Reinforcement should be provided for confinement about these locations. The diam- eter opening of 3.5 in. is used to provide a moment arm of 12 in. and still keep the duct flush with the top and bottom flanges. Shear keys extend the full horizontal length at the top of the girders and the full vertical depth every 8 ft along the member. Figure C.8 shows how the end girders must be made flat at the exterior face of the bridge in order to pro- vide an even surface for bearing of the posttensioning materials. The high-strength steel rods should be cut at even-length increments of 4 ft in order for couplers to be placed at the joint between adjacent members. This is done for ease of construction. It is also recommended that an 8-ft length be used so that members can be placed two at a time, thereby reducing problems with construction tolerance. Experimental Program Introduction An experimental program was developed to test the perfor- mance of the posttensioned system. A test specimen was designed using four 48-in.-wide × 27-in.-deep × 8-ft-long box beam elements that were positioned side by side and connected transversely by grouting and posttensioning. This created an overall specimen 16 ft long × 8 ft wide × 27 in. deep for testing. The length of the specimen consisted of the four box sections connected on each side. The effectiveness of the developed system was analyzed based on its joint performance under fatigue and ultimate loading. Fatigue can be represented by a 5,000,000-cycle load applied to create tension in the top and bottom trans- verse ties of the 16-ft × 8-ft × 27-in. specimen, which was used to represent a 52-ft × 64-ft × 27-in. prototype bridge. The performance of the joint was evaluated by data col- lected from strain gauge, deflection, and leakage monitoring at critical locations of tensile stress along the transverse connections. Figure C.6. Narrow joint connection details. 3.5'' 6" 5 8" 3' 6" 2" 2" 2'-3'' to 3'-6'' 112" 114" 8 ft Spacing 4' 6" 15.5'' 5.5'' Figure C.7. Posttensioned system box dimensions.

164 Construction Overview Construction of the 16 ft × 8 ft × 27 in. specimen mentioned in the system analysis began by tying together the reinforce- ment and building the interior and exterior wooden forms. Tying the bottom cage reinforcement began by laying out No. 4 U bars at an 11-in. spacing and then overlaying the rebar with No. 4 straight longitudinal bars. Markings on the steel were used to achieve the desired clear cover and spacing. A top bar served to stabilize the structure and keep the bars posi- tioned at a 90° angle from the floor. This bar was removed when joining the top and bottom cages because it interfered with the top reinforcement and transverse duct. No. 3 C bars were attached 3 in. up on the legs of the No. 4 U bars for ease of construction. In the final step, the second layer of the bottom No. 4 longitudinal reinforcement was tied to the No. 3 C bars after having been placed before the No. 3 C bars had been tied. Figure C.9 shows a complete view of the bottom cage rein- forcement. A crane was used to move the cages from the work- station to a safe location for storage. After completion of the bottom cage reinforcement, the top cage reinforcement was tied. Straight No. 5 longitudinal bars were attached to No. 4 A1 bars at 9-in. spacing, and additional No. 4 straight longitudinal bars were used on the legs of the reinforcement to provide an attachment for the No. 3 D bars at 12-in. spacing. Chairs held up the No. 4 A1 bars during cast- ing so that enough clear cover existed between the interior box form and the No. 3 D bars. Figure C.10 shows a complete view of the top cage reinforcement. 714" 734" Spherical Nut Dished PL Coupling Nut Threaded Rod @ 8 ft Spa Plan View Elevation View Interior Duct @ 8 ft Spa Exterior Duct @ 8 ft Spa Grout Shear Key Figure C.8. Posttensioned connection details. Figure C.9. Bottom cage reinforcement.

165 Once reinforcement preparations were complete, exterior supports for the interior form were built to connect the top and bottom cages. Interior forms had been constructed before this part of the experimental setup was executed. Supports were made by cutting end boards and screwing them to wooden 2 × 4s for extra bearing against the ground. These 2 × 4s were meant to provide the support for the interior forms and top reinforcement until the specimen was cast and could handle the loads. Lines were sketched to ensure screws became fully embedded into the interior form. All sides toward the inte- rior of the form were covered with oil before construction began. A saw was used to cut the end boards so that post- tensioning could be viewed from within the girder voids. Eventually this form was lifted and placed over the bottom reinforcement. The stabilizing bar at the top of the bottom reinforcement was removed to complete the section by low- ering the top reinforcement cage. One-inch individual chairs were used to hold up the top and bottom reinforcement and provide enough clear cover for the bars. The No. 3 D bars of the top reinforcement had to be supported by ties to prevent the bars from bowing in the middle and reducing the clear cover to the interior form. Foam tubes were added to create the 3.5-in. ducts for the transverse posttensioning located at the middle of each girder (Figure C.11). The tubes were cut at an angle to fit the interior box chamfer. Three-eighths-inch-diameter, 7-in.-long lag screws held the tubes in place until after the concrete was cast and had strengthened enough for pullout. These screws had to be embedded at least 4.5 in. to avoid interfering with the side forms. Additional reinforcement was added at the center of each girder to prevent cracking about this duct. The reinforcement includes No. 4 longitudinal bars above and below each foam tube and running the full length of the girder, as well as No. 4 bars cut to fit vertically along either side of the tube. The easiest way to place this steel was to drill holes in the end sections of the exterior form and remove the tubing for the bottom duct. A wood block was used to prevent concrete from flowing out of the holes. The tubing was then reattached to the structure. Lifting points consisting of 4-in. steel hoops were attached as necessary handles for rearranging individual girders after the concrete was cast and had hardened enough to withstand the transverse posttensioning. These hoops were positioned close to the end zones of each girder. Because stronger lifting points were required to lift all four girders as a single speci- men, two bent No. 6 bars were added as supports located near the inside face of each exterior girder. These bars were placed over the top of wood planks spanning the width of the girder’s exterior side forms. The main function of this board was to keep the sides from bowing when concrete was placed, as well as to hold down the interior form, which experienced uplift pressure during casting. Wood was also placed along the top and bottom edges of the side forms to help with bowing. Side forms were simplified by eliminating the chamfer originally detailed at the bottom edges. Corners where the side forms intersect with the end boards were connected by L-shaped brackets and liquid nail. Screws fixed the form sides into the concrete bed and were covered with duct tape as a precaution against concrete spillage, which would hinder form removal. Additional braces against bowing were located at the middle of each girder. Ready Mix Concrete Co. delivered concrete for casting at 9:00 a.m. on September 13, 2010. The order was for 8-ksi concrete with a 10-in. slump. Admixture was used to increase the flowability to greater than a 22-in. spread. Concrete was poured along one side of the forms until it flowed under the interior form and could be seen from the opposite opening. Problems with flowability of the first two specimens resulted in the need to cut holes in the girders so that the concrete could be spread by vibration. Leftover concrete was used for strength specimens that gave a 28-day compressive strength Figure C.10. Top cage reinforcement. Figure C.11. Box girder without side forms.

166 of 12 ksi, which is much higher than the expected 8 ksi ordered. Burlap sacks were used to provide moisture as the specimen cured. Removal of the forms began after 1 day. All girders appeared fully developed. Pictures representing the girders immediately after casting and form removal can be viewed in Figure C.12 and Figure C.13, respectively. An unforeseen issue with bearing on the end girders required a second form setup to grout the exterior shear keys. This pro- cess was similar to the initial side form setup. Issues occurred during casting because the grout leaked heavily from the form. In addition, not enough bracing was provided, and as a result the end girders look bowed. Grout was intended to have at least 8-ksi strength and reach a 28-day strength of 10 ksi. Achieving this strength is important because the girders are not meant to fail as a result of crushed bearings. It should be noted that with time the grout at the bearing location experienced excessive cracking; however, the actual specimen had only a few cracks due to shrinkage. Cracking at the bearings was most likely due to an inadequate ability to provide moisture during curing, as the majority of the section was encased by the form setup. The final step in forming the specimen was to position the box girders adjacent to each other so that preparation could begin for pouring the interior shear keys. A continuous poly- vinyl chloride pipe was inserted into the duct to keep the hole open for posttensioning. Posttensioning was tightened to 5 kips before grouting. This was done as a way to get the boxes as close together as possible before grout was poured. Post- tensioning materials included a 1.5- × 6- × 6-in. dished plate, spherical nuts, couplers, and 8-ft–1-in.-diameter threaded rods with 150 ksi ultimate strength. This material was manu- factured by Williams Form. Grout used for the shear key was a semifluid, nonshrink mix that did not contain corrosive chemicals that could pose a risk to the posttensioning. The grout was the same as used at the bearing locations and specified with the same design strength as the girders (8 ksi). Similar to the bearing grout, the joint grout reached a 2-day strength of 10 ksi. Once the joint grout reached 8 ksi, the posttensioning was jacked to 0.7 × fpu × Aps (amperes), or 90 kips, according to AASHTO specifications. Jacking started at the top and was done in cycles of 30 kips plus the load already applied to the bar not undergoing additional stress. Figure C.14 and Figure C.15 represent the shear key grout- ing and posttensioning process, respectively. Test Setup Instrumentation The final stage before proceeding with the experiment was to set up the actual test frame. Both top and bottom tension were tested for the specimen, meaning two test frame setups were necessary. Top tension was deemed the critical case to test first as it results in cracking at the top of the member Figure C.12. Box girder immediately after casting. Figure C.13. Box girder without interior and exterior forms. Figure C.14. Specimen after grouting shear keys.

167 where an actual girder is most susceptible to environmental corrosion. Cracks were monitored by leakage from a dam built around the middle joint, as well as white latex paint to aid with visibility. Beam supports for both cases were placed at stiff locations, at webs and joints. For top tension, the beams were placed at the middle joint and edge of the specimen so that a load could be applied at the center of the connection between the two cantilevered girders. Bottom tension was created by placing the supports at either end of the specimen with the load applied at the center. A neoprene bearing pad was placed between the specimen and the beams to provide additional flexibility for the system. The specimen was able to handle its own weight as it was lifted to the initial top tension test frame setup. No visible dam- age took place. After placement, a beam had to be set on top of the specimen to prevent it from lifting off its support under loading. This beam was placed as close to the end as lab condi- tions would allow. No visible damage took place when the spec- imen finished the top tension test and was lifted again in order to rearrange the supports for the bottom tension setup. An actuator was used to apply 5,000,000 cycles of loading to a 15- × 15-in. bearing plate on the specimen. Five strain gauges were placed to monitor changes in strain near the load and middle joint. Pictures of the test frame setups are shown in Figures C.16 through C.18. After fatigue testing, the specimen was moved to concrete blocks so that it could be tested for ultimate capacity. Bearing was provided by 4- × 8-in. planks at each end of the specimen. The specimen did not crack during lifting and placement. A loading cell was placed directly over the center joint to produce tension once more in the bottom flange (see Figure C.19). A deflection gauge was placed beneath the specimen to take measurements along with the five strain gauges at the joint. Figure C.15. Jacking the specimen to AASHTO specifications. Figure C.16. Initial specimen lifting. Figure C.17. Actuator at exterior joint for top tension setup. Figure C.18. Actuator at interior joint for bottom tension setup.

168 Loading The loads for testing were determined by using the theory of the previously developed 3-D models of the adjacent box girder system. The first model was used to determine the axial force produced by ultimate and fatigue loads in the transverse direc- tion of a 52-ft × 64-ft × 27-in. bridge (Figure C.20). Bridge loads included an assumed curb and rail load equal to 0.48 kips/ft and 12-ft lanes with standard HL-93 design vehicular and fatigue truck loading. Because the weight of the structure is uniform and does not affect the transverse direction, it was not included in the design analysis. Curb and rail loads were applied to the exterior edge of the exterior box girders. Ultimate loading con- ditions included single and multiple lanes; fatigue conditions considered a single lane only. Single lanes were placed at the center of the bridge width to create maximum tension in the bottom flange of the box girders and at the edge to create maximum tension in the top flange. LRFD specifications include a 33% dynamic load factor for ultimate analysis and a 15% dynamic load factor for fatigue analysis. Long-term effects are accounted for with an infinite life factor of 2 for fatigue. This translates to load and resis- tance factor design load combinations of 1.25 D + 1.75 L for ultimate and 1.25 D + (0.75)(2) L for fatigue. Ultimate axial and fatigue demands were 80.288 and 15.318 kips, respectively. The ultimate demand is comparable to the estimated effective posttensioning force of 82 kips required by the previously developed design charts. The 82 kips is based on a span-to-depth ratio of 30, and the SAP2000 model represents a span-to-depth ratio of 28.4 (i.e., 64 ft/27 in. = 28.4). The posttensioning system for the experiment is a threaded rod, with a yield strength of 120 ksi. AASHTO estimates the effective force per duct after losses to be 0.80 × fpu × Aps = 0.80 × 120 ksi × 0.85 in.2 = 81.6 kips. This is enough to handle the anticipated loads. The second model developed was used to determine the loads required to reach fatigue and ultimate capacity of the experimental 16-ft × 8-ft × 27-in. system. Modeling dif- fered from the large-span bridge in several ways. The sup- ports were along the transverse direction, thus making the dead weight of the structure act in the transverse direc- tion. Bottom tension was created by a simple-span setup with load at the center connection between middle girders. Top tension resulted from a simple span with cantilever setup in which the load was applied at the connection of the two cantilevered end girders (see Figure C.21). Supports were assumed at 7 in. from the edge of each exterior girder for bottom tension and 7 in. from the end with supports at the connection between the two middle girders for top ten- sion. An additional support was placed at the center of the top flange near the end support to represent the actual lab conditions. Figure C.19. Actuator at interior joint for ultimate capacity test. Figure C.20. 52-ft  64-ft  27-in. adjacent box girder SAP2000 3-D model. Figure C.21. Model for top and bottom tension setups of 16-ft  8-ft  27-in. specimen.

169 Analysis showed that the axial forces due to dead weight for the bottom and top tension setups were 4.9 and 21.4 kips, respectively. A load equal to 17.4 kips brings the axial force to fatigue conditions for the bottom tension setup; however, the fatigue load is already surpassed by the top tension setup. Fatigue for this condition will instead be represented by a fatigue truck wheel load equal to 18.4 kips, which creates a tension in the top flange equal to 51.6 kips. This value is acceptable because it is less than the ultimate axial force. Ultimate capacity of the 16-ft × 8-ft × 27-in. specimen was determined by strain compatibility and found to be 280 kips/ft. A single-point load causing positive moment on a simply sup- ported beam is equal to four times the moment divided by the span length (i.e., M = PL/4). Span length, assuming a 4-in. bearing on both sides, was 15 ft 4 in. This relationship indicates that a total load of 73.0 kips will induce failure. Part of this load was due to the weight of the box (approximately 1.5 kips per linear foot). The corresponding moment due to weight would be 48.5 kips/ft. The actual load to cause failure would produce 231.5 kips/ft of moment. Solving for the load gives a 60-kip applied force to break the joint between adjacent boxes. Loads were not factored when solving for the resultant forces of the experimental setup because the analysis is meant to determine the loads required to reach the design conditions. A summary of results is provided in Table C.1. Test Results The box girder experienced a deflection of 0.10 to 0.12 in. dur- ing the top tension fatigue test. No cracking occurred in the joint; however, cracking was noticed in the top flanges of the box girder. The cracks were due to shrinkage and occurred before loading. Testing continued because the focus of the test was on the joint and not the flange. The dam began to leak over the joint near the end of the test. This formed a water stain that made the specimen joint appear failed when it was not. Strain gauge data suggested that the stress on the box girder was the same before and after top tension loading. Two of the gauges had overlapping initial and final strains. These occurred under the load and on the northwest side of the joint. The strain gauge on the southeast side of the joint initially read nearly half the average value of the other strain gauges (36 µ€). At the end of the test, however, this strain gauge had increased to the 36-µ€ average. The northeast strain gauge also doubled, suggesting that the load was being unevenly distributed to the east. The last strain gauge broke, and accurate data were unobtainable from the specimen. As mentioned earlier, leakage was noticed along the flange of the box girder. This leakage occurred mostly at a crack on the northeast side and parallel to the joint. Smaller leaks were due to cracks located northwest and southwest of the joint. These cracks occurred where the top boards for bracing had been placed during the initial casting of the individual girder specimens (perpendicular to the joint). As the measurement on the west side did not show any strain loss, it is reasonable to conclude that parallel cracking caused the strain loss on the east side. It can also be inferred that the crack caused some strain loss in the specimen that otherwise would not have occurred. Strain gauge data are shown in Figure C.22. Figure C.23 and Figure C.24 show interior and exterior views, respectively, of the northeast crack and leakage. The box girder experienced a deflection of 0.08 to 0.11 in. during the bottom tension fatigue test. No cracking occurred in the joint, and leakage decreased because the top flange was under compression. Again, strain gauge results suggested zero loss of strain. Two of the gauges had overlapping initial and final strains. These occurred under the load and on the southeast side of the joint. The strain gauge on the northwest side of the joint lost approximately 5 µЄ, and the strain gauge on the north- east side increased by approximately 5 µЄ. The southwest strain gauge appeared to overlap, but the initial readings showed a loss of strain when the load was constant. This is probably because the strain gauge should have been replaced after the top tension test. The adhesive for the strain gauges did not hold up well after being submerged in water a second time. (a) 52 ft  64 ft  27 in. Fatigue Ultimate Maximum axial force 15.3 kips 80.3 kips (b) 16 ft  8 ft  27 in. Axial Force Due to Weight Applied Load Total Axial Force Top tension 21.4 kips 18.4 kips 51.6 kips Bottom tension 4.9 kips 15.3 kips 15.3 kips (c) 16 ft  8 ft  27 in. Weight Applied Capacity Moment 44.1 kips/ft 235.9 kips/ft 280.0 kips/ft Load 23.0 kips 61.5 kips 84.5 kips Table C.1 Results of (a) Full-Scale, (b) Fatigue, and (c) Capacity Analyses

170 Both north strain gauges and the southwest strain gauge had to be replaced before obtaining their final readings. Once more, results suggested that the parallel crack caused the changes in strain. It makes sense that the northeast gauge would experience a greater strain because it was on the com- pression side of the crack, while the northwest gauge would lose strain as a result of the increasing tension on the other side of the crack. Strain gauge data are shown in Figure C.25. The ultimate load taken by the specimen was 67 kips applied and 90 kips total. This was greater than the expected 61.5 kips applied and 84.5 kips total loading from hand calculations. The joint cracked evenly across the bottom and extended up to where the web indents and forms the shear key that runs the entire length of the specimen (see Figure C.26). Deflection stayed evenly below 0.1 in. for approximately 60 kips before increasing exponentially to 3.31 in. between 40 and 67 kips (see Figure C.27). Conclusions and Future Work Conclusions The general objective of this research topic was to improve the performance of the transverse connections currently used in adjacent box girder bridge applications. This was done by devel- oping a detail capable of transferring moment and shear in the transverse direction. Benefits of the modified system, which allows for posttensioning, include significantly simplifying box Figure C.22. Initial and final strains on specimen (top tension). Figure C.23. Interior view of cracking at flange along northeast side of joint. Figure C.24. Exterior view of cracking at flange along northeast side of joint.

171 production, improving the rate of construction, allowing easier inspection of voids, and reducing projects costs. It was found that posttensioning also increases the capacity and efficiency of the section because joints are placed under compression and are less likely to experience reflective cracking and leakage. The experimental program under this research topic con- sisted of applying an 18.4-kip load for 5,000,000 cycles at an exterior joint of the four-box specimen with supports at the center and opposite edge. This produced tension at the more critical top flange, where the system is vulnerable to environ- mental chemicals and corrosives. Cracking was monitored by a dam over the joint and five strain gauges placed next to the load and on either side of the center joint. No cracking or strain loss occurred during the course of the experiment. Supports were then rearranged to both ends of the member, and the load was repositioned to the center joint for tension in the bottom flange. A 17.4-kip load was applied for 5,000,000 cycles, and again no cracking or strain loss occurred. Ultimate capacity was calculated as 280 kips/ft using strain compatibility and found to be slightly higher during testing. An applied load of 67 kips, in addition to the self-weight of the specimen, reached 300 kips/ft before failure. Cracking did not propagate until after reaching a 60-kip total load, and then deflection increased exponentially from 0.1 to 3.31 in. over the next 30 kips applied. Based on the test results, it can be concluded that a post- tensioned transverse connection without diaphragms or a concrete overlay can be designed and detailed to have compa- rable performance to typical connections while being more economical and practical. The tested specimen had excellent performance under both static and cyclic loads. Future Work It is proposed that the transverse, posttensioned connection developed in pursuit of this research be applied in a full-scale application for observation of the joint performance under actual weathering and loading conditions. Figure C.25. Initial and final strains on specimen (bottom tension). Figure C.26. Side (bottom left) and underside (bottom right) cracking.

172 Figure C.27. Total load versus deflection of specimen.