In March of 2005, an inspection team led by Weidlinger Associates Inc. completed an in-depth inspection of the segmental portion of the Red River Bridge in Boyce, La., for the Louisiana Department of Transportation and Development (LADOTD). The overall bridge length is 3,067 ft with the central 1,797 ft 6 in. constructed as a segmental structure. It was Louisiana’s first concrete segmental bridge and was constructed in 1985. The project was originally designed as a precast segmental bridge but was changed by the contractor, J.A. Jones, to a cast-in-place segmental bridge by means of a value-engineering proposal.
The structure was built using the balanced cantilever method employing form travelers to cast the segments. The six segmental spans form a continuous unit of 1,797 ft 6 in. in length and feature a maximum span length of 370 ft.
The bridge, a single-cell box girder, carries one lane of traffic plus a shoulder in each direction. The road is designated as LA 8.
As is typical for segmental bridges built using the balanced cantilever method, the Red River Bridge is post-tensioned with two types of tendons: cantilever tendons and continuity tendons. The continuity tendons are further subdivided into bottom-slab and draped tendons. Cantilever tendons are located in the top slab and are installed during the construction of the cantilevers. In the finished structure, they provide the negative-moment resistance at, and adjacent to, the piers.
The continuity tendons are installed after adjacent cantilevers are completed. The bottom-slab continuity tendons make the structure continuous and provide the positive-moment resistance within the span. The draped continuity tendons assist in the positive-moment resistance, provide additional negative-moment resistance as they drape up and over the piers and, due to the drape, help counteract vertical shear.
In addition to a general inspection of the bridge’s condition, a special task was added to the scope. This task addressed a detailed inspection of the post-tensioning strands to determine if any corrosion or voids in the grout protection were present. Most post-tensioned bridges utilize a multistrand tendon composed of anywhere from nine to 19 strands. Strands are composed of seven wires twisted together. The completed strand is either 0.5 in. or 0.6 in. in diam., with 0.6 in. being the more common usage for bridges.
The wires of the strands are the singular element that provides the strength of the bridge, therefore even a partial loss could have serious consequences. As discussed below, wire corrosion became a major concern across the industry in 2000 and prompted many owners to perform an in-depth inspection of their post-tensioned bridges.
Scope it out
The Red River Bridge inspection started with a review of available information that included current maintenance reports, inspection records, as-built drawings, shop drawings, casting dates and post-tensioning system details. After a period of familiarization, a one-day site visit was made in conjunction with LADOTD personnel in charge of previous inspections of the bridge. The purpose was to assess the general condition of the bridge and its accessibility and to determine the confined-space program requirements.
During this visit, the Weidlinger team tested the atmosphere in the interior of the bridge for potential hazards. A specially equipped rescue service team that was trained in confined-space environments was available at the site to assist the inspector.
Prior to starting the inspection, a time-dependent analysis (TDA) of the segmental bridge was performed to determine which tendons were critical to the bridge’s structural integrity and serviceability and which were no longer essential. Tendons that were essential during the balanced cantilever construction may not be essential to the load-carrying capacity of the completed bridge.
The Bridge Designer II (BD2) version 3.05 software was used for the TDA. BD2 uses basic matrix structural analysis formulation, combined with time-dependent material properties, to carry out a time-step simulation of a structure as constructed. It is an appropriate tool to evaluate time-dependent behavior of concrete structures since it is built upon a time-dependent stiffness solution and provides a flexible environment for construction simulation.
Weidlinger’s engineers followed the TDA methodology used by the Florida Department of Transportation (FDOT) in determining the critical tendons to be inspected during the repairs of Bridge 860510, I-95/Sawgrass Interchange near Fort Lauderdale. The Weidlinger analysis allowed a refinement of scope and a reduction in the field inspection effort by examining only those tendons judged to be critical to the bridge’s structural integrity. The bridge contains 194 top-slab tendons (cantilever tendons), 78 bottom-slab tendons (continuity tendons) and 20 web tendons. The analysis showed that about 70% of the top-slab tendons and all of the bottom-slab tendons were critical to the bridge’s structural integrity. None of the web tendons was judged to be critical.
The anchorages and tendon locations selected for inspection on the Red River Bridge were determined statistically from the critical tendon database. The inspection plan called for inspecting 98 top-slab (cantilever) tendon locations (46 of which were anchorages). Approximately 30% of the top-slab critical tendons were examined. For the bottom-slab (continuity) tendons, 54 locations (27 of which were anchorages) were examined. This represented approximately 38% of the bottom-slab tendons.
Inspection responsibilities included inspection of the deck, deck geometry and longitudinal profile, exterior superstructure, interior superstructure, bearings, expansion joints and drainage system.
For the deck, a close visual inspection was made to detect any signs of staining, cracking or delamination in the underside of the deck. The top surface, which is an overlay, was visually inspected for any signs of distress. Additionally, a longitudinal survey was performed to compare the existing profile with the original design and constructed profile. The profile will be used as a benchmark for future bridge surveys.
A close visual inspection was performed of the bridge’s exterior including the webs, the bottom slab and the underside of the wings for any signs of cracking or other distress. All the interior surfaces were visually inspected for signs of cracking, delamination or other distress.
The bearing configuration of the six concrete box girder spans consists of guided sliding elastomeric bearings on the outermost piers, unguided elastomeric bearings on the next pair of inner piers and fixed concrete pintles on the three center piers. The bearings were visually inspected for lift-off, separation between plates and PTFE or elastomer, cracking or corrosion of the PTFE and stainless-steel sliding surface, bulges and tears in the elastomer and excessive deformation or misalignment of the elastomer or other components.
The impulse radar method was used to exactly locate the embedded tendon locations and anchorages. Anchorages at interior diaphragms and bottom-slab blisters were exposed for inspection by endoscopy. Selected tendons in the top and bottom slab were evaluated by drilling and borescope inspection to determine the grout quality around the strand and to determine if there were any voids and, if so, if there was any strand corrosion.
Location of inspection points: The tendon geometry shown on the as-built plans was used to find the approximate location of the inspection points. Thereafter, impulse radar was used to pinpoint the embedded statistically selected tendon locations and anchorages.
Anchorages: The selected anchorages in the top and bottom slab were evaluated by drilling and borescope inspection to ascertain the quality of the grout around the strands and to determine if there were any voids. The anchorages at the top slab were located using the tendon geometry shown on the as-built plans.
Tendons: The selected tendon sections (in the free length) in the top and bottom slab were evaluated by drilling and borescope inspection to determine the grout quality around the strand and to determine if there were any voids.
The determined condition of the bridge is excellent. A total of 152 tendon sections were exposed and examined for grout voids and strand corrosion, 54 in the bottom slab and 98 in the top slab. A total of 46 sections of the 98 were inspected near anchorages. All the locations inspected, with the exception of one, were found full of grout. Only one location was found with a small grout void at the north end of Span 14. However, there was no evidence of corrosion on the exposed strands.
In the case of the bottom-slab tendons, 54 locations were inspected, 27 near anchorages at blisters. The inspection showed no sign of voids or corrosion in any of the 54 locations.
Protection in triplicate
The post-tensioning strand corrosion issue came to light during the course of three bridge inspections in Florida in 1999 and 2000. FDOT discovered corrosion in a tendon in the Niles Channel Bridge and in several post-tensioning tendons of the Mid Bay Bridge and the approaches for the Sunshine Skyway Bridge. These bridges are precast segmental structures that utilize external tendons. The bridges were erected using the span-by-span method. The tendon problem found on the Skyway Bridge was a vertical application used in a precast segmental pier column.
The discovery prompted FDOT to undertake an in-depth inspection of all of its post-tensioned bridges in 2000 and 2001. Florida has the most segmental bridges in the U.S., according to the Federal Highway Authority’s (FHWA) 2004 National Bridge Inventory.
Additionally, it caused concern throughout the bridge community wherever grouted post-tensioned tendons were used in bridge construction.
Internal tendons, in other words, tendons embedded inside the concrete, are protected by three levels of corrosion protection: ducts, grout and the concrete in which they are embedded.
External tendons also have three levels: ducts, grout and the dry interior of the box girder.
The fundamental problem was a compromise of the strand protection system. For corrosion to take place, an exchange with air and water is necessary. If the integrity of the anchorage area and the duct is maintained, then this exchange cannot take place. Several inspections revealed voids in the grout around the anchorage area. But because the protection system was not compromised as it relates to air and water intrusion, no strand corrosion was detected. The void was simply injected with grout and the system was restored to its original function.
An interface of strand with air or water can arise from leakage at a joint—either the joint that seals segments to each other (epoxy) or the expansion joint—and a break in the duct. While not desirable, a leak at the deck joint can be overcome if the anchorage protection system remains intact.
One other breakdown in the protection system was observed: a separation of the water and cement that made up the grout mixture. Since water has a lower density than the cement, the water rose to the high point of the tendon profile while the cementitious material settled away from the high point. If the consolidation of the cement was extensive, it could leave the strand exposed within the duct. In other words, the strand would not be encased in a homogeneous grout and one of the defenses against corrosion would be lost.
For the corroded tendons, FDOT found an air-water interface present through either a break in the duct or a compromise in the anchorage protection system at a joint that leaked. The fundamental problem was identified to include the properties of the grout mix and its ability to separate; poor workmanship; deficient post-grouting inspection of the problem areas; insufficient protection details at the anchorages; and chemical composition of the HDPE duct material.
Changes in practice
FDOT took a leadership position in resolving the problem with the support of FHWA and the American Segmental Bridge Institute (ASBI; www.asbi-assoc.org). FDOT formed a task force of its own staff and consultants. After issuing an interim design memorandum, the work of the group was compiled into 10 volumes. These volumes are available from FDOT’s Structures Design Office webpage (www.dot.state.fl.us/structures).
The Florida specifications and details have pretty much become the standard for detailing, grouting and protection of post-tensioning strands. The major changes include:
- Use only plastic ducts except for galvanized-steel pipe for areas of sharp curvature; mechanically attach ducts to anchorages; positively seal all duct connections;
- Dry grout ingredients shall be pre-bagged and pre-approved; no onsite mixing of dry ingredients as was previously practiced;
- All anchorages shall be accessible for inspection after grouting; after grouting, probe all high points and anchorages for voids;
- Grouting and inspection shall be performed by certified personnel. ASBI has developed a certified grouting program that is accepted by most owners;
- Protect anchorages with a four-level system including grout, a permanent grout cap with sealing ring, an applied coating and enclosure in a concrete structure or pour-back;
- Seal ducts at all times prior to introduction of strand;
- Use epoxy on both faces of segments to be joined; and
- Requirement for a pressure test of each duct system prior to commencement of grouting operations.