Repair and Maintenance of Post-Tensioned Concrete Bridges (2021)

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.

2 Post-Tensioning Basics Post-tensioning (PT) systems are used in bridge applications in superstructure and sub- structure members to induce a prestressing force to the concrete section; this prestressing force causes a precompression of the concrete, which is beneficial in resisting induced tensile stress. The precompression of the concrete section—generally introduced prior to service—improves crack control, deflection behavior, and durability. Post-tensioned bridge structures are an eco- nomical construction type for spanning long distances. PT is a form of prestressing that is accomplished after some initial curing of the main concrete section. In this method, a conduit or duct is cast into the concrete cross-section during place- ment of the concrete. A prestressing strand or bar is passed through the duct and anchored at each end to the concrete section. Prestressing strands are typically helically wrapped, seven-wire steel (see Figure 1). A force applied to the prestressing steel delivers the precompression to the main concrete section through the anchor points, or anchorages (see Figure 2). Prestressing strands are terminated in an anchorage (typically a steel plate) and held in place with wedge fittings. All force delivery occurs at the anchorages. The PT tendon applies both a compression force aligned with the axis of the beam and a transverse force, depending on the tendon profile. For example, angular changes of the tendon profile, created by deviator blocks or through a parabolic profile, can cause an upward or downward component of force that can be calibrated to balance structural loads. After installation and stressing of the tendon, the space in the duct typically is filled with a cementitious material—commonly referred to as grout. Grout provides both corrosion pro- tection to the prestressing steel and—in the case of internal tendons—bond to the primary cross-section, which is an important condition considered during design. Early post-tensioned bridge structures used mixtures of cement and water, or other nonproprietary grout mixes, as filler materials. Since the early 2000s, most agencies have shifted to specifying the use of pro- prietary, prebagged thixotropic cementitious grout blends (Federal Highway Administration 2013). Alternatively—although much less common in the United States than in some countries in Europe—the duct can be filled with a non-cementitious material (termed flexible filler in this document) that provides protection without bond. Several non-cementitious flexible filler materials have been used for the corrosion protection of prestressing steel in post-tensioning tendons: waxes, greases, and polymer gels (Brenkus et al. 2017b, 2018). Internal tendons are in direct contact with the primary concrete member, while external tendons are placed outside of the cross-section of the primary member and are anchored directly at the ends. The standard practice in the United States to construct both internal and external tendons has been to use prestressing strands, sheathed or placed in plastic ducts, with cementitious C H A P T E R 1 Literature Review and Background

Literature Review and Background 3   grout as a filler material. Both types of tendons may experience durability issues associated with construction practice and filler material quality; rarely have internal tendons been identified with evidence of corrosion. Lacking complete encasement in the primary concrete member, external tendons may experience an increased occurrence of durability issues. Post-tensioning systems used in bridges are usually one of several commercially available proprietary products. A PT system is a multiple component construction composed of pre- stressing strands, an encompassing duct, a filler material, and an anchorage. The anchorage, which in post-tensioned systems is the location of all force transfer from the prestressing steel to the concrete, is an assembly of components: a cast or machined metal body with a bearing surface, a wedge plate, and strand wedges. In most post-tensioning applications in the United States, the utilized prestressing steel is a seven-wire, helically wrapped strand. Threaded post- tensioning bars are another commonly used option. Post-tensioning tendons are used exten- sively in segmental and I-girder bridges, as well as in box girders, pier caps, bridge decks, and slab bridges. Figure 2. Typical post-tensioning multistrand anchorage. prestressing steel filler materialduct Figure 1. Typical post-tensioning tendon cross-section.

4 Repair and Maintenance of Post-Tensioned Concrete Bridges Post-tensioning tendons can be either bonded or unbonded—a categorization based on the tendon’s contact with the surrounding concrete section and indicative of stress-strain com- patibility between the tendon and the surrounding concrete. Tendons filled with cementitious materials can be either bonded, as in the case of internal tendons, or unbonded, as in the case of external tendons (such as those used in some segmental construction). The use of flexible filler materials in the ducts, such as greases and microcrystalline waxes, is a recently introduced con- cept to bridge construction in the United States. The use of these materials results in unbonded tendons that are not bonded to the surrounding concrete section in such a way to ensure stress- strain compatibility (Brenkus et al. 2017a). Although PT systems and components must pass acceptance testing prior to their use, rigorous studies of in-service PT systems and their specific components have been limited. Isolated cases of repair-requiring issues in PT bridges have been documented, but given the uniqueness of each situation, responses vary widely. Without a national guidance document, post-tensioning repair efforts largely rely on the expertise of the coordinating agency, consultants, and the con- tractors involved in the repair. This document provides a summary of current PT practices used in bridge design across the United States. History of Post-Tensioning in the United States Post-tensioning as a technology has been used in the United States since the 1950s, when it made its first appearance in lift-slab building construction. From its initial use to address issues with deflections and reduce member weights in lift slabs, post-tensioning expanded into other types of construction with significant technology transferred from established practices in Europe. In 1954, the first concrete bridge using post-tensioning bars was constructed in Florida. Addi- tional projects soon followed, with post-tensioning used to improve serviceability and lengthen spans. In 1962, the first strand post-tensioning system was introduced in the United States, an important evolution advancing the technology. The load-balancing method was popularized by T.Y. Lin starting in 1963, easing the burden of design by conceptualizing the PT tendon as “just another load.” The founding of the Post-Tensioning Institute (PTI) led to many improvements in the corrosion resistance of PT systems (Bondy 2018). Prestressing was added to the American Concrete Institute’s Building Code Requirements for Structural Concrete (ACI 318) in 1963; many improvements specific to prestressing were added to ACI 318 in 1977 and 1989, including minimum requirements for bonded reinforcement, dis- tillations of recent test findings completed at the University of Texas at Austin and the Univer- sity of Washington, and greater consideration of indeterminate structures (most post-tensioned structures). Although ACI 318 is the guiding document for building design, these advancements contributed to the development of expertise and the use of PT for all structures in the United States (Bondy 2018). In 1979, two additional advancements in post-tensioning were introduced to U.S. bridges: draped tendons in girder webs and segmental structures. Loss of prestress force is the principal concern in PT structures and can occur through several mechanisms, including thermal effects, creep, relaxation, or corrosion. Creep refers to compression-induced strain in the concrete that occurs over the life of a structure. Prestress relaxation occurs when prestressing steel experiences tensile strain over the life of the structure and loses some of its prestressing force. Creep and relaxation are critical factors in the design of PT structures because of the constant application of prestressing, but they are not fully understood for large-scale members. Excessive creep and prestress losses can lead to excessive deflections in the member and can cause cracking.

Literature Review and Background 5   Issues are usually identified through the observation of crack formation and growth. A recent example of rapidly occurring structural degradation and “exponentially growing” cracks in a PT segmental structure is the West Seattle High-Rise Bridge. Though currently the bridge is under investigation and the cause of the cracking is unknown, the rapid crack growth observed is thought to be creating a collapse mechanism (Banks 2020). It is hypothesized that the crack growth has been accelerated by creep and thermal effects. By far, the most significant problem facing post-tensioning has been tendon corrosion. Although significant advancements have been made to PT technology in general, corrosion remains an issue requiring careful consideration during design, detailing, construction, and maintenance. Early tendon sheathing and grease materials used in PT building construction were eventually deemed inadequate for aggressively corrosive environments; significant efforts to address these issues have led to improvements. Materials specifications and guidance docu- ments developed by PTI for sheathing materials (including an important switch from paper wrapping to plastic conduits), strand coatings, and tendon encapsulation techniques have largely resolved early corrosion problems encountered in buildings (Bondy 2018). The concepts and the lessons learned from this construction type can, in part, translate to understanding and improvements in the bridge industry. Similar approaches—improvements in materials specifi- cations, detailing, and comfort with the technology—have improved the finished products and durability of bridge structures in states that have invested in these efforts. Tendon Corrosion Tendon corrosion is regarded as the most significant problem encountered in post-tensioned structures because strand loss can compromise structural integrity and can lead to catastrophic failure (Bondy 2018). Corrosion can take many forms, including electrochemical corrosion, stress corrosion cracking (SCC), and hydrogen embrittlement. Corrosion of concrete reinforcement most commonly occurs as an electrochemical process when corrosive agents, such as chlorides, lower the pH of the concrete surrounding the reinforc- ing steel and the ions react with the steel reinforcement to produce corrosion byproducts. Electro- chemical corrosion can occur in PT tendons with inadequate grout, corrosion protection, or both. Electrochemical corrosion cells can form in reinforced concrete as macrocells, in which the separation between the anode and cathode is large (such as between two strands or rebar), or as microcells, in which the separation is small (such as along the same strand or rebar). Corrosion is significant because it can reduce the capacity of the reinforcing steel, as well as cause delamination of the surrounding concrete or grout. The two primary sources of chlorides are deicing chemicals and seawater (Smith and Virmani 2000). The minimum concentration of chlorides required to initiate corrosion is referred to as the chloride threshold level. When the chloride threshold limit is reached within a concrete structure, the conditions are such that local corrosion initiation is possible (Anania, Badalà, and D’Agata 2018). A comprehensive literature review of chloride threshold values for grouted PT tendons conducted by the Federal Highway Administration summarizes historical chloride threshold values for metals in cementitious materials as reported in literature and state inspection reports (Virmani and Ghasemi 2012). Post-tensioned structures are not necessarily more susceptible to chloride intrusion; chlorides are believed to diffuse through PT structures the same way as in other structures. However, post- tensioned concrete is more sensitive to the presence of chlorides than reinforced concrete because prestressing steel has been shown to corrode at a faster rate than mild steel. Further, the con- sequences of corrosion from chloride intrusion of PT structures can be catastrophic. Because PT design relies on the development and long-term maintenance of a specified prestress, the loss

6 Repair and Maintenance of Post-Tensioned Concrete Bridges of tendons, the loss of prestress due to corrosion, or the loss of both can lead to serviceability concerns (i.e., excessive deflections) or tendon failure. Extreme instances have the potential to lead to structural collapse. Post-tensioning tendons are also vulnerable to stress-related corrosion. Because of the high stress state of PT tendons, limited corrosion can cause brittle failures in the steel due to pure stress corrosion or hydrogen embrittlement (Schupack and Suarez 1982). The interaction between corrosion and mechanical stress can cause SCC—a brittle corrosion failure that exhibits qualities resembling fatigue failures. Prestressing steel is vulnerable to SCC because of the cold forming manufacturing process that creates its high tensile strength. The process applies stress to the steel to induce plastic deformations and modifies the material’s microstructure (Toribio and Ovejero 2005). Constant prestressing force provides the stress necessary, in the presence of corrosion inflections, to initiate intergranular microcracks in the steel (Perrin et al. 2010). SCC is serious because it can cause a brittle failure—a loss of strength without evident metal loss—making it difficult to detect (Arup and Parkins 1979; Ashby and Jones 1986). Hydrogen embrittlement often occurs in conjunction with SCC. In the low-pH environment required for corrosion, hydrogen penetrates the steel; this can have a large impact on fracture mechanics (Perrin et al. 2010). The high stress level in post-tensioning tendons is known to accelerate the steel corrosion process (Cavell and Waldron 2001; Vu, Castel, and François 2007). Issues with PT systems in bridge structures were initially identified in 1971, when localized corrosion was identified on anchorage blocks on the old Skyway Bridge spanning the Tampa Bay in Florida. Several post-tensioned bridges have had issues with severe corrosion of the pre- stressing steel, with some cases of individual tendon failure. The Niles Channel Bridge in Florida started several research efforts to investigate issues with grout-related corrosion. Issues related to deficient grout and other pathways leading to corrosion (such as improperly filled vents or anchorage caps, or the use of dissimilar grouts) warrant serious investigation. One lesson learned by the PT building industry is that significant cost savings and conve- nience can be had by replacing single unbonded tendons with flexible filler material, rather than replacement of entire slabs (Schwager, Schwager, and Schwager 2018). Other lessons have been learned in PT buildings as the first structures, built in the 1970s, have aged and some have required repair. These lessons may improve PT bridge performance and durability if they can be transferred from the building industry. Grout There are several sources of damage to post-tensioned structures—many of which are com- mon to other types of bridge construction—including vehicle impact, environmental distress caused by saltwater or moisture exposure over time, construction error, poor maintenance prac- tice, load-related damage, natural disaster, and extreme events (Harries, Kasan, and Aktas 2009). Specific to PT structures, issues related to the commonly used cementitious grout materials used as filler in tendons have been identified as significant contributors to damage. Grout issues include poor grout quality, poor/incomplete filling of the PT tendon with grout, contaminants in grout, and underweight bags. Grout deficiencies, such as those encountered prior to 2000, have not been completely resolved through the use of prebagged, thixotropic grouts. Also, newer grouts have demonstrated the potential for grout segregation, soft grout, excessive bleed water, and high chloride and sulfate contents (Gee 2011; Theryo, Hartt, and Paczkowski 2013). Grout deficiencies, including their causes and remediation actions, have been investigated by several researchers and agencies. Notable research efforts have been funded by the Florida Department of Transportation (FDOT) and the Minnesota Department of Transportation.

Literature Review and Background 7   Evidence of soft grout—unset grout with unhydrated cement products as described by Randell et al. (2015)—has been uncovered in several Florida bridges, including the Wonderwood Con- nector and the Mid-Bay Bridge (Hartt and Venugopalan 2002). Soft grout is generally located at the top of a duct (Randell, Aguirre, and Hamilton 2015). Work by Randell et al. (2015) suggests that grout storage conditions and the presence of inert filler materials may play a role in the formation of soft grout. The researchers also found that proprietary grouts exhibit observable sensitivity to water-to-cement ratios outside of the manufacturers’ recommended value. When combined with the possibility of underweight prepackaged grout bags, this finding could have significant implications in the quality of the final product. Several practical recommendations to reduce the occurrence of soft grout include storing grout away from water sources (to prevent unintentional cement hydration before use); ensur- ing the quantity of mix water reflects accurate bag weight [or test the range of mix water beyond the range recommended by the manufacturer to ensure an understanding of the materials’ sen- sitivity (Randell, Aguirre, and Hamilton 2015)]; and carefully considering the use of inert filler materials in plain grout formulations (Hamilton et al. 2014). Grout bleed, or bleed water, is the water that rises to the surface of grout while it is still in a plastic state. Bleed water is created when the grout’s constitutive materials not used by the hydration reaction separate and settle, allowing excess water to rise toward the surface. Because PT tendons are theoretically closed systems, the bleed water is unable to escape the system. Grout bleed can become a major issue if excess water floats to the top of the grout and results in a pocket of water (or a void if the water is able to evaporate) that may lead to corrosion of PT ducts (Pielstick and Peterson 2002). The quantity of bleed water increases as the water-to-cement ratio increases, and has been observed to be influenced by wicking action. In 2002, H. R. Hamilton conducted grout bleed testing at the FDOT Structures Research Center on four prepackaged grouts in an effort to improve agency specifications concerning post-tensioning corrosion protection. The project was initiated because of problems with an earlier generation’s quality that had sparked the development and availability of new products for grouting post-tensioning ducts. This new generation of products required thorough testing before their use. The research effort investigated multiple variables, including types of cementi- tious grout, types of corrugated ducts, number of strands, and elevation change of the tendon profile (Alvarez and Hamilton 2002). Revealed through field observations, bleed water prevalence in tendons with a primarily hori- zontal profile is distinct from its prevalence when the tendons are vertical. This difference was investigated for cementitious grout with the goal of limiting bleed water in PT tendons. To broaden specifications to address both horizontal and vertical applications, several tests were performed: wick-induced bleed tests, inclined bleed tests, and a relative fill evaluation on hori- zontal corrugated ducts of different types. Schupack pressure tests were conducted to examine the effect of temperature and mixing time on bleed water formation, as well as to compare grout performance. It has been demonstrated that the generation of bleed water in grouts is influenced by wicking action, or capillary action; prestressing strands embedded in grout can serve as “wicks” in the tendon, facilitating water draw from one location to another. The intent of the wick-induced bleed test is to simulate this mechanism to evaluate a grout mixture’s sensitivity to wick-induced bleed. In the wick-induced bleed tests used for this study, clear, vertical polyvinyl chloride (PVC) pipes up to 25-ft tall were used to mimic post-tensioning ducts. Each vertical pipe con- tained either a single or a bundle of three ½-in. diameter, seven-wire prestressing strands. The setup height ranged from 10 to 25 ft with strands centered in the pipe. Grout was mixed and injected into the pipes; several grout manufacturers’ products were included in the test program. Following grout injection, each pipe was visually inspected and marked, and lengths were recorded periodically for bleed water at the top and along the length of the grout column.

8 Repair and Maintenance of Post-Tensioned Concrete Bridges Schupack pressure tests are used to examine the effect of temperature and mixing time on bleed water generation. In a Schupack pressure test, the mixed grout is placed in a pressurized vessel and the pressure is incrementally increased (see Figure 3). Bleed water quantity is measured at specified intervals and is reported as a percentage of the original sample volume. Alvarez and Hamilton (2002) conducted Schupack pressure tests on four grouts, evaluating relative performance of the materials, as well as the influence of temperature and mixing time on bleed water formation. A comparison was conducted to examine the effect of duct corrugation on the ability of the grout to completely fill the duct. Three distinct, currently available styles of ribs in corrugated duct were evaluated. The first type of duct had parallel ribs, oriented perpendicular to the run- ning axis of the duct. The second type had spiral ribs. The third type had parallel ribs similar to the first, but with four additional longitudinal ribs that are parallel to the axis of the duct and are equally spaced radially around the circumference. Three 50-ft-long ducts, containing no strand, were filled with the same grout. Air was trapped in all three corrugation configurations (Alvarez and Hamilton 2002). In addition to horizontal bleed testing, two inclined bleed test setups were also investigated. The inclined bleed test simulates the inclined duct conditions associated with field-observed instances of grout bleed. The setup consisted of two clear PVC pipes oriented on an approxi- mately 30° slope. Each pipe contained a bundle of prestressing strands resting on the bottom of the pipe. In the wick-induced and Schupack pressure tests, several conditions were shown to influence bleed water formation. Relative bleed water quantity was shown to increase with the height of the duct and number of strands. Strand bundles were also shown to generate more bleed water Source: Adapted from Alvarez and Hamilton 2002. Figure 3. Laboratory setup for the Schupack pressure test.

Literature Review and Background 9   through wicking action than increased height with a single strand, suggesting that wicking action occurs between strands as well as between interstitial spaces between individual wires. Increased temperatures during mixing of the grout increased the bleed quantity in most of the tested grouts when performing the Schupack pressure test. Varying the mixing time affected the creation of bleed; most of the tested grouts developed more bleed water when mixing times were decreased. In general, it was shown that the generation of bleed in proprietary grouts was sensitive to both environmental conditions and procedural variables, and that the manufacturer’s recommenda- tions for mixing were significant to control bleed (Alvarez and Hamilton 2002). Torres et al. (2018) performed an evaluation of shelf life of PT grout with two goals. The pri- mary goal was to develop a test method to indirectly relate particle size growth or other cement prehydration mechanisms to the potential for soft grout development, such that PT quality could be tested for acceptance; this effort led to the creation of a modified inclined tube test (MITT). The secondary goal was to explore the sensitivity of PT grouts and their constitutive cements, supplementary cementitious materials, and admixtures to exposure conditions likely to be experienced in the field (Torres et al. 2018). The sensitivity to elevated temperature and moisture of the various constituents of PT grouts, including portland cement, supplementary cementitious materials, and powdered admixtures, was investigated. To evaluate prehydration mechanisms of grout, a MITT was developed and performed along with a number of bench-top laboratory tests to monitor changes in material characteristics in different exposure conditions and at different ages. A MITT is useful for evaluating a grout’s susceptibility to soft grout formation. In the studies to develop the MITT, several PT grouts and individual constituents were subjected to selected temperatures and humidity to simulate various exposure conditions. Four exposure conditions were assessed: (a) extreme exposure at 95°F and 95% relative humidity (RH), (b) field exposure at 85°F and 85% RH, (c)  lab- oratory exposure at 65°F and 50% to 70% RH, and (d) control exposure at 65°F and 45% to 65% RH. Small-scale samples were tested for mass gain, particle size analysis, Blaine fineness, loss on ignition, thermogravimetric analysis, and microwave moisture content, all bench-top tests used to characterize the exposure time and conditions necessary to generate soft grout (Torres et al. 2018). All of the tested grouts exhibited physical changes, chemical changes, or both during exposure. The primary mechanism of soft grout formation appeared to be the premature hydration of portland cement. All of the tested grouts, if given sufficient exposure time after injection into the tube, eventually formed soft grout. Field exposure resulted in an average time required to form soft grout of 8 days, with values ranging from 4 to 13 days; extreme exposure resulted in an average time-to-formation of 3 days, with values ranging from 1 to 7 days. This behavior was attributed to a combination of excess mixing water and the segregation of partially hydrated portland cement particles and low-density, low-reactivity fillers. Very fine particles of these materials were suspended in the bleed water, which was displaced as the larger particles settled due to gravity and collected in the high point of the tube (Torres et al. 2018). Pour Backs Pour-back details are an essential part of protecting post-tensioned concrete. Pour-back details, or “pour backs,” refer to the material cast to cover the anchorage or vent assemblies of PT tendons after the filler injection process. These details serve as the “first line of defense” at key locations in the duct where access was previously required during the filler injection. In PT bridges, there is concern that the pour-back area surrounding a tendon anchorage may be susceptible to cracking, perhaps providing a location for moisture intrusion. These regions

10 Repair and Maintenance of Post-Tensioned Concrete Bridges are often filled with a cementitious nonshrink grout to provide protection, but some agencies have reported the contractor’s use of conventional concrete as pour-back material, or the com- plete absence of a pour back (Ahmad et al. 2018). Few studies have focused on this particular region, despite the critical role pour backs play in protecting PT components from corrosion, especially at anchorage locations. Alvarez and Hamilton evaluated an epoxy pour-back material’s susceptibility to cracking under thermal cycling. A full-scale mockup representing the end of a bridge beam was constructed with a combined anchorage zone, consisting of four anchorages covered in a single pour back of epoxy grout, and subjected to thermal cycling. Following thermal cycling tests, no cracking was noted with impact-echo results, suggesting that the bond between the epoxy and concrete remained unaffected (Alvarez and Hamilton 2002). Historically, anchorage and vent pour backs have been problematic. A survey conducted in the early 1980s of transportation officials and post-tensioning professionals recommended better detailing for those regions because of their particular vulnerability to corrosion. Another study of an out-of-service concrete girder revealed more corrosion at PT anchorages than anywhere else in the tendon despite having been encased in concrete (Tabatabai, Ciolko, and Dickson 1995). A recent study analyzed the effects of effective prestress, clear cover depth, and chloride exposure on the rate of corrosion of PT anchorages using concrete pour backs and noted the serious consequences of anchorage corrosion (Li, Luo, and Liu 2017). Furthermore, the pour backs covering tendon vents can potentially expose the PT tendon to corrosive agents. Vents are located at the high and low points of PT tendons to allow the evacuation of air and bleed water during tendon grouting operations. Recent borescope investigations on a bridge in South Carolina revealed that insufficient grouting left voided vents, severely limiting the tendon’s cor- rosion protection (see Chapter 3). Historically, those regions have been filled with concrete or a grout material. One of the chief concerns about the durability of pour backs is their potential for shrinkage. Shrinkage of the pour back can result in cracks that allow moisture and contami- nants into the anchorage or vent. Current standard details for PT bridges often include four layers of protection for anchorage hardware: (a) a permanent grout cover, (b) a plastic grout cap cover, (c) epoxy grout pour back, and (d) for external pour backs, a waterproof coating covering the area of the pour back. Although epoxy grout is currently considered a durable material for pour backs, it can be vulnerable to shrinkage effects, allowing corrosive agents to penetrate (Ahmad et al. 2018). Durability Issues in Florida’s Post-Tensioned Bridges The Florida Department of Transportation has encountered and handled a number of PT durability issues, including tendon failures and severe tendon degradation. As highlighted in several of the case examples, other states’ bridge owners referred to Florida’s experiences when faced with a PT bridge issue. Further, the FDOT’s well-developed research program has funded several investigations to improve the durability of these structures. For these reasons, a specific consideration of Florida is deemed worthy of inclusion in this report. Pre-2000 By 1999, corrosion concerns were identified in segmental bridges spanning between the Florida Keys. These corrosion concerns, and a significant number of those that followed, were linked to poor grout quality. Becoming more aware of corrosion issues, the FDOT began a concerted effort to improve PT durability. Policies and methodologies related to post-tensioning are often defined by their relationship to this key turning point. Pre-2000, filler materials used in post- tensioning systems were commonly field-blended cementitious grouts consisting of cement,

Literature Review and Background 11   water, and expansive admixtures. Grout issues identified in the pre-2000 period were asso- ciated with • Presence of excessive bleed water, • Presence of voids in the filler material, • Splitting of the encapsulating polyethylene ducts, and • Water infiltration of the duct (water recharge). These issues were ultimately linked to an increased potential for corrosion of the prestress- ing strand and, in rare cases, tendon failures. Tendon failures have catastrophic implications, negatively impacting the load-carrying capacity of the bridge. Remediation and subsequent maintenance efforts to address such events have substantial costs, both monetary and intangible. Post-2000 Post-2000, the Florida Department of Transportation moved to implement changes to improve post-tensioning system performance and durability. Many of these changes are dis- cussed in New Directions for Florida: Post-Tensioned Bridges (Corven Engineering 2004), and include the following: • Improved PT details and hardware, including coupler improvements and anchorage protec- tion systems • Enhanced inspection protocols, with required training and certification of inspectors • Required use of prebagged thixotropic grouts • Installer training and certification requirements, including minimum experience for personnel Industry groups made parallel moves to improve grout integrity and overall system perfor- mance. Revisions to both the PTI M50 Guide Specification for Grouted Post-Tensioning and the PTI M55 Specification for Grouting of Post-Tensioned Structures aimed to further improve grouted tendon durability (PTI 2012; PTI/ASBI 2012). Recent Issues Recently, the FDOT has encountered several issues in its PT structures, including seg- regation of the grout (sometimes visually observable as different colors in the dried grout product), contaminants (including chlorides and sulfates), soft grout, voids, and excessive bleed water. Some of these issues resulted in detrimental impacts to tendon durability. In isolated cases, tendons have failed, losing all or partial force. Tendon issues require extensive remediation actions, including—at times—tendon replacement or the addition of strengthening tendons to accommodate force loss in compromised tendons. In recent years, additional causes for poor grout quality have been investigated, including inadequate quality assurance/quality control, improper or prolonged storage of prepackaged grout, excessive water added to the mix during the grout mixing process, insufficient mix time, uncontrolled pump pressures, grout sensitivity to extreme conditions, and mixing different brands of prepackaged grouts. Each of these contributed to further issues with grout materials used in post-tensioning systems. In response to current grouting and tendon issues (at, for example, the Mid-Bay Bridge, Sunshine Skyway, and Wonderwood Connector, among others), the FDOT has considered a number of actions. A moratorium on all post-tensioned bridge structures was considered. Another in-depth statewide maintenance inspection was performed to evaluate the existing inventory. The current materials and practices were evaluated to identify areas for improve- ment. Further, the agency considered alternative PT filler materials for two reasons: to address concerns with grout materials specifically, but also to provide tendon replacement options.

12 Repair and Maintenance of Post-Tensioned Concrete Bridges Following much consideration, the agency implemented several important changes to improve PT structure durability: (a) adoption of flexible fillers, (b) use of diabolo-shaped voids at diaphragms and deviator blocks, and (c) detailing to accommodate full tendon replacement (Robertson 2014). Two characteristics of flexible filler are viewed as advantageous: flexible filler provides corro- sion protection, and tendons utilizing flexible filler can be detailed to allow future tendon removal and replacement, facilitating later maintenance and repair of the tendon, if necessary. The agency’s motivations are reflected in its current guidance: “Design and detail all tendons that utilize flexible filler to be unbonded, fully replaceable, meet anchorage clearance requirements . . . and have clearance at the anchorages for jacking and future tendon replacement operations” (FDOT 2020). Current FDOT specifications only allow microcrystalline flexible filler materials that are heated to facilitate their injection, though other “flexible” fillers exist, such as thixotropic gels (that can be pumped without applied heat). The FDOT-approved products are stable, non- separating products, with good adhesive qualities, which are semisolid at typical ambient tem- peratures in Florida (reducing opportunities for leakage). In 2016, additional criteria and guidance were added to the FDOT specification to clarify requirements for post-tensioning system approval for use with flexible filler. Post-tensioning tendon systems are evaluated and approved through a battery of tests to ensure optimal perfor- mance characteristics; characteristics and tests performed are different from those conducted to approve PT systems for grout use. The majority of the test requirements are modeled off fib Bulletin 75, the European guidance document for technical approval of polymer (plastic) post- tensioning ducts. In general, the prescribed tests ensure that PT components (duct and duct couplers) demonstrate 1. Sufficient longitudinal load resistance, 2. Leak tightness, 3. Flexibility of the duct system, and 4. Integrity of the duct couplers. Duct couplers, in particular, are of concern because these connections are sometimes made with “heat-shrink” materials, with implicit heat sensitivity that may be detrimental during injection of heated filler materials. PT system assemblies, which are tested with all components from anchorage to anchorage, are also required to demonstrate 1. Leak tightness of the anchorage-duct assembly, 2. Full-scale duct system assembly, and 3. Leak tightness of the assembled duct system. These requirements are intended to ensure that the PT system, if properly assembled, will remain intact from the time of initial installation to the injection. History of Flexible Filler Use In 2014, the Florida Department of Transportation adopted the use of flexible filler materials as alternatives to the commonly used cementitious grouts (Robertson 2014). Specific tendon types were identified for which flexible filler would be used. The agency also required that tendon detailing accommodate future tendon replacement; a specific detailing change to facilitate tendon replacement was the requirement of the use of diabolo-type voids at deviation points. These changes were introduced as a means of improving PT durability in light of FDOT’s expe- rience with PT issues (Brenkus et al. 2017b). At least one other state (Virginia) has used flexible filler in PT tendons (in the Varina-Enon Bridge).

Literature Review and Background 13   Flexible filler materials are not new. Since World War I, flexible filler products have been used for corrosion protection of naval machinery and coastal artillery. In addition, they have been used by several industries; in the nuclear industry, they have been used in PT tendons in nuclear containment structures. The materials have been used by the U.S. nuclear industry since 1969 (Bhatia et al. 2017; Brenkus et al. 2017b). Applications Targeted for Flexible Filler in Florida The specification of flexible filler materials in FDOT projects is informed by the agency’s past experience with grout issues. The tendon profiles targeted for flexible filler are those pro- files with demonstrated past evidence of increased incidence of grout issues, or those tendons that, if lost, would cause severe consequences to the structure. Tendons with significant profile drape, for example, have been identified by the agency to have exhibited more grout issues. In addition, particularly long tendons, and PT tendons made of strands (instead of bar), are more susceptible to grout issues and, therefore, require flexible filler. Flexible filler materials are now required for all external strand tendons and all continuity strand tendons. Some applications permit the use of either flexible filler or grout, including in straight strand or parallel wire tendons—other than continuity tendons in U-beams and girders—and in bar tendons such as those found in vertical or horizontal orientation in a bridge substructure or superstructure. Some applications still require the use of grout filler material. In Florida, grout fillers are still required for transverse top slab tendons. There is concern that a reduced cross-section in the top slab—by creating “voids” at the duct locations, as would occur if flexible filler were used— would lead to cracking or susceptibilities in the deck. A solid, rigid filler is preferred in these locations to transfer the live load to the main structure. At this time, all tendons adjacent to the riding surface are filled with grout filler materials, including top slab cantilever or transverse tendons in segmental box girders. Further, tendons with a profile drape of 2 ft or less in slab- type superstructures are also required to include bonded, grout-filled tendons. Post-tensioning systems using bars are also permitted to use grout as the filler material because these systems have not demonstrated the same corrosion concerns as strand tendons. Current Practices and Lessons Learned As a preemptive measure to improve PT tendon quality, FDOT requires contractors perform full-scale tendon mockup injections to demonstrate a successful injection procedure plan on the planned tendon profiles. This requirement ensures that the proposed post-tensioning system, filler material, equipment, and injection method can fill the tendon ducts fully. Mockup injec- tions provide an opportunity for contractors and injection personnel to identify both preferred practices and potential issues prior to the actual injection, minimizing issues at the jobsite. Further, these injections are advantageous because they add to the body of knowledge regarding the use of flexible filler, including identifying lessons learned and improving future specification revisions and contractor practices. FDOT identified practices and lessons learned during the recent construction of the first structures to use flexible fillers. These practices are in addition to the formalized guidance included in the FDOT specifications (FDOT’s Structures Design Guidelines 2020). The following are current practices and lessons learned: • Worker safety protection (such as gloves, Tyvek protection, face shields) is required for heated materials. • Duct coupler/connections should not be located in a curved region of the duct. • Pre-bending of the high-density polyethylene (HDPE) duct for placement in curved regions may facilitate its placement.

14 Repair and Maintenance of Post-Tensioned Concrete Bridges • To facilitate remediation during flexible filler injection, it is useful to have the following on hand: – Buckets with water and wet rags to cover small leaks – Heat gun or torch to address clogs in plumbing – Clean-up supplies, such as sand and shovels, in case of flexible filler spillage – Tarp or other covering for areas adjacent to the inlet and outlet points (to keep these areas clean in case of material leaking) – Barrier or other protection at key areas in case of blowout • Ensure that all components of the pumping system are clear prior to injection, including plumbing on the pump and between filler reservoirs. All hoses should be cleaned and cleared immediately before injection. Clogs in hoses, if not cleared, can be transferred from equip- ment into the tendon and can lead to void formation. If using multiple barrels in parallel, ensure that all components of the pumping system are included in the recirculation proce- dure to prevent clogs. • Regarding pumping pressure increases: Like with grouted tendons, pressure spikes during injection are more likely in internal tendons than in external tendons because of the confine- ment provided by the encasing concrete. • Incidence of tendon crossover can be minimized by vacuum and pressure tests prior to the injection. Design Strategy The International Federation for Structural Concrete (fib) suggests a tiered approach to strategize the design of the PT system’s corrosion protection. The guidelines outlined in the fib Bulletin are the same as those presented in the PTI M50 Guide Specification for Grouted Post- Tensioning (PTI/ASBI 2012). A structure is designed with a designated protection level (PL), which guides design decision making of protective elements. The PL is selected on the basis of the exposure of the structure and the aggressiveness of the structure’s environment (see Fig- ure 4). Alternatively, the PL may be selected, in part, on the basis of the post-tensioning system’s potential exposure to corrosion-inducing chlorides (see Figure 4). PL-1 is designated for post-tensioning systems in mild environments with a high level of protection afforded by the structural components (see Table 1). PL-1 is the minimum protec- tion that should be provided for a post-tensioning system. Ducts in PT systems at this level can be made of either corrugated metal or plastic, but they must be completely filled with a filler material, leaving no voids. PL-1 is further delineated into PL-1A, and PL-1B. PL-2 encompasses the same design and detailing protection measures as PL-1, but the ducts must be plastic and watertight and have specific testing done to ensure airtightness. PL-3 is for the most aggressive environments, such as those encountered in marine environments or in locations where deicing salts may contact the structure. PL-3 encompasses the same considerations as PL-2, but implies the use of monitorable, electrically isolated tendons for additional protection. Post-Tensioning Durability Post-tensioned bridges are efficient in spanning long distances. When properly designed and constructed, these systems offer a durable design solution for longer bridge spans. Durability is an imprecisely defined characteristic in the field of structural engineering. In general, it is understood to mean the long-term performance of the structure during its intended design life, including its resistance to environmentally influenced degradative forces. Though occurring rarely, damage to PT systems can have significant implications. Corrosion of a post- tensioning tendon, for example, can result in a loss of structure integrity, reduction in structural safety, and the need for costly repairs such as tendon replacement.

Literature Review and Background 15   Source: Fuzier et al. 2005. A gg re ss iv ity /E xp os ur e Figure 4. Recommended protection levels on the basis of environment. Table 1. Protection levels according to PTI M50. Protection Level Aggressiveness of Environment Basic Requirements PL-1A Low Basic grout or engineered grout; nonreactive filling material; grout filling leaving no voids PL-1B Low PL-1A measures, plus engineered grout and permanent grout cap PL-2 Medium PL-1B measures, plus an envelope providing a permanent leak-tight barrier PL-3 High PL-2 measures, plus electrical isolation of tendon or encapsulation that is monitorable or inspectable at any time

16 Repair and Maintenance of Post-Tensioned Concrete Bridges Protection of PT Systems In general, ensuring durability for PT systems is related to the protection of the system against water intrusion. Bulletin 33, the fib guidance for improving durability of PT systems, identifies potential sources of durability concerns and breaks them down into two categories: failure of the external barriers, and failure of the tendon corrosion protection system. Identified pathways of water intrusion (which could carry detrimental chlorides) include the following (note that the numbers in the following list correspond to those in Figure 5): 1. Defective wearing course 2. Defective waterproofing membrane 3. Defective drainage/pipes 4. Incorrectly placed or defective outlets (reducing drainage) 5. Leaking expansion joints 6. Leaking construction joints 7. Inserts 8. Defective concrete cover 9. Incorrectly filled vents/inlets/outlets 10. Leaking metallic ducts (not shown) 11. Cracked or porous pocket concrete 12. Grout voids at high and low points Post-tensioning systems rely on multiple layers of defense to protect the prestressing strand against moisture and corrosion mechanisms. Protection is afforded by the structural concrete, the encapsulating duct material, and the filler material. Filler materials provide different cor- rosion protection approaches, depending on their type. More than simply serving as a physical barrier, cementitious grout materials provide a passivating layer. Flexible filler materials such as petroleum-derived waxes provide corrosion protection through their inherent hydro phobic Source: fib Bulletin 33. Figure 5. PT system vulnerability.

Literature Review and Background 17   nature. They may also be formulated with corrosion-inhibitor additives to further facilitate resistance to mechanisms of corrosion. Corrosion Protection of PT Anchorages PT anchorages are the most critical part of the PT tendon because they are the locations at which all prestressing force is transferred to the concrete member. Four levels of corrosion pro- tection are provided at post-tensioning anchorages to guard against corrosion at these locations. Tendon filler material, which may be either a cementitious grout or a flexible filler material, provides a first level of corrosion protection. A second level of protection is provided by the anchorage cap, occasionally called a grout cap, which remains in place when the structure is in service. A third level of protection depends on the construction type: for tendons terminated at an exterior surface, protection is provided by a pour back, or for tendons internal to the main concrete cross-section, such as those in a segmental box girder, protection is provided by the concrete section. Finally, a seal coat provides a fourth level of protection. Industry Standards and Certification Programs Notable improvements have been made over the years as the post-tensioning industry has matured. Several industry groups are engaged in educating and promoting the use of post- tensioning in some form, including the Post-Tensioning Institute, the American Segmental Bridge Institute, the Precast/Prestressed Concrete Institute, and the American Concrete Insti- tute. Although post-tensioning is not necessarily the primary, specific interest of some of these organizations, all of them are engaged in developing consensus-driven guidance that ultimately benefits the PT industry. The PTI M50 and M55 documents are frequently referenced by state departments of trans- portation (DOTs) in their specifications for post-tensioned structures (see survey, Chapter 2). PTI/ASBI M50.3 Specification for Multistrand and Grouted PT, a consensus specification with stakeholders from the owners (DOTs), bridge designers, PT system suppliers, component manufacturers, and academia, requires the following personnel qualifications: “7.1 — Supervision Post-tensioning operations: • The Direct Supervisor of Post-Tensioning Operations shall be certified as PTI Level 2 Multistrand and Grouted PT Field Specialist; • The Foreman of each installation and stressing crew shall be certified as PTI Level 2 Multistrand and Grouted PT Field Specialist; • The Foreman of each grouting crew shall be certified as PTI Level 2 Multistrand and Grouted PT Field Specialist and ASBI Certified Grouting Technician; and • At least 25% of each crew shall be certified in PTI Level 1 Multistrand and Grouted PT Installation.” The PTI M55.1 Specification for Grouting of Post-Tensioning Structures, also a consensus document with the same stakeholders, outlines requirements for grout materials and grout- ing procedures. Both the M50.3 and M55.1 specifications are to be used together for bonded post-tensioning. Important improvements to the end-product quality and durability of PT structures are, in part, attributable to the development and increasing use of certification programs. Certification programs are aimed at training, educating, and standardizing industry expectations to improve the quality of PT concrete structures by certifying field personnel, inspectors, tendon fabricating plants (more applicable to the building industry), and plants producing prestressing steel. Some states require the personnel to have attained such certifications to work on their PT projects,

18 Repair and Maintenance of Post-Tensioned Concrete Bridges while other states (such as Florida) go a step further and have their own Construction Training and Qualification Program in conjunction with recognized certification programs. Certification programs offered by PTI and ASBI serve to educate and train personnel engaged in PT construc- tion. Certification renewal for both PTI and ASBI is required periodically (every 4 years and 5 years, respectively). Field personnel certification programs relevant to bridge and infrastructure projects offered by PTI include separate specialist and inspector programs. PTI offers Level 1 and 2 Multistrand and Grouted PT Specialist programs—workshops aimed at educating personnel engaged in the installation of multistrand and bar PT systems. These programs cover aspects related to installation procedure, as well as safety concerns, grout materials, and grouting techniques of multistrand applications. Level 1 certification is attainable without any prior fieldwork experi- ence; Level 2 certification requires a minimum of 1,500 hours of relevant fieldwork experience (of which 250 hours are required in each of the following: installation, stressing, and grouting, or inspecting these operations). The PTI Level 1 and 2 Multistrand and Grouted PT Inspector program is a workshop designed to certify inspectors involved in the inspection of multistrand and bonded post-tensioning sys- tems in multistrand applications, such as bridge infrastructure. This program assumes familiar- ity with post-tensioned multistrand applications, reflected in the requirements for participation in the program. A prerequisite for participation in the Level 1 or the Level 2 inspector certifi- cation is either current certification as a Level 1 Multistrand and Grouted PT Installer, or as a Level 2 Multistrand and Grouted PT Specialist. Fieldwork experience is not required to achieve Level 1 inspector certification. To achieve certification as a Level 2 inspector, the participant must have 500 hours of relevant experience in inspection of PT systems (of which 100 hours are required in each of installation, stressing, and grouting inspection). The American Segmental Bridge Institute offers personnel certification through its ASBI Grouting Certification Program to provide supervisors and inspectors of grouting operations with appropriate training. The ASBI Grouting Certificate can be attained by participation in the training; certification as an ASBI Certified Grouting Technician requires both workshop participation and 3 years of experience in the construction of grouted post-tensioned structures. Since 2017, ASBI and PTI have jointly worked with the Florida Department of Transporta- tion to provide the Flexible Filler Certification Training program, offered annually, to educate personnel, contractors, designers, and inspectors on aspects specific to the use of flexible filler materials in post-tensioned bridge projects. The training includes instruction in the unique injection process; material characteristics of flexible filler; required equipment; and relevant FDOT policies, specifications, and design standards. FDOT currently requires construction and inspection personnel working with flexible filler to have this certification. Industry groups also have published key guidance documents. Two frequently referenced documents are the PTI/ASBI M50 Guide Specification for Grouted Post-Tensioning and PTI M55 Specification for Grouting of Post-Tensioned Structures. The PTI/ASBI M50 Guide Specification for Grouted Post-Tensioning is a guidance and speci- fication document for post-tensioning applications in buildings, bridges, storage vessels, and other structures, except stay-cable structures and rock anchors. It was first published in 2012 and is a joint effort of the Post-Tensioning Institute and the American Segmental Bridge Institute. The document provides general guidance for post-tensioning systems, including their handling, installation, stressing, grouting, and protection. It also sets minimum standards for post- tensioning work and components, defines testing requirements to qualify PT systems, and provides guidance on performance (or protection) level selection by the designer. It applies to buildings, bridges, storage structures, and other structures using grouted post-tensioning tendons, excluding

Literature Review and Background 19   stay cables and rock anchors that are already covered by other PTI documents. The protection levels outlined in the M50 guide are the same as those found in fib Bulletin 33 (PTI/ASBI 2012). The PTI M55 Specification for Grouting of Post-Tensioned Structures is a separate guid- ance and specification document that outlines current practices specific to cementitious grout filler materials and their use in post-tensioned structures. It was first published in 2003 and is in its third edition. It provides minimum requirements for selection and design, injection of grout, and installation of ducts in post-tensioned systems. The document is focused on ensuring “essentially complete filling of the duct”: materials, design, quality assurance and control, equip- ment, and construction. The M55 document limits its scope to PT systems with steel prestressing elements and cementitious grout filler materials and does not address vacuum grouting. It also provides some troubleshooting for grouting problems (PTI 2012). Inspection Inspections of in-service post-tensioned bridges, like other bridge structures, are subject to the National Bridge Inspection Standards (NBIS) requirements for inspection (FHWA 2020). NBIS requires inspection of all publicly owned highway bridges greater than 20 ft in length. In most cases, routine inspections are required every 24 months and consist of visual observa- tions. For bridges with known deficiencies, inspection frequency is stipulated to occur more often. In addition to NBIS, FHWA has mandated that element-level condition ratings (based on the AASHTO Manual for Bridge Element Inspection) be submitted for all bridges on the National Highway System. Also, some states have their own inspection requirements, often with require- ments for increasing frequency and complexity if issues arise. Inspection requirements specific to PT structures are not explicitly described by NBIS; at least one state (Minnesota) has pub- lished research efforts to improve inspection practices and remedial grouting contracts specific to PT structures (Chauvin 2017; Schokker and Berg 2012). Most agencies employ inspectors with general bridge expertise that may or may not include PT bridge experience. As some of the case examples illustrate, developing issues in PT structures may not be caught in regular inspection. Visual inspection is the primary means of assessing the condition of a bridge, including post- tensioned bridges. Such inspections are usually conducted on a periodic basis, with frequency dependent on the agency and documented history of the structure. Inspections conducted after a rainfall event (fib Bulletin 33) may provide particular insight into potential damage or areas of concern in a system, including the potential of water flow through the structure. Several agencies (Virginia DOT, South Carolina DOT) have documented water penetration occurring during or shortly after rainfall events. Inspection of post-tensioning systems requires a detailed, thoroughly considered approach because issues with or damage to post-tensioning systems are not always immediately evident. Damage to PT systems, at the same time, can have significant implications; corrosion of a post- tensioning tendon can result in a loss of structure integrity, a reduction in structure safety, and the need for costly repairs, such as tendon replacement. Inspections of PT systems are conducted to identify and gather information on the following (but are not limited to): cracking, discoloration, joint leakage, spalling and delamination, water flow or other evidence of moisture, honeycombing, rust staining, and duct damage. Inspection methods for post-tensioned bridge systems vary in complexity and impact to the structure; inspection methods are chosen by agencies on the basis of evidence of potential concern or issues. In general, visual methods are relied on for routine inspections. Inspection approaches requiring invasive or destructive methods are not a first action and are usually specified only if the owner has reason to suspect the PT system has been compromised.

20 Repair and Maintenance of Post-Tensioned Concrete Bridges Project-specific detailing, construction errors, and material selection can affect the ease of inspection. Internal post-tensioning tendons—and especially grout-filled post-tensioning tendons—pose a difficulty for inspectors. In cases in which these types of tendons require repair or replacement, these tendons are particularly difficult to address because of their bond with the main concrete member. In other cases, access for inspection has unintentionally been blocked during construction; for example, the Varina-Enon Bridge in Virginia lost access to its PT columns when concrete was poured over access holes (see Chapter 3). Guidance for inspecting and restoring tendons with defective grout has been published by FHWA (Theryo, Hartt, and Paczkowski 2013). This document guides state departments of transportation in their inspection of grouts used in post-tensioned structures, including the post-2000 prepackaged, thixotropic grouts that have been the subject of only limited study. Nondestructive Evaluation for Post-Tensioning Nondestructive evaluation (NDE) techniques can be used for the inspection of post- tensioning systems. Many different technologies exist for aiding PT system inspection, including many proprietary systems. Each NDE technology has its own unique efficacy and limitations; no NDE approach is universally capable of inspecting all aspects and all types of the variety of post-tensioning systems present in PT bridges. Various techniques have been developed and numerous efforts have been made for defect detection (for example, strand breakage) and for monitoring prestress levels in both bonded and unbonded post-tensioning tendons. Some of those techniques include the following: • Ground penetration radar • Radiography • Global dynamic approach • Magnetic flux leakage • Magnetic permeability • Magnetostrictive sensing • Electrochemical methods • Electromechanical impedance-based technique • Time-domain reflectometry • Impact-echo method • Pulse-echo technique (ultrasonic testing) • Guided wave ultrasonic technique • Acoustic emission technique Several factors influence the performance of NDE techniques, including elements of the post-tensioning system (such as the duct material and filler material), the geometry of the main structural element (geometry of the member), and even the type of defect or damage. In general, a combination of NDE techniques is required to adequately assess the condition of a post-tensioned system (Azizinamini and Gull 2012). These techniques can be broadly categorized into five classes of noninvasive NDE techniques: electrochemical methods, electromagnetic methods, mechanical wave or vibration methods, radiological methods, and magnetic methods. In addition, options for invasive, low-destructive evaluation (LDE) exist. Efforts to develop and implement these approaches have produced mixed results and levels of success. Many are not suitable for in-situ applications because of various limitations, such as susceptibility to environmental changes, high scattering of data, and insensitivity to relatively small defects. However, a few of them show promise and have several advantages over the other methods. A brief overview of the most common types follows.

Literature Review and Background 21   Electrochemical Methods Electrochemical methods are inexpensive methods used to monitor active corrosion of strands in concrete structures by measuring the electrical properties of the reinforcing steel. Most electrochemical techniques use the same measurement setup that consists of a reference electrode, a working electrode, a counter electrode, and a voltmeter. A closed electrical circuit is required, so direct electrical connection to the inspected steel strands must be established, as shown in Figure 6. To create this direct connection, sensors must be placed inside the duct, which poses a problem for monitoring existing PT ducts that do not have those sensors. Another major limitation of electrochemical methods is that these techniques do not measure corrosion activity in voided areas. This limitation is significant because several of the identified strand/wire breaks in post-tensioned structures have occurred at void locations, as seen in the Varina-Enon Bridge (see Chapter 3, Case 4). Four different types of electrochemical techniques have been developed, including half-cell potential, linear polarization resistance (LPR), electrochemical impedance spectroscopy (EIS), and electrochemical noise. The half-cell potential measurement is a technique widely used to evaluate active corrosion in reinforced steel and prestressed concrete structures. The half-cell potential is a quick measurement that is taken by using a voltmeter to find the potential dif- ference between a reference electrode and steel reinforcement. This method is less reliable for prestressing steels and large concrete covers, and for concrete with certain constituents that inhibit voltage potential. Alternatively, LPR determines the instantaneous corrosion rate in metal. In LPR, steel is slightly perturbed electrically from its equilibrium potential. Once the electrical potential is changed by a known amount, the current decay is tracked and related to the condition of the reinforced concrete structures. These measurements can then be compared with the measurements of the EIS tests. EIS uses a small amount of alternating current with a particular frequency applied to a metal electrolyte interface to calculate the impedance of the interface between steel and concrete. Impedance is calculated for various frequencies to find values of parameters that can be used to fit the measured data. Finally, the electrochemical noise technique monitors fluctuations in open circuit potential and current to infer how the system corrodes and it is mainly used to detect pitting corrosion. Of the four mentioned techniques, half-cell potential and LPR are the most commercially accessible. The former gives information on the probability of corrosion and the latter is related Figure 6. Fundamental measurement setup for electrochemical techniques.

22 Repair and Maintenance of Post-Tensioned Concrete Bridges to the corrosion rate. The half-cell potential method is practically and widely employed to iden- tify the presence of corrosion. Electromagnetic Methods The most commonly used NDE method (not considering visual inspection, hammer sound- ing, or manual excitation of the tendon) for post-tensioned structures is ground penetrating radar (GPR), which uses high-velocity electromagnetic waves to sense material changes in solid surfaces. It can be used to locate rebar in a deck or used on the interior of a box girder to find a tendon profile. Most often, this technique is used to locate internal tendon profiles, such as in the Wonderwood Connector Bridge (see case examples in Chapter 3). GPR can also be used to identify defects in tendons. Grout voids can be detected using GPR (Giannopoulos 2005), but the resolution of the data depends on additional factors, such as the presence of mild reinforce- ment and the depth of the duct. High frequency GPR antennae can gather images up to a depth of approximately 24 inches. GPR is often used in conjunction with other NDE methods to corroborate findings. Hurlebaus et al. 2017 list the two most commonly used complementary techniques as ultrasonic echo and impact echo, which are both mechanical wave methodologies (Hurlebaus et al. 2017). GPR has also been used in conjunction with radiological techniques. Another common electromagnetic NDE technology is infrared thermography (IRT). This technology uses the changes in materials’ thermal properties to develop a subsurface image. In general, IRT can detect grout deficiencies and voids within ducts, but it cannot assess strand condition. IRT is a relatively easy technique compared to others, requiring only an area to set up an infrared camera and application of a thermal gradient. Mechanical Wave or Vibration Methods Mechanical wave and vibration methods rely on a physical process of sending either an acoustic or seismic wave through a material and measuring the reflection of those signals off of voids in the materials. These techniques are useful in identifying grout defects but they are not useful in identifying strand condition. As such, they will likely not provide meaningful information when employed on unbonded tendon. In fact, in an investigation of an existing unbonded post-tensioned parking structure, Puri and Moser (2007) decided against mechanical wave methods in favor of magnetic and radiological methods to supplement GPR data. Ultrasonic echo sends a mechanically induced pulse through a material and measures the reflection/response of the material, which is helpful in identifying potential grout deficiencies. However, this technique does not identify problems with prestressing strands. It can be used on either metallic or nonmetallic ducts, and it is useful for depths of 2 to 12 inches. Impact echo (IE) is a technique in which a metal ball is dropped on the concrete surface. Seismic waves pass through the material as far as they can go before hitting an “edge” and trav- eling back to a sensor. The time taken for the wave to travel back to the sensor is measured. The technique has been extensively used on concrete decks to detect delamination, but it can also be used in post-tensioning to detect voids and evaluate grout in tendon ducts. IE can work for both metallic and nonmetallic ducts as long as there are not any shrinkage cracks or voids between the concrete and the duct. Improvements have been made to the technique, such as the stack imaging of spectral amplitudes based on impact echo (SIBIE) technique (Muldoon et al. 2007). SIBIE stacks multiple IE measurements to develop a more robust method of observing material surrounding an air void. In the early 2000s, an idea was proposed that used an ultrasonic method called C-Scan to detect various corrosion states in post-tensioning steel (Iyer, Schokker, and Sinha 2002). The

Literature Review and Background 23   tool presented accurately detected corrosion in the strand, but the testing apparatus that was used did not lend itself well to in-field examinations. Radiological Methods Using X-rays to assess the condition of internal tendons is not common because of safety concerns and equipment expense, but it has been demonstrated to effectively identify damaged strands. Puri and Moser used radiological methods to accurately depict severed strands in existing post-tensioned structures (Puri and Moser 2007). The radiological method backed up data from other electromagnetic and mechanical wave methods. However, the time, effort, and expense required to conduct testing limit its potential to be used on a large scale. Magnetic Methods Mechanical wave methods tend to provide useful data only about grout deficiencies; magnetic methods tend to only provide useful data about strand condition. These techniques provide detailed estimates of the amount of lost steel in cross-section from corrosion and breakage. Magnetic flux leakage (MFL) is one such technique presented by Ghorbanpoor et al. (2000) that induces a magnetic field near prestressing steel to detect changes in cross-section from corro- sion or breakage. Other reinforcing steel near the tendon can cause signal interference and affect results. This system works for tendons that are within 6 inches of the sensors, and as such works best with external tendons. In 2010, this technique was field investigated in service conditions to assess the Varina-Enon Bridge. MFL is consistently capable of detecting loss of metallic area greater than 5% for many different types of tendon damage (such as corrosion, breakage, etc.), and it can be used in single or multistrand tendons (Karthik et al. 2019; Hurlebaus et al. 2017). Puri and Moser (2007) used magnetic scanning to accurately detect severed strands in an existing parking structure with unbonded monostrand that were not detected by GPR. This method may prove useful in detecting issues in unbonded tendons in bridge structures (Puri and Moser 2007). Low-Destructive Evaluation Methods LDE methods require invasive sampling to expose the post-tensioning tendon and these methods are the only testing alternative that provides accurate data on the degree of degrada- tion of prestressing steel. LDE methods are common practice when severe deficiencies are evident on visual inspection of the outside surface of a structure (such as rust staining, leak- ing grease, etc.). One common method of low-destructive evaluation is drilling tendon high points to look for voids in the grout material, using a borescope to visually inspect and identify corrosion. To inspect internal tendons, the concrete cover is carefully removed to expose the post-tensioning tendon. Observations can be taken about the degree of corrosion and the presence of broken strands. The “screwdriver test” is a simple test in which a thin, flat implement is wedged between wires in a strand to detect if any are broken. If any of the wires move, it indicates a severed wire. Gupta (2003) presents the pull-off method for calculating the remaining prestress in monostrand tendons after being exposed through statics. Roughly 2 feet of monostrand is exposed, and a hydraulic jack is used to pull on the strand at a specified force. The deflection of the strand from its horizontal profile is then measured and used to calculate the remain- ing prestress in the element. Both the screwdriver test and the pull-up method have been successfully demonstrated in the investigation of an unbonded post-tensioned parking deck (Gupta 2003).

24 Repair and Maintenance of Post-Tensioned Concrete Bridges Nondestructive Evaluation for Flexible Filler The effectiveness of many NDE methods for PT inspection is influenced by the concrete cover, duct material, and rebar congestion, all parameters that are affected by the use of flexible filler. For example, the smooth HDPE pipe used with flexible filler is thicker than the typical corrugated duct used for grouting. In the particular case of the Wekiva Bridge, the bottom slab thickness and haunch were sized (thicker) to accommodate the internal bottom slab continuity tendons with flexible filler. Using smooth HDPE duct alone—without corrugations—may hinder the bond/create an air pocket between the encompassing concrete and the tendon, making some NDE methods inaccurate or inappropriate. Repair Problems with post-tensioning systems can necessitate various repairs. Repairs may be required of the PT tendon specifically (including the duct, filler material, or the strand); the anchorage and associated hardware; or the pour-back details that serve as part of the corrosion protection system, including pour backs at anchorages and at grout vents and ports. In addi- tion, issues with a structure’s PT system may require repair of the deviator blocks, diaphragms, or other points of contact between the PT tendon and the concrete section. Force loss in a PT structure may necessitate repair to the structure because of secondary effects, including struc- ture cracking and joint issues. Repairs, modifications, and issues with other parts of the structure, however, may have unintended effects detrimental to the PT system. Issues with deck integrity and water-tightness can lead to water penetration of the PT system. Modification of the deck in routine, planned deck-replacement operations has caused damage to the PT system because of inadequate concrete cover, the over-milling of the concrete surface, or both (see the case example of the Veterans’ Glass Skyway). Post-tensioning tendons rely on multiple layers of defense to protect the prestressing strand against moisture and corrosion mechanisms. A tendon repair may address deficiencies in the tendon duct material (sheathing), the filler material, or the prestressing strand. Sheathing/Duct Early U.S. PT bridge structures used metal duct; since the early 2000s, most PT tendons have been constructed with plastic duct. Issues encountered with a post-tensioning tendon’s duct are influenced by the chosen duct material. Typically, internal tendons are detailed with corrugated polypropylene; external tendons are now commonly made with HDPE. Each material has its own characteristic resistance to damage. Duct issues requiring repair include duct corrosion, duct misalignment, duct kinking at deviation points, duct splicing failures, duct bursting during filler injection, and duct damage of unknown origin. Most duct issues occur during construction. Issues encountered during construction, if caught prior to filler injection, are repaired to pass tendon pressure tests, which are often set as a project requirement before the injection may proceed. Duct bursting has been documented in multiple structures during filler injection. In some cases (such as the issues described in the case example of the Veterans’ Glass Skyway deck replacement), duct damage can occur during later structure modification. Duct damage also occurs during invasive tendon investigation and must be repaired to ensure adequate tendon protection.

Literature Review and Background 25   Currently, repair recommendations to address duct issues encountered in bridge construc- tion are not available from the Post-Tensioning Institute. However, its repair committee, DC-80, has published recommendations and guidance documents for the repair of sheathing of the more uniform scenarios encountered with the single-strand tendons used in building construction. Filler Issues with cementitious filler materials, including nonproprietary and proprietary grouts, have been well documented since the early 2000s. Grout deficiencies in external tendons have been documented in several structures (for example, the Sunshine Skyway in Florida and the Wando River Bridge in South Carolina). Problems with filler materials have included the pres- ence of excessive bleed water, the formation of soft grout (Wonderwood Connector, see Chap- ter 3), and the presence of contaminants (Veterans’ Glass Skyway, see Chapter 3). Causes of poor-quality grouts have been identified through many owner-instigated investigations and research efforts, and have included inaccurate bag weights affecting the mix proportions in proprietary grouts, improper mixing technique, mixing of different grout materials in the same tendon, insufficient clear passage in the duct potentially causing grout separation (Wonderwood Connector, see Chapter 3), and improper injection procedures (including issues with venting, pressure, injection speed, and other methodological causes). Deficiencies in grout in internal tendons pose significant problems when they occur. As internal tendons are contained entirely within a portion of the concrete cross-section, access is limited for the purposes of both inspection and repair. The difficulty of identifying problems in internal tendons has necessitated research efforts to identify effective methods. When problems with the grout in internal tendons have been identified, the removal and replacement of the tendon is not feasible. The current in-progress repair of the Wonderwood Connector in Florida (see case examples) is a prime example of the unique challenges and actions required to remediate internal tendons with deficient grout. The severity of the repair required when filler material is deficient has instigated a number of responses. Several research efforts have been undertaken to identify and address deficient grout (Alvarez and Hamilton 2002; Hamilton et al. 2014; Randell, Aguirre, and Hamilton 2015; Torres et al. 2018). Problems with grouts have compelled some agencies to consider other options for filler material and prestressing strand material to ensure soundness and robust corrosion pro- tection of the tendon. Further, issues identified in a few in-service PT bridges have compelled significant response by industry groups and grout manufacturers to improve grout quality and injection procedures. Strand Damage to the prestressing strand, when encountered, is often a side effect of the compro- mise of some other protection layer of the PT tendon. Types of strand damage include electro- chemical corrosion (rusting) of the strand, mechanically caused damage (nicking or severing of wires or the strand), and hydrogen embrittlement. Strand damage may be identified through rust staining of adjacent grout or nearby concrete surfaces. In extreme cases, relaxation of the tendon may be evident, indicative of force loss; in some cases, force loss may occur without obvious visual cues. Because some grouted tendons can permit force redevelopment (depending on their contact with the main concrete cross-section), strand damage can remain unidentified until it reaches a critical threshold.

26 Repair and Maintenance of Post-Tensioned Concrete Bridges Advances in Post-Tensioned Bridge Design in the United States Since the introduction of post-tensioned bridge structures to the United States, several advances have improved the durability of these systems. Key changes include the following: • Improved quality control of grout, including mixing and injection procedures • Improved grout materials, including thixotropic prebagged grouts • Improved detailing, particularly at connections • Introduction of grout certification programs • Introduction of flexible filler materials • Enhanced inspection methods (by some states) • Provisions and detailing for tendon replacement (by some states) The use of alternative filler materials, such as microcrystalline waxes, has constructability, design, and maintenance implications, but this use can provide corrosion protection to tendons while permitting later tendon removal. International Experience The introduction and evolution of post-tensioning in bridge structures have always been tied to international practices and technology transfers, including both the early contributions of Magnel and Freyssinet (both French engineers) and the early examples of post-tensioning issues and repairs. In 1985, the collapse of the Ynys-y-Gwas Bridge, attributed to elevated chloride levels at seg- ment joints, instigated a serious reconsideration in the United Kingdom of the durability of post-tensioned bridges, despite good service performance of these structures and few docu- mented cases of severe corrosion (Raiss 1993; Woodward 1989). By 1991, the Department of Transport’s preferences in design practice had shifted from the use of internal tendons to external tendons (Robbins 1991). In 1992, the U.K. Concrete Society and the Concrete Bridge Development Group established a working group to investigate and address concerns related to concrete bridge durability. Simultaneously, a temporary ban was placed on “new bridges of the grout-duct post-tensioned type” (Department of Transport 1992). The group worked to develop interim and revised guidance on PT design, detailing, specification, and construction methods, and on testing methods to evaluate PT structures (Raiss 1993). A 2009 inspection of the Sawasokogawa Bridge, a post-tensioned concrete composite girder bridge in Japan, revealed corrosion and concrete spalling of the deck near the intermediate piers. An extensive repair was performed to remove the deck slab “cables,” to replace the deck with a reinforced concrete deck, and to add strengthening reinforcement by additional external tendons. Causes of the initial corrosion were attributed to water and contaminant intrusion into the structure near the fulcrum, exacerbated by the heavy use of deicing salts. Salinity test- ing of the concrete found chlorides well exceeding the recommended limits (2.7 kg/m3 versus 1.2 kg/m3) (Nagatani and Tajiri 2019). The Petrulla Bridge was a viaduct built in Italy in the early 1980s that was devoted to vehicular traffic, mostly vehicles transporting raw agricultural materials, including fertilizer. The structure collapsed in 2014. The failure was attributed to several factors: corrosion instigated by delay between stressing and grouting, exposure to fertilizer, a lack of gap between the ducts, and inappropriate grout for tendons. Investigations found that grouting occurred after a sufficient time lapse that allowed water to accumulate in some locations, leading to aggressive corrosion. Testing of the grout revealed a high chloride content in the tendons: 0.24% by weight of cement versus the limit (per AASHTO-LRFD, PTI M55) of 0.08% (Anania, Badalà, and D’Agata 2018; Virmani and Ghasemi 2012).