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.
CHAPTER 2 FINDINGS 2.! Corrosion Performance of Precast Segmental Bridges In order to provide a comparative evaluation of the corrosion performance of precast segmental bridges, first the experience in the United Kingdom and to a lesser extent Europe are reviewed, followed by a review of the experience in North Amenca. 2.. United Kingdom and European Experience 2.~.~.! Literature Review. There is little information specifically about the corrosion performance of precast segmental bridges. However, related data on post-tensioned bridges provides some insight into the corrosion issues with grouted post-tensioned segmental bridges. The first serious problem with corrosion of bonded post-tensioned bridges in the U.K. was the collapse of the Bickton Meadows Footbridge in Hampshire in 1967.3 The only other collapse due to corrosion in the U.K. was the Ynys-y-Gwas Bndge in West Glamorgan, 1985.3 4 Both of these structures were of segmental construction with thin mortar joints. The collapse of the Ynys-y-Gwas Bridge was due to the corrosion of longitudinal tendons at the segment joints. The mortar at the joints was highly permeable and allowed moisture, chlorides, and oxygen ready access to the tendons. The structure was 32 years old with no evidence of distress prior to failures. The Bickton Meadows Footbridge collapsed as a result of severe corrosion ofthe top tendons.5 However, both the precast units and the thin mortar joints were of extremely poor quality. It is reported that the precast units were cracked and honeycombed when delivered to the site to the extent that grout appeared at the surface of the units during the grouting operation. In addition, it is reported that the bridge was overstressed. The structure was ~ 5 years old at the time of collapse. A bridge in Belgium over the River Schelde collapsed in 1992. It had been reported3 that corrosion of the post-tensioning through a hinged joint which was part of the end frame of the 6
structure led to the collapse. However, although there was corrosion ofthe post-tensioning, it was also reported5 that a petrol tanker collided with the bridge and caught fire prior to the collapse. Obviously, vehicular collision and/or fire damage could have contributed to the collapse. Numerous other non-segmental, post-tensioned structures have experienced different degrees of corrosion to the post-tensioning.356 Many ofthe problems have occurred near anchorages although problems have also been documented where poor grouting exposed the strands to air, moisture, and chlorides. Additionally, there have been a number of cases in Europe involving problems with hydrogen embrittlement due to the use of quenched and tempered and "exotic" steels and contamination with sulfides, cyanide, and other poisons.5~7 Also, chloride contamination of the grout has led to problems; although the ingress of chlorides from deicing salts has occurred, often the contamination has been from the use of calcium chloride as an admixture, seawater, or chloride contaminated aggregates. Overall, the vast majority of segmental and post-tensioned structures in the U.K. and Europe are performing well.5-8 The problems that have occurred have been due mainly to deficient construction practices and poor design choices rather than intrinsic deficiencies with segmental and post-tensioned structural systems. 2.~.2 North American Experience 2.~.2.! Literature Review. A durability survey of segmental concrete bridges was conducted under the sponsorship of the American Segmental Bridge Instituted The survey identified 109 precast segmental bridges, most built in the U.S. Overall survey results indicated good durability performance with the structures with no significant corrosion problems with the post- tensioning. However, the survey results were based on visual inspection reports only. Unlike in the U.K., there have been no reported cases of corrosion of post-tensioning in segmental bridges. A number of summaries of case studies have been reported on the durability of non-segmental post-tensioned concrete structures.57~3 In Canada, two grouted post-tensioned structures experienced corrosion problems where chlorides had penetrated into the structure due to deficient expansion joints, cracking and low cover to the post-tensioning ducts. Most of the corrosion observed was corrosion 7
of the metal ducts although significant tendon corrosion was noted in a few cases. In the U. S., few problems have been reported with corrosion of grouted post-tensioned tendons in the bridges. Improper grouting and detailing was blamed for corrosion problems with the post-tensioned Walnut Lane Bridge. The Sixth South Street Viaduct experienced corrosion distress ofthe post-tensioning, but the tendons were in a galvanized steel duct without the presence of grout; thus the tendons were unhanded and unprotected.~3 Overall, there have been no reported problems with corrosion of precast segmental structures in North America. Again, similar to the experience in the U.K., the few reported problems with grouted post-tensioned tendons was due to inadequate construction practices rather than intrinsic deficiencies with segmental and post-tensioned structural systems. 2.~.2.2 Survey. Ninety questionnaires were sent to members of the AASHTO Subcommittee on Bridges and Structures and select consulting engineers. A sample questionnaire and more comprehensive details of the responses are contained in the Appendix. Forty-four percent (44%) of those contacted responded; the level and detail of response varied. A total of 90 bridges were documented in the responses in addition to typical design and construction practices and known durability problems. Portions of the survey focused on the types of segmental joints utilized. Results are summarized in Table 2-1 and Figures 2-1 and 2-2. Eighty-six percent (86%) of the bridges contained internal tendons while fourteen percent (14%) had external tendons only. Of the bridges with internal tendons, ninety-two percent (92%) used match-cast epoxy joints with eight percent (8%) using cast-in- place joints. Mortar joints or match-cast dry joints were not used. Of the bridges with external tendons only, sixty-nine percent (69%) of the joints were match-cast epoxy and thirty-one percent (3 1%) were match-cast dry. Cast-in-place and mortar joints were not used. Overall, match-cast epoxy was by far the most popular joint type with an eighty-nine percent (89%) utilization with match-cast dry and cast-in-place joints being used four (4°/O) and seven percent (7°/O), respectively. 8
Segment Joint Conditions Internal Tendons Cast-in-Place 8°/0 Match-Cast Epoxy 92% Figure 2-] Joint Conditions for Segmental Bridges - Internal Tendons 9
Segment Joint Conditions External Tendons 1~1 Match-Cast Dry 31°/0 Match-Cast Epoxy 69% Figure 2-2 Joint Conditions for Segmental Bridges - External Tendons 10
There were very few durability problems reported. The problems mentioned were more concerned with serviceability issues. Of the six (6) structures with reported problems, none were related to corrosion. While most reported problems related to cracking, one response indicated a problem with an overlay wearing surface and another dealt with the addition of post-tensioning due to a design error. This data should be used cautiously, however, because extensive comprehensive evaluation techniques for tendon corrosion have not been developed. In the words of one transportation official, "[there are] no known problems [with tendon corrosion] but no significant testing or evaluation of [the] condition of internal tendons has been made." The current state of design practice with segmental bridges was also queried in the survey. All those who responded to the question indicated that the AASHTO Guide Specifications for Design and Construction of Segmental Concrete Bridges was used. Table 2-1 Segment Joint Types (from survey) ~ Number ~ ~r ~ Tendon Types of Match Cast, Match Cast, Casts Place Mortar Bridges Dry Epoxy Internal 77 O External Only 13 4 Total Bridges 90 4 2.2 Current Design and Construction Practice 92% 1 61 8% 1 0 1 0% 69% O0% O ~ 0% l 89% 6 7°/o O 0% 2.2.1 United Kingdom. In September of 1992, the U.K. Department of Transportation (U.K. DoT) announced that it would not commission any new bridges of the "grout-duct post- tensioned type", pending a review of standards in place at the time in the U.K.3 In effect, a moratorium was put in place on new post-tensioned bridge construction until better design/construction practices and specifications were developed and implemented. Prior to the moratorium, in June 1992, a working party was set up by the U.K. Concrete Society to study the problem of durability of bonded post-tensioned bridges and prepare recommendations; a final technical report was issued in 1996.3 It is believed that this report contains 11
the current state of design and construction practice for all bonded post-tensioned concrete bridges in the U.K.. Subsequent to the issuing of this technical report in 1 996, the moratorium on grouted post-tensioned bridges was lifted by the U.K. DoT except that precast segmental construction using internal tendons is still not allowed. 2.2.~.! Segmental Construction. The Working Party believes that "sufficiently wide insitu concrete [segment] joints, and match-cast [segment] joints properly sealed with epoxy resin, are satisfactory in durability terms".3 Thin mortar joints are specifically not allowed. However, although the Working Party states that match-cast epoxy joints are acceptable from a durability standpoint, they add that "special consideration has also to be given to the continuity of the ducts across the joints".3 Until the Working Party is satisfied with a detail to guarantee duct continuity, or equivalent corrosion protection can be assured by other means, the ban on precast segmental construction with internal tendons will remain in effect in the U.K. i4 2.2.~.2 Duct Type. Unlined ducts and ducts that are lined with cardboard or any other biodegradable material are not allowed. Grouting specifications require the use of a corrosion resistant duct, such as high density polyethylene or polypropylene, that can pass an air- pressure test prior to grouting to demonstrate that the complete duct system (vents, anchorages, anchorage caps, couplers, connections) forms a "complete encapsulation for the tendons which is resistant to the passage of air and water".3 Additionally, vents must be provided in the anchorages and at low and high points in the ducts. The purpose of vents is to allow bleed water and/or air to escape · ~ during grouting. 2.2.~.3 Grouting. The U.K. specifications were revised in order to assure that the ducts will be adequately filled w. ith a quality grout. The current specifications require that the grouting contractor conduct full-scale trials of the grouting to insure that the grouting scheme will result in adequate grouting. This is assured by cutting or coring the trial duct in order to expose transverse and longitudinal cross-sections. Other requirements are listed in Table 2-2. It is anticipated that superplasticizers and expansion agents will need to be used to meet the workability requirements and the requirements in Table 2-2. 12
Table 2-2 Select Current U.K. Grout Requirements Property T Common Grout ~SpeciaIGrout Maximum w/c ratio 1 0.40 0.35 max l Volume change Bleeding Strength at 7 days -1%tO+5°/° 0 to +5% less than 1% none 3900 psi 3900 psi _ 2.2.2 United States. Currently, all bridge design and construction is governed by the AASHTO Standard Specifications for Highway Bridges.~5 Precast segmental bridges also fall under the jurisdiction of the AASHTO Guide Specifications for Design and Construction of Segmental Concrete Bridges. Although this document is only a guide specification, results from the survey indicate that, as a rule, designers and owners are adopting and using the document as part of the design and project specifications. 2.2.2.! Segmental Construction. The segmental bridge guide specification clearly states that "Type A (match-cast epoxy) joints shall be utilized for all bridges utilizing internal tendons and for all bridges exposed to severe climatic conditions..." Type B (match-cast dry) joints can be used with external post-tensioning when the bridge is not subject to severe exposure such as freeze-thaw exposure or chloride exposure from deicing chemicals. Cast-in-place closure pours are allowed. Although not specifically disallowed by the specification, the intent of the document seems to preclude the use of thin mortar joints since joints other than Type A and B joints (match-cast) are required to have adequate width to permit the coupling of tendon ducts. This would in essence preclude the use of a mortar joint. 2.2.2.2 Duct Type. Minimal guidance is provided on the ducts for internal or external tendons. The duct may be manufactured from galvanized steel or high density polyethylene. Polyethylene is normally used for external ducts. Minimum sizes and wall thicknesses are based on the size of the tendon. 13
2.2.2.3 Grouting. The segmental guide specification references the Post- Tensioning Institute's (PTI) "Recommended Practice for Grouting of Post-tensioned Prestressed Concretei7. As with the specification items relating to the ducts, there is minimal guidance on the grout to be used to protect the tendons against corrosion. Table 2-3 documents select required properties of the grout to be used. Currently, the PTI Committee on Grouting Specifications is finalizing a comprehensive document on grouting of post-tensioned structures. Examination of draft copies of this specification indicate that it will provide more definitive guidance than is currently available in the PTI Post- Tension~ng Manual that is currently referenced by the AASHTO Guide Specifications for Design and Construction of Segmental Concrete Bridges. Although there is not a comprehensive specification available, there are a number of papers that document research on the testing and development of high performance grouts for the corrosion protection of post-tension~ng.~~27 The studies attempt to optimize both grout material properties and placement techniques. In general, the studies recommend grouts with lower water/cement ratios, improved bleed characteristics, and placement schemes which provide complete encapsulation ofthe prestressing steed and a reduction in the potential for voids within the grouted ducts. Table 2-3 Select Current U.S. Grout Requirements | Property T Common Grout l Maximum w/c ratio | 0.45 l Volume change Bleeding* Strength at 28 days Not specified Expansive admixtures may be used 2% at 3 hours; 4% maximum Not specified; approximately 4000 psi should be reached by use of specification * suggested approximate limits only. 14
2.3 Current Inspection and Repair Practice 2.3.1 Inspection. One of the issues that apparently causes the most consternation among bridge engineers and owners is the difficulty with inspection of the main load carrying elements of a precast segmental bndge, that is the longitudinal internal and external tendons. 2.3.~.! Grouting. The main corrosion protection for grouted tendons is the grout. If the tendon ducts are not completely filled with grout or if the grout is absent, the tendon is more susceptible to corrosion. Thus, it would be useful to evaluate a tendon duct to determine if there are voids within the grout or if the grout is absent. The best non-destructive test (NDT) method for internal bonded tendon ducts seems to be impact-echo which utilizes transient stress waves to locate voids in grout. Both laboratory research and field testing has been successful in locating voids in bonded metal tendon ducts. 28-3 t Plastic ducts and external unhanded tendon ducts cannot currently be evaluated with impact-echo. One other disadvantage of the method is that a skilled operator is required to interpret the test signals. The first large-scale application of this NOT method to evaluate grouted tendon ducts was completed in 1997 on a 14-span, precast segmental bridged The superstructure consisted of precast grouted post-tensioned cantilever beams connected by drop-in precast, pretensioned AASHTO sections (see Figures 2-3 and 2-4). Thus, although this bridge was precast segmental and contained internal tendons, it did not have internal tendons crossing segment joints. Although solid grout was found in some instances (see Figure 2-5), partially grouted tendons (see Figure 2-6) and ungrouted tendons (see Figure 2-7) were found in40% of the beams (156 beams total). Most of the voids found were classified as partial voids because there was some grout present, although in many cases, the ducts were nearly empty. In many beams, more than one duct contained voids. In a few cases, the ducts were completely empty, as if the contractor had failed to grout the duct. Impact-echo results were verified at selected locations by drilling a small hole and using a horoscope to observe the interior of the duct and to examine if any significant corrosion of the prestressing steel had occurred in the non-grouted region. Despite the high incidence of the voided tendons, no significant corrosion of the prestressing steel had occurred in some 30 years (see Figure 2-7). The tendons were re-grouted to provide corrosion resistance and ultimate strength as a bonded tendon. 15
Figure 2-3 Post-Tensioned Cantilever Beams Located at the Pier Sections of the Bridge ~. 3.77~ _O "I )r()l,-ill" (~;irlJer ( -it . (a) l 0 ~ .2 11) 1.18 In , ' 1 . Section A-A (in) Figure 2-4 Post-Tensioned Cantilever Beam: (a) Elevation of One-Half of a Cantilever Beam, Showing the Parabolic Profile of the Tendon Duct Tested; (b) Cross-Section Showing Tendon Locations in the Region of Testing 16
Figure 2-5 Exposed, Pully-~outed Duct Flare 2-6 Palely Routed Tendon Duct Found by ImpacLEcho 17
Figure 2-7 Ungrouted Tendon Duct Found by Impact-Echo (Note Animal Corrosion After 30 Years of Exposure) Impulse radar has been used to locate grouting voids in tendon ducts. It has had limited success, and some have concluded that its best application may be the location of ducts/tendons for further testing or invasive inspection.30 One disadvantage is that impulse radar signal interpretation must be made by a skilled operator, and signal interpretation becomes difficult in areas with closely spaced reinforcement. Closely spaced reinforcement is often found in anchorage zones in segmental bridge structures. Currently, a test is being developed in England to determine the presence of air voids within a freshly grouted duct.3 2i If successful, testing the grouting operation immediately after the duct is filled will allow the opportunity to add additional grout, if necessary, while the existing grout is still plastic. 2.3.~.2 Corrosion-Induced Tendon Damage. It would be beneficial in an existing structure to be able to assess the level of corrosion damage (wire section loss, fracture), if any, to the tendons. Currently, radiography has been the most successful.30 32 33 Although the radiographs can be interpreted by an engineer with average familiarity with segmental bridges, highly skilled operators are required to conduct the testing. An additional drawback is that a large area in the 18
vicinity of the testing must be evacuated during the testing because of the high radiation output required. A1SO7 access to both sides ofthe tendon is required. A new radiography system, called the Scorpion, has been developed by French Engineers and used successfully.3O 32 33 The system is vehicle mounted and contains a telescopic arm which positions the source and detector on either side of the beam being tested. In addition to its high cost, the same safety procedures are required as for conventional radiography. Italian and Swiss Engineers have developed an electrical reflectometer test known as Reflectometric Impulse Measurement (RIMT). A short duration electrical pulse is applied at the anchorage, and the return signals are interpreted. Trials of the method in England were not successful, and the method is not recommended at this time.30 Continuous acoustic monitoring of post-tensioned structures has been used successfully in Canada to monitor fractures in unbended tendons in post-tensioned structures since 1994.34 In addition, the authors have used the system on a recent unbended post-tensioned building project in the United States. Acoustic monitoring relies on the fact that a wire or strand failure results in the release of strain energy which in turn sets up transient stress (acoustic) waves in the structure. Accelerometers are mounted throughout the structure to detect the acoustic energy from wire or strand breaks due to on-going corrosion. In order for acoustic monitoring to be effective, it must be possible to differentiate signals generated by wire fractures from other ambient noise in the structure. Previous work in Canada has shown the method to be successful with unhanded systems. Trials were carried out in the U.K. on bonded systems and results were encouraging. More work on evaluating fully grouted tendons needs to be conducted.35 Although acoustic monitoring shows great potential as a management tool for continuing corrosion damage, it cannot provide information on existing damage to the post-tension~ng of a structure. In tendon systems utilizing metal ducts, half-cell corrosion p.otentials36 and corrosion rates using the 3LP linear polarization method37 may be used to determine if corrosion is occurring and its relative rate. These techniques have limitations, though, since the prestressing steel is typically electrically interconnected to the metal ducts, at the ends of the tendons and to the anchorages. Thus, 19
the measurements cannot differentiate between corrosion of the prestressing steel, the duct, or the anchorage. 2.3.2 Repair Techniques. The survey results clearly indicated that no repairs had been effected on precast segmental bridges due to corrosion problems. Precast segmental bridges that had been repaired were reportedly for cracking problems and design deficiencies. The only known repair to a precast segmental bridge with internal tendons that had a potential for tendon corrosion was by retrofit grouting of voided tendons.29 Although numerous partial and full voids were discovered in tendons, there were no signs of significant corrosion after 30 years. The retrofit grouting was done by cutting windows every 5 If. into the beam webs to gain access to the tendon voids. The tendons were regrouted for long-term corrosion protection and because the ultimate strength of the beams required the tendons to be bonded. 2.4 Past and Current Research 2.4.! University of Texas Study. Starting in 1992, an on-going research project sponsored by the Texas Department of Transportation at the University of Texas at Austin (UT Austin) is studying the corrosion protection of internal tendons across segmental bridge joints. The work was begun by Vignos38 and is being continued by West39 under the direction of Dr. John E. Breen. The lack of a continuous duct across segment joints was identified as a potential "weak link" in the corrosion protection of internal tendons in precast segmental bridges. 2.4.. Test Program. The first step in the research was the development of an accelerated test method to evaluate the performance of different segment joint types and corrosion protection schemes at the joint. A modified form of the standard corrosion macro ceil developed in ASTM GI09 was adopted. Figure 2-8 shows a typical test schematic.39 Both match-cast dear (no epoxy) and match-cast epoxy joints (with and without gaskets) were investigated. Additional variables included: duct type, level of Recompression across joints, and type of grout. The duct materials studied were galvanized steel and polyvinyl chloride (PVC). The level 20
ma% NaD solution plexiglass dam duct ~ grout ~ \ Concrete Segment ~ hi, . \ . , \ r--_-~7~ -my. I Try, ~-,,v; ~ R:lOOOhms 5 in. / 7: ...... - - .~. ~,d, ;.,-.,.,0;.;] i. . . i..] .... , . ~ t I ~ gasket (when applicable) ~ . ~ -. . , 0.5. dia. 7-wire I r strand ·. . a , ~ · - ~ ~ - match cast seomentol joint 2-~4 bars ~ Longitudinal Section 3 in. 6 in. _. _ ~r~ . ~, .75 in. ~ . ., ~ ~ ~ D, .5 in. . it. .: ~ By. 1 in. ... , a L 4.5in. -l End View Figure 2-S Typical Test Schematic for Segmental loins 21 r 12 in. \ \ -I Epoxy paint - typ. 40.25~ end cover R
of precompression was varied Proms psito 36cpsi (~190 psi for this test). Three "rout types were studied: a standard Portland cement (PC) grout, a PC grout with silica fume, and a PC grout with a calcium nitrite based corrosion inhibitor. Exposure conditions for the specimens consisted of four week wet/dry cycles with a three percent sodium chloride (NaCl) pending solution 2.4.1.2 Preliminary Test Results. Preliminary test results from the study are presented by West39. After 1,500 days of accelerated exposure, the data indicates that 12 of 38 specimens have experienced an initiation of corrosion. Nearly all the specimens (1 1 of 12) with signs of corrosion have dry segment joints. However, corrosion rate calculations indicate that the overall magnitude of corrosion of all the specimens in the study is very low to negligible.39 Additionally, some ofthe data indicates that some specimens are experiencing "reverse macrocell" corrosion; that is, the embedded mild steel is corroding preferentially to the strand. Specimen autopsies began in February 1998 and are on-going. Preliminary results have confirmed that even with the specimens showing the highest corrosion rates, the corrosion damage to the strand is very low to negligible.40 Photographs from a dry joint specimen autopsy in Figures 2-9 and 2-10 are illustrative. The photograph in Figure 2-9 shows the condition of the galvanized steel duct; note that the duct is severely corroded in the vicinity of the joint. This emphasizes that, even if it were continuous across the joint, a metal duct will not likely provide an effective long-term barrier to the ingress of chlorides to the strand. Despite the perforation ofthe steel duct, the condition of the strand was good upon removal of the grout as shown in Figure 2-10. The white line in the photographs indicates the location of the segment joint. No corrosion of the strand was noted in the vicinity of the joint. A few areas exhibiting very mild surface corrosion were noted outside the test area both on uncoated portions of the strand and in locations originally covered with epoxy paint. Areas away from the joint were coated with epoxy paint so as to limit the exposed length of the anode and cathode to 5 inches as per ASTM G-109. It seems that the alkaline environment pf the normal Portland cement grout may have provided better protection to the strand than did the epoxy paint. 22
Figure 2-9 Metal Duct Showing Severe Corrosion (White Line Indicates Location of Segment Joint) . Figure 2-10 Strand in the Vicinity of the Segment Joint Showing No Corrosion (White Line Indicates Location of Segment Joint) 23
Overall, the preliminary autopsy results to date have shown that, while the level of corrosion of the galvanized ducts has been severe in some cases, the corrosion of the strand and/or mild reinforcing has been very low to negligible for all specimens in this accelerated corrosion test.40 In addition, no corrosion of the prestressing strand was found for specimens with epoxy joints and galvanized steel ducts. 2.4.2 Non-Segmental Related Corrosion Studies. A number of studies have been conducted on the durability of post-tensioned concrete structures and elements. The U.S. Army Corps of Engineers conducted long-term exposure tests on post-tensioned beams exposed in a tidal flat at Treat Island, Maine since 1956. Overall, the test specimens performed well, with no indication that the integrity of the post-tensioning has been compromised after decades of exposure to chlorides and cyclic ~eezing.l2 An NCHRP report was issued in February, 1989 based on a one year study on corrosion protection systems for prestressed concrete structures.4i Severe accelerated corrosion tests were conducted on various corrosion protection systems. The author found that epoxy coated ducts, hardware, and strands performed the best. However, the "post-tensioning industry standard consisting of a bare anchorage, galvanized steel duct with duct taped joints, bare prestressing strand and normal cementitious grout adequately protected the encased bare strand from corrosion when embedded under only about 1 inch of concrete cover.~341 A comprehensive study was conducted at the University of Texas at Austin on improving the durability of bridge decks using transverse prestressing.42 In general, grouted galvanized duct was found to adequately protect the tendons between the anchorages. Thin concrete covers at anchorages were found to provide inadequate corrosion protection. The authors recommended full encapsulation of all components of the post-tensioning system. Additionally, cracking in the concrete was found to greatly promote the corrosion of the mild reinforcement in the slabs; prestressing had a significant effect in reducing the penetration of chloride ions at crack locations when the cracks were limited in width. 24
