National Academies Press: OpenBook

Inspection and Maintenance of Bridge Stay Cable Systems (2005)

Chapter: Chapter Two - Stay Cable Systems and Materials

« Previous: Chapter One - Introduction
Page 8
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 8
Page 9
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 9
Page 10
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 10
Page 11
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 11
Page 12
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 12
Page 13
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 13
Page 14
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 14
Page 15
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 15
Page 16
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 16
Page 17
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 17
Page 18
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 18
Page 19
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 19
Page 20
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 20
Page 21
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 21
Page 22
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 22
Page 23
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 23
Page 24
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 24
Page 25
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 25
Page 26
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 26
Page 27
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 27
Page 28
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 28
Page 29
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 29
Page 30
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 30
Page 31
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 31
Page 32
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 32
Page 33
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 33
Page 34
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 34
Page 35
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 35
Page 36
Suggested Citation:"Chapter Two - Stay Cable Systems and Materials." National Academies of Sciences, Engineering, and Medicine. 2005. Inspection and Maintenance of Bridge Stay Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/13689.
×
Page 36

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.

STAY CABLE SYSTEMS An overview of various stay cable systems is presented in this section to familiarize readers with the terminology and technical aspects of various cable designs, and the issues re- lated to inspection and maintenance of stay cables in cable- stayed bridges. In very general terms, a stay cable can be described as a tension element composed of a single or multiple longitudinal MTEs, which is connected at one end to the bridge pylon and anchored at the other end at the bridge deck. Over the years, there have been two fundamentally dif- ferent and competing philosophies regarding design of stay cables. In the first approach, which dominated early German, British, and Japanese designs, the stay cables were designed based on the well-developed suspension bridge technologies; that is, those of the main suspension cables and the hanger cables, and wire rope technology from industrial applications. In the second approach, which more or less began with the Brotonne Bridge in France and dominated the U.S. stay cable designs until late 1990s, the cables were designed based on the post-tensioning tendon technologies. There were also variations in each of the two main philosophies. The concepts underlying these philosophies and their significant evolution over the last 30 to 40 years will be discussed later in this chap- ter. The motivations for these system evolutions were based on the field performance of the systems, technology develop- ments and, above all, economic factors. Main Tension-Resisting Elements There are several different arrangements of the MTE compo- nents in the free length of the cable. The free length refers to areas of the cable that are not in the vicinity of the anchorages. The MTE could be a single bar, multiple parallel bars, multi- wire helical strands (wire ropes or bridge strands), a bundle of parallel wires, or a bundle of parallel seven-wire strands. Fig- ure 5 shows some of the MTE systems. The locked coil cable was very common in early European and Japanese cable-stayed bridges. There is a central core of parallel round wires surrounded by spirally wrapped layers of interlocking z-shaped (and in some cases trapezoidal) wires. This arrangement makes a denser more compact cable (with reduced voids), with a smooth outer surface and less sensitiv- ity to side pressures (Walton 1996; Ito 1999). Helical wire 8 ropes have been popular in the United Kingdom. The spiral wires reduce the modulus of elasticity and strength of cable compared with equivalent parallel wire cables, but are much easier to handle (Ito 1999). The locked coil cable and spiral strands are examples of applications of suspension cable tech- nology to stay cables. The single or multiple bar system typically consists of one or more thread bars with a diameter of 26 to 36 mm (1–1.375 in.). The Dame Point Bridge in Florida and four pedestrian bridges in Calgary include bar cables. Worldwide it is believed that three other cable-stayed bridges with bars have been built; one each in Malaysia, Germany, and Chile. The parallel wire cables are typically made of 5 to 7 mm (0.19–0.27 in.) wires. Unlike the main suspension cables, the parallel wire stay cables do not include closely wrapped external spiral wires to maintain the shape of cable. The Pasco–Kennewick Bridge in Washington State and the Hale Boggs/Luling Bridge in Louisiana are examples of parallel wire cables in the United States. The parallel seven-wire strand system is the most com- mon MTE used in the United States. The survey results indi- cated that 75% of U.S. bridges included parallel seven-wire strands (see Figure 6). In contrast, only one bridge in the Canadian survey had seven-wire strands. The only Canadian bridge with parallel strands is also the newest one (opened to traffic in 2003), pointing to a possible move toward parallel strands. The majority of the Canadian bridges surveyed (54%) have steel wires. There are however four bridges with parallel bars in Canada. The guaranteed ultimate tensile strength of seven-wire strands is 1860 MPa (270 ksi). The wire and strand stays are continuous from anchorage to anchorage because they are produced in long lengths and trans- ported on reels, but bar systems require splicing with couplers, because the maximum length of individual bars is on the order of 18 m (60 ft). Figure 7 shows a bar cable with couplers. The factors that typically drive the decision on the choice of MTEs have generally included the geographic preferences of the designers, suppliers, and owners (based on adopted design philosophies and available materials), perceived notions of long-term durability (i.e., potential for corrosion and fatigue), and cost. More recently however, issues related to inspectabil- ity, feasibility for nondestructive evaluation (NDE), and possi- bilities for MTE replacements and additions have also entered CHAPTER TWO STAY CABLE SYSTEMS AND MATERIALS

9(a) (b) (d) (c) FIGURE 5 Various MTE cross sections: (a) locked coil, (b) helical strand, (c) parallel wire, (d) parallel seven-wire strands (Gimsing 1998).

10 the decision process. The arguments that are generally made for, or against, one MTE system or another involve some of the factors listed here: • Cost, • Implications of corrosion exposure including surface- to-volume ratio, • Fatigue performance including implications of crack propagation, • Redundancy, • Interwire fretting, • Notching at anchorages, • Stiffness, • Tightness of MTE bundle (void areas), • Implications of vibrations, • Ability to adjust MTE force, and • Ability to remove and augment MTE. Anchorage Systems There is a great variety of anchorage systems used for stay cables, depending on the choice of the MTE and the cable manufacturer. In bar systems, threaded nuts (matching the large threads on the bars) and anchor plates are used for anchorage. In this section, a brief discussion of generic cate- gories of cable anchorages for parallel wires and strand cables is presented. For the sake of brevity, only anchorage systems common to the United States are discussed. However, such systems are commonly used worldwide. In the cable free length, the parallel wires or strands are bundled together, thus making contact with each other. As the cable approaches an anchorage, the wires or strands must separate from each other to achieve proper anchorage. The distance from the point that the strands (or wires) splay out to the anchorage point is gen- erally referred to as the anchorage length. There are three fundamental approaches to cable anchorage design. The first is to individually anchor each splayed wire or strand at a single point on an anchorage plate. That anchorage point would exclusively carry all dead and live loads imposed. The second is to transfer all loads through a conical steel socket. The force in the MTE transfers by bond through a filler mater- ial inside the conical socket. The third is a combination of the first two approaches; that is, transfer dead loads through the anchorage point and carry live loads through the socket action. Point Anchorages Figure 8 shows the point anchorage concept. Typically, a two- or three-piece conical wedge with a toothed center hole grabs on the outside of the seven-wire strand and anchors it. This is essentially a modified version of the wedge system used in post-tensioning applications. Examples of this type of anchor- age include the Charles River Bridge in Boston and the C&D Canal Bridge in St. Georges, Delaware. When individual wires are used, they are generally terminated at a “button head” that is formed at the ends. In the multistrand system with point anchorages, the cable can be assembled in the field by stressing all strands at the same time, or it can be stressed one strand at a time (using a system to ensure equal distribution of stress). The grip- ping wedges create notches on the strands, which could become fatigue initiation points. However, stay cable systems 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 7-wire steel strand steel wire steel bar or threadbar other no answer MTE Type P er ce nt o f B rid ge s U.S. Canada FIGURE 6 Types and frequency of main tension elements used. FIGURE 7 Typical bar couplers in stay cables.

11 go through fatigue qualification tests that they must meet. The other consideration regarding point anchorage systems is the performance of such systems in a rapidly detensioning cable during an earthquake. The concern is that in such a case the wedges could potentially exit the anchorage plate, result- ing in the loss of anchorage. There have not been however any reported cases of such an occurrence, and there is no informa- tion available regarding cable performance in such scenarios. “Hi-Am”-Type Anchorages Figure 9 shows a “Hi-Am”-type socket. The strands or wires splay out at the entrance to a steel socket that is cylindrical on the outside and conical on the inside. The socket is typically filled with epoxy and small steel balls as well as zinc dust. The MTEs terminate at a locking plate. An example of this type of design is the Luling Bridge in Louisiana. This type of anchorage has to be assembled to the right length at a plant and brought to the site, usually on reels. The load transfer between MTE and socket occurs over the length of the socket and not at a single point. The cable must be stressed as a whole. Bond Socket-Type Anchorages Figure 10 shows a bond socket anchorage. In this type of anchorage the strands are terminated at an anchorage plate with wedges, but there also exists a conical pipe (conical out- side and inside) that is filled with either cement grout or epoxy compound. Examples of this type of anchorage include the Clark Bridge in Illinois (with grout-filled socket) and the Cochrane Bridge in Alabama (with epoxy-filled socket). When cement grout is used, the wedges carry the initial stresses on the cable before grouting (or epoxy filling) operations are com- pleted (similar to point anchorage). Following grouting, the socket would be expected to resist changes in cable load. Therefore, the intent of this system is to minimize stress changes at the point anchorages. This type of anchorage can be assembled in the field if the grout or epoxy compound is injected after the initial installation of strands and stressing. Figure 11 shows the results of the survey as related to the types of anchorage systems. In the United States, the conical socket with wedges and the point (wedge) system were dom- inant. It is clear however that the respondents did not simi- larly understand the anchorage characterizations, and some misidentifications may have occurred. Recent Trends in Anchorage Design In recent years, the differences between the approaches to anchorage design of various cable manufacturers’ have nar- rowed to some extent. Currently (in 2005), all of the major stay cable manufacturers in the United States have at least one sys- tem that more or less falls within the point anchorage system described in Figure 8. Shop or Field Cable Fabrications There are two different approaches regarding the assembly and erection of stay cables. In one approach, the stay cables “Hi Am ” Socket Locking Plate ~ Stay Pipe Filler or empty ~ Epoxy-Steel Ball Compound MTE FIGURE 9 “Hi-Am” type anchorage. Stay Pipe Anchor Plate Filler or empty ~ End Cap Ring Nut Anchorage Pipe MTE Anchorage Length Free Length FIGURE 8 Point (wedge) anchorage concept.

12 are shop-fabricated and installed in the field as a unit. This is typical of the Hi-Am-type anchorages. However, field fabri- cations have become more common in the United States, especially in the last decade. For field assembly, the cable sheathings are first welded together on the bridge and lifted into place (Figure 12a). Then, strands are typically individu- ally inserted into the anchor plates at the bottom anchorage and fed through the stay pipes towards the top anchor plate (Figure 12b). The strands can be collectively stressed with large hydraulic jacks or, as is more commonly done today, they are individu- ally stressed with small jacks as they are inserted in the cable. Different cable suppliers have their own procedures and meth- ods to achieve equal force in all strands. Shop fabrications are still very common in Japan. The Alex Fraser (Annacis) Bridge in British Columbia has shop-fabricated cables (Saul and Svensson 1991). The Burlington Bridge in Iowa and the Lul- ing Bridge in Louisiana are examples of shop-fabricated cables in the United States. Saddles The costliest components of a stay cable are the anchorages. Therefore, some designers elect to eliminate anchorages at the pylon by providing a continuous cable through the pylon. The curved saddle at the pylon is typically a steel pipe that re- directs the cable force through the pylon. Another reason given is that the pylon can be smaller (narrower) when there are no anchorages (Figure 13). However, large transverse forces are generated on the cable and individual strands in the saddles, especially at the entrances to the pylon. As the strands enter the saddle, they begin to move to the bottom of the pipe and large interstrand forces can develop, particularly when bare strands are used. Changes in cable tension can result in fretting and fatigue. Such fatigue fractures have been ob- served on at least one qualification test of a saddle system (Tabatabai et al. 1995). In that test, bare strands were used and fractures were initiated at oval-shaped fretting marks at interstrand contact points. To address these issues and reduce interstrand contact, coated strands (such as epoxy-coated) have sometimes been used. In the case of the Maumee River Bridge in Ohio, the engineer designed a “cradle” system in which each strand passes through its own stainless steel sleeve within the cradle assembly (Harris 2002). An FHWA Technical Advisory released in 1994 (“Cable Stays . . .” 1994) discussed a number of factors related to sad- dles and discouraged the use of saddles at that time. However, the use of saddles has continued in the United States. Among the factors cited by the FHWA advisory were: • Stressing of cables with saddles requires simultane- ous stressing from both anchorages (during and after construction); 0.0 20.0 40.0 60.0 80.0 100.0 wedges conical socket w/ wedges cylindrical sockets w/ wedges "Hi-Am" Type other not known no answer Anchorage Type Pe rc en t o f B rid ge s U.S. Canada FIGURE 11 Type and frequency of anchorage used. Anchor Plate ~ ~ Grout or Epoxy Compound Stay Pipe Filler or empty ~ End Cap Ring Nut Bond Socket MTE Grout or Grease FIGURE 10 Bond socket anchorage system.

13 • A more difficult cable removal and replacement process would be required should that become neces- sary; • Precluding slip at the saddles would require special considerations; • In large single saddles, the application of protective tape may become difficult in the vicinity of large sin- gle saddles as spaces between cables are reduced; • Steel pipe at large single saddles should not participate in load transfer to the pylon (i.e., tension in the steel pipe controlled); and • Geometric control through cable length would be more difficult. In 1993, a worldwide survey of stay cable practitioners by Hamilton and Breen (1995) indicated that the majority of respondents did not favor the use of saddles, with European respondents having the highest rate of objections at 76%. The results of the questionnaire in this study showed a total of seven bridges with saddles (21%), six of which were in the United States and one in Canada (Figure 14). STAY CABLE MATERIALS In this section, the materials used in cable systems are dis- cussed, and the importance of detailing and issues of material suitability and compatibility are presented. MTE Materials Steel Today, steel is the predominant MTE material used for stay cables (100% of cable-stayed bridges in the United States and Canada). According to the latest edition of the PTI Rec- ommendations for Stay Cable Design, Testing and Installa- tion (2001), steel wires used as MTEs must conform to the requirements of ASTM A421/A421M, Standard Specifica- tion for Uncoated Stress-Relieved Steel Wire for Prestressed Concrete, Type BA. Strands must conform to ASTM A416/ A416M, Standard Specification for Steel Strand, Uncoated Seven-Wire for Prestressed Concrete, and must be weldless, low-relaxation grade. Bars must conform to ASTM A722/ A722M, Standard Specification for Uncoated High-Strength Steel Bar for Prestressing Concrete. Fiber-Reinforced Polymers In recent years, a number of exploratory efforts have focused on the use of fiber-reinforced polymers (FRPs) in prestress- ing applications and stay cables. These investigations have generally focused on glass, aramid, or carbon fiber-reinforced polymers (GFRP, AFRP, and CFRP). Epoxy-based resins are typically used as the matrix for the composite, and the FRP is made using a pultrusion process. Fisher and Bassett (1997), Christoffersen et al. (1999), Roos and Noisternig (1999), and Noro et al. (2001) provided information on the properties of FRP composites and their comparison to steel. Cable goes thru saddle Cable terminated at pylon Saddle Cable Pylon Deck Pylon Deck Pylon FIGURE 13 Saddles in stay cables. (a) (b) FIGURE 12 Field assembly of cables for the Cape Girardeau Bridge (courtesy: Missouri DOT).

14 Tables 3 and 4 show reported comparisons of different material properties. The main advantages of FRP composite cables are corro- sion resistance and lighter weight. For CFRP, the coefficient of thermal expansion is much lower than steel (approximately one-sixtieth), and the strain at rupture is reported to be 1.6% as compared with 6% for steel (Roos and Noisternig 1999). The main disadvantages of FRP composites are their high cost and very low shear strength (both transverse and inter- laminar shear strengths). The low shear strength seriously affects the gripping ability at anchorages (Christoffersen et al. 1999). Fisher and Bassett (1997) reported that although com- posite materials do not rust, “they can corrode when inte- grated into structures with incompatible materials.” They report that carbon fiber can be subjected to galvanic corro- sion with metals and thus should be insulated from metallic anchorage components. Similarly, glass fiber prestressing tendons “can be susceptible to corrosion under sustained loads when exposed to water or salt water.” Various manufacturers have devised anchorage solutions. These solutions, an example of which is shown in Figure 15, are typically based on a conical steel socket filled with a pot- ting material such as epoxy. However, a wedge-type anchorage system for the carbon fiber composite cables of a pedestrian bridge (the Laroin Bridge in southern France) has been devel- oped. There is a cushioning layer used between the jaws and the rods (Bridge Design and Engineering 2005). A number of demonstration projects have been built with FRPs. However, there are currently no known cable-stayed bridges in the United States and Canada with FRP cables. According to Seible and Burgueno (1997), the first all-com- posite cable-stayed pedestrian bridge was built in Aberfeldy, Scotland, in 1993, with aramid fiber stay cables. These authors also reported on the design of a vehicular cable- stayed composite bridge on the campus of the University of California, San Diego (I5/Gilman). However, this bridge has not been constructed. Christoffersen et al. (1999) reported on the construction of a CFRP cable-stayed bridge in Denmark. To protect against possible damage to cables from fire, impact, or vandalism (saw cutting), the designers used stainless steel sheathing over an extruded HDPE sheath. The design was also based on the ability to sustain static failures of two adjacent cables or a sud- den failure of one cable. Provisions were made for periodic replacement of an original cable at 5-year intervals. The Storchenbrücke (Stork) Bridge in Winterthur, Switzer- land, incorporates two CFRP stay cables, each consisting of 241 parallel pultruded CFRP rods of 5 to 6 mm in diameter (Hooks et al. 1997). The other 22 stay cables on this bridge TABLE 3 TYPICAL PROPERTIES OF THE MOST COMMON FRP MATERIALS AND STEEL Tensile Strength ksi (MPa) Material Young’s Modulus ksi (GPA) Density lb/ft 3 (kg/m 3 ) CFRP (carbon) 245–435 (1700–3000) 20300–43500 (140–300) 100 (1600) AFRP (aramid) 175–305 (1200–2100) 7250–17400 (50–120) 81 (1300) GFRP (glass) 218 (1500) 7250 (50) 150 (2400) Steel 270 (1860) 29000 (200) 490 (7850) Source: Christoffersen et al. (1999). 0.0 20.0 40.0 60.0 80.0 100.0 yes no no answer Saddle Use Pe rc en t o f B rid ge s U.S. Canada FIGURE 14 Use of saddles.

15 have steel MTEs. The stiffness of the anchorage filler material was varied along the length of the anchorage by adding alu- minum oxide pellets with varying thicknesses of epoxy coat- ing. The cables passed qualification fatigue and static testing (Hooks et al. 1997). Roos and Noisternig (1999) reported on fatigue and static testing of CFRP stay cables with up to 91 wires using PTI rec- ommendations. The cable sustained two million cycles of fatigue loading without wire failure, but reached only a maxi- mum of 78% of nominal capacity and thus did not meet the requirements. MTE Coatings Various MTE coatings are available worldwide. These coat- ings are mainly provided for the corrosion protection of MTE. In earlier stay cable designs when uncoated strands and cement grouts were used, it was assumed that grout would provide the necessary protection. However, given that the time between stressing of strands and grouting could be several months or years, it soon became clear that the strands would be left unprotected and could corrode within that time period. One of the early steps taken to address this issue was to use water- soluble oil sprays on the strands (Funahashi 1995). Later, a protective/lubricant coating (a petroleum microcrystalline wax based product) was applied to the strands. Figure 16 shows the results of the survey with respect to the type of MTE coatings used, if any, on bridges in the United States and Canada. Strands and wires that are coated with temporary protection oils, as described previously, are considered bare in the survey. Cables with bare MTEs repre- sent 43% of the survey bridges in the United States, whereas no Canadian bridges use bare MTEs. Galvanizing A very common coating for strands that is used extensively in Europe and Japan is zinc coating (hot dip galvanizing). Galvanizing is a sacrificial form of cathodic protection against corrosion and can be consumed with time, especially in an aggressive environment. In the United States, however, gal- vanized MTEs have not been used very often for stay cables, except for the Sacramento River (Meridian) Bridge in Cali- fornia and the two early bridges in Alaska, including the old- est cable-stayed bridge in the United States, the Sitka Harbor Bridge. The main concern has been that the galvanizing process, especially with strands in contact with grout, could lead to hydrogen embrittlement. Corrosion and other electro- chemical processes can lead to evolution of hydrogen. Ab- sorbed hydrogen can reduce the ductility of steel, through a phenomenon known as hydrogen embrittlement (Barton et al. 2000). On the other hand, 61% of Canadian bridges in the survey used galvanized MTE members. However, none of the Canadian bridges included galvanized MTE in contact with cement grout. It was also believed that the process of galvanizing would degrade the tensile strength of strand and its fatigue life (“Cable Stays . . .” 1994). The concern for contact between galvanized strand and cement is widely held (Ito 1999). However, PTI recommendations state that “galvanized prestressing strand may be used in contact with cement grout provided the steel has been manufactured in accordance with the latest ASTM A416, BS 5896, or EN 10138 standard. Experience has shown that strand manufactured to these standards is not susceptible to hydrogen embrittlement” (Recommendations for Stay Cable Design . . . 2001). The PTI document does not refer to other references that form the basis for that statement. The PTI recommendations also include the following: Galvanized strand is made from either as-galvanized wires (in Japan) or drawn-galvanized wires (in Europe). The advantage of as-galvanized wire is heavier coating weight (300 g/m2) or more for better corrosion protection. The advantage of drawn-galvanized wire, on the other hand, is improved fatigue performance and tighter control on tolerance. In Europe, galvanized wires and strands are routinely used for ungrouted stay cables, and special manufacturing processes are adopted that reportedly ensure compliance with the strength and fatigue requirements including those of the standards listed by the PTI document. In the United States, the market FIGURE 15 One type of CFRP anchorage (Roos and Noisternig 1999). TABLE 4 QUALITATIVE COMPARISON OF FRP PROPERTIES Properties GFRP AFRP CFRP Environmental resistance — + + Tensile strength + + ++ Fatigue strength 0 — ++ Young’s modulus — — ++ Creep/relaxation — 0 ++ Stress fatigue — — ++ Density + ++ ++ Material price ++ — — Notes: — = not good, 0 = neutral, + = good, ++ = very good. Source: Christoffersen et al. (1999).

16 conditions have reportedly not yet justified local production of galvanized strands of sufficient quality (fatigue and strength) for use in stay cables. The Buy America Act enacted by the U.S. Congress in 1933 has so far effectively prevented the importation of stay cable-quality galvanized strands. Accord- ing to the Buy America Act, all federal construction contracts that are undertaken within the United States must use domes- tic construction materials, subject to a few exceptions. There- fore, galvanized wires and strands are currently not being used in U.S. stay cables. Suzumura and Nakamura (2004) studied environmental factors affecting corrosion of galvanized steel wires for sus- pension bridges. They concluded that galvanized steel wires did not corrode when kept in an environment with a relative humidity of less than 60%. The corrosion rate increased sig- nificantly with temperature. They reported that for a wire kept in a wet environment the zinc layer (350 g/m2) would be con- sumed within 10 years. In 100% and 60% relative humidity environments, the consumption of zinc would be complete in 34 years and 211 years, respectively. Figure 17 shows the effects of relative humidity and sodium chloride on the cor- rosion rate. Figure 18 shows the effect of temperature on the corrosion rate. Tarui et al. (2001) reported that galvanized wires with strengths of 256 ksi (1770 MPa) for 7 mm/0.276 in. wires and 284 ksi (1960 MPa) for 5 mm/0.197 in. wires have been developed in Japan. They reported good fatigue and low- temperature response and elongations of 6% to 7%. Tauri et al. (2001) attributed the loss of strength in galvanized wires to the “spheroidizing of cementite, resulting in the col- lapse of the lamellar structure of ferrite and cementite.” They reported that the silicon and chromium elements can suppress this loss of strength. It should be noted that galvanized strands individually sheathed with HDPE are also available and have been used overseas. According to a worldwide survey of the stay cable industry performed by Hamilton and Breen (1995), the galvanized-and-sheathed strand is the most highly rated by the respondents. Individually Sheathed Strands with Corrosion Inhibiting Coating PTI provides detailed recommendations for such strands, which are typically referred to as greased-and-sheathed or waxed-and-sheathed strands (Recommendation for Stay Cable Design . . . 2001). The grease or wax is believed to reduce potential for fretting fatigue resulting from interwire contact (Frank and Breen 2004). These strands are individually coated and then covered with HDPE or high-density polypropylene FIGURE 17 Corrosion rate for galvanized wire (Suzumura and Nakamura 2004). FIGURE 18 Relative corrosion rate—wet condition without chloride exposure (Suzumura and Nakamura 2004). 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 MTE Coating greased- and- sheathed epoxy- coated outside only epoxy- coated in and out galvanized stainless steel other no answerbare Pe rc en t o f B rid ge s U.S. Canada FIGURE 16 Types of MTE coatings used.

17 (HDPP) that is extruded over the strand. These systems have been common in all recently constructed cable-stayed bridges in the United States. Examples of bridges using these types of strands are the Cape Girardeau Bridge in Missouri and the Sixth Street Viaduct Bridges in Wisconsin. Epoxy Coating The use of epoxy-coated strands became popular in the United States in the early to mid-1990s, and was used on at least four bridges in the United States in that decade. Three types of such strands were originally available for stay cables. In one, an epoxy coating with a smooth surface was applied on the out- side perimeter of the seven-wire strands, thus leaving air voids in the interstitial spaces between the six outside wires and the center wire (Figure 19). The second type of epoxy-coated strand produced was similar to the first, except for a grit- impregnated surface to improve bond with grout. The third type of strand had epoxy in all interstitial spaces in addition to the outside surface. The FHWA advisory (“Cable Stays . . .” 1994) and the current PTI provisions (Recommendation for Stay Cable Design . . . 2001) recommend that only epoxy- coated strands with filled interstices should be used for stay cables. Qualification tests in the early 1990s on unfilled strands indicated that pressurized grout water could infiltrate the void spaces inside the strands and remain there as free water, resulting in extensive corrosion and fatigue fractures in the time period of the test (Tabatabai et al. 1995). Although a complete determination of the path of water was not made, it was clear that one likely source was the penetration (and breach) of the epoxy coating at the wedges. Corrosion tests by Hamilton et al. (1998b) also showed corrosion inside unfilled epoxy-coated strands, but no corrosion was found in the filled strands. Saul and Svensson (1991) reported that during the installa- tion of stay cables for the Quincy Bridge in Illinois, “it became apparent that the ends of the strands must be sealed with an epoxy coating in order to prevent moisture rising due to capil- lary pressure through the full height of the cables in the interstices between the individual seven wires forming each strand.” According to the questionnaire response received, moisture has been found in the cable anchorages on this bridge. Cable Sheathings and Wraps Options for cable sheathings include HDPE, steel, stainless steel, or aluminum (Ito 1999). The most common cable sheath- ing is HDPE (Figure 20). Seventeen U.S. cable-stayed bridges included in the survey responses (61%) have HDPE pipes around the cables. However, other bridges such as the Dame Point Bridge (Florida), Maumee River Bridge (Ohio), and the Sunshine Skyway Bridge (Florida) have steel pipes. The new Maumee River Bridge is designed with stainless steel pipes. The cable sheathing, when used, serves as the first line of defense, a barrier against damage or intrusion of harmful sub- stances from the outside. In cases where grout or other fillers are used, the sheathing also serves as a container for the filler. The survey indicates that three bridges in the United States and nine bridges in Canada do not have any external sheathing. The HDPE pipes include approximately 2% to 3% carbon to protect against ultraviolet radiation (Saul and Svensson 1991; Ito 1999). However, considering that the coefficient of thermal expansion of HDPE is much higher than the grout or steel (Funahashi 1995), and that the basic color of HDPE with carbon is black, the issue of increased surface temperatures had to be addressed. Saul and Svensson (1991) reported that the surface temperature of black pipes can reach more than 149°F (65°C) owing to direction solar radiation, whereas the surface of a white pipe under the same condition would reach only 104°F (40°C). Paint does not adhere well to HDPE. Until recently, new HDPE-covered cables were commonly wrapped with a light color self-adhesive polyvinyl fluoride (PVF) tape (mostly referred to by the commercial name Tedlar®), which was spirally wrapped around the HDPE pipe. Typically, a 50% overlap is provided. Some damage has been reported on wrapped tapes in some bridges (based on the survey results). In one case, the Pasco– Kennewick Bridge in Washington State, the damage was reported to be extensive. In that case, polyvinyl chloride (PVC) tapes were first used over the pipes, and these tapes became brittle after several years and began to flake off (Saul and Svensson 1991). In other cases, the damage was reported to be minor. A laminated tape consisting of a trans- lucent Tedlar tape with a color PVC tape backing was also developed (Saul and Svensson 1991). In recent years, co- extruded HDPE pipes with bright surface colors have entered the market, and recently constructed bridges use this approach in lieu of the PVF tape. Figure 21 shows the results of the sur- vey with respect to damage to the protective tape. In tests performed by Hamilton et al. (1998), clear HDPE sheathing was used to allow assessment of grout condition inside the sheathing. There is no information available that would indicate if clear HDPE sheathing has ever been used FIGURE 19 Epoxy-coated strand (Funahashi 1995).

0 10 20 30 40 50 60 70 80 yes no not known not applicable no answer Wrapping Tape Pe rc e n t o f B rid ge s U.S. Canada FIGURE 21 Damage to wrapping tape. 18 on stay cables in the field to facilitate inspections. An impor- tant challenge would be the required resistance to ultraviolet radiation. Figure 22 shows the results of the survey with respect to the cracking of cable sheathing or sheathing connections. Four bridges or 14% of the respondents in the United States (and none in Canada) reported problems with the sheathing or con- nections. In the C&D Canal Bridge in St. Georges, Delaware, cracking of the steel sheathing was noted on one of the stay cables. This cracking was attributed to the position of a grout vent hole at a high-stress location near a pylon. The respon- dents to the questionnaire also identified two bridges that had splitting of HDPE [Quincy (Illinois) and Luling (Louisiana)]. In field-fabricated cables, the HDPE pipe segments are typ- ically assembled and welded together (HDPE welding) by spe- cial machines on the bridge deck before being lifted into place. In some shop-fabricated cable systems, the cable assemblies (including HDPE) are assembled, coiled, and then shipped on large reels. The coiling and uncoiling of HDPE pipes at low temperatures can lead to cracking (Funahashi 1995). In newer shop-fabricated Japanese cable systems, the HDPE is extruded over the MTE bundle, thus creating a tight fit between the sheathing and MTE. When a cable with HDPE sheathing is grouted, the pipe must resist grouting pressures. This would increase the required thickness of the pipe. On some of the early bridge projects, such as the Zarate–Brazo Largo Bridges in Argentina and the Luling Bridge in Louisiana, there have been problems with grouting operation that reportedly contributed to the cracking of the HDPE pipe (Saul and Svensson 1991). The authors discussed HDPE stresses as a result of coiling and uncoiling, effects of grouting pressures, and effects of high temperatures at the time of grouting. Steel pipe segments are typically welded together in the field. The external pipe is generally bolted to the anchorage pipe. The axial and flexural stiffness of the steel pipe is far greater than that of the HDPE. In a grouted system, sufficient bond between the grout and the steel pipe can be developed, thus transmitting some of the fluctuating cable stresses into the sheathing (owing to strain compatibility). In some quali- fication tests, steel sheathing connections developed fatigue fractures as a result of this unintended effect (Tabatabai et al. 1995). Saul and Svensson (1991) also reported that “some welded connections have failed in the past” without elaborat- ing. Steel sheathings must also be periodically painted to pro- tect against corrosion. Fillers and Blocking Compounds Fillers refer to materials placed inside the sheathing and around the MTE. In United States practice, the most common 0 10 20 30 40 50 60 70 80 HDPE steel Pipe no sheathing other no answer Cable Sheathing Pe rc en t o f B rid ge s U.S. Canada FIGURE 20 Types of cable sheathing used.

19 type of filler within the HDPE pipe and the MTE has been the cement grout. Table 5 shows the results of the survey with respect to the type of fillers used, if any, in the stay cables. Fifty-four percent of U.S. bridges in the survey include some type of cement grout in the free length of the cable. None of the Canadian bridges use cement grout. Table 6 shows responses to the survey with respect to the filler materials used in the anchorage zones. Portland Cement Grout There have been a variety of opinions on the merits of cement grouts for stay cables. As stated earlier, the practice of grout- ing stay cables comes from the post-tensioning technology, and not from the suspension cable technology. Grouting has not, in general, been very popular in Europe (Hamilton and Breen 1995). The main advantages typically given for cement grout in stay cables are: • Cement grout provides a physical barrier for the MTE that is not easily breached. • Grout provides an alkali environment for the bare steel and protects it against corrosion. • The increased mass owing to the grout helps with damp- ing and vibration control. The disadvantages are: • Stress fluctuations in the cable and grout shrinkage can result in the cracking of the grout. This cracking can lead to intrusion of moisture if the external sheathing is breached. • Grouting adds to the cost of cables. • Grouting could complicate many types of NDT and inspections. • Grout water and bleed water could present internally driven corrosion danger when not properly controlled. Voids could be introduced inside grouted ducts. Tabatabai et al. (1995) performed qualification tests on some grouted stay cable specimens with uncoated (bare) strands. The dissection of cable specimens after fatigue and static tests indicated transverse cracking in the grout. Corro- sion was noted on the strand at the intersection of the grout cracks and the strand, some with surface pitting. Fatigue frac- tures were also noted at those locations, thus establishing that the cracking occurred early in the fatigue test and not in the subsequent static test. Figure 23 shows corrosion at the loca- tion of grout cracking. Ito (1999) refers to the presence of grout cracks and how they may be associated with potential for “fretting corrosion” of steel wires. U.S. % U.S. Canada % Canada % total Grout not used 6 21.4 12 92.3 43.9 Cement–water 5 17.9 0 0.0 12.2 Cement–water– admixtures 9 32.1 0 0.0 22.0 Commercial pre- packaged grouts 1 3.6 0 0.0 2.4 Not known 6 21.4 0 0.0 14.6 Not applicable 0 0.0 1 7.7 2.4 No answer 1 3.6 0 0.0 2.4 TABLE 5 SURVEY RESULTS—TYPE OF GROUT USED? (Question 4.7) TABLE 6 SURVEY RESULTS—ARE FILLER MATERIALS USED IN THE ANCHORAGE ZONE? (Question 4.8) U.S. % U.S. Canada % Canada % total Yes—grout 6 21.4 0 0.0 14.6 Yes—grease 10 35.7 0 0.0 24.4 Yes—other 7 25.0 2 15.4 22.0 No filler 0 0.0 11 84.6 26.8 Not known 4 14.3 0 0.0 9.8 No answer 1 3.6 0 0.0 2.4 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 yes - sh ea thi ng yes - co nn ec tio ns yes - bo th no no t k no wn no t a pp lica ble no a ns we r Cracking Pe rc e n t o f B rid ge s U.S. Canada FIGURE 22 Cracking of cable sheathing and connections.

20 There are conflicting results from examination of grout conditions on four bridges. In these bridges, the Cochrane Bridge (Alabama), Pasco–Kennewick Bridge (Washington State), Talmadge Memorial (Georgia), and the Fred Hartman Bridge (Texas), the HDPE sheathing was partially removed (windows cut) during inspections to allow examination of the condition of grout and wires (Grant 1991; Tabatabai et al. 1998; Dowd et al. 2001; and survey results). In the Cochrane Bridge, no cracking of grout or corrosion of MTE was reported (Tabatabai et al. 1998a). Grant (1991) also reported no grout cracks for the Pasco–Kennewick Bridge. However, Frank and Breen (1994) reported that in the Pasco–Kennewick Bridge there were small, closely spaced grout cracks perpendicular to the stay (Figure 24). The inspection of grout for the Fred Hart- man Bridge indicated “fine, intersecting transverse and longi- tudinal cracks, with spacing between 12 and 19 mm.” These cracks were not readily evident. It is not known whether longitudinal cracking of HDPE pipes in some bridges (such as Zarate–Brazo Largo in Argentina) affected the integrity of the grout, because no examination of the grout was reported in the literature for these bridges. Ito (1999) reported that grout cracking has been observed on some cable-stayed bridges, but does not provide additional information. Saul and Svensson (1991) also reported grout cracking on cable test specimens. Hamilton et al. (1998a,b) reported on accelerated corro- sion tests done on eight grouted stay cable specimens, five of which were uncoated (bare) strands, one with epoxy-coated strands (some filled and some unfilled), one with a galva- nized strand, and one with a greased-and-sheathed strand. These specimens were loaded and grouted inside clear pipes to allow visual examination of grout surface. Windows were cut into the pipes to simulate damage to the HDPE. Wet and dry saltwater ponding cycles were initiated to represent long- term ingress of chloride-laden air and moisture through the openings. The main objective was to determine if the cement grout can provide positive secondary protection if the first protective layer (HDPE) is breached. Hamilton et al. (1998a,b) concluded that a relatively low level of loading above the grout injection load would result in grout cracking. The salt solution was able to reach almost any location in the speci- mens, and the primary mechanism for corrosion was crack- ing of grout at sheathing breaks. According to the authors, galvanized, greased-and-sheathed, and filled epoxy-coated strands provided vast improvement over the bare strands. Corrosion was observed inside the unfilled epoxy-coated strands. Therefore, the authors concluded that the traditional grout-bare-strand HDPE system could no longer be consid- ered a redundant system. Frank and Breen (2004) concluded that the use of portland cement grout has not proven to be an effective corrosion protection barrier. When wires of bare strands are encased in grout fracture, the force in the broken wire redevelops a relatively short distance away from the fracture. Some qualification tests have shown multiple fractures on the same wire over a length of a few inches (Tabatabai et al. 1995). Therefore, the overall cable axial stiffness would not necessarily change when limited numbers of individual wire breaks occur, especially when those breaks are spread over some distance. This can be viewed as both pos- itive and negative; positive because cross-section strength at locations away from fracture would remain unchanged and negative because monitoring of cable force changes (or deck profile deflection changes) would not indicate loss of section because stiffness has not been affected substantially. The global stiffness of the cable would remain essentially unchanged even when moderate wire section losses occur. Other Fillers Ito (1999) reported that cement grout plasticized with poly- urethane has been used in some bridges. A synthetic resin material based on polybutadiene was used on two Japanese bridges (Ito 1999). Grease and wax have also been used. In the Alex Fraser (Annacis) Bridge in British Columbia, petro- leum wax blocking compound was used inside the sheathing. FIGURE 23 Corrosion of strand at transverse crack in grout. FIGURE 24 Exposed cable on the Pasco–Kennewick Bridge (Frank and Breen 1994).

21 neoprene ring) and its reliability is subject to debate, it is clear that problems with neoprene rings can exacerbate cable vibra- tion problems. Telang et al. (2000) reported on problems with the washers on the Cochrane Bridge that likely contributed to excessive rain–wind vibrations. The cable was not centrally located in the middle of the steel anchor pipe (or the guide pipe); thus, the thickness of neoprene around the HDPE was variable. Also, there were gaps between the neoprene ring and the cable along the perimeter (see Figure 26). This would reduce the effectiveness of the washer both in reducing bend- ing stresses and in damping vibrations. The steel rings that typically hold the washers in place (“keeper rings”) can fail and result in the dislocation and mis- alignment of the neoprene ring (Figure 27). FIGURE 25 Anchorage detail including neoprene washers for the Cochrane Bridge (Alabama) (Telang et al. 2000). Wax is injected at high temperatures and solidifies when cooled, resulting in shrinkage and cracking. Ito (1999) reported that a type of petroleum wax that could be applied at ambient temperature has been developed. Hemmert and Sczyslo (1999) reported that red lead is com- monly used in locked coil cables. A coating of paint is some- times used over the locked coil cables. One option is not to have any fillers inside the HDPE pipe. That is the approach used on the Charles River Cross- ing Bridge in Boston, Maumee River Bridge in Ohio, Sixth Street Viaduct Bridges in Wisconsin, and Cooper River Bridge in South Carolina, where individually coated and sheathed strands (or epoxy-coated strands) are used inside HDPE pipe without cement grout. The Charles River Bridge is believed to be the first ungrouted parallel strand stay cable system built in the United States and marks a major shift in the stay cable technology in this country. In response to the ques- tionnaire, the cable manufacturer for the Charles River Bridge, Freyssinet LLC, stated that the ungrouted system would improve inspectability and allows for future replace- ments. All of the major cable suppliers in the United States currently offer cable systems with the no-grout option. Little et al. (2001) discussed fungal-influenced corrosion of post-tensioned tendons. They reported that bacteria have been implicated in corrosion of tendons in structures. Their experiments showed the fungal degradation of lubricating grease, which produced formic and acetic acids resulting in corrosion of steel cables. Fusarium sp., Penicillium sp., and Hormoconis sp. were isolated from corroding tendons in a post-tensioned structure and used in testing. The test speci- mens were coated with “metal soap hydrocarbon grease” be- fore insertion into PVC sheathing. There were no indications of chlorides in the energy-dispersive X-ray analysis system spectra of the grease. This article did not refer specifically to stay cables. Neoprene Rings A stay cable is subjected to lateral movements as a result of vibrations. These vibrations create bending stresses at the two ends of the cable, thus increasing the potential for fatigue. To address this issue, most cable-stayed bridges in the United States have what are termed neoprene rubber “washers,” “rings,” or “donuts” placed around the HDPE pipe within the guide pipe (anchor pipe) near the ends of the cables. Figure 25, a diagram of the anchorage for the Cochrane Bridge in Alabama, shows the typical position of the neo- prene washer with respect to the other components of cable anchorage. In addition to reducing bending stresses at the anchorages, the neoprene rings also contribute to the vibration damping of the cables. Although the level of damping (attributed to the FIGURE 26 Cap between cable and neoprene washer (Telang et al. 2000).

22 Figure 28 shows the results of the survey with respect to the use of neoprene rings. Most cable-stayed bridges in the United States (64%) use neoprene rings, whereas 31% of Canadian bridges have neoprene rings. Responses to the questionnaire indicated that seven bridges in the United States and two bridges in Canada had problems with movements of the rings out of position for various rea- sons, indicating that this is a relatively common problem (see Figure 29). At least one cable supplier has developed a proprietary vis- coelastic damping system that also serves the purposes of the neoprene washer. The topic of vibration damping is discussed later in this report. Neoprene Boot Neoprene boots are generally used to cover the gap between the cable sheathing and the end of the guide pipe near the neo- prene ring. Figure 30 shows a typical neoprene boot that is in good working condition. Typically, hose clamps are used to tighten the boots against the sheathing and the guide pipe. In some cases, it has been observed that the clamps become dis- placed and rainwater can enter the guide pipes. Bloomstine and Stoltzner (1999) reported on water intrusion into the top and bottom neoprene boots. They recommended using sili- cone filler under the boot before clamping. Responses to the questionnaire for at least three bridges indicated problems with neoprene boots. Table 7 shows the survey responses related to problems with neoprene boots. Four bridges in the United States and one in Canada reported problems with neoprene boots. However, two of the four U.S. cases referred to neoprene boots that are not as described above. STAY CABLE DESIGN CHALLENGES Aside from structural strength, the design of stay cables also must address the challenges of corrosion, fatigue, vibration, inspectability, and maintainability. More recently, considera- tion of extreme events such as fire, ice, blasts, impacts, and earthquakes are attracting more attention in the design of stay cables. In this section, the mechanisms for corrosion, fatigue, and vibrations (including rain–wind vibrations) are first dis- cussed, followed by a discussion of the challenges of design- ing stay cables for inspectability and maintainability. The PTI recommendations, including qualification tests, are reviewed. Finally, a brief outline of issues related to extreme events is discussed. It should be noted that although these issues are presented separately, they are highly interrelated and cannot be con- sidered independent. For example, corrosion and vibrations could have major negative influence on fatigue performance. The ability to inspect and maintain also influences durability of cables in all areas. These major structures must safely carry traffic for a long time. Therefore, a clear understanding of the durability limits and issues is very important. Corrosion Corrosion protection for stay cables is understandably one of the primary concerns of designers, suppliers, and owners involved in cable-stayed bridges. According to the PTI Rec- ommendations for Stay Cable Design, Testing, and Installa- tion (2001), a minimum of “two nested qualified barriers” 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 yes no not known not applicable no answer Neoprene Rings Pe rc en t o f B rid ge s U.S. Canada FIGURE 27 Failure of the keeper ring and dislocation of neoprene washer (Telang et al. 2000). FIGURE 28 Use of neoprene rings.

23 must be provided for the corrosion protection of the MTE. By clearly specifying minimum acceptable levels of protection and setting performance requirements with respect to corro- sion, these provisions are a major departure from earlier prac- tices. A two-tier system is established in which the individual barriers must first be qualified through testing, followed by the testing of the nested barriers as a system. As the number of nested qualified barriers (that are compatible with each other) increase, the system redundancy and reliability is expected to improve. It is important to realize that corrosion can be either inter- nally or externally driven. The primary mode of protection has naturally and rightly been against externally driven corrosion (i.e., moisture and other harmful substances entering from out- side). However, internally driven corrosion mechanisms have also been observed and must be addressed in design and main- tenance. Examples of these corrosion mechanisms include corrosion resulting from the presence of free grout water in different components of cables. Design of cable components for corrosion resistance should consider, when applicable, the effects of extreme tempera- tures, solar radiation, shrinkage or expansion of fillers, age, vibration, and fatigue on the effectiveness of the system. As will be seen later in this report, recent trends have been toward inclusion of additional features such as vibration control and force measurements as part of the cable design. This consti- tutes a “system approach” to the design of stay cables. Embed- ded corrosion monitoring systems can also be considered as technology develops further. There have been a number of debates over the years on the issues of corrosion and the overall health of stay cables. In a 1988 article, “Cables in Trouble,” Watson and Stafford (1988) presented an alarming picture of the condition of stay cables, indicating that cable-stayed bridges were in serious danger as a result of corrosion. The authors pointed to corro- sion (of the Kurt Schumaker Bridge in Germany), vibration (of the Brottone Bridge in France), intersliding of wires, and long-term creep behavior of cables as evidence of serious challenges for cable protection. In response, in a 1991 article, “Cables Not in Trouble,” Grant (1991) countered that cable- stayed bridges were not in danger of failure from corrosion of cables. Grant reported tests on the Sitka Harbor Bridge in Alaska involving removal of six galvanized structural strand cables and their examination by magnetic, ultrasonic, radi- ographic, and X-ray methods. All cables were reported in “nearly new condition.” Tests were also performed on the Meridian Bridge in California and the Pasco–Kennewick Bridge (Washington State), and the steel elements were reportedly found to be without corrosion. Saul and Svensson (1991) discussed some of the damage observed on cable-stayed bridges. In the case of the Kohlbrand Bridge in Germany, they reported on the detection during inspections of 25 broken wires on the nongalvanized locked coil cables that were protected with red lead and linseed oil 0 20 40 60 80 yes no, not known not applicable no answer Neoprene Ring Problems Pe rc en t o f B rid ge s U.S. Canada FIGURE 29 Movements of neoprene rings out of position. FIGURE 30 Neoprene boot (courtesy: Indiana DOT). TABLE 7 SURVEY RESULTS—PROBLEMS WITH NEOPRENE BOOTS (Question 4.24) U.S. % U.S. Canada % Canada % total Yes 4 14.3 1 7.7 12.2 No 19 67.9 6 46.2 61.0 Not known 1 3.6 0 0.0 2.4 Not applicable 1 3.6 3 23.1 9.8 Other 2 7.1 0 0.0 4.9 No answer 1 3.6 3 23.1 9.8

24 (Figure 31). The bridge was under construction between 1969 and 1974, and the wire breaks were found in 1976. The authors attributed failures to many factors including misalignment of cables, missing protection at the sockets, cable vibrations, and possible effects of deicing salts. The Lake Maracaibo Bridge in Venezuela also suffered corrosion of its galvanized locked coil cables after fewer than 18 years of service, and all of its cables were subsequently replaced in 1980 (“Cable Stays . . .” 1994). Saul and Svensson (1991) reported that the causes included inadequate mainte- nance and painting in the hot marine climate and a mistake made in not reinstalling neoprene boots during inspections, which resulted in a humid microclimate. Figure 32 shows the fracture of wires on the Lake Maracaibo Bridge. Sarcos-Portillo et al. (2003) reported that inspections car- ried out in 1997–1999 revealed “corrosion in both cables and sockets, as well as considerable settling in the sockets” of the new cables. A “significant” amount of water was also found in most sockets. Vibration-based tension force measurements indicated major force changes. Deck profile changes were also noted. The cables were retensioned, and they recom- mended painting the cables and waterproofing and lubricat- ing the sockets. The response to the questionnaire for the Fox Hollow pedestrian bridge in Calgary, Canada, indicated failures of two galvanized bars used as MTEs, and replacement of a third bar. There were no sheathings or grout used on these cables. On further inquiry, the respondent reported that the failures were without any sign of prior problems. An evalu- ation has reportedly been performed by outside experts and the failure mode was reported as “corrosion induced fatigue.” The remaining bars were examined and a third bar was iden- tified with a corrosion pit and replaced. Wire rope cross cables were installed after the failures. No further information was available at this time. As discussed earlier, Saul and Svensson (1991) reported on the cracking of the grouted HDPE pipes on the Luling Bridge in Louisiana and the twin Zarate–Brazo Largo Bridges in Argentina. The longitudinal cracks in the pipes were attributed to high strains owing to grouting during hot temperatures. Sub- sequent cooling against hardened grout creates stresses in the pipe. Both bridges used shop-fabricated cables that were deliv- ered on reels. In the case of the Argentine bridges, they were left on reels for up to 3 years. In the case of the Luling Bridge, failures of the butt welds between HDPE segments were also noted, which were attributed to malfunctioning welding equip- ment and uncoiling at low temperatures. Repair of HDPE in both of these bridges included filling cracks with polyurethane grout and wrapping them with filament tape and PVF tape (Saul and Svensson 1991). FIGURE 31 Corrosion and rupture of locked coil cable on the Kohlbrand Bridge in Germany (Frank and Breen 2004). FIGURE 32 Corrosion of locked coil cables of the Lake Maracaibo Bridge in Venezuela (Frank and Breen 2004).

25 coming out of one end cap as bolts are loosened. Various degrees of corrosion were noted in the cables (see Figure 35). Telang et al. (2004) concluded, based on vibration-based measurements of cable forces, that “the cables have not suf- fered any significant damage.” They do not however discuss whether corrosion damage would necessarily result in global stiffness changes in grouted cables resulting in force changes. Further testing is planned for the Luling Bridge. In the case of the Zarate–Brazo Largo Bridges, Saul and Svensson (1991) stated “five years after the repair the cables were inspected by the Argentine Federal Highway Adminis- tration and found in good condition.” It is estimated that the inspection was probably performed around 1987. In November 1996, the first ever rupture and complete fail- ure of a parallel wire stay cable occurred on the Guazu Bridge in Argentina, one of the two Zarate–Brazo Largo Bridges (Andersen et al. 1999). These bridges were built in the early FIGURE 33 Neoprene washers on the Luling Bridge at the exit point of cables from the steel box (Telang et al. 2004). Telang et al. (2004) reported on inspections of the cables of the Luling Bridge in 2002 and 2004. They reported that, at least in one location, exposed and rusted stay cable wires were detected. The original “epoxy repair” had deteriorated and resulted in the rupture of the protective tape and filler grout and corrosion of wires. Extensive water leakage inside sock- ets of deck level anchorages was observed. Water dripping from the split rings and shims was observed at most locations. It was suggested that rainwater entered the steel box at the cable exit locations through gaps in neoprene washers. It should be noted that the neoprene washers on the Luling Bridge are different from those described earlier. They sur- round the sheathing and are caulked to the opening at the top of the box girder (Figure 33). The caulk that was used around the washers was weathered, cracked, or missing at some locations. Neoprene washers were removed and a video boroscope (videoscope) examination was performed. Accumulated water was found surrounding the cable inside the box. The end caps of sockets were removed to expose the button end of the wires. Figure 34 shows water FIGURE 34 Water exiting the end cap of one anchorage (Telang et al. 2004). FIGURE 35 Corrosion at the end plate of one socket with wire button ends (Telang et al. 2004).

0.0 10.0 20.0 30.0 40.0 50.0 60.0 yes no not tested not known not applicable no answer Presence of Moisture Pe rc en t o f B rid ge s U.S. Canada FIGURE 36 Occurrences of moisture inside stay cables. 26 There has been some work done on the corrosion and embrittlement of high-strength wires for suspension bridges, which can be relevant to stay cables as well. Laboratory work by Barton et al. (2000) reported that “corrosion degradation of high-strength wires exceeds mere loss of load-bearing mate- rial.” Wire strength was reduced more than the cross-sectional area suggesting that “cracking or pitting effects may be pres- ent, whether induced by corrosion or by hydrogen interaction, or both.” Their studies indicated that hydrogen was absorbed into the corroded wire, with hydrogen retention being higher in galvanized wire. Corrosion results in higher embrittlement of both galvanized and nongalvanized wires. Mayrbaurl and Camo (2004) reported on a study of struc- tural safety of suspension bridge parallel-wire cables. They dis- cussed issues related to corrosion of galvanized wires, in- cluding categorization of wire corrosion in four stages. They also presented cable strength models based on field assessments of wire data. However, unlike stay cables, the primary tool for inspection in main suspension cables is the removal of outside wrapped wire and the physical opening of the cable (insertion of wood wedges) to visually inspect the interior of the cable. Despite some similarities, suspension main cables and stay cable have major differences in design, materials, inspection processes, deterioration mechanisms, and anchorage systems. However, information related to long-term deterioration of gal- vanized wires is still valuable to the stay cable community. In 1992, the U.S. Patent and Trademark Office issued Patent No. 5,173,982 to inventors T.G. Lovett and S.L. Stroh for a corrosion protection system for stay cables (“Immersion of Stays . . .” 1993). It is designed to keep the stay cable immersed in a lightweight, corrosion-resistant fluid within the cable sheathing. It is not known if this concept has been used on any actual stay cables. Kitagawa et al. (2001) reported on a dry-air injection sys- tem used to reduce humidity levels inside the main cables of the Akashi Kaikyo Suspension Bridge and other bridges in Japan. The system includes salt filters to remove chlorides. Humidity measurements inside the cable reportedly show the effectiveness of the system. Figure 36 shows survey responses with respect to moisture found inside the stay cable components. Respondents for 25% 1970s. The cable consisted of grouted nongalvanized parallel wires within HDPE pipes and anchored within Hi-Am-type sockets (Andersen et al. 1999). According to the authors, “a combination of corrosion and fatigue has been found to be the cause. The corrosion has taken place due to insufficient cor- rosion protection of the non-galvanized wires. The likely cause is that the cement grout, which was supposed to be the main corrosion protection, was insufficient in the anchorage zone due to the presence of a non-protecting epoxy tar.” They also stated that “following intrusion of water through defects in the PE pipe or due to condensation of water inside the PE pipe, corrosion has been initiated.” A complete rehabilitation of the bridge was planned for 1999/2000. The cable had failed in an area near the entrance to the bot- tom anchorage. Subsequent ultrasonic testing on other anchor- ages revealed damage to other cables, with up to 62% wire breaks. The cable with 62% wire breaks had adjacent cables with 41% and 20% breaks. Damage to bottom anchorages was significantly greater than to top anchorages. Cable force mea- surements reportedly indicated that forces in the cables had changed by as much as 20% when compared with forces at the inauguration of the bridge. This however appears to have in- cluded the effect of the lost cable, and it is not clear whether the forces at the inauguration of the bridge were actually measured or estimated by the designer. Large amplitude cable vibrations (reportedly not rain–wind vibrations) had taken place on this bridge. During emergency repairs, 13 cables were replaced. Prato et al. (1997, 1998) reported on the replacement of all locked coil cables of the Chaco–Corrientes Bridge in Argentina. The locked coil cables had external galvanized wires. This bridge was built in 1973. Failure of several z-shaped wires (in the external layer of wires) on four cables occurred in 1986 and the cables were replaced in 1996. Reinholdt et al. (1999) reported on the replacement of all wire rope cables of the Luangwa Bridge in Zambia in 1997. The bridge was built in 1968. The shop-fabricated cables were originally made longer than required resulting in a dip in bridge deck surface. This was addressed by installation of “cable clamps” to reduce cable length by approximately 135 mm (5.3 in.). Severe corrosion and pitting of cables was noted in 1997, resulting in replacement of all cables.

27 of bridges in the United States indicated that moisture has been found inside cables. Another 21% either have not tested or do not know if moisture exists. Figure 37 shows survey responses with respect to MTE corrosion. Only one U.S. bridge and two Canadian bridges were reported to have evidence of MTE cor- rosion. The corrosion status of five other U.S. bridges and three Canadian bridges was reported to be unknown. As will be discussed later, assessments of MTE corrosion in cable- stayed bridges are, in many cases, very difficult. Fatigue The PTI Recommendations for Stay Cable Design, Testing, and Installation (2001) provide detailed fatigue and static qualification testing requirements for stay cables. Three cable specimens are typically tested for each bridge. These tests include two million cycles of loading, with a stress range of 28 ksi (159 MPa) and a maximum stress equivalent to 45% of the cable’s nominal strength. The number of wire breaks dur- ing fatigue tests should not exceed 2% of the total number of wires in the cable. After fatigue tests, cables are loaded stati- cally to achieve a target load of 95% of the nominal strength or 92% of the actual strength of the strands. Some European codes such as the SETRA/CIP require fatigue tests that include a small angle change (rotation) induced at the anchorages. The PTI requirements do not have this provision at this time. The PTI recommendations also specify procedures for axial and flexural tests involving cable saddles. It should be noted how- ever that the PTI qualification tests do not specifically address fatigue issues related to cable vibrations. In response to the observed rain–wind vibrations on two bridges in Texas, Dowd et al. (2001) began a research proj- ect aimed at developing a set of procedures for evaluating fatigue damage in stay cables resulting from large amplitude and rain–wind-induced vibrations. This effort includes test- ing of cable specimens in the laboratory as they are subjected to axial loads and simultaneous cyclic lateral loads at the mid-point of the cable. The authors reported that similar tests were done in Japan on cables with 163 parallel and galva- nized wires with Hi-Am-type sockets and PE pipes (without grout). In the Japanese tests, angle changes of ±1.35° pro- duced fatigue failures at 0.26 million cycles, whereas no fatigue failures were observed after 10 million cycles for a ±0.9° angle change. Frank and Breen (2004) discussed stay cable bending fatigue test results in which performance of grouted and ungrouted stay cable specimens were compared. Bare strands were used and the two cable types were identical except for grouting. The number of wire breaks recorded was much higher in the grouted specimens. The authors suggest that the grout acts as an abrasive that reduces fatigue life resulting from fretting. Prato and Ceballos (2003) studied dynamic bending stresses near anchorage sockets for grouted cables with HDPE pipes, but with bituminous epoxy replacing grout just before the anchorage (Figure 38). The authors show that the dynamic stresses in wires are higher, and stress concentration occurs, when such a discontinuity is present (i.e., grout is replaced by bituminous epoxy). They noted that shear deformations in such cases would not be negligible, and the dominant discontinuity would be that of shear stiffness and not bending stiffness. Fig- ure 39 shows the results of the survey with respect to fatigue. Vibrations Since the mid-1980s, bridge owners and researchers have reported large-amplitude stay cable vibrations with increas- ing frequency. This has resulted in increased concern for the fatigue performance of cables. Figure 40 shows vibrations recorded on the Cochrane Bridge in Alabama, and witnessed by this writer. Categories of Vibration The primary types of stay cable vibrations are as follows (Irwin 1997): • Rain–wind induced vibrations, • Sympathetic vibration of cables with other bridge com- ponents excited by wind (parametric excitation), • Inclined cable galloping, • Vortex excitation (single cable or groups of cables), 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 yes no not known not applicable no answer Corrosion Pe rc en t o f B rid ge s U.S. Canada FIGURE 37 Incidence of MTE corrosion.

28 • Wake galloping, and • Buffeting by wind turbulence. The rain–wind-induced vibrations are by far the most widely reported, large-amplitude (up to a few feet) vibration phenomenon in stay cables. It was first reported on the Meiko– West Bridge in Japan in 1986 (Matsumoto 2000), and has since been reported on many bridges worldwide. This phe- nomenon occurs in moderate wind and rain conditions, and is believed to be caused by an aerodynamic instability result- ing from the formation of water rivulets on the surface of the cable. However, uncertainties still exist regarding this phe- nomenon (Matsumoto 2000). When vibrations are occurring, the speed of the wind is sufficient to maintain the upper rivulet within a critical zone (Irwin 1997). Larose and Wagner Smitt (1999) discussed the results of their wind tunnel studies and reported that rain–wind vibrations were reproduced in the laboratory for a single cable and for cables in tandem configuration. They also reported that the rivulet changes its position with wind speed and also cable motion. Miyazaki (1999) reported that the lower rivulet is formed at lower wind speeds, and both rivulets appear at higher speeds. This is consistent with what this writer observed on the Cochrane Bridge in 1998. In this case it was the lower rivulet that appeared first; however, it was the subsequent formation of the upper rivulet that initi- ated large amplitude vibrations. Also, the rivulets appeared to oscillate up and down within a “wet” band as they moved down the cable (Figure 41). Larose and Wagner Smitt suggest that the “wetability” of the cable surface is important in the formation of rivulets. They noted that a slightly eroded surface with dust particles is more “wetable,” and thus can form the rivulets more eas- ily. This may be the reason why some bridges do not experi- ence rain–wind vibrations for the first few years of their ser- vice. According to Swan (1997), “a very smooth surface may initially avoid the problem, until atmospheric deposits allow just enough roughness to hold the rivulet.” 0.0 20.0 40.0 60.0 80.0 100.0 yes no not known not applicable no answer Fatigue Pe rc en t o f B rid ge s U.S. Canada FIGURE 39 Incidence of fatigue of MTEs. FIGURE 38 Deformations near cable anchorages with discontinuous grout (Prato and Ceballos 2003).

29 Jones and Porterfield (1997) reported on the instrumenta- tion and long-term vibration monitoring of the East Hunt- ington Bridge in West Virginia. They reported random buf- feting response, locked-in vortex-induced vibration, and rain– wind oscillations. They noted that significant displacement responses are in the lower modes of the structure. High acceleration values at higher modes do not mean high dis- placements at those frequencies (acceleration amplitudes are equivalent to displacement amplitudes multiplied by fre- quency squared). Main and Jones (2000) also reported on the instrumentation and long-term vibration monitoring of the Fred Hartman and Veterans Memorial Bridges in Texas. Figure 42a shows a sample histogram of dominant modes for one stay cable, and Figure 42b shows vibration amplitudes versus wind speed for the same cable. Main and Jones (2000) concluded that the highest ampli- tude responses (which occurred during rainfall) were in the lower modes and “seemed to ‘lock-in’ to a specific mode of vibration over a wide range of wind speeds.” High-frequency vibrations over narrow wind ranges were also observed, which were attributed to vortex-induced vibrations. Tabatabai et al. (1998a) and Lankin et al. (2000) have reported on vibration measurements and mitigation efforts for the Cochrane Bridge in Alabama. In these studies, the level of damping in all cables was measured and studies were performed to determine and optimize mitigation solutions. Irwin (1997) recommended the following equation for controlling rain–wind vibrations: where Sc = Scruton number, m = mass per unit length of cable, ζ = damping ratio, ρ = density of air (1.225 kg/m3), and D is the cable diameter. This equation has been adopted in the PTI Recommendations for Stay Cable Design, Testing, and Installation (2001) for control of rain–wind vibrations. Tabatabai and Mehrabi (2000) used cable information from 16 cable-stayed bridges to determine the level of damp- ing required based on Eq. 1. Figure 43 shows a histogram of required cable damping for all stay cables in those 16 bridges. These data indicate that 90% of the cables would meet the requirements of Eq. 1 with a damping ratio of 0.7%. The authors suggested that the typical first mode damping ratios for cables are in the range of 0.05% to 0.9%. Similar data for control of inclined cable galloping is also provided. Incidences of large amplitude cable vibrations have also been reported when there is no rain, and typically at higher wind speeds. There is debate and uncertainty regarding the exact nature of all of the events that fall under this category of vibrations. It is known that cable vibrations can occur when deck or tower vibrations are occurring at frequencies close to the cable frequency (Stubler et al. 1999; Wu et al. 2003). This is also called “parametric vibrations” or “local parametric vibrations” by some investigators. Wu et al. (2003) reported that parametric vibration has been confirmed on three bridges in Japan, including the Tatara Bridge. Irwin (1997) discussed the possibility of inclined cable galloping based on the work of Saito et al. (1994) in Japan. Although circular cross sec- tions do not gallop when aligned normal to wind (Starossek 1994), Irwin provides a possible explanation in that the wind would “see” an inclined cable as an elliptical section, and thus be able to gallop. This phenomenon has been investigated in wind tunnel tests and it was determined that separate require- ments to address this phenomenon are not necessary. Until recently, there were no vibrations reported on the Sunshine Skyway Bridge, which joins St. Petersburg and S m Dc = ζ ρ ≥ 10 Eq. 1 FIGURE 40 Large amplitude vibrations of the Cochrane Bridge (Alabama) (Telang et al. 2000). Top Rivulet Wind Bottom Rivulet FIGURE 41 Position and movements of water rivulets during rain–wind vibrations.

30 Bradenton in Florida, whether rain–wind or otherwise. This bridge was opened to traffic in 1987, has grouted parallel strand cables with steel sheathing, and has two-dimensional viscous dampers (shock absorbers) installed on each cable. On April 12, 2004, Florida DOT personnel noted small- amplitude vibrations on one of the longest cables of the bridge (personal communication, S.D. Womble, April 14, 2004). It was reported that, according to National Oceanic and Atmo- spheric Administration records, a sustained wind of 72 kph (45 mph) and gusts of up to 96 kph (60 mph) were present in the area. Wind was blowing at 90 degrees to the struc- tures (perpendicular to cable plane). There was no rain, and estimated vibration amplitudes of up to 75 mm (3 in.) were reported. It should be noted that the reported amplitudes in this case are far smaller than amplitudes typically reported for rain–wind vibrations in other bridges. Hartman Stay AS16 (Dominant mode indicated by symbol) (b) Hartman Stay AS16 (5-minute RMS in-plane acceleration > 0.25 g) (a) Mode 87654 0 20 40 60 80 100 120 321N um be r o f o cc ur re nc es as do m in an t m od e 0 2 4 6 8 10 12 14 0 0.5 1 1.5 2 2.5 3 Deck-level wind speed – 5-minute mean (m/s) In -p la ne a cc el er at io n – 5- m in u te R M S (g ) mode 2 mode 3 mode 4 mode 5 mode 7 FIGURE 42 Vibration data from Fred Hartman Bridge: (a) histogram of modes, (b) vibration amplitudes (Main and Jones 2000).

31 Vortex excitation is likely the most common form of cable vibration, with the cables vibrating at lower displacement amplitudes and higher frequencies (mode 5 and higher) (Main and Jones 2001). Therefore, this mode of vibration is not as significant a risk to stay cables as rain–wind vibrations. Vortex-induced vibrations have been noted on the Tatara Bridge in Japan (Yamaguchi et al. 1999). When cables are positioned in the wake of towers or other cables, they can have large amplitude wake galloping vibra- tions. However, the wake galloping that could occur in stay cables is typically characterized by very small cable spacing, on the order of six cable diameters (Miyazaki 1999). Bruce et al. (1987) reported on the aerodynamic monitor- ing of the Luling Bridge in Louisiana 3 years after the open- ing of bridge to traffic. They noted vortex shedding and wake-induced effects. However, they also reported a first mode response of stays to “either galloping or bridge deck motion.” The responses to the questionnaire indicated that a sizable number of cable-stayed bridges included in the survey have experienced rain–wind vibrations. These bridges are the Cochrane Bridge in Alabama; Talmadge Memorial over the Savannah River in Georgia; Clark in Alton, Illinois; Burlington in Iowa; Veterans Memorial between Bridge City and Port Arthur in Texas; and Fred Hartman in Houston, Texas. In Canada, the Prince’s Island and Fox Hollow bridges (Alberta, Calgary), and the Hawkshaw, Longs Creek #1, and Nackawic River bridges (New Brunswick) have reportedly been affected. Figure 44 shows the results of the survey as related to rain–wind-induced cable vibrations. It is interesting Damping Ratio (Percent) Frequency Cumulative Percentage 0 50 100 150 200 250 300 350 400 0% 20% 40% 60% 80% 100% 120% 0.2 0.1 0.4 0.5 0.7 0.8 1.0 1.1 1.3 1.4 1.6 1.7 1.9 2.0 N um be r o f C ab le s Cu m u la tiv e Pe rc en ta ge FIGURE 43 Histogram of required damping ratio for controlling rain–wind vibrations (Tabatabai and Mehrabi 2000). 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 yes no not known not applicable no answer Rain-Wind Vibrations Pe rc en t o f B rid ge s U.S. Canada FIGURE 44 Rain–wind induced cable vibrations.

32 to note that the survey response for the Fox Hollow Bridge indicated rain–wind vibrations even though the stay cable is made of exposed threadbars, which would not likely promote the formation of water rivulets. It may be that, in this case, other vibration types have been mischaracterized as rain–wind. Figures 45 and 46 show survey results with respect to the use of viscous dampers and cross cables. It appears that the most popular method of vibration control is the use of cross cables. Nearly one-third of the bridges in the United States and about one-quarter of the bridges in Canada have cross cables for vibration control, either installed from the beginning or retrofitted later to control vibrations. Viscous dampers are used in the United States by six bridges (21.5%) and in Canada by three bridges (23.1%). In some bridges such as the Fred Hartman Bridge in Texas, both viscous dampers and cross cables are added (retrofitted) to control vibrations. In the Cooper River Bridge in South Car- olina, viscous dampers will be installed, but provisions for future installation of cross cables are made in case they are needed. Figure 47 shows the survey results with respect to the use of other types of dampers. Extreme Events There are a number of extreme or unusual events that could affect the performance of stay cables including earthquakes, fire, blasts, impacts, and ice build-up. The earthquake design issues are generally handled through a global analysis of the entire cable-stayed bridge. However, during the fall 2004 meeting of the PTI cable-stayed bridge committee, David Goodyear noted that there potentially are cases when during an earthquake the tension force in a cable can rapidly decrease to zero or even compression. This impact loading, in a direc- tion opposite to how the cable anchorage is designed to resist may result in permanent dislocation and damage to some cru- cial anchorage components, potentially rendering them in- effective and resulting in failures. Specifically, wedge sys- tems could be affected where there is no significant resistance to forces that would push the wedges out of their positions within the anchorage plates. This issue may be considered by the stay cable community and studied further. However, there have not been any reported cases where this scenario has materialized. During an oral presentation at a stay cable seminar, Zoli and McCabe (2004) reported on issues related to fire, ice, and impact on stay cables. They reported that there have not been major fire incidents involving cable-stayed bridges. However, six major Interstate highway fires have occurred, resulting in significant cost and extended closures of major arteries. Zoli and McCabe suggest that wedge anchorage systems would be more resistant than some other anchor- ages. Zinc-filled sockets are temperature sensitive and con- tain materials with low melting points. Possible mitigation measures include utilization of fire-resistant cable sheathing near deck level, intumescent paints, ablative coatings, ceram- ics and composites. According to Zoli and McCabe, there are currently no code provisions in the United States address- ing fires on bridges, although the Eurocode includes some provisions. Regarding the effects of icing on cables, Zoli and McCabe noted that ice formations on a major suspension bridge have been periodically removed as a safety precaution. They reported on research being done on the issue of icing of cables, including assessments of sheathing performance and icing wind tunnel tests. The effects of icing on galloping vibra- tions of stay cables need to be studied. They discussed “ice- phobic” coatings and ultrasonic deicing systems. Regarding impact, Zoli and McCabe discussed a number of approaches including cable “armoring” involving hybrid ceramic FRP materials. Inspectability and Maintainability Question 11 in the survey questionnaire asked agencies the following: what do you see as the single most important prob- 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 yes - from the beginning yes - retrofitted to correct vibrations no no answer Viscous Dampers Pe rc en t o f B rid ge s U.S. Canada FIGURE 45 Percentage of bridges using viscous dampers.

33 lem in stay cable maintenance? The great majority of answers mentioned accessibility and inspection problems, especially in the anchorage areas. The general consensus of the respondents points to a concern by the owners about difficulties in access for inspections and maintenance. It should be noted that although there is general agreement by the stay cable commu- nity about the need to address the stay cable maintenance issue and problems, there is no universal consensus on this issue, especially with the characterization of the subject as a “prob- lem” as indicated by one respondent. The following are some of the answers provided by respondents: • Access and rain–wind induced oscillation. • Access to upper anchorage. • Inspection and condition evaluation of anchorages. • Effective corrosion barriers that do not interfere with the ability to adequately inspect and assess the health of the cable stay system on a regular interval and within practi- cal means. • Accessibility for inspection and maintenance. • Access to the cable anchorages. • Uncertainty of cable condition and anchorages. • Inspection, access, testing, and cost. • Inability to inspect the elements inside the cable and anchorage areas. • Inspecting the cable anchors and grout-filled cables. • Hidden nature of the system. • Access for inspection. • Integrity of the stays. Grouted cables are impossible to inspect with a nondestructive technique (i.e., one that does not require removal of sheathing and grout); therefore, it is impossible to identify corrosion prob- lems early. • The largest “problem” with stay cables is that they are widely perceived of as “a problem” rather than just another bridge member with specific needs and charac- teristics. Stay cables have been placed unnecessarily “on a pedestal.” Although they are a very important bridge member, in current designs they are highly redundant, overtested, and (relatively) easily replaced. There is no other major bridge member that fits into all three of these categories. Let us not promote the feeling that stays are “a maintenance problem.” • Provide end caps that are easily removed and fully pro- tected against corrosion. • Ability to determine the effectiveness and remaining life of corrosion protection systems for main tension elements. The configuration and construction techniques make evaluation and inspection using nondestructive techniques almost impossible. 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 yes - tuned mass dampers yes - other dampers no not known no answer Other Dampers Pe rc en t o f B rid ge s U.S. Canada FIGURE 47 Percentage of bridges using other types of dampers. 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 yes - from the beginning yes - retrofitted to correct vibrations no no answer Cross Cables Pe rc en t o f B rid ge s U.S. Canada FIGURE 46 Percentage of bridges using cross cables.

0.0 20.0 40.0 60.0 80.0 100.0 yes no not known Methods Effective? Pe rc en t o f B rid ge s U.S. Canada FIGURE 49 Respondents reporting that inspection, testing, monitoring, and repairs are effective and adequate. • Lack of familiarity with this type of construction by the department’s staff, which requires assistance from the consultant community in the inspection of these elements. • Cannot inspect cables without pulling strand every 10 years. • Access for inspection and actual testing. • Detection of corrosion in cables; maintenance of sheaths and boots. • Corrosion at the anchorages. • Migration of water into cable strands. • Fatigue. • Inspectability in the anchorage area. The anchorages are typically unreachable except from the deck by special “reach-all” trucks (see Figure 48). Some newer bridges (such as the Cooper River Bridge) incorporate anchor- ages that are at about deck level. The end caps are generally difficult to remove, especially when filled with grout or epoxy. Even when the end caps are removed, the condition of MTEs within the anchorage area and beyond cannot be examined visually. If moisture were to enter the cable along its length, gravity would likely force it down to the bottom anchorage. There is currently no easy way to check for the presence of moisture or corrosion in the bottom anchorage, except through removal of the cap. Massive reinforced concrete or steel superstructure elements that are designed to resist anchorage forces typically surround the anchorage zones. Therefore, the sides of the anchorage zones are generally neither visible nor accessible all the way up to the top of the neoprene rings and boots. Some recent anchorage designs (such as the 6th Street bridges in Wisconsin) have incorporated individually coated and sheathed strands that reportedly allow for future replace- ments of individual strands (one by one). Some recent bridges also include additional strands in the cables that are designed for removal at 10 to 15 year intervals for inspection. In some cases, allowance is made in the cables to add new strands, if needed. Permanent access platforms for use by inspectors are also an important consideration. Question 5 in the survey asked whether the current inspec- tion, testing, monitoring, and repair methods were effective 34 and adequate. Figure 49 summarizes these responses. The U.S. respondents were far less certain than their Canadian counterparts, with less than 40% believing that they have effective and adequate methods available. One of the respon- dents indicated that for cables with steel sheathing the inspection methods available are limited. Another respon- dent referred to problems in inspection of anchorage areas and expressed the need for a technological breakthrough to address this problem. One important question in the maintainability of a cable- stayed bridge is whether the cable (or individual strands) can FIGURE 48 Access to cable bottom anchorage for ultrasonic testing.

be replaced, if needed. In the opinion of respondents for 79% of U.S. bridges and 62% of Canadian bridges, the answer to this question is “yes.” Figure 50 summarizes the survey results for this question. Another question in the survey asked whether there is an inspection and maintenance manual for the bridge. Figure 51 shows the results of the survey for this question. The great majority of U.S. bridges (71%) have maintenance manuals; however, more than 92% of Canadian bridges do not. As will be discussed later, there is a wide variation in topics discussed in individual maintenance manuals. Survey question 10 asked whether an up-to-date resource such as a national database of information on stay cable inspection, repairs, and testing would be a useful tool. Fig- ure 52 summarizes the responses. An overwhelming major- ity of responses (approximately 90%) in both the United States and Canada responded in the affirmative. There was a wide variety of answers provided to the survey question on what the cable suppliers should incorporate into their systems to make them accessible and inspectable. The following are some of the suggestions: • Transparent outer pipe, eliminate grout. • Current grouted and sheathed systems do not allow for visual inspection. New stay systems (perhaps ungrouted, unsheathed systems consisting of bare corrosion- resistant tension members) need to be developed that allow for inspection of the entire stay length. Research is also needed to develop rapid, economical nondestruc- tive evaluation (NDE) methods to determine conditions of stay cables. • Access is a very sharp two-edged sword. If you can more easily access the cable, so can corrosive elements (not to mention potential terrorist/security considerations). • Include a maintenance manual with clear instructions for both specific wires or full cables. • Perhaps a permanent load cell that would permit real- time readings of cable forces at any time during the life of the bridge. • Our cables are reasonably accessible, inspectable. Pos- sibly a closeable drain at the lower end of the cable to allow visual inspection, sample collection, testing for corrosion product of any water in the cable sheaths. • Different corrosion protection system at the anchorages that permits easier visual inspection. Removable sec- tions of the HDPE and Vandal Tubes would make it easier to inspect strands near the anchorages. 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 yes no not known no answer Replaceable? Pe rc en t o f B rid ge s U.S. Canada FIGURE 50 Can the cables (or strands) be replaced? 35 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 yes no not known no answer Maintenance Manual? Pe rc en t o f B rid ge s U.S. Canada FIGURE 51 Bridges with an inspection and maintenance manual available.

36 • I like the idea of hermetically sealed, ungrouted cables with fiber optic sensors throughout or exposed wire that can be directly inspected. • Provide access on inside and outside of tower anchorage. • Fiber optic strain gauges and redundant systems. • We would recommend the development of individual strand monitoring capabilities that encompass the strands from anchorage to anchorage. • For non-box bridges, an inspection traveler should be in- stalled on the cable-stayed bridge. This should be done by the owner. • Not possible that the cable suppliers can do any more. • Ability to detension, inspect, and retension individual strands; ability to add strands to each cable or cable group (5%). FABRICATION AND ERECTION OF STAY CABLES As discussed in the previous section, stay cables can be either shop- or field-fabricated. In United States practice, and espe- cially in recent years, field fabrication has been more com- mon. The shop-fabricated cables come with the entire cable, including anchorages and sheathing pre-assembled and coiled on reels. The cables are then uncoiled and lifted into place. If fillers are required, they are generally placed or injected after stressing of cables. In this arrangement, the entire cable must be stressed with one large hydraulic jack. The field- fabrication method generally involves inserting strands one by one into the wedge plate in the bottom anchorage and, through various methods developed by cable suppliers, the strands are pulled through the top wedge plate. The strands are typically stressed one at a time using a single-strand jack. If required, the cable fillers (e.g., grout) would then be injected into the anchorage zones (in the case of bond socket) or the entire cable. 0.0 20.0 40.0 60.0 80.0 100.0 yes no not known Database Useful? Pe rc en t o f B rid ge s U.S. Canada FIGURE 52 Would a national database of stay cable information be useful?

Next: Chapter Three - Inspection and Monitoring Techniques »
Inspection and Maintenance of Bridge Stay Cable Systems Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s National Cooperative Highway Research Program (NCHRP) Synthesis 353: Inspection and Maintenance of Bridge Stay Cable Systems identifies and explains various inspection and maintenance techniques for bridge stay cable systems. It discusses both short- and long-term approaches. The report information on methods for inspections and assessments, including nondestructive testing and evaluation procedures; repair and retrofit; methods for control of cable vibrations, including rain–wind vibrations; stay cable fatigue and failure; effectiveness of various inspection and repair methods; limitations of available technologies; and trends and recommendations for future study.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!