Click for next page ( 8

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement

Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 7
7 CHAPTER TWO BACKGROUND DEVELOPMENT OF SPECIFICATIONS FOR STEEL BRIDGES In 1967, the Point Pleasant Bridge over the Ohio River (con- structed in 1928 and also known as the Silver Bridge) col- lapsed, resulting in 46 deaths (Figure 1). The collapse was the result of a brittle fracture of one of the nonredundant eyebars supporting the main span (57). As discussed later in this chapter, although there is disagreement about what should be classified as an FCM, there is no doubt that this eyebar was an FCM. There are several reasons why this catastrophe was extraordinary and is not likely to be repeated. The small flaw in the eyebar may have been caused by stress-corrosion crack- ing (SCC) (5,8), which is discussed further in Appendix A. Stress-corrosion cracking should not occur in modern bridge steel; however, in this instance, the eyebar steel was 1928 vintage, heat-treated AISI 1060 steel, which is substantially different than today's steel. The fracture toughness of the FIGURE 1 Collapse of Point Pleasant Bridge. eyebar was also marginal and a relatively small crack led to the brittle fracture of the eyebar, which in turn led to the collapse of the bridge. This collapse was the catalyst for many In addition to CVN requirements, these provisions restrict changes in material specifications, design, fabrication, shop the choice of details as well as control weld flaws and other inspection, in-service inspection, and maintenance of steel crack-like defects. These provisions have reshaped industry bridges. practices and result in an acceptably low probability of fatigue cracking and brittle fracture in new bridges. Material, Design, and Fabrication Specifications, However, many older steel bridges built before the imple- and Effect of Bridge Design Date mentation of modern fatigue design provisions in the mid- 1970s possess poor fatigue details, such as cover plates that In 1974, in part as a result of the Point Pleasant Bridge col- can develop fatigue cracks (Figure 2) (11), which if not lapse, mandatory Charpy V-notch (CVN) toughness require- repaired, can grow and lead to fracture of the member and ments were initiated for welds and base metal to ensure ade- possible collapse of part or all of the bridge. quate resistance to fracture; that is, fracture toughness (9,10). The greater the CVN at a particular temperature, or the lower Other factors that make these older bridges susceptible to the temperature at which the CVN is required, the larger the fracture include: critical crack size that can be tolerated at lowest anticipated service temperature without fracture. The present CVN Marginal fracture toughness of the steel and weld metal; requirements (for non-FCM) are essentially the same as the Detailing, fabrication quality, and shop inspection below CVN requirements implemented in 1974. The CVN require- modern standards; ments were the result of significant debate and some com- Severe corrosion problems, especially at open or failing promise during their development, which is discussed fur- expansion joints; ther in Appendix A. Higher traffic volumes and truck weights than the bridge was originally designed to handle. Presently, material selection, design, and fabrication of steel bridges are governed by In light of these factors, periodic in-service inspection is particularly important for older bridges to provide an oppor- AASHTO LRFD Bridge Design Specifications (1) and tunity to detect cracks and corrosion before they grow to a AASHTO/AWS-D1.5, Bridge Welding Code (2). critical size. In 1970, partly in reaction to the collapse of the

OCR for page 7
8 Concrete Deck Girder Flange Crack Crossframe Girder Web FIGURE 2 Development of fatigue crack at cover plate ends on FIGURE 3 Typical web-gap fatigue cracking. the multibeam Yellow Mill Pond Bridge in Connecticut in 1976. (Courtesy: John W. Fisher.) rected in 1985 by a change in AASHTO specifications that mandated the attachment of the connection plate to both Point Pleasant Bridge over the Ohio River, the NBIS (4) were flanges. established. Title 23, Code of Federal Regulations, Part 650, Subpart C sets forth the NBIS for all bridges of more than Hence, it is important to distinguish three different age 20 ft (6 m) span on all public roads. Section 650.3 specifies ranges of steel bridges: inspection procedures and frequencies, indicates minimum qualifications for personnel, and states reporting, inventory, 1. Steel bridges built before the implementation of modern load posting, and inspection recordkeeping requirements. The fatigue design provisions in the mid-1970s. current NBIS mandates a 2-year inspection interval for all 2. Steel bridges designed after the mid-1970s, but before highway bridges carrying public roads. 1985, which have fewer fatigue problems but remain susceptible to distortion-induced fatigue. However, modern steel bridges are not nearly as suscep- tible to fracture as older bridges (12,13). As a result, the ways 3. Modern steel bridges designed after 1985 that should modern bridges are managed could possibly be evaluated not be susceptible to fatigue at all. differently than older bridges. This could be studied further, with considerable potential benefits. Fatigue is virtually unheard of in modern steel bridges as a result of improved design specifications, more fatigue- For example, problems with severe corrosion have been resistant details, and improvements in shop inspection reduced. In the last 20 years, durability of weathering steel (15,16). Fatigue problems that have occurred in modern and coating systems has improved. Expansion joints have bridges were typically the result of unintended behavior been improved if not eliminated through the use of continu- or a design error that is not consistent with the intent of ous jointless bridges (14). present specifications and usually manifest in the first few years of the life of the bridge. Although the bridges that In addition, there have been few if any cases where weld are referred to herein as "modern" are less than 20 years defects or low-toughness steel has been an issue for modern old, there is substantial confidence that these structures will steel bridges, owing primarily to improvements in details, continue to perform with few problems resulting from fabrication practices, and fracture toughness of the steel and fatigue. weld metal (12,15). If spontaneous fracture from weld defects is ruled out, then fracture can only occur if preceded by As a final note, in changing from load factor design (LFD) fatigue (15). Therefore, in this case, it is essentially sufficient to LRFD, two-girder bridges will be designed more con- to control fatigue to prevent fracture (12,15). servatively than before. According to Dr. Dennis Mertz, in calibrating the LRFD specifications, loads were increased Distortion-induced fatigue cracking, discussed further slightly to compensate for improved and less conservative in Appendix A, continued as a fatigue problem in typical distribution factors. However, the distribution factors for plate girder bridges designed before the mid-1980s (11,12, two-girder bridges were always reasonably accurate, so they 15,16). A common example of distortion-induced fatigue did not get the benefit of improved distribution factors that cracking is web-gap cracking, which occurs in the gap when multigirder bridges did. This should be kept in mind when a connection plate is not attached to a flange and is subject considering the reliability of these two-girder and two-line to out-of-plane distortion (Figure 3). This problem was cor- truss systems.

OCR for page 7
9 Additional Material, Fabrication, and In-Service designed before implementation of the FCP in 1978 (22). Inspection Requirements for FCMs and Cost Impact The survey (described in chapter three) indicated that approx- imately 75% of FCBs in present inventory were designed FHWA led the development of a fracture control plan (FCP) before the FCP. to provide a higher level of safety for FCBs. In the broad sense, an FCP includes everything that affects the potential However, the fabrication provisions and the CVN require- for fracture--in-service inspection and maintenance as well ments for the materials of the fabrication FCP increase costs. as design, fabrication, and shop inspection (17). The idea is One major bridge fabricator reported that the approximate that trade-offs can be made between components of the plan increase in initial costs for new FCMs relative to non-FCMs without compromising reliability. For example, better tough- is on the order of 8% of the cost of fabricated steel. ness could be required to compensate for relaxed in-service inspection standards, because better toughness would lead In 1983, the Mianus River Bridge on I-95 in Connecticut to a larger critical crack size that would be easier to see from (built in 1957) collapsed, killing three persons (Figure 4). a distance and that would take longer to develop. Efforts to Packout corrosion in a nonredundant pin and hanger assem- make FCBs more conservative were largely the result of bly pushed one of the plates partly off the pin, eventually experiences with cracking in tied arches, mostly owing to leading to a fatigue crack and collapse of the suspended span large fabrication defects because of difficulty associated with (between two cantilevers) (16,25,26). This event can be fur- welding A514 steel (11). ther attributed to poor maintenance, because a clogged drain was partly responsible for the packout corrosion. Through- The American Iron and Steel Institute initiated a research out the country there are numerous other bridges with similar project to develop an improved FCP for fabrication of non- pin and hanger details; however, this type of suspended span redundant structures. This work (9,1721) ultimately resulted is rarely if ever used in new designs. As with the eyebar of the in the 1978 publication of the AASHTO Guide Specification Point Pleasant Bridge, this bridge collapse demonstrated that for Fracture Critical Non-Redundant Bridge Members (22). these pin and hanger assemblies are also clearly FCMs. (Also A key element was more stringent CVN requirements for base similar to the Point Pleasant Bridge collapse, the Mianus metal and weld metal for FCMs. River Bridge collapse was the result of extraordinary circum- stances, in this case corrosion, and not just fatigue or fracture.) This Guide Specification has now been withdrawn. The CVN requirements for base metal, including the greater In part because of this failure, the NBIS were revised in requirements for FCMs, are now included in the AASHTO 1988, requiring among other things a hands-on inspection of LRFD Bridge Design Specifications (1), as well as the ASTM FCMs. This requirement significantly increases life-cycle and AASHTO specifications for the steel (23,24). Most of costs relative to non-FCMs, which may be inspected from the remaining material from the Guide Specification is now the ground, through, in most cases, binoculars (2729). This included in Section 12 of AASHTO/AWS-D1.5, "AASHTO/ requirement is particularly onerous for box girders, because AWS Fracture Control Plan (FCP) for Non-Redundant it requires the inspectors to enter the boxes, which signifi- Members" (2). Note that in AASHTO/AWS-D1.5 the defini- cantly increases costs. The frequency and extent of inspec- tion of an FCP is narrower, including only fabrication and tion are not clear in the current NBIS. Consequently, there is shop inspection--not base metal selection, in-service inspec- disagreement on what constitutes a fracture-critical inspec- tion, and maintenance. Therefore, these provisions will be tion and how often it is done. (When the word inspection is referred to as a fabrication FCP. (Unfortunately, yet another completely different meaning for the term "fracture control plan" has arisen and that is the plan- or elevation-view draw- ing identifying all FCMs for use in in-service inspection.) The differences between the provisions for the fabrication of FCMs in Section 12 and the provisions for non-FCMs elsewhere in AASHTO/AWS-D1.5 are primarily more strict fabrication and shop inspection requirements to control weld flaws and other crack-like defects in FCMs. For example, transverse groove welds are required to be inspected in the shop with both radiographic testing (RT) and ultrasonic test- ing (UT), whereas only RT is required for non-FCMs. The fabrication FCP and the more stringent CVN require- ments result in an even lower probability of brittle fracture in new FCMs than for typical non-FCM members. Note that this FIGURE 4 Collapse of Mianus River Bridge. (Courtesy: John additional fracture reliability does not apply to older FCMs W. Fisher.)

OCR for page 7
10 used throughout the rest of this report, it is intended to mean fracture-critical, hands-on field inspection, unless otherwise noted.) These increased life-cycle costs for FCMs are signifi- cantly greater than the approximately 8% increase in initial materials and fabrication costs for FCMs discussed earlier. According to the survey (and as described later in chapter three), many owners believe that inspection costs associated with FCBs consume a large portion of their entire inspection budget. Owners were asked to estimate the relative increase in costs when inspecting an FCB relative to inspection and non-FCBs. There was substantial variation in the response; however, most agencies indicated increases of between 200% and 500%. The most common reasons indicated for these FIGURE 6 Hands-on inspection from manlift. increases were: Specialized access equipment such as a snooper (Fig- Primarily because of these increased life-cycle costs, there ure 5), manlift (Figure 6), or rigging required for hands- is a general reluctance to design new FCBs. Fewer FCBs have on inspection (30). been proposed since the fabrication FCP went into effect in Traffic control to close lanes to permit the access equip- 1978 (22). FCMs, such as steel pier caps and cross girders, are ment to be placed on or below the bridge (see Figure 5). still frequently designed, although usually only if they cannot Additional employee-hours required to conduct a detailed be avoided. In some circumstances, bridge designs with FCMs, hands-on inspection. such as tied arches, two-girder bridges, and trusses, may be the More frequent use of nondestructive testing (NDT) most efficient and cost-effective structural system. Although (described in Appendix A). the more stringent CVN requirements, the fabrication FCP, Greater frequency of inspection for FCBs. and the additional inspection requirements for FCMs are beneficial, if they are overly conservative for modern bridges These hands-on inspections have revealed numerous fatigue they can become an obstacle to the savings gained in using and corrosion problems that otherwise might have escaped more cost-effective designs. notice. Many of these problem details are discussed in Appen- dix A. Twenty-three percent of respondents to the survey indi- International scanning tours for bridge management (32) cated that they found significant cracks that could have become and fabrication (33) have noted that Europe does not have much worse, possibly averting collapses (see chapter three). special policies for FCMs. A risk-based approach, coupled Similar examples may be found in trade magazines [e.g., see with more rigorous three-dimensional analysis techniques, is Zettler (31)]. used to ensure that a sufficient level of structural reliability is provided. Consequently, steel bridge designs that would be considered fracture-critical in the United States are still com- monly built without prejudice. However, they have also had failures of what we would consider FCBs. The following is from the fabrication scanning tour report (33): Perhaps the most significant design-related observation of the scan team was the rest of the industrialized world's more liberal view of the importance of redundancy. Two-girder bridges, as well as other structure types considered nonredundant and fracture critical in the United States, are not discouraged and, in fact, are used extensively as safe and cost-effective bridge designs. Kawada Industries cited redundancy studies it performed to demonstrate adequate redundancy of its two-girder systems with widely spaced, mid-depth cross beams. No special design, fabrication, or inspection requirements for such bridges were apparent. The U.S. design philosophy for nonredundant bridges should be reconsidered, based upon these observations and improvements in steel toughness. FIGURE 5 Snooper used for hands-on inspection from bridge Twin-girder railroad bridges are common in Germany. The deck. single-cell box girder, commonly used for elevated roadways in