National Academies Press: OpenBook

Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems (2023)

Chapter: CHAPTER 2 Literature Review and Synthesis

« Previous: CHAPTER 1 Introduction
Page 5
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 5
Page 6
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 6
Page 7
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 7
Page 8
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 8
Page 9
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 9
Page 10
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 10
Page 11
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 11
Page 12
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 12
Page 13
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 13
Page 14
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 14
Page 15
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 15
Page 16
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 16
Page 17
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 17
Page 18
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 18
Page 19
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 19
Page 20
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 20
Page 21
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 21
Page 22
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 22
Page 23
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 23
Page 24
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 24
Page 25
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 25
Page 26
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 26
Page 27
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 27
Page 28
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 28
Page 29
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 29
Page 30
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 30
Page 31
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 31
Page 32
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 32
Page 33
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 33
Page 34
Suggested Citation:"CHAPTER 2 Literature Review and Synthesis." National Academies of Sciences, Engineering, and Medicine. 2023. Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems. Washington, DC: The National Academies Press. doi: 10.17226/26861.
×
Page 34

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.

5 C H A P T E R 2 Literature Review and Synthesis Literature Review Purpose of Literature Review NCHRP 534 represents the current guidelines followed by bridge owners for planning and conducting inspections of main cables of suspension bridges and for estimating their remaining strength. The literature review for the present project used this previous NCHRP effort as a starting point. Rather than repeating all the information gathered during the development of NCHRP 534, attention was focused on the developments that have occurred on inspection methods and assessment of the remaining cable strength of main cables of suspension bridges from the publication of the NCHRP 534 report to the present. An effort was made to highlight the knowledge gaps between state of the-art and current guidelines. Literature Review Process The literature review effort started from the body of work contained in the NCHRP 534 Guidelines and built on it, looking at the published work produced from 2004 until the present. The review was conducted using the library search engines of Columbia University together with inspection reports and studies available to the members of the research team. The search focused on international peer- reviewed journals like the American Society of Civil Engineers (ASCE) Journal of Bridge Engineering and the ASCE Journal of Structural Engineering, as well as bridge-specific conference proceedings. Rather than including it here, an annotated bibliography, e.g. the full list of literature reviewed, along with a summary of each document, is provided in Appendix A. With regard to the assessment of the main cable's strength, it appears that, from the results of the literature review, the vast majority of the published work has been done by members of this research team. The focal point appears to be the 2007 paper by Shi, Deodatis and Betti where a random field methodology was introduced as a more rigorous and accurate alternative to the one presented in the NCHRP 534 Guidelines. Although not officially applied in the field, this methodology has been published in peer-reviewed journals and validated by comparison to currently used methods. It has been used in various publications to study a variety of related problems (e.g. inter-wire friction, time- dependent corrosion rate, thermal effects, etc.). It represents the springboard for the development of the new methodology proposed in this study. The literature search revealed the existence of an additional method for computing the remaining strength of a cable: it is the BTC method, a proprietary method proposed by Bridge Technology Consulting. With regard to cable inspections, few studies were found where Nondestructive Techniques (NDT) were used in the inspection of main cables. The FHWA Report (FHWA-HRT-14-023), produced by

6 members of this research team, contains an up-to-date review of NDT technologies, e.g. Acoustic Emission, Magnetostrictive, Flux Leakage, used in practice and research and was used as the starting point for the work done in this study. Identification of Knowledge Gaps Following the general approach presented in NCHRP NCHRP Report 534, the tasks required for a proper assessment of the remaining strength of a parallel wire cable involve the following steps: 1) Inspecting and mapping the corrosion grades across the cross-section of the cable, 2) Extracting wire samples from the various corrosion grades, 3) Developing wire stress-strain diagrams using tests of extracted wires, 4) Finding the minimum strength of a given length of wire, 5) Calculating probabilistic characteristics describing how the strength of wires vary within the cross-section and along the length of the cable, 6) Estimating the number of cracked and broken wires within a panel, 7) Determining the effectiveness of the cable bands in redeveloping the strength of a wire broken at some distance from the point at which the strength is being determined, 8) Estimating the effect on the cable strength of a given panel of the deterioration in panels near the one being evaluated, 9) Estimating the strength of the cable in a given panel based on this data. While Steps 1 through 3 are well defined and regulated by precise procedural standards, Steps 4) and 5) require some clarifications. Steps 4) and 5) involve the description of a wire "of a given length". If this "given length of the wire" is made of n unit segments (specimens), the strength of such a length of wire can be calculated to be the minimum of the n strengths of the n unit segments (weakest link model). This computation can be done either by using some extreme value distribution (exact or asymptotic), or through equivalent Monte Carlo simulations. In most of the publications on this topic, we found that there is an inherent assumption made in this standard procedure: the n random variables that model the strength of the n successive unit length segments along the wire "of a given length" are independent (e.g. the corresponding n strengths are uncorrelated). This is an assumption that it is acceptable when dealing with brittle and quasi-brittle materials but it is highly questionable and potentially wrong for a ductile material like steel. It is with the pivotal paper by Shi et al. (2007) that the spatial correlation of the wire strength along its length was introduced in the estimation of the strength of suspension bridge cables. In the approach proposed by Shi et al. (2007), the strength of a wire of length equal to n unit length segments is modeled as a Random Field (e.g. as n correlated random variables). This approach was successfully applied to the estimation of the remaining cable strength of the Williamsburg Bridge (New York City, NY). From the literature review, we found that there are four basic variations of the standard approach, based on four different ways to estimate the strength of a wire consisting of n unit length segments: 1) using the exact extreme value distribution (EVD) of the smallest value of the strength (Exact EVD), 2) using Type I asymptotic distribution of the smallest value (Type I EVD), 3) using a Type Ill asymptotic distribution of the smallest value (Type Ill EVD, also known as the Weibull distribution), and

7 4) using Monte Carlo simulations reflecting the EVD (simulating independent and identically distributed random variables). This last approach is basically equivalent to numerically computing the results of the Exact EVD approach: it requires that: 1) the wire "of a given length" be modeled as n uncorrelated random variables, each describing the strength of a unit length wire segment, 2) select an initial distribution (e.g. lognormal, beta) of the strength of a unit length segment, 3) generate n unit length segment strengths from the selected initial distribution, 4) determine the smallest strength among the n generated strengths and this will be the strength of the wire "of a given length", and 5) repeat this procedure a large number of times to generate the statistics of the strength of a wire "of a given length". All the previous studies and guidelines for assessing the remaining strengths of suspension bridge cables have been based on such methodologies. In 1988, Steinman et al. attempted to estimate the strength of the Williamsburg Bridge main cable by assuming, as the initial distribution for the strength of 1-ft-long wire segment, a Gaussian distribution and then used the assumption of n independent and identically distributed random variables to estimate the strength of a wire of length equal to n units. The shortcoming of this approach was the use of a Gaussian distribution, since it allows the strength of a unit length segment to assume negative values, something that is physically impossible. However, while from a purely theoretical point of view, only distributions that assume nonnegative values should be considered, the Gaussian distribution provides reasonably good estimates (the left tail of the Gaussian distribution becomes negligibly thin for negative values of the strength). Following the same approach, Matteo et al. (1994) used the Gaussian distribution as the initial distribution and then used a Type I EVD approach to find the strength of a wire of a given length. A similar approach based on Type I EVD was used by Haight et al. (1997) to assess the remaining strength of main cables for four suspension bridges. It was in 1998 that Perry introduced the Type Ill EVD model for the reliability analysis of the Williamsburg Bridge cables and this type model, together with the Type I EVD, was set into standard in the NCHRP NCHRP Report 534 (2004). It should be mentioned that from a practical point of view, the aforementioned two distributions (Type I and Type Ill) yield almost identical results. It was Shi et al. (2007) that established a new approach in the way cable strength should be calculated. Looking at the laboratory results from the testing of 330 wire segments, they found that the probability distribution function that best fits the data is a beta probability distribution function (limited to positive values only). In addition, having the sequential order of the segments along the long samples removed from the bridge cable, they were able to estimate an autocorrelation function for the wire strength as a function of the separation distance along the length of the wire. A Monte Carlo approach was then used to generate a large number of realizations from which it was possible to obtain the mean and the standard deviation of the cable's strength. This approach, by accounting for the correlation of the wire strength along the length of the cable and its inherent non-Gaussianity, provided estimations of the mean strength of the cable that were about 10%-12% higher than those obtained from the previous methods. Estimating the number of cracked and broken wires in a cable has a pronounced impact on the estimation of cable strength. Very little information is contained in the current literature about how to

8 establish the percentage of cracked and broken wires. The methodology contained in NCHRP 534 is, at present, the only approach that is being used. A review of inspection reports indicates that little data has been collected as to the observed distribution of cracked and broken wires. Developing an inspection procedure and statistical methodology for a more reliable method of establishing these parameters and how they impact cable strength was a goal of this project. In computing the remaining strength of a parallel wire cable, an important parameter is represented by the redevelopment length of a broken wire. This is the length that it takes for a broken wire to regain its full strength through friction with the surrounding wires. Hence, it is strongly dependent on the level of compaction of the cable, as induced by the cable bands. Until recently, the redevelopment length was estimated by the retraction distances when samples were cut and was usually considered equal to two panel lengths on each side of fracture surface (NCHRP NCHRP Report 534, 2004). In the past few years, work done at Columbia University (Noyan et al., 2010, Betti et al., 2011) has focused on the friction among wires and on how it affects the strain (stress) transfer from a broken wire to the surrounding wires. Using neutron diffraction technology on small strands of parallel (steel and aluminum) wires, Noyan et al. (2010) have shown that the length from the fracture surface to the point where the wire has regained its full strength is much shorter than the one previously estimated. These results have been incorporated in various finite element analyses to assess the friction effect on the estimation of the cable strength (Betti et al., 2011). The effects of this conclusion are further investigated for potential changes in the guidelines. With respect to the inspection of a parallel wire main cable of a suspension bridge, very little information is available in the current literature. The current practice of visual inspection, consisting in unwrapping, wedging and visually classifying the exposed wire into 4-5 categories, is still the only one that is used in field investigations for assessing the cable strength. However, even in the case of unwrapping the entire length of a cable, this practice provides only a limited amount of information because the amount of wires exposed (and consequently inspected) is still a quite small percentage of the entirety of wires. Consequently, the effectiveness of the various methodologies for estimating the remaining cable strength is affected by the reliability and completeness of such a limited information. In recent years, thanks to the advances in sensor technologies, a variety of nondestructive evaluation technologies (e.g. acoustic emission, magneto-strictive, fiber optic gauges, etc.) has appeared on the stage of cable inspection but the results, so far, have been inconclusive. The major stumbling problem for their success is represented by the size of a main cable of a suspension bridge. While all of these technologies are valid and cost-effective for small bundles of wires, like a strand at the anchorage point, they have difficulties when presented with an enormous bundle of wires such as a main cable. A recent report from FHWA (FHWA-HRT-14-023), released in May 2014 and prepared by members of this research team, provides a comprehensive assessment of the state of the-art in main cable inspection technologies, highlighting the strengths as well as limitations of each method. The following NDE technologies were evaluated: • Acoustic Emission (AE) • Magnetostrictive (MS) • Fiber optics (pH, humidity, chloride content, strain, etc.) • Electromagnetic (magnetic flux leakage) • Linear Polarization Resistance (LPR) • Electrochemical impedence spectroscopy The results of this research were promising. They showed that commercially available sensors are capable of providing useful information about the environmental conditions within the interior of a

9 main cable, and that they are sufficiently durable to survive the rigors of field installation. The ability to monitor variables such as temperature and humidity can be used as indirect indicators of the potential for corrosion activity. Several direct sensing systems were also identified as having a strong potential for future applications, these included the main flux method and a system based on a combination of AE and MS technologies (the latter being applicable to smaller strands, such as those found in anchorages). Based on the literature review performed by the research team, there have not been any noteworthy advances in the NDE/SHM technology from the results published in 2014. It should be noted that the scope of the literature review was to identify information on commercially available NDE technologies, as well as SHM and other bridge preservation technologies such as dehumidification, that are relevant to suspension bridges in general and main cables in particular. It should be further noted that the development of new NDE methods for cable inspection and/or evaluation is outside the scope of this project. Synthesis of Survey Purpose of Survey The previous NCHRP effort, which led to the development of NCHRP 534, that represents the current guidelines followed by bridge owners for planning and conducting inspections of main cables of suspension bridges and for estimating their remaining strength, was based upon data from primarily two bridges with partial information from one other. Now that the methodology developed under NCHRP 534 has been in use since 2004, and numerous bridges have been inspected and evaluated using NCHRP 534, it was felt that most data was available than was the case in 2004. In addition, the use of NCHRP 534 for many more bridge inspections could reveal any shortcomings in the present methodology as well as establish areas that the present approach could be improved. An effort was made to highlight the knowledge gaps between state of the-art and current guidelines. Survey Process and Summary of Results A survey was developed by the research team and was sent by e-mail to the owners of approximately 70 suspension bridges located in North America, Europe and Asia. Responses were received addressing 21 bridges located in North America and Europe. Sixteen bridges were in North America and five were in Europe. No responses were received from Asian bridge owners. Of the 21 bridges, 17 utilized parallel wire cables and four had helical strand cables. A copy of the questionnaire is included in Appendix A. Due to security concerns, many of the bridge owners requested that the data received be “masked” in such a way that the data could not be associated with any specific bridge. For this reason, the RT made the decision to summarize the data in ranges and charts in such a way that no direct association can be made of the data and a specific bridge. A summary of the results are as follows and is presented in the same order and question numbers used in the questionnaire. Q1.0 Organization. (21 Responses) Of the 21 bridges for which data was received, 15 of the bridges were owned by Authorities, five were owned by departments of transportation and one was from a private owner. Q2.0 Keep Information out of the Public Domain?

10 As stated above, for reasons of security, most of the bridge owners requested that the data provided not be specifically associated with their bridge. Q3.0 Description of Suspension Bridge. (21 Responses) Q3.1 Name of the Bridge. (21 Responses) For reasons of security, most of the bridge owners requested that the bridge name not be used due to concern that the data provided might be specifically associated with their bridge. For this reason, no bridge names are presented in the data summaries. Q3.2 Date Opened to Traffic. (21 Responses) Date Opened to traffic. Of the responses received, the oldest bridge opened to traffic in 1921 and the newest opened to traffic in 1997. Figure 1. Period Bridges Opened to Traffic Q3.3 Roadway Carried. (21 Responses) Roadways carried included interstate highways, state highways, local streets and privately-owned roadways. Q3.4 Length – Anchorage to Anchorage. (19 Responses) Of the responses received, the total bridge length (Anchorage to Anchorage ranged from just under 2,000 feet to 9,570 feet).

11 Figure 2. Total Length of Bridge (in feet) Q3.5 Span Arrangement. (20 Responses) Q3.5.1 Number of Suspended Spans. All but two of the bridges had three suspended spans. Two of the bridges had side spans not supported by the main cables. Q3.5.2 Length of Main Span. The length of main span of the bridges reported ranged from 644 feet to 4,260 feet. The distribution is shown below. Figure 3. Length of Main Span (in feet) Q3.5.3 Length of side spans. As stated above two of the bridges did not have side spans supported by the main cables. The side spans ranged in length from 176 feet to 1800 feet.

12 Figure 4. Length of Side Spans Supported by Main Cables (in feet) Q3.6 Width of Bridge curb to curb. (17 Responses) As illustrated in the graph below, the width of the bridge decks from curb to curb varied from 26 feet to 87 feet with one bridge having two roadways: one with a width of 80 feet and the other a width of 74 feet. Figure 5. Bridge Width (in feet) Q3.6.1 Number of traffic lanes. Four of the bridges carry two lanes of traffic, nine of the bridges carry four lanes of traffic, two carry six lanes of traffic, one carries eight lanes of traffic and one carries 12 lanes of traffic (two levels at six lanes each). Q3.6.2 Width of lanes. Of the bridges for which data was suppled, the lanes widths vary from a minimum of 10 feet to a maximum of 13 feet.

13 Q3.7 Sidewalks. (21 Responses) Fourteen of the bridges reported having sidewalks and seven did not have sidewalks. The widths of the sidewalks ranged in width from 2’-6” to 6 feet. Q4.1 Number of Cables. (21 Responses) Of the 21 bridges for which responses were received, 20 of the bridges had two main suspension cables with one bridge having four cables. Two of the bridges have supplementary cables. Q4.2 Cable Sag. (21 Responses) The cable sag is a measure of the vertical distance from a straight line between the tower saddles to the lowest point on the cable, usually at midspan. The cable sag for the 14 bridges for which the information was provided is shown below. The sag is most often expressed as a sag ratio which is the cable sag divided into the span length. Sixteen respondents reported the sag ratio for their bridge. The sag ratios for the bridges reported as a function of their date of construction are shown in Figure 7. Figure 6. Cable Sag

14 Figure 7. Sag Ratio versus Age Q4.3 Cable Diameters. (21 Responses) Seventeen of the respondents provided the diameter of cables on the bridges. Only the main span diameters are shown below. Two of the bridges have more wires and hence a slightly larger diameter in the side spans due to the span ratio. One of the bridges has four cable all the same diameter. Figure 8. Diameter of Cables (inches) Q4.4 Parallel Wires or Helical Strand? (21 Responses) Of the 21 bridges reported, 17 had parallel wires and four had helical strand comprising the cable. 4.5 Cable Protection System. (21 Responses) 4.5.1 Paste beneath the Wrapping Wire and Paste Material: Of the 21 responses, 19 bridges contained red lead paste with 15 of the bridges reported having had a zinc paste applied in panels that

15 had previously been opened. The remaining two bridges are reported to have a zinc chromate paste. These were both helical strand cables. 4.5.2 Wrapping Wire Used? Characteristics. Of the 21 responses to the questionnaire, 19 bridges have No 9 (0.148” in diameter) wrapping wire. Two of the bridge outside of North America use wrapping wire with diameters of 0.12” and 0.14” in diameter. 4.5.3 Have the cables been oiled? If yes, what material was used? Five bridges were reported to have had their cable oiled, two with raw linseed oil and three using a proprietary Prelube 19 material. Two of the bridges have had both cables oiled for their full length. The remaining bridges have only had the panels that have been opened for inspection oiled. Some of the bridge owners have oiled the strands in the anchorages. 4.5.4 If the cable has been oiled, when was it oiled? Most of the oiling has taken place in concert with cable inspections. The latest oiling of the cables has taken place in 1993, 1994, 1996, 1998, & 2002. 4.5.6 Does the cable have a Flexible Wrapping System? If yes, what material? Six bridges reported have a flexible wrapping material in addition to wrapping wire. One bridge reported use of a “polyethylene” material, three utilized ”Cable Guard” , one has part of its cable wrapped with Cable Guard, and one with a material as being “elastomeric.” 4.5.7 Does the cable have a dehumidification system? If yes, how long has it been in service? Five bridges were reported as having a cable dehumidification system for their main cables. Two bridges had dehumidification systems in service less than 1 year at the time of reporting, one was installed in 2009, one in 2008, and one installed in 2005. Q4.6 For Cables Comprised of Parallel Wires. (17 Responses) Q4.6.1 Diameter of wire used in the main cable: Of the 17 bridges reported to have parallel wire strand, all but one bridge was reported to have galvanized wire that was 0.196”in diameter. The one exception was reported to have a wire diameter of 0.207”. Q4.6.2 Number of Wires in the Main Span Main Cable. The Number of wires reported in the questionnaire ranged from 7,068 wires to 26,108 wires. The distribution is shown in the graph below.

16 Figure 9. Number of Wires in Parallel Wire Cables Q4.6.3 Number of wires in the side Spans. All the bridges containing parallel wires, with the exception of three bridges, were reported to have the same number of wires in the main span and the side spans. Three of the bridges had a small percentage of additional wires in the side spans due to the main span to side span length ratio. Q4.6.4 Mechanical Properties of the Wire? Ten bridge owners provided data regarding the strength of the wires. As can be seen in Figure 10, in most cases the mean tested strength of the various stages of the wire exceeded the original specified strength of the wire. Though not reported in the questionnaire, the standard deviation in the higher stages is typically greater. Figure 10. Mean Wire Strength by Corrosion Stage

17 Q4.6.5 Are the Wires galvanized? Eighteen respondents reported that the wires in their main cables are galvanized. The remaining three did not respond to this question. Q4.7 For Cables Comprised of Helical Strands. (3 Responses) Q4.7.1 For cables comprised of Helical Strands – Type of Strand? On the four bridges reported to have Helical strand two were reported to have Structural Strand – round wire, one was Locked Coil and one did not respond to the question. Q4.7.2 For cables comprised of Helical Strands – Diameter of Helical Strand The strand diameter of one of the three bridges that responded was reported to be 1.47 inches, while the second bridge had a combination of 1.75-inch and 1.25-inch strand. The third bridge had Locked Coil with a diameter of 2.37 inches. Q4.7.3 For cables comprised of Helical Strands – Number of Helical Strands in each Cable? Of the three bridges reported, one bridge was reported to have 61 strands with no diameter reported, while the second was reported to have 31 strands at 1.75 inch and six strands @ 1.25 inch. The bridge with locked coil had a total of 31 strands with no diameter reported. Q4.7.4 For cables comprised of Helical Strands – What Type of Filler Elements were used to make the cable round? One bridge was reported to use Cedar wood fillers; two were reported to have extruded aluminum and one was reported to use a combination of tubular aluminum, solid PVC, and hollow PVC. Q5.0 Main Cable Monitoring System. (21 Responses) Q5.1 Acoustical Monitoring System on the Bridge? Of the 21 responses received, 7 bridges are reported to have an acoustical monitoring system on the main cables. Figure 11. Bridges with Acoustic Monitoring

18 Q5.2 What Acoustical Monitoring System Is Used? Of the seven bridges reported to have an acoustical monitoring system three were installed by Pure and four were installed by MISTRAS. Q5.3 How Long Has the Acoustical System Been in Service? Of the seven bridges reported to have an acoustical monitoring system the oldest was installed in 2001 and the newest installed in 2015 Q5.4 Comments on the Effectiveness of the Acoustical Monitoring System Only one comment was made from one respondent. “Recent upgrades in sensors, cabling, and software have improved the performance of the acoustic monitoring system. Additionally, a wireless network connection over Verizon network has improved communications and reliability.” Q5.5 Health Monitoring System Installed? Of the respondents, none reported the installation of an operating health monitoring system. 5.0.1 Name of health monitoring system used and performance. One bridge had a test installation of various sensors but most of the sensors failed during compaction and rewrapping. No manufacturer was provided in the response. Q6.0 Main Cable Inspection History. Q6.1 Date of Last Cable Inspection. (21 Responses) Of the 21 responses, 19 reported the year of the last inspection, which varied from 1996 to 2018. Two did not respond to the question. See Figure 12 for the distribution of inspection years. Figure 12. Year of Last Inspection Q6.2 How Many Times Has the Cable Been Opened/Inspected? (21 Responses) Only 20 responders provided information on the number of times their cables have been inspected for each bridge ranges from one to nine inspections. One did not respond to the question. See Figure 13 for the response distribution.

19 Figure 13. Number of Times Cables Have Been Inspected Q6.3 Is the Cable Inspected on a Regular Basis? (21 Responses) As shown in Figure 14, of the 21 respondents, only 10 stated that they inspect their bridge cables on a regular basis. Figure 14. Regular Inspections Performed? Q6.3.1 If inspected on a regular basis, on what interval? Of the respondents that stated they performed regular inspections, six stated they inspected the cable every 10 years, three every 5 years and one that stated the next inspection was based on the findings of the previous inspection.

20 Q6.4 Number of Panels Opened During Each Inspection. (20 Responses) Two bridges were reported as having all the panels opened and inspected. In the case of both bridges, this took place prior to the publication of NCHRP 534. Subsequent inspections on one of these two bridges reported opening eight panels for each inspection. Figure 15. Number of Panels Inspected Q6.5 on What Basis Were the Panels Selected for Opening? (20 Responses) As illustrated in Figure 16, location was the primary criteria for selecting panels to open with 13 respondents citing location as at least one reason for selecting a certain panel. The explanation provided in the questionnaire responses stated that understanding of the behavior of a suspension bridge and accessibility were the two main factors that drove the location selected. Results from previous inspections was the primary criteria for four of the bridges while exterior condition of the cable was the primary criteria for four other bridges.

21 Figure 16. Selection Criteria for Opening Panels Q6.6 Were Field Inspection Forms Prepared for Your Bridge? (19 Responses) All 19 respondents stated that inspection forms were prepared for the cable inspection. Those that provided forms closely followed the examples in NCHRP 534. Q6.7 Total Number of Samples Taken for Testing? (17 Responses) Of the 17 parallel wire cables for which responses were received, only 14 reported the number of samples taken with the number ranging from 3 to 122. Q6.8 How Many Specimens Were Tested? (17 Responses) Only 10 of the 17 respondents provided the number of specimens that were taken from the samples. As shown in Figure 17, the number ranged from 11 to 1208. It does not appear that most of the inspections followed the number of samples recommended in NCHRP 534 though it is not possible to emphatically come to that conclusion without knowing if stage 3 and stage 4 wires were present at the time of inspection.

22 Figure 17. Number of Samples and Specimens Q6.9 What Tests Were Conducted on the Specimens? (10 Responses) It appears based on the 10 responses to the questionnaire that the majority of the specimens tested used most, if not all, of the tests recommended in NCHRP 534. Tests reported to have been performed were chemical analysis, tensile tests; coating weight; Preece test; surface chemistry; microscopic examination; fractographic examination; fatigue tests, hydrogen content, analysis of water in the cable, and bacterial testing. It appears that following the initial inspection the testing focus changed to primarily tensile tests, microscopic examinations, and fractographic examination. Q6.10 Please Describe the Sampling Technique Used? (15 Responses) Based on the responses to the questionnaire, most bridges sampled from eight wedge lines located at the 12:00, 1:30, 3:00, 4:30, 6:00, 7:30, 9:00, and 10:30 clock positions. Responders reported samples were typically take from rings no deeper than approximately 2 inches into the cable. Most of the responders reported following NCHRP 534 as far as technique. Only one helical strand cable reported taken a few samples from broken wires on the surface of the strands. Q6.11 Were New Wires Spliced to Replace the Samples? (17 Responses) Respondents stated that in the case of 15 bridges, new wires were spliced to replace those taken for samples. Q6.12 Were New Wires Spliced to Replace Broken Wires? (15 Responses) The returned questionnaires reported five bridges for which new wires were spliced to replace the broken wires where conditions allowed; five bridges stated that the broken wires were not spliced; five reported no broken wires were found, and two provided no response. Q6.13 If Yes, to What Depth in the Cable? (14 Responses) Those replying regarding the depth to which they replace wires ranged from 1 inches (about five rings) to 2 inches (about 10 rings). Two bridges only replaced wires on the surface and one bridge had no response.

23 Q6.14 Describe the Method Used to Tension New Wires? (17 Responses) Twelve respondents stated that a new wire was inserted with a pressed-on ferrule at one end and a pressed-on turnbuckle on the other end. Holding on to the two wire ends, a come-along was used to tension the wires to the desired load so that the installed turnbuckle could be tightened. The tension was also checked by recording the deflection of the wire under a fixed load over a fixed length. Five respondents did not answer the question. Q6.15 Method Used to Inspect and Evaluate the Condition and Strength? (21 Responses) All but two respondents answered this question. Of the remaining 19 bridges, one reported using a method that predated NCHRP (though an upcoming inspection was to use NCHRP), and the remaining 17 bridges used NCHRP. Two of the 17 bridges used both the NCHRP and the BTC method. Three of the four helical strand bridges stated they used NCHRP with modification for their cable inspection. Q6.16 If NCHRP Was Used Are There Any Suggestions or Recommendations? (3 Respondents) Three respondents, some owning multiple bridges, offered up the following suggestions: 1. Better define the inspection frequency and number of panels to open based on past inspections and condition findings. Incorporate any appropriate NDE methods. Consider level of maintenance, or changes in maintenance methods. Determine how dehumidification affects inspections. Expand the suspender rope inspection guidelines. 2. Need to include the details of inspection and strength evaluation of the main suspension cable that consists of the helical structural strands. Need to specify adequate NDT methods to detect in situ corrosions and wire breaks. 3. The third group of respondents from the owner of multiple suspension bridge offered the following: Cable Inspection, Sampling and Testing a. Consider length of cables when determining number of panels to be opened. Does it make sense to open the same number of panels on a small/short bridge as on a big/long bridge? If different bridges have fewer/more panels (or length of cable), why open the same number of panels? Might the number of openings be a percentage of the number of panels (or length of cables)? Every bridge cable (s) is unique as to the cable length, diameter, number of cables and wires, their condition, maintenance and inspection history, propensity for corrosion and level of deterioration, types of corrosion, propensity for cracked wires, number of broken wires, etc. The inspection plan must take these and other considerations into account and develop a recommended inspection program that is specific to the bridge and its cable(s). This would avoid opening up more panel points than necessary as it is unknown what harm cable openings may cause to long-term cable performance. b. Evaluate sample sizes. (Table 2.4.3.5.1-1). If the sample mean calculations are eliminated it may result in needing a different sample size where a smaller sample sizes can be used. Correlation to be developed for the number of wire samples to be taken vs. the cable wire condition found upon cable opening. If no wire breaks are discovered wire sampling requires cutting and splicing unbroken wires. This results in splicing back wires with some uncertainty as to the performance of a spliced wire and may be causing more harm to the cable. If numerous broken wires are found, then sampling of such wires limits the harm to the cable as they must be spliced. However, if no broken wires are found the number of wires to be cut and removed for sampling must be carefully considered so as to minimize cable wire damage. c. The current practice of wire sampling should be revisited and the use of statistical methods to improve the current wire sampling protocol should be considered. d. Wire sampling from interior of cable: Most samples come from the outer few layers of wires due to practical limitations on access. Recommend the collection of information from different owners

24 on methods used for wire cutting and wire repair for the interior wires in order to come up with a recommendation on how to utilize more wires from the interior of the cable in the wire sampling. e. Evaluate the inspection intervals (Table 2.2.4-1). 5 years between internal inspections is short. Consider including the safety factor and adding acoustic monitoring system at this level of corrosion. f. Consider adding safety factor as a criterion for scheduling the next opening. (Table 2.2.4-1). Why base the next scheduled opening on the level of corrosion? The level of corrosion doesn’t determine the safety factor (i.e. strength of the bridge). The capacity of the cable and the load (demand) on the cable encompasses both aspects unique to each particular bridge and should be considered when deciding the timeframe for the next cable opening. g. Include inspection methodologies for anchorage strands. The strength calculation for the anchorage splay have specific issues that are not covered by the report. For example, how does a broken wire effect a full strand, vs. a half strand, vs. a quarter strand? h. Address suspension bridges with four cables. Inspection and sample selection is based on “per cable”. Does the same hold true for bridges with four cables? Can the quantity of panels to be wedged *per cable* be reduced when the number of cables is increased? i. Conduct independent inspections. To measure the sensitivity of the inspection and calculations due to human objectivity, an experiment could be done by using two set of independent inspectors for a panel. The two sets of inspectors shall also independently rate the specimens before testing. This will help determine if there is an effect on the inspection subject to human objectivity. j. Conduct two sets of inspections with two sets of wedge lines. To evaluate the extrapolation practices, inspect eight wedge lines, then re-drive the wedge lines a few degrees (and only a few degrees) off the previous wedge line. This will gauge the sensitivity of the inspection results as a factor of the specific location of where the wedge line is driven. This is important to determine if changes from one inspection to another are the result of actual changes in cable condition or just change in the precise wedge line location. k. Evaluate/Compare results for bridges inspected since NCHRP 534 was implemented. Evaluate/Compare results for bridges inspected since NCHRP 534 was implemented. Possibly also compare to results from pre-NCHRP analyses. Cable Strength Evaluation a. Recommend the re-evaluation/further development of the methods used for cable strength evaluation to better account for variations between corrosion grade and ductility using statistical methods. b. The way the percent cracked wires determined from the sample population generally leads to an overestimation of the percent cracked wires, which leads to an overly conservative estimation of cable strength. c. Redevelopment length needs to be revisited, from experience it appears that a broken wire redevelops its full capacity sometimes within the cable panel point, and certainly at the nearest cable band. d. Fix calculation formulas that are incorrect. (4.3.1.1-1 & 4.5.2-4). A couple of formulas appear to have incorrect operation symbols (+/-). e. Reevaluate the calculation of the lowest strength of a wire (4.4.3). There has been some experiments that have indicated that a longer piece of wire might break at a lower level than the lowest break value of its smaller pieces. This conclusion is not fully understood or conclusively proven. However, the Guidelines tries to capture this by finding the lowest breaking value of a wire by using a distribution to define the minima of the distribution as the hypothetical minimum breaking strength. The strength of a longer wire should be further tested. Current practice might be overly conservative and demands lots of sample wires, it might be sufficient to assume that a wire breaks at its weakest point as a chain breaks at its weakest link. Experienced inspectors should be able to find the worst

25 pieces of wire for testing. The Guidelines calculations also assumes (3.2.1) the specimens shall be of the same stage of corrosion. This is seldom the case and will influence the accuracy of these calculations. f. Move Simplified Model and Limited Ductility Model to appendix. (5.3.3.1 & 5.3.3.3). Brittle Wire Model is predominantly used. g. Change the calculation on redeveloped broken wires. (5.3.4). Improve the calculation of redeveloped broken wires. The broken redeveloped wires are simply added as a force. The redeveloped wires have a force at zero load in the cable and also a force in the cable when all wires are broken in the Guideline calculations which is unrealistic. Calculation can become closer to reality and more intuitive. With the use of computer programs the simplification of this calculation in the Guidelines is not necessary. h. Change the calculation on redeveloped cracked wires. (5.3.2.4.2 & 5.3.3.2.3 & Appendix B). Improve the calculation of redeveloped cracked wires. The cracked redeveloped wires are added back into the calculation along a 95% Stage 2 normal distribution. The redeveloped cracked wires never do break in the Guideline calculations which is unrealistic. Calculation can become closer to reality and more intuitive. i. Evaluate the development length (4.5). Strength calculations are highly sensitive to the effective development length, and the development length is highly sensitive to retraction measurements of the sample wires. Even with extreme care while measuring, retractions can have large variations. Some of the larger retractions can be due to the absence of wire wrapping in inspected panel. j. Remove the groups (4.4.1). In the calculation the wires in the stages are rebranded into groups with their separate number, this can be confusing. We suggest keeping the same nomenclature throughout. For example, groups would be called “Stage 4 not cracked” and “Stage 4 cracked”. If you have Stage 3 cracked wires the group numbering will be offset from the norm. k. Include a fifth stage (1.4.2.2). Stage 4 has the biggest effect on the main cable strength. In cables with significant amounts of Stage 4 corrosion the lower tier of the stage can significantly reduce the characteristics of the stage group. A suggested fifth stage would include severely corroded wires that have 100% corrosion over 3 inches and/or section loss. This will give a better representation of the actual properties of the wires. This suggestion will also more accurately estimate the correct number of cracked wires as those are more prevalent among the much-corroded wires. l. Evaluate the use of Monte Carlo Simulations. Monte Caro simulations may prove to be a useful tool to model the strength of the cable. It has a great flexibility and can quantify uncertainties in calculations. Appendices a. Use Normal distribution for Stage 1 and 2 and Weibull distribution for the rest (Appendix A.5). b. Consider using Normal distribution for the lower stages since they tend to have a shorter left tail due to minimal strength requirements during construction. c. Develop methods to calculate adjacent inspected panels (B.4.1 & B.4.2). Briefly describe how to treat one adjacent inspected panel. Now only single panel or all panels inspected are discussed. d. Change the calculations from stress based to force based (Appendix C). Wires with section loss can have high breaking stress when the area is taken into account. Ultimate force per wire will be more appropriate and will not lead to confusion. The guidelines does use a nominal area in its calculations but it is not explicitly explained. e. Fix Appendix C Contents. Remove chapters C.7 and C.8 since they are not included in the appendix. Reliability-Based Evaluation Methodology

26 a. Though pointed out in the Problem Statement of the second top priority “Development of models to predict the strength of deteriorated cables” at the NCHRP-sponsored “Workshop on Safety Appraisal of Suspension Bridge Main Cables” (2) held in Newark, New Jersey, in 1998, that “the Factor of Safety commonly calculated following a cable inspection appears to have little or no meaning…..A cable rating concept is needed, along with a statistically based model to account for the effects of various forms of wire deterioration”, the main cable remaining strength evaluation portion of the NCHRP 534 solely focused on the uncertainties of the main cable wires affecting the cable capacity. However, a cable rating concept won’t be complete without addressing the uncertainties of cable loading demand. Moreover, the current industry practice still adopts the concept of Factor of Safety (FOS) (ratio of capacity over demand) using the traditional allowable stress design method in evaluating the main cable remaining strength, which can be misleading as it does not sufficiently address the uncertainties of the applied load effects on the demand side. For example, the current American Association of State Highway and Transportation Officials (AASHTO) Load and Resistance Factor Design (LRFD) bridge design specifications recognize the uncertainties of the dead load effect by using both 1.25 and 0.9 as the applied dead load factors in Strength I design and such variation in dead load can affect the FOS significantly especially for suspension bridge main cables as the dead load to live load ratio is much higher than that of the conventional highway bridges (5 to 1 vs 2 to 1). In addressing the non-applicability of Load and Resistance Factor Rating (LRFR), the FHWA Primer indicates that “further research is required in the area of LRFD and LRFR, before specific recommendations regarding load factors and rating strategies can be made for long span suspension bridges”. Although no calibration has been performed for LRFR for long span suspension bridges, a reliability-based evaluation methodology consistent with LRFD and LRFR philosophy to verify that the cable target failure probabilities are not exceeded with the consideration of the main cable remaining strength as well as the applied load effects shall be established as a priority due to the criticality of the suspension bridge main cables. A more conservative target reliability index higher than what’s used in AASHTO LRFR procedures may be appropriate due to the consequences in the event of failure. NDEs and SHM a. The cable inspection, wire sampling and wire testing are needed because of the lack of reliable NDE and SHM techniques. However, the cost, complexity and uncertainties of doing cable inspections with current procedures clearly point out the need for better NDE techniques that can look through the various materials used for cable covering and penetrate far enough into the body of the cable to provide meaningful information. Consideration should also be given on how SHM for cables with the use of sensors could supplement cable inspections and possibly reduce the potential for cable openings in the future while providing information on cable corrosion and humidity. Q6.17 Calculated Factor of Safety & Remaining Life? (14 Responses) Of the 17 parallel wire bridges for which responses to the questionnaire were received, no FOS was provided for four. For the FOS of the remaining 13 bridges see Figure 18. None of the respondents reported a projected remaining life.

27 Figure 18. FOS as of Last Inspection Q6.18 Location of the Panel With the Lowest Factor of Safety? (16 Responses) A slight majority of the respondents stated the controlling FOS was in the general vicinity of the quarter-points of either the main span or side span. Only one reported the controlling FOS at the midspan. Q6.19 Highest Stage of Corrosion Observed During Inspection? (17 Responses) Of the 17 parallel wire bridge responses to the survey, all but one reported having stage 4 wires in their cable. Its highest level of corrosion was stage 3. Information was not provided for this item for one bridge. Q6.20 Have Broken or Deteriorated Wires Been Observed at the Strand Shoes, Tower Saddles, Cable Bents or Spay Saddles? If so, Where and How Many Wires Were Broken? (17 Responses) Of the 17 parallel wire cables, only four were reported to have wire breaks in the strand shoes and the splay saddles. Most did not know the exact number. Four other bridges reported having no breaks in these areas and the remaining nine bridges did not respond to the question. Of the helical cable bridges, one reported having one broken wire but it does not state where, two report no broken wires and one did not respond. Q6.21 Was a Formal Report of Findings Prepared? (21 Responses) Of the 21 respondents, 18 stated a report was prepared, one said a report was not written and two did not respond to the question. Q6.21.1 If so, can you provide a copy of the final report? Twelve respondents stated they would provide a copy of their report. That being said, only seven did so. Six respondents said no and three did not respond. Q6.21.2 If no, will you allow the report to be reviewed? Of the six that responded no to sending a copy of the report, five stated they would allow the report to be reviewed on site. Q6.22 Were Any NDE Methods Used to Inspect the Cable? (21 Responses)

28 Of the 17 parallel wire bridges none reported any use of NDE on the cable. One owner reported field instrumentation of selected eye-bar elements. Of the four helical strand cables, none reported any use of NDE on the cable. One reported the use of “magnetic testing to measure zinc coating thickness.” Q6.23 If Yes, Please Describe the NDE Approach Used. (21 Responses) As stated above, one reported field instrumentation of eye-bar elements and one reported the use of “magnetic testing to measure zinc coating thickness”. Q6.24 If Your Cable Is Comprised of Helical Strands Have You Inspected It? (4 Responses) Of the four helical strand bridges, all report having inspected their cable. Q6.25 If Yes, Describe the Approach Used to Inspect It? (4 Responses) Two of the four respondents described their process. 1) Exterior wire wrap, cedar wood fillers and a lead powder /linseed oil coating were removed to expose the outer surfaces of the strands of each of the cables for inspection. Cable faces have been labeled for ease of locating recorded observations. The strands on the six outer faces were inspected and then the six outer strand layers were separated from the inner bulk of the strands with ultra high molecular weight plastic wedge shims to expose the inner surfaces of the exterior strands and reveal the condition of the subsequent underlying layers of strands. Two opposite faces of the cable bundle were exposed during each phase of this portion of the inspection. 2) The process assigned pre-defined categories of corrosion to the galvanized wires. It was not considered advisable or possible to attempt to open the lay of the locked coil strands over this panel length while under tension and in close proximity to other strands. Therefore, the condition was noted for the external wires visible on the outside of each of the strands. The inspection results for each strand were summarized for each wedge line at three positions along its length, spaced at approximately 6 foot intervals. Q7.0 Anchorage and Tower Tops Q7.1 Are Your Anchorages Dehumidified? (12 Responses) Twelve respondents stated that their anchorages were dehumidified. One respondent did not provide the age. For the ages of the systems in place, see Figure 19.

29 Figure 19. Age of Anchorage Dehumidification System Q7.2 Do you inspect the strands in the anchorages as part of your biennial inspection? (21 responses) Sixteen respondents stated they inspect the strands as part of their biennial inspection. One said no and four did not respond to the question. Q7.3 Have you wedged your strands in the anchorages? (21 Responses) Seven respondents state that they have wedged the strands in the anchorages, nine state that they have not and five did not respond to the question. Q7.3.1 If so, can you provide a copy of the final report? Only one respondent said they could provide the report, four said they could not and 16 did not respond to the question. Q7.3.2 If no, will you allow the report to be reviewed? Four responded yes to this question, 17 did not respond. Q7.4 Have you shown any losses in the steel anchorage elements at the concrete interface? (21 Responses) There was no response from four of the bridges. Eleven bridges were reported as having no losses in the anchorage elements. Six bridge reported losses ranging from minor to moderate loss of section. Two bridges reporting cracking in the concrete around the eyebars. Q7.5 What protective measures have you taken at these locations? Five respondents reported that they applied grease to the concrete eye-bar interface while four reported other protective coatings. Four reported they did nothing while eight did not respond to the question. Two respondents stated they had injected cracks that had occurred in the anchorage. Q7.6 Are your tower tops dehumidified? Twelve respondents stated that their tower tops are not dehumidified, while five said theirs were. Four did not respond to the question.

30 Q8.0 Suspender Information Q8.1 Panel length between suspenders and diameter of suspenders? (21 Responses) The panel lengths from 17 respondents ranged from a maximum of 65.62 feet to a minimum of 20.83 feet. Four respondents did not provide a panel length. Eighteen respondents provided suspender diameters from a minimum of 1.875 inches to a maximum of 2.75 inches. One respondent provided a suspender diameter but no panel length. The distribution of panel lengths to strand diameter is shown in the figure below. Figure 20. Suspender Diameter versus Panel Length Q8.2 Number of Suspenders at each Cable Band? (21 Responses) Of the 21 respondents, 18 reported two suspenders at each cable band, two reported one, and one bridge reported four. Q8.3 Are the suspenders ropes or strands? (21 Responses) Of the 21 respondents, 19 reported the use of wire rope for the suspenders and two reported the use of locked coil. Q8.4 Do they loop over the cable or are they pinned? Of the 21 respondents, 17 reported the suspenders were looped over the cable bands, two reported they were pinned, and one reported they were socketed to the cable band. Q8.5 Are they inclined or vertical? Of the 21 respondents only one reported the use of inclined suspenders. Q8.6 Are the suspenders Jacketed? Of the 21 respondents, only two reported their suspenders jacketed and the composition of that information is classified.

31 Q8.7 Have any suspenders been replaced? If so When? Fourteen responded that they had removed and replaced suspender ropes. Most were removed for testing. One bridge has removed and replaced all of the suspenders. A table showing the dates of removal are shown below: Table 1. Date of Suspender Removal/Testing Arbitrary Bridge Designation Year Suspenders Replaced 1 2004 2 2008, 20016, 2017 3 2007-2008 4 2005,20018-19 5 1970 6 Ongoing 7 2011,2013 8 1987-1989 9 1998 10 2003-2006 11 2005 12 2005 13 2008 14 2007 Q9.0 Additional Comments Four respondents provided additional comments. Each of these are listed below: Respondent # 1 1. The current document (NCHRP 534) has provisions that make proper interpretation and application of the provisions impractical or prohibitively expensive for owners (if one were to faithfully follow the text of the guidelines). 2. One of the questions I raised when we were evaluating our options, was how could one say, with confidence, that the sampling performed in accordance with NCHRP 534 was “statistically significant” and how could we say that the results obtained from such sampling were adequate and appropriate to draw conclusions about the entire length of a main cable. 3. A risk-based guideline for inspection and strength evaluation of main cables should attempt to identify and document the biggest risks faced by these critical structural elements – corrosion, impact from accidents/terrorist events, long-term creep and fatigue etc. 4. We all know that we calculate factors of safety of aging structural elements at weakest (or worst) sections (I don’t believe the current document talks about the criteria for identifying such sections on a main cable), but if the sampling used to calculate such factors of safety is itself not statistically significant, how could we then say that the minimum FOS arrived at by evaluating a limited number of panels and the conditions observed therein is also the minimum across the entire length of the main cable? 5. While I realize that main cables are important structural elements of suspended bridge spans, they are by no means the weakest links on the suspended spans. In other words, failures that result in-service interruptions on suspended bridge spans will occur on structural elements other than main cables far before the main cables themselves start to experience failures that result in-service interruptions. As such, the industry needs to be careful when it recommends expensive dehumidification systems to

32 protect the main cables, more so, because higher risks of service interruptions exist and can manifest themselves on other structural elements far sooner. Respondent #2 New/updated guideline should include details of helical structural strands inspections, testing and strength evaluation techniques, reliable NDT methods to determine in situ cable conditions, any technique to conduct internal inspections at saddles, vibrations based tension measurements techniques and applications, wire sample collections and repair techniques for helical strands, condition rating and remaining life estimation techniques, protection system-wrap-paint vs dehumidification - cost-benefit approach, inspection frequency and timing for helical strands. We have used cable rating definitions (abbreviated): a. Excellent – pristine condition, no concerns. b. Good – may have pitting and minor corrosion loss with no wire breaks c. Fair – minor material loss, some broken wires d. Poor – significant material loss and many broken wires. Requires immediate attention and possible load restriction. Respondent #3 Difficult applying NCHRP 534 to our cable type. Stands are layed [sic] up with bituminous filler material between stands - this makes dehumidification virtually impossible although - bituminous filler has done an excellent job of preventing corrosion. Outer z wires on strands have also contributed to protecting overall condition of strands NDT inspection methods to determine internal conditions of individual strands would be very useful. Respondent #4 Applicability: The NCHRP 534 Guidelines offer the only published guidance on the inspection and assessment of parallel wire, aerially spun, main cables for suspension bridges. They were developed in the United States from experiences on American suspension bridges. However, there are key differences between US and British suspension bridges which might cast doubt on the strict applicability of the Guidelines to the UK Bridges: • US bridges that have been investigated are generally much older (in the approximate range 65 to 120 years old). • Environment and climate could be significantly different, e.g. New York summers tend to be hotter and more humid. • Standards of inspection and maintenance are probably better at some UK bridges than for some of the US bridges, especially those that are not operated as toll bridges. Wire quality might be better for newer bridges and more consistent than wire from older suspension bridges. • Some of the bridges being evaluated are younger than those bridges used as case studies when the guidelines were developed. Some possible consequences of the above differences might be: • The older bridges in the USA are more likely to present similar corrosion conditions in many or all specimens along each wire sampled for testing. This will affect the standard deviation of specimen tensile strength within wire samples. • Different environments might lead to very different corrosion rates. • Different wire quality might affect the numbers of wire cracks and crack-like corrosion pits. • The NCHRP 534 guidelines emphasize that, in the USA bridges, the presence of cracked wires is the most important parameter in determining the strength of the main cable. This does not appear to dominate the calculations in quite the same way in the UK. Reservations about methodology: •We have several reservations with the NCHRP 534 wire model, viz:

33 • Wire samples containing mixed corrosion stage specimens provide lower values of the theoretical sample minimum strength between cable bands than wires in worse condition throughout. • Wire samples are classified according to the worst corrosion stage found anywhere along their lengths. Thus, most of the tensile test specimens in the least corroded condition are analyzed in association with other specimens from their sample which are in worse condition, and we find that there are relatively small samples of test data relating to wires in the best condition even though, at this site, such wires are in the majority. • We also have some reservations about certain aspects of the manner in which NCHRP 534 separates cracked wires from the data set and re-introduces them and treats them differently from other wire types. We understand that others have criticized similar aspects of NCHRP 534, although no alternative procedures are currently recognized. • Cable strength is very sensitive to small differences in assumed numbers of cracked wires; and these numbers must be deduced from small numbers of sample observations which are open to different methods of interpretation. • We note that NCHRP 534 employs wire strength models which do not appear to adequately model the correlation between the strengths of specimens all drawn from the same sample wire. The fact that apparently arbitrary adjustment (even though probably realistic ones) have been made for this by some consultants suggests that similar issues exist for other wire corrosion stages. Observations and Identification of Knowledge Gaps It is apparent from the responses to the questionnaires that virtually all inspections of parallel cable bridges are following the procedures outlined in NCHRP 534. There is variation in the number of panels opened and there is variation in the number of samples and specimens. The conduct of the inspections, methods used to take samples, and the splicing of new wires appear to be similar across the board. Though the data summarized in the previous section vastly expands the levels of information available at the time NCHRP 534 was developed, the RT found several areas that required further study: • The responses provided by bridge owners indicate that virtually no bridge owner is inspecting the number of panels required under NCHRP 534. This is being driven by the costs associated with cable opening and inspection as well as the concern on the part of a very small number of owners that such inspections may exacerbate cable condition problems. Statistical methods need to be developed that will allow for fewer panel openings and sampling/testing. • The responses also indicate that the number of samples taken from the cable and specimens tested vary widely. The number of samples needs to be justified and more well defined. • There is widespread confusion regarding the classification of samples versus specimens leading some expressing questions regarding “inspector bias.” Clarification of sample versus specimen classification is in order. • Even though NCHRP 534 goes into great deal regarding the statistical approach used and the confidence level resulting from that approach, there is concern by some that improvements should be made to account for the failure of most inspections to closely follow sampling and testing requirements. • On the basis of the data provided by bridge owners, It appears that the range of wire strengths for uncracked wires between stages 1 and 4 is quite small with most exceeding the originally specified ultimate strength. This appears to support the assertion by most investigators that the stage of corrosion isn’t the greatest impactor on cable strength. • The impact of and determination of cracked wire percentages needs to be revisited and addressed as well as how the effects of cracked wires are quantified.

34 • Inspection frequency and number of panels to be opened need to take into consideration any previous findings and extenuating circumstances such as the use of a dehumidification system or an acoustical monitoring system. Should the number of panels be based more on length of cable rather than a fixed number of panels per bridge? Should the existing FOS be taken into consideration rather than basing the next inspection on the presence of stage 4 and broken wires? • Many have questioned if other bridge owners have developed a means of sampling wires from deeper in the cable. From the questionnaire responses, it is apparent that all inspection samples have been taken from the outer two to three inches of the cable and repairs to broken wires have been made to the same depth. Methods of re-tensioning are the same for all inspections. • Several owners raised the issue of inspection of strands in the anchorage. There are no present guidance provided in NCHRP 534. • The question was raised regarding the impact the number of panels to be inspected has on bridges having more than two cables. This is not addressed in the present guidance. • One responder suggested doing a test of independent inspections on the same panel to determine the impact of inspector bias. This was reported as having been done in the reports provided for two bridges. Though there were minor differences in the wire classification, the resulting strength was almost the same value. • Several responders raised questions regarding the redevelopment length of broken and cracked wires contained in NCHRP 534. Based on some simplified field testing and limited laboratory testing, it is generally felt that this should be revisited. Methods that are more intuitive and closer to reality should be developed. • It was suggested that modifications be made in the wire classification system using subsets of the present classification system. • It was suggested to add the use of Monte Carlo simulation to quantify some of the uncertainties in the calculations. • The responders that have cables comprised of helical strand expressed the need to have some guidance as to how to inspect and determine the load capacity of the cable. • One responder suggested that the uncertainties of the loading demand on the cable should be addressed using an LRFR approach. No calibration has been performed for LRFR for long span suspension bridges. Given that the primary purpose of this project is a means of determining the strength of the cable, nothing developed in this project would preclude conversion to a reliability-based method once calibration for dead load and live load factors are developed at a later date.

Next: CHAPTER 3 Review of NCHRP 534 Approach »
Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Most suspension bridges in use today have cables composed of thousands of steel wires and most of these bridges are aging and carry high volumes of traffic. Deterioration of the elements of the suspension system is a problem, replacement of these elements can be expensive and problematic, while failure could be catastrophic.

The TRB National Cooperative Highway Research Program's NCHRP Web-Only Document 353: Risk-Based Inspection and Strength Evaluation of Suspension Bridge Main Cable Systems helps develop guidelines for inspection and evaluation of suspension bridge main cable systems using probabilistic approaches.

  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!