U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Based on its findings during the scanning study, the team recommends the following:
In summary, the scan team found many similarities, as well as significant differences, between the United States and the host countries in bridge design and analysis practices and bridge management and operating procedures. The team identified several key findings that it considers best practices, outlined in this report. The team believes that the best practices should be mainstreamed into practice in the United States by making the information available on Web sites, seeking demonstration or pilot projects, and holding workshops in association with the pilot projects. In addition, the team has planned papers and presentations at national and local meetings and conferences over the next several years. The purpose of the papers and presentations is to describe the overall results of the scanning study and details of specific technologies that participants should consider implementing in their States.
The results of this scan will support ongoing activities by FHWA, the AASHTO Subcommittee on Highway Bridges and Structures, and TRB/NCHR P to improve U.S. bridge design and analysis codes and specifications. This scan report contains many detailed findings that will enhance U.S. understanding of bridge design safety and serviceability and will lead to pursuit of further practices that will improve bridge design, analysis, and operations nationwide. The scan team is convinced that implementing the key findings of this study will improve design and operational safety standards of U.S. bridges, enabling them to provide longer service life with less maintenance. Changes to the bridge design and analysis codes will provide operational improvements that will increase mobility and help preserve the Nation’s highways.
Risk evaluation considers the likelihood and consequences of failure. Bridge safety is measured in terms of the risk level rather than the conventional failure probability level. In assessing risk to public safety, relevant factors such as the consequence of failure, structural system, indications of distress, possibility of hidden distress, bridge hits, extreme event data, traffic load history of the structure, and level of previous assessments completed should all be taken into account. Standards should provide guidance on appropriate inspections, safety assessment measures (load ratings, fatigue), intermediate mitigation measures (load posting, monitoring), and long-term strengthening or replacement strategies that may be used to manage the risks associated with structures.
The decision to take interim measures should be based on an assessment of the risks associated with the continued use of the structure without imposing any interim measures. The strengthening or replacement of all substandard structures is an ongoing process, and the work needs to be prioritized. Prioritization of strengthening or replacement should take into account the relative risk of each structure to public safety. A further enhancement would be to adopt a whole-life risk approach to maintain an acceptable level of risk over the life cycle of the bridge.
Of specific interest to scan team members was information the European hosts provided on assessment and prioritization of their existing bridge stock. The following is derived and combined from countries visited during the scan and attempts to capture their best practices in bridge assessment. It is described as an implementable process that includes several worthwhile concepts instead of presented as stand-alone ideas. This will be an iterative process.
Assessment of an existing structure should be carried out in stages of increasing complexity tied to the level of risk associated with the structure and with the objective of efficiently determining its adequacy. Early stages may contain conservative means of evaluating force effects. Provided that a structure is shown to be adequate at early stages, no further analysis is required. However, if a structure is found to be inadequate at an early stage and is considered to pose an unacceptable level of risk, assessment work should continue and later stages should seek to remove any conservatism in the assessment calculations.
The levels of assessment introduce increasing sophistication with the objective of assessing the safety of a bridge more accurately.
Each additional level of assessment may involve considerably more time and cost. The bridge owner should consider these implications and approve the progress of the assessment through the various levels.
Level 1 is the simplest level of assessment, based on a conservative estimate of load capacity. At this stage, only simple analysis methods are necessary.
Level 2 assessment involves the use of more refined analysis and better structural idealization. More refined analysis may include grillage or finite element analyses whenever these may result in more accurate capacities. Nonlinear and plastic methods of analysis may also be used for the substructure; actual measured material properties may be used for the superstructure.
Level 3 assessment includes the option to use bridgespecific live loading. Recent WIM data could be used to characterize truck load models (or calibrate load factors) specific to the site. Use of bridge WIM systems should be investigated on small structures because more accurate dynamic amplification factors can be obtained. Level 3 assessment may use material testing to determine characteristic strength or yield stress.
In Level 4 assessment, probability-based system methods are used in conjunction with an owner-specified level of safety. Such methods require in-depth knowledge of and expertise in reliability analysis techniques. (Levels 1 through 3 account only for element failures in bridge assessment. However, in many cases, element failures may not cause system failures. In other words, a bridge may have a smaller chance of failure than the corresponding system value.) A technical approval process should be implemented for the owner and assessment team to concur on the method of analysis and how the uncertainties of the specific bridge condition and the local traffic situation are considered. Structures believed to pose an immediate or high risk to the public may be candidates for a Level 4 assessment.
Note: In Level 1 and 2 assessments, extremes of normal traffic are represented by notional load models. Site-specific load models are used for Level 3 and 4 assessments.
Note: Traffic WIM data can be obtained by mounting sensors in the road pavement or on an existing bridge structure and estimating the corresponding static loads using appropriate algorithms. It is clearly desirable to collect as much data as possible, but 1 or 2 weeks of continuously recorded data may be sufficient for assessment purposes. It is important to ensure that these data are representative, so consideration should be given to seasonal variation patterns when scheduling a measurement period. The COST (European Cooperation in Science and Technology) 345 report does not specify the required accuracy of WIM data. However, some guidance is given by specifying the required accuracy with reference to the COST 323 WIM specification. Bridge loading is not overly sensitive to WIM system accuracy, and a system with accuracy that corresponds to about 95 percent of gross vehicle weights within 15 percent of the exact static value is considered sufficient. Extreme value distributions, such as those contained in the Gumbel family, are fitted to measured data recorded over a period of time. Subsequent extrapolation of these fitted distributions for a specified return period yields the characteristic value.
Note: Level 1 to 3 assessments, as described, are based on code-implicit safety levels, incorporating the nominal values of loads and resistance parameters and the corresponding load and resistance safety factors. To ensure that the assessment rules are simple for routine use, the format and values of the load and resistance factors are chosen to accommodate a wide range of structure and component types. Level 4 is a departure from these sometimes conservative assessment techniques.