Calibration of AASHTO LRFD Concrete Bridge Design Specifications for Serviceability (2014)

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.

1 BACKGROUND AND RESEARCH APPROACH 1.1 Background The AASHTO LRFD Bridge Design Specifications (AASHTO LRFD) represented a refinement of bridge design practices as compared to past American Association of State Highway and Transportation Officials (AASHTO) specifications. A primary goal of using the Load and Resistance Factor Design (LRFD) philosophy is to achieve uniform reliability which can be achieved through statistical calibration. However, during the development of AASHTO LRFD the service and fatigue limit states were not statistically calibrated. These limit states were calibrated against previous AASHTO design requirements to achieve component proportions comparable to past practices. This process does not achieve uniform reliability. This project was initiated to address the lack of statistical calibration of the service and fatigue limit states for the design of concrete structures in the AASHTO LRFD. 1.2 Special Challenges Related to Service Limit States The Strength Limit States (ULSs) of the AASHTO LRFD are calibrated through structural-reliability theory to achieve a certain level of safety. They are intended to achieve similar component proportions to those of the Standard Specifications for Highway Bridges. These ULSs consider the extreme loads the bridge is expected to be subjected to during its design life. They do not consider the integration of the daily, seasonal, and long-term service stresses that directly affect long-term bridge performance and subsequent service life. The current Service Limit States (SLSs) of the AASHTO LRFD are intended to ensure a serviceable bridge for the specified 75-year design life. These limit states are based upon the traditional serviceability provisions of the Standard Specifications for Highway Bridges. These SLSs are not calibrated using reliability theory to truly achieve a determined life with a specific level of certainty as the tools and data to accomplish this calibration were not available to the AASHTO LRFD code writers. The background on the current AASHTO LRFD SLSs are presented in Chapter 2 and in Appendix A. Some of these SLSs may relate to a specified design life; others do not. Many are presently very deterministic, such as some owners’ wish to limit the tensile stresses in prestressed concrete components to a level which is thought to result in a crack-free component. This SLS could be calibrated to achieve a certain probability of a crack-free component, but this calibration would include a service life only in the determination of the live load the component must resist, for example, a 75-year live load. To achieve the objective of developing the appropriate tools, candidate SLSs have to be evaluated against a set of criteria. This applies both to the retention of some of the existing SLSs in the AASHTO LRFD, as well as any new limit states which may be developed as part of this project or in the future. The criteria include: • Is the limit state quantitatively and qualitatively meaningful? – Does the limit states tell something that can be used to maintain a structure in service and continue or extend its service life? 3

• Can the limit state be calibrated? – Can limit state functions be developed and can a means be developed to determine the data necessary to perform a calibration? Where no such data exists, expert elicitation (Delphi process) may be useful in at least determining a range of data and the relative importance of certain characteristics in the data, including uncertainty, so that some calibration can proceed. Consideration of SLSs requires a different input data than the ULSs. In ULSs, the limit state function is defined with two variables, resistance, which was considered constant in time, and loads. For SLSs, a different approach is needed because: • The definition of resistance can be very difficult. • Defining the acceptance criteria is difficult as exceeding the set limits for the service limit state does not necessary lead to immediate change in the resistance or the performance. • Acceptable performance can be subjective (full life-cycle analysis is required). • Resistance and load effects can be and often are correlated. • Load is to be considered as a function of time, described by magnitude and frequency of occurrence. • Resistance and loads can be strongly affected by quality of workmanship, operation procedures and maintenance. • Resistance can be subject to changes in time, mostly but not only deterioration, with difficult to predict initiation time and time-varying rate of deterioration (e.g. corrosion, accumulation of debris, cracking). • Resistance can depend on geographical location (climate, exposure to industrial pollution, exposure to salt as deicing or proximity to the ocean). In general, the consequences of exceeding SLSs are an order, or even several orders, of magnitude smaller than those associated with ULSs. Therefore, an acceptable probability of exceeding a SLS is much higher than for a ULS. If the target reliability index for ULS is βT = 3.5 to 4.0, then for SLS, βT = 0.0 to 1.0 might be quite acceptable. Further to special challenges, it may be found that changes to materials or construction practices are the more effective way to deal with what appears to be a serviceability issue. This could be the case, for example, with deck cracking where changes to mix proportions and/or curing practices designed to reduce shrinkage may be as effective as limit states based on strain calculations. 1.3 Problem Statement and Research Objective The objectives of this research, as stated in the project’s Request for Proposals (RFP), are to develop new concrete service and fatigue limit states as needed, calibrate new and existing concrete service and fatigue limit states, and prepare specifications and commentary for consideration for adoption by the AASHTO Highway Subcommittee on Bridges and Structures. 1.4 Scope of the Study The scope of the study was generally determined by the tasks identified in the RFP as the tasks anticipated to be encompassed by the research. The task description, copied from the RFP, is provided below. 4

Task 1. Collect and summarize information on design practices and research findings related to concrete bridge serviceability, including fatigue. Review and evaluate state Department of Transportations’ (DOTs’) technical issues pertaining to application of the AASHTO LRFD Bridge Design Specifications to serviceability. The details of owner- specified special design and permit vehicles shall also be collected. This information shall be assembled from technical literature; case histories; and unpublished experiences of engineers, bridge owners, and others. Review pertinent serviceability criteria in structural design specifications from other countries and disciplines. Task 2. Review the relevance of existing service and fatigue limit states and identify any new limit states needed to accommodate loading and performance criteria. Provide recommendations on new limit states to be developed and calibrated and the existing limit states to be calibrated. Task 3. Identify concrete bridge member designs used by state DOTs and select representative designs for calibration. Consider the variability in design and construction practices in the design cases. Task 4. Define the calibration procedure to be used for concrete bridge service and fatigue limit states. Task 5. Identify statistical data needed for calibration of the limit states (type, quantity, and quality). Identify data sources and assess their applicability to reliability-based calibration for the limit states. Task 6. Develop an updated and detailed work plan to complete Tasks 8 through 11. Task 7. Submit an interim report describing the findings of Tasks 1 through 6 for panel review. National Cooperative Highway Research Program (NCHRP) review and approval of the interim report will be required before proceeding with the remaining tasks. The contractor should anticipate meeting with the project panel to discuss the proposed work plan. Task 8. Following approval of the work plan submitted in Task 7, assemble the databases and determine the statistical parameters required for the calibration. Task 9. Determine target reliability indices for each of the service and fatigue limit states. Submit the recommended values with justification in a letter report for NCHRP review and approval. Task 10. Using the approved target reliability indices and calibration procedure, calculate service and fatigue limit state load and resistance factors for concrete bridges. Task 11. Draft specifications and commentary with complete details for panel review and comment. Prepare design examples that demonstrate the application of the proposed specifications. Task 12. Revise the specifications, commentary, and examples consistent with panel review comments. 5

Task 13. Prepare a final report documenting the entire research effort. The recommended specifications, the calibration procedure, calculations and data, and the design examples shall be included as appendices to the report. 1.5 Relationship to Project SHRP R19B Several members of the research team were also involved with the Strategic Highway Research Program (SHRP) project SHRP R19B; Bridges for Service Life Beyond 100 Years: Service Limit State Design. The two projects were running concurrently. The goal of the SHRP R19B project is to develop framework for calibration of the service and fatigue limit states in the AASHTO LRFD. SHRP R19B scope included applying the calibration procedures to the existing service limit states and, where need exists, develop new calibrated limit states for incorporation in the AASHTO LRFD for a variety of bridge materials, foundations, joints and bearings. With the broader scope of R19B, significant overlap exists between the area of concrete structures in SHRP2 R19B and this project. All aspects of the development of the live load models, calibration processes and the application of the calibration process to concrete- related limit states are fully applicable to both projects. The work on the live load model and the basic calibration process were developed jointly by the two projects. Significant portions of the calibration process for the fatigue limit state were originally developed in SHRP R19B and were incorporated herein. 1.6 Research Approach To accomplish the stated objectives of the research and to cover the work on the tasks of the project, the following approach was followed: • Existing AASHTO LRFD and some other major bridge design specifications were reviewed to identify existing service and fatigue limit states. • To determine the current state of practice of design for the SLSs, a questionnaire was developed and sent to major bridge owners across North America including all 50 states Departments of Transportation and other major authorities. The questionnaire included questions covering design loads, design provisions and general questions related to the performance of bridge structures in service. • An extensive literature search was performed to identify and review relevant past research and the background of existing service and fatigue limit states. • An extensive set of weigh-in-motion (WIM) data was assembled and analyzed to determine the live load model appropriate for use for fatigue and SLSs. • The SLSs appropriate for calibration and the variables to be included in the statistical calibration for each limit state were identified. • Databases of existing and simulated bridges were developed for use in the calibration. • A calibration process was developed. • The inherent reliability indices of existing and simulated bridges were determined and a target reliability index was selected for each limit state. • The calibration process was refined and the load and resistance factors required to produce the target reliability indices were determined. • Proposed revisions to the design specifications were developed. 6