Fracture-Critical System Analysis for Steel Bridges (2018)

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.

4 1.1 Problem Statement and Research Objective The adoption of the first fracture control plan (FCP) in 1978, as the structural engineering community’s response to the 1967 collapse of the Silver Bridge connecting Point Pleasant, West Virginia, and Gallipolis, Ohio, resulted in the first specifi- cation targeting design for fatigue and fracture in steel bridges (AASHTO 1978). The objective of the FCP was to avoid fatigue and fracture in steel bridges, particularly in structures in which redundancy was not demonstrated. This subsequently led to the development and application of minimum fabrication, design, and inspection requisites. The 1978 FCP has evolved over time. Currently, fabrication requirements are specified in the AASHTO/AWS D1.5M/D1.5 Bridge Welding Code (AWS D1.5) (AASHTO and AWS 2015), design requirements are specified in the AASHTO LRFD Bridge Design Specifica- tions (AASHTO LRFD BDS) (AASHTO 2014), evaluation procedures are described in the AASHTO The Manual for Bridge Evaluation (AASHTO MBE) (AASHTO 2011A), and inspection requirements are specified in the National Bridge Inspection Standards (NBIS) as a part of the Code of Federal Regulations (CFR) (FHWA and U.S. Department of Transpor- tation [U.S. DOT] 2017). The specific requirements in each document, however, are not explicitly integrated. One of the most significant effects of the FCP is the cre- ation of the fracture-critical member (FCM) term. An FCM is a nonredundant primary steel tension member; its fail- ure will likely result in the collapse of the bridge or loss of serviceability (AASHTO and AWS 2015, AASHTO 2014, AASHTO 2011A, FHWA and U.S. DOT 2017). FCMs are required to be fabricated in accordance with Section 12 of the AWS D1.5 and must undergo FCM hands-on inspections every 24 months (AASHTO and AWS 2015, AASHTO 2014, AASHTO 2011A, FHWA and U.S. DOT 2017). Although the fabrication requirements of FCMs are a moderate initial cost premium, the additional inspection requirements constitute very high expenses to be added to the life-cycle cost of the structure. These requirements discourage the construction of several steel bridge configurations, such as two-girder, tied- arch, or truss bridges. Further, despite their excellent service record, the public perception is that these bridges are unsafe as the concept of fracture criticality is typically associated and even confused with “structural deficiency.” However, there are multiple cases of steel bridges consid- ered to be nonredundant that underwent failure of a primary steel tension member and did not result in collapse or loss of serviceability. For example, the Neville Island Bridge in Pittsburgh, Pennsylvania (Fisher et al. 1980B); the Lafayette Bridge in Saint Paul, Minnesota (Fisher et al. 1977); the I-26 Bridge over the Green River in Henderson County, North Carolina (McGormley et al. 2000); US 442 Bridge over the Schuylkill River in Pottstown, Pennsylvania (Kaufmann et al. 2004); and the Diefenbaker Bridge in Prince Albert, Saskatchewan, Canada (Ellis and Connor 2013) are two- girder steel bridges that experienced failure of a girder and continued to carry their dead load and some level of live load. Other types of FCMs have failed without resulting in collapse or loss of serviceability, such as the pier caps (or bents) of the Dan Ryan Transit Structure in Chicago, Illinois (Fisher et al. 1980A) or the lower tension chord of the Mathews Bridge in Jacksonville, Florida, which was severed by a ship (Judy 2013). The aforementioned structures were considered to be non- redundant with tension members designated as FCMs, based on the configuration of the bridge. In other words, an analysis was not performed to determine whether the failure of a primary steel tension member could lead to collapse or loss of service- ability. Although design and evaluation procedures for fatigue are defined in the AASHTO LRFD BDS and the AASHTO MBE, the same cannot be said for the evaluation of fracture and redundancy (AASHTO 2014 and AASHTO 2011A). Typi- cally, the fracture limit state is handled through fabrication and material requirements, but structural engineers do not have any codified guidance to design for or evaluate redundancy. C H A P T E R 1 Background

5 In 2012, the FHWA issued a memorandum that re-stated and clarified the definition, fabrication, and inspection require- ments for FCMs (Lwin 2012). Additionally, the term “system- redundant member” (SRM) was introduced and defined. An SRM is a member that would traditionally be designated as an FCM but, through refined analysis, its failure has been shown not to result in collapse or loss of serviceability. However, as indicated in the memorandum, there are no redundancy evaluation procedures, and the requirements of such refined analysis are to be agreed upon between the owner and the engi- neer. Additionally, the memorandum states that SRMs require fabrication—according to Section 12 of the AWS D1.5—but do not need to be considered as FCMs for in-service inspec- tion. The inspection requirements of SRMs are still to be set. Based on experience revealing that structures with FCMs have shown redundant capacity after the failure of an FCM and the need for analytical procedures to assess redundancy in steel structures, the objectives of this research are to (1) develop a methodology to establish whether a member is an FCM or an SRM, based on loads, existing conditions, material properties, and bridge configurations; and (2) recommend specifications to AASHTO, based on such methodology for design of new bridges and evaluation of existing bridges. 1.2 Scope of Study The research conducted for NCHRP Project 12-87A, “Fracture-Critical System Analysis for Steel Bridges,” summa- rized in the current report, focused on the development and application of a methodology to evaluate system-level redun- dancy in steel bridges. As such, the analytical framework, load models, and evaluation criteria that have been developed are based on the assumption that the failure of a primary steel tension member takes place. In other words, it is not the intent of the methodology to provide guidance for modeling the ini- tiation and progression of the mechanism that produces the member failure. Rather, the intent is to address whether the fail- ure of such a member results in collapse or loss of serviceability. One tenet of this work is that the need for fabrication require- ments of new FCMs and SRMs must be considered differently than the need of inspection requirements. The fabrication and material requirements of Section 12 of the ASW D1.5 invoke procedures that reduce the probability of fatigue and fracture failure by providing higher toughness materials and quality control standards (AASHTO and AWS 2015). Even if system analysis shows that member failure will not cause collapse or loss of serviceability, fracture is still a limit state that should be avoided. For new bridges, fabrication of FCMs and SRMs per Section 12 of the ASW D1.5 is still beneficial to the life- cycle performance of the structure. However, the methodology is intended to extend the current applicability of the memo- randum published by the FHWA in 2012. At present, the 2012 FHWA memorandum only permits system analysis to those bridges that were fabricated to the FCP (Lwin 2012). The pro- posed methodology was crafted specifically with the objective that it could be applied to bridges, regardless of whether they were fabricated in accordance to the AASHTO AWS D1.5 FCP. Thus, the 2012 FHWA memorandum would need to be revised to permit the full use of the provisions proposed herein. A second tenet is that the analysis methodology is only to be applied to bridges in “satisfactory” condition. Although the analysis methodology may be used to investigate the conse- quences of member failure in a problematic structure, the rec- ommendations from the research are not intended to be used to defend relaxation of the existing inspection requirements in structures with corrosion, cracking, or other maintenance problems. In the same way, structures with questionable con- nections or details—for example, those susceptible to con- straint-induced fracture (CIF)—must not have their inspection requirements lessened as a result of the application of the rec- ommendations in the current report. The evaluation of system-level redundancy requires con- sideration of all possible load paths, including mechanical phenomena not typically considered in design (e.g., partial composite action). Hence, an analysis methodology based on traditional structural analysis methods, such as line-girder analysis, is not practical and cannot be reliably applied. Instead, a finite element analysis (FEA) methodology was developed. As the FEA methodology must capture nonlinear behavior, alternative load transfer mechanisms, and other complex phenomena, the engineer must have previous proven experi- ence with finite element models of multicomponent assem- blies. Additionally, engineers must take into account that load and resistance factor design (LRFD) procedures employed in design and evaluation of bridges are not directly applicable to evaluation procedures that require or used nonlinear FEA. Therefore, engineers must be able to understand the load combinations and performance requirements proposed so that they may apply them to redundancy evaluations that follow the described FEA methodology. Finally, the proposed analysis requirements and recom- mendations are complementary to any governing design and evaluation specifications. The proposed guide specification is not developed to substitute all of the provisions in the AASHTO LRFD BDS or the AASHTO MBE. A redundancy evaluation per the FEA methodology described in the current report is to be used solely for the identification of FCMs and SRMs. The bridge must still meet all applicable limit states in the nonfaulted condition. The results from the system analysis may be used to modify inspection requirements at the dis- cretion of the owner, with the approval of the appropriate agency, and in agreement with the NBIS in the CFR.