News Detail

FIU Bridge Collapse—A Tragedy at Many Levels

1/2/2020

I suspect that almost everyone in the concrete industry has heard about the March 15, 2018, collapse of the pedestrian bridge that was under construction over a busy roadway at Florida International University (FIU) in Miami, FL. The collapse killed six individuals and injured at least 10 others. The National Transportation Safety Board (NTSB) issued its report detailing the probable cause of the collapse on October 22, 2019.

The full text of the report can be found at:

www.ntsb.gov/investigations/AccidentReports/Reports/HAR1902.pdf.

In full disclosure, I was retained as an expert for one of the parties, but in the end, I did not provide any opinions to the attorney for that party.

The NTSB wrote a blistering account of the failures by various parties. It stated that the probable cause of the collapse was design errors by the Engineer of Record (EOR) and that contributing to the collapse was an inadequate peer review that failed to detect the calculation errors in the bridge design. Severe cracking was observed during construction in the nodal region at the intersection of concrete truss members, and this was characterized on multiple occasions by the EOR not to be significant. In fact, at the time of the collapse, the structural member in question was being retensioned at the direction of the EOR, presumably in an effort to impart compression in the nodal region of the cracking. Notably, the NTSB also stated that the severity of the outcome of the collapse was due to the failure of various parties, including the EOR, the inspection engineer, the contractor, the university, and the Florida Department of Transportation (FDOT).

Further, the NTSB noted that the EOR underestimated the demand that was imparted in the nodal region at the intersection of the truss members, and the NTSB faulted the EOR for not computing the capacity of the nodal zone where the concrete truss element intersected the bottom flange. In load and resistance factor design of structural concrete such as used in the ACI 318 Structural Concrete Building Code and in the AASHTO Specifications, an underestimation of demand and an overestimation of capacity drives the capacity-to-demand ratio (C/D) toward a value of 1.0. A value less than 1.0 indicates that there is a high probability of failure of the structural element. Without a redundant structural system that would allow a redistribution of forces, structural collapse will occur. The structural system used for the FIU concrete truss pedestrian bridge had no redundancy that potentially would have allowed a redistribution of forces in the nodal region that failed.

Various news reports and even the NTSB suggest that the FIU pedestrian bridge was “complex” in accordance with the FDOT requirements for the project. However, the FIU bridge was not a complex structure. I will agree that the bridge design was unique, as noted in the NTSB report, because concrete was being used for a truss in which some members are in tension and others are in compression. What is clear is that the significant cracking and distress that had been noted by various parties during construction of the nodal region presaged imminent collapse. As documented in the NTSB report, the cracking observed by various parties was the type and magnitude observed when failure occurs.

While the EOR, the peer-review engineer, and the Federal Highway Administration on behalf of NTSB all used structural analysis programs to facilitate the rote determination of forces in the members for the multiple load cases that needed to be considered, the forces can be determined by simple hand computations. This truss bridge is an exemplar of using the strut-and-tie method as detailed in Chapter 23 of the ACI 318 Structural Concrete Building Code. There are very detailed requirements in ACI 318 concerning design of nodal zones, and had these provisions been followed, there would not have been a failure of the FIU bridge.

The FIU bridge project was clearly a failure at many levels, including an apparent lack of communication between all parties involved and inadequate focus on the safety of workers and the public during critical stages of construction. Although ACI has published numerous papers and special publications and held seminars specifically on the utility of the strut-and-tie method since its introduction into the 318 Code in 2002, it apparently has not been enough. ACI and universities need to make a significant effort to better educate engineers on the strut-and-tie method, especially in the detailing of nodal regions, and practicing engineers need to step up and learn the method.

Randall W. Poston

ACI President