The FIU Bridge Collapse. April 5, 2018

Copyright © 2018                                      John F. Oyler

April 5, 2018

The FIU Bridge Collapse

The abrupt collapse of the FIU bridge has generated a lot of excitement in the structural engineers’ community. If it weren’t for the fact that at least six people lost their lives, I suspect we would be happy that it occurred and provided us with the opportunity to act as amateur forensic investigators and attempt to understand what went wrong.

Actually, as a structural engineer, I am embarrassed on behalf of our profession. We are obligated to provide safe structures for their users and should be technically capable of doing that.

The first problem was sorting through the available information and determining what was correct and what was misunderstanding. An early headline was the apparent oxymoron, “Fourteen million dollar pedestrian bridge collapses”.  

Sure enough, some well-meaning persons suggested the construction of a bridge over the busy Tamiami highway to permit students at Florida International University to safely reach their living quarters in the Sweetwater neighborhood. It seemed like a reasonable suggestion, especially after a student was killed crossing that busy artery.

A conventional pedestrian bridge ten feet wide would have been adequate to satisfy this requirement and could be constructed for less than two million dollars. One of my colleagues at Pitt, Dr. Max Stephens, was able to acquire a copy of the successful proposal for this project; a brief review of it easily explains the escalation in cost.

In retrospect, it is tempting to be sarcastic about all the ironies comparing the language in the proposal with the photographs of the collapsed structure; we will attempt to resist the temptation. Turns out it is not just a pedestrian bridge after all – it is an “Event Venue, Linear Park, and a Place to Linger and Gather”, with a platform thirty feet wide.

Page three of the one hundred seventy three page document highlights the provision of “hammock hooks between trusses” for students “resting on the bridge”. To be fair to the proposal writer we must point out that these are “Future amenities not included in the base bid”.

Another early report blamed the collapse on the fact that the bridge that failed was incomplete, merely a portion of a cable-stayed bridge that was positioned over the busy highway before the cables were installed. The renderings of the finished structure do indeed make it appear that it is a legitimate cable-stayed bridge, which gives this theory credibility.

A review of the proposal found no instance of it being called a cable-stayed bridge. In reality, the pylon and the sloping stays apparently are cosmetic; the truss that failed was designed to support the entire load independently. But, the “Contemporary iconic cable structure provides an aesthetic gateway attracting people to enjoy a unique experience.”  So much for that theory.

In truth, the proposal attempts to justify the pylon and stays as providing stability to prevent resonant vibration induced by pedestrians walking in step across the bridge. I would like to see the calculations that support that conclusion. It would take a lot of soldiers marching in step to excite a bridge weighing nine hundred and fifty tons.

Another potential villain identified as the culprit was the “Accelerated Bridge Construction” (ABC) concept, an approach to construction which has been implemented throughout North America in recent years. Ironically, FIU’s University Transportation Center has been the focal point in its development.

For this project, the ABC approach was used to minimize shutting down the eight lanes wide Tamiami Trail. The main span is a concrete truss, one hundred seventy-four feet long, with a canopy sixteen feet wide as the top chord and a deck thirty feet wide as the lower chord. It was formed, cast, and cured in a staging area close to its final location and weighed nine hundred and fifty tons when finished.

Five days before the collapse it was jacked up, moved to the piers that would support it by four self-propelled modular transporters (SPMT), and set in place, during a shutdown of the highway for six hours. Certain critics are convinced that something in this process created problems that led to the disaster and are advocating termination of the ABC approach. We have, so far, found no evidence of this.

The “high performance” concrete in the truss has an ultimate compressive strength (8500 psi) more than double that of conventional concrete and should have been adequate for this application. Nine hundred and fifty tons of concrete would have required about two hundred mixer truck loads; it is easy to wonder if they were able to achieve consistency throughout such a large pour.  

Other critics have focused on the innovative, “leading-edge” technology attributed to the design and construction of the bridge and are advocating reverting to “tried-and-true” methods. It is certainly true that concrete trusses are rare, because of concrete’s weakness in tension. Normally this is overcome by the addition of reinforcing steel to carry the tensile loads. On the FIU bridge the designers elected to employ a technology called post-tensioning.

Before a concrete component that is intended to be post-tensioned is cast, conduits big enough to permit the passage of high strength steel rods called tendons are positioned in the formwork. After the concrete has been cured and acquired its desired strength, tendons are inserted and attached to bearing plates at end of the component. A large tensile force is applied to them, which then compresses the concrete, generating high compressive stresses in it.

As a result, any tensile load applied to the component will merely reduce the compressive stress in it, avoiding failure by tension. This is mature technology, although applying it to diagonal members in a concrete truss may well be novel. There is a report that workers were adjusting the load in tendons near the point where the truss began to fail, at the time of the collapse.

The video record of this move seems to indicate that one of the transporters was located eleven or twelve feet away from a truss panel point, leading to conjecture that this induced longitudinal bending into the deck, which would have produced additional sufficient additional tension in the outer fibers of the deck, exceeding the compression from the post-tensioning. An interesting possibility that could only be investigated by having information currently unavailable.

It is well documented that cracks had been observed in that area a day before the collapse. Someone has postulated that this adjustment was an effort to close up the cracks. Another report is that someone heard a loud “pop” just before the collapse, leading to speculation that one or more tendons had fractured.

Another major criticism is the current use of the “Design-Build” concept. Traditionally major public works projects were implemented in two distinct phases. The owner hired an engineer to design a bridge, then hired a construction firm to build it. This put the owner in the middle of disputes between the designer and the builder.

Consequently owners, primarily governmental agencies, have embraced the idea of single responsibility, which they call “Design-Build”. Based on the proposal document it appears that the responsibility for this project falls on the construction firm, despite the appearance that it is a joint venture between designer and builder.

Our senior design class at Pitt has been studying responsibility and accountability for projects of this type. This specific disaster will certainly be an appropriate case study for future classes. Who is responsible for constructability? Should the builder be required to have licensed engineers on its staff if it is accountable for a successful project?

The intriguing thing about this problem is the fact that it encompasses a large number of technology topics that are of great interest to us. Is this a truss with abnormally wide flanges, or is it a massive flexural member with an abnormally open web? Does shear lag play a part in transferring load from the truss/web into the deck and canopy? How does the distortion energy failure apply in compression? To paraphrase Yul Brunner, “et cetera, et cetera, et cetera”.

Now that the National Transportation Safety Board team is investigating this tragedy, much of the information is no longer available to the public and will not be released for many months. The litigious environment that pervades our society is a major deterrent to access to real data, thus encouraging the dissemination of unfounded rumors and ill-founded theories.

Our attempts to evaluate the design, of course, are based on information available in the proposal; we have no access to “as-built” drawings. Nonetheless the proposal design does reflect the original intent of the designer and his/her capability.

My personal, albeit ill-founded, speculation is that the connection of diagonal to the deck at the pylon end sheared off, triggering the collapse. This is based on an early photograph that I recall seeing, though I can no longer locate it. I acknowledge that could well have been a result of the collapse, rather than its cause. I suspect I will agonize over this for months till we get the NTSB report and then I will disagree with it. By the time this column is published I will certainly have gone down several false alleys and encountered dead ends.

Regardless of the specific error that someone made, I am convinced that the overall approach to this project is the true culprit. Somehow the team charged with the responsibility for building a simple, functional pedestrian bridge lost its focus and became enamored with the opportunity to create an icon, emphasizing aesthetics and a commitment to sustainability.

The massive proposal document goes into remarkable detail describing non-technical characteristics of the project. Want to know how deep the mulch should be to plant a cigared sabal palm tree? Go to page 93 of the proposal. Want reassurance that the project meets Criterion PD-15 (Historic, Archaelogical, and Cultural Preservation) so it can achieve a LEED Platinum rating for sustainability? Go to page 66. Want to determine the specification for the tendons that make the concrete useable? Good luck – it’s got to be there somewhere, but I haven’t found it yet.

It is ironic that some of us are obsessed with the problems of funding infrastructure construction and maintenance while others are spending fourteen million dollars on what could have been a routine pedestrian bridge.

For me the moral of the story is an old one. Let’s get back to fundamentals and a comprehension of basic mechanics and get things right, then worry about all the frills. Let’s revert back to the first Canon of the Engineers Code of Ethics, “Engineers shall hold paramount the safety, health and welfare of the public”, and think about aesthetics and sustainability later.