16 Mar, 2018 By Mark Hansford and Emily Ashwell

A 862t bridge deck still under construction has collapsed onto a Florida highway, crushing cars and killing at least six people.  

The cable stayed bridge was being built in the city of Sweetwater by Florida International University (FIU) as a safety measure so that students did not have to cross the busy highway to get to and from its Modesto A. Maidique campus.

The bridge was to be cable stayed with one central pylon. The cables were to support a 862t deck post tensioned prestressed concrete deck structure with a bottom flange serving as the walkway and a top flange providing cover from bad weather. At time of collapse on Thursday afternoon the 53m long deck structure had been installed over the main highway, but the back span, spanning over a river, the pylon and the cables were yet to be installed.

It is possible that work was being carried out on the post-tensioning cables at time of collapse. Florida senator Marco Rubio has told local media that the bridge’s “cables had loosened” and that “engineers had ordered tightening”. UK experts said this could have nothing to do with the stay cables as they were not in place yet, and that the senator was referring to the post-tensioning.

The $14.2M (£10M), 53m long main span was built offline using a method being dubbed Accelerated Bridge Construction (ABC). It was moved into position by a rig in just six hours on Saturday 10 March.

The bridge is designed by Figg Bridge Engineers and was being built by Munilla Construction company (MCM). Barnhart Crane and Rigging operated the Self-Propelled Modular Transporters that placed the bridge on its permanent supports.

UK bridge experts who have spoken to New Civil Engineer have suggested compression failure caued by poor concrete in the top flange could be a cause. The Guardian newspaper has obtained and posted CCTV footage that captures the moment of the collapse which appears to shows the failure beginning in the top flange, close to the central pier.

Independent bridge consultant Simon Bourne said: “It has clearly failed in an un-stayed condition, but this was also (very clearly) a condition that it must have been designed to accommodate. So, with the back span, tower and cables not even built yet, it must either be a compression failure in the top flange due to poor concrete or a temporary works failure related to one of the temporary bearings.”

“While the CCTV footage is not hugely clear, it does seem to show a possible failure starting in the top flange. But the location is quite close to where the central tower is to be built, which is odd, as the stresses would be relatively low there.

“Overall, it looks like the top flange is prestressed as well as the bottom flange. But globally at this stage, the self-weight is clearly compressing the top flange, but any prestressing is actually compressing the bottom flange and reducing compressions in the top. I would estimate that a truss of this nature would probably have higher compressions in the bottom flange at this stage, which then makes the top failure look more odd, unless there’s some poor concrete in the top,” he added.

“Collapses are invariably due to poor concrete in the highly stressed zones, or poor concrete anywhere, in fact. So, poor concrete does look to be the main culprit, but I wouldn’t rule out the fact that the whole span might simply have slipped off its temporary bearings.”

He added that it would not be unusual for the contractor to be stressing some more prestressing cables in the bottom deck at the time, and that such activity is always tricky and potentially dangerous, as the loads are enormous - hundreds of tonnes per cable). But he said that this rarely causes a collapse.

”Even though you can get cable failures on rare occasions, it never usually causes a collapse,” he said. ”So, I get back to the same basic issues - it’s either a major compression failure in the concrete of the span somewhere, or it fell off its temporary bearings.

Bridge designer Ian Firth said it was too early to draw any firm conclusions, but one area to examine would be how far the construction methodology had been scrutinised by the designers.

He pointed out that following the collapse of a steel box girder bridge under construction in Milford Haven, South West Wales in 1970, the investigation by the Merrison Committee of Inquiry recommended that designers should supervise construction. This rarely happens today and may be a factor here, he said.

“Maybe it was something in the construction stressing sequences. Some have said that stressing was taking place at the time. How was the bridge being stressed? Did the stressing system fail? Did the jack fail? When you stress a strand, you first take the load on a jack, and the temporary system needs careful checking,” he said.

Firth added Figg was a well-known firm in the bridge design community.

The ABC method is being advanced at FIU’s Accelerated Bridge Construction University Transportation Center (ABC-UTC). 

MCM said the bridge was still under construction and it would undertake a full investigation.

FIU has previously said that the bridge was the first in the world to be constructed entirely of photocatalytic “self-cleaning” concrete. When exposed to sunlight, titanium dioxide in the Italian-developed cement catalyses chemical reactions that convert surface pollutants to soluble compounds that are washed away by the rain, reducing maintenance costs.

New Civil Engineer technical editor emeritus Dave Parker said this unusual concrete could also be a factor if its properties were not properly understood.

Figg Bridge Engineers said in a statement: “We are stunned by today’s tragic collapse of a pedestrian bridge that was under construction over Southwest Eighth Street in Miami. Our deepest sympathies are with all those affected by this accident. We will fully cooperate with every appropriate authority in reviewing what happened and why. In our 40-year history, nothing like this has ever happened before.

”Our entire team mourns the loss of life and injuries associated with this devastating tragedy, and our prayers go out to all involved.”

Florida bridge artists impression close up

I suspect failure occurred in the junction between the lower end of the diagonal and the deck at the tower end. The horizontal force component would be very high in this temporary state, possibly high enough to shear the connection to the deck. If it’s true they were trying to adjust the prestress in that diagonal, that may well have contributed. All pure speculation of course. By the way, the view you show of the finished bridge is incorrect. That was an initial design submitted at the planning stage. It’s a pity it wasn’t the one built.

Sad news. Certainly it won’t do any good to our profession. Footage at 1:26 suggets a punching failure too, perhaps not enought concrete depth on the top flange.

Viewing the video indicates failure being initiated by a hinge forming in the top slab at its node with the inclined member. This, in turn, suggests that the end reaction of the deck weight was being transferred through the stub tower to its connection with the top slab – causing that member to act as a cantilever (something I doubt was intended by the designers). This could only occur if the connection between the bottom slab and the foot of the stub tower was insufficient to generate truss action.

Looks like a sequencing problem, and mismatch between designer’s intentions and construction procedure. The span as installed was carried at third points on trestles on transporter (primarily hoggng), then changed to a full span (sagging), then future intention to support by an array of cables (fairly neutral in terms of hoging or sagging). Reports of cable tensioning works at time of incident imply adjustment to pretensioning of prestressed concrete bottom and/or top slab. A highly sophisticated load-transfer procedure which went horribly wrong.

Than you for your great writing. Our company, IPC is transforming 50 year old manual inspections to bring modern technology and robotics to infrastructure inspections nor just for post tensioned tendons but for infrastructure inspections worldwide. https://www.infrastructurepc.com/a-post-tension-primer/

Sign in or Register a new account to join the discussion.