Monday, March 19, 2018

The FIU Bridge Collapse: More Questions Than Answers

On Thursday, March 15, workers on a newly installed span of a pedestrian bridge across the busy seven-lane Southwest Eighth Street that divides Florida International University (FIU) from the adjacent town of Sweetwater were adjusting some tensioning cables in the span.  The Saturday before, Barnhart Crane Rigging Company had lifted the span from the side of the road where it had been built over the preceding months, carried it a short distance along the closed street, and lowered it into place successfully.  FIU officials were proud that a novel technique called accelerated bridge construction (ABC) was being used for the bridge, because the school itself has a center that promotes and studies that technique.  Instead of blocking traffic for weeks or months while hazardous lifting operations put a bridge into place piece by piece, workers using accelerated bridge construction build the bridge offline, so to speak, and then shut down traffic only briefly as entire spans are lifted into place. 

The technique has been used frequently in the last few years with no major problems, so no one expected any issues this time.  As it turns out, though, those expectations were disappointed.

At 1:47 PM Thursday, without apparent warning the entire bridge collapsed onto the busy highway, crushing cars and killing at least one construction worker who was on the bridge at the time.  Six people died in the accident. 

Some of the questions raised by this tragedy can't be answered yet.  For example:

Why wasn't the road closed during an operation such as cable tensioning that might have endangered the stability of the bridge?

The whole idea of ABC is to keep transportation going as much as possible.  Evidently the engineering firm which designed the bridge, FIGG Construction of Tallahassee, had reason to believe that the span was going to support itself safely during tensioning operations, so no one issued orders to close the road. 

What is the significance of some cracks that one engineer reported seeing on Tuesday, two days before the collapse?

Right now, we don't know.  Cracks in concrete mean that wherever the crack shows up, tensile (stretching) forces locally exceeded compressive (squeezing) forces.  Concrete is a material that can withstand a lot of compression, but hardly any tension on its own.  That's why large concrete structures contain carefully designed and placed steel reinforcement such as "rebar" and sometimes cables, which were evidently used in the ill-fated structure as well.  Some cracks are only skin-deep, so to speak, and do not indicate a structural problem, just a cosmetic one.  Others may go all the way through a structure and are signs of a major problem.  Until forensic experts piece together the remains of the 174-foot span and figure out where the trouble started, we won't know whether the cracks were superficial or significant.  At a meeting of construction personnel Thursday a couple of hours before the collapse, the cracks were discussed and the consensus was that they were not a safety issue.

If the final design of the bridge included a tower and suspension-bridge-like pipes connecting it to the span that collapsed, why weren't those parts of the structure in place before the road under the span was opened to traffic?

I am not a mechanical or civil engineer, and so I'm strictly an amateur compared to someone with professional training in those fields.  Reportedly, the university website about the bridge stated that the tower and suspension pipes were not needed for static support, and were there simply to cut down vibrations and add esthetics.  Allow this amateur to beg to differ. 

A sketch of the intended completed bridge can be viewed here, and shows that several concrete struts or trusses that connect the concrete roof of the bridge to the lower walkway part are straight in line with the planned suspension pipes.  It certainly looks like the trusses were intended to carry the weight of the walkway through tension (probably by inner tension cables) up through the roof to the suspension pipes.

It's possible (more than possible, if the FIGG engineers knew what they were doing) that even without the tower suspension structure, maintaining proper tension on cables inside the trusses would keep the whole concrete structure in compression sufficiently to counteract the tension that the walkway part would experience once it was put in place over the roadway.  What may have happened (and this is purely my speculation) is that when a construction worker began to adjust one of the tension cables, he might have done something as simple as turning a nut the wrong way.  Such an action might have sent an already marginally stable structure over the edge of failure, and once such a delicately balanced system has one part fail, the rest of it goes too.

I do not envy the work of forensic engineers who now have to transport the messy wreckage somewhere so they can pore over every identifiable piece, figuring out what was where and what the exact positions of tensioning adjustments were.  From such small details a picture should emerge that will let us figure out what went wrong last Thursday that led to such a dismal outcome of something that was supposed to be a point of pride.

In the meantime, this disaster should serve as a warning for every construction firm doing accelerated bridge construction.  Maybe we should ease off on the accelerator a little, at least until we find out what happened at FIU last week.

*NOTE added after posting:  A comment posted by the Happy Pontist, a professional bridge designer, contradicts my admittedly amateur opinion that the tower and suspension pipes might carry a load.  He confirms what the university website claims, which is that they were for vibration damping, and their alignment with the truss members was for esthetic reasons only.  Thanks to the Happy Pontist for this correction. 



  1. More on this discussion over at I seriously doubt that Figg altered their preliminary design to use the stays as a strength component. Photos of the connection to the top of the truss show bolts very similar to what is shown on the preliminary design drawings. The key issue is stiffness: it's hugely unlikely that the stays could ever be made stiff enough to carry a useful share of load: their extension under axial load will greatly exceed the bending deflection of a truss as deep and stiff as this one.

  2. Three points:
    1) To build a 25 m high concrete tower and 200 m 16” pipe stays wih no statical purpose seems to be very exaggerated
    2) The fist north diagonal which is supposed to be under bar tensioning at the instant of collapse is designed in the project as “no post tensioning bar”. And that is reasonable since the diagonal is stressed by a very hard compression , not traction.
    3) A very heavy crane with hook hanging just over the collapsed node can be seen in the video of the falling bridge. You don’t need a several hundred tons crane to tighten tensioning bar! What was the crane for? The crane can be seen in the very first picture after the collapse and then in a few minutes it disappears . Why so an hurry ???