Miami Pedestrian Bridge, Part III 99

[JStephen]

"Site management has requested that we limit the length of the other thread by forming a new one. This subject may require III, IV...."
So here's Part III. Please don't post any more in Parts 1 and 2.

Part I thread815-436595
Part II thread815-436699

 

[jrs87]

VSL employee was tied off to horizontal safety line. The safety system became wrapped around the exposed post tension rod. See attached photos. This unfortunate worker seems to have been flung by this while bridge fell. It looks like he dangled from crane because vertical crane line held up the horizontal safety line.

The north end fell fast in part because the center of gravity falls at 1 g, but like a falling smoke stack, the free end goes faster. The marks on the pylon made by the falling deck may help with investigation. The marks begin well below the top of pylon base.

The central web truss does not look right, it looks more like a concrete canopy support system would look in a walkway/plaza on solid ground, not a support system for a 174 foot span. It looks like they were going for clear vistas from sides of walkway, so they put only one web in the middle. It's sad, but FIU/MCM put in writing (the proposal) all the constraints that added together doomed this project. I.E. ABC, a pedestrian overpass that resembles a magistic full size cable-stayed bridge, concrete to be cool in summer, wide ribbon deck...

This bridge will not be completed with this design, you can bet the new design will be highly redundant.

I will keep on open mind until report from NTSB is issued, but I must say now, I'm shocked engineers did not block this project and even more shocked rod adjustment was done while traffic was below.

This bridge failed because deck separated from joint 11-12. Half the steel of 11 went with deck, the other half went with 12. Joint burst and beam zippered, no PT-rods or anchors broke. I agree with other posters the north end depended on support from other section that was not yet in place or the stress path was bad. The "last minute" 11 foot increase in span is scary to me.

PS. There is a public FIU student dropbox with time-lapse of formwork phase. You can clearly see the rods and joints before concrete was poured.

 

[Meerkat 007]

Comparison between bridge collapse movie and theoretical simulation of truss without member #11

Deck fail between member #11 and #12 #10 and #11 is close too.
Courtesy to Zac Doyle for video and to Ivan Markov for online simulator.
Data input : (0, 100)(87, 100)(154, 100)(221, 100)(289, 100)(358, 100)(412, 100)F0, 139)(80, 139)(157, 139)(240, 139)(327, 139)R0412, 139)(0-1)(1-2)(2-3)(3-4)(4-5)(5-6)(7-8)(8-9)(9-10)(10-11)(11-12)(0-7)(7-1)(1-8)(8-2)(2-9)(9-3)(3-10)(10-4)(4-11)(11-5)(5-12)(12-6)

 

[gwideman]

There is a public FIU student dropbox with time-lapse of formwork phase.

Care to let us in on the address?

 

[jrs87]

Meerkat007

Can you set simulator to no connection between deck and 12?

 

[Meerkat 007]

For that simulator is same if you remove deck between #11 and #12 #10.
Delete "(11-12)" from data input.
UPDATE: You can remove member #12 entirely together with canopy.

 

[OSUCivlEng]

I simply disagree. I think this entire design was irresponsible. Test pilots, astronauts and soldiers at war push boundaries and put their lives at risk. Innocent motorists shouldn't be put at risk with unproven designs built right over their unsuspecting head.

From the MCM-FIGG proposal:

"The superstructure shape for the new signature pedestrian bridge is innovative and one-of-a-kind. We have re-invented the traditional I-beam in a magnificant scale with a special transformation of an open truss down the middle, improving both its functionality and form for a 30’ wide path."

Somebody has been reading too much of their own press clippings.

 

[jrs87]

Reply to gwideman:

https://www.dropbox.com/sh/u8qicuu22xmfhg1/AADPXPe1obmF_0JwDpcqLJeMa?dl=0

 

[jrs87]

OSUCivlEng just made some very good points. Jerehmy is right that new is often unfairly ridiculed. But it is never unfair to ridicule failure. Maybe Figg did not push boundary, maybe they went past it. Or the end result was such via chain of events. We all learned in engineering school the right thing. We forget as time goes by... And it's the same old story: engineers not enough power. But physics always prevails.

Anyway, no one wants to demonize anyone in this forum. Let the NTSB investigate.

 

[Alfie.Tx]

Meerkat007, the simulation clearly showed the movement of the roller support, akin to the tore out. I am newly registered here, and found this discussion really interesting, with lots of great responses. Thanks a lot.

 

[jrs87]

Starting enumeration of code standards:

1. Redundancy, no single plane of truss
2. Symmetry and no faux in super-structural members
3. No 30 foot wide width for 174 foot span without curve in horizontal plane or true wire stays
4. No ribbon deck so it looks like wire-stay but is box truss (wait it's not even a box truss)

This is not meant to be mean, just a thinking exercise.

PS This accident is very interesting because we know how and why this design came about, it's so obvious. But the cause of sudden collapse is very perplexing. After all it was unloaded when it failed. And there is no doubt about it, some trust of engineering as been lost in America now. There as always been accidents and failures, but this one seems different... Maybe because something so simple as a pedestrian bridge failed in the hands of those with much bigger capabilities.

After this collapse we could say big bridge designer demoted to pedestrian bridges. Oh... sorry.


I wonder too how the thin edge of deck was expected to respond to accidental strike from overhigh cargo.

 

[Jerehmy]

OSUCivlEng:

How do you prove a design without building it?

FIGG shouldn't be ridiculed for trying to innovate, they SHOULD be ridiculed for not doing their due diligence in ensuring everything went correctly. I don't think they did. If you're going to do an innovative structure, you need to put extra time and effort so that everything goes correctly.

Was there a FIGG engineer on site? If not, WHY? In this book I'm reading on building/bridge history in USA, the design engineer was on-site every single day. He led the project. I feel we've gotten far away from that and I'm not sure why. Especially on a project that is innovative.

I'm also curious into their R&D in using a concrete truss. Did they do any scale models of concrete trusses? Test them to failure?

 

[jrs87]

No single element doomed this bridge. ABC is okay but rules out wire-stay, concrete is okay but rules out thin, central web is okay but rules out no support from wire stays and ABC... Now we need a Venn Diagram and we will see I suppose this bridge will not be in any circle. Engineers are taught that "never-been done before" is a red flag, because maybe it has not been done because it is NOT an improvement.

Jerehmy, are we in the USA anymore? USA used to mean diligence. Not anymore. You make good points, but I think innovative should not be done just to be innovative in and of itself. Not everything needs to be innovative. Be innovative in the Utah bridges, that's okay. Care to share the name of the book?

I know FIU tests elements, but maybe only steel. And then only small parts. This bridge failed as a system that cannot be tested. This bridge was the lab. Right? Too many variables...

 

[WinelandV]

Re-posting from the last thread:

A defense of the design:

To the structurals here, who hasn't, at the behest of an architect, worked their tail off to make something superflous work? Is that not part of our job? Ok, so the "cable stays" really aren't. So?

Any of us who has come into a project halfway through has had the thought, "This, this, and this look really strange and I hope that this, this and this were accounted for." Same situation here - we're all coming in after the fact and slowly learning what the design really is. I, for one, am willing to give Figg some slack here on all the design issues: the un-symmetry, the not-really cable stays, the non-redundancy, the kinda-a-truss but kinda-a-beam-with-holes issues. That firm isn't just 2 people in a garage somewhere. They have designed many, many bridges throughout the United States (including the I-35W replacement in Minneapolis).

Maybe, when it's all said and done, we'll know one way of the other if the final design was safe (and maybe not if everything gets settled and everyone signs an NDA). Until then, all we're doing is throwing thoughts around. Yes, it failed, but there's only a couple hundred other factors that could have ultimately caused the collapse.

Yes, many people here have identified challenges with the design side. SURELY you don't actually think that FIGG didn't try and account for those challenges? To those saying that the design is completely wrong - it did stand for 5 days - surely that's at least proof of concept? Short of gaining access to the design team's notes, hand calcs, and computer models, along with the rebar shop drawings, and the constructor's construct and erection plans, there's a giant pile of unknowns as to the ultimate cause of failure. I'm not saying that there isn't a fatal flaw in the design (as I don't know either way), but let's slow down the blame game - there's enough talking heads on cable handling that. And goodness, the last thing we need is to start revamping AASTHO's spec for bridge design before we have all the facts.

Additional thoughts:
-if you want to see a cable stayed bridge, go look at the conceptual pre-bid documents that TY Linn put together.
-originally, this bridge only had $6,000,000 allocated to it and it was all coming out of the TIGER grant - where did all the other money come from?
-what happened at the army corps that the 11' pier change 1) was necessitated and 2) came so late in the process?
-what was the full sequence of PT stressing on member 11 over the life of the bridge? Was the work being done at collapse to tension or de-tension the rods?
-who in their right mind allows a crew to work on 150ksi PT rods on a non-redundant structure without closing traffic????

 

[jrs87]

Meerkat 007, I would rather not have to remove all those elements, need bending resistance. I know Java model cannot do that as is.

winelandv, six people died, 10 injured, it's like a nightmare. We know Figg is good and asked the same things we are. This was just a pedestrian bridge. This forum is a form of peer review. There is nothing wrong with that. Raw blame game is bad. I wish NTSB all the best with this one.

I have commuted several times under skyways (elevated freeway supported from median of existing freeway) during construction, 110 in Los Angeles and the Metro Manila Skyway. That is darn scary. So we must have trust in our builders and engineers.

 

[3DDave]

As I mentioned earlier, they had all the time in the world to test the bridge while it was on the construction site and before moving it. Since the largest point of the process was to highlight their engineering program, they could have fully load tested the bridge via a couple of semesters of classes, but were no doubt rushed by the pedestrian death and funding concerns and skipped it.

The horror I imagine is the bridge survived this tweak and instead they had a huge party some time, bridge crowded with new graduates and their families, when the overloaded section blew out, pulverizing dozens and crippling hundreds. Any claim that the fact it didn't collapse instantly proves the concept was OK is a bad one.

I also don't understand the repeated references to 'redundant' as if it would fix this. The I-35W bridge had multiple trusses and still collapsed because one side failed. Redundant structures are very difficult to create for spans. An overload failure of one element typically adds that element weight and the load it was carrying to the load of other elements, which should have been equally loaded to the one that failed, causing them to fail as well. It seems like only in special configurations where a defect in an element allows it to fail far below it's design limit for the current load that there is a chance for load sharing to work. And, worse, a really good redundant structure will mask the original failure until such time as the backup also fails.

 

[pellucidar]

LittleInch, the NTSB only said the workers were adjusting tension on #2 and #11 on the day of the collapse, not that this was the original, somehow neglected, destressing. On the day the bridge was moved, the project manager said they were going to detension "the two bars on top of the bridge" immediately after removing the SPMTs and some ropes. (twitter)

There's another interview with him in the dropbox linked upthread, where he says they have the whole weekend for the move if they need it; it doesn't sound to me like the destressing was skipped in the rush or forgotten.

 

[jrs87]

3DDave, you are correct, redundant term is overused. It would be more precise to say design and engineer to be best practices available. What really bothers me is all the risks taken, no falsework, no closing traffic during tension adjustment. But then again, redundant engines on airliners is nice. Right? I think what sparked this forum is the fear non-engineering influences are getting stronger and stronger and USA standards are receding.

I have noticed on this forum no one dwells on the cracking. And they should not, amateurs worry about that. So we should give ourselves some credit for that. We went into engineering because we know how things work, we should use that knowledge.

The party scenario you mentioned is truly a nightmare averted.

 

[jrs87]

I wonder if those dropboxes stay open because the NTSB ordered no deleting of data.

 

[3DDave]

Engines on airliners exemplify what I mentioned. Each engine is more than capable of taking the full load needed to keep the plane airborne, but they are helped because aircraft can also shed load at will by changing their rate of climb. Even so, there have been a number of cases where they all failed because of lack of fuel or because identical failed service crippled them all or because they all ingested birds at the same time.

 

[waross]

Respectfully gentlemen:
I see two questions here:
1> How did the bridge fail?
2> Why did the bridge fail.

I suggest that the why is the root cause that must be addressed to prevent future "hows".

Why did the bridge fail?
Regarding the calculations of the various forces in the structure:
Calculations depend on values. Some of the values are known, some are reported and some are assumed.
Many of those with a lot of field experience may share anecdotes relating cases where reported values were in error anwhere normally safe assumptions were in error for special cases.
Safe assumptions:
When the bridge was being transported the force exerted on member 11 by the PT rods had an adequate safety factor.
When the bridge was in the final position the force on member 11 due to the weight on the bridge had an adequate safety margin.
A special case where the assumption may be unsafe.
The force on member 11 is the sum of the PT force and the force due to the weight of the bridge.
Now add the unknown force of the de-tensioning procedure with an already reduced safety factor.
Another suspect assumption:
Over tensioning a PT rod is not dangerous.
I assume that this was an experienced crew. They were sent to work alone without direct supervision.
Normally a structure is supported by some type of false work until post tensioning is completed.
A failure of a PT rod or anchor may be inconvenient, it may be dangerous to the workers and it may be expensive.
However the false work will prevent a catastrophic collapse.
I think that it is safe to assume that this was an experienced crew. They have done this many times.
They have probably over stressed PT rods a number of times without any issues.
They may even have stressed rods and/or anchors to failure a number of times without a collapse.
We may never know truthfully what the instructions and warnings were given to the crew before the work began.
It is doubtful that the crew was aware of the reduced safety margin.
It appears that one worker falls from his safety harness. If true, that says something about this workers profesionalism.
The man who will take shortcuts with important personal protective equipment may also be the man who will overtension a PT rod without a second thought.
I hope that this is not the case.
I believe that the root cause of the wye may be a tiny grain of sand or other foreign matter jamming the threads of the PT nut and the following cause was over tensioning the rod in an effort to free the jammed nut.
Yes the "How" is important, but identifying the "Why" is the best way to avoid future "Hows".
I hope that someone has checked the PT nut to see if it is free to turn or if it is jammed. After the failure this will not be conclusive but if accurate information makes its way into the investigation may provide a strong indication of a possible root cause.
All the design calculations may come to naught if the field crew imposes a tension 250% of design tension on a structure.
Bill
--------------------
"Why not the best?"
Jimmy Carter