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

 

[dik]

Hear, hear! All the minutiae about which member failed and how, while ultimately valuable, rather sidesteps the issue: this was a horrible design.

I thought it was extravagant and really unnecessary. It did not have a proper function and the hype with all the useless amenities was a travesty. It was, however, attractive (to each his own). I enjoy an engineering challenge and wait for years before something 'pops up'. I would love to have to undertaken a similar design, a cable stayed unequal flanged beam/truss. As grim as the outcome for the design of this one was, doesn't make it horrible. I think it could have been a design challenge... anyone can design boxes. Doing this is the sort of thing that caused humans to leave their caves.

Dik

 

[dik]

epoxybot:

Notice the new style of hardhat...

Dik

 

[emba]

Dik - the people in those photos are search and rescue, not construction. Those are standard search and rescue helmets. Note the chinstraps - Emba

 

[MDEAus]

I've registered just to put my two cents in. I think that many other comments have already suggested the likely cause so I'm not sure how much more I will add to the conversation. In my opinion the PT bar in 11 was preventing the shearing of the 11,12,deck node by transferring the shear forces to the deck. When the tension is released from this member the shearing is greater than the node can handle, as a result the node has failed first. what we then see is a failure at the 10,11,roof node. As the 11,12,D node shears there would be a moment reaction required at the 10,11,R node which it is unable to support and thus the roof buckles at this node. With the first two nodes failing we then need a moment reaction at the 9,10,D node which it is unable to support and thus the deck buckles(at this point the northern end of the deck is still on the pylon).

Now as to the PT bar being so far extended I believe due to the length of member 11 reducing as the bridge collapsed it was able to be pushed out (part of why the blister moves to it's final resting place) Rather than pinging out if the PT bar had snapped during tensioning/detensioning. I haven't had any experience with PT bars but wouldn't one breaking under tension be energetic enough to result in the PT bar coming clean out? I don't think that the lower PT bar has failed in anyway, there are too many observations that require it to be intact. With the lower PT bar in 11 fixed into the deck a below the 11,12,D node the deck pulling away from members 11, and 12 still on the pylon would have produced the zipper out the bottom face of 11 (as others have noted).

I hope that isn't to ramblely.

-Will

 

[dik]

emba... thanks, there's a new style of hardhat coming to the job site, and I thought it was one of them. Catch the link:

https://www.constructionjunkie.com/...ies-are-ditching-hard-hats-for-safety-helmets

Complete with chin straps...

Dik

 

[Jerehmy]

Far and away what bothers me the most about the failure is the lack of ductility. The near instantaneous collapse due to the brittle failure guaranteed injuries. A more ductile failure would have at least given the victims a chance.

A structure with no redundancies, brittle failure limit states, and located in a high traffic area just doesn’t sit well with me. The ductility can and should be addressed in code changes.

 

[dik]

Jerehmy: Agreed... but, this was a structural failure, likely, due to an error in design. Codes require things to be strong enough, and, this for whatever reason failed. Most engineers consider alternative load paths and ductility in design... I don't know how this was designed, and hopefully, this will come out. It may have been the original design had strength and ductility incorporated, and, something went horribly wrong. Intelligent design should not have to be mandated by code.

Dik

 

[appster]

For background please review my post in Thread II

In addition to my post in thread II, I believe the following could be relevant :

1) According to the design notes the entire structure with regular rebar was to be completely poured and cured. The section was then to be post tensioned by completing the bottom flange first. If this was carried out in one step then the PT loads would shorten the bottom flange inducing shear along the entire web being higher in both end panels. Was this stress allowed for in the design and did it result in micro cracking which could reduce the reliability of the plain concrete taking shear.
[tt][/tt]
2) PT of either the web or top flange after the bottom flange would introduce stresses to either the web or flange accordingly. After all of the PT was the section overall stress fairly balanced or was there a noticeable bias in deformations between the bottom flange, web and top flange which resulted in an unevenly stressed structure with excessive built in stresses aside from other loads.

3) Had most shrinkage taken place before PT was started, shrinkage in the flanges may have been slightly higher that the web due to being relatively less bulky and the flanges would only be formed on the bottom rather than all 4 sides. Overall creep effects should be fairly equal if PT stresses in the concrete over the section were fairly balanced.

4) Since there were strain gages on at least some members is there strain history available to the NTSB which might assist their investigation.

5) De-stressing the bottom two bars in 11 seems counter intuitive because any additional shear capacity at the deck provided by confinement of the concrete at 12 would be lost. There does not appear to be adequate milt steel reinforcing in the evidence available to transfer the compression from 11 into the deck.

Just a few more factors which confirm my belief that the design concept is more complex that it might appear at first glance.

Ductility in a simple span structure must be available for safety. Regardless of what failed first this was not a ductile failure.

 

[Jerehmy]

Dik:

Yes, it may in fact had been designed with adequate ductility. We’ll have to wait and see. It just didn’t seem so in the failure. I also understand the sentiment that the codes are already burdensomely large, and that he bigger the codes get the less an engineer gets to engineer. But, I can’t say I’ve ever seen a concrete truss before. I don’t see the harm with code requirements for concrete trusses such as minimum mild steel at joints. New member types should warrant codes requirements.

 

[Ron]

steveh49....my point was simply that engineers should never be in a subservient position to those making money decisions without technical validation.

 

[appster]

"Yes, it may in fact had been designed with adequate ductility. We’ll have to wait and see. It just didn’t seem so in the failure. I also understand the sentiment that the codes"

Use of ultimate strength design was to insure ductile behaviour but it assumes that adequate reinforcing be provided. There does not appear to be more than nominal stirrup reinforcing either in the top blisters, intersection of web members or the bottom flange. The two bars which provided some confinement at the bottom of 11 and 12 were being de-stressed which in my opinion actually reduced the shear capacity. The bottom of member 12 is not even fully connected to the deck as per "gwidemans" earlier posted pictures.

 

[emba]

To Dik - just for interest value, on the rescue photo you can see the shirt on the lead person says "FLTF1", that's the "Florida Urban Search and Rescue Task Force 1". Same pic in wider format shows a headlamp on the helmet (

https://www.cbsnews.com/pictures/de...national-university-fiu-today-2018-03-15/14/)

Rear rescuer with dog looks like she has a Sierra cup hanging from her belt. -- Emba

 

[VolsCE84]

I'm with you Dik. We learn from our failures. We pick up the pieces and fail foward. We will all learn from this.

This design was beautiful. I liked the simplicity in the way the they used the canopy as the top CHORD...the sleek deck as the bottom CHORD. Keep in mind the south end had more loading in member 2 than member 11, because it had a smaller angle to the horizontal and the diagonal was longer. Member 2 and its connections survived the impact of the collapse. What happened at the node of 11 & 12?

Did they spend a heck of a lot of money for this???...you bet your sweet silde rule they did.

 

[Tomfh]

I guess beauty's in the eye of the beholder, because to my eye it was an ugly phoney kludgey design.

 

[VolsCE84]

tomfh i did not like the fake stays...even though they were to dampen.

 

[appster]

hokie66

Correct after looking at the structure 2 is at a flatter angle than 11 and would have the higher horizontal component, so why the difference.

Joint at 11 is fixed?
An error occurred in de-stressing no. 11?
Was one rod in 11 over tensioned and lifting it off caused it to rupture with the released energy propelling part of it out of the channel?
Was there a particular problem with the cap at the top of 11? or did it punch through due to the collapse of the canopy onto the road?

Don't really know except the failure was not elastic.

 

[hokie66]

You must be looking at the bridge backwards. 2 is the shallow one, 11 is steeper. Yes, the vertical component would be the same for both, but the more shallow one would have the greater force.

 

[appster]

hokie66

Gave you an extra star for this one. Also meant that the bearing at 11 is fixed not the joint. 2 appears longer than 11 and therefore 11 would be stiffer than 2 if similar size. I don't have all the sizes but from the only drawing that I have they appear to be similar in depth, if so then 11 could transfer more unbalanced moment into the deck and canopy.

 

[OSUCivlEng]

The ductility can and should be addressed in code changes[/b].]

Code changes? What part of the AASHTO Bridge Specs are written for concrete trusses? I'm curious how many people on this forum have seen a concrete truss, let alone a concrete truss bridge? I found 3 on bridgehunter.com, the newest built in 1934. Fewer still are those who have designed a concrete truss bridge.

FIGG pushed the design of this bridge to far past the edge of what was possible. Unfortunately, some paid the ultimate price for their arrogance.

For $16.6 million you could have built several pedestrian bridges with proven designs and technology. The USDOT and FDOT share some of the blame for providing funding and going along with such a foolish endeavor. Pride goeth before destruction, and an haughty spirit before a fall.

 

[Jerehmy]

OSUCivilEng:

Exactly my point. If concrete trusses are going to become more common, they should specifically be referenced in the code.

Also, on your point about pushing boundaries; people said the same thing about truss bridges when the transition from masonry arches to trusses was occuring, suspension bridges bridges when they first came around, steel frames in buildings instead of masonry in the late 1800s, and so on and so on. To demonize FIGG for pushing boundaries when we don’t have all the answers isn’t right. Our profession’s history is full of us pushing boundaries.