Miami Pedestrian Bridge, Part II 55

[LonnieP]

Tom, got that from the side view rendering. Failure occurred at the end diagonal with the lessor internal load. curious.

 

[miecz]

The dash cam video has a crane obstructing the joint at the top of member 11. But the short “CCTV” video posted by Tomfh on 18 Mar 18 01:39 appears to show the space between members 10 and 11 diminishing before the collapse. This looks like a shear failure at the top of member 11.

 

[NITTANYRAY]

If you look at the 5 second video posted by MattOhio in the first thread, you can see the initial collapse point was at the bottom chord on the canal side of the bridge (the node where 9 & 10 intersect in the diagram near the beginning of this thread). If the failure was due to bending the collapse would have occurred near mid-span. This clearly indicates that the collapse was due to a shear failure in a global sense. The failure probably occurred in one of the web members at this joint. The web(s) could have failed in either shear, tension or compression. Initially member 11 was in tension during transport since the end was cantilevered beyond the support trolley. After the truss was lowered onto the piers the stress in members 10 and 11 would have been reversed. Proper procedure would have probably required the tendons in 10 to be stressed and some or all of the strands in 11 to be released on the day the bridge was placed. Apparently some or all of the strands were not done on Saturday because after the job meeting on Thursday morning, workers went up on top of the truss to do what ever was needed. In that video posted by MattOHIO it appears that there were 2 or 3 workers on the roof of the bridge above the node where 10 & 11 intersect.

 

[appster]

I have read many, but not all, of the comments here and just for a pause would like to go back to the basics of the structure. The bridge is basically a simple span girder/truss with a fairly narrow top flange and extremely wide bottom flange. As a girder/truss it is a torsionally weak section which is supported on each end on the bottom edge and open to possible lateral torsional failure and sever shear lag problems in the wide bottom flange. The top flange is composed of a two thin curved sections presumably with solid concrete diaphragms at the termination of the diagonals. The truss members are concrete which is in itself regardless of strength a relatively brittle material with low tensile and shear strength unless reinforced and/or pre/post tensioned.

The top flange when exposed to bright sunlight will partially or fully shade the bottom flange and web members resulting in an average temperature gradient between the top flange and the bottom flange and web members. Because the web consists of concrete members the joints must be assumed to be fixed. The support of the entire bridge on the bottom of the section requires that all torsional stability is provided by the web members of the girder. This results in members that ultimately are subject to tension and compression plus biaxial bending based on rigid joints and extra stresses imposed by temperature differentials.

It is my opinion that the truss is much more like a beam than a truss mainly because from looking at the truss joints is does not appear to me that the streses from the tension and compression diagonals at the top or the bottom can be transferred directly to each other but require shear flow transfer to occur into the top and bottom flange. The top flange is hollow and I assume with diaphragms at the junction of truss diagonals. There does not appear to be any significant stirrup arrangement in the diaphragms or in the bottom slab so in effect it is then plain concrete in shear. The bottom flange is thin and wide and only a small fraction would be effective in transferring shear flow due to major shear lag effects in the bottom flange.

In summary I believe that the structure is much more complex than a "normal" truss but may have been designed as simply a truss and that the joint design was not suitable to transfer all of the loads between truss diagonals without complex stresses at the joints which were not designed to fully transfer the loads without involving substantial local effects on the top and bottom flange. Substantial shear stresses in the top and bottom chords and additional tensile and compressive stresses in the diagonals led to failure and with the maximum shear in the truss panel where member 11 lived due both to the end bearing reaction, culmination of any temperature gradients and maximum torsional stability stresses. This would also be the location of maximum shear flow if we look at the bridge as a girder where shear flow = Vq/I. The complexity which I believe is there plus the lower load factors taken for dead load do not allow in my opinion for making any simplifying assumptions.

Sorry for going on so long but for me to get this thing straight I need to logically go through this thing to conclude that it is an ill conceived design which is much too costly and in my opinion led to inevitable failure of the structure.

appster

 

[housemover]

I noticed I can no longer find any of the vidio of the bridge being turned 90 degrees prior to placement, it has all been removed. It appeared to me in that vidio they were basically skidding a transporter with 450 tons around 90 degrees, the pivot transporter wheels did not appear to be following a circular radius although the outboard transporter appeared to be on radius. I wonder if using the bridge as a giant lever was included in the design requirements as obviously the spreader beams were not set up to handle the kind of twist that would be generated if this were the case. It also appears the walkway initially collapsed near that piviot point. Just observations.

 

[JAE]

A short article from 2013 on FIU's participation in the project:

https://www.chronicle.com/article/Engineer-Connects-His-Research/142787

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[rapt]

Nittanyray,

Agreed, that is the location that I suggested. But I think the failure is actually a compression failure in the concrete in the slab at the connection. You have the very concentrated high compression stress in the concrete in the slab at the node from 9 combined with the precompression from the prestress tendons in the bottom slab that is also relatively high.

At the 3 - 4 frame in the video you see something like crushing in this area with an explosion of material upwards from the top (not as in caused by explosives). The diagonals still seem to be intact at this point. Next frame, the top drops very quickly and you get the rotations in the 10/11 node at the top that have been noted previously as the bottom slab at this location crumbles.

 

[dik]

Also, I'm having a hard time believing an Engineer would CHOOSE to have the last diagonal in a concrete truss in compression, with a tendency to push it's way off the end if something failed.

If all is detailed correctly... that's what concrete is best at... compression is good.

Dik

 

[dik]

LonnieP, the member orientation is to make them look like hangers aligned with the phoney cables above..

Those phoney cables are connected with 8 - 1-1/2"dia bolts (grade unknown) That could have a capacity of 400K or 500K. Not at all a slip connection.

Dik

 

[JAE]

Those phoney cables are connected with 8 - 1-1/2"dia bolts (grade unknown) That could have a capacity of 400K or 500K. Not at all a slip connection.

Even with the dreaded APPENDIX D!!!!!!!????? I'm not so sure but would have to plug it into Profis to see.

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[bimr]

FIGG Bridge Engineers, Inc. is the designer of the bridge, working for MCM. The project is part of a $19.4 million U.S. Department of Transportation TIGER grant, and was bid competitively using design-build procurement by Florida International University.

http://www.figgbridge.com/FIU_Statement.html

https://heavy.com/news/2018/03/florida-international-university-bridge-collapse-photos/

https://heavy.com/news/2018/03/denney-denny-pate-figg-florida-fiu-bridge-collapse/

General Plan and Elevation

https://heavy.com/news/2018/03/muni...ement-mcm-built-florida-fiu-bridge-collapsed/

 

[Lnewqban]

@ Lnewqban

I have a bit of experience with SPMTs. They are really a wonderfully designed piece of equipment, there are several manufacturers now but they all operate the same way. All of the axle lines during a normal transport are on the same hydraulic circuit which allows each axle to stroke up and down independent of adjacent axle lines while maintaining a constant bearing, so crossing normal bumps and slight elevation changes are no problem. But the axles only have 1ft of up/down stroke so there are limits. Also they talk about them in terms of the number of axles but there really are no axles, they have independent, two tire hubs on either side of an axle line.

And just because I think they are so cool, here's some more. Any hub can be isolated and raised, for instance if a tire blows out. They can turn the hubs on either side of the axle line independent from each other so you can walk the trailer nearly 90 deg from its long axis, or almost pivot in place, wonderful things.

Thank you very much for the detailed explanation, EPCI-Steel.

Additional questions about the tensioning process for anyone that may know:
1) Can the plastic flow of the steel, prior snapping, be seen in the gauge or be informed to the operator by some kind of alarm?
2) If so, does the hydraulic machine have any automatic device that prevents it from applying additional tension to the cable/tendon/tensioner beyond its yield point?
3) Can the machine fail in a way that excessive force is applied (failing switch, inaccurate gauge)?

"Where the spirit does not work with the hand, there is no art." - Leonardo da Vinci

 

[3DDave]

housemover - this shows the arrangement of the crawlers

https://amp.local10.com/news/florid...-installs-first-of-its-kind-pedestrian-bridge

I think there is no chance there was misalignment during transport.

 

[Tomfh]

you can see the initial collapse point was at the bottom chord on the canal side of the bridge

My money is on the top of #11.

 

[dik]

Even with the dreaded APPENDIX D!!!!!!!????? I'm not so sure but would have to plug it into Profis to see.

I don't know how anything is anchored or what grade the steel anchor is or if there is any confinement... the attachment is anything but a 'non-load' transferring connection, and, with this capacity, it has the potential for throwing a real wrench into the analysis of the walkway.

I've never used a 1-1/2 dia Kwikbolt...

Dik

 

[dik]

The owner is holding back payments...

I don't know how this is undertaken contractually... whether a Certificate of Payment can be negated due to on site damage, or what. I've never encountered this. I don't know how to 'claw' back money once it has been released. I've withheld funds because work has been rejected (one of the few ways payment can be withheld, and, still honour the contract.

Dik

 

[dik]

Can the plastic flow of the steel, prior snapping, be seen in the gauge or be informed to the operator by some kind of alarm?

It's likely that the rod failed at a thread; there would be little ductility or indication of failure before the rod 'broke'.

2) If so, does the hydraulic machine have any automatic device that prevents it from applying additional tension to the cable/tendon/tensioner beyond its yield point?

With yielding materials, hydraulic machines are very good for relaxing load and the load falls off quickly.

3) Can the machine fail in a way that excessive force is applied (failing switch, inaccurate gauge)?

Yes, but the equipment is generally very reliable. The devices are robust and simple, and nothing can go wrong, go wrong, go wrong, go wrong, go wrong, go wrong, go wrong...

Dik

 

[Settingsun]

The owner is holding back payments...

No, really?

I don't know how this is undertaken contractually... whether a Certificate of Payment can be negated due to on site damage, or what. I've never encountered this. I don't know how to 'claw' back money once it has been released.

That would depend on the contract. But you're in a better position if the money's in your bank account than theirs. D&C probably helps too since it doesn't matter whether it was a design or a construction error, it all 'your' problem from the client's perspective.

 

[Ingenuity]

No, really?

Nothing quite like Aussie sarcasm!

 

[EPCI-Steel]

@Lnewqban,

To further Dik's replies to your questions.
I have used these type of hydraulic power units before to operate other devices, not a tensioning ram but other types of cylinders.

1) Can the plastic flow of the steel, prior snapping, be seen in the gauge or be informed to the operator by some kind of alarm?

No. These are very simple devices. The power pack consists of a pump, mounted onto a steel box which serves as a reservoir for the hydraulic fluid required, an inline pressure gauge, and a valve. The ones I have used are for operating a double acting cylinder so the valve has three positions, Open to line 1, closed, and open to line 2. One line will extend the cylinder, one line will retract the cylinder. If you want to derive useful force information, you must know the area of the piston where you are applying pressure (retraction is less because you must subtract the rod area) and multiply that by the hydraulic pressure from the gauge. From an untrained or inexperienced operator's perspective the only thing he will see once the steel is fully plastic is that the pressure on the gauge in not increasing.


2) If so, does the hydraulic machine have any automatic device that prevents it from applying additional tension to the cable/tendon/tensioner beyond its yield point?

No, the only limit is the ability of the pump to apply pressure and the individual components of the system. Usually the pump is the limting factor in the system and the cylinder is pressure rated to at least the pump pressure. 3000 and 10000 psi are the normal maximum operating pressures I have seen for these types of units.

3) Can the machine fail in a way that excessive force is applied (failing switch, inaccurate gauge)?

Everything can fail, in almost any conceivable combination, but a runaway pump scenario is very unlikely. Gauges can be inaccurate but they are supposed to be tested annually and given a certificate of calibration. Human error is much more likely.