Miami Pedestrian Bridge, Part XIII 81

[JAE]

A continuation of our discussion of this failure. Best to read the other threads first to avoid rehashing things already discussed.

Part I
thread815-436595

Part II
thread815-436699

Part III
thread815-436802

Part IV
thread815-436924

Part V
thread815-437029

Part VI
thread815-438451

Part VII
thread815-438966

Part VIII
thread815-440072

Part IX
thread815-451175

Part X
thread815-454618

Part XI
thread815-454998

Part XII
thread815-455746


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

 

[Kestrel42]

Reading through FIGGs analysis as well, Vance. I find it supremely unconvincing.

They state several times that MCM/BP did not accurately inform them that the cracks in 11/12 and the pier diaphragm we’re getting worse, and then they admit in the details that they knew this information but were not given specific crack reports with detailed crack progression maps. They complain that pictures taken late Tuesday prior to the collapse were not sent to them in a timely fashion and that had they seen these pictures their response would have been different. But FIGG had already instructed MCM to place the center pier shim and get the crew out to retension the PT bar in 11. From what I have read, it also appears that the EOR had already prepared the slide presentation (the one given on March 15th) by the end of day on March 12th.

Importantly, this presentation was not shared with the other FIGG engineers. They all state that the EOR liked to keep this kind of thing under wraps?!? None of the other FIGG engineers indicate they were tapped to look into the cracking or run any calculations.

Once the EOR was on-site and viewed the cracks in person there were no apparent changes to the recommendations or presentation. Clearly, seeing the cracks in person made no difference (mind was made up, no safety concern because...it hadn’t fallen yet?). The fact that FIGGs models could not account for the cracks didn’t seem to bother the EOR(I disagree, some of the modeling very clearly indicated the cracks in the diaphragm were predictable and it was in danger of failure).

FIGG has a point that the monitoring of the cracks by BP/MCM fell off during the weekend and wasn’t documented very well on Monday/Tuesday. Had BP/MCM continued to measure progress and take pictures that clearly showed the sizing of the cracks, as they did in the first crack report in February, the situation would have been a lot clearer to everyone.

There were so many communication break downs on this project. MCM and BP believed (so they say) that they could not close the street under the bridge without FIU/FDOT involvement. BP specifically mentions that they got push-back from FDOT just for closing the lanes for retentioning of the 11 PT bars. Jose Morales of BP said that while he was at the March 15th meeting he didn’t really follow the technical presentation and relied on the EOR when it came to the safety of the structure. Same for MCM folks. They had every reason to trust that what the EOR said was true, they say. No one noticed that the presentation given by FIGG used different results for the force and capacity calculations than on the original plans! Even FIGG could not replicate their earlier results, which should have been a red flag.

Several times during the project, FIGG hurried through a stage and forced other’s to hurry as well. FIGG decided that they didn’t need a second independent peer review of their bridge design and construction plans. They used their own people (in a different office) until FDOT reminded them that was not permitted by the LAP project requirements. In order to stay within their prescribed timeline, FIGG had to hurry and hire LB to complete the review. But they also had to reduce the amount of time allotted for the review from 10 weeks to 7 and lowered the budget substantially. And they never bothered to check that LB was precertified to conduct this review for a complex bridge structure in concrete. I think the hurry to get it done helped create this gap where so many double-checks were missed.

Then, once FIGG decided on Tuesday to retension the PT bars in 11 they pushed MCM to get the VSL crew out ASAP. The only reason the retensioning project was delayed until after the Thursday meeting was because VSL wasn’t available. FIGG did not ask for more information or wait until they could view the cracks in person before this step was completed. It was only luck that prevented the VSL crew from doing the retensioning earlier on Wednesday. BP didn’t even know the retensioning was going to happen until the meeting!

Notably, the BP inspector certified for PT operations was not called in to monitor the retensioning efforts. He had been at all of the other tensioning/untensioning operations. So BP asked another of their guys to watch the retensioning. He admitted he had never been asked to do this before and was only there to have “eyes on” the project.

Regarding FIGGs statement that the only reason the bridge came down was due to the failure of the MCM team to roughen the concrete in the cold joints to the 1/4 inch amplitude...I call BS. Agree with the Mad Spaniard that this is a complete red herring. FIGGs “testing” is so rife with errors it really has little bearing on the bridge as it was constructed. FIGG dropped the ball in their plans by not specifying how all the cold joints were to be treated but in the end, the faulty design was the proximate reason for the collapse.

I am curious about one thing, though. Typically, would a contractor close a job site or street for safety concerns if their EOR had stated clearly and multiply that there was no reason for concern? How likely is it that a contractor like MCM would override their EOR? Would it matter if the EOR was a big name in the region? I’m trying to figure out for myself if the reputation of the EOR played a roll in everyone just taking what he said at face value and not challenging it.

 

[hokie66]

Yes, I think FIGG's previously stellar reputation played a big part in everyone else involved bowing down and accepting what they were advised to do.

 

[Earth314159]

I finally had time to sit through the NTSB board meeting video.

I found most of it was as to be expected and correct but there lingering questions and lost opportunities I noticed.

They only alluded briefly to design procedures with computer analysis and I think it was a contributing factor. When doing an analysis, you need to start with rough hand calculations, proceed to simple models and then do complex models. You need to understand when the computer is giving bogus numbers and you are making an incorrect interpretation of those numbers. Hand calculations give good estimations of the shears. When computer models give you something else, you have to stop and figure out why that is the case. You have to see the forest before you start delving into the trees. They totally missed that point in the recommendations.

The comment about attention to details and the typos was underhanded and a distraction of the critical issues. In thousands of pages of documents, you are going to get typos that are not caught. That was totally unnecessary. You can use the same trick in any similar investigation. I found it to be underhanded and rubbing salt in the wound.

I wanted to hear more about redundancy. In particular, how do they plan on defining redundancy. Redundancy is important but it is not easy to codify. They showed a steel bridge with two trusses as an example of redundancy but it didn't look very redundant to me. If a critical element in one of the trusses failed, that truss would fail. If one truss fails, the bridge fails. Also, if the FIU bridge didn't have some redundancy, it would have failed after the initial cracks but it didn't. The extra steel in the #12 allowed the bridge to survive until the #11 PT was re-tensioned. So what is redundant and what is not redundant?

 

[epoxybot]

My compliments to the project manager at WJE for populating their contract with some major billing and validating the concerns of BPA, who called into question the utility of the specification Figg cited. It was very generous of Figg to pay for the investigation.
Figg would have people believe that the point of departure is their reiteration of the FDOT spec but the point of departure is that the CEI, knowing full well the specification, advancing concerns that the specification is not up to the field conditions. If you want to blame MCM, it can certainly be stated that MCM should have been the party concerned at the cold joint. If they were actually concerned with the product of their work, they would have at least seconded BPA's concern.

The 1-2 Node makes for an interesting comparison.

 

[hokie66]

Yes, that joint is broken like the one on the other end, and while still on the ground. The deck bottom flange is shortening, and the diagonal wants to resist. No amount of roughness is going to stop that separation. Right there is where the project should have been stopped for further evaluation.

 

[Vance Wiley]

Under normal conditions I kinda doubt a contractor would close a roadway if the EOR had observed the condition and repeatedly stated there were no "safety concerns". That points out what a heavy statement that was. That statement bought the whole thing for FIGG as regards personal safety - workmen and public. There were others who could have made the call, but the EOR was supposed to be the person who knew the structure best.
Now, if the contractor had a lot of experience with concrete and understood that the cracking in the deck near the end was actually just as critical or more so than a large shear crack at the end of a 174 foot long beam-31 foot wide and 18 foot deep weighing 950 tons, they might have sent up a real flag.

To illustrate the importance of member 11 and node 11/12, I have thought for months about sketching a 90 degree rotated image of 100 feet of this truss with the deck oriented vertical and on the north end and 18 feet above a well traveled roadway with some unseen solid support maybe behind and 10 feet above the bottom of the deck. The 100 feet gives an end reaction of 1100 kips to equal the shear connecting node 11/12 to the deck. Support the 1100 kip thing with a 21 X 24 column with a sloping shear plane and reinforced with 8 - #7 bars connected right at the bottom. Then observe cracking forming across the bottom as they did on the top of the deck, in the diaphragm, and in member 11.
Somehow that seems to be something most people would probably think was dangerous and critically urgent.
I should think engineers would.
This truss was no less critical.
But it was a truss - whooda thot something sliding sideways could make it fall DOWN?
As to the cold joints and the lack of "intentional roughening to an amplitude of 1/4" - I think the jury is out on whether calling for a FDOT preparation is eligible for a coeff of friction of 1.0 - instead of 0.6. That point could have made the deck surface friction capacity much greater. Whether it could have lasted for 6 months to complete the backspan is questionable.
I think we need a lab test of the FDOT prep condition.
The FIGG submittal to NTSB doubletalks the testing results and their direction to the jobsite regarding the cold joints. And the drawings specifically call for roughening to 1/4" amplitude on the SIDES of member 12 before casting the intermediate pylon section. That location is SO much less critical than any node to deck connection.
"You don't tug on Superman's cape,
You don't spit into the wind.
You don't pull the mask off the Lone Ranger,"
Jim Croce 1972

 

[Vance Wiley]

One of the many incompatible conditions inherent in a concrete truss with prestress and normal reinforcing.
Add the need to pick the thing up and drive it down the road.
But it was a "signature project" for sure, and will be remembered for a long time.

 

[Earth314159]

The shortening of the deck from the PT really shouldn't contribute much to the shear force. The shortening tends to camber the truss rather than be resisted by the diagonals. With a determinant truss, there would be zero shear from the diagonals due to the shortening.

 

[hokie66]

Earth,

Not so, and this frame was far from determinate. That would be the case with a ductile material like steel, or a solid section like a slab, but not with a concrete truss. And the shortening is not just due to PT, probably 70% is due to drying shrinkage.

 

[Earth314159]

How is shear going to be added with a determinant truss?

Everyone was complaining that there was not enough redundancy on this truss. How can it be so indeterminant that the PT in the deck would contribute to a high shear (or shortening from shrinkage)?

 

[hokie66]

It is a rigid frame. No rigid frame is determinate.

But we are not seeing any truss or frame action when the structure is still on the ground in the casting yard, as shown in the first two photos epoxybot showed above. Those photos clearly show differential movement between the deck and the diagonal, which can only be explained by volume change in the deck.

 

[3DDave]

Much larger structures have been moved far greater distances; some decisions for this move were poor, but moving as a concept is not inherently an insurmountable problem. It's possible that had the original plan been used, the initial ruptures would have happened during the initial lift at the fabrication site causing a halt of movement at that point.

 

[Brian Malone]

I am amazed at the claims made by the FIGG Group concerning what and who was responsible for the failure of the FIU bridge. The way I have interpreted the NTSB preliminary findings and the OSHA report FIGG's design was underdesigned for the conditions and the cracks and indications of failure of the structure were not properly interpreted. FIGG is throwing everyone else under the bus when to a layperson like me their assertion the alleged improper preparation of the surfaces for the cold joint at the 11/12 note caused the bridge to fail makes me question how they claimed earlier the design had redundancy.?? Failure at one collapses the structure seems like a single-point failure condition. A side question on the pads that were placed by MCM at the direction of FIGG - how was this operation done after the bridge was set down off the transporters? Were workers under the span placing the pads unknowingly at great risk?

 

[Earth314159]

Hokie,

I wouldn't call that a rigid frame. it is far more a truss and not a very redundant truss at that. The differential movement is predominantly caused by the shear from the self weight and not volume change.

Even with shrinkage, you need differential shrinkage and indeterminacy to create any shear at all.

To cause shear from the PT and creep, you need a stiff canopy which it is not compared to the truss as a whole. The canopy would have to act as a strong back.

 

[Hiway.1]

I wanted to hear more about redundancy. In particular, how do they plan on defining redundancy. Redundancy is important but it is not easy to codify. They showed a steel bridge with two trusses as an example of redundancy but it didn't look very redundant to me. If a critical element in one of the trusses failed, that truss would fail. If one truss fails, the bridge fails. Also, if the FIU bridge didn't have some redundancy, it would have failed after the initial cracks but it didn't. The extra steel in the #12 allowed the bridge to survive until the #11 PT was re-tensioned. So what is redundant and what is not redundant?

It shouldn't be too difficult for an engineer to figure out redundant load paths that the NTSB discussed. There were no redundant load paths in this structure resulting in an immediate catastrophic failure when the non-redundant truss member No. 11 was re-stressed and the 11/12 node failed.

And to add insult to injury, there were cars waiting at a red light underneath when it occurred and 6 people lost their lives.

 

[hokie66]

3DDave,
Look at the photos which epoxybot posted. The "initial ruptures" happened even before the lift. That didn't stop them.

 

[hokie66]

Earth,
Then how would you explain the differential movement when it was still in the casting yard? Do you think those cracks were sufficient cause to stop and rethink?

Are you confusing redundant with determinate? Totally different things.

 

[Earth314159]

Hokie,

The shores were removed in the casting yard before the cracks showed up.

Redundancy and indeterminacy do have a relationship. They are interrelated. The only indeterminant non-redundant structure are 2nd order structures. A 1st order structure that is redundant is also indeterminant.

 

[hokie66]

Earth,
Do you know that? Do you know how it was supported at the time of cracking?

My opinion is that this was an indeterminate, non-redundant structure, so we disagree.

 

[Vance Wiley]

The heights of the shims that were placed before erection was something like 3 inches. So there was a space under the center to accept the shims. This space was what allowed the cracking of the diaphragm to develop from vertical loads.
I am going to hazard a guess here as to how the added shims were installed - a big hammer. Or left loose until the diaphragm cracked a bit more. There are procedures to be used to replace the low friction pads at the south bent and north bent. They use a short jack (or several), lift enough to get the old ones out, put in new, and let it down.
All space under the north diaphragm was to be grouted, and grouts can set quickly - hours. But they wanted shims because (I suspect) they knew it was cracking and moving so grout would not set.
And yes, the workers were in serious danger. A couple of days later and they might have been casualties. So were anyone who might have been placing shoring under the thing to try to catch it or hold it up. So were every unsuspecting person who drove under it.