Miami Pedestrian Bridge, Part VI 31

[dik]

If it's not spec'd, don't expect it.

Unless spec'd, load testing of new construction is not common, and, the owner would not want to pick up the ticket for the load testing. It's not uncommon to spec it. Load testing costs many thousands of dollars and affects schedule. This is particularly true with design build; they won't do it on their 'dime'.

Dik

 

[TheGreenLama]

I never really believed they could be stressing the PT rods at the time of collapse, but after seeing those cracks prior to the move, both in member 11 AND at the deck interface, it's now in the realm of possibility that stressing operations were being performed at the time of collapse. But to be "potentially" stressing member 11, in an effort to essentially SAVE the bridge, without first installing temporary falsework and closing traffic is beyond comprehension.

One set of data that will be interesting to see, once NTSB releases its report, is how the structure was behaving during the stressing of all post-tensioning prior to the move. Were they getting the expected camber? How did this construction stage deflected shape match calculations? If it wasn't behaving as expected, what changes, if any, were performed? I hope all this comes out.

 

[dik]

But to be "potentially" stressing member 11, in an effort to essentially SAVE the bridge, without first installing temporary falsework and closing traffic is beyond comprehension.

...and, not running up a 'flag' when the cracks were first noticed... the engineer could be in for a heap of heartburn.

Dik

 

[epoxybot]

I'm suffering from "Too Much Information". I felt certain that I had read, within the FIU drawings, that the temporary shoring of the diaphragms shall support the structure only where the shims & bearing pads would contact the pylon or bent. I've been through the drawings twice now and can't find the note. Perhaps I read something similar in a ABC bridge power-point. The south end deck/diaphragm had plinths but the north end did not. Even with the plinths the top of the shoring would have required some kind of plate to evenly support & spread the load. In the time-lapse, when the bridge is moved off the temporary shoring, there isn't any thing to suggest any final bearing footprint had been mimicked prior to the move. Unless the bridge was supported on the temporary shoring in the same manner it would be supported on the pylon & south bent, there isn't any conclusion or hypothesis one can project on to the performance of the cracked 11/12 node.

I did find a note that might clear up some confusion regarding the temporary supports that shall remain in the middle of the bridge during SPMT transport. A later note states that temporary supports of the deck shall remain inside the diaphragm. I believe this is in regards to the SPMT blocking, since the diaphragm is the portion or the deck with longitudinal PT reinforcing.

 

[Cutterhead]

My take...
1 Deck was poured.
2 Truss was poured.
3 Canopy was poured.
4 Blisters were poured. At his point, all concrete, & steel is at equilibrium, with no forces to cause cracks.
5 Once concrete was aged, post tensioning was started while still in supported mode.
6 Then shoring was removed, leaving only end bearing supports. I believe this is when #11 cracks 1st showed up, due to the flexible #11 rebar in shear
interface to deck. It cracked, & jammed #11 into #12. #12 was more solidly restrained in place due to #12's to deck rebar in tension formation.
You can still see #12 to deck rebar at the end of the deck, with the j-bends on them that turned upward into the column. These, combined with the visible
Box out area under #12, made for a more rigid deck connection. The rebar ran in a more parallel interface to deck, which put it in tension, vs #11 rebar
in shear.

6 Once it was time to move, diagonal/post tendons were pre-stressed to allow inward load bearing on transporters. At this point, the crack most likely
stayed open.
7 Bridge was traversed across uneven road ( curb on south end , plus divider curb at mid/northern side) with several moments of inward load bearing, while
they shuffled bridge off center. It's evidenced in one of the time lapse videos. In effect, this may actually have closed the cracks. Due to the extreme
cantilever, it resulted in more weight than the tendons could bear. The tendon calc. was base on an outer bearing point.
8 Bridge is set in place, with crack still closed up, due to additional surface friction from #11 tendons at 560 kips.
9 I'm thinking at this point that once bridge was set, the next step was to de-tension #11 tendons. If this was not the case, why else would have #11's
blisters have been grouted? It was the planned course of action by design.
Remember, they had to jack hammer the opening to get to the tendons for the tensioning, that Rubio later said they were going to do to close the
cracks.
10 Jack hammered out grout, & reapplied tension to close the crack.
11 Bridged collapses while applying tension to last #11 tendon.

The logic behind applying tension to #11 to close cracks, may have come from the likelihood that it probably did just that when it was moved, & set. So,
it worked before, why not again!
The basic push/pull loads from shoring, to transport, to being set, to releasing tendons, to applying tendon tension, at #11's base contributed to bridge
collapse. In the end, I believe that #11 short fillet bore the brunt of horizontal loads, exploded.

 

[chris snyder]

not running up a 'flag' when the cracks were first noticed... the engineer could be in for a heap of heartburn.

Is what I was thinking when I first saw those photos - there would have been a lot of people who knew about it. Four independent engineers gave opinions about the photos.. three said cracks were a sign of imminent failure - from one, "There's nothing they could have done short of starting over and redesigning the structure,' said Howell, whose firm conducts bridge inspections for the state of Texas."
No one would have wanted to do this, or even suggest it, but two said they could have tried repairs. The worst was not closing the road while they were working on it, and I'd assume the cracks got worse after the move.

Does anyone think a repair could have been made?

 

[dik]

Chris Snyder: If all else is 'up to snuff' then an epoxy injection is possible; it appears that this is not the case. I think the flaws were too deeply seated for a simple repair to be made. It would have been a messy redesign and reconstruction job. If you are trying to develop a large bar in compression, bends are not effective and the section depth did not look like you can develop a large bar... but, I don't know. I need more information.

Dik

 

[dik]

I'm suffering from "Too Much Information".

Agreed... too much and not enough of the right kind...

Dik

 

[TheGreenLama]

Does anyone think a repair could have been made?

That's just it. Nothing is obvious from the pictures included here. One possibility could be to make the temporary tendons within 11 permanent. [Edit: Only as a means of attacking the cracks in the short term. There still would exist serious design issues with the overall design.] But it's hard for me to fathom how an engineer could sign off on a change order like that. I won't even speculate on what that hypothetical wording might be.

 

[dik]

I'm hoping that once the data is released, that, because of the type of collapse. The cause of the failure will 'stick out' like a sore thumb... Need information: 'design forces in the web members, including shears and moments. Require how much reinforcing and how it was placed. Need information on how it was constructed. Need information on the 'strand/bar' and sequence of post-tensioning. Not much information, and, may have missed a couple, but essential to run some numbers through. I don't know how ACI or AASHTO handles the 'triaxial' mess at the nodes.

Dik

 

[hokie66]

dik
I think the cause already stands out like a sore thumb...your "triaxial mess at the nodes". Trusses need axial forces to be coplanar and connect, which wasn't the case here. Neither did the nodes cope with the inevitable bending moments.

 

[hokie66]

Cutterhead,

As to your point 4 above, you forgot shrinkage and temperature forces.

 

[dik]

your "triaxial mess at the nodes"

I've thought that was the case since the beginning... I need member forces (axial, shear, and moment) to just confirm. My suspicions have been that there has been an improper development of these... and need data to confirm. I don't have a lot of confidence in the NTSB coming to the correct cause... I've seen how Ministry of Labour works in Canada...

Dik

 

[hokie66]

Not sure what capabilities the NTSB has in house for analyzing this. I think it more likely that they will use a consultant like WJE, SJH, or somebody, then take credit for the report.

 

[dik]

Hokie, I'd think that they would use an external consultant, too.

Dik

 

[Sparweb]

dik,
I've had nothing to contribute technically, but I have read with interest all along.
What is stopping you from carrying out a truss analysis to get the loads to a rough order of magnitude?
(Again I am likely exposing my profound ignorance of bridge building).
If FIU intended the bridge to function structurally like a truss, and oriented the members to behave that way, can it not be worked out what the member loads were expected to be?
One would have to disregard any failure on anyone's part to actually accomplish a true truss with pin-joints, members aligned with nodes, no secondary moments, etc. etc.
The design report details all cross-section dimension, member lengths, and material densities, therefore is this enough data to carry out a "back of the envelope" analysis of the loads?


STF

 

[dik]

I don't have the time, or, the applicable codes for loading combinations. It would be handy to have a listing of the applicable loads for analysis. I can work them out, but, it would take a week of solid work. It would be easier to be provided with the loading and to do a quick check to see if the loading is appropriate. This information should be available since it occurred before the NTSB embargo.

Should have added... I don't have real experience with bridges, either.

Dik

 

[dik]

Trusses need axial forces to be coplanar and connect, which wasn't the case here.

It's nice to have, but, if it's not that way, it's possible to design to incorporate this. I had a 18 or 20 storey building that the owner wanted the one building column moved about 6' on the second floor... had to transfer the loading over 3 floors, but, it's still standing.

Just so much information not available, yet.

Dik

 

[dik]

Just realised it... I'm surprised that Figg or Berger has not called in their own independent consultant to review the design.

Dik

 

[SFCharlie]

Why don't y'all take a' look at
Chris Snyder (Electrical)'s great post (many stars) above at 30 Apr 18 21:20
,in it there's a link to the full analysis?
SFCharlie