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

 

[Archie264]

TME, very gratuitous response on Part II of this topic.

 

[Ron]

Either the risks of this work weren’t appreciated by the team, or there was some sort of decision made based on risk vs cost/schedule/hassle/whatever.

This is exactly why an engineering firm should NEVER be a subconsultant to a contractor in design-build.

 

[IEGeezer]

Compression of member #11, before deck fall, could be pushed out PT rod, as someone suggested already. (Meerkat 007 20 Mar 18 18:02)

Numbering nodes 1 through 6 on the bottom chord beginning with the bottom of Members 11 & 12 as node 1 and the bottom of Members 1 & 2 as node 6, the distance between all nodes on the video Meerkat 007 posted seemingly remain constant before impact with the ground, except for the section between nodes 1 and 2. Therefore, it seems reasonable that the failure was in the bottom chord between nodes 1 and 2. After collapse, what is left of the bottom of Member 11 is pointing at the top of the pier and the deck portion is resting at the bottom of the pier. Thus, it seems reasonable that the point of failure would have been in horizontal shear between the deck and the bottom of Member 11 at node 1.

There does not seem to be any shear reinforcement between nodes 1 through 6 and the deck. Nodes 2 through 5 appear to be confined horizontally, but not vertically. Nodes 1 and 6 appear to be confined vertically and transversely, but not longitudinally in the direction away from the deck. The PT tendons in the deck begin about 8 inches outside of the width of the concrete "truss" and node 1 portrudes about 10 or so inches beyond the anchors for the deck PT tendons. Was the horizontal shear between node 1 and the PT tendons in the deck being carried in (great?) part by the PT rods in Member 11? Did the concrete fail at the anchor in the deck for one of the Member 11 PT rods when tightening, thus leading to shear failure at that point, followed by the deck falling off the pier? I wish I understood post-tensioned structures better.

 

[waross]

Has anyone seen Juan Brown's video?
He is talking about the failure of truss #22 but pointing to truss #11.
A bit of a hit to the credibility.
Link
Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[3DDave]

Now my head hurts. Numbering nodes backwards to members.

How about deck between #10 and #11, canopy between #11 and #12. node common to #11 and #12, node common to #10 and #11? No added numbering to confuse people.

waross - he needed reading glasses, and the ability to count left to right.

 

[bridgebuster]

Unfortunately DB is the way of the future.

Ron, what you said reminds me of something the chief engineer at a bridge firm I worked for in the 80"s said "engineers are like prostitutes fighting over a customer".

 

[waross]

I hope he never has to land on runway 11

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Latexman]

Either the risks of this work weren’t appreciated by the team, or there was some sort of decision made based on risk vs cost/schedule/hassle/whatever.

Maybe the early completion bonus was too large to resist?

EDIT - Guess not. See below.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[oldrunner]

This is interesting. Something I thought about at the very start.

http://www.foxnews.com/us/2018/03/2...hedule-bridge-in-deadly-florida-collapse.html

 

[dik]

Has anyone seen Juan Brown's video?

made my tummy hurt... not a good presentation, IMHO.

Dik

 

[Tomfh]

What was the span and depth of the bridge? In metres if possible.

 

[Ingenuity]

What was the span and depth of the bridge? In metres if possible.

In 'pagen' units: 175' span, 15.5' depth ==> L/D = 11

In 'gods' units: 53.3m span, 4.72m depth ==> L/D = 11

 

[JAE]

Some real crazies sure come out with engineering mumbo jumbo after major failures.
This one seems to take the prize: FIU Bridge Collapse: Why Müller-Breslau Matters



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

 

[Lnewqban]

@ dik, EPCI-Steel and Ingenuity: I really appreciate your answers to my questions about the tensioning process and equipment.
At the office, we were discussing about the fact that one of the tensioning bars of member 11 remained in place while the one having the hydraulic cylinder attached to it came out several feet.
If over-tensioning of that bar up to the point of its violent rupture is not probable, what other thing could have caused that different reaction of both bars during the collapse?
Could under-tensioning and subsequent buckling of member 11 have had a similar result?
Is it possible that the combined loads of compression, bending at nodes and shear endured by member 11 were so high that additional increasing compression from the tensioning bar fractured the concrete of the member?

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

 

[Ron]

BB....that Chief Engineer was probably right!

 

[gwideman]

Details that stand out most for me: From various photos, and also NTSB footage on YouTube (links below), the following observations seem fairly solid (and some have been pointed out be others previously -- I'm consolidating here):

1. Member 11's two tensioning rods do not appeared to have broken, but rather remain connected in some way to their intended end points:

-- The upper rod remains connected at top end, and at the bottom to the base of vertical member 12.

-- The lower rod remains connected at the bottom to the deck, and at the top is protruding several feet from its intended attachment point, with the blue hydraulic jack attached.

2. Member 11 appears to be substantially in its normal relationship to member 12 at the base, and at the other end, relative to the last canopy section where the canopy-11-12 joint would be. However, it is not connected by concrete at either end, and now rests in that position primarily by virtue of the upper tensioning rod, and maybe some rebar. The lower end of 11 is disconnected from the (lower) deck despite the lower tensioning rod, which zippered out of the underside of 11.

3. The tensioning rods or cables in the (lower) deck appear to be substantially intact, probably being responsible for dragging the northernmost segment of deck off the supporting pillar during the final milliseconds of the collapse (as shown in the dashcam videos). So, I'm dissuaded that this was a failure of the lower deck at some intermediate location.

3. At both ends of member 11, the concrete is broken up. But is this a cause or effect of the overall failure? Certainly, if crushing of one end or the other of #11 occured, it would have resulted in a collapse as seen. But did it?

The purpose of member 11 is to transfer the very high horizontal compression force in the top canopy, into a diagonal compression in member 11 (and tension in #10). Then #11's compression vector translates into a downward compression in the pillar, and a large horizontal tension in the deck.

I think it is very telling that the entire #11 tore away from the deck, except for one of its tension rods. The rest of its rebar and the upper tension rod remained attached to the vertical member #12. Even if there was some other cause for the overall failure, I would have expected the base of 11 to remain attached to the deck. Does this not indicate that the connection of #11 to the deck, specifically to the deck's tensioners, was inadequate?

The base of #11 has to deal with a high level of compression, and evidently some shear forces, for both its vertical and horizontal duties. If the rebar is not judiciously arranged to maintain the concrete's integrity and ability to sustain the compression load, then the diagonal compression will not be transferred to the deck tensioners and the pillar, and #11 fails.

A similar, though perhaps less severe, situation occurs at the top of #11, though here, the top of the member, while crushed, did not split in two different directions (so far as we know).

A further factor: The joint of #11 to the deck and the end wall #12 looks to be a joint that would suffer the most flex between its role as the furthest cantilevered point during transportation, and the reversal of all forces as the bridge was placed into position. As others have mentioned, concrete joints aren't pins. Such flexing could disrupt the integrity of the concrete, and make it less capable of withstanding the compression it is normally so good at.

In theory, the tensioning of the rods prior to transport, and detensioning during placement, should have compensated for that. But one wonders how precisely that was done.

In short, my speculation is that the joint at the bottom of member 11 to the deck will receive a lot of attention (as a couple of others here also favor).

NTSB:

https://www.youtube.com/watch?v=B1PPqa9cS-k,

https://www.youtube.com/watch?v=aeJKqojmHgY

CBS some detailed helicopter footage:

https://www.cbsnews.com/news/florid...hind-schedule-over-budget-updates-2018-03-20/

 

[wiktor]

It appears to me that Ron pointed to the right mode of failure - punching shear dislodging entire blister out of the canopy. Lack of mild reinforcement is quite apparent and entire blister cracked out intact. On one side of the blister there are few #4 rebars visible..
As an oldtimer, I'm sticking to the classical rule - and not relying on the shear capacity of the concrete in this situation. Mild steel reinforcement well anchored in the canopy and sized to handle the total shear would save the bridge.
Another contributing factor, which should be discussed, is apparently low overall safety factor for the span. Assuming that the bridge was designed using AASHTO LRFD, DL=11 kips/ft will have a factor of 1.25, while the LL, which is only 2.9 kips/ft, has a factor of 1.75. Adding it together we are ending with the safety factor of 1.35 for a fully loaded pedestrian bridge with no other loads. Fake stays will add some extra capacity, but not to the blisters.
STRENGTH I: 1.25(DC1 & DC2) + 1.75(PL) + 0(WS)
STRENGTH III: 1.25(DC1 & DC2) + 0(PL) + 1.40(WS)
SERVICE I: 1.00(DC1 & DC2) + 1.00(PL) + 0.30(WS)
DC = dead load of structural components
PL = pedestrian live load
WS = wind load on structure

 

[human909]

Most people are in agreement about member #11. Those that are suggesting other alternatives I would suggest you look at a simple analysis of the hinge line that the bridge breaks around. Gross failure in the deck or the canopy would very likely lead to different breaking points and hinge points.

In short, my speculation is that the joint at the bottom of member 11 to the deck will receive a lot of attention (as a couple of others here also favor).

I initially more persuaded by punching shear of the canopy. However what has caught my attention from the dash cam video and is otherwise hard to explain is the rapid failure of member #12. Member #12 would be under minimal load during both static and during failure. Yet it very quickly collapses. This ties member #11 failing at the deck as failure here would compromise the connection of #12.

 

[waross]

I am sure that the compression forces on member 11 by the tensioning rods, during travel and placement had an adequate safety factor.
I am sure that the compression forces on member 11 due to the weight of the bridge had an adequate safety factor.
However, how much would the safety factor be reduced when both the weight of the bridge and the PT rods were acting together on member 11 when the bridge was placed and before the rods were de-tensioned?
Any comments?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[waross]

JAE That's an interesting article.
The gentleman claims, more than once, that due to the lack of a center support, the bridge failed in the center.
When someone points out that the bridge failed at the end he again claims that the bridge failed in the center.
One feels pity for his students.
He may be suffering from a defect in his central nervous system.
It may be that impulses from his brain do not go directly to his voice box, but are somehow detoured through his bowels on the way to his lips.

Bill
--------------------
"Why not the best?"
Jimmy Carter