Miami Pedestrian Bridge, Part XI 32

[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


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

 

[Tomfh]

I agree the horizontal protrusion is most likely just the corner of the deck.

You can see it earlier in the video, well before the collapse:

Then the corner passes in front of the tree, which camouflages it:

The corner is emerging from the tree at the exact moment the bridge collapses.

And then the corner vanishes as collapse continues and deck falls:

So it tricks you that it suddenly appeared and vanished at the moment of collapse, but really it was just camouflaged by the tree just before the collapse.

 

[Sym P. le]

I have focused more on whether debris shows up north of the pylon. There is a dirty spot there but it could just be the algorithm or it could be the algorithm because some debris is there. Either way it's the algorithm.

 

[saikee119]

End view of the north end of bridge showing the limit of the temporary fencing. No obvious crack was seen then at the Diaphragm II and the deck.

 

[SFCharlie]

There is a dirty spot...

Yes, it lingers for a least 0.4 seconds. Could be very fine dust, but no aggregate, as it would have fallen in that time...

SF Charlie
Eng-Tips.com Forum Policies

 

[Tomfh]

I have focused more on whether debris shows up north of the pylon.

Yes, it lingers for a least 0.4 seconds. Could be very fine dust, but no aggregate, as it would have fallen in that time...

You can see a similar object earlier on:

I think it may be the rag tied to the reo:

 

[Sym P. le]

Well, that takes care of that one. I'll strike it off the list. It's looking more like OSHA's easy answer isn't supported by the evidence.

Edit: I guess now I can say it's the algorithm because a rag is there.

Edit 2: This also supports the value of the video, rough as it is.

 

[3DDave]

I expect what looks like curvature of 12 is an artifact due to the use of a rolling shutter video capture.

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

 

[Tomfh]

I expect what looks like curvature of 12 is an artifact due to the use of a rolling shutter video capture.

Yes, there are similar kinks along the other members too.

 

[saikee119]

I accept the horizontal protrusion, with a rag hanging underneath, can be just the corner of the deck too. Although the reason why it appeared triangular instead rectangular in the 5 frames of approaching camera still needs to be explained.

It's looking more like OSHA's easy answer isn't supported by the evidence.

Can you point out which part of OSHA's easy answer isn't supported by evidence?

3DDave (Aerospace) and Tomfh (Structural),

The rolling shutter effect in the video is valid. Do you guys have a view on shutter rolling vertically or horizontally?

The curvature effect, easily seen in the shots extracted by MikeW7 from a stationary camera, would support the rolling took place horizontally possibly from south to north. The south face of Member 12 could have been captured slightly earlier than the north face which was already moving when the shutter reached it.

 

[Tomfh]

Saikee,

Regarding the rectangle/triangle/dirt, I’m not sure what it is. I suspect we’re seeing the corner of the bridge and the rag for many of the frames. Could we also be seeing signs of collapse? Or are we just chasing shadows? It’s so hard to say.


I was assuming the rolling shutter was vertical travel, but again couldn’t be sure.

There’s also the possibility of the phone adding rolling shutter? Phones typically have rolling shutters.

Can you tag exactly what you mean by the “curvature effect, easily seen”. I want to make sure I’m on the same page as you guys.

 

[MikeW7]

ROLLING SHUTTER EFFECT

The image sensors of most digital cameras are continuously active, but they can only record a portion of the image at a time because of data transfer limitations (limited bandwidth). The image is "taken" by sequentially reading rows (or columns) of the sensor and transfering the data stream to a storage medium to create a file. When one file is complete, the next read-write cycle begins.

The rolling shutter effect results when there is rapid motion perpendicular to the camera's scanning direction. If the data scanning is by rows, top-to-bottom, the top portion of the image sensor is read before the bottom portion, so any part of an object that actually looks like "|" is recorded as "/" if the object moves rapidly to the left, or as "\" if it moves rapidly to the right.

This phenomenon also causes problems when taking pictures of a video display, since displayed images are created using a similar technology, in reverse fashion - a source data stream is "drawn" onto the screen one line at a time.

 

[MikeW7]

Can you tag exactly what you mean by the “curvature effect, easily seen”.

If your image viewer allows you to view consecutive images using the L-R arrow keys, you can bounce back and forth between Frame 049 and 057 and note that the top of member 12 seems to move to the left (south) in Frame 057, making it appear to "bend" as the canopy collapses.

If the traffic camera is scanning top-to-bottom this cannot be explained by the rolling shutter effect I explained in my post dated 9 Jul 19 15:04 since it is the top of member 12 that appears to move, not the bottom which remains stable against the background. If the scanning is bottom-to-top then the "bend" at the top may be due to rolling shutter if the top of member 12 suddenly moves to the left.

 

[SFCharlie]

In what now seems like a long time ago, several of the more experienced members of this forum were discussing and drawing pictures of how much of 11 and 12 were missing. It seemed like there was more missing than was accounted for on the top of the pier and the deck. Someone suggested that NTSB should check the bottom of the canal. I think this may still have merit.

SF Charlie
Eng-Tips.com Forum Policies

 

[MikeW7]

It seemed like there was more missing than was accounted for on the top of the pier and the deck.

The missing stuff was probably in the big debris cloud visible in dash cam frames 079-080. I think it was caused when the 12-deck joint literally exploded.

(Apologies in advance if somone figured this out earlier - this thread is getting way too long.)

When the canopy and deck both hinged north of member 10, the bridge was divided into 4 sections:
[ul]
[li]A solid truss between members 1 and 10, rotating counter-clockwise (when viewed from the west) about a pivot point on the south pier.[/li]
[li]A solid "L" shape formed by member 12 and the north canopy.[/li]
[li]The deck section north of member 10.[/li]
[li]The remains of member 11.[/li]
[/ul]
EDITED to add analog clock analogy, with truss pivot point located at center of the clock:
The "L" shape and north deck may have initially remained connected in some way, but as the 1-10 truss fell:
[ul]
[li]The north canopy and north deck were pinched together at their southern ends (bending the "L" shape if member 12 is still firmly anchored to the deck).[/li]
[li]The north deck was pulled southward because its hinge point is below the truss pivot (big clock hand moving from 9 o'clock toward 8).[/li]
[li]The north canopy was pushed northward because its hinge point is above the truss pivot (big clock hand moving from 10 o'clock toward 9).[/li]
[/ul]
These effects spring-loaded the "L" shape, and 12-deck joint ruptured explosively (based on the rapid growth of the debris cloud, visible in the dash-cam video that was running at 5fps) as the north deck was literally pulled out from beneath member 12, and the bent "L" shape straightened itself out.

The spring-loaded "L" theory might also explain the apparent "bent" shape of column 12 seen in traffic cam Frame 057, which was taken at about the same time as dash cam Frames 077-078.

 

[Vance Wiley]

I think I recall an excavator dredging the canal at the time of the collapse. ?? Did that activity continue?

 

[MikeW7]

I think I recall an excavator dredging the canal at the time of the collapse. ?? Did that activity continue?

After reviewing the North View timelapse, it appears the excavator parked immediately, remained stationary for a few days, then all the equipment from that area was removed.

 

[MikeW7]

Here are before/after views of the canal from the North View timelapse, enlarged 400% by nearest neighbor algorithm (each original pixel becomes a 4x4 block of pixels). I couldn't crop any further north because the bottom of the images is the actual bottom border of the original video.

I can see some debris on the concrete landing north of pier, but no dust poofs in the dirt or splashes in the canals - plus the canal is still roiled after the excavator lifted a bucket of sludge.

Cropped Frame 26422 from timelapse video "Bridge-109 Mar 8-19 2018 600X-1080.mp4"

Cropped frame 26423 from timelapse video "Bridge-109 Mar 8-19 2018 600X-1080.mp4"

 

[Tomfh]

I can see some debris on the concrete landing north of pier, but no dust poofs in the dirt or splashes in the canals

It may be ejected debris that landed north, or it may not. It could just be shadows


edit: ok, it's probably debris:

 

[MikeW7]

The OSHA investigators evidently never looked at the dash cam or traffic cam videos:

Diaphragm II experienced a blow-out of concrete at the junction of diagonal 11 and column 12 creating a hole shown below in Figure 58 through Figure 70. As a result, column 12 lost support over the pylon and failed with the top tilting approximately 80 degrees towards the south, as shown in Figure 52. The base of column 12 shifted a few feet towards the north but remained on the top of the pylon. Thereafter, collapse of the canopy, diagonal 11 and the deck followed.

Let me rewrite that in simpler terms (with my comments):
[ol 1]
[li]Diphragm II suffered a blow-out at the junction of members 11-12, as shown by a bunch of pictures. (Some of the damage shown in the pictures may have been caused when member 12 separated from the deck and fell behind it, *ADD* and when 12 slipped off the deck and sandwiched 11 between the canopy and deck - see dash cam Frame 81.)[/li]
[li]As a result, column 12 fell over 80 degrees to the south. (How? The top of 12 was firmly attached to the canopy.)[/li]
[li]The base of 12 shifted a few feet north but remained on top of the pylon. (It slid completely off the pier, dropped 3-4 feet, and was dragged back on top by the truss canopy.)[/li]
[li]After column 12 fell over, the remainder of the bridge collapsed. (When member 11 failed, the deck hinged north of the member 10, and member 12 began it's wild ride midway through the collapse.) [/li]
[/ol]

I wonder how their interpretation of the collapse affected their explanation of its causation?

 

[Earth314159]

As a result, column 12 fell over 80 degrees to the south. (How? it was still attached to the canopy!)

There was a plastic hinge created in the deck at the #11/#10 joint. Once the base of #11 can't transfer the horizontal shear, there are large bending forces in the canopy (hinging at the joint) and in the deck where the moment is maximum at the #10/#9 joint. The canopy and deck can not withstand these bending forces (not even close to withstanding the bending moments) which causes the plastic hinging. As the truss collapses, the canopy/#11/#12 triangle rotates the noted 80 degrees. The large bending forces in the canopy can also make it curve (assuming plane sections, the curvature in the non-plastic region is M/EI but you also get curvature and even complete rotation in the plastic region).