Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations waross on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Miami Beach, Champlain Towers South apartment building collapse, Part 05 111

Status
Not open for further replies.

SFCharlie

Computer
Apr 27, 2018
925
US
 
Replies continue below

Recommended for you

Thermopile, Good point. It also seems pretty clear that the pool deck closest to the building collapsed while the pool deck on the other side of the column stayed partially up.. or we'd be able to easily see column 28/M12.1. But it does seem like with the damage down there something very large would have had to of fallen to be the trigger.
 
julootamu said:
Has little to do with the why the structure collapsed.

Nothing to do with the collapse, but the electrical discussion started when people questioned how some lights were still on in the video when the building had already fallen 20 feet. The bluish colored lights are most likely arc flashes from the feeder wires getting pulled out.
 
Thermopile said:
warrenslo where are you???

This guy?

warrenslo - July 6 said:
Alan knows it was the roof anchors but cant prove it yet, boom:
Link

Allyn “Alan” (sic) Kilsheimer - July 12 said:
“We don’t know enough yet. This is in the infancy of this particular thing. And, all these armchair quarterbacks — both engineers and non-engineers — out there that have all these ideas about why this has happened, they don’t have any basis of fact for it,” Allyn Kilsheimer, founder and chief executive officer of KCE Structural Engineers, said in an interview Saturday.

“We’re basing everything we do on facts, and we don’t have any inkling right now on why this occurred,” he said. “Right now, I don’t know what caused it. I didn’t know what caused it the first night. I still don’t know what caused it now.”

Hired engineer: Still ‘no inkling’ to why Surfside building collapsed
 
jbourne8 said:
But how these get triggered is by percentage of movement.. so they won't pick up palm trees, etc, swinging in the breeze

My IP cam has a 0-100 sensitivity setting for the trigger.

You can also set how the triggered activity gets recorded. I can record up to 20 seconds *before* the detected movement (not yet flushed from camera memory by new data) if I set it as such.

 
Electrical folks - good info on the phases.

Could the units be split-single phase, or would this require an extra transformer on each floor?

The main answer I’m trying to get to is how many different phases - that is, 120 degree phases off the 480 system - left the electrical room on each floor, and in what combinations.

Was each floor balanced across all three phases? Was each floor given 2 of the 3 and the system balanced across all floors but not each floor? Or did each floor have a split-phase 120/240 and only one of the 480 phases was active for each floor?

Because it looks like some of the floors go out at the same time, leading me to think they shared a phase. Ultimately, if we knew which phases fed which units, it may be possible to positively identify the floors by which other floors they were sharing a phase with, based on when the lights went out.

What got me started on this was the light in an x10 unit being turned on during the collapse. I was hoping to compare that to which of the x10s were occupied at the time. Unfortunately, all of the upper ones were, so it’s no help. But maybe the electrical outage sequence could lead to pairing up two sets of lights and confirming they are floors 9 and 11 for example, as opposed to 8 and 10.
 
The more I look at the slab step from pool to patio level in the K-M 9.1-11.1 area, the more concerned I get. The construction details show a couple of bottom bars and a couple of top bars making a beam of sorts at these slab transitions. However, the pool level slab is basically hanging from the bottom of the 'beam' which means that it is really just relying on its own steel and concrete in tension for support. To complicate that, heavy planters are installed right on that edge in really the worst possible location.

Considering some of the other alarming things we have seen in the reinforcement steel, it is possible and maybe even likely that the pool slab was poured first and then the top half of the 'beam' and patio slabs were poured on top of the pool deck. This would pretty much eliminate any support along that edge. Also, although there are small beams along the K-M column lines, these are also stepped with the deck. Depending on how the steel was laid and how the slabs were poured, these beams may not provide much support.

My theory is that the pool level slab lost support midway between lines 9.1 and 11.1, probably right at the step-up. This overloaded M11.1 and L11.1 for certain, and probably also K11.1 to start. I am 90% certain that the stills from the garage entrance show that the planter that was on top of M11.1 punch sheared over the column and this caused the column to collapse. I think that the lower yellow part of the column is inside of the planter and the top part is broken and laying on top of the planter. The next bay to the north (M-N) does not have a patio level slab and appears clear of debris.

If, as I suspect, the failure was midspan, you would then have the patio slab hanging off the the 9.1 columns and likely attached with a few hook bars. While the moment load would be large, I don't think it would have been fatal. However, it is not hard to imagine the hook bars taking large chunks out of columns K, L, & M on the 9.1 line. The building might withstand one damaged column, but three in line on the outside line of the building would likely bring it down. Had everything been on the same level, the bars would have been continuous over the columns and this probably would not have been an issue.

There was also progressive failure of the slab from north to south, which from witness accounts probably happened before the 9.1 column line failure. However, none of this really mattered, once the slab failed between 9.1 and 11.1 the building was doomed. Similar column failures did not happen along column line I on the southeast side of the building because the slab was one level and dropped as a unit instead of being cantilevered and damaging the column as hook bars pulled out.

Potential Causes:

1) Design - While the design my have met all RCC codes at the time, the slab steps midway between column lines is less than ideal, especially with the planters added right on that step. Drainage for the exposed slabs appears to have been inadequate.

2) Modifications - It appears that planters were added over the years that may have made matters worse. These is a flat planter south of the original taller planters and the large square planters on line 11.1 appear to have been added as well. These contained palm trees for several years and added considerable weight.

3) Construction - Rebar placement does not appear to have been done according to plan and there are several disturbing pictures of atypical punch shear, evidence of very poor column/floor attachments, etc.

4) Maintenance - Allowing water infiltration for years was clearly an issue and needed to be dealt with more aggressively. It also appears that some of the work was not done properly and this all contributed. Not that one top bar was the difference, but I would not like to have to defend coring through it when methods exist for avoiding that issue.
 
In this instance it is mostly just an indication of poor rebar installation. BUT a closer look seems to show the bar from the left passing behind the bar from the right and all may be OK.
IF it is two bars terminating in a gap, the problem becomes one of column capacity, particularly after 40 years.
Columns suppor axial load, and both concrete and steel shorten under compressive loads. Initially they share loads according to their ED ADD relative area and elasticity. Add the creep of concrete when under continuous loads for long periods and the concrete begins to transfer some of that load to the steel, which does not creep. The transfer from bar to bar is usually thru bond and concrete shear between the bars.
The gap would not allow that transfer.
In the perceived case of a gap, at a slab connection, either bar would have no development length beyond the gap, and therefore no stress development to provide anchorage and resist moments induced at the joint. In short, if all bars were 'gapped' there would be no stabilizing effect at the joint and the column would likely loose capacity.
EDIT ADD When a bar in compression MUST be aligned above and below the splice, mechanical couplers are used to maintain the alignment and contact between the bars spliced. That maintains the loads and passes that load to the bar below. The couplers come in several forms - some are fitted and fused in place by igniting a powder to melt and lock in place, some are crimped in place just like a butt connector in an automotive wire, and some thread the bars using a steep taper and short coupler threaded together, similar to pipe.
 
Garage image folks - if the contention that the building never left its footprint to the south is true, then all of the objects in the tiktok video may actually be the same objects visible in broad daylight after the collapse. Perhaps these could be mapped in terms of how they would appear to the tiktok user and we’d get a better perspective.

For example, are any of these objects visible?

D3B2B52B-6733-4D9F-B291-2F33035C711B_bw8uyx.jpg
 
Js5180, I think in that USA today garage video it looks like you can see the top of the pool slab. It looks something like this:
Untitled_drawing_g-1_kft72m.jpg


It's very hard to make out though, so I'm not sure that's correct. But if that's the case.. we wouldn't be able to see any of those objects. Since we can't clearly see column 28/M12.1 it's hard to say though. I think that white object on the ground might just be one of their planters too, not the missing column.
 

I can understand that KCE wouldn't want to speculate at this point in the investigation but its not quite right to say that its too early to form a hypothesis. We have the original plans which are LOADED with red flags that strongly suggest the original design was flawed. A serious investigation of the failure begins, logically, with a careful analysis of the loads on the building and the structural capacity at each critical location based on the original design. You cannot assume that because the building stood for 40 years that the design was correct, whether or not it was done to code. If after this analysis, you don't find a flaw then you move on to other causal factors, including concrete and rebar deterioration, construction errors, additional loads added later, catastrophic loads immediately prior to the collapse, etc.

As far as this talk about various roof elements causing the collapse, I think most structural engineers, myself included, are not that interested in this discussion, maybe it triggered the collapse, or maybe not. What really concerns me is that Morabito inspected the building and did not really find anything that would make any engineer think that the building would collapse 3 years later. The worst damage he shows is to the pool area which stayed completely intact throughout the oollapse.
 
Ok, JBourne, that vertical bar in the gate, where the diagonal ornamental bars join, right of the “C” - from the camera’s point of view, where would that line fall on the pool deck in my picture?

It should be an exact line, and from that, maybe some of the other details match (or don’t match). But the line can be drawn exactly if you have a few known points, which we do. Every object in the Tiktok video can be placed within a triangle in the pool deck, and for one dimensional objects like the vertical gate bars, you can draw exact zones across the pool deck.
 
tmwaits1,

IMO, we only care about the trigger, in order to understand the design, construction, or maintenance issue it exploited, or in reality the combination of factors that led to the perfect storm. With this information, designers and code agencies can determine requirements that prevent tis tragedy from happening again. Collective brainstorming aids the process, but unfortunately it brings excess baggage too. If I am Alan, I want us to shut up on one hand, but on the other, I bet he has a groupie reading this forum.

Catch 22
 
So I've fleshed out my completely speculative deterioration hypothesis. In short, the drop beam at the planters is not detailed to transfer positive moment across the beam even though the reinforcement suggests that it was designed as a 2-way slab (top bars match adjacent slabs top bars). It took 40 years because the #3 stirrups needed to deteriorate before failure started.

Questions that would serve to illuminate:
- Is there any rusting of the #3 bars at the slab drop beam?
- How far are the #5 hooked bars developed in the slabs at the transitions?
- Did they hook the "Beam A" bars into the columns?
 
 https://files.engineering.com/getfile.aspx?folder=325ca65d-6749-4a61-ba79-65e0cd8091f6&file=8777_Collins_-_Deterioration_Hypothesis.pdf
tmwaits1, I also find the roof discussions a bit boring with such a lack of evidence. Not to mention the building inspector being up on the roof 14 hours before collapse and not seeing any problems up there. They're almost as bad as the car discussions.

Js5180, I think if the pool deck were still standing, it'd be about here.. it's hard to place within the rubble though since you can't see it. I'm not sure if it's a diagonal cut either, seems like it could be straight and just sloping down in the direction of the V during the video.
Untitled_drawing_g-2_cbjhc6.jpg
 
jbourne8,

But the PE SI had thoroughly inspected the facility, and was managing current phase of work. So it appears he saw no issues that would lead to this being an emergency. You can not see hidden problems with mostly a vision analysis.
 
Thermopile, Yea, mostly it seems relevant that the roof was inspected because if anything was in danger of falling off the roof, it seems like he would have noticed. But he also only inspected the roof, and not the garage.. which is where the collapse started.

Teguci, Nice write up! That white line I drew where the slab looks collapsed in that garage video is also suspiciously close to where you suspect the failure occurred. The rest of the slab nearest to the building seems to be on the floor of the garage, and M12.1 seems like it might be already puncture sheered. It all lines up really well with your thoughts on it too. It's nice to see the deterioration theory clearly written up.. although, very scary too.
 
I was thinking more like this - forgive the hand drawing.

Basically, tracing rays from the camera, through each bar in the gate, and out to the pool deck. We have exact blueprints and dozens of fixed objects to use. I may be able to work on it later, but this is the idea:

1948CC4B-F294-4610-9723-9167A7B4912D_yeyk7j.jpg


Then you can compare objects in each zone of the pool deck with objects in the tilt ok and try to determine if any of them match up. I’m thinking that white planter box is a candidate.
 
jbourne8, So you can prove it started at bottom. I am impressed.

Why are you ignoring facts like the fully developed crack in the cantilever/parapet over the 12 th floor. I assume because we don’t need to worry about suspended slab problems, because you already know it was solely bottoms up?
 
What if the crack just appeared due to loads being placed on cantilever by roofer?

Actually it might have been just over sliding door line?
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top