Miami Beach, Champlain Towers South apartment building collapse, Part 11 54

[SFCharlie]

A continuation of our discussion of this failure. Best to read the other threads first to avoid rehashing things already discussed.
Miami Beach, Champlain Towers South apartment building collapse
Miami Beach, Champlain Towers South apartment building collapse, Part 02
Miami Beach, Champlain Towers South apartment building collapse, Part 03
Miami Beach, Champlain Towers South apartment building collapse, Part 04
Miami Beach, Champlain Towers South apartment building collapse, Part 05
Miami Beach, Champlain Towers South apartment building collapse, Part 06
Miami Beach, Champlain Towers South apartment building collapse, Part 07
Miami Beach, Champlain Towers South apartment building collapse, Part 08
Miami Beach, Champlain Towers South apartment building collapse, Part 09
Miami Beach, Champlain Towers South apartment building collapse, Part 10


SF Charlie
Eng-Tips.com Forum Policies

 

[SFCharlie]

----Sorry Maud---- I wrote this while you were posting. Your right.
I have been thinking, (dangerous, I know.) that since to pool deck, itself, remained remarkably intact, with the columns punching through one at a time. That would result in a series of banging...

 

[MaudSTL]

Sorry Maud

Not to worry. I am just trying to make sure we are all on the same page where these warning signs are concerned. I have always thought the knocking was rebar popping, but I am not qualified to have an opinion.

The tricky part of the first collapse at 1:10 is that if it occurred in the garage, the Vazquezes either couldn’t or didn’t see it as they parked, walked to the elevator, and boarded it to go up to the lobby. As you know, they were in the elevator when the deck collapsed. And all their statement says is that they heard loud cracking sounds a few seconds before getting on the elevator.

 

[w-streng]

We all speculating of course but for interest sake, my thoughts have been that the sounds are of the masonry units taking up compressive stresses/strains as the vertical load on then slowly increases until they finally fail in crushing and/or buckling and then 'crash' onto the floor slab above the Nirs.

The fact that this precedes the collapse is what intrigues me most.

 

[MaudSTL]

The fact that this precedes the collapse is what intrigues me most.

More than one engineer has commented that the building was issuing warnings that no one understood.

Your theory is the first I’ve heard that implies that the first collapse actually could have come from above. Most folks have been been thinking part of a slab collapsed in the garage. If part of a partition on the second floor actually did collapse, I would imagine that could have contributed to a column buckling, right?

 

[w-streng]

The video evidence points to the first actual structural failure/collapse being the pool deck area, no disputing that from my side.
I don't have a theory, far too many unknowns, I'm just trying to figure out how the pool deck collapse was preceded by creaking in the walls above the Nirs. It's easy to say that load was being re-distributed but WHY. The high-rise portion hadn't lost any structure at 11pm the night before.

Your Q: The partition wall collapse would very unlikely buckle a column. It shouldn't be taking any load and collapsing anyway, something would already have gone very wrong for it to be taking load.

 

[Spartan5]

The sounds of structural distress, wherever it is, will be transmitted through all of the interconnected pieces of the building. The floors, having the most coverings (flooring, furniture, clothes, rugs, etc. will appear effectively transmit less sound than the walls and ceiling (which has the least amount of coverings on it). Hence, “noises from above.”

 

[CE3527]

I had two points about over reliance on the direction of noise in the forensic analysis, but since Spartan covered the first I only have one.

The physical evidence suggests a ground level failure originating at the garage. If a ground level column failed at the exterior face of the building, which appears to be the case, the remaining column directly above the failure ceases to carry the structure load and begins transferring load to all adacent, remaining columns. This progression overstresses those remaining columns at all levels, which would explain noise and failure at every floor of the building. It is entirely possible for a garage level failure, to ultimately result in a collapse that begins at one of the upper levels.

 

[IEGeezer]

I recalculated the weight of the planters based on the suggestions given herein. Geofoam was specified on Sheet A2C-1.0 (Level 1 Floor Plan) (pg 4) from 8777-collins-avenue---preliminary-review-plans-for-40-year-re-certification.pdf available at

https://www.townofsurfsidefl.gov/de...rk/champlain-towers-public-records-documents.

It is unknown if geofoam had actually been installed in the planters, but I assumed for the puirpose of these calculations that it was. On drawing LP-1 (pg 73 from the same source) (the lanscaping plan) the specified soil was to be installed to 18 inch depth, but 36 inches surrounding tree root balls. There is a disconnect because on Sheet A2C-1.0, the engineer specifies 1 to 6 inches and no more than 80 lb/f3. Waterlogged soil can weigh up to 130 lb/ft3 regardless of what the engineer wants. Therefore, I assumed 18 inches of waterlogged soil for the short planters and 36 inches for the tall planters.

On Sheet S2C-2.5 (pg 19 from the same source and titled Plaza Details) the planter walls are specified as 8 inch CMU wall with reinforcement, so I changed the width from 10 inches to 8. In addition, I used page 11 from the same source, titled Level 1 Slab Reinforcement and Framing Plan, for remeasuring the planter dimensions using the scaling and measurement features of Acrobat. I set the scale to the one shown on the plan and there was almost perfect agreement when I measured a labeled dimension. Accordingly, I changed all the size of the planters to the new measurements and I noticed that I had missed some. I am attaching and embeding the new plan with the measurements as well as a pdf of the calculations. Maybe this might shed some light on the collapse.

 

[MaudSTL]

…noise and failure at every floor of the building.

From 11 PM through 1:10 AM, we have reports of increasingly intense knocking noises heard on the first floor, and at 1:10 we have reports of a loud crash heard on the first floor, resulting in dust particles in 111. Other than the five people who were extracted from the rubble, the only other survivor in the part of the building that collapsed, Ileana Monteagudo up in 611, was asleep until around 1:15.

I’m trying to understand whether we can dismiss any direction of the sound origins as a possibility, and also whether there is a certain direction that is logically more probable based on what we believe are the origins of the building collapse.

So please check my understanding. Is it correct to say that both the sounds—knocking and crash—could have originated on any floor, even though we agree that the building collapse itself was caused by a ground level failure?

If that is correct, then can we say that the directionality perceived by the survivors may or may not be accurate…that the sounds could have come above, below, or off to one side or another?

 

[IEGeezer]

Based on the measurement of the weight of the planters I think that the probable point where punching shear first developed was column K13.1, followed by K12.1 as shown on the attached Sheet titled Level 1 Demo Plan and identified as Sheet D2C-1.0 from the Morabito bid documents referenced in my prior post.

Previously I9.1, K9.1 and L9.1 (as identified on the attached pdf; the numbering may vary on other drawings) had been identified as probably being the first columns to develop punching shear. However, based on the principle that a chain breaks at its weakest link, I would like to suggest that punching shear most likely developed at columns K13.1 and K12.1 before it developed at I, K and L9.1. Columns K13.1 and K12.1 are 12 in by 16 in type N columns, whereas I, K and L9.1 are type C columns which are 16 in by 16 in. It would stand to reason that the skinnier columns failed first. Column K13.1 is almost directly behind the elevators when one is in the vicinity of the office. Was that the loud sound that the security guard heard from the direction of the elevators? Maybe. Just food for thought. I have also labeled the weights of each planter shown on the drawing based on my previous post showing the recalculated weights for illustrative purposes. There are about 24.8 calculated metric tons of weight from the planters in the vicinity of columns K13.1 and K12.1. The video from Building Integrity influenced these thoughts.

 

[IEGeezer]

RE: Sounds

There is a difference between the planned demolition of the part of the building that remained standing and the part that fell. The columns of the part that was demolished suffered compression failure (due to explosives). What is striking about the debris field is the number of columns from the collapsed portion that still look relatively intact. Consequently, the columns from the collapsed part did NOT suffer compression failure. They failed by buckling.

When the punching shear developed, the columns in the center portion tried to buckle, but the building was still braced by the north and south sides. The swaying (or attempted swaying) movement must have caused very large stresses throughout the building and would likely explain noises from above and possible structural failures in the upper portions of the building before the collapse. Just a thought.

 

[12345abc6ttyui67]

RE: Sounds

I do not have any numbers to back this up. But just trying to visualise things I would expect that as soon as the lower (say ground floor) portion of the column is experiencing any distress (lets go with it's beginning to buckle because the pool slab has punching sheared off from the column and doubled the columns unrestrained length) the column loses some/most of it's load bearing capacity.

Most/all of the loads once carried in that column are now trying to find any alternative load path and redistribute to parts of the building which can carry them.

Since the column has failed at the bottom, but was supporting every level of the building; these alternative load paths could be for instance all floor slabs dropping into catenary action between the remaining exterior columns and cracking along the bottom face; or all slabs attempting to cantilever from the interior columns putting hogging moments (and cracking) over the top of the slabs at interior columns; or all slab to failed exterior column joints trying to hold together as a Vierendeel truss and taking up moments they are not designed for. Evidently things got so bad none of the possible alternative load paths, even acting in whatever combinations were possible, had enough capacity to hold the load being redistributed from the failed column.

In reality the building will have tried to hold the buckled/failed column loads any way it possibly could, so there will be a whole heap of these possible alternative load paths all in play at once. Maybe the CMU walls even tried to take some load (explaining the finger wide cracks in 611[?]) during the later stages of the attempt to redistribute the loads.

Obviously the load will take stiffest (stronger) paths first, progressively moving into less stiff (weaker) load paths until either the load is finally redistributed, or there are no load paths left to try.

I have been holding my tongue at the comments trying to suggest "sounds from above" = "collapse initiated above" due to this reasoning.

I expect this explains the strange noises occurring all round the building.

[Edit] The only thing I am unsure about is relative timescales. I expect most of what I have described above would only really take seconds to go from intact column > attempted redistribution > collapse; which then doesn't marry well with the several minutes of reported sounds. The only explanation I have is that the building was really really close to having enough capacity to redistribute the load, but was ever so slightly overloaded, hence cracking progressed slowly bit by bit, little by little at first. As capacity was lost though, this cracking would speed up as more and more load paths were comprised. Like I say, nothing to back this up, just my pragmatic view of what could be happening.

 

[Vance Wiley]

Very well described.
I had actually begun to visualize that scenario sometime in the last two days and you have explained the concept quite well. Thank you for posting when you did.
With a lower column failing, there could have been brief instances where the remaining sections of the column were in tension, pulling downward on the slabs above as the building developed alternate load paths in all the remaining slabs above, however momentarily.
The brief loss or even reversal of shear stresses in the slab around the perimeter of the columns may explain the seemingly clean punching shears seen along some columns.
Thanks,

 

[StructuralMadness]

I think this is where Bifurcation analysis of column buckling instability can help understand the timeline. So a lot of engineers don't know about this because it is hardly ever taught in colleges or have to wonder about in practice (I was surprised to know about this analysis mainly because I just assumed that columns buckle as soon as the critical buckling load is reached, the column should collapse). But when we calculate the critical buckling load (CBL) of any column, we assume that it will fail immediately as soon as the load exceed the CBL. In reality the column can handle a bit of additional load beyond the CBL even if it has buckled. Based on the load and slenderness of the column, the rate of this buckling deformation varies. It is like an inverse exponential curve where load is on the Y axis and deformation on the X. What buckling equation predicts is the point of initiation of buckling of column at this CBL. Now as soon as load exceeds CBL, the rate of column buckling and there by instability and lateral deformation, starts really slowly. As the lateral buckled deformation increases, the column sees additional P small delta moments. And you get a small differential deformation difference between this buckled column and surrounding columns that are intact. But initially the slab would not have redistributed the load as the differential deformations are relatively small.

Then as the time progresses, this rater of lateral deformation starts increasing, and it reaches to the point of no return. This is the bifurcation point. Which explains that now even if the load reduces a bit, the column is never going to be stable. It is on the verge of collapse. As this lateral deformation rate increases, the differential deformation between columns increase and the slab tends to redistribute the load. Again for slab it takes time, because it is a flexible body (8" slab spanning 20-24 feet). As it redistributes, there will be cracks formed in the plaster / ceiling finishes. This can be observed in the ring video of one of the apartments. Like there were some initial spalls of plaster as this was happening.

As the column literally buckled significantly, now the stresses redistributed significantly in the slabs, but since these were flat slabs with no integrity reinforcement, and the brittle nature of punching shear failure, the collapse chain of series of punching shear failures was triggered and it was unstoppable unless it was encountered by a wall / bigger columns / columns with less tributary areas. I think this could explain why there was a pause of 7-8 minutes between pool deck failure and building collapse.

Now typically it is not that easy to buckle an RC column. So I am also interested in understanding what was happening at the step - beam column joint. Were there any ties provided in the column through the region of the step beam or not. And based on construction sequencing, I am assuming that for this 2.5' height of column at the step beam column joint, the contractor may have used 4000 psi concrete instead of 6000psi for columns. As there were no requirements for puddling back in ACI 318-77. So if these joints were not constructed and detailed properly, then it could explain why an RC column failed so easily.

 

[MarkBoB2]

It’s me again. But before you begin the ridicule again, please hear me out.

The photo below shows the debris pile that was left at the end of the ramp for two days after everything else was removed. I believe that this was done because the Miami Dade recovery crew believed that the pile contained important evidence about the collapse. The importance of this evidence is clear from their using red paint to identify some of the objects and by their use of red cones connected by red tape on the ramp to keep away curious people who might tamper with the evidence.

I encourage you to zoom in on the objects marked in the photo to see what they are. To help you to do so, I have enclosed a digital PNG version of the photo that allows zooming in by rotating the mouse wheel forward after opening the document in Windows 10. You can also re-center the objects as you zoom in by clicking on the window and then moving the mouse from side to side or up and down. I have also enclosed in the following post a PowerPoint document containing the same photo that allows one to zoom in by using the slide button in the lower right hand corner. This document also allows one to selectively remove any annotation that obscures an object and to add new annotation as desired by the viewer.

What I found recently by inspecting this photo, and what you can see yourself by zooming in on it, is shown in the following figure.

One sees that there is a large amount of wire cable on the pile that does not seem to be evident in debris photos taken at other locations. I was able to trace at least 20 feet of this cable, although there is much more, perhaps over a hundred feet. Most importantly, I found that one end of it goes into the circular top of an object of interest on the debris pile as shown in the following figure.

I believe this object is the condenser of an air conditioning unit. Near this object lies a square piece of sheet metal with rounded corners that has a circular hole inside. I believe that this the metal top of the air conditioning unit. One would expect the cable to go through this circular hole as well. But one can see a slot in one side of the top that has allowed the cable and top to be separated. This slot could have been made by the recovery crew to allow easier inspection of the top, or could have been made by pulling on the cable by a power shovel while the top was still covered by the debris pile.

Now, I would wager that it is no accident that this cable extends into the internal condenser part of an air conditioning unit. Normally, the top of an air conditioning unit is covered by a wire grating that would prevent a cable from randomly penetrating inside. However, if one wanted to raise or lower an air conditioning unit with a cable, instead of putting it on a platform or using a lifting sling in some manner, it is easier to just remove the circular grating in the metal top and pass the cable hook through the top opening, then around a compressor support bracket at the bottom of the condenser, and then reattach the hook to the cable again. This alone would indicate that a human being was responsible for attaching the air conditioning unit to the cable.

So, why is the object of interest (whether an AC or some other object) still intact and still attached to the cable after the building has collapsed? Clearly, something went wrong during the raising or lowering of the object that prevented the person involved from detaching the cable. This something that went wrong could not have happened near the top of the building, because then nearly all of the cable would still be in the winch and not strewn all over the debris pile with the object still attached. Therefore, it must have happened near the pool deck. But if the winch malfunctioned in some way to cause the object to fall freely to the pool deck with the cable still attached, then the object likely would have been dented beyond all recognition by the collision with the deck floor. Clearly, the TikTok video shows that this object is still relatively intact. But what if the something that went wrong occurred just a few feet above the pool deck? This could happen if the cable was not long enough. In this case, the users involved might be tempted to gain a few more feet to safely put the object on the deck by tilting the crane at the top a little bit. But when this was done, the change in the crane’s center of gravity could have caused the crane to press against the parapet, causing a moment arm of over 7 feet (4 feet of parapet height plus 3 feet of clearance) with over 300 pounds of force to be applied. This may have caused part of the parapet to give way, sending it crashing down onto the most sensitive part of the building’s structure; namely, the beam between column 27 (column M10) and column M11.1. In this case, the object of interest would fall only the distance of one floor plus a few more feet, which would enable it to stay relatively intact. It would also have resulted in the object of interest falling midway between the hoist counterweights on the debris pile as observed in other photos because the object would have been pulling down on the hoist structure, thereby keeping it centered on the object.

Some further thoughts about this scenario are as follows:

1) It is likely that there was a second person at the bottom of the building as well as one at the top. This is likely because someone is required at the bottom to remove the hoist hook from the object and cart it away, while someone at the top controls the hoist. This second person was likely in cell phone communication with the one at the top. This would have allowed almost silent operation under normal circumstances after 12:00 AM at night. But this second person might have been seen by a surveillance camera having a line of sight over the patio area of the pool deck. This person at the bottom would have been exposed to the falling parapet and may have been injured or even worse. Nothing has been mentioned by the recovery crew about finding a person near the object of interest.

2) The amount of cable required would have been at least 120’ 8” to cover the distance from the penthouse roof floor to the pool deck below. An additional 8 feet or more would have been required to get over the penthouse parapet (4 feet tall), the height of the air conditioner (3 to 4 feet), and either a cable pulley or a winch (1 foot), thus adding an additional 8 to 9 feet. Therefore, the total cable length required would have been 129 to 130 feet. Most commercial winches have only 120 feet of cable, some have 130 feet, but very few have over 130 feet. Therefore, the amount of cable needed was right at the limits of most available winches.

3) With the pulley or winch needing to be 8 to 9 feet above the penthouse floor, the hoist would have been quite tall. Also, it would have needed to extend at least two to three feet over the parapet to clear a 3 foot square object like an air conditioner, so it would have required a counterbalance over 8 feet long with weights greater than 300lb x 3/8 = 100 lbs to stabilize the hoist in two directions. Two hexagonal weights on rods were found on the debris pile that could serve this purpose. Also, this large hoist would likely have been present when the building inspector was on the roof the day before the collapse, so he should be able to give a description of the hoist.

4) A winch would have been attached to the hoist structure. This winch would have fallen to the pool deck below with the hoist structure, probably causing the winch to be seriously damaged. An object was found on the debris pile that appears to be this winch, with the motor and cable spool separated from the housing. However, no other parts of a hoist structure have been observed except for the two counterbalances and possibly a pulley.

5) Operation of a hoist on a roof requires that the hoist have a tie-down cable to a secure building structure to prevent it from falling. There was no such secure building structure on the penthouse roof near the hoist at the time. Therefore, the workers may have used the aluminum framework of the air conditioner mounts as a tie-down structure. Evidence of an aluminum framework with a second air conditioner attached can be found on the debris pile at the end of the ramp.

6) Could the falling of the penthouse parapet wall above and its crashing to the deck correspond to Maud’s first stage of collapse? The second stage of collapse has been identified as being caused by the falling of the pool deck, which might have taken place after a short pause.

7) Could the “knocking sounds” or “hammering on the roof” heard by a tenant and by the night watchman before the collapse have been caused by compressed air guns being used to remove air conditioners from their mounts prior to moving them?

8) Could the weird smells noticed by some tenants have been caused by the refrigerant being released from the air conditioners that had fallen? This refrigerant is a gas that normally has a faint smell like either ether or chloroform that some people might find to cause a disagreeable odor. Each air conditioner has about two to three pounds of this gas inside that takes about a half hour to release completely. With a total of about sixty or more air conditioners on the roof falling at the same time, this would have caused over 150 pounds of this gas to be released all at one time, which would have been noticeable throughout the collapsed building.

9) There may be a witness who can tell us who was on the roof that night. The security desk employee should have seen anyone who entered the elevator after hours while wearing working clothes. And this same security desk employee may also have seen them exit the elevator to leave the building after the pool deck collapsed, which he/she should also have seen by watching one of the security cameras. Has anyone asked the security employee about these topics?

I will leave it to others to explain who may have been on the roof that night to trigger this sad sequence of events because of potential liability concerns. Suffice it to say that the trigger for the collapse was likely an industrial accident created by carelessness and stupidity, and not merely by structural deficiencies as bad as they were to everyone’s agreement, including my own. Without this trigger, no one knows how long the building could have remained standing with all its deficiencies. The case is equivalent to an old and unhealthy person being run over by an automobile driven by a careless driver. No matter how old or how unhealthy the old person is, the cause of death is still an automobile accident caused by a careless driver. One can introduce new health measures to keep an old person functioning longer, but only by stopping careless driving can one prevent automobile accidents that cause the death of older people.

Enclosure 1 (1st post below):
PowerPoint document: Overheads of debris pile at end of ramp (zoom with slide in lower RH
Enclosure 2 (2nd post below):
PNG photo: Overhead of debris pile at end of ramp (zoom with mouse wheel & re-center by moving mouse)

 

[MarkBoB2]

Here is the PowerPoint version of the same photo:

 

[MarkBoB2]

Here is the PNG version of the same photo:

 

[waross]

I'm waiting for a report from someone who was on-site.
In the meantime, I'll see your collective speculations and raise you two guesses and a maybe.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[Vance Wiley]

The Vierendeel truss concept is most interesting. Losing a lower column in a structure like this could leave a V truss with 12 levels of horizontals and columns forming a large grid. Loss of a single interior column could be balanced and not create shears in the single remaining column stack above. Negative moments at the perimeter of the unsupported bay would cause shears in the remaining columns which would be distributed to the remaining columns thru direct tension and diaphragm action at each floor.
This could supply the "redundancy" not evident in the system. It likely failed in this case because the slab reinforcing was spliced at the column strips and continuity reinforcing was inadequate. And the columns appear to have little capacity for shear and moments.
A comprehensive study of this failure should investigate this redundancy thru V truss action.
One can envision designing this capacity in future designs.
Good call.
Thanks,

 

[SFCharlie]

It’s me again. But before you begin the ridicule again, please hear me out.

Please, Not ridicule, but analysis.

The "AC" seems to be the size of the cones. Either large cones or small AC?

edit: It looks as if this is a screen shot from a video. Could you please post a link to the video? Thanks very much!