# Towards A Replicable Physical Model Illustrating Aspects of the Collapse of The WTC Towers on 9/11

Discussion in '9/11' started by Mick West, Mar 15, 2016.

Tags:
1. ### Polly MathMember

Would the slenderness of a cantilever beam matter if we were testing that rather than the model of the towers? Of course it would.
Why a cantilever ? Because that's what the tower in effect was. It matters for a beam, it matters for the building and it matters for the model.

• Agree x 1
2. ### Mick WestAdministratorStaff Member

What matters? Of course if you change anything in the model, then it will change what happens. But I'm still not clear exactly what you think could be improved.

Could you draw two labeled diagrams. One showing things as they are now, and the other as you think they should be?

3. ### jaydeehessSenior Member

I don't follow. I was asking you what the slenderness of the vertical model (you brought up height to width ratio)has to do with the way this model behaves with respect to the progression of collapse. You now bring up cantilevered beams?
The trusses in the real building were supported at each end, the trusses held the floor. There was no cantilever action there.

Even if all you have is MSPaint you could draw a simple diagram of how you think the model could be improved. I know Mick has already addressed expanding the model to 3-d so you need only work in 2-d.

4. ### jaydeehessSenior Member

Sure I get that.

Even my old ### SD Leitch Velocity NLE (come to think of it, it was a pretty good system in 2001, though getting old even then) can adjust the play speed of a video clip to be pretty much anything over a wide range. I can stretch or compress the x and y size too, or skew it. I can adjust the colour and brightness in various ways too for that matter I fail to understand how any of that would help understand anything wrt your modeling versus the actual tower collapse. You could drop a load of sand from 12 feet up and match it to the speed of the tower collapse.

5. ### Polly MathMember

No that's not what I said. I said the tower was in effect a cantilever beam.

• Agree x 1
6. ### Polly MathMember

No need. I already explained this in an earlier post.
If you half the width of your model, you will have more of a 6:1 ratio (TT being 6.5). You can then estimate for example that the floor slats in your model are roughly 9 floors apart.

EDIT - Do you think the resultant model will take the same lateral hit as your current one ?
I don't.

7. ### Mick WestAdministratorStaff Member

Draw a diagram to remove confusion.

8. ### deirdreModeratorStaff Member

or on a piece of paper (might be easier) and take a pic with your camera or phone.

9. ### Polly MathMember

Drawing it isn't happening right now. Your model is 3x higher than it is wide. The towers were 6.5 x higher than they were wide. If you half the width of your model the height will be 6x the width so closer to the towers.
This means that you can estimate how far in proportion the floors are dropping. It comes out to be around 9 floors that your slats are dropping.
Why would you not want to apply the scale of the tower to your model ?
Estimate for me how many floor each slat is dropping in the model. In the towers it was obviously 1.

10. ### Mick WestAdministratorStaff Member

Because it's impossible.

• Winner x 1
11. ### OysteinActive Member

So what? If it helps you, imagine the model is only the top half of the tower. That doesn't make the model any better or worse, but perhaps it helps you get around your confusion.

This is nonsense. 1 floor in the model represents 1 floor in a hypothetical large tower that share some design properties with the WTC twins. Skipping floors would defeat the purpose of the model.
If anything, you might want to complain that the ratio between story height and floor span is too large, and ask for (much) longer floor planks. Floor trusses in the WTC were 60 and 35 feet. Floor height was 12 feet. So for the short spans a ratio of 3:1 (e.g. 3 ft long planks, 1 ft vertical spacing) would do.
But really, Polly, you need to explain why and how this is relevant to the purpose of Mick's model.

It appears to me that you believe a model is only valid if every ratio between any two features is the same as it is in the original building. That is nonsense, for three reasons:
1. It would be overkill, since many purposes can be achieved with non-scaled models, and even with features missing
2. It would be plain WRONG - because of cube-square law and others
3. It would be practically impossible

12. ### Polly MathMember

Yeah ok. Scale has no bearing in your world mayb. The distance proportionally between the floor in Micks model is what exactly?
I note not ONE PERSON made a estimate when I asked. You included.
The tower was in effect cantilever beam. of course scale matters.

13. ### OysteinActive Member

Stop digging.

14. ### Polly MathMember

Oy, does how stiff the floors are make a difference to the collapse progression ?

15. ### EfftupSenior Member

People didn;t feel it was relevant.

Unlike 1) Richard Gage's Model of cardboard boxes that just has 2 ridiculously strong hollow shells to represent separate parts of the building, Or 2) Cole's Model where he had no core, but very strong single piece outside walls and floor connections that were stronger than the floors themselves or 3) Cole's other models with Pizza thingies that effectively gave each floor many many small vertical columns in between each floor, We instead have 4) Micks;s Model.

Mick's model possesses a strong inner core made out of sections, where the weakest point is the connectors between sections. He has outer walls, also made in sections where again the weakest point is the connectors between sections. He then has floors which connect the core to the walls.
The core is providing much of the structural strength of the building where the floors themselves hold the core and walls together and give lateral and torsional stability to the core.
The weakest point here is again the connectors that connect the floor to the walls and the columns.

The floors are strong enough for entire building sections to be picked up by the floors.
The whole building is stable and sways a little but doesn't collapse from sideways forces. It can take a force of a relatively large weight being smacked into the side of it.

Each floor can hold ( I think it was ) 10 floors of static load. and certainly 6 floors of dynamic load if the load is only dropped from a very small distance.

HOWEVER, if the weight of 2 floors is dropped from the height of one floor, the dynamic load causes the floor to collapse, and all this weight then crashes down by another floors height onto the next floor Which then collapses for the same reason, causing a progressive collapse all the way down the building, unless way more than one floor's worth of material is ejected for each floor travelled.

The walls and core, now having lost the stability the floors were giving them then collapse too, breaking at the connectors also.

It is completely irrelevant how many floors there are in the building, as this mechanism will happen however many there are. Therefore the height to width of the building model is not relevant either.
As far as Mick's model is concerned, one floor is one floor.

• Like x 1
16. ### akaMember

That is not entirely true. You can't stack 30 of these assemblies on each other as-is - the lower ones would need progressively stronger connections to prevent shear and P-Delta; making progression more difficult to ensure.

17. ### Polly MathMember

Correct and thank you. The slenderness and weight to height ratio is crucial as you obviously know.
This is common sense. If you took one of the towers and turned it upside down it would crush itself. Sitting here in a bubble pretending that these ratios and scales do not make a difference is naive. The model is fine and again I applaud Mick for doing it - it is very interesting and I am enjoying watching the journey.
Would be interesting to derive what the potential of the top 10 or 11 floors of the towers would be upright and then inverted. Unlike in this model, buildings in the real world are not uniform.

EDIT I wanted to add that if this model did have a height:width ratio anything near to the towers, when Mick pushed from the side it would have fallen. The only reason it did not is because it has double the width (or 4x the footprint if 3d). If the model had been to scale, the fragility and precarious nature of it would have been evident. The model is currently biased toward progression of the collapse. I don't think deliberately so, but I do think that further more in depth study of the buildings themselves with particular note to the huge safety factors that existed particularly at the upper outer columns is required.

Last edited: Apr 16, 2016
18. ### Mick WestAdministratorStaff Member

Could you draw a diagram indicating which connection in the model would need to be stronger, and how much stronger, and in which direction?

19. ### Mick WestAdministratorStaff Member

But conversely the fact that the falling mass is accelerating at a significant fraction of g makes progression much easier to ensure. i.e. the falling mass get both bigger and faster as it falls. It would seem quite plausible that this amplification of the kinetic energy in the system would be more than enough to overcome any increase in "connection strength" - especially as there seems no good reason for the floor slab connections to be stronger.

20. ### Polly MathMember

Mick, I genuinely do not see where you are making any point particular to the twin towers or any other real world building here. I can see that you are proving a point about the inevitability of collapse in your model, but how exactly are you making a connection between this and any aspect of the twin towers?

21. ### akaMember

Full ack.

I had typed a lengthy post with a few musings about what the model shows and what it doesn't, what shortcomings it has, which can be amended and which can't -- and finally decided against posting it; mostly because it is not entirely clear yet whether Mick is going to continue experimenting or whether this is already the "final product" and last word. Earlier posts seemed to indicate that a better representation of the core may be expected - but instead of replacing the 4x4 with something representing the core structure that took up ~30% of the floor area, it was replaced with an even thinner "column" to allow for its self-disassembly...

It seems different "truthers" have different problems with the model. I honestly find some of the objections not entirely fair (although the "debunkers" surely deserve it, after all the complaints about "truthers'" models) and try to be objective and genuinely generous towards the argument "debunkers" are making.

Mick's stated aim was to "debunk" the Gageian cardboard box model, and it does a great job at it. It visualizes what "debunkers" mean when they describe "zippers" and "pancakes" and how the unbraced core also went down. But prove that the "collapse" of the Twins was a natural one, or demonstrate it could have been, it does not. My interpretation is that it shows how theoretically, in principle, the towers would have to be designed to allow for a purely gravitational collapse.

Scientifically speaking, we would have to somehow devise a way to control the variables to discern the intent of making the building stand up and the intent of allowing it to self-disassemble in a Rube-Goldberg-machine fashion.
I will, but first let me share another small piece of data I could derive from the clips:

I tracked the uppermost edges of the three columns of the three-level tower and their response to the "impact test". Five periods are ~4.7 seconds long, so the three-level tower oscillates at ~1.06Hz (a comparison with the four-level tower would have been mighty interesting). "Natural Frequency" is a concept that often comes up in structural analysis, so I thought someone might be interested in this.

BRB, will return with a display of my mighty inkscape skills.

22. ### Keith BeachyActive Member

http://www.nist.gov/el/disasterstudies/wtc/faqs_wtctowers.cfm see number 12.
Progression of the floor failure is the same; the connections of WTC floor don't get stronger. Thus progression gets less difficult, "easier" (as in more impossible to stop) due to gravity acting on the debris mass longer, and the debris mass getting bigger.

23. ### akaMember

So the left side shows the model as is and what keeps it together. The green vector shows mg, it is easily transmitted into the ground, and any lateral forces resulting from sway and/or "Euler buckling"/P-Delta are easily neutralized by the attractive force of the connectors/magnets (blue and red vectors). The right side shows what I suggest and predict will happen if you stack the same assembly further (if it not outright shears and falls over): the increased pressure from above (red vector) will more easily translate into a lateral motion due to "Euler buckling"/P-Delta, and this force (blue vector) will eventually be greater than the force of the connectors. Here, they will "rip", the "shell" columns move outward (think of a blob of jello) and the whole thing come down unless you make these connections (violet circles) stronger. However, as Keith points out, these connectors were basically the same all over the real towers, so they would all have to be equally strong... what that means for what would happen right after initation should be clear.

Last edited: Apr 16, 2016
24. ### Polly MathMember

Increase the live load on the model to failure as is on the external. Bet it doesn't take much.

25. ### Mick WestAdministratorStaff Member

It's an illustration of the likely mode of collapse, and a refutation of claims along the lines of "you can't destroy the lower part of any stable structure by dropping the upper part on it".

26. ### Polly MathMember

So increase the live load on the external column of your model and find out how stable a structure you actually have then.
ETA Why would you possibly not want to do that

27. ### Mick WestAdministratorStaff Member

Thank you for the diagram, I think there's a number of issues here. Firstly the "strength" of the connections varies in different directions, and obviously the magnets can't model all of them. What they DO model is the supporting strength of a floor (allowing it to carry six floors of weight when suddenly applied, or 10-12 if very carefully applied). They also have a very strong tensile strength (pulling in the side walls), and a near infinitely strong compressive strength (pushing out). They also have a moderate moment resisting strength about the vertical and in/out axes, and a weaker moment resisting strength about the left/right axis.

So simply saying that the connections need to be stronger is a little unclear.

But assuming they do. It would seem like a reasonable test of this to take two 2' sections and assemble them, and then see how much weight they can support piled on top of them?

On a broader note though, this seems something of a different argument. Shifting from "the rapid progressive collapse is obviously impossible without explosives" to "It might be possible, but intuitively I think it's not".

28. ### Mick WestAdministratorStaff Member

I think you mean the static load? Again, a diagram would be helpful.

29. ### Polly MathMember

Look at what the page you already read says - LHS page 3rd column 2nd paragraph.

File size:
4.5 MB
Views:
241
30. ### Mick WestAdministratorStaff Member

This? I'm unsure what your point is. Please explain.

31. ### Polly MathMember

The point is that there are 57 or 19 of the 3'3" or 9'9" column modules per side on the building. The tower had a reserve to take 400 to 2200% increase in live loads. The outer columns would be pretty near the high end of that scale. "stable" is a very general term and you are presuming that because your model stands up you can deem it "stable". Not the case, and not nearly stable enough to make a comparison with anything real world.

32. ### Mick WestAdministratorStaff Member

So can you quantify how "stable" it would need to be? How would I test it? How would you build a model that replicated the stability of the WTC?

33. ### Polly MathMember

At 1:29 in the last video you posted. Before the failed columns of the top section have impacted the floor below, your outer columns are both failing left at the base of the structure. Let's start by defining that as "highly unstable" and face the reality here that you have proven that columns need to be big.
Edit - but preferably slightly less so on top.

34. ### Mick WestAdministratorStaff Member

So here's the model with some additional static weight

One 3 floor section is 7 pounds. I added 31 pounds of additional weight (I say 23 in the video, but missed one eight pound weight). So that's like another four sections, or equivalent (in mass) to a 12 foot by 2 foot structure.

It survives a little swaying, but then collapses when I give it a good poke in the center of mass.

The failure point was the splice plates. Those could certainly be stronger, I kind of messed it up by grabbing the top after it started to collapse (as I wanted to add more weight). But I think it's clear that the connections that need to be stronger are the column-column slice plates. Notable I did not have any on the center column.

• Informative x 1
35. ### Polly MathMember

Try them on the 3rd up. If it survives I will buy you a beer via paypal.
Edit, all of a sudden now there's beer on it, i wish I'd said 4th.

36. ### Mick WestAdministratorStaff Member

You mean this? It's failing because of the falling debris. The model is stable under normal conditions, not while it is collapsing. You seems to be defining stable as something that will not progressively collapse. Rather a circular argument.

37. ### Mick WestAdministratorStaff Member

Survives what? The 31lb of extra weight being carefully placed on top? Me poking it? Throwing a wrench at it?

38. ### Polly MathMember

No. I mean these 2 screenshots from 1:28 of your video. Flick between the images.

39. ### Mick WestAdministratorStaff Member

Keep watching, the columns at the base sway left, but they don't fail. Before the mass hits them, they are swaying back to the right.

40. ### Polly MathMember

I dare you to measure from your bottom RHS lowest outer column splice connection to the line in the middle of your doors from both these shots and scale it up to anything like the towers. The structure is extremely unstable.