GeekFurious
Member
You shouldn't abandon it. But you should consider accurate analogs for what you're trying to simulate.
Buckled Structural Steel in Building Fires
- Thread starter MikeC
- Start date
-
- Tags
- buckled steel
- Status
You shouldn't abandon it. But you should consider accurate analogs for what you're trying to simulate.
Oystein
Senior Member
Whatever: They are obviously strong enough to keep Mick's structure standing even against some lateral load.
What makes you think they are too weak? Your inexpert gut feeling?
Gamolon
Active Member
No you don't. First example. You complain about Mick's column joints. You're making a model that doesn't even have them. How is that holding your models to the same standards?
You've also stated that some characteristics of the model are too complicated to build and will therefore leave them out. How is an inaccurate model going to satisfy your understanding of how the collapse mechanics worked? That's not helping you at all.
Gamolon
Active Member
If we were to put small pieces of Scotch tape on the outside of Mick's columns to join each joint, would that suffice?
Mendel
Senior Member.
See how the WTC facade columns broke at the panel edges? This looks more like Mick's model than yours.
https://www.villagepreservation.org...wers-with-cranes-and-smoke-in-the-foreground/
"Small" meaning scales like 1:4, not 1:1000. And they're not testing structural properties.
Article: A mid - late 1990's original development scale model of a Footwork/Arrows Formula One race car. [...] Car Dimensions: 177 cm/69¾ inches long x 78.5 cm/31 inches wide.
No, that's exactly how one would intuitively construct a column. Ask any toddler or Jenga
player.The reason is that these columns don't need to withstand a lot of lateral force.
That's also why the actual WTC facade panel connections broke when subjected to a large lateral force (see my above photo for proof) while leaving the "vertical strength" of these panels intact.
Mick's model works for what it does, because it only models a single aspect of the destruction, and that's the progressive collapse. Mick doesn't try to model the buckling that took place before that. And that's because the buckling and the progressive collapse do not scale in the same way.
Last edited:
I basically think Mick's columns shouldn't have joints. The columns were assembled so that the joints were staggered, which, as I undertand, should make them basically irrelevant when thinking about the vertical strength of the columns. The were essentially continuous. There should be no particular "weak points" along their length.
As I understand it, there are "dowels" in the middle of the "drum" sections of classical columns.
No, I would want a couple of good piece of dowel inside the joints and wood glue.
Gamolon
Active Member
But columns of the towers snapped at the joints did they not? Why would you remove a characteristic of the real collapse from a model of it? That makes no sense.
Models can be simplifications without being "inaccurate". But you're right that, at the end of the day, it's going to be a judgment call. Was it reasonable to leave some aspect of the actual buildings out?
I think the basic idea here is to make sure that the structure that is destroyed in the collapse is also the structure that provided its standing strength. In Mick's model, the column fails at two points along their length (where the boards meet) but the strength of the boards is otherwise unaffected. That's not a lot of destruction compared to what happened to the WTC.
I said above that I'd want the columns to be connected with a strong dowel joint and glue. I would also like them to be cantilevered at the base, so that they could stand up without lateral support from the floors. This raises an issue that we've discussed before:
Should the columns be self-standing? Or should they be subject to Euler slender-column buckling? The models I'm working on now are trying to live up to he later, which is Mick's concept of the perimeter facade: they would crumple in on themselves if not braced laterally by the floors.
But they must be cantilevered at the ground because, obviously, they will have able to stand up to some specifiable height.
To get back to the dowel joint: this may still be the place where it happens to fail above the critical height. But you don't just want it to fall off because it loses its balance. You want it to self-buckle at that height. So the joint has to be almost as strong as the column itself.
Last edited:
Mendel
Senior Member.
It has been explained to you that the design principle of the WTC's tube-in-tube design requires horizontal connections between the outer and inner tube to ensure its "standing strength". If these connections are removed, the structure is no longer stable. They are absolutely critical to the strength of the structure.
That's exactly what happens with Mick's model.
I'm not sure that empirically true of all the failures along the length of the columns. Some of the columns were bent out of shape. Some of the joints didn't fail.
But, yes, they were probably the weak points in many cases. (The relevantly weak point also depends on the exact impact force, i.e., where the torque pivots, not using that phrase very precisely).
The buildings would not have stood any stronger on 9/11 if the joints had been stronger. Or so I am told. They weren't decisive in the collapses. The columns would just have failed at less "clean" points.
You didn't read the rest of that post, did you?
Mendel
Senior Member.
Collapse initiation was from "not a lot of destruction".
Mendel
Senior Member.
The WTC columns were not able to do that, by design.
You can stand a board on edge without it being cantilevered.
Mendel
Senior Member.
From the aircraft impact, where they had horizontal bracing from the floors.
And tap it hard with a wrench at the top?
Gamolon
Active Member
If the joints weren't decisive in the collapse, why do you care if Mick's model has any connected joints between them in the simulated vertical columns?
Like I say, I don't think that's, empirically, the only reason. A lot of steel was bent lower down in the building. (I'm happy to look for examples if you think it's very decisive. That is, if you will have to revise your view should such evidence be available. But I think you can just grant it, can't you?)
Because the joints are decisive to the way his model collapses.
Gamolon
Active Member
And what way should his model collapse? I though were were trying to replicate the way the towers collapse so you could understand better?
Furthermore, your paper model has no joints or even separate vertical columns so that is going to be decisive in the way your model will collapse.
Mendel
Senior Member.
Again, not something the WTC columns could do.
The horizontal bracing is crucial for both the model and the WTC.
A column could stand a lone to some specifiable (using Euler) height. You showed a picture of some of them doing just that.
Like I say, a single perimeter column would obvioulsy self buckle at a height of 1700'. Almost as obviously, so would a single face. Mick thinks it's just as obvious that a hollow tube of four faces, joined at the corners (where they would now provide each other with lateral bracing) would crumple in on itself for lack of the bracing provided by the floors. That's not as obvious to me, but it's my working hypothesis in my current attempts to model them in paper. If Mick is right, I should be able to build a tower that falls down by itself without the floors (and their loads) and is very stable with them (and their loads) in place.
Note that all this easy to demonstrate with strips of paper of variously lengths clamped at 90 degrees to the ground.
Mendel
Senior Member.
As a wager, it's unfair, because I don't expect you're going to revise your view when you don't find evidence of vertical facade panels that were bent during the progressive collapse phase (and before some other rubble fell on them). And I'm not going to grant something when I have photographic evidence and all you have is incredulity.
Mendel
Senior Member.
With a corner brace, anchored by a heap of rubble. I can do that with 2 boards and tap them with a wrench.
Gamolon
Active Member
I think I'm beginning to see what you're after here.
You want to create a model that doesn't globally collapse so you keep changing the parameters to fit that want instead of creating an accurate model to replicate the tower's construction to see it THAT model globally collapses.
Like some of us have said. You're doing this backwards.
You're basing your design criteria of your model based on the fact that you DON"T want it to globally collapse instead of creating a model that replicates the actual tower's design to see it THAT would globally collapse.
You need to completely rethink this.
Edit: Added the word "don't".
Last edited:
Take a WTC perimeter column to a height of 5 storeys. You could hit it 1 meter from the top with a force strong enough to sever it and you would not knock it over. You would simply set off an oscillation. It would yield to the impact (perhaps breaking off the top) and then bounce back (and forth and back). That's because it is cantilevered deep into the ground.
Gamolon
Active Member
Ok.
Build a 1:100 model of the tube using paper and tape. That would be about 25" wide at each face and about 13' tall.
So it's decisive to you? If structural steel was bent (not just broken at the joints) in the lower portion of the towers, then you're wrong about something important? If not, then I am wrong about something important?
Mendel
Senior Member.
.... and revolutionize architecture if that worked.
Mendel
Senior Member.
not what I wrote
Why not let me do it at 1:1000?
Okay. I don't understand what point you're making then.
Gamolon
Active Member
Why?
Shouldn't matter right? If it stands at 1:1000, it'll stand at 1:100 according to your understanding.
This is literally my initial reaction to Mick's conception of the WTC tubes. Somehow they were structures that couldn't stand on their own but could stand when loaded with at least twice their weight (because some of that load also provided lateral bracing). I'm trying to understand it by building such a structure..... and revolutionize architecture if that worked.
Gamolon
Active Member
That model would be about 15.6" high and the four faces of the façade would be about 2.5" wide.
Again. According to your understanding, both models will act the same.
1:1000 is much easier to build on my desk. So if you're happy with those conditions, I'd be grateful.
Mendel
Senior Member.
The bottom part of the columns was constructed differently, instead of the grid it had these narrow arches. The panel section starts above that, and then the steel dimensions change as you go up the tower. But these panels were not "cantilevered" against each other, they were just joined at some edge points, and braced via the floor joists.
If you hit one of the panels in my photo with a wrecking ball, the joints would sever, and it would break off.
Yeah, like the challenge I specified above?
Mendel
Senior Member.
then re-read what i actually wrote
- Status
Similar threads
