The Twin Towers at free fall speed

[Furcifer]

Why all the resistance to a real-life test of the FEA models?

Isn't it reasonable to question whether the simulations correctly modelled their tests? What better way to verify them than to test for repeatable failure modes in a live test?

Why is this such an unreasonable standard?

.

The "real life" test of a model wouldn't be as detailed as a FEA. It would never be as comprehensive. The complexity involved in creating a model that even approaches that of the SAP model would be outrageously difficult, even if you managed to scale it properly. I can't, for the life of me even begin to think how you could scale fire??? My rough guess is that a 1/10th scale model might approach the requirements. Maybe. Ask NB, but I'm guessing an 11 story tall replica model would cost around $10 Million just to build.

Even if you built it, the slightest issue while running the test would render it invalid. That's the problem with running a simulation outdoors.

Finally, even if it did fail people will continue to call to question the building methods, the builder, the materials, the conditions, the scale, the lack of thermite, the lack of explosions, etc.

This is why it is an unreasonable standard. It just doesn't make sense. To me! And I'm willing to entertain your ideas. Most people won't. Can you imagine what would happen if the general public found out the Obama administration set aside $20 Million or so to burn a model???

Give your head a shake.

 

[A W Smith]

They do those tests on girders before they build the building. That's how they know it will stay up.

Exactly, There is no need to construct an entire test building to verify the proposed building will perform as required.

 

[JimBenArm]

How do you scale gravity?

With a sky hook, of course!

 

[jaydeehess]

A test that replicated the results of the FEA models would suffice. I'm not suggesting building the entire structure in real life. Just see if the finite structures react the way the simulation predicts.

The method of using an FEA has been demonstrated to be accurate in the design of structures for quite some time now, as explained by NB. You seem to wish to move backwards. Are you usually a Luddite?

That was the whole point of NIST performing separate simulations for Global (i.e. the whole kit and kaboodle) failure, and Finite Element Analysis (i.e. individual members/joints) failures.
.

NIST tested a few things just to confirm the behaviour of those elements. this was done in cases where sufficient previous testing or fire history did not give them a good base. You seem to expect such testing on every element of the structure. You claim you do not want an entire physical model built but you want all elements tested?

 

[aggle-rithm]

Why all the resistance to a real-life test of the FEA models?

I have no objections. Please, proceed with your real-life tests. Let us know how it comes out.

 

[jaydeehess]

Originally Posted by jay howard
Again, there's an unusual amount of protestation about a call for a real-life test. Some people protest that "It's too expensive!"
Others claim "It's all been accounted for by other simulations..."
Some say "It's too complicated to run a real-world test on."
While others say "All the tests have already been run."

Get your excuses straight.

Those are all valid and correct reasons. Which logical fallacy is this? False dilemma?

Seems right.

It is more expense than can be justified. That is to say there is little or nothing that can be shown to be gained by this expenditure.

Other simulations came up with the same result(aside from some details) and thus the NIST sim conclusions are borne out

It would be horrendously complicated to attempt to devise tests that mimic evey failure mode possible in order to determine whether or not the failure modes depicted in the sims were correct.

Physical testing and steel structure fire history do indeed supply a vast amount of information on which to base an analysis and simulation.

The only dilemma then is why JH continues to protest that all elements should be 'real world' tested, not why we protest them being done.


Given that I am a Canadian I would not be seeing my tax dollars mis-spent however, every million dollars that the USA mis-spends makes life harder for the rest of us, especially on the country with which the USA does the vast majority of its trade.

 

[jaydeehess]

I have no objections. Please, proceed with your real-life tests. Let us know how it comes out.

As I keep pointing out, engineers, architects, lawyers and pilots all make pretty good money so all in those professions who belong to organizations that would be interested in such testing could throw in $1000 (what's that, 1% to 2% of their pre-tax earnings even on the low side of the pay rate for those professions) and they would have quite a good kitty with which to get started not to mention that if the engineers threw in some expertise gratus it would cut costs.

 

[Newtons Bit]

The "real life" test of a model wouldn't be as detailed as a FEA. It would never be as comprehensive. The complexity involved in creating a model that even approaches that of the SAP model would be outrageously difficult, even if you managed to scale it properly. I can't, for the life of me even begin to think how you could scale fire??? My rough guess is that a 1/10th scale model might approach the requirements. Maybe. Ask NB, but I'm guessing an 11 story tall replica model would cost around $10 Million just to build.

A 1/10th scale model can't be done, at least not accurately. For example, the floor system is both the floor trusses and the light-weight concrete deck. The two require each other for both stability and strength. It's not possible to make concrete 1/4" thick.

I don't even want to think about the fire movement aspect.

You could do a full-scale model. However it'll run you ~$10 million per floor just for the super-structure and the tenant improvement (interior partitions, cubicles, flammables, etc) will probably cost another couple mil per floor.

 

[jaydeehess]

yep, I would wonder how you convince fire to scale down properly for a 1/10 model.

 

[jaydeehess]

Now JH isn't, he says, calling for a full scale model but frankly there is no getting around it to satify the TM. Even if you build only the top 20 stories on a mount that sims the sway of a structure wit another 90 floors below and carry out this testing there will be many who cry foul.

 

[tsig]

Now JH isn't, he says, calling for a full scale model but frankly there is no getting around it to satify the TM. Even if you build only the top 20 stories on a mount that sims the sway of a structure wit another 90 floors below and carry out this testing there will be many who cry foul.

Even if you built a full scale model and it worked it wouldn't prove anything because it would not be identical to the towers unless you had people in it./TM mode off

 

[R.Mackey]

Good grief.

Astute observers will note that the Truther claim has silently morphed from "NIST isn't falsifiable" to "NIST isn't testable", then took a huge leap to "NIST wasn't tested", and now has hit bottom at "NIST wasn't tested to my personal satisfaction" whatever that might be. Truther arguments almost always devolve to personal incredulity.

As a reminder, I tackled the problem of scaling in Part III of my Hardfire lecture, here. It's not a trivial undertaking at all. As for testing in full scale, if we wish to also test the aircraft impact, we will have to build the structure in controlled airspace -- so multiply cost by two -- and using the FSTCID test as a baseline, the cost to conduct the flight and crash is around $100M, more if the aircraft must be purchased. Far too much money to spend merely to indulge paranoid fantasies.

Despite the difficulties of testing in subscale, some of it has been done, including the fire and temperature curves and the collapse initiation of WTC 1. Also supporting our theory. Not that it will matter for our argumentative friends.

Finally, an erratum in my previous post -- Dr. Quintiere is at University of Maryland, not Northwestern. This is also a fully falsifiable and testable proposition.

 

[ElMondoHummus]

Again, I don't think it's controversial to say that in a complex system like the crash site of the towers, there can be all kinds of factors that the model designers may be leaving out or over-or underestimating. That's why a particular principle should be tested, i.e.: "steel girders with k static force from x,y directions and t temp over m time will result in these failure modes f and j."

Newton's Bit, Ryan, Architect, or anyone else: Is the above not a description of what the Cardington tests were?

@jay howard: Why do you presume that some "particular principle" hasn't already been tested? The existence of fire codes and standards for steel construction demonstrates that there is already a body of knowledge in existence of how steel assemblies behave under both stress and fire conditions. That body of knowledge will be recorded in various works easily available to researchers. And indeed, that is the case here; NIST cites the works it relies on in their report. For example, directly relating to your example "steel girders... will result in these failure modes f and j": There are several works cited in NIST's report that speaks specifically to both steel's behavior under load and temperature. I opened up NCSTAR 1-3E (the subreport that dealt with the mechanical analysis of the WTC steel), went right to the end, and found those citations. My point, of course, is that such principles are already established and documented. Asking for them to be reaffirmed is arguing from a position of ignorance about what knowledge already exists.

 

[TruthersLie]

I'm saying someone could build a real life test corroborating the computer models IF we knew the specific parameters to apply.

Again, it should be something like "steel trusses will exhibit f failure modes at exposure to t temp for m minutes." Undoubtedly it will be more detailed, but the more detailed, the more precisely a test can be constructed to verify it.


.

By all means... Feel free to build one then and show us that NIST is wrong. Don't forget all the complications of scale.

When you finish you will have earned a Phd in structural engineering. I eagerly await your thesis and complete model which shows that NIST is wrong. YOu might want to get cracking on it since it is only the most important engineering story of the century. Or is that just too much work?

 

[TruthersLie]

Oh jesus, stop whining. I'm saying if the computer simulations demonstrate failure of certain connections at certain temps over a specified time, then that info should be translatable to a real-life test.

Now, it should probably be scaled to a reasonable size, but the results should be repeatable in real life, don't you agree?

.

By all means... feel free to scale it yourself. It should be rather easy to do.

Don't forget how scaling affects everything else. I for one eagerly await your model of the most important engineering fact of the century.

 

[TruthersLie]

If the computer models are already done, then all that needs to be done is replicating the conditions (scaled appropriately) in real life. Why is that so unreasonable?

Then feel free to do so. I know Ryan Mackey explained in great depth about how scaling works... so keep that in mind and build a scale model which shows that NIST is wrong.

Of course you might want to make sure you get
1. permits to create a building that you want to then destroy
2. get the funds to create said building
3. let the FAA know that you are going to crash a jumbo jet into it at 500 mph
4. get said funds to do so.

Feel free. Get cracking.

The FEA models are where this data would come from. The FEA's showed failures of certain joints and complexes, so it shouldn't be hard to build a live model of those conditions and see if it in fact fails as the simulation predicts. If it doesn't, then there's reason to believe the simulations did NOT accurately recreate the circumstances in the towers. If it does fail as predicted, then the theory is corroborated.

argument from incredulity and ignorance noted. Again, please provide JUST ONE peer reviewed engineering journal from ANYWHERE IN THE WORLD which says and shows that NIST is wrong. I eagerly await it.

I'm not talking about increasing the complexity of the simulations: I'm talking about recreating an FEA in real life and testing its veracity. In light of the lack of corroborative forensic evidence, wouldn't you want to corroborate NIST's theory?

Not really. They have the experience and education to make their claims, and since there claims have been out there for well over a year for the last one with NO uproar or public disagreements, I'll accept the scientific consensus.

Since you seem to have the problem, then feel free to prove that they are wrong.

 

[TruthersLie]

Why all the resistance to a real-life test of the FEA models?

Isn't it reasonable to question whether the simulations correctly modelled their tests? What better way to verify them than to test for repeatable failure modes in a live test?

Why is this such an unreasonable standard?

Your lack of knowledge of scaling, building codes and basic general ignorance have been shown to everyone.

Again and again, if you think they are wrong, then feel free to create your own scale models and show us that NIST is wrong.

 

[Furcifer]

Not a single person says "sure, that's not a bad idea. Why not just take the FEA models and see if a real test looks like the simulation."

Hey, I was pushing hard for the 1/10th scale model but NB says you can't pour concrete 1/4" thick, which is true. A sand and cement mixture just won't behave the same way, plus the air bubbles in the cement don't know what scale means. I'm sure you will agree this is very important. We are demanding accuracy here.

No one is lying here. Not a single person. (You'll note that people will correct one another if they speak out of turn or misrepresent the facts) I think we've all entertained your objections rather "skeptically". Despite your appeal you've been shot down on all accounts. Your logic is faulty, your accusations are unfounded and your methods are antiquated.

The big step now is coming to terms with the reality we've presented. I now refer you back to an earlier post where you take this info and go to a college or university and independently verify what's been said here.

Go now, you don't want to be accused of pseudoscience now do you?

 

[sylvan8798]

Again, there's an unusual amount of protestation about a call for a real-life test. Some people protest that "It's too expensive!"
Others claim "It's all been accounted for by other simulations..."
Some say "It's too complicated to run a real-world test on."
While others say "All the tests have already been run."

Get your excuses straight.

Not a single person says "sure, that's not a bad idea. Why not just take the FEA models and see if a real test looks like the simulation."

Seriously, what possible harm would come to anyone to verify the simulations with a real world test?

You seem to be confused. The idea is that we do tests and then come up with models (equations) which simulate the behavior, then we try those out on various cases. If we are confident that they work, then we don't have to keep testing. We can design using the models.

The FEA comes FROM the knowledge engineers and scientists and mathematicians already have about the material behavior, not the other way around. They didn't just pull a bunch of parameters and criteria and limits out of their butts and stuff them into a computer program.

 

[dafydd]

You do not understand the real world event is a full up test with not only a model, but the actual building. You can't understand 911 so you want a real life test; oops, u missed it twice, 2 times for Balsamo, on 911.

You fail to understand the goal of NIST and seem to prefer full blown nut case delusions made up by idiots in 911 truth, a failed movement; 8 years of failure.

What do you have on the free fall nonsense as you drift off topic into displaying a complete lack of knowledge in engineering, models, and more?

What are you qualifications in this field Jay?