'What about building 7'?

[LSSBB]

No it isn't. That's just so wrong. The condition is satisfied when the residual is less than the criterion. I even posted that earlier, and you clearly didn't comprehend it at all.


Yes I agree. In another algorithm for another type of model measuring something else the condition might be something different. And in another life your post may be relevant to the model we are discussing.


Wouldn't they first try graduating the increment of the load? I presume you would be talking about the first iteration here.

I am talking about whatever iteration, until the criteria is met. And I am not wrong. Your quote was "The difference between external and internal loads is called the residual", equivalently (External Force) - (Internal Force) = residual. " The goal is to iterate until the residual becomes acceptably small", i.e. approaches 0. How exactly the algorithm does this depends on the algorithm.

Here is an algorithm that uses an ANSYS module as a subroutine in a loop and uses element removal:

http://www.hindawi.com/journals/ame/2013/413463/

ANSYS itself has multiple algorithms:

ANSYS can carry out advanced engineering analyses quickly, safely and practically by its variety of contact algorithms, time based loading features and nonlinear material models.

http://www.figes.com.tr/english/ansys/ansys.php

If you want to know how the ANSYS algorithm is constructed, contact ANSYS. If it's not proprietary, you might get something useful.

 

[MileHighMadness]

You have a lot of questions about ANSYS.
First of all, do you agree with me that "load balance" is a good a 2 word summation of optimising convergence in such an FEA ?

Wow...avoid my questions, and ask a new question.

But to answer your question...no

 

[AZCat]

You have a lot of questions about ANSYS.
First of all, do you agree with me that "load balance" is a good a 2 word summation of optimising convergence in such an FEA ?

I don't think LSSBB's original question limited you to a 2 word summation in response. Why the additional constraint?

 

[gerrycan]

Explains why the engineering world pays so much attention.

Why do you think that is?

I merely agreed with the statement that you "have no idea what this means".
Were you talking generally?

 

[JayUtah]

You have a lot of questions about ANSYS.
First of all, do you agree with me that "load balance" is a good a 2 word summation of optimising convergence in such an FEA ?

No, they seem to be questions about your understanding of and facility with ANSYS. They seemed to be formulated to elicit answers that distinguish someone who actually knows from someone who's largely bluffing. Consequently I for one would like to see you answer them rather than evade and deflect.

 

[gerrycan]

No, they seem to be questions about your understanding of and facility with ANSYS. They seemed to be formulated to elicit answers that distinguish someone who actually knows from someone who's largely bluffing. Consequently I for one would like to see you answer them rather than evade and deflect.

Ahhh, Jay the FEA expert. Did you like my description "load balance" ?

 

[JayUtah]

Ahhh, Jay the FEA expert. Did you like my description "load balance" ?

Would you please answer MileHighMadness's questions without further deflection and evasion?

 

[DGM]

I merely agreed with the statement that you "have no idea what this means".
Were you talking generally?

I was referring to your statement:

Strange. You exhibit all the symptoms of a sufferer.

I'm sure you know this. Pave on. You do know people can look back and see this.

 

[LSSBB]

Here is a good breakdown of how the ANSYS FEA is worked with, and some details of how it works

http://www.ansys.com/staticassets/A...elibrary/confpaper/2004-Int-ANSYS-Conf-24.PDF

 

[Ziggi]

Correction: the time frame for column 79 displacing is 3.5 to 4 hours; the time frame for the girder walk-off is 3.7 to 4 hours. The latter is not mentioned as a time frame for the column movement. Based on that alone, the probabilities are reversed with respect to your belief.

This is not about probabilities, this is about what NIST actually concludes in the report having itself assessed its own data and "probabilities," and that conclusion was to NOT specify the column displacement as a factor in the walk off event.

Any rational person would infer that the conclusion is: in order to know for sure, ask the source (NIST).

I already asked the source and the report´s conclusion is to not specify the column displacement as a factor in the walk off. "Making sure" is something you should have done before you claimed as a FACT that this column displacement was a factor in the walk off.

I "conjured up", as you say, that explanation when looking for ways in which ANSYS could have showed walk-off as NIST reports it did. Again, assuming innocence as a premise, rather than starting with the conclusion that the report is flawed and walking backwards trying to find mistakes at all costs, which is the truthers' approach. why?

The flaw in the report is obvious: it does not give evidence for the possiblity of the alleged 6.25 inch displacement. And you obviously noted this flaw yourself because you went out "looking for ways in which ANSYS could have showed walk off as NIST reports it did."

But instead of just accepting the problem you made up a remedy to the flaw and passed it off as NIST´s story. Your leverage story was not a way in which the ANSYS "could have" shown walk off because it is based on assumptions that contradict NIST´s data, namely that the beams were still connected to the girder so that this phenomenon you claimed was in fact impossible according to NIST´s data.

Who, and why?

You and others that tried to pass off the false column displacement and the false leverage story in a matter of fact manner as NIST´s conclusions. Some people here fabricated these strawmen and passed them off as NIST´s work, and then you ridiculed gerrycan for not including them in his assessment of the 6.25 inch displacement.

Without these made up fantasies, you are again stuck with what NIST actually says caused the walk off: the expansion of those floor beams.

 

[MileHighMadness]

So Ziggi, if the girder didn't walk off, how did the connection fail? We know the connection had to fail, because the building collapsed?

 

[MileHighMadness]

Delete...double post

 

[JayUtah]

The flaw in the report is obvious: it does not give evidence for the possiblity of the alleged 6.25 inch displacement.

I think you mean it doesn't give the kind of presentation you were expecting. I've asked you several times what the presentation should have looked like instead, what sorts of things should have been considered, and what the values were for the applicable unknowns. The problem with your line of reasoning is that the standards you intend to impose are not self-evident and seem to come from nowhere except your imagination. As the evaluator, you have the burden to show that your analysis is based on reasonable standards.

 

[gerrycan]

Here is a good breakdown of how the ANSYS FEA is worked with, and some details of how it works

http://www.ansys.com/staticassets/A...elibrary/confpaper/2004-Int-ANSYS-Conf-24.PDF

And look what is says.............
"In ANSYS the Newton-Raphson method is implemented to solve a nonlinear problem iteratively"

Now what was the question you asked that Jay was so keen for me to answer. Something about ANSYS being linear or non-linear?
Can't think why I didn't answer that one, I mean there were only 2 choices. I could have taken a 50/50 shot.

 

[JayUtah]

Now what was the question you asked that Jay was so keen for me to answer. Something about ANSYS being linear or non-linear?

Actually there were several questions MileHighMadness asked. That was one of them.

Can't think why I didn't answer that one, I mean there were only 2 choices. I could have taken a 50/50 shot.

I assume it's because you had to Google it first, after it became clear that evasion and deflection wasn't going to work.

 

[gerrycan]

Would you please answer MileHighMadness's questions without further deflection and evasion?

I already did, he just wasn't paying attention, and neither are you.
You should be lapping this stuff up Jay, after all, you are a self proclaimed expert in this stuff, claiming that others on here are "lay" as opposed to you being the consummate pro engineer.
Time to speak up Jay, no more hiding.

 

[LSSBB]

And look what is says.............
"In ANSYS the Newton-Raphson method is implemented to solve a nonlinear problem iteratively"

Now what was the question you asked that Jay was so keen for me to answer. Something about ANSYS being linear or non-linear?
Can't think why I didn't answer that one, I mean there were only 2 choices. I could have taken a 50/50 shot.

Point being?

 

[JayUtah]

I already did, he just wasn't paying attention, and neither are you.

His post:

So gerrycan, is the ANSYS analysis linear or non-linear? How was the structural elastic modulus modified for changing steel temperature? How are you differentiating between critical and non-critical elements? Should these connections be modeled as shear or full moment connections?

 

[JayUtah]

You should be lapping this stuff up Jay...

How quickly you personalize the debate.

 

[beachnut]

And look what is says.............
"In ANSYS the Newton-Raphson method is implemented to solve a nonlinear problem iteratively"

Now what was the question you asked that Jay was so keen for me to answer. Something about ANSYS being linear or non-linear?
Can't think why I didn't answer that one, I mean there were only 2 choices. I could have taken a 50/50 shot.

You can google, up ANSYS, and that is the limit of your engineering skills.

Where is your evidence to support your CD theory? Was it silent explosives in your CD fantasy, or magic thermite which does not leave evidence on steel?