Non-Linear, number of iterations, export of results

[userP8ZHE]

Hello everyone.

got a question, please advise if you can.

doing non-linear (bi-linear) analysis. few different cases for vector of load application. same model, same meshing, same all environment, just changing vector.

number if iterations in "non-linear set up" is same as well.

one test (one angle) gives to me 10 increments.

another angle gives to me 230 increments.

another - 20 increments.

from my understanding it is about approximation and fitting in tolerances of "measurement results". is that correct? kinda "optimization".

if I am right, would that be possible to guide me where to read about this please?

then,

I got few data sets/measurements with "230 increments". great.

by now, I am just writing manually into my measurements XLS table for further calcs. if that is 10 increments - well, not perfect but still doable. but to right down 230 increments with 6-8 digits after comma is quite painful.

would that be possible to export results somehow to Excel?

can someone to guide on this or give a ref where to read about this?

and another then,

I see that increments counter is increasing but "load scale factor" is not always increasing, in set of "230 increments" it is moving up and down. consistent increase in begin of results and in the end of results. middle section of results - load scale factor is jumping.

what that could mean? I will sort them anyway by load scale factor, but, for my understanding, what does it mean? again, straight answer of ref to study will be very much appreciated.

thanks in advance.

[John_Holtz · Accepted solution]

Hi @userP8ZHE 

It would be helpful to see what you have entered for the Nonlinear Setup. I suspect you have it setup like figure 1 in this article except you have set Intermediate Output to "All" instead of "On". See Understanding convergence in a Nastran nonlinear analysis. Convergence is about the results following the changing stiffness of the model. 

You can output results using the third-party app FNO Reader.

You may need to clarify what you mean by the load scale factor jumping around in the middle of the analysis. For example, where are you seeing this? (Maybe include a screenshot showing the value.) What is the load in the analysis? What did you enter in the Nonlinear Setup?

John

[userP8ZHE]

thanks for answer and links, I should read these...

yes, in set-up "all", not "on". 

in past, it was working ok.

for "jumping":

I did no save that calc, sorry, but I still will try to explain

increments "INCR" are increasing, it is counter.  ok.

LOAD is "load scale factor", should be increasing too. but, what I had, was for ex, INCR = 10, LOAD = 0.25. INCR = 11, LOAD = 0.12. INCR = 12, LOAD = 0.44. INCR = 13, LOAD = 0.3 and so on, so on...

[userP8ZHE]

ok, I found some samples for "jumping"..