[Abaqus] Why does Cutback not appear even when it failed an Attempt in one Increment? 1

[Bingtong]

Hello everyone:
I'm doing a simulation for a Peel-Off-Test with VCCT in Abaqus.
Now I've met a problem:
Normally, the increment reduces itself if an Attempt is failed and try again, like this:

(It looks a littel strange as well because normally it will be reduced with factor 0.25 everytime)
But now it doesn't reduce anymore, like this:

As for the *Control I have only changed the I_A value from 5 to 25, in order to allow more Attempts in one increment.
Other Parameters are with Default values.
** CONTROLS
**
*Controls, reset
*Controls, parameters=time incrementation
, , , , , , , 25, , ,
**
** OUTPUT REQUESTS
... ...

So why doesn't the increment reduce itself, and what stranger is, why sometimes it would at last SUCCEED with the same increment, which had been tried and failed again and again?
Could you please help me, ladies and gentlemen?

 

[Mustaine3]

That's really strange. Did you had a look at the .msg file? Maybe there is an info when the cutback happens. Also check the beginning of the file, where all the used settings are printed out.

 

[FEA way]

Job Diagnostics tool available in the Visualization module will let you examine convergence progress of the solution. There you can find all the necessary details and create plots of residuals. It should give you an insight on what actually happens during the analysis.

 

[Bingtong]

Hello, Mustaine3:
Thank you for your advice. I have found a clue in the .msg file.
This "Attempt once more without cutback" might be caused of no convergence problem but that the programm figured out that more Nodes than estimated would be debonded, so it will try another time (one more Attempt) to calculate. That is to say, by a VCCT DEBONDING test it is totally normal to happen.
But the reason why it can figure out that, which results show that more nodes would be debonded, I couldn't understand jet.
After all, I have to say thanks to you for guiding me to read the details in .msg file. Thanks, freind!

Yours,
Bingtong