Problem in setting the Halfrol in dynamic implicit step in abaqus

[mveeram]

Hi,

I have a problem in setting the half step residual tolerance for dynamic implicit step in Abaqus standard. In the manual it says that it can be either set to 0.1 P or P or 10 P depensing upon some conditions , where P is the load applied/reaction forces sustained by the model. But for what ever value i give it the analysis just stops running i.e., in the sense it doesnt abort or give some errors but just stops without any further increments.

Can some one shed some light on my issue.

Thanks in advance,

Regards,
Mveeram

 

[Drej]

First of all, check that HAFTOL isn't the problem. It may be something else causing the analysis to fall over. You can make HAFTOL extremely large (relative to expected model forces) and try going from there. I would, however, carry out a static and modal analysis on the model before progressing on to dynamic analyses to make sure the behaviour is as expected. You should ALWAYS do this anyway.

The manual is poorly covered in this respect (HAFTOL) however. After you are sure that the model is behaving reasonably under static/modal conditions, as a first trial I would check your static analysis (a gravity run should be sufficient initially) using the boundary conditions of your dynamic analysis, and note the total reaction force in the model. Base the HAFTOL on these forces and then reduce the HAFTOL until the analysis results converge (within a practical time frame of course).


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[mveeram]

Thanks for the reply. I already did a static analysis of the model and I got some reasonable results. I am applying a load of 3000N to my model and the reaction force i get from it is the same. I have set a haftol of 3000. So what can i do to improve the accuracy of my solution. Is it either setting a low haftol or big value to it?