Tiebreak contacts in LS-DYNA have undergone significant enhancements in recent versions. A brief overview is provided in the PDF document below. Special thanks to Jim Kennedy for requesting this and also to Dr. Brian Wainscott, LSTC, for sharing some intricate details about TIEBREAK contacts.
Hosting Sponsored by:
Hi Suri,
On page 6 of these notes, there is:
“LS-DYNA uses a small increase in the area of the segment in which it scales the master segment area by an additional 2% in an attempt to collect nodes that may lie near the edges”.
Is there any way to control this 2% to a larger value?
I often encounter a problem when tying objects that were meshed separately, yet their original CAD surfaces were perfectly coincident. For example, imagine two cylinders being tied end-to-end. In the worst case scenario of these being meshed separately with cross sections of a single quad element (that just happen to be misaligned by 45 degrees). In this situation, all of the nodes on both quads will be unable to find a master segment to tie to, and the tie will not be successful. As the parts are meshed more finely, interior nodes will find a segment to tie to, but perimeter nodes are often still left hanging.
How can this best be avoided?
I think the easiest solution would be if the 2% value were editable (or perhaps, a larger value used for non-surrounded elements (or even more sophisticated, only increased on outside ‘naked’ edges)).
Beyond that, I guess that nudging of all exterior nodes on one of the tied surfaces (probably the least stiff surface) to ensure it will ‘hit’ a segment when projected would work, but it would of course be labour-intensive.
Cheers,
Sven.
Hi Sven,
I think the MAXPAR is supposed to just that. Its defaulted to 1.02 which is a 2% increase in the segment size. Larger values may result in increased CPU time or instabilities under rotation. Mesh refinement is probably a safer approach.
Suri Bala
Hi Suri,
in one reply of yours, you attached a PDF file. At the end of that file some examples are mentioned. Where could one download them?
Thanks.
J
Hi J,
I will try to upload these in a few days.
Thanks,
Suri Bala
Hello Suri,
which option of tiebreak contact can i use to consider only the failure in tension not in compression, because i see that tiebreak options always consider compression like tension.
Thanks,
Tammam
Tammam,
TIEBREAK only considers failure in Tension and Shear.
Compressive loads are always transmitted unless a death time is provided that is less than the termination time of the problem.
Suri Bala
Hello Suri,
TIEBREAK considers the failure only in tension, what about the shear failure with cohesion when accompanied with compression. Is the TIEBREAK contact follow Mohr-Coloumb failure criteria with cohesion, or only consider the friction?
Thanks,
Tammam
Tammam,
TIEBREAK considers failure in both tension and shear. I will check on the cohesion part.
Suri