Automatic Tiebreak – Some Quick Notes

I posted a detailed document on the TIEBREAK contacts. Here are a few things that I noticed after the I wrote the article.

1. Scale down the penalty stiffness down to 0.1 if you see instability.
2. It is important to have the two surface to be “just in contact” and not “too far” or “penetrating”.
3. Do not have the parts in TIEBREAK in any other contact if possible to avoid duplicate contact treatment.
4. If you still see instability, remove the TIEBREAK and see if the regular a3 contact works.
5. Also, just for verification, a quick Eigenvalue extraction will ensure the right nodes are tied to right master segments.

  • Hi Suri,

    I’m modeling delamination in CFRP composites using the tie-break contact in LS-DYNA (automatic one way tiebreak with option=8).
    To have a feel about the material properties the in-plane modulus is around E=50,000 MPa, and for the delamination model, failure normal stress is around 150 MPa and critical crack opening is around 0.015mm!

    So here is my problem:

    – tie-break is based on penalty method, so “tied” surfaces before the initiation of damage, have some relative displacement!
    – In the traction-opening law used, the code “assumes” that before initiation of damage there is NO relative displacement. (refer to page 31 of the PDF file you provided)
    -at the time that failure criteria is met, code starts looking at the damage curve to calculate the traction between surfaces, now due to the relative displacement before initiation of damage, transferred traction between the layers DROPS suddenly.
    – In the other words, to calculate the post failure behavior, the code assumes that before failure, there is absolutely no relative displacement, whereas due to penalty method, there is some!, so it decreases the transmitted stress because of pre-failure relative displacement!!
    – This problem becomes very significant when CCRIT is comparable with (PeakStress/ContactStiffness)!!!, which is the case for the material that I’m working on.
    – This sudden drop in the transferred traction causes oscillation, noise and more importantly, inconsistency of absorbed energy in crack opening process!

    The solution to this problem is:
    – Increasing the penalty stiffness, For example, in one case, I had to increase the penalty stiffness scale (SLSFAC) to 5 which is 50 times more than the default value!!!
    – but as you mentioned in your posted notes, there are some other considerations regarding the contact stiffness too, so in most of the cases you cannot increase the contact stiffness due to its consequences!

    So what’s your recommendation to solve this problem? and is it hard for LS-DYNA developers to fix this problem?


Leave a Reply

Your email address will not be published. Required fields are marked *