I want to create a high order element in abauqs through abaqus vuel subroutine, with mechanical and electrical DoFs. But I read in documentation that explicit allows only degrees of freedom 1-6, 8 and 11(user defined elements 32.15.1-5). Where 1-6 are mechanical, 11 is thermal and 8 is about pressure (conventions 1.2.2). I understand that through this comments I basically I answer my question. But I am asking it, just in case someone has found a loophole.
Thank you
In the paper below:
http://onlinelibrary.wiley.com/doi/10.1002/nme.4706/full
we've implemented a coupled mechanical / charge-dipole UEL in both Abaqus standard and explicit. So yes, it is possible; have a look - hope it's useful!
Hello
Your question is a common dilemma. While it is true, we have implemented the above MDFEM in both UEL and VUEL. Typically, we hide the electrical charge and dipole in DOF 11-14. However, in the explicit form, we have never managed to obtain stable convergence with DOF 11-14 active, which we have activated in the input file, ignoring manual section 32.15.1 .
So my advice is:
a) to just try activating more degrees than "allowed", the manual in subroutines is not always the most up to date and has discrepancies along it. If it converges, heavy validation and verification is nonetheless required.
b) Otherwise, does your element also need the rotational DOF active? DOF4-6. If not, you can store your electrical DOF in there. Abaqus is oblivious about the physical meaning of the content stored in a DOF, you just have to keep an eye on non-linear integration and stability criterion that the solver applies to different DOF, i.e. incremental addition for rotational values such as DOF 4-6.
Finally, a word of advice/experience for a fellow subroutine programmer. In VUEL, i found that if you have more than one VUEL type, i.e. VUEL1 and VUEL2 in the same model, we never obtained a running program process (obscure windows error codes 193 etc...) unless VUEL1 and VUEL2 had the same number of nodes and same DOF active at those nodes. So we solved it by upgrading the VUEL with the smaller number of nodes, to have the same "shape" and DOF of the larger VUEL. And then writing the smaller VUEL, just assign 0 to all entries that the VUEL does not apply to. The solver is just carrying zeros around, but doesnt stop. And if the larger VUEL is accurately formulated, Stiffness Instabilities etc... , will not occur.
Best of Luck to you, if you do find a workaround, please let us know.
Best Regards,
Andre Wilmes
P.S.Oh, forgot to mention that in our case, we dont require the mechanical rotational DOF 4-6, so we can use them for other purposes, i.e. electrical DOF.
- Badges
- Science topic