Hello everyone!
My goal is to model large plastic deformations of polycrystalline metals using crystal plasticity and ALE adaptive meshing. In order to do that I want to use VUMAT subroutine in Abaqus/Explicit, since as far as I understand Abaqus/Standard is not good for large deformations and it doesn't support ALE adaptive meshing for that purpose.
My concern is how to apply different material properties in different integration points or even elements. In UMAT it is straightforward, but in VUMAT it is not. I have tried setting different values of material properties using ghost shell function which I have found on the Internet and the simple VUMAT of isotropic elasticity. It works fine when I use one element, but when I use more than one, it says that there is zero or negative dilatational modulus. I have even checked what happens, when I apply the same material properties in every material point (e.g. my props was 2.1e11, 2.1e11, 2.1e11,..., .3, .3, .3, ...), but it still gives the same message.
Do you know what could happen? Or maybe you know better ways for dealing with different material properties in differnet material points in VUMAT? It is important for me, that I can use ALE adaptive meshing, so I think it is not possible to just define different materials inside my model domains.
Best regards,
Karol Frydrych
You can use the "call material" option when you need it any time or step
Maybe it is the following problem. Besides the difference in the indizes between UMAT and VUMAT the main difference between an UMAT and a VUMAT is the processing. For an UMAT each integration point is called from Abaqus and the material law is calculated.
For a VUMAT you have this call inside the FORTRAN routine. You have to write all your user material routine within a loop with length "nblock" (cf. 1.2.17 VUMAT). That is why one element works. You call your routine and send one result at integration point 1 back. If you have two points and no loop, the second material point get a zero for the tangential stiffness matrix and you will get the error you described.
I am agreed with Christian. you need to check the VUMAT code where you have used WHİLE and FOR loops.
Thank you, Christian and Shahram, I made simple mistake in my loop. Now I corrected it and it works fine.
how to apply different material properties in different integration points or even elements in umat? it is very easy? material properties are saved matrix by matrix as a txt file (because it is anisotropic property). How to do that assigning the different matrix to the cooresponding points? thank you
- Badges
- Science topic