Can UMAT be used along with *Soils???

[rm447]

Hi everyone,

I am trying to model fluid flow through a porous material. I have modeled the solid matrix in umat as a type of hyper elastic material, and would like to use the soils consolidation to do the analysis, however, even though according to the following thread it seems doable, it seems like the solver does not communicate with my umat when I use *Soils.

http://www.polymerfem.com/forums/sho...ght=umat+soils

I have examined a couple of test cases:

-I tested the input file using Abaqus built-in elastic material for the solid matrix, solving with *Soils, and it works fine and I get what I expected.

-I have checked the umat using the same input file only with a simple static step, and replacing the elastic solid matrix with the user defined material. That works fine too.

I put in some write statements and output the deformation gradient, elastic stress and strain values, and the stresses are zero everywhere, and strain values are wrong (I have prescribed the displacement, thus strain is known). The strain does not seem to be the actual strain because 1) the deformation gradient and Cauchy-Green tensors remain identity all throughout, 2) stresses remain zero.

This makes me wonder if it is impossible after all to use a user-defined material in the soils analysis, even though it does not make much sense. I could not find anything in the documentation regarding this. In the *Soils documentation it looks like you can only define the material properties using *elastic, *bulk modulus, etc. which would imply that umat can not be used? Not sure if lack of that information means this is not feasible.

Please let me know if you have any experience with this.
Thanks
RM

[Jorgen]

That is a good question, and I agree that the Abaqus manuals are not quite clear about that.

I am not 100% sure about it either, but I suspect that a UMAT subroutine does not work with the *Soils step. I think you can use a CREEP subroutine with *Soils, however. Have you tried that?

[rm447]

Dear Dr. Bergstrom,
Thanks for your response. No, I have not tried *creep, but I don't see why using *creep is going to help me since I need to define the constitutive equations of the hyperelastic solid matrix using a suitable strain energy function. Is there a reason why you suggested that, besides the fact that it can work with *soils?
I just looked at the UMAT documentation again and it says UMAT "can be used with any procedure that includes mechanical behavior", which I think should include *soils as well, right? Because on the other hand the documentation for *creep specifies only a few procedures: "if used to define a material behavior, the subroutine can be used in the coupled-temperature displacement, soils, and quasi-static procedures.".
I am confused

[Jorgen]

I had no particular reason expect as a simple "oh by the way" statement.

To find out for sure, I would perform a simple test. Which it sounds like you have already done. It it does not work, then UMAT + *soild probably does not work

-Jorgen