UMAT anisotropic elasto-plastic material

[luiz]

Dear everydoy,

I’m trying to simulate a material behaviour that reacts different in tension and compression.

At the moment I am trying to develop the compression behaviour for the material that should be an elasto- plastic one. Additionally it is an anisotropic material and solid 3D elements are used.

I am using the plastic Hill yield criterion for anisotropic materials. Such kind of behaviour can be modeled with Abaqus itself, so that I can compare the stresses and strains with the Abaqus results.

For testing the Umat, I modeled a cubus and in a STATIC general step this cubus the forces are applied via boundary conditions, so that this is displacement controlled. (I attatched the input file)

As it can be seen from the plots (attached jpg), the material starts to yield, stresses are kept constant, but strains, too. Does anyone has some hints or does anybody have such kind of UMAT? I guess I have problems with my Jacobian Matrix. It would be great if somebody could help me.
Thank you very much in advance.
Best Regards
luiz

[Jorgen]

If you are getting results that do not seem right, then the problem is probably not caused by the Jacobian. The Jacobian only influences the convergence, not the actual results.

-Jorgen

[luiz]

Dear Jorgen,

thank you for your answer. Do you have any ideas how to fix this problem? Should I use some hardening? Maybe you know where I can get such a UMAT?
Thank you for your help again!
Regards
luiz

[Jorgen]

Yes, why don't add hardening. Perhaps that will help.

I don't have any UMAT to share
Why don't you try to write it yourself...

-Jorgen

[Aloulou]

Hi Luiz,
your routine is a bit strange, your arbitrary parameter 'a' which reduces the Jacobian automatically when your stress is beyond the limit of elasticity looks artificial to me... It might be good if you have a look at "mathematical theory of plasticity" by Hill

[ar_chatt]

hello eveyone,
I am relatively new with abaqus subroutines. Can anyone please let me know how Isotropic material with different elasto-plastic behavior in tension and in compression can be modelled in ABAQUS. I just need to define different yield strength in tension and in compression. I tried with *NO COMPRESSION elastic behavior but that is not compatable with plastic behavior, so I always get an error which says "*NO TENSION or *NO COMPRESSION options may not be used with Plastic, Creep or Viscoelastic options". Does anyone have any idea how to avoid this error.

Thanks for your kind reply.

[matsgd]

You need to use one of the material models with pressure dependent yield surfaces / flow rules. Examples are the Mohr-Coulomb and Crushable foam models.

Mats

[ar_chatt]

Thank you for your quick reply.

But the problem I am dealing with is a variety of steel which has high yield strength in tension and very low in compression. Thins like cohessive yield strength or friction angle do not seem relevant. Can you please suggest any other option to model this behavior?

The material properties I have are:
Youngs Modulus (E) = 210000
Poisson's ratio = 0.3
Yield stress (tension) = 350
Plastic Strain (tension) = 0
Yield stress (compression) = 10
Plastic Strain (compression) = 0

The geometry is a kind of bracing arrangement where alternatively the tenssion and compression comes on the (truss/beam) member.


Thanks for your kind reply.

[matsgd]

This is steel? With a factor of 35 between tensile and compressive yield stresses? Do you have some more info on it?

Mats

[Frank]

*NO TENSION and *NO COMPRESSION can be combined with ELASTIC exclusively.

This should work:

*ELASTIC
210000,0.3
*PLASTIC
10,-100.0
10,-0.000000001
350,0.000000001
350,100.0

This should make the yield stress constant, coming from a large negative strain to almost zero negative strain, and the like for positive strain.