Notifications
Clear all

UANISOHYPER_INV and SIGINI working together

5 Posts
3 Users
0 Likes
837 Views
sdn0503
Posts: 3
Topic starter
(@sdn0503)
New Member
Joined: 10 years ago

Greetings,

I am a new user and I would firstly like to say hello, and thanks in advance, to everyone here! I am working in research in the biomechanics field and I am using a few subroutines in a FEA simulation with ABAQUS. However, I have experienced some problems, which are described below.

I am trying to implement a transverse isotropic hyperelastic material with an UANISOHYPER_INV and I also need to declare initial stresses in my simulation (I do it with a SIGINI subroutine). Although both subroutines work alone, when coupled together, they seem not to work. I experience no stress at all, even during the very first moment of the simulation. Does anyone else experience this issue? May it be a bug?

I created another simulation using a simple hyperelastic law (Neo-Hookean) with a UHYPER, and it works perfectly with the SIGINI.

I am currently writing a UMAT. I have finished the Neo-Hookean version but it is quite unclear to me if it is possible to implement initial stresses in the UMAT, without using the SIGINI. I have tried using UMAT and SIGINI, and again although both work alone, they do not when used together. However, I get the correct stresses in the first moment of the simulation, but then they disappear during the first step although the movement is completely restricted (I think it is because of the stress definition in the UMAT). Does anyone have some information on this?

Thanks for everything,

Levrero.

PS: Happy new year!

Topic Tags
4 Replies
1 Reply
mapenghui
(@mapenghui)
Joined: 5 months ago

New Member
Posts: 1
Posted by: @sdn0503

Greetings,

I am a new user and I would firstly like to say hello, and thanks in advance, to everyone here! I am working in research in the biomechanics field and I am using a few subroutines in a FEA simulation with ABAQUS. However, I have experienced some problems, which are described below.

I am trying to implement a transverse isotropic hyperelastic material with an UANISOHYPER_INV and I also need to declare initial stresses in my simulation (I do it with a SIGINI subroutine). Although both subroutines work alone, when coupled together, they seem not to work. I experience no stress at all, even during the very first moment of the simulation. Does anyone else experience this issue? May it be a bug?

I created another simulation using a simple hyperelastic law (Neo-Hookean) with a UHYPER, and it works perfectly with the SIGINI.

I am currently writing a UMAT. I have finished the Neo-Hookean version but it is quite unclear to me if it is possible to implement initial stresses in the UMAT, without using the SIGINI. I have tried using UMAT and SIGINI, and again although both work alone, they do not when used together. However, I get the correct stresses in the first moment of the simulation, but then they disappear during the first step although the movement is completely restricted (I think it is because of the stress definition in the UMAT). Does anyone have some information on this?

Thanks for everything,

Levrero.

PS: Happy new year!

Hello,Excuse me,I want to know, have you solved the problem that UMAT and signi cannot be used at the same time.

Reply
kiranckst
Posts: 93
(@kiranckst)
Trusted Member
Joined: 13 years ago

I havent used SIGINI much but according to the manual stresses from SIGINI should be available in the UMAT in the stress array at the beginning of the analysis. If you dont propagate this stress yourself in the UMAT it will naturally disappear. Also the manual says it is not possible to define initial stresses for fabric materials, which might explain SIGINI not working with UANISOHYPER_INV.

Topic Tags
4 Replies
sdn0503
Posts: 3
Topic starter
(@sdn0503)
New Member
Joined: 10 years ago

Thanks for the quick answer!

So I guess I should forget about using SIGINI and UANISOHYPER together. However, I am not using a fabric material since you need to specifically define it through a *FABRIC line in the input file and I have not done it. My material is an anisotropic hyperelastic with a main direction.

As you mentioned, in my analysis with UMAT and SIGINI I obtain the initial stresses in the first increment of the simulation but they disappear, most likely because I did not propagate the stresses. Does anyone know how can I propagate these initial stresses in the UMAT, using the initial stress values from the SIGINI?

Reply
kiranckst
Posts: 93
(@kiranckst)
Trusted Member
Joined: 13 years ago

For the UANISOHYPER it was only a guess that it didnt work due to the limitation on initial stresses in fabrics (I originally thought the manual said fiber materials actually so had to edit!). You could try looking at the UA term in UANISOHYPER, maybe the stress is somehow passed in through that.

For the UMAT, if your material is elastic maybe you can save the initial stress state in an SDV and then superimpose it on the stress due to your material law.

Reply
Share: