problems with VENER and general energy question

[gunkerle]

Hello everyone,

I have some problems with respect to the output variable
VENER. I simulate a visco-elastic material using time domain proy series data to simulate the viscous behaviour. The aim is to simulate the thermo-mechanical behaviour of the material. I am using the C3D8T elements. As it is a dynamic simulation I am using the explicit solver. For output purposes all energies in the Field output and all kinds of energies in the history output options are toggled on. I am using fully coupled simulations. The simulation consists of a cyclic and non-cyclic loading part.

My problem is as follows:
I get the output for ALLVD. Thus Abaqus seems to calculate the time integrated value for the the viscous energy.

But I get no value, more precisely only the value "0" for the variable VENER.

I looked into the manual in order to find out if I need something else and I found some discrepancies in the description regarding the output for viscous energies in Abaqus:


Due to the explanation in the ouput variable index it is valid:

VENER (=viscous energy density)
Energy dissipated by viscous effects, per unit volume.
.fil: no .odb Field: no .odb History: yes

Thus the variable VENER should only exists as HISTORY OUTPUT. But There exists no variable VENER.



As mentioned before I have the problem, that I get results for

ALLVD (VD= viscous dissipation)
Energy dissipated by viscous effects.
.fil: yes .odb Field: no .odb History: yes

but no results (only results "0") for the variable VENER.

On a different site in the manual the following statement is written:

The energy ALLCD is the output variable for visco-elastic dissipation,
but ALLCD stands for CREEP-Dissipation this ist - to my best knowledge NOT identical with the visco-elastic dissipation of a pure viso-elastic material (no plasticity involved).

Is this correct?

ALLCD,
Energy dissipated by viscoelasticity. (Not supported for hyperelastic and hyperfoam
material models.)
.fil: yes .odb Field: no .odb History: yes

To make things worse:
Only the ALLCD (but not the energy part for viscous dissipation ALLVD !!) energy is included into the total strain energy ALLIE (see below).


ALLIE
Total strain energy. (ALLIE=ALLSE + ALLPD + ALLCD + ALLAE + ALLDMD+ ALLDC+ ALLFC.)
.fil: yes .odb Field: no .odb History: yes

But ALLVD IS part of the total energy ETOTAL. As you can see ALLVD is an additional part of the TOTAL ENERGY.

ETOTAL
Energy balance defined as: ALLKE + ALLIE + ALLVD + ALLFD + ALLIHE ? ALLWK ? ALLPW ? ALLCW
? ALLMW ? ALLHF. (Available only for the whole model.)
.fil: yes .odb Field: no .odb History: yes


I now have the following qustions:

1.)
Why is no contribution of the viscous part included into the strain energy part?

2.)
How can I get results for the variable VENER?
Meaning: Do I need other variables than the energy densities (in Field Ouput) or Energies in History Output or do I need special elements to do this (other than C3D8T)? I am calculating a fully coupled simulation, because I need the interaction between the themical and mechanical behaviour.

3.) How is it possible to include the ALLVD part into the total strain energy, menaing how can I define a contribution on my own?


If I am defining my own heat source I need at least the contents of the variable VENER and possible and a modified version of the strain energy ALLIE.
In the moment in Abaqus Explicit there IS NO viscous dissipative energy contribution defined as a heat source. In order to be able to code my own heat source using Abaqus explicit I need access to the viscous variables in Abaqus, but as stated above in the moment it does not work as it should.


I hope someone can help me

ciao gunkerle

[Jorgen]

Here are a few comments:

1.) I suppose that the strain energy should only contain stored energy, the viscous energy is dissipated

2.) I'm not sure about that. Perhaps you can try without full thermomechanical coupling, and/or using an implicit simulation. This could help determine when VENER is available.

3.) For Abaqus/Standard you can write a UVARM subroutine. There is no corresponding feature for Explicit.

- Jorgen

[gunkerle]

I know now a little bit more about what happens with the variable VENER.
This variable is definitively not available using the field output option. Nevertheless Abaqus is initializating this variable with zeros, even though it does not compute anything with this variable.


What I do not now further on is:
When is Abaqus using the variable VENER and in which situations the variable CENER (or ALLCD, ALLVD, respectively) to compute the dissipative part of the energy. If I habe a thermo-mechanically coupled system the dissipated energy should heat the material and thus should produce thermocally induced stress & strain. Due to this it should be part of the strain energy not directly but via the induced stress & strain.

In order to use the dissipated part as a heat source I have to know what really Abaqus is doing meaning which of the offered energies are the "real" dissipation energies.

If I am using hiperelastic models in conjunction with visco-elastic models I have to know the following:
Especially I have to know why the variables CENER & ALLCD are not usable if one is using hyperleastic material models in conjunction with e.g. visco-elastic material models. Meaning what are the dissipative energies due to the material behaviour in this case??

Hoping for help,
ciao Gunnar