Question to coupled temperature-displacement calculation - material parameters

General Questions
1

Hello Comunity!

At first i want to introduce myself: i studied material science at Montanuniversität Leoben 24 years ago, and since them i’m working in a steel producing factory.
since some weeks i have to do small FEM calculations, and so i found calculix and prepomax → i have to say “Thank you” for your work!

So now my problem: i want to do coupled, or uncoupled temperature-displacement calculations. for testing reasons i tried a small geometry and did the calculation with one of
the steels in the program - no temeprature dependence of the material parameters - it worked.
But: when i created a “new” steel with temeprature dependent parameters (calculatet with JMatPro) i get unbelivable results; somtimes the calculation stops, or the
diplacements are “exploding” at some points.

so i have questions: can you give me a hint and are there constraints concerning the temperature-dependent parameters of the material ( i have the calculated parameters
from ~1400°C down to 25°C).

Thanks a lot for your time!
Peter

2

Do you mean temperature-dependent elasticity or some other properties ? There are no special rules apart from the interpolation between the data points and extrapolation outside of the specified range. It should work normally, the only related bug I’ve heard of is for temperature dependent stress exponent when modeling creep: Creep properties interpolation problem? - #2 by flani - CalculiX (official versions are on www.calculix.de, the official GitHub repository is at https://github.com/Dhondtguido/CalculiX).

Maybe there’s some mistake/typo in your data. Can you share it ? Possibly also with the PrePoMax model. You could use some hosting website and paste the link here.

3

Hello and thank you for the quick answer.

The link to the simple test-model is below - hope it will work.

4

A few remarks:

  1. I’m surprised that CalculiX allows this, in Abaqus temperature and other independent variables have to be specified in ascending order and I would stick to that approach.
  2. Instead of heat transfer step followed by coupled temperature-displacement one, I would use two coupled steps.
  3. Is that supposed to simulate a specific process or is it just an abstract test ?
5

is just an abstract Test, but following i want to do a more complex structure and at first i wanted to get shure, that the calculation do not stop or bring “satisfying” results when it runs in the night;-)

6

I see. Temperature-dependent properties should work fine and the values you have look good. Maybe apart from this abrupt change in elastic properties which may cause convergence issues:

*Elastic
1.99E-16, 0.4999, 1575
2.75E-12, 0.4999, 1475
9950, 0.422, 1425

Specific heat also has some sudden peaks.

Still, I would define those properties in an ascending order of temperatures.

7

Looks like a phase transition ¿isn’t it?

8

Yes, steel getting almost zero Young’s modulus at 1475 degrees is clearly liquefying. Iron-carbon phase diagram can show it for different types of steel.

9

Thank you very much for the help!

and yes - it’s a phase transition. I will cut the properties at high temperatures to eliminate the liquid phase :wink: But i have a second transition (solid-solid) from ~550 to ~650°C. I think there i have to smoothen the properties to get convergence in the calculation.

Thank you again very much for your help!!