About documentation and some questions I met

[xibaiw]

Hello
I am currently attempting to simulate a thermal cracking test on a ring-shaped stone material, and I have encountered some issues during this process.

1. It seems that the documentation may lack intuitive descriptions for certain settings in the input file. For example, I want to perform some calculations in my boundary conditions, such as square roots, but I couldn't find detailed instructions in the documentation on what keywords to use or how to format these expressions in the value field.
2. I combined the thermal flow and thermal expansion models for the calculation, but the results show that all material points are oscillating back and forth. This phenomenon does not occur when using the two models separately.

[CWillberg]

@xibaiw
If you send me your example, I can try to check if something is wrong.

Regarding the thermal flow and thermal expansion model in combination.

If you have only thermal flow, no mechanical coupling occurs. Therefore, no vibration is okay. If you have the thermal expansion, vibrations should occur a little bit, because of the Verlet solver.

If you couple them and you give the complete thermal field as in a thermal expansion problem, the result should be identical. If you have a temperature boundary condition and the temperature rises, the vibration should occur for Verlet.

But thats the theory. Send me your model and I will check the issue.

[email protected]

[CWillberg]

@xibaiw I checked you yaml. The issue is with the initial temperature. The initial ttmperature will compute you the deformation. This leads to the shock. We have to be more precise. Temperature means temperature difference. Zero is your reference. If you need a reference value for convenience, write an issue I will add it.

[xibaiw]

Thank you Mr.Willberg, I understand. Regardless of how I set my boundary conditions (BC), PeriLab's initial temperature, or the starting point at 0 seconds, is always 0°C.
Additionally, a guide on how to set up calculation functions(like sin cos or srqt) in the "value" field might be more user-friendly for users.

[CWillberg]

If you want to use cos, it should work like this.

x,y,z are the placeholder for the coordinates and t for the time.

BC_1:
  Variable: "Temperature"
  Node Set: "Node Set 3"
  Value: "cos(100*t)+sin(x*5.0)+sqrt(2)"
  Type: Dirichlet