Thermo-mechanical modeling using 2 modelling (pure thermal + mechanical)

General Questions
1

Hi,

In the following test case, i’m using 2 files:

  1. pure thermal simulation
  2. mechanical calculation including the thermal expansion

The temperature increases from the room temperature up to 600°C between t=1s to t=2s; i’m not figuring out why in the mechanical simulation, T=600°C at 1s => what i’m missing (maybe be in the step definition)?

Thanks

Paul

Following screenshots follows data from the “spy” node

screenshot_purethermal
screenshot_purethermal1096×654 52.6 KB

screenshot_thermomeca
screenshot_thermomeca1098×652 71.9 KB

pure_thermal.pmx (3.7 MB)
thermo-mechanical.pmx (7.5 MB)

2

I guess CalculiX wants to apply 600°C at the end of the step. Your static step ends at 1 s. If you extend it to last 2 or 3 s then 600°C will be applied at 2 or 3 s:

image
image561×412 13.8 KB

P.S. I would test it on a simple cube first.

3

But as in the thermal modelling, i want to have 20°C at t=1s …

4

I did such a simple test and compared it with Abaqus. The behavior is different. Abaqus maps the whole time variation of the temperature and sets 20°C at 1 s while CalculiX simply takes the value from the last increment (there’s no BINC parameter here) and applies it to the last increment of the mechanical analysis.

5

whatever i’m doing, I cannot apply the correct loading at 20°C :tired_face: in practice the preloadinf is performed at 600°c and not 20°C :scream:

I also testec “AMPLITUDE=STEP” feature to “*Step”, but it failed as well (.inp file run with ccx in standalone)

Link (available until 03/12/2025): link

6

Do you really need that thermal analysis here ? You could apply the temperature directly in the mechanical analysis:

*Amplitude, Name=amp
0, 20, 1, 20, 2, 600
*Temperature, Amplitude=amp 
All_nodes, 1
7

If I understand correctly, would it help to split the step into two? Step1 until 1s/20°C, Step2 until 10s/600°C?

8

I tried to split into 2 steps … but with no sucess after several trials; i’m missing something …

9

Hi

If i set aside thes issue in using thermal .frd files, i’ve modified the model using “Coupled temperature-displacement step”; i think all inputs have been defined (including the gap conductance) but the FEA does not start and i do not understand why?

I’ve the following warning but i’m wondering if the issue comes from: any suggestion?

 *WARNING in calinput: no conductivity 
          constants were assigned to material SURFACE_INTERACTION
          in a thermo(mechanical) calculation

thermo-mechanical3.pmx (685.7 KB)

Thanks

10

hmm… i never tried this before but I have the feeling that no matter what step number is entered for the temperature field, always the results of the first step will be used.
At least I would have expected an error if the entered step number cannot be found in the .frd file:

t
t344×370 5.91 KB

11

There’s a generic error message “Job failed - no results exist”. It usually means some modeling error. I would simplify the model to find which feature causes it. As I said before, such tests should be done on simple models first before proceeding to the actual usecase.

Btw. your pre-tension is applied incorrectly. It should be applied to interior face of a bolt split into two parts with boundary layer defined: https://www.youtube.com/watch?v=7ucoBTqCcWE