Hello everyone,
I am trying to solve a rather simple thermal case : a 5x10x20mm cube heated through a 1W heat flux, and cooled thanks to a 10W/Km2 surface convection. From the same input, CalculiX solver gave me a temperature raise of 140°C (what I was expecting), while Elmer solver returned me an unrealistic 60000°C temperature increase.
After a quick investigation the issue seems to come from the .sif generation. The 1W boundary condition was written as "Temperature Load = Real 1000000.0". I could not find much documentation about this temperature load constraint, so I replaced it with a "Heat Flux = Real 20000" (20000 is the conversion of my 1W heat flux to W/m2 as I have a 5x10mm surface). A new calculation with this change gave me the exact same result I found with CalculiX.
My information:
OS: Windows 10 (10.0)
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.19.20802 (Git)
Build type: Release
Branch: master
Hash: f4279f1bea8377a7ef63f7f4c69b67159d498965
Python version: 3.8.2
Qt version: 5.12.5
Coin version: 4.0.0
OCC version: 7.4.0
Locale: French/France (fr_FR)
Elmer 8.4
Also, I'm new to FreeCAD and this forum, so I'm not sure this is the best way to report an issue/bug. Therefore, please feel free to tell me the best way to do it!
Antoine
Thermal analysis : difference between Elmer and calculiX results
Moderator: bernd
Forum rules
and Helpful information for the FEM forum
and Helpful information for the FEM forum
Re: Thermal analysis : difference between Elmer and calculiX results
would you post the FreeCAD file?
Re: Thermal analysis : difference between Elmer and calculiX results
Where did you know the expected result from? Is it that simple you could do it with a calculator? Did you do the analysis with another software?
Re: Thermal analysis : difference between Elmer and calculiX results
Hi,
I got the approximate result from a quick calculation :
I have a 1W input, and a 10W/m2K convection output through a 700e-6m2 exchange surface. All that put together ends up to a 142.8°C temperature increase. This meets the Calculix result.
I joined to this message the case. For file size considerations, I had to remove the Calculix and Elmer results, so solvers should be run again.
I got the approximate result from a quick calculation :
I have a 1W input, and a 10W/m2K convection output through a 700e-6m2 exchange surface. All that put together ends up to a 142.8°C temperature increase. This meets the Calculix result.
I joined to this message the case. For file size considerations, I had to remove the Calculix and Elmer results, so solvers should be run again.
- Attachments
-
- cube_light.FCStd
- (723.81 KiB) Downloaded 72 times
Re: Thermal analysis : difference between Elmer and calculiX results
Hi
In Elmer "Heat Flux" is the BC one would usually want to set. In SI units it is given in W/m^2. The "Temperature Load" is the nodal heating given in W. The latter is mesh dependent. The "Temperature Load" is actually directly the r.h.s. term in the matrix equation. It is a library feature i.e. you can apply "Varname Load" as a nodal r.h.s. load to any physical module.
-Peter
In Elmer "Heat Flux" is the BC one would usually want to set. In SI units it is given in W/m^2. The "Temperature Load" is the nodal heating given in W. The latter is mesh dependent. The "Temperature Load" is actually directly the r.h.s. term in the matrix equation. It is a library feature i.e. you can apply "Varname Load" as a nodal r.h.s. load to any physical module.
-Peter
Re: Thermal analysis : difference between Elmer and calculiX results
Thanks!raback wrote: ↑Fri May 08, 2020 11:46 pm In Elmer "Heat Flux" is the BC one would usually want to set. In SI units it is given in W/m^2. The "Temperature Load" is the nodal heating given in W. The latter is mesh dependent. The "Temperature Load" is actually directly the r.h.s. term in the matrix equation. It is a library feature i.e. you can apply "Varname Load" as a nodal r.h.s. load to any physical module.
Nobody reacted yet but we have a bug in FreeCAD in this respect.
I will fix this for FreeCAD 0.20.2.
Re: Thermal analysis : difference between Elmer and calculiX results
This is now fixed. To benefit from this right now:
- get FreeCAD 0.20.1: https://github.com/FreeCAD/FreeCAD/releases/tag/0.20.1
- apply this package of patches: https://forum.freecadweb.org/viewtopic.php?f=18&t=71058