Thermal steady state analysis

Hengre

Hello
I am playing around with some thermal calculations (first time).
I could poduce some results, but I am wondering about some values I got trying to validate the result. I used a coarse mesh to speed up. The attached PDF shows where the part is heated up and where I want to cool it. Cooling is done by waterflow throug holes. Only the lower 2 holes are used in this case. I modeled that using convection.

On page 3 I compared the energy flow in and out of the body an got a difference. I would expect the values should be almost equal, isn't it?
The calculation was made using CCX as solver.

Any kind of help or guidance is apreciated.

Regards

Victor

It may be a bit too rough to integrate the heat flux magnitude since it can include some component tangent to the surface or in the opposite direction which won't be counted properly.

The error seems to be in the wrong direction for that though so perhaps it just needs mesh refinement. The heat flux values in the solution are obtained by integrating temperature over the whole element so they suffer from discretization error.

For the flat surface, at least use the component normal to it (Z?) instead of the magnitude.

Confirm the surface areas are right using Tools -> Surface area.

For the surfaces of the holes, I think the *SECTION PRINT card in CCX calculates heat flow rate, so hopefully you can just add that to CCX -> custom step contents. Like below then look in the .dat file for the result.

*SECTION PRINT, SURFACE=A_E-Halterkühlbohrungen, NAME=SP1
FLUX

Hengre

Thank you Victor,
so my thoughts were right.
I will try your suggestions tomorrow . I'll got to bed now :-)
Have a nice day

Hengre

Hello Victor,
I have placed the following to the coustom step contents

*SECTION PRINT, SURFACE=A_E-Halterkühlbohrungen, NAME=SP1
FLUX
*SECTION PRINT, SURFACE=A_Grundplattenkühlbohrungen, NAME=SP2
FLUX
*SECTION PRINT, SURFACE=A_Schlittenplattenkühlbohrungen, NAME=SP3
FLUX

From the DAT.File I got
total surface flux (q) for set A_E-HALTERKüHLBOHRUNGEN
2.171730E+03 (4168 W)
total surface flux (q) for set A_GRUNDPLATTENKüHLBOHRUNGEN
1.567979E+02 (199 W)
total surface flux (q) for set A_SCHLITTENPLATTENKüHLBOHRUNGEN
3.268216E+02 (469 W)

The values in brackets i git using Solution -> Surface integral with heat flux magnitude and the appropriate surface.

But I am still worried about the differences. Is Calculix calculating in a different way compared to Mecway? Which value is supposed to be the mor correct one?

Regards

Victor

I expect CCX will be more accurate. If it's treating them analogously to external force, then it should be accurate regardless of mesh density, but if it's using the temperature gradient like Mecway does, you need to check for mesh convergence by refining. Both ways are more correct than integrating heat flux magnitude, especially on the complicated tube surfaces where I expect heat flux magnitude would give a completely wrong result.