[PSander]

I am working with chtMultiRegionFoam on OF7 and trying to solve a transient calculation. The geometry is one fluid surrounded by an inner (tube) and an outer solid (soil) and I want to simulate heat expansion between the fluid and solids.

So far the model is working, but if I check the results I face the problem that the temperature expansion inside the fluid is not matching the fluids velocity.
The internal temperature of the fluid is 283°K and the inflow from the inlet has 303°K with a velocity of 1 m/s.

So my idea was that after 100 seconds of flow with a velocity of 1 m/s the fluid with 303°K should have travelled approximately 100 meters.
But it seems to have travelled only about 20-30 meters.

I checked already the scale of the mesh, it is definitively 100 meters long and not for example 1000m, the velocity is 1 m/s.
As well I played with gravitation, changed it between x, y and z axis with different directions and values but without any change.

I tried to figure out if T and U are not connected, but as well this seems not to be the problem.

Do you have any idea/hint where to look at? It would be great help to me.
In the .zip you find my case.

To create the mesh I use
"blockMesh" and "splitMeshRegions -cellZones -overwrite".

Thank you in adavance!

[jherb]

I might be related to the under-relaxation you are doing (relaxation factors < 1). Then the result is not really reached for each time step, so your fields get smeared out.



Try using (much) more then 1 nOuterCorrectors. This has to be set in the system/fvSolution file (and not in the system/fluid/fvSolution file for chtMultiRegionFoam.

[PSander]

Originally Posted by jherb

I might be related to the under-relaxation you are doing (relaxation factors < 1). Then the result is not really reached for each time step, so your fields get smeared out.



Try using (much) more then 1 nOuterCorrectors. This has to be set in the system/fvSolution file (and not in the system/fluid/fvSolution file for chtMultiRegionFoam.

Thank you for your help!
Increasing nOuterCorrectors as you said fixed my problem. I choosed now 50 instead of 1.

Do you recommend to increase the relaxation factors? Because with values of 0.9 or higher I get after a short time the error of negative initial temperatures. So far I am working with 0.5, 0.75 should be possible without collapsing.

[jherb]

Lower relaxation factor means more stable solution, but also might need more time. Higher factor might crash your solution. So it is always a trade-off.



Take a look at the tutorials for the outerCorrectorResidualControl settings. Then the PIMPLE loop is stopped, if the required resdiual settings are reached. So you do not have to wait for all nOuterCorrectors iteration.

[PSander]

Thank you a lot, this worked perfect for me.