Field values diverge in convective HT simulation

[ssantoz96]

Hi people,

I am doing a convective heat transfer analysis in engine - convective_HT by ssantoz96 | SimScale & convective_HT_2 by ssantoz96 | SimScale - cooling jacket.

The main idea was to specify a fixed power heat flux [W] on few faces, others to keep adiabatic and see, how much heat would be absorbed by the water flowing through the channels. Results should provide input for pressure drop calculation and wall temperature load determination.

Boundary conditions could be seen in the mentioned projects, and also on the pictures below:

– INLET

– OUTLET

– HEAT FLUX

– ADIABATIC

.

I have some troubles during simulations

and

.

I don’t know if the problem really lays in geometry or maybe in B.C. conditions.

I did offline calculations in OF23x (using buoyantBoussinesqSimpleFoam solver) on my local machine and didn’t see the problem, that occurs here, but I had a problem concering the overall temperature difference of water.

My problem was, that I followed basic physics, that means

power_heat_flux[W]=mass_flow[kg/s]*termal_capacity[J/kgK]*temperature_difference[K]

but when I integrated phi on inlet and outlet and multiplied by density to get the mass_flow and also averaged temperature on inlet and outlet and put in the equation above, I got much bigger temperature difference from simulation, then by hand.

This was the main reason, I came here. I wanted to verify that I did some mistakes locally, but a whole lot different problem occured.

Kind regards,
Sergej

[Ricardopg]

Hi,

I had a look at your cases. The first one seems to be a very old geometry from an old project - I’d recommend, before running any simulation, to follow the instructions provided:

In any case, could you please clarify if you are intending to define a total power of 1325 W to all faces, or 1325 W to each face?

Please note that, for power definition, the value is per assignment (see here for example). Currently, you are defining a power of almost 170 kW, which is probably not intended.

Cheers

[ssantoz96]

Hi,

first of all, sorry for my late reply and thank you for your suggestions and help.
I will update the outdated geometry and it will probably do the job.
Great for pointing out the power loading asssignment procedure. I have to be honest, I haven’t read the documentation and obviously it got me there. I did some modifications and now it works perfectly.

I wish you joyful, pleasant, and relaxing Christmas holidays!

Take care