[zfaraday]

Dear Foamers,

I'm working in a case where a sample is heated inside a cylindrical heater by radiation. Convection should be neglected since the heater is placed within an enclosure in vaccum conditions. The heater is made of a very thin metal (0.1mm thickness) in which an electrical current is injected to make it generate the required thermal power to heat up the sample. Below you can see my geometry.

You can see that I generated it as a cylindrical mesh using blockMesh because all the sistem has cylindrical symmetry.

As I don't need to compute anything in the air but the energy transference (only by radiation) I used frozenFlow option in the SIMPLE subdictionary within fvSolution file, so that no velocity nor pressure are computed to speed up the simulation.

As I said, the heater is really thin (0.1mm) and for this reason at the very beginning I tried to create it as a baffle with not much success. I created the shields as a baffle too (2mm thickness) but it all gave me a lot of trouble and finally I decided to create the thinner regions as solid elements. To simulate the power generated by the heater I used the fvOptions framework.

I ran some simulations and the results I got, when it comes to temperature distributions, are quite satisfying: the ranges of temperatures are very similar to the ones obtained with another software for heat transfer. Maybe the distributions are not exactly the same, but the mean values are quite close compared with the results I first obtained when I created the thinner regions as baffles... That made me a little glad!

However, not totally satisfied with that, I used som function objects and wallHeatFlux utility in order to compute all heat fluxes in the air region (some fluxes in the solids also) to check heat balance. And here comes the problem... Even though my case has converged (the solver is not iterating for any region for the last few time steps), no balance is gotten when it comes to energy.

I computed radiative power in all patches belonging to the air region as

Qr*Area

Thus, if I sum all of the values gotten from all the patches I should get a value of 0 (energy cannot be created nor destroyed!) if I'm not wrong... Well, then I don't get any 0 but a value around 6-9 depending on the case. May anyone explain me why is it happening to me?

Not only the balance in the air is not balanced, but also is unbalanced for the solid regions, some of them are more or less balanced, but others have a magical gain or a loss of energy. Here you can see the balance for the sample as an example (extracted from the wallHeatFlux utility):

Wall heat fluxes [W]
sample_to_air 0.05049047
sample_to_support -0.0600329

Another point I don't understand is the following: If I set the value of the generation in the fvOptions file as, say, 50W, am I not supposed to get the same value when I compute the power in the post-processing stage? I ask that because it's not happening to me. As an example I share some numbers with you. I set the generation in one of the tests I've done as 16W in the fvOptions. After the simulation I got these values with wallHeatFlux and function objects:

• wallHeatFlux
  

  Wall heat fluxes [W]
  heater_to_air -17.56851

• function object to compute radiative power
  

  Expression Qr_air_patches on air_to_heater:  sum=-17.56811

Why is not the same value of power I set as a generation?

At this stage, I'm a little lost because I ran a few simulations using different values but I always got an unbalanced system. I don't know if the problem may be related to the geometry and the thin regions I need to use for this simulation or maybe it is something regarding the fvDOM specification (since I'm not very experienced in the use of this method). I would appreciate any word you can give me regarding this matter since I can't go further on my own...

Many thanks in advance.

Best regards,

Alex

[jaydeep]

too little too late from me

https://bugs.openfoam.org/view.php?id=2759