CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Radiation at interface between fluid and porous domain (http://www.cfd-online.com/Forums/cfx/76927-radiation-interface-between-fluid-porous-domain.html)

Hitch8 June 8, 2010 13:57

Radiation at interface between fluid and porous domain
 
Dear CFX experts, I need help!

I try to model a reactor which consists basically of a porous a fluid and a solid domain. Because I've to deal with very high temperatures radiation is very important. I use the Monte Carlo Model to simulate the radiation inside the fluid and the solid domain. From my point of view I don't need to simulate the radiation heat transfer through the porous media, I consider it by an increased thermal conductivity. But what happens at the interface between porous and fluid domain. Is the porous domain emitting radiation inside the fluid domain and absorbing radiation? When I enable the Beta features it says that the 'opaque' option for non-wall boundary types is enabled.

It would really help me a lot if anyone knows the answer!

Regards,
Philipp

stumpy June 8, 2010 14:56

First, you said you had radiation inside the solid domain. Is it glass, or some other transparent media? If not, turn off radiaiton inside the solid domain since it cannot travel through it.
Radiation does not interact with a porous domain. So if you have radiation turned on inside the porous domain, radiation just sees this as a fluid domain and will pass straight through it. If you have radiation turned off inside the porous domain, then I think the fluid-porous interface is just seen as opaque by the radiaiton, so there's no emission of radiaiton by hot porous media (or absorbtion). So there's currently no way to allow porous media and radiation to interact correctly.

Hitch8 June 8, 2010 15:24

Dear Stumpy, thank you very much for your answer.

If I want to simulate the heat transfer in a fluid domain between two solid domains at different temperatures (opaque solids), I don't need a radiation model in the solid domains? Is CFX able to simulate the radiation heat transfer from one solid domain to the other (emitted and absorbed radiation)?

Regards,
Philipp

ghorrocks June 8, 2010 18:34

Quote:

If I want to simulate the heat transfer in a fluid domain between two solid domains at different temperatures (opaque solids), I don't need a radiation model in the solid domains?
If radiation is a significant source of heat transfer then you need it. If not you don't.

Quote:

Is CFX able to simulate the radiation heat transfer from one solid domain to the other (emitted and absorbed radiation)?
Yes.

Hitch8 June 9, 2010 02:41

Dear Glenn,

Thank you very much for your answer! So CFX is only able to treat the radiative heat exchange between a fluid and a solid domain (opaque) correct if in both domains a radiation model is enabled. Is that correct?

Many thanks and regards,
Philipp

ghorrocks June 9, 2010 06:25

Yes. If the solid is transparent then enable radiation in the solid and the radiation can penetrate the solid. If the solid is opaque then radiation interacts with the external interface only, by absorbing and emitting radiation.

Hitch8 June 9, 2010 07:47

Glenn, thank you very much!!

Hitch8 June 10, 2010 05:01

Just one more question:
My solid material is opaque, I just want to model interaction at the surface, is it correct to use in this case the option surface to surface? Or should I choose participating media and define the radiation properties (absorption coefficient...) in the material definitions?

Thank you very much

ghorrocks June 10, 2010 08:05

If the material is opaque then turn the radiation model off in that domain. Then it does not matter what radiation properties you define.

Hitch8 June 10, 2010 09:11

I'm still not sure if I got it correct:

In the case of opaque solid domains, radiation models inside the solid domains are not needed to simulate the radiative interaction between a fluid and a solid domain (emitted and absorbed radiation at the interface of the solid and fluid domain)? I just need to enable a radiation model in the fluid domain? I'm only interested in what happens at the surface of the solid domain. I don't want to simulate the radiative heat transfer through the solid.

Radiation models inside the solid domains are only needed to simulate the radiative heat transfer inside the solid? Is that correct?

I'm sorry to ask you that many questions...

Thank you and regards,
Philipp

ghorrocks June 10, 2010 18:23

Yes, you are correct. If you turn off radiation in the solid then heat transfer will still occur in the solid, but just by conduction. The radiation will only affect the outer surface.

Hitch8 June 11, 2010 09:56

Glenn, thank you very much! You helped me a lot!
Philipp

eto June 15, 2010 06:20

thx
 
Thanks Glenn, it helped me very much too.

mpeppels April 17, 2015 06:46

Hi everyone,

sorry to reanimate this thread but maybe someone will read this even though its last been updated 5 years ago:

I am currently investigating a model consisting of a fluid domain reaction chamber separated from fluid domain surroundings by adiabatic walls and one conductive wall.

Radiation is incident on this emitter, the wall is opaque but not black.
The way I understand it from this thread, I have no need for a radiation model in the solid plate.

However, if I do not employ a radiation model in the plate, I seem to be unable to specify surface emissivities on both sides of the plate for the surface-to-surface radiation models in the fluid domains to incorporate.

Is there a way of doing this without giving the solid domain (plate) a radiation model (because if I do, the linear solver delivers a fatal error) ?

Thank you very much in advance,
Max

ghorrocks April 17, 2015 07:56

I feel like a bit of reanimation sometimes. Some slightly strange individuals are still hanging about on the forum even 5 years down the track :).

You only need to have a radiation model in a domain where radiation either passes through it or radiation heats/cools it internally. If the heating/cooling is external (such as a hot metal block radiating heat, but no radiation occurring inside the block) then you do not need a radiation inside the block but might need one in the domain around the block for the interaction of the block with its surroundings.

mpeppels April 17, 2015 08:32

Thanks Glenn,

the plate should emit into the reaction chamber at high temperatures. These tepmperatures would result from incident radiation on the plate, so I guess a radiation model is required then.

Is it even possible for the plate to experience heating by radiation with a surface to surface MC model?

Whatever configuration I use with radiation in the plate, as soon as writing the radiation balances in the first iteration the solver aborts with this general ERROR #001100279.

I cannot really identify the problem from that, all I know is that the solver stopped while being in the directory /RADIATION/RG5. Does anybody know what is stored in there? Is there anyway to examine the solution directories in CFX without putting together a UserFortran routine?

Thanks,
Max

ghorrocks April 18, 2015 02:14

Quote:

These tepmperatures would result from incident radiation on the plate, so I guess a radiation model is required then.
I suspect you misunderstood what I said. If occurs on the other faces of a solid block but the internal thermal field is determined by conduction (that is, the block is opaque, like most blocks would be) then you do not use a radiation model inside the block. There is no radiation acting inside the block.

Quote:

Is it even possible for the plate to experience heating by radiation with a surface to surface MC model?
Yes. The discrete transfer model can do this as well and it is much simpler and quicker.

Quote:

Whatever configuration I use with radiation in the plate, as soon as writing the radiation balances in the first iteration the solver aborts with this general ERROR #001100279.
FAQ: http://www.cfd-online.com/Wiki/Ansys...do_about_it.3F

Don't worry about where the solver stopped for a floating point exception. All it means is it is probably something to do with the radiation model. Run the model again without radiation and if it runs then you know you have a problem with your radiation model.

mpeppels April 18, 2015 08:53

Quote:

I suspect you misunderstood what I said. If occurs on the other faces of a solid block but the internal thermal field is determined by conduction (that is, the block is opaque, like most blocks would be) then you do not use a radiation model inside the block. There is no radiation acting inside the block.
It was not so much the incident thermal radiation on the plate that gave me the idea of employing a radiation model there but the fact that on the other side I need to accurately capture the effects of the emitted radiation from the plate into the second fluid domain on the other side.

I attached a picture of the geometry setup (I didn't manage to embed it with the picture button here, but the link should be sufficient: http://www.directupload.net/file/d/3...5xhsse_png.htm), its basically the red solid plate being 'sandwiched' by the dark blue surroundings fluid domain (including a cavity geometry above the plate) and the cyan reaction chamber fluid domain.
The issue whether or not to employ a radiation model for the solid is mainly dependent on the question if the latter case will still impose radiative heat transfer from plate to reaction chamber.

And if so, where can I tell the fluid domain MC to account for a certain emissivity at the plate-fluid interface (the IF definition itself in pre clearly won't let me define any of the kind if there is no radiation model employed in either of the domains.)

Quote:

Yes. The discrete transfer model can do this as well and it is much simpler and quicker.
I'm pretty sure it is impossible to use DT on solid domains?

Thank you for your patience!

EDIT: I figured some more explanation for the setup might be good: in the top center of the surroundings domain there is a directional radiation source with a strong negative z-direction component. Radiation mainly hits the plate and heats it. Below the plate there is the reaction chamber which receives energy mainly by reradiation from the plate (puttting a window instead of a reradiating plate fails as the in the actual setup reaction products like soot will reduce transmissivity to zero in no time)

ghorrocks April 19, 2015 06:44

I am not completely sure whether DT can handle solids - I suspect it can. But it certainly cannot handle refractive index causing the radiation to bend, or absorption in the solid or the gas. You will need MC to handle this.

To set the emissivity at the gas/solid interface, open the interface boundary in the domain level. You will find the radiation boundary controls there.

What material is the red solid plate? Your final comment suggests it is opaque. In that case, you only have one choice - no radiation model in the solid. The top domain interacts with the red solid via the interface boundary condition, heat is transfer through using conduction only, and another interface radiation boundary condition causes reradiation inside the inner chamber.

mpeppels April 20, 2015 06:24

Yes, it is opaque, the heat transfer inside the plate is purely conductive.

Thank you for hinting me to the domain level interface definition, I think I achieved the setup I intended now.


All times are GMT -4. The time now is 11:39.