Heat transfer - Radiation

Monin · August 31, 2017, 21:32

Trying to make a heat transfer simulation of a filament in a vacuum chamber.

H2 gas inside.

Setup

[IMAGE]

Black Squares - Shields

25C - Chamber wall with cooling - Set as wall at the end of Fluid domain with 25C

Green Square - Filament - Set all wall with a temperature at 2500C

Gases cannot be omitted in this case.

------

Not entirely sure how to model this correctly, which simplifications can be made?

Does anyone have a good example on Fluid Solid Radiation?

How to correctly do the Domain Interfaces between the plates, chamber and fluid?
Feels like I made it extra difficult due to the Filament simplification.

Any pointers which assist this case are greatly appreciated.
Cheers!

ghorrocks · September 1, 2017, 06:14

You are modelling a vacuum. This means no fluid. CFX is a CFD code, and the F stands for fluid. So CFX is not an ideal software for modelling this as it is purely a heat transfer simulation with no fluid dynamics. Having said that CFX can model this, but it is the hard way to do it in most cases.

You can do a pretty good approximation of this using hand calculations. Is this suitable? If not, why not?

Opaque · September 1, 2017, 09:01

A bit confused with the setup. For starters, steel is not a participating media; therefore, there is no need to activate the radiation modeling within those domains.

Not clear which domain is supposed to be "vacuum".

JuPa · September 3, 2017, 07:01

When you say vacuum chamber, what is the pressure exactly? I work with people who describe 50 millibar as a vacuum.

I agree with Glen - unless you introduce a fluid this will be difficult. I also agree with his other statement, this looks like something you can do with a pen, paper and a heat transfer text book.

Monin · September 3, 2017, 19:55

Quote:

Originally Posted by ghorrocks

You are modelling a vacuum. This means no fluid. CFX is a CFD code, and the F stands for fluid. So CFX is not an ideal software for modelling this as it is purely a heat transfer simulation with no fluid dynamics. Having said that CFX can model this, but it is the hard way to do it in most cases.

You can do a pretty good approximation of this using hand calculations. Is this suitable? If not, why not?

Hi Glenn,

Yes, I said this is vacuum and you are probably right. The pressure might be too low for the fluid to be included in the simulation. At 20 [Torr]. This is a Hot Filament Chemical Vapor Deposition (HFCVD) and that is why I want to try and include the gas. We can use the number to compare with out experiments setup.

About the heat transfer, yes, can be done by hand. We have already paid some people a large sum for this to be calculated in 2D. This is my pet project to prove them that, it is a waste of money and it can be done in Ansys and still get reasonable results.

Also one of my supervisors love colors, hence CFD (ColorFul Display)

Monin · September 3, 2017, 20:18

Quote:

Originally Posted by Opaque

A bit confused with the setup. For starters, steel is not a participating media; therefore, there is no need to activate the radiation modeling within those domains.

Not clear which domain is supposed to be "vacuum".

Hi Opaque,

Think I am a bit confused.

Interior of the "Box" is the vacuum. See image.

I did a simulation without all the plates, only Fluid and set the absent space of the plates to adiabatic. It went pretty ok, however I want to include the plates and their heat transfer through them.

Ah, I see, I have misunderstood how these domain interfaces work. I believe I fail because of this.

Monin · September 3, 2017, 20:19

Quote:

Originally Posted by JuPa

When you say vacuum chamber, what is the pressure exactly? I work with people who describe 50 millibar as a vacuum.

I agree with Glen - unless you introduce a fluid this will be difficult. I also agree with his other statement, this looks like something you can do with a pen, paper and a heat transfer text book.

Hi JuPa,

The pressure is set to 20 Torr, also see my reply to Glenn.

ghorrocks · September 3, 2017, 20:41

Make sure you have a clear objective for this work. If you are just trying to get the heat transfer right then work out the heat transfer caused by the gas. If it is significant then include it, if it is not significant then do not include it.

Or are you trying to find what the gas flow inside the chamber is? This is a different objective entirely, and you would take a different approach to model it.

On whether the vacuum is too low to model in CFX - read up on the Knudsen number to determine if your gas is suitable for modelling by the Navier Stokes equations as used by CFX.

Monin · September 3, 2017, 22:34

Quote:

Originally Posted by ghorrocks

Make sure you have a clear objective for this work. If you are just trying to get the heat transfer right then work out the heat transfer caused by the gas. If it is significant then include it, if it is not significant then do not include it.

Or are you trying to find what the gas flow inside the chamber is? This is a different objective entirely, and you would take a different approach to model it.

On whether the vacuum is too low to model in CFX - read up on the Knudsen number to determine if your gas is suitable for modelling by the Navier Stokes equations as used by CFX.

Hi Glenn,

I've read about the Knudsen number. It is withing the continuum regime. So it shouldn't be an issue to include the fluid in this case.

Agreed, a clear objective here is important.

Actually both. Heat and flow. I have indications that the buoyancy plays a role in this and altering the flow in the chamber.

Made some smaller simulations which points to this and now I am trying to do a more large scale simulation to validate my claims.

I did some gas only simulations and skipped heat transfer completely.
How would you model this considering only gas flow? Your approach is most likely better than mine, always want to learn more!

ghorrocks · September 3, 2017, 23:14

I will leave the questions of the objective up to you and answer your direct questions:

OK, so if you are in the continuum region then you should be OK to model it in CFX.

If you choose to do this I recommend you ignore the thermal conduction in the outer case and just model the fluid flow by itself. Get this working before you combine it with thermal conduction to get a CHT simulation.

You will need the gas properties, including the buoyancy properties to do this. Do you have these properties? Is the gas totally transparent, or does it have significant opacity?

Then it is a matter of setting up the boundary conditions and simulating it. Note that there is highly likely to be significant amounts of complex flow, like transient behaviour, plume flapping and other complex behaviour.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
chtMultiRegionSimpleFoam: Thermal Conduction + Surface-To-Surface Radiation	Zeppo	OpenFOAM Running, Solving & CFD	16	May 18, 2017 18:04
Interphase mass transfer of a reaction	cfx_ws1992	Main CFD Forum	0	May 15, 2017 21:42
Difficulty In Setting Boundary Conditions	Moinul Haque	CFX	4	November 25, 2014 17:30
Conjugate heat transfer and radiation modeling questions	shankara.2	FLUENT	0	April 21, 2009 15:55
Radiation Heat Transfer	The IFRF Heat Transfer Team	Main CFD Forum	0	January 31, 2000 14:02

September 1, 2017, 06:14		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,723 Rep Power: 143	You are modelling a vacuum. This means no fluid. CFX is a CFD code, and the F stands for fluid. So CFX is not an ideal software for modelling this as it is purely a heat transfer simulation with no fluid dynamics. Having said that CFX can model this, but it is the hard way to do it in most cases. You can do a pretty good approximation of this using hand calculations. Is this suitable? If not, why not?

September 1, 2017, 09:01		#3
Opaque Senior Member Join Date: Jun 2009 Posts: 1,815 Rep Power: 32	A bit confused with the setup. For starters, steel is not a participating media; therefore, there is no need to activate the radiation modeling within those domains. Not clear which domain is supposed to be "vacuum".

September 3, 2017, 07:01		#4
JuPa Senior Member Mr CFD Join Date: Jun 2012 Location: Britain Posts: 361 Rep Power: 14	When you say vacuum chamber, what is the pressure exactly? I work with people who describe 50 millibar as a vacuum. I agree with Glen - unless you introduce a fluid this will be difficult. I also agree with his other statement, this looks like something you can do with a pen, paper and a heat transfer text book.

September 3, 2017, 20:41		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,723 Rep Power: 143	Make sure you have a clear objective for this work. If you are just trying to get the heat transfer right then work out the heat transfer caused by the gas. If it is significant then include it, if it is not significant then do not include it. Or are you trying to find what the gas flow inside the chamber is? This is a different objective entirely, and you would take a different approach to model it. On whether the vacuum is too low to model in CFX - read up on the Knudsen number to determine if your gas is suitable for modelling by the Navier Stokes equations as used by CFX.

September 3, 2017, 23:14		#10
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,723 Rep Power: 143	I will leave the questions of the objective up to you and answer your direct questions: OK, so if you are in the continuum region then you should be OK to model it in CFX. If you choose to do this I recommend you ignore the thermal conduction in the outer case and just model the fluid flow by itself. Get this working before you combine it with thermal conduction to get a CHT simulation. You will need the gas properties, including the buoyancy properties to do this. Do you have these properties? Is the gas totally transparent, or does it have significant opacity? Then it is a matter of setting up the boundary conditions and simulating it. Note that there is highly likely to be significant amounts of complex flow, like transient behaviour, plume flapping and other complex behaviour.