Modeling a vacuum chamber in Fluent

sagila · November 2, 2012, 02:18

Hello All, How can a vacuum chamber be modeled in Fluent?
Thank you.

RodriguezFatz · November 2, 2012, 04:13

You first need to estimate the Kundsen number. That's the mean free path (for neutral-neutral collisions) divided by the physical length scale of your chamber. If it is much smaller than 1, you can use Fluent. If not, your gas is no continuum.

sagila · November 2, 2012, 04:26

Thank you so much for your reply.
Actually I need to model s2s radiation in a vacuum room. Could you please explain more because I am not familiar with the “mean free path” and “physical length scale”?
Thank you.

RodriguezFatz · November 2, 2012, 04:37

I am not familiar with the S2S model, but if you want to simulate any flow, you need to ensure, that fluid equations are suitable for your problem.

Mean free path is the average distance that a neutral gas atom/molecule travels, before it collides with another atom/molecule.
Physical length scale means the typical dimension of your vacuum chamber. If you have a diameter of 0.1m and a height of 0.2m, it can be about 0.1 to 0.2m.
Now, if you divide these two numbers you get an estimate of how many collisions occur, if a particle moves from one side of your volume to another.
If there is just a small number of collisions (high Knudsen number), the normal fluid equations become incorrect, so you can not use Fluent.

Edit: Which gas and pressure?

sagila · November 2, 2012, 04:57

Thank you for your reply and complete explanation.
I need to simulate just radiation from surface to surface inside a room with the width of 5 meter and the height of 3.5 meter. So, the heat needs to be transferred only via radiation from surface to surface.
For this reason, void cells are needed to be created in the room! Or omission of the air inside the room needs to be happened!

RodriguezFatz · November 2, 2012, 05:00

Ok, so there is no flow in your room? Then forget all i wrote

sagila · November 2, 2012, 05:05

Thank you again for your time

Could you please let me know what I could do to solve this problem

RodriguezFatz · November 2, 2012, 05:14

As I wrote, I am not familiar with the S2S model.
Did you have a look at the Fluent manual? The S2S section seems to be pretty large...

sagila · November 2, 2012, 05:28

I have checked Fluent manual and I’m sure that s2s model is suitable for my problem.
(It says that s2s model is suitable for enclosure radiative transfer with non-participating media.)
I think I just need to create vacuum or omit the convective heat transfer inside the room, but unfortunately I don’t know how!
I would appreciate if you could help me.

Benfa · November 2, 2012, 05:42

Hi,

You should define the inner volume of the chamber as air and set the heat conductivity of air to a very low nummber (for example 1e-5) so that no heat conduction in air will occur (like in vacuum). Do not use gravity because it couldlead to natural convection effect. Then you just need to specifie all inputparameters for the s2s model and the boundarys (thats something you need to figure out!)

sagila · November 2, 2012, 06:00

Hi Benfa,
Thank you very much for your reply and solution.
I’ve done the same as your advice. But, my energy residual doesn’t decrease to 10e-6 (it’s about 0.1). Also my net imbalance for “total heat transfer rate” and “radiation heat transfer rate” is not about 1% of the lowest one. Therefore, I’m not sure whether it has converged on not! And, I don't know how to double-check whether this solution is true or not!

Benfa · November 2, 2012, 07:10

Hi,

from my experience 10e-6 as residual is not necessary but 0.1 is for sure not good enough.

1. Check your boundaries again (by the way what are the BC: heat flux, Temperature,...?)
2. Create Monitorpoints that plot you the wall temperature over the iteration number and see how it develops
3. Plot the residuals in CFD-Post and refine the mesh in the crictical regions
4. try a transient calculation (but I do not expect any influence)

If you specifie heat flux or power you will have an steady increase in temperatures till infinity. If you use such bc you need to specifie heat sinks to reach a steady state of temperature over time.

sagila · November 2, 2012, 07:32

Thank you again for your reply.

B.C. for three surfaces: WALL type “temperature” = constant
B.C. for one surface: WALL type “heat flux” = constant

Could you please explain more about how I could specify heat sink to reach a steady state of temperature over time?

Benfa · November 2, 2012, 08:27

Just for testing I would propose to set the bc of the heat flux to temperature too. But use a higher or lower temperature so that heat radiation can occur. With this setup you should at least see if you get a converged solution. If yes you need to think about the bc for the other unheated wall bc. I think there is a bc where you can set the outer temperatur, the thickness of the wall and the heat transfer coefficient. During your heating with the heat flux the other walls will get the radiated heat but will also loose heat because of the sourinding air. Hope this helps. For solving your problem you definitly need to think more about the happening physics and think about solution strategies.

November 2, 2012, 02:18	Modeling a vacuum chamber in Fluent	#1
sagila Member sagila Join Date: May 2012 Posts: 39 Rep Power: 14	Hello All, How can a vacuum chamber be modeled in Fluent? Thank you.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
radiation modeling in vacuum	mirko	OpenFOAM Programming & Development	7	May 20, 2016 09:31
chemical reaction modeling using FLUENT	wlt_1985	FLUENT	6	April 16, 2014 08:45
Modeling a vacuum chamber in Icepak	L093N	FLUENT	2	December 25, 2013 09:39
Multiphase flow modeling: CFX or FLUENT?	Paul	Main CFD Forum	8	January 24, 2012 10:25
Turbulence BCs for flow into vacuum chamber	eric.m.tridas	OpenFOAM Pre-Processing	0	December 20, 2011 14:18

November 2, 2012, 04:13		#2
RodriguezFatz Senior Member Philipp Join Date: Jun 2011 Location: Germany Posts: 1,297 Rep Power: 27	You first need to estimate the Kundsen number. That's the mean free path (for neutral-neutral collisions) divided by the physical length scale of your chamber. If it is much smaller than 1, you can use Fluent. If not, your gas is no continuum.

November 2, 2012, 04:26		#3
sagila Member sagila Join Date: May 2012 Posts: 39 Rep Power: 14	Thank you so much for your reply. Actually I need to model s2s radiation in a vacuum room. Could you please explain more because I am not familiar with the “mean free path” and “physical length scale”? Thank you.

November 2, 2012, 04:37		#4
RodriguezFatz Senior Member Philipp Join Date: Jun 2011 Location: Germany Posts: 1,297 Rep Power: 27	I am not familiar with the S2S model, but if you want to simulate any flow, you need to ensure, that fluid equations are suitable for your problem. Mean free path is the average distance that a neutral gas atom/molecule travels, before it collides with another atom/molecule. Physical length scale means the typical dimension of your vacuum chamber. If you have a diameter of 0.1m and a height of 0.2m, it can be about 0.1 to 0.2m. Now, if you divide these two numbers you get an estimate of how many collisions occur, if a particle moves from one side of your volume to another. If there is just a small number of collisions (high Knudsen number), the normal fluid equations become incorrect, so you can not use Fluent. Edit: Which gas and pressure?

November 2, 2012, 04:57		#5
sagila Member sagila Join Date: May 2012 Posts: 39 Rep Power: 14	Thank you for your reply and complete explanation. I need to simulate just radiation from surface to surface inside a room with the width of 5 meter and the height of 3.5 meter. So, the heat needs to be transferred only via radiation from surface to surface. For this reason, void cells are needed to be created in the room! Or omission of the air inside the room needs to be happened!

November 2, 2012, 05:00		#6
RodriguezFatz Senior Member Philipp Join Date: Jun 2011 Location: Germany Posts: 1,297 Rep Power: 27	Ok, so there is no flow in your room? Then forget all i wrote

November 2, 2012, 05:05		#7
sagila Member sagila Join Date: May 2012 Posts: 39 Rep Power: 14	Thank you again for your time Could you please let me know what I could do to solve this problem

November 2, 2012, 05:14		#8
RodriguezFatz Senior Member Philipp Join Date: Jun 2011 Location: Germany Posts: 1,297 Rep Power: 27	As I wrote, I am not familiar with the S2S model. Did you have a look at the Fluent manual? The S2S section seems to be pretty large...

November 2, 2012, 05:28		#9
sagila Member sagila Join Date: May 2012 Posts: 39 Rep Power: 14	I have checked Fluent manual and I’m sure that s2s model is suitable for my problem. (It says that s2s model is suitable for enclosure radiative transfer with non-participating media.) I think I just need to create vacuum or omit the convective heat transfer inside the room, but unfortunately I don’t know how! I would appreciate if you could help me.

November 2, 2012, 05:42		#10
Benfa Member Benny Join Date: Apr 2012 Posts: 40 Rep Power: 14	Hi, You should define the inner volume of the chamber as air and set the heat conductivity of air to a very low nummber (for example 1e-5) so that no heat conduction in air will occur (like in vacuum). Do not use gravity because it couldlead to natural convection effect. Then you just need to specifie all inputparameters for the s2s model and the boundarys (thats something you need to figure out!)

November 2, 2012, 06:00		#11
sagila Member sagila Join Date: May 2012 Posts: 39 Rep Power: 14	Hi Benfa, Thank you very much for your reply and solution. I’ve done the same as your advice. But, my energy residual doesn’t decrease to 10e-6 (it’s about 0.1). Also my net imbalance for “total heat transfer rate” and “radiation heat transfer rate” is not about 1% of the lowest one. Therefore, I’m not sure whether it has converged on not! And, I don't know how to double-check whether this solution is true or not!

November 2, 2012, 07:10		#12
Benfa Member Benny Join Date: Apr 2012 Posts: 40 Rep Power: 14	Hi, from my experience 10e-6 as residual is not necessary but 0.1 is for sure not good enough. 1. Check your boundaries again (by the way what are the BC: heat flux, Temperature,...?) 2. Create Monitorpoints that plot you the wall temperature over the iteration number and see how it develops 3. Plot the residuals in CFD-Post and refine the mesh in the crictical regions 4. try a transient calculation (but I do not expect any influence) If you specifie heat flux or power you will have an steady increase in temperatures till infinity. If you use such bc you need to specifie heat sinks to reach a steady state of temperature over time.

November 2, 2012, 07:32		#13
sagila Member sagila Join Date: May 2012 Posts: 39 Rep Power: 14	Thank you again for your reply. B.C. for three surfaces: WALL type “temperature” = constant B.C. for one surface: WALL type “heat flux” = constant Could you please explain more about how I could specify heat sink to reach a steady state of temperature over time?

November 2, 2012, 08:27		#14
Benfa Member Benny Join Date: Apr 2012 Posts: 40 Rep Power: 14	Just for testing I would propose to set the bc of the heat flux to temperature too. But use a higher or lower temperature so that heat radiation can occur. With this setup you should at least see if you get a converged solution. If yes you need to think about the bc for the other unheated wall bc. I think there is a bc where you can set the outer temperatur, the thickness of the wall and the heat transfer coefficient. During your heating with the heat flux the other walls will get the radiated heat but will also loose heat because of the sourinding air. Hope this helps. For solving your problem you definitly need to think more about the happening physics and think about solution strategies.