How to simulate a fan without the geometry of fan?

swtbkim · November 10, 2014, 02:59

Hi all.

I'm simulating a fan inside a pipe.
But it is sort of cumbersome to simulate a fan directly.
And I'm interested in a flow outside a pipe, not inside.
So I'd like not to use frozen rotor or something, but to simulate only air-sending function of a fan.

How can I do it?

I thought that momentum source is enough at first,
but I think it can not reflect the effect of pressure change of a fan because momentum source is defined as force per unit volume acting on the fluid.

How about using momentum source and porous media together?
Velocity and pressure decrease passing through porous media, and only velocity will increase passing through momentum source zone.

Any idea will be highly appreciated.
Thanks in advance.

swtbkim · November 10, 2014, 03:02

I forgot attaching an image.

swtbkim · November 10, 2014, 03:15

Oh one more thing lol

I heard that there is a function of "Fan modeler" or something.. I don't know exact name.

This maybe what I want: don't make fan geometry, but just designate specific zone as a fan.

Is there anyone who heard about this function?

ghorrocks · November 10, 2014, 04:13

The fan model might be a Fluent thing. You have to define it yourself in CFX.

There are two main options to implement this:
1) Use a momentum source (and no porous region is required)
2) Use an outlet/inlet to specify the flow.

1 is better if you want the air to recirculate. 2 is better if you know parameters definable by the boundary conditions, such as the inlet and outlet pressures or flow rates.

swtbkim · November 10, 2014, 04:56

Quote:

Originally Posted by ghorrocks

The fan model might be a Fluent thing. You have to define it yourself in CFX.

There are two main options to implement this:
1) Use a momentum source (and no porous region is required)
2) Use an outlet/inlet to specify the flow.

1 is better if you want the air to recirculate. 2 is better if you know parameters definable by the boundary conditions, such as the inlet and outlet pressures or flow rates.

Thank you, Glenn.
Your replies are always really helpful.

I think 2) is suitable.
Is this figure what you suggested?

Far field will be treated as an opening

ghorrocks · November 10, 2014, 16:04

Yes, that is what I was talking about.

Some minor points:
* There is no need to have such a long duct into and out of the inlet and outlet. A short duct is all that is required. For the inlet probably no duct at all is acceptable (so just flush with the end of the pipe).
* The mesh in the duct can also be coarser. You are not modelling this flow accurately so do not waste mesh elements in it. You need enough to get a basic flow profile, but don't worry about boundary layers or fine details like that.

swtbkim · November 11, 2014, 00:10

Thank you for your considerable advice.

1. The location of inlet should be over there for next calculation.
2. Thank you. It's a good idea.

And, can you explain about details of boundary conditions at inlet and outlet?

I tried as follows:
inlet : velocity inlet, 3m/s
outlet : pressure outlet, -100Pa
opening : opening, 0Pa

Is this nice?

ghorrocks · November 11, 2014, 00:18

Do you mean what is the meaning of inlet, outlet and opening; or do you mean the boundary conditions you have implemented in your model?

If it is your model please post an image of your whole domain showing where these boundary conditions are located.

November 10, 2014, 02:59	How to simulate a fan without the geometry of fan?	#1
swtbkim Senior Member Sangwoo Kim Join Date: Jul 2014 Location: Seoul, South Korea Posts: 115 Rep Power: 11	Hi all. I'm simulating a fan inside a pipe. But it is sort of cumbersome to simulate a fan directly. And I'm interested in a flow outside a pipe, not inside. So I'd like not to use frozen rotor or something, but to simulate only air-sending function of a fan. How can I do it? I thought that momentum source is enough at first, but I think it can not reflect the effect of pressure change of a fan because momentum source is defined as force per unit volume acting on the fluid. How about using momentum source and porous media together? Velocity and pressure decrease passing through porous media, and only velocity will increase passing through momentum source zone. Any idea will be highly appreciated. Thanks in advance. __________________ Best regards

November 10, 2014, 03:15		#3
swtbkim Senior Member Sangwoo Kim Join Date: Jul 2014 Location: Seoul, South Korea Posts: 115 Rep Power: 11	Oh one more thing lol I heard that there is a function of "Fan modeler" or something.. I don't know exact name. This maybe what I want: don't make fan geometry, but just designate specific zone as a fan. Is there anyone who heard about this function? __________________ Best regards

November 11, 2014, 00:10		#7
swtbkim Senior Member Sangwoo Kim Join Date: Jul 2014 Location: Seoul, South Korea Posts: 115 Rep Power: 11	Thank you for your considerable advice. 1. The location of inlet should be over there for next calculation. 2. Thank you. It's a good idea. And, can you explain about details of boundary conditions at inlet and outlet? I tried as follows: inlet : velocity inlet, 3m/s outlet : pressure outlet, -100Pa opening : opening, 0Pa Is this nice? __________________ Best regards

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
methodology to simulate fan performance analysis	ayothicfd	Main CFD Forum	4	July 3, 2013 20:56
How to simulate on geometry with curvature	seboxx	OpenFOAM	1	January 18, 2012 05:43
Simulation of Flow through Complex 3D Geometry	EmersonKB	CFX	5	July 2, 2009 08:17
How can I simulate a fan in a cavity	Elvia Martinez	CFX	1	September 7, 2005 09:49
Simulate geometry in different parts	Ralf Schmidt	FLUENT	0	July 29, 2005 10:08

November 10, 2014, 04:13		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	The fan model might be a Fluent thing. You have to define it yourself in CFX. There are two main options to implement this: 1) Use a momentum source (and no porous region is required) 2) Use an outlet/inlet to specify the flow. 1 is better if you want the air to recirculate. 2 is better if you know parameters definable by the boundary conditions, such as the inlet and outlet pressures or flow rates.

November 10, 2014, 16:04		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	Yes, that is what I was talking about. Some minor points: * There is no need to have such a long duct into and out of the inlet and outlet. A short duct is all that is required. For the inlet probably no duct at all is acceptable (so just flush with the end of the pipe). * The mesh in the duct can also be coarser. You are not modelling this flow accurately so do not waste mesh elements in it. You need enough to get a basic flow profile, but don't worry about boundary layers or fine details like that.

November 11, 2014, 00:18		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	Do you mean what is the meaning of inlet, outlet and opening; or do you mean the boundary conditions you have implemented in your model? If it is your model please post an image of your whole domain showing where these boundary conditions are located.