CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   CFX (https://www.cfd-online.com/Forums/cfx/)
-   -   Modeling a compartment with circulation (by a fan) (https://www.cfd-online.com/Forums/cfx/68903-modeling-compartment-circulation-fan.html)

neilpan October 5, 2009 21:33

Modeling a compartment with circulation (by a fan)
 
Hi guys,

I am working on a hollow compartment (Its essentially a steel box made from 2 mm sheet metal). I have 1 inlet and 1 outlet and a heat source inside the compartment. The circulation inside the compartment is done via another fan which I have no clue on how to model. Should I just create a surface (its a 120mm x 120mm fan) and impose flow conditions? If so, what specific conditions do I need to impose?

I am really new to CFX and I have gone through a bunch of tutorials supplied with the software (heat-coil one is very useful and applicable to my situation).

Kindly advise.

Thanks!

ghorrocks October 6, 2009 17:42

You say you have no clue how to model the second fan. Is that because you don't know how to model fans in CFX or because you know nothing about the fan's performance? If it is the former we can help you. If it is the latter we cannot help you - obviously you need to know the conditions you are simulating before you start.

Assuming you have a fan performance curve or some other fan performance data then you can put this as a boundary in the model with the performance curve (eg flow rate as a function of average pressure at the fan). Alternately if the fan is inside the domain somehow you can use a momentum source to move the fluid, again using the fan performance curve.

As you can see it all comes from the fan performance curve. Do you have that? Or at least a fan performance point?

neilpan October 6, 2009 17:49

Thanks for your reply. I don't think I have a fan performance curve but I do have the fan air flow in CFM (cubic feet per min) which is 75 CFM through an annular 120mm area. Would that be enough to model? Please enlighten.

ckleanth October 6, 2009 18:52

no because thats a mass flow rate with no reference to a change of pressure; you need to have the fan characterisic so that your system characteristic will match to a flow condition on the fan curve

ghorrocks October 7, 2009 06:18

Well the simplest approach is to assume the fan's flow rate is 75 CFM regardless of the pressure difference. This is obviously a big simplification but until you get a fan performance curve that's all you can do.

So if the fan is outside the region of interest then you can simply model this constant flow fan as a specified flow rate inlet or outlet. If the fan is inside the region of interest (ie you want to model both the incoming and outgoing flow of the fan) then you will need a momentum source.

neilpan October 7, 2009 18:23

Thanks guys! Well, the fan is inside the region of interest. So, could you please exemplify as to what a "momentum source" means? I am also going to look it up in the help.

flattie October 8, 2009 17:12

Quote:

Originally Posted by ghorrocks (Post 231739)
Well the simplest approach is to assume the fan's flow rate is 75 CFM regardless of the pressure difference. This is obviously a big simplification but until you get a fan performance curve that's all you can do...

I find that sometimes this may also be needed to get things started even when you have a performance curve. I had a sim where I needed an initial guess on the fan flow rate even with the curve, else the flow went to zero over time.

neilpan November 11, 2009 00:14

Quote:

Originally Posted by ghorrocks (Post 231739)
Well the simplest approach is to assume the fan's flow rate is 75 CFM regardless of the pressure difference. This is obviously a big simplification but until you get a fan performance curve that's all you can do.

So if the fan is outside the region of interest then you can simply model this constant flow fan as a specified flow rate inlet or outlet. If the fan is inside the region of interest (ie you want to model both the incoming and outgoing flow of the fan) then you will need a momentum source.


Hi guys,

So, I did find a fan curve. Volumetric flow-rate versus static pressure.
http://img694.imageshack.us/img694/3417/77751288.th.png
http://img21.imageshack.us/img21/469...rafikgr.th.png

Now, given that, what is the approach for modeling a fan inside the area of interest? Do I need to model the inlet and outlet conditions using the fan curve? I am sorry if I sound like an idiot but I am really new at CFD analysis. I see that there is an option for momentum source when I create a subdomain. My question is how do I tell CFX that my flow rate changes as the backpressure increases (characterized by the fan curve)? I know it involves creating an expression.


All times are GMT -4. The time now is 20:37.