Modelling an air filter

siw · October 2, 2009, 03:57

Hi,

I need to model air flow through a pipe and then into a radial filter and finally out through an exit pipe. Luckly, CFX has a tutorial for a porous domain modelling, which I shall run next. However, my filter is not a catalytic converter so the numerical data in the tutorial would be different to my case.

Does anyone have experience in modelling air filters, my past CFD work has always been on external aerodynamics, so some pointers would be useful. I doubt relative data will be available from the filter manufacturer - such as volume ratio as mentioned in the CFX Solver Theory guide.

Thanks,

WalterW · October 2, 2009, 08:39

I do have some experience with flat filter panels.

Usually, we measure the pressure drop of the cartridge in a rectangular or circular duct as a function of mass flow. It is then relatively safe to assume a uniform velocity distribution on the filter, allowing us to convert the (pressure-drop/mass flow) curve into a (pressure drop / mm filter height / velocity) curve, which usually looks like a 2nd grade polynom: Voilá, there are your linear and quadratic isotropic loss coefficients.

Something similar should work for radial filter cartridges, too.

siw · October 3, 2009, 06:18

Great, thanks for the information. I have a few extra questions now that I have finished the Catalytic Converter tutorial.

Do you always set the Reference Pressure to the required value and keep the Relative Pressure zero? I found that for my external aerodynamics work this produced wrong answers and had to set the Ref. Press. to zero and the Rel. Press. to a value.

In the Porosity Settings Loss Model do you use the Superficial or the Tru Velocity Loss Velocity Type? The tutorial uses the Superficial but the guides don't say much about them.

Cheers.

WalterW · October 5, 2009, 08:54

Whoa.

Is the "Porous Domain" Domain type new to CFX 12?
I was taught to use subdomains to model a porous medium... and there is no option for superficial velocity and the like.

ckleanth · October 5, 2009, 10:21

No the feature was available in v11. if you use a momentum loss in a fluid domain you have to use superficial velocity, for full porous domain model you can chose the velocity (superficial or true velocity) that your experimental values are based on.

ljacobs12 · September 13, 2011, 13:48

I don't know if this is an OK place to post my question because I am not sure what modeling an air filter is. To be honest, I don't know anything about air filters and I have really dumb question to prove it. Let me start out by saying that I am a girl and I know nothing about cars. For the past month my husband has been talking about how his car needs a new air filter and I'm going to surprise him by buying one and installing it in his truck! Step one was to start searching for where to buy them. I looked online and found this website that sells discount filters. I have no idea if these filters are the same kind of filters that go in vehicles. Can someone tell me if there is a difference between house air filters and car air filters? Does that place have the car kind? My second question is how to install them into the truck? Seriously, thank you so much for any help! Try to hurry with your answers, I want to have this done by this weekend!

October 2, 2009, 03:57	Modelling an air filter	#1
siw Senior Member Stuart Join Date: Jul 2009 Location: Portsmouth, England Posts: 742 Rep Power: 26	Hi, I need to model air flow through a pipe and then into a radial filter and finally out through an exit pipe. Luckly, CFX has a tutorial for a porous domain modelling, which I shall run next. However, my filter is not a catalytic converter so the numerical data in the tutorial would be different to my case. Does anyone have experience in modelling air filters, my past CFD work has always been on external aerodynamics, so some pointers would be useful. I doubt relative data will be available from the filter manufacturer - such as volume ratio as mentioned in the CFX Solver Theory guide. Thanks,

October 5, 2009, 10:21		#5
ckleanth Senior Member George Join Date: Mar 2009 Location: Birmingham, UK Posts: 257 Rep Power: 18	No the feature was available in v11. if you use a momentum loss in a fluid domain you have to use superficial velocity, for full porous domain model you can chose the velocity (superficial or true velocity) that your experimental values are based on. __________________ Top 4 tips 1. Knowledge is everything and Ignorance is dangerous. 2. Understand your limitations and try to eliminate them. 3. Get yerself a bike and hoon the chuffer. You will soon learn why dogs like to hang their heads out the car window. 4. Please before asking any questions on how to run simulations in CFX, go though all the tutorials

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
mass flow in is not equal to mass flow out	saii	CFX	12	March 19, 2018 06:21
air bubble is disappear increasing time using vof	xujjun	CFX	9	June 9, 2009 08:59
Air bearing modelling	Dave Larkham	CFX	1	October 19, 2008 20:10
Filter modelling	Armando	FLUENT	0	June 27, 2006 10:43
Air filter modeling	Mohamad Tajali	Main CFD Forum	1	December 5, 2005 11:16

October 2, 2009, 08:39		#2
WalterW New Member Join Date: Oct 2009 Posts: 11 Rep Power: 17	I do have some experience with flat filter panels. Usually, we measure the pressure drop of the cartridge in a rectangular or circular duct as a function of mass flow. It is then relatively safe to assume a uniform velocity distribution on the filter, allowing us to convert the (pressure-drop/mass flow) curve into a (pressure drop / mm filter height / velocity) curve, which usually looks like a 2nd grade polynom: Voilá, there are your linear and quadratic isotropic loss coefficients. Something similar should work for radial filter cartridges, too.

October 3, 2009, 06:18		#3
siw Senior Member Stuart Join Date: Jul 2009 Location: Portsmouth, England Posts: 742 Rep Power: 26	Great, thanks for the information. I have a few extra questions now that I have finished the Catalytic Converter tutorial. Do you always set the Reference Pressure to the required value and keep the Relative Pressure zero? I found that for my external aerodynamics work this produced wrong answers and had to set the Ref. Press. to zero and the Rel. Press. to a value. In the Porosity Settings Loss Model do you use the Superficial or the Tru Velocity Loss Velocity Type? The tutorial uses the Superficial but the guides don't say much about them. Cheers.

October 5, 2009, 08:54		#4
WalterW New Member Join Date: Oct 2009 Posts: 11 Rep Power: 17	Whoa. Is the "Porous Domain" Domain type new to CFX 12? I was taught to use subdomains to model a porous medium... and there is no option for superficial velocity and the like.

September 13, 2011, 13:48		#6
ljacobs12 New Member Laura Jacobs Join Date: Sep 2011 Posts: 2 Rep Power: 0	I don't know if this is an OK place to post my question because I am not sure what modeling an air filter is. To be honest, I don't know anything about air filters and I have really dumb question to prove it. Let me start out by saying that I am a girl and I know nothing about cars. For the past month my husband has been talking about how his car needs a new air filter and I'm going to surprise him by buying one and installing it in his truck! Step one was to start searching for where to buy them. I looked online and found this website that sells discount filters. I have no idea if these filters are the same kind of filters that go in vehicles. Can someone tell me if there is a difference between house air filters and car air filters? Does that place have the car kind? My second question is how to install them into the truck? Seriously, thank you so much for any help! Try to hurry with your answers, I want to have this done by this weekend!