strainer in a flow duct

Marcopolopeich · January 28, 2012, 12:55

Hello to everyone,
I already saw something similar in the forum, but I think this could be an interesting question for a new topic anyway.
I should simulate, unfortunately very fast, some kind of strainer or sieve in a flow duct.
The modelling of such a sieve would just be a crazy thing (unless I have the possibility to use 20 millions nods and 20 months to meove the nodes in icem...).
Is there a possibility to give just the effect of this strainer, like an interface, a porous medium, a momentum source? And how accurate could be such a modelling strategy?
Any help and any suggestions in the discussion would be very appreciated.
Grüßi

ghorrocks · January 29, 2012, 04:58

Yes, a porous body or a momentum source term is the standard way of doing this sort of thing. If you know the resistance of the material well it is very accurate.

Marcopolopeich · January 29, 2012, 06:45

I am studying the topic .
I have very few information (only a foto) for the strainer and I have to set the conditions very carefully...

Is the tutorial with the catalytic converter a good reference?

ghorrocks · January 29, 2012, 17:45

Yes, the cat converter example is a good place to start.

But to do anything useful with your strainer you need pressure versus flowrate curves. If you do not have this you are just guessing. You can approximate the pressure versus flow rate curve from the geometry (assume the sieve is an oriface plate and use the resistance from that, or assume it is a bunch of cylinders in cross flow) but these approaches have significant error.

It all depends on how accurate you want to be and how important this seive is in the device.

Marcopolopeich · February 28, 2012, 14:03

I am going further with the simulation and I am slowly understanding the physics.
Unfortunately I have received from the constructor of this duct very few details and I have to search for some parameters.
Doyou have any idea where I can find suitable example?

evcelica · February 28, 2012, 16:03

If its a strainer is is usually a standard mesh size, perhaps you can find information about the mesh.

I don't know that much about your specific geometry but what I have done in the past is just model a single representative area of the strainer (one hole or a fraction of a hole using symmetric boundary conditions) and determine your pressure loss vs. flow velocity curve. I have found You can represent this curve pretty well using a true velocity porous membrane and inserting your volume porosity and a loss coefficient, which you can also make small simple model of before defining it in your larger model.

Marcopolopeich · February 29, 2012, 16:35

My problem is, I have only few gemetrical information about the mesh in the duct and no curve at all.
THe bst would be, just to see a similar case and use some suitable loss coefficient, just to get some simulation running. Then I can see if I can model the single hole.
I am searching for literature, but I am afraid I am searching for the wrong words.
Thx for your help.

ghorrocks · February 29, 2012, 17:10

If you do not know anything about the mesh then how are you expected to model it?

Surely you can get some basic information on it - wire diameter and spacing, open area, something like that. Even if you have a photograph of the mesh or have simply had a quick look at it you should be able to estimate those parameters. That can give you a starting point.

Marcopolopeich · March 2, 2012, 13:52

Hi glenn,
I have few data.
I am trying now to run a simulation with a sigle hole and using some translational periodicity or simmetric conditions.
I know what you mean with the accuracy of the modelling and the information about the geometry. This is the task that I received, I can't help about it.

What I meant, I can still run some simulation in order to get the pressure drop. But surely someone already modeled this case and has some numbers that I can use as validation for my modeling.
I am searching in my archives for the words Strainer duct and pressure loss but I get few answers.
Am I searching for th wrong words?

evcelica · March 2, 2012, 21:32

you could do an "orifice" calculation using your hole size, though I would trust my model results more than an orifice calc.

Marcopolopeich · March 3, 2012, 04:26

I am doing your model. Thank you again for your advice.
I am just trying to adapt some things. The value of the coefficient are pretty low, I am tring to understand if they are suitable.

January 28, 2012, 12:55	strainer in a flow duct	#1
Marcopolopeich New Member Hugo Heinz Peich Join Date: Oct 2010 Location: Switzerland Posts: 11 Rep Power: 15	Hello to everyone, I already saw something similar in the forum, but I think this could be an interesting question for a new topic anyway. I should simulate, unfortunately very fast, some kind of strainer or sieve in a flow duct. The modelling of such a sieve would just be a crazy thing (unless I have the possibility to use 20 millions nods and 20 months to meove the nodes in icem...). Is there a possibility to give just the effect of this strainer, like an interface, a porous medium, a momentum source? And how accurate could be such a modelling strategy? Any help and any suggestions in the discussion would be very appreciated. Grüßi

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
on fully developed duct flow	yfyap	Main CFD Forum	6	November 7, 2012 09:59
convergent duct water flow	kkiitm	CFX	0	January 2, 2011 11:11
Duct Flow Analysis	benbru	Siemens	1	September 26, 2007 15:53
Turbulence flow in a rectangular duct	Watchapon	Main CFD Forum	0	April 7, 2007 07:34
Question about flow in ducted fan in another duct	Jane	FLUENT	0	June 21, 2004 13:57

January 29, 2012, 04:58		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	Yes, a porous body or a momentum source term is the standard way of doing this sort of thing. If you know the resistance of the material well it is very accurate.

January 29, 2012, 06:45		#3
Marcopolopeich New Member Hugo Heinz Peich Join Date: Oct 2010 Location: Switzerland Posts: 11 Rep Power: 15	I am studying the topic . I have very few information (only a foto) for the strainer and I have to set the conditions very carefully... Is the tutorial with the catalytic converter a good reference?

January 29, 2012, 17:45		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	Yes, the cat converter example is a good place to start. But to do anything useful with your strainer you need pressure versus flowrate curves. If you do not have this you are just guessing. You can approximate the pressure versus flow rate curve from the geometry (assume the sieve is an oriface plate and use the resistance from that, or assume it is a bunch of cylinders in cross flow) but these approaches have significant error. It all depends on how accurate you want to be and how important this seive is in the device.

February 28, 2012, 14:03		#5
Marcopolopeich New Member Hugo Heinz Peich Join Date: Oct 2010 Location: Switzerland Posts: 11 Rep Power: 15	I am going further with the simulation and I am slowly understanding the physics. Unfortunately I have received from the constructor of this duct very few details and I have to search for some parameters. Doyou have any idea where I can find suitable example?

February 28, 2012, 16:03		#6
evcelica Senior Member Erik Join Date: Feb 2011 Location: Earth (Land portion) Posts: 1,167 Rep Power: 23	If its a strainer is is usually a standard mesh size, perhaps you can find information about the mesh. I don't know that much about your specific geometry but what I have done in the past is just model a single representative area of the strainer (one hole or a fraction of a hole using symmetric boundary conditions) and determine your pressure loss vs. flow velocity curve. I have found You can represent this curve pretty well using a true velocity porous membrane and inserting your volume porosity and a loss coefficient, which you can also make small simple model of before defining it in your larger model.

February 29, 2012, 16:35		#7
Marcopolopeich New Member Hugo Heinz Peich Join Date: Oct 2010 Location: Switzerland Posts: 11 Rep Power: 15	My problem is, I have only few gemetrical information about the mesh in the duct and no curve at all. THe bst would be, just to see a similar case and use some suitable loss coefficient, just to get some simulation running. Then I can see if I can model the single hole. I am searching for literature, but I am afraid I am searching for the wrong words. Thx for your help.

February 29, 2012, 17:10		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	If you do not know anything about the mesh then how are you expected to model it? Surely you can get some basic information on it - wire diameter and spacing, open area, something like that. Even if you have a photograph of the mesh or have simply had a quick look at it you should be able to estimate those parameters. That can give you a starting point.

March 2, 2012, 13:52		#9
Marcopolopeich New Member Hugo Heinz Peich Join Date: Oct 2010 Location: Switzerland Posts: 11 Rep Power: 15	Hi glenn, I have few data. I am trying now to run a simulation with a sigle hole and using some translational periodicity or simmetric conditions. I know what you mean with the accuracy of the modelling and the information about the geometry. This is the task that I received, I can't help about it. What I meant, I can still run some simulation in order to get the pressure drop. But surely someone already modeled this case and has some numbers that I can use as validation for my modeling. I am searching in my archives for the words Strainer duct and pressure loss but I get few answers. Am I searching for th wrong words?

March 2, 2012, 21:32		#10
evcelica Senior Member Erik Join Date: Feb 2011 Location: Earth (Land portion) Posts: 1,167 Rep Power: 23	you could do an "orifice" calculation using your hole size, though I would trust my model results more than an orifice calc.

March 3, 2012, 04:26		#11
Marcopolopeich New Member Hugo Heinz Peich Join Date: Oct 2010 Location: Switzerland Posts: 11 Rep Power: 15	I am doing your model. Thank you again for your advice. I am just trying to adapt some things. The value of the coefficient are pretty low, I am tring to understand if they are suitable.