CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   Need help on perforated plate with less than 2% open area (https://www.cfd-online.com/Forums/fluent/91968-need-help-perforated-plate-less-than-2-open-area.html)

 sosososo1114 August 28, 2011 23:22

Need help on perforated plate with less than 2% open area

HI, dear all,

I have problems to decide the effect of perforated plate on the velocity component parallel to it and on elimination of turbulence.

recently I am working on a project requiring simulating the down stream air flow which flows cross the perforated plate with less than 2% open area. That is to say, the hole one the plate has a less than 2mm diameter while each hole space with each other at 2.54cm. I know that usually screen or thin perforated plate will be simulated as porous jump in FLUENT.

However, it seems porous jump seems not affect velocity component (v ) parallel with the plate obviously. As this perforated plate is installed in a diffusor with large turbulent up stream flow and its open area is rare, I quite doubt weather the down stream flow which flows in a rectangle chamber will have an obviously velocity in the direction which parallels with the plat as its upstream flow does. So would it work to simulate this plate as a porous jump? and Will this plate guide the flow into a verticle parallel flow ( the plate is located horizontally)?

Could you offer me any hints. I will appreciate them.

Best regards,
Hongjin

 -mAx- August 29, 2011 01:56

Why don't you simply import your perforated plate, mesh it, and compute it?

http://www.cfd-online.com/Forums/mem...fice-plate.jpg

 sosososo1114 August 29, 2011 11:05

Quote:
 Originally Posted by -mAx- (Post 322006) Why don't you simply import your perforated plate, mesh it, and compute it? http://www.cfd-online.com/Forums/mem...fice-plate.jpg
Thanks Max, I would like to try. But I just have no idea on how to mesh the plate if I directly implant it into my model. the thickness of the plate is just 2mm while the height of the chamber where the plate is installed in is 0.8m, so what kind of mesh function should I use? Could you give me some suggestions on details? And how about the memory size and computation time? I will really appreciate it.

 -mAx- August 29, 2011 11:45

 sosososo1114 August 29, 2011 12:19

Quote:
 Originally Posted by -mAx- (Post 322116) I can help you if you are working with Gambit
Yes, sir, Please. I am currently meshing with Gambit, but I am not quite sure what kind of interval size I should use and what the ratio of mesh size interval is appropriated for a quantitative analysis. I will really appreciate your help.

 -mAx- August 29, 2011 12:54

Post a picture to see how it looks like

 sosososo1114 August 29, 2011 23:30

1 Attachment(s)
Quote:
 Originally Posted by -mAx- (Post 322125) is your perforated plate already modelised? Post a picture to see how it looks like
Yes, Max, thx , the picture below is the symmetric half of the model, and the points you see are in fact holes on the perforated plate

 -mAx- August 30, 2011 01:40

ok you don't have so many holes, so it will be quite easy.
For your plate, you only need to isolate the small cylinders with splits.
For this do following:
*create surface from top of cylinder with wireframe by picking the edge circle
*create surface from bottom of cylinder with wireframe by picking the edge circle
Now you can isolate your cylinder by splitting it with the 2 surfaces (split volume with surface >> select the 2 surfaces you just created).
If it is successful, you will be able to select the cylinder, and mesh it with hexa.
That's the goal for this stuff: isolating and meshing the orifices fine sufficient for getting flow rate inside each hole.
Then once all your holes are meshed, you may apply a size function from those volumes, and let grow your mesh from them.
If you pay attention to my picture you can see the hexa mesh inside the holes, and the growing mesh from them.
In my case I had some x-thousand holes, and you are quick limited from your hardware... :)

 sosososo1114 August 31, 2011 00:28

Quote:
 Originally Posted by -mAx- (Post 322186) ok you don't have so many holes, so it will be quite easy. I assume you already extracted the fluid domain from your geometry. For your plate, you only need to isolate the small cylinders with splits. For this do following: *create surface from top of cylinder with wireframe by picking the edge circle *create surface from bottom of cylinder with wireframe by picking the edge circle Now you can isolate your cylinder by splitting it with the 2 surfaces (split volume with surface >> select the 2 surfaces you just created). If it is successful, you will be able to select the cylinder, and mesh it with hexa. That's the goal for this stuff: isolating and meshing the orifices fine sufficient for getting flow rate inside each hole. Then once all your holes are meshed, you may apply a size function from those volumes, and let grow your mesh from them. If you pay attention to my picture you can see the hexa mesh inside the holes, and the growing mesh from them. In my case I had some x-thousand holes, and you are quick limited from your hardware... :)
Thanks Max, your suggestions help a lot on processing meshing, but still I get problems with meshing. The gambit says I have an entity unmeshed when I export mesh, but I used the filter found no unmeshed entity, so have you ever encountered such conditions?

 -mAx- August 31, 2011 01:33

use filters and check unmeshed volumes, unmeshed faces and unmeshed edges

 All times are GMT -4. The time now is 00:55.