Simulation of air flow in a chamber

Issa · November 22, 2009, 12:08

Hi,

I would like to inform you that I'm using CFX to simulate an air flow in a chamber, the air inters the chamber through a (0.5 m X 0.5 m) flat faced displacement ventilation (DV) diffuser (photo attached). The flat faced DV diffuser is consist of 3944 holes, each hole is 2 mm diameter and these holes represent only 5% of the total area of DV diffuser. The air velocity measured at the inlet is 0.1 m/s, (so the flow rate = 0.1 X0.5X0.5X5%). I think it is very difficult to draw this DV diffuser with 3944 holes. So, instead I reduced the air velocity which comes out from the diffuser from 0.1 m/s to (0.1*5%=0.005 m/s). But unfortunately the predicted air velocity and air temperature is different than the measured ones. Then I tried to keep the air velocity at 0.1 m/s but in this case I reduced the inlet area from (0.5 m X 0.5 m) to (0.1118 X 0.1118 m) which is 5% of the total inlet area but still there is discrepancy between the predicted and measured air velocity and air tempertature. Hence, I will be very grateful if you advice me what to do to get an accurate prediction of the air flow ? Thank you

Regards,
Issa

ghorrocks · November 22, 2009, 17:53

By the way: The image was not attached.

It is a tricky thing to model. On a micro-scale, each little hole creates an air jet at a high velocity which then enters the main room. These jets will travel out for a little way, then they will start to diffuse and turbulent action will chop them up. After sufficient diffusion adjacent jets will start to merge and it will become just a bulk air flow.

I would recommend you consider putting the inlet at the point where the little jets have merged and become a simpler bulk air flow. Your simple approach fails to capture the details of this jetting and merging effects and that is why it is not matching results. So best not even attempt to model it and move the boundary to somewhere where the flow is better known.

Issa · November 23, 2009, 09:40

Dear Horrocks,

Thank you for your reply. However, actually I don't know where is the point where the little jets have merged and become a simpler bulk air flow. I think there is another way to simulate the displacement ventilation unit with many holes but I don't know how. So, if possible would you please recommend another solution? Thank you

Regards,
Issa

ghorrocks · November 23, 2009, 18:16

My first suggestion is not good enough? You are the one doing the analysis so you must decide.

But in general the concept is that you MUST put the boundary somewhere you know enough about the flow to describe it with the functions available. If you don't know the details of the flow at the grill then you have to go somewhere else to somewhere you do know.

November 22, 2009, 12:08	Simulation of air flow in a chamber	#1
Issa New Member Join Date: Nov 2009 Posts: 12 Rep Power: 17	Hi, I would like to inform you that I'm using CFX to simulate an air flow in a chamber, the air inters the chamber through a (0.5 m X 0.5 m) flat faced displacement ventilation (DV) diffuser (photo attached). The flat faced DV diffuser is consist of 3944 holes, each hole is 2 mm diameter and these holes represent only 5% of the total area of DV diffuser. The air velocity measured at the inlet is 0.1 m/s, (so the flow rate = 0.1 X0.5X0.5X5%). I think it is very difficult to draw this DV diffuser with 3944 holes. So, instead I reduced the air velocity which comes out from the diffuser from 0.1 m/s to (0.1*5%=0.005 m/s). But unfortunately the predicted air velocity and air temperature is different than the measured ones. Then I tried to keep the air velocity at 0.1 m/s but in this case I reduced the inlet area from (0.5 m X 0.5 m) to (0.1118 X 0.1118 m) which is 5% of the total inlet area but still there is discrepancy between the predicted and measured air velocity and air tempertature. Hence, I will be very grateful if you advice me what to do to get an accurate prediction of the air flow ? Thank you Regards, Issa

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Two-phase air water flow problems by activating Wall Lubrication Force	challenger85	CFX	5	November 5, 2009 06:44
calculate volume flow from a 2D simulation	SimonH.	OpenFOAM	0	October 27, 2009 05:39
Internal flow simulation boundary conditions	Kishore	FLUENT	1	July 10, 2007 12:42
moving water from a chamber using air	Matt Mason	CFX	0	December 2, 2003 20:46
simulation of mixing flow	willams	Main CFD Forum	6	October 18, 1999 01:46

November 22, 2009, 17:53		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	By the way: The image was not attached. It is a tricky thing to model. On a micro-scale, each little hole creates an air jet at a high velocity which then enters the main room. These jets will travel out for a little way, then they will start to diffuse and turbulent action will chop them up. After sufficient diffusion adjacent jets will start to merge and it will become just a bulk air flow. I would recommend you consider putting the inlet at the point where the little jets have merged and become a simpler bulk air flow. Your simple approach fails to capture the details of this jetting and merging effects and that is why it is not matching results. So best not even attempt to model it and move the boundary to somewhere where the flow is better known.

November 23, 2009, 09:40		#3
Issa New Member Join Date: Nov 2009 Posts: 12 Rep Power: 17	Dear Horrocks, Thank you for your reply. However, actually I don't know where is the point where the little jets have merged and become a simpler bulk air flow. I think there is another way to simulate the displacement ventilation unit with many holes but I don't know how. So, if possible would you please recommend another solution? Thank you Regards, Issa

November 23, 2009, 18:16		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	My first suggestion is not good enough? You are the one doing the analysis so you must decide. But in general the concept is that you MUST put the boundary somewhere you know enough about the flow to describe it with the functions available. If you don't know the details of the flow at the grill then you have to go somewhere else to somewhere you do know.