[JulienV]

Hi everyone,

I'm new on this forum, but i saw that simulations about perforated plates had been discussed already but i didnt find solution to my problem.
I am working on CFX on part-time, so i didnt explore most of possible options...

Let me explain... (and sorry if my english is approximate sometimes...)

I want to simulate a thin perforated plate in a HVAC application.
I would like to use perforate plates as inlet, outlet and as "connexion" between two domains (representing 2 different rooms for example).

I am looking for the good method to obtain the right flow and the right velocity on the total surface

For example,
1x1m plate, 50% open surface.
A flow through the plate of 1m3/s.

The resultant velocity at the plate outlet would be 2m/s.

Is it possible to obtain this velocity with a 1m² inlet/outlet in my model ?


I tried to simulate the plate as a porous domain between 2 rooms. But my first results show an incorrect air profile at the outlet of the porous domain.

I tried to understand how porous domains work with the CFX tutorial (catalytic converter)...

If my description is not accurate enough or hard to understand, dont hesitate to ask me more informations.

Waiting to read you...

Thanks !

Julien

[ghorrocks]

You need to have a think about what the porous material is doing. If the flow is 1m3/s over a 1m2 plate with a 50% open surface, then yes the local flow next to the open area is approximately 2m/s. But next to a closed area it is 0. What the porous model does is average this all together - so you get a net flow of the 2 m/s and 0m/s averaged over the whole area which is 1m/s.

In other words, when the open areas are small the little 2m/s jets of air flowing through them will quickly dissipate and become a bulk flow of around 1m/s.

Another thing - think about the boundary conditions you are applying. You are saying that you have a porous plate at the inlet and outlet. You know the flow rate at the inlet - so why not just use a mass flow rate boundary at the inlet? The inlet porous plate does not contribute to anything in your model (other than lowering the pressure a little due to pressure drop) so why model it?

[JulienV]

Thanks for your answer Glenn,

Originally Posted by ghorrocks

You need to have a think about what the porous material is doing. If the flow is 1m3/s over a 1m2 plate with a 50% open surface, then yes the local flow next to the open area is approximately 2m/s. But next to a closed area it is 0. What the porous model does is average this all together - so you get a net flow of the 2 m/s and 0m/s averaged over the whole area which is 1m/s.

In other words, when the open areas are small the little 2m/s jets of air flowing through them will quickly dissipate and become a bulk flow of around 1m/s.

Now that you explained me that, i understand the results i had on my first tests.
But, we often use perforated plate in order to create a more homogeneous flow (or even a laminar flow for low velocity diffusion).
So, if i'm right, the porous domain modelling can't describe this phenomena accurately in the direct proximity of the grid, but it does accelerate the flow ?

Originally Posted by ghorrocks

Another thing - think about the boundary conditions you are applying. You are saying that you have a porous plate at the inlet and outlet. You know the flow rate at the inlet - so why not just use a mass flow rate boundary at the inlet? The inlet porous plate does not contribute to anything in your model (other than lowering the pressure a little due to pressure drop) so why model it?

I would like to consider the local acceleration of air due to the open surface %age.
So, if i want to consider the right velocity i have to set a higher mass flow than reality.

[ghorrocks]

But, we often use perforated plate in order to create a more homogeneous flow (or even a laminar flow for low velocity diffusion).

Then there is no need to model the inlet porous plate and just start your simulation with a homogeneous flow as the inlet.

You place boundary conditions at locations where you know the flow conditions. If you cannot define precisely what the flow is doing at that location you cannot impose a boundary condition there. It sounds like you know the flow downstream of the inlet plate better than at the inlet plate - so put the inlet boundary downstream where the flow is simpler and you can accurately define it.

I would like to consider the local acceleration of air due to the open surface %age.
So, if i want to consider the right velocity i have to set a higher mass flow than reality.

The local acceleration of the air will quickly dissipate to become a bulk flow, as you mentioned.

If you increase the flow velocity then you will increase the flow rate!

If the details of the air jets coming out of the plate are important then you should forget a porous flow model and directly model the plate with its holes. Then you will have the air jets which dissipate downstream.

[JulienV]

I agree, modelling this perforated plate as a simple inlet of 1x1m dimension and 1m/s homogeneous velocity would provide good results far enough from the inlet.
All i would have to do is verify that this distance is acceptable for my application. This have to be done by modelling the holes or with experimental data.

Did you already work on that for HVAC applications ?
I saw that you are very active on most of threads i read on this subject...

If you increase the flow velocity then you will increase the flow rate!

Of course but that is what bugs me in this case ^^


Anyway, i surely was wrong, i thought that this kind of application was common, but it's not as easy actually...
I will continue to make some test and reading documentation to try to fully understand the porous domain assumptions.


Thank you for you help Glenn.

[ghorrocks]

I have never done HVAC work. I just know a bit about fluid mechanics.

[Mandar saraf]

i want to simulate spray dryer, in that there are two inlet one of liquid spray and one of air.. which is through prefforated(air), this prefforated plate is 3mm thick and having holes of 3mm dia, this hole are arrange in traingular pitch of 5mm, this plate is made up of steel, i trie to model this but its very expensive, please tell me, wether it is possible to solve this problem using porous media?? if so please trell me what parameters required to have equivalent porous media as of this sheet,
dia of sheet is 1290mm, and mass flow rate of air is 2.12kg/s
please reply,
thanks,

[ghorrocks]

Yes, you can use the porous model to model this. You could also use a momentum source term (which is what the porous model is anyway).

[Mandar saraf]

thanks for your reply, can you please tell me how to start with this ? i have done catalatic convertor and spray dryer tout of cfx, for my model i use this two concept but still m stuck. even iam not sure wether iam applying right physic, am attaching some details of my model. if possible please tell me how can i start with this problem.
I iam creating 3 serprate domain for this upper one is air distributar, middle one is preforated plate with porous media and bottom one is chamber,
waiting for reply,
thank you,

[ghorrocks]

I would recommend you do a simple model to work out the porous region. Maybe a simple pipe flow with a porous bit in the middle. Then you can easily vary the flow rates and pressures and check that you are getting the pressure drop you expect in the porous region. And the model should take second to run rather than hours.

Once a simple model is behaving as expected then you can transfer the setup to your full model.

[ghorrocks]

Please do not post multiple posts again. I have removed the other duplicate posts.

[Mandar saraf]

Originally Posted by ghorrocks

Please do not post multiple posts again. I have removed the other duplicate posts.

ya sure sorry for the mistake