[jins9158]

I used pressure change interface to consider pressure difference about thin filter instead of using porous model.

First, I made 1D function by using userfucntion (reference image1) and I made Interface between 2 domain.

Finally I applied 1d function's about pressure drop in surface of interface between 2 domains and I simulated.

When I reviewed, I knew that the result is wrong!

Inlet velocity is 3m/s(reference image3) and It means that interface pressure of difference should be about 70 ~ 80Pa.
(Inlet velocity 2.8m/s > -57[Pa], Inlet velocity 3.5m/s > -81[Pa])

But interface pressure of difference was very small (Pressure is about 3Pa.. reference image3)

I reviewed why the result is wrong for days. But I couldn't find pressure change interface information in google, customer portal, ansys cfx example, you-tube, CFD online and so on.

So, I really need to help.

Please give me advice

Thank you


1. image1 > cfx pre image & userfunction image
image1.jpg

2. image2 > interface pressure change
image2.jpg

3. image3 > Velocity & Pressure result image
image3.jpg

4. output file
CFX_001.txt

[ghorrocks]

Your links do not work. But please do not post images on third party sites, post them directly on the forum: https://www.cfd-online.com/Wiki/Ansy...n_the_forum.3F

To answer your question: Please post your output file.

[jins9158]

Originally Posted by ghorrocks

Your links do not work. But please do not post images on third party sites, post them directly on the forum: https://www.cfd-online.com/Wiki/Ansy...n_the_forum.3F

To answer your question: Please post your output file.

I posted images again according to your adivce.

I am sorry to bother you

[ghorrocks]

Please post your output file.

[jins9158]

Originally Posted by ghorrocks

Please post your output file.

I am sorry

I uploaded my outfile

Thank you for your concern~

[ghorrocks]

You have defined a positive pressure change across the filter. The pressure is higher on the downstream side of the filter than the upstream side. If this is what you intend then you might have invented a perpetual motion machine.

But I suspect you did not want to do this. The pressure drop across the interface is defined as a positive number to give you a pressure drop in the direction of flow.

[jins9158]

Originally Posted by ghorrocks

You have defined a positive pressure change across the filter. The pressure is higher on the downstream side of the filter than the upstream side. If this is what you intend then you might have invented a perpetual motion machine.

But I suspect you did not want to do this. The pressure drop across the interface is defined as a positive number to give you a pressure drop in the direction of flow.

Thank you for relpying my question

I have some questions.

Is the local coordinate also applyed in this matter? When I tested, I realized that local coordinate was not applyed.

I need to simulate about ventilation system that have several filters

So I need to use local coordinate to apply difference of pressure.

[ghorrocks]

I have never used the local coordinate option with that. You will have to test it and work it out, or read the documentation.

[jins9158]

Originally Posted by ghorrocks

I have never used the local coordinate option with that. You will have to test it and work it out, or read the documentation.

I reviewed and I thought that the local coordinate not work

Thank you for relying my question

[Gert-Jan]

If I'm correct you can use this interface with pressure specification only in a translational periodic setup. Imagine you have pipe with 100 obstacles in line with fixed distance, then you model a section with 1 obstacle. Then add a periodic interface, allowing you to couple the outlet with the inlet.
Then you can choose a pressure drop over the 1/100-th pipe section, which results in a massflow. Or you specifcy a massflow and you get a pressure drop.

[jins9158]

Originally Posted by ghorrocks

You have defined a positive pressure change across the filter. The pressure is higher on the downstream side of the filter than the upstream side. If this is what you intend then you might have invented a perpetual motion machine.

But I suspect you did not want to do this. The pressure drop across the interface is defined as a positive number to give you a pressure drop in the direction of flow.

could I give some question?

you said that the pressure is higher on the downstream side of the filter than the upstream side.

This answer made me to confuse becasue the porous domain's result about tutorial is different.

When I used porous domain to appy pressure's drop, the pressure is higher on the upstream side of the porous domain than the downstream side.

Aren't they option about pressure loss in both?

why are they different?

Please give me advice

[ghorrocks]

I think the pressure is higher downstream of the interface in this case because you defined a negative pressure jump over the interface. If you made it a positive number you would get lower pressure downstream which I suspect is what you want to do.

[evcelica]

I'm confused, you only use "filter" but do not define the independent variable "velocity".
shouldn't you have to put in filter(Velocity) or filter(Velocity w) or whatever the correct variable is?
Or does it just assume velocity is the independent variable?