[Dani86]

Hi,

I'm studying mechanical engineering and I'm really new with the fantastic CFD world.
I'm modelling my first CFD model and I have a question (which may be silly but it's something that I want to understand).
I'm modelling a ventilation system for a room where there is a supply system and an extract system.
Each system has a dedicate fan. (please see attached a sketch).

With Solidwork Flow Simulation I'm modelling only the room with two openings (which represent the supply and extract grilles).
For the supply opening I have set up the flow rate= 100l/s
For the return I would like to do the same (because at the end what I know is that the flow rate for the inlet is = to the flow rate at the outlet. The same should be for the velocity right?)

Unfortunately I can't run this model because it seems that I can't define the inlet and outlet flow at the same time (see attached screenshot).
What I understand is that I should define the Inlet flow and the Outlet pressure if I want to run this model.
Can you explain me why?
If I have to set the Outlet pressure what should be the value? I can't use the environment pressure because the extract fan is creating a negative pressure and I don't know what is the value.
Can someone explain me how can I calculate the pressure at the extract grille?

Thanks for your answers.
I really appreciate your help!

[FMDenaro]

Sorry, I am not an expert of this software but if you are solving incompressible flow model, either velocity or pressure can be mathematical prescribed at the outflow. That is a general issue but I don't know if both possibilities are implemented.
Have you tried to set the environmental pressure at the exit to see if the run converges?

[Dani86]

Hi FMDenaro,

Thanks for your answer.
Yes, I have set an incompressible flow as the grade of compressibility for the flow should be not relevant (the air density shouldn't change too much).

If I set the boundary conditions as below, I can run the CFD:
Supply opening=> Flow Rate = 100 l/s
Extract Opening => Pressure = Environment Pressure = 101325 Pa

The question is:
Are the results corrects? I don't think so because the pressure that I see at the Extract opening is NOT the environmental pressure. There is an extract fan after the extract grille so there should be a negative pressure. (Please correct me if I'm wrong).


Does anyone have my same problem?
How can I analyse a CFD model where I have a supply system and a return system with dedicate supply and extract fans?

I have seen the following tutorial:
https://www.youtube.com/watch?v=kI02JEwwvyQ&t=713s

But unfortunately there is only a supply system so I don't understand how should I set up the boundaries conditions for the outlet because I don't know the value of the negative pressure).

Thanks for your help guys, really appreciated

[FMDenaro]

Originally Posted by Dani86

Hi FMDenaro,

Thanks for your answer.
Yes, I have set an incompressible flow as the grade of compressibility for the flow should be not relevant (the air density shouldn't change too much).

If I set the boundary conditions as below, I can run the CFD:
Supply opening=> Flow Rate = 100 l/s
Extract Opening => Pressure = Environment Pressure = 101325 Pa

The question is:
Are the results corrects? I don't think so because the pressure that I see at the Extract opening is NOT the environmental pressure. There is an extract fan after the extract grille so there should be a negative pressure. (Please correct me if I'm wrong).


Does anyone have my same problem?
How can I analyse a CFD model where I have a supply system and a return system with dedicate supply and extract fans?

I have seen the following tutorial:
https://www.youtube.com/watch?v=kI02JEwwvyQ&t=713s

But unfortunately there is only a supply system so I don't understand how should I set up the boundaries conditions for the outlet because I don't know the value of the negative pressure).

Thanks for your help guys, really appreciated

yes, of course the setting of the pressure environment is not physically congruent to the setting of your problem but I was just interested to see if the run had a convergent solution...unfortunately I don't know this software and I cannot give a correct help.
I can suggest to ask in dedicated forum.

A possible trick, that I don't know if can work, is to treat the outlet in a similar way as you did in the inlet, just reversing the sign of the velocity to allow flow to go out from the domain. But it is likely that the software will not accept such setup with only inlet conditions.

[Dani86]

Hi FMDenaro,
I don't think the problem is generated by the software. I'm doing some researches and I've found that for Autodesk CFD and (maybe) for Ansys Fluent I have to set the same conditions:
Flow rate (or velocity) for one grille
Pressure for the other grille.

I can't put a negative value for the velocity/flow rate. The software has an option where you can decide if the opening is an inlet or an outlet (so you have to put always a positive value).

I'm wondering if anyone has crashed with the same issue because I do believe someone has modelled a similar model previously. If you think about how we distribute ventilation in a big building there will be always a dedicate supply system and a dedicate extract system.

Thanks for who could give me a help. I'm really struggling to find a solution...

[LuckyTran]

If you want to set it to the right pressure. Then you need to know the pressure drop across the grill and head rise through the fan (the fan curve).

If you are using the incompressible model then it doesn't matter what you specify as the pressure, so set it to anything. Pressure doesn't mean anything in this case so it doesn't matter if it's less than or greater than the environmental pressure. Just recalibrate your thinking to interpret the results in terms of pressure drop and not the absolute pressure.

It is quite easy to see why an inlet outlet is not the correct boundary pair in general. It's exactly the problem you are having now! If I tell you there's 3 m/s velocity at the inlet and 3 m/s velocity at the outlet (or L/s if you like volumetric flow), you'd still never be able to find out what is the pressure and temperature anywhere. For the incompressible case, you (mathematically) don't care anymore what is the pressure and that's why that case is special.

[FMDenaro]

Originally Posted by LuckyTran

If you want to set it to the right pressure. Then you need to know the pressure drop across the grill and head rise through the fan (the fan curve).

If you are using the incompressible model then it doesn't matter what you specify as the pressure, so set it to anything. Pressure doesn't mean anything in this case so it doesn't matter if it's less than or greater than the environmental pressure. Just recalibrate your thinking to interpret the results in terms of pressure drop and not the absolute pressure.

It is quite easy to see why an inlet outlet is not the correct boundary pair in general. It's exactly the problem you are having now! If I tell you there's 3 m/s velocity at the inlet and 3 m/s velocity at the outlet (or L/s if you like volumetric flow), you'd still never be able to find out what is the pressure and temperature anywhere. For the incompressible case, you (mathematically) don't care anymore what is the pressure and that's why that case is special.

I don't understand clearly this issue, why in a room for an incompressble flow (divergence-free constraint + momentum equation) you could not set an inflow profile and an outflow profile? Of course physically you cannot prescribe the real profile (at least for real problems) but mathematically is just sufficient that both profiles have the same integral to satisfy the continuity. Pressure (absolute value) is never necessary.

Turning back to the original question, also the inflow should be a dynamic condition depending on the fan.

[Dani86]

Dear LuckyTran, FMDenaro

Thanks for your messages. I finally get to understand what is going on on my model. I'm sorry, I know that those questions for you may be very stupid but I really would like to do the correct things and have a better understand of CFD.

From my understanding if I have an incompressible flow then the air flow rate that is coming inside the room should be equal to the air flow rate that is going outside the room. Am I wrong to say that?

I think I can compare my room with a pump for water: what is coming inside the pump is going out of the pump right?
So it's easy to say that the flow rate that is going inside the pump is = to the flow rate that is going outside the pump. That is why I have used flow rates for my boundary conditions.


I haven't checked that with my model but if I put any pressure value at the outlet do you think that the air flow rate at the outlet will be the same as the one at the inlet?
last question:
I've found few articles that states that the Air can be considered incompressible as long flow velocities are less than 100m/s (Ma=0.3).
Is that correct? is there any influence with temperature/Humidity?

Many thanks,
Daniele

[FMDenaro]

The incompressible assumption M<0.3 is valid for the compression due to the acustic waves but you can have density variation with temperature while still considering the divergence-free constraint on the velocity.
If you consider the integral of this constrain in your volume and use Gauss, it writes

Int [S] n.v dS =0

Therefore the (volumetric) flow rate in inlet must be equal to the flow rate in outlet. A condition on the pressure in outlet still ensures such conservation by solving the pressure equation with proper BC.s-

[LuckyTran]

Originally Posted by FMDenaro

I don't understand clearly this issue, why in a room for an incompressble flow (divergence-free constraint + momentum equation) you could not set an inflow profile and an outflow profile?

No you are right, I just didn't emphasize enough that I meant the general compressible case. For the incompressible case you should be able to. But unfortunately the software does not allow it in this case. And since we cannot hack the solver and have it do what we want it to do, I don't discuss it any further.

Originally Posted by Dani86

From my understanding if I have an incompressible flow then the air flow rate that is coming inside the room should be equal to the air flow rate that is going outside the room. Am I wrong to say that?

For the constant density case yes. But your software doesn't give that as an option so it doesn't help to think this way, although you should expect it in the result! Btw a temperature dependent density still gives a divergence free velocity (which we can also call incompressible).

Originally Posted by Dani86

I haven't checked that with my model but if I put any pressure value at the outlet do you think that the air flow rate at the outlet will be the same as the one at the inlet?

Yes. There is a continuity + momentum equation in Solidworks which is solved using FVM and automated mesher.