[eSKa]

Hello guys.

I have a tunnel with a fan running inside. I´m trying to imitate a natural wind trough the tunnel, it shall work as a basic velocity, and in addition to this velocity the fan will increase the pressure in the tunnel.
I´m working with two openings.

My idea is to put on every opening a static pressure, calculating it this way:
p= 1/2*rho*v.

Will this be right? Can the solver work with this boundary condition as an addition to the flow, or will this condition stands constant over the whole time.


(I´m also using the function "Global Initialization" as an starting Velocity to help the simulation for a better convergence. I hope this won´t disturb my calculations. It should only work as a starting point.)




I´m thankfull for every idea or comment. I hope my question is comprehensible.

[ghorrocks]

If the flow can be assumed incompressible then you can just sum all the pressure increases and decreases together and apply them as a single pressure difference over the entire tunnel. This will give the correct overall flow through the tunnel, but be aware the pressure could be offset compared to the actual values at places as you are not modelling the individual fans.

If the flow cannot be assumed incompressible then you need to model all the inlets, outlets and fans individually.

[eSKa]

Thank you for the reply.

In my case incompressible.
I´m a little bit confused.

To prevent a missunderstanding. My tunnel with a fan is modelled, with its rotating domain. So I need only the basic velocity of the tunnel, the pressure increase of the fan is unknown. This will be calculated in the simulation. I´m scared that a boundary condition could work wrong.

I learned it that way. A boundary condition can be set and it will persist the whole calculations. E.g. an inlet velocity.
In my case I have an inlet velocity, but now I turn on my fan, and the velocity on the inlet increases. How can I set this basic velocity which must be added to the impact of the fan. This basic velocity may not be changed by the calculations.

[ghorrocks]

This question is sounding just like the last one you asked.

You need to understand some very fundamental concepts in CFD, which is that you model a certain region, and the modelled region has boundary conditions on the outside. Those boundary conditions must describe everything which can possibly happen from the rest of the world outside your modelled region.

So that means, in your case, the boundary conditions must have the same system characteristic as the tunnel you are not directly modelling. If you do not know the system characteristic you do not have enough information to perform your model.

I recall saying this on your previous thread. What is different about what you are asking here?

[eSKa]

hi,

do you mean my thread about the loss coeficcient?

I understand the logic of the boundary condition, but I cannot imagine how to solve this problem.
In this case I have a tunnel, and outside is nothing, only air. But their is a wind.
In a normal case Im setting the pressure in the Openings to 0 Pa...the relative pressure.

So can I now set the pressure on the opening 1 to 0 Pa and on the opening 2 on some other Pa? To create a velocity? But will the rotating domain (fan) maybe disturb the solution completly or will the pressure difference do his work.
It sounds ok to do this so...but how can I be sure that this will work.

[eSKa]

I´m searching for solutions. But its partly confusing.
I´m at the point that I need a pressure difference on the openings. I have no other option. Do I?

Without my wind I have a difference of 0 Pa (relative pressure). But when the wind is working in the tunnel, the pressure on the inlet must be over zero.

p(total)=p(static)+p(dynamic)=p(static)+1/2+rho*v^2

I could set the opening static pressure, e.g. 1m/s wind --> 1/2*rho*1(m/s)^2.
But will this work, where could be problems or is this solutions total waste?



edit: The tunnel length and the pressure losses in the tunnel are important for calculating the pressure on the inlet. To get a wind of 1m/s. So it is possible to implement my idea? I´m totally confused. Now I´m thinking it is not possible, because of the unknown characteristics inside and outside.


edit: maybe is bernoulli an option to calculate a pressure for the inlet?

[ghorrocks]

It is time to go back to basics here. Unless the fundamental setup of this simulation is correct it is going to go nowhere.

Please draw an image of what you want to model and what is inside it. Show the inlets and outlets and any fans you want to model. Also indicate what is upstream and downstream of the inlets and outlets and what you know about them.

[eSKa]

Ok. I´ll work on this.
Thank you for ur help!