[enginpower]

Hi all,
I'm a newbie in the use of OpenFOAM and I've just followed some basic tutorials.

Now I have to solve a more important CFD problem.
I have a control valve and I need to find its flow cofficient when I set a difference of pressure between inlet and output. In this case p is known at inlet and outlet, velocity is what I'm looking for.

I imported my mesh model in openFOAM from Salome without problems.
My douts and problems are how to set the simulation in simpleFoam.
I'm using this solver because my flow is incompressible and turbolent.

As boundary fields, they are inlet, outlet and wall.
I'm following these criteria:

• file p: setting a fixed value on inlet and outlet, no specification for wall
• file U: setting a fixed value on wall (i.e (0,0,0)), no specification for inlet and value

Is it correct?

What else should I set? Turbolence parameters k and epsilon? To consider that velocity is unknown, and I can set it just as initial conditions and not as boundary conditions. The problem is that I don't understand where to set it in OpenFOAM. Which file I have to specify in initial and which I have to specify in boundary conditions? I'm a bit confused!!

And what about the internalField parameter in each variable file in 0 directory? I have to set uniform for each properties?

I hope you can help me.


Thanks in advance!!

[jmdf]

Hi,

I've never tried such setup, however it seems ok.

The fields on 0/ folder are the initial conditions. If you define a boundary as fixedValue, for example, it won't change throughout the run. Although if you use a different type, the values on that boundary will be calculated by the solver.

The internalField are all the cells of your mesh that there are not boundaries. The internal volume of your model. Is usually initialise as uniform.

Hope it helps.

[enginpower]

Thanks to answer!

Then, to sum up, using fixed value, I'm setting the boundary conditions (they will never change), others are as initial. That's right?
Obviously internalField is about initial conditions, because I want to study what happens inside.

And what about setting all other parameters for my specific problem: incompressible steady-flow with RAS turbolence model? I just need k and epsilon (plus p and U, obviously), I suppose.
Why in 0 directory there are so many other files?

Reading all around the web, I got that for k-epsilon turbolence model I need the variable p, U, epsilon, k, omega. Could I delete the other files with nut, nuTilda etc.?

Then, I'm writing these setting apart from p and U already discussed above.
I set for inlet, outlet and wall zeroGradient. Is it correct? Actually velocity is unknown, what I know is pressure. What could you suggest me?
As internalField I set a value calculated considering an expected velocity in inlet and outside, even if inside is not the correct one, but I hope it will be useful to start the problem.

By the way, omega is the vorticity? What is it exacly?

In a fist simulation I get this error:
Floating point exception (core dumped) What does it mean? How could i solve?

Thanks again for who wants to help

[enginpower]

I supposed the solver diverged.

So is there someone could just indicates which methods I should follow to determine initial condition for k, nut, omega etc... and other settings to balance with relaxation parameters, upwind and so on?

Thanks