[sorockin]

Dear Foamers, I need some help.

I try to simulate flow in open channel with interFoam. I use setting from wigleyHull tutorial. I only changed the geometry of channel and inlet velocity (0.25 m/s). But I see instability in simulated results. Tel me please what need to change to get flow without instabilities. Screenshots of alpha, U and case setup are attached.

Thanks a lot!
Maxim

[sorockin]

I have moved form LTSInterFoam to interFoam and changed turbulence values setting. I think that problem with stability was in time constraints or wrong turbulent settings.
Now I use interFoam with maxCo=0.5 and maxAlphaCo=0.5.
I set internalField and inlet value for k from expression k=1.5*(U*I)^2. U=0.25, turbulence intensity I put 1%. So k=9.375e-6.
For mega - from expression omega=Cm^(-0.25)*k^0.5/L, Cm=0.09, turbulence length I put 7% of channel depth. So omega=5.324e-3.
I deleted nut file in 0/ folder to OpenFOAM calculate nut values.
But I still have flow instability.
Help me please I am new in OpenFOAM and don't know what needs to do!
New case setting are attached.

Maxim

[sorockin]

If I turn off turbulence and simulate laminar flow I still obtain instability. What is the cause? Solution scheme or boundary condition? But I took boundary conditions from wigleyHull tutorial. I am confused.

Maxim

[David*]

The second case doesn't look bad in my opinion. You set turbulence at the inlet, and you get turbulence.

In the third case the reason might be some BC.

If you say instability one thinks of numerical instability. You can check it via the residuals. In the forum there is a script for gnuplot that processes the values given in your .log-file (if you write a .log-file).

[sorockin]

Thank you for reply David!

The second case doesn't look bad in my opinion. You set turbulence at the inlet, and you get turbulence.

I used initial setting from "wigleyHull" tutorial. "wigleyHull" used turbulence model too but flow is much more uniform. I simulated this case with flow3D (k-epsilon turbulence model, the same mesh) and had uniform flow.

In the third case the reason might be some BC.

How nice it would be to know exactly what it is - BC or solution scheme.

If you say instability one thinks of numerical instability. You can check it via the residuals. In the forum there is a script for gnuplot that processes the values given in your .log-file (if you write a .log-file).

It is good idea, thank you!

Maxim

[sorockin]

I changed setup for laminar simulation. I moved from 3D to 2D (xz plane) and divided inlet plane. Onto inlet and wall.

So now I have channel 85x1x17 m with mesh 340x1x170. Water depth is 15m.
With BCs:

inlet
U: fixedValue; value uniform (0.25 0 0);
alpha1: fixedValue; value uniform 1;
p_rgh: zeroGradient;

inletWall
U: symmetryPlane;
alpha1: symmetryPlane;
p_rgh: symmetryPlane;

outlet
U: zeroGradient;
alpha1: zeroGradient;
p_rgh: zeroGradient;

atmosphere
U: pressureInletOutletVelocity; phi phi; value uniform (0 0 0);
alpha1: inletOutlet; inletValue uniform 0; value uniform 0;
p_rgh: totalPressure;

bottom
U: symmetryPlane;
alpha1: symmetryPlane;
p_rgh: symmetryPlane;

I still have "instability" at inlet. It may be connected with wrong inlet BC. I will be glad any advice. The case is attached.

Maxim

[sorockin]

I made another changes.
My mesh had dx/dz ratio 2.5. I made dx=dz to remove mesh as one of cause that may induce instability.

Then I changed BCs, accordingly Matthew Denno
http://www.cfd-online.com/Forums/ope...lowing-up.html

inlet
U: fixedValue; value uniform (0.25 0 0);
alpha1: fixedValue; value uniform 1;
p_rgh: buoyantPressure; value uniform 0;

inletWall
U: fixedValue; value uniform (0 0 0);
alpha1: zeroGradient;
p_rgh: buoyantPressure; value uniform 0;

outlet
U: inletOutlet; inletValue uniform (0 0 0); value uniform (0 0 0);
alpha1: inletOutlet; inletValue uniform 0; value uniform 0;
p_rgh: inletOutlet; inletValue uniform 0; value uniform 0;

atmosphere
U: pressureInletOutletVelocity; value uniform (0 0 0);
alpha1: inletOutlet; inletValue uniform 0; value uniform 0;
p_rgh: totalPressure;

bottom
U: fixedValue; value uniform (0 0 0);
alpha1: zeroGradient;
p_rgh: buoyantPressure; value uniform 0;

And I still have "instability" that arise at inlet and then result strongly nonuniform flow.
I tried different "working" BC set from another threads but "instability" didn't disappeared.
May be deal in solution schemes. Can anybody help me? Please!

Maxim

[sorockin]

I took fvSolution setting from Matthew case and looks like "instability" disappeared.
Thank you Matthew for helpful post!

Maxim

[JFM]

Hi Maxim

You refer to Matthew's case in the post above, can you provide a link to this.

Thanks
JFM