NACA profile - problem with stability of computation
 

Hi!

I am quite new in OpenFoam and his world of pain and frustration, therefore excuse me if you find my problem silly.
I would like to compute separating cavitation cloud over NACA hydrofoil. I have decidet to start with simplier singlephase simulation using pimpleFoam and realizable k-eps model of turbulence. Unfortunatelly, I met some problems with stability. I am not sure if I set the case properly, so there are my dictionaries at the end of my message. Could you give me some advices? Thank you ;)

Simulation description

Single phase flow over NACA hydrofoil
Velocity inlet, pressure outlet
full 3D geometry, Hexahedral mesh
realizable k-eps model

Problem description:
Velocity exceeds reasonable magnitude in the area of the trailing edge. Magnitude of the inlet velocity is 2 m/s. Pictures of the velocity field and detail of the mesh near the hydrofoil are attached to the message.

Attachment 35440

Attachment 35441

Attachment 35442

Attachment 35443

Hello,

try another turbulence model, e.g. kOmegaSST.

H

I will try to do following steps
1. checkMesh - dimension convert m/mm
2. check your BC - especially for turbulent quantities
3. change divScheme to 1st order and time scheme to Euler
4. decrease time step turn from pimple algoritm to piso

VB

Thank you! :)

-I have checked the quality and scale of the mesh, it should be OK
-BC should be ok too
-I tried to change div schemes and few other things
-then I will try k-omegaSST

What is the difference between pimple and piso, is there any crucial advantage of PISO algorithm?

Have a nice day :)
JK

I will tell you about my progress as soon as I will have some results.

In general: Increase nOrthogonalCorrectors up to max 2
or nCorrectors in Piso Loop, depends on your skewness angles


Explanations for Piso vs Simple here:

http://openfoamwiki.net/index.php/Op...hm_in_OpenFOAM

http://openfoamwiki.net/index.php/Op...hm_in_OpenFOAM

http://openfoamwiki.net/index.php/PISO