[bieshuxuhe]

Hello , everyone !
I found a Physical Anomaly in tutorials/interFoam/laminar/damBreak case !

I didn't modification any other thing except for the regions in setFieldsDict ! That is :

regions
(
boxToCell
{
box (0 0 -1) (1.1461 0.292 1);
fieldValues
(
volScalarFieldValue alpha1 1
);

This just fill up the tank about halfway with water. Then began to compute.
Certenly, it is in a condition of rest ! So the Pressure and Velocity should be arround zero .
But I found the wave amplitude of Pressure and Velocity is bigger and bigger in the course of time! Though the amplitude of Pressure and Velocity is still very small, it becomes bigger when there is no energy input !
I think it is a wrong numerical result!

could you help me ? How to solve this problem!
thanks!

[Cyp]

Hi,

this problem is very probably due to spurious currents which are intrisinsic to the VoF formulation as implemented in OF. Some scientists have proposed more or less efficient solutions to remove this indesirable phenomenon, but they have never been included in the official release.

They are several threads on the forum (look for "parasistic currents").

Best,
Cyp

[bieshuxuhe]

thank you!
I wish the opencfd could improve it! because the spurious currents will lead to divergency in fluid and strcucture interaction numerical computation, especially when the elasticity modulus of the structure is a little small. If the problem couldn't be solved , I have to give up the twophase fluid !

[wyldckat]

Originally Posted by bieshuxuhe

thank you!
I wish the opencfd could improve it! because the spurious currents will lead to divergency in fluid and strcucture interaction numerical computation, especially when the elasticity modulus of the structure is a little small. If the problem couldn't be solved , I have to give up the twophase fluid !

Quick answer: Don't blame the software, if you don't understand the physics in play by the solver due to the mesh and boundary conditions .
Because this exact experimentation has already been explained here: http://www.cfd-online.com/Forums/ope...ity-field.html - it's a long read, but it should clear up any misconceptions of "it should just work, because it worked in a commercial software".

You might also want to check this blog post of mine: OpenFOAM: Interesting cases of bad meshes and bad initial conditions

[bieshuxuhe]

Thanks for your help !