[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [ff3d-users] Navier-Stokes problem
From: |
Stephane Del Pino |
Subject: |
Re: [ff3d-users] Navier-Stokes problem |
Date: |
Tue, 5 Jun 2007 08:04:55 +0200 |
User-agent: |
KMail/1.9.7 |
Hello Thomas.
Le dimanche 3 juin 2007, Thomas B a écrit :
> Hi!
> I tried to use FreeFEM3D to simulate an object being pulled through water.
> First I used the example "navier-stokes.ff" provided with FreeFEM3D but
> I had several problems with the "eps" parameter. After searching in the
> archive of this mailing list I found the thread started by Dominique
> Deloisin just a few days ago and the nice solution by Stephan.
> But I still have some problems:
> 1) Is it possible to simulate the case described above by assuming that
> the object rests while water is moving? The flow is assumed to be
> stationary.
If you mean that in this case, the object is not moving, the strategy used by
Dominique is good I think. But I don't understand why your flow should be
assumed stationary.
> 2) If yes, how can you compute the force acting on the object if the
> algorithm forces the pressure to be quite random in the object itself?
> It seems to be impossible to accurately interpolate the pressure on the
> actual surface of the object.
Yes. The force on the object can be approximated by the integral over the
object of 1/epsilon*U :
double fx = int[M](1/epsilon*ux);
double fy = int[M](1/epsilon*uy);
double fz = int[M](1/epsilon*uz);
where U = (ux,uy,uz) and the epsilon is used for the penalty term P in the
example for Dominique, 1/epsilon = 10^3.
If you need to compute moving objects, I can also give you a solution.
Best regards,
Stéphane.