boundary conditions

[Bruno Deremble]

Hello basilisk users,

I have two issues with boundary conditions:

(1) I can't reproduce the vorticity pattern in
examples/karman.c

I get positive and negative vorticity at the top and bottom boundary. This may
be related to recent changes in the treatment of boundary conditions?



(2) In the same example (karman.c), I was also trying to change the boundary
condition for the tracer: dirichlet f=1 at the top and bottom boundary. I added

f[embed] = fabs(y) > 0.25 ? dirichlet(1.) : neumann(0.);

and also added tracer diffusion:

#include "diffusion.h"

event tracer_diffusion (i++) {
mgf = diffusion (f, dt, muv);
}

but the tracer doesn't seem to penetrate in the domain.

I guess I am missing something here but I can't see what. I'd appreciate if you
have any clue to solve these issues (I am using the most updated version of basilisk).

Best regards (and happy new year)
Bruno

[j.a.v...@gmail.com]

Hallo Bruno,

1) The vorticity appears in the movie because it is not diagnosed correctly near the boundary.

2) The embedded flux is only computed for partial cells. It is listed as bug...

http://basilisk.fr/sandbox/bugs/marginal_embed.c

FIX: choose a less unfortunate Y0 value. 

Happy new year to you too,

Antoon

Op maandag 10 januari 2022 om 13:09:28 UTC+1 schreef Bruno Deremble:

[Bruno Deremble]

Thank you Antoon!

Bruno