Problem with snapshot in bview2D

[Alexandre Lecoanet]

Hello,

I am trying to use masks to model an axisymmetric constrained domain in which a jet impacts the bottom of a pit. The bottom andside of the pit have no slip and no penetration conditions for the velocity u.

I attach the .c file.

The code runs and gives out snapshots and a movie. (It fails after some time but it is not a problem for me).

The thing I don't understand is that it produces a movie but I cannot open the snapshots with bview2D (using bview2D snapshot-0.1) it gives me the following error :

restore(): error: expecting 'flags'

I attached the code as well as the movie and a snapshot.

Does anyone have an idea of what I might be doing wrong ? I need help.

Also I would like to use a slanted mask (with the mask defined as a slope to imitate a conical shape in axisymmetric) but when I replace, in the file attached,

the expression in mask by :

mask (x < y/2 ? top : none);

I get the following error message signaling a floating point exception error. I don't understand why.

/home/lecoanet2/logiciels/basilisk/src/tension.h:44:error: [Thread debugging using libthread_db enabled]
/home/lecoanet2/logiciels/basilisk/src/tension.h:44:error: Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
/home/lecoanet2/logiciels/basilisk/src/tension.h:44:error: [New Thread 0x7ffff13ee700 (LWP 22495)]
/home/lecoanet2/logiciels/basilisk/src/tension.h:44:error: Thread 1 "ja_m3" received signal SIGFPE, Arithmetic exception.
/home/lecoanet2/logiciels/basilisk/src/Makefile.defs:45: recipe for target 'ja_m3.tst' failed

Best regards,

Alexandre LECOANET

 

[Youssef Saadeh]

Hi Alexandre,

Masking is the old way of creating boundaries inside your domain and will soon be obsolete, leaving the stage for the new embedded boundaries, which don’t work as of yet for multiphase flows.

Having said so, dumping is not compatible with masks. You also lose some other functionalities when you mask, such as MPI. Generating movies on the fly works as you have observed, but indeed you cannot restore your dump files and open them interactively, using bview2D. I suggest that you output your fields into files and post process them elsewhere (gnuplot, matlab, python, etc.).

Regarding the second point, what you’re doing is that you’re masking the region left of the line x=y/2 which is what you want I suppose; however you’re leaving no place on the left boundary out of which you can actually inject your fluid, since you’re starting your mask directly from (x,y)=(0,0).

I suggest something like:

mask (y > 2*x + R ? mur : none);  // with R a certain radius for your nozzle to let something flow.

I also understand that your conical shape is a nozzle so why the top boundary condition instead of the “wall” you used in previous simulations?

Best,

Youssef

--
You received this message because you are subscribed to the Google Groups "basilisk-fr" group.
To unsubscribe from this group and stop receiving emails from it, send an email to basilisk-fr...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/basilisk-fr/2dc9853d-e73e-4f99-885d-c590b5ee9a00%40googlegroups.com.

[Alexandre Lecoanet]

[Alexandre Lecoanet]

Hi Youssef,

Thank you for your reply. I indeed tried to use the mask as embedded boundaries do not work for multiphase flows. Do you know if embedded boundaries with multiphase flows will be implemented in Basilisk ?

In fact what I want to do is to make the jet impact on a convexe conical shape (like in a conical glass).

Best regards,

Alexandre LECOANET

Youssef

To unsubscribe from this group and stop receiving emails from it, send an email to basil...@googlegroups.com.

[Youssef Saadeh]

Yes, at some point, I guess it will.

In fact what I want to do is to make the jet impact on a convexe conical shape (like in a conical glass).

Well, in this case, you’re masking your jet in what you wrote. Check it and make sure you initialise your fractions accordingly as well.

Best,

Youssef

To unsubscribe from this group and stop receiving emails from it, send an email to basilisk-fr...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/basilisk-fr/4b8d7736-1674-41e8-bac2-a331e51956e8%40googlegroups.com.