All-mach & VOF

[wmoste...@gmail.com]

Dear all,

The file momentum.h, which combines two-phase flows into a momentum-conserving scheme, works most excellently for me in the default, incompressible form in which all-mach operates. Is it possible to run all-mach with a defined equation of state so that it is no longer solving the incompressible constraint (to which the pressure evolution equation devolves in the absence of an EOS) in the presence of the VOF solver? I see on the compatibility matrix VOF and compressible flows are incompatible, and I also notice that on vof.h, it remarks:

"We follow Weymouth & Yue, 2010 and use a step function which guarantees exact mass conservation for the multi-dimensional advection scheme (provided the advection velocity field is exactly non-divergent)."

 - the proviso in which is not satisfied in the presence of an EOS.

This is aside from practical constraints such as limitations on the CFL number for compressible flows, etc.

Best,

Wouter

[Stephane Popinet]

Hi Wouter,

> Is it possible
> to run all-mach with a defined equation of state so that it is no longer
> solving the incompressible constraint

The short answer is yes, however this is very much work in progress.

See here for an example:

http://basilisk.fr/sandbox/popinet/collapse.c

For incompressible flows, see also:

http://basilisk.fr/src/navier-stokes/conserving.h

cheers,

Stephane

[wmoste...@gmail.com]

Hi Stephane,

Thanks for the links - that great collapse example has been very useful in particular and looking at it has much improved my rather naive first attempt!

Best,

Wouter

[Zhang 张Yuchen 雨辰]

Hi, I am trying to use Basilisk to model the collapse of cavitation bubble, and I build my simulation based on this your code of collapse.c. The code can work in 2D coordinates but velocity explodes for 2D axisymmetric simulation. 

The attached code simulates a bubble located  in the left lower corner. If #include "axi.h" is commented out, the code works fine, otherwise, velocity explodes and time step becomes very small. I also generate volume fraction field, and it seems the interface does not move near the axis.

Thank you,

Yuchen Zhang

On Tuesday, January 30, 2018 at 1:50:18 AM UTC-5, Stephane Popinet wrote:

[josé López-Herrera]

Hi Yuen,

some metric factors are missed in all-mach.h see

http://basilisk.fr/sandbox/lopez/rising.c

Also check Fuster's sandbox.

Cheers

Jose

[Zhang 张Yuchen 雨辰]

Hi,

I fixed the problem in the metric factor. The resulting volume fraction is attached in f_16.mp4. In the movie, I did mirror the images and It seems that the bubble is not fully spherical. I am wondering is there any problem in code in axi.h.

I also attached my code in rayleigh.c

Thank you,

Yuchen Zhang

[Zhidian YANG]

Hi Yuchen,

I am also doing the collapse of the bubble, based on the file collapse.c. After I tried to fix the problem in the metric factor,  I also ran into the problem that the bubble is not fully spherical.

I am wondering if you have solved this problem yet? It would be so nice of you to give me some information about it.

Thank you,

Zhidian Yang