Changing simulation parameters and conditions

[SaumiliJana]

Greetings Aphros team! I was trying out the example 205_multivof/clustering and could successfully run it in its default configuration (except for changing the number of processors and mesh resolution). I was trying to change some simulation parameters and conditions and have some queries regarding the same.

1. How to change the size of the box in which the bubbles are rising, without changing the mesh resolution and other parameters? For example, if I wish to increase the height of the box by 3 times how to do it?
2. How to change the size of the bubbles? Also, is there a way to change the number of bubble sources?
3. If we wish to change the boundary conditions from no slip to periodic, how to do the same?
4. How to change the fluid properties to Non-Newtonian fluid?
5. If we wish to dump the information like velocity field, volume fraction etc into the output files (.vtk) so that we can visualize using tools like Paraview, how to do the same? (Some of these data are present in the traj.csv files, but not in the .vtk files which are generated)
6. Is it possible to directly calculate time-averaged RMS quantities and turbulent energy budget using aphros? If yes, how to do it? It would be really nice if you could respond to the above queries. Thank you.

[pkarnakov]

Hi, The file deploy/scripts/sim_base.conf contains descriptions of configuration parameters along with their default values.

1. the maximum domain size is set by extent, to make the box three times higher, you need to add to add.conf

set double extent 0.06

and then use make 'm=192 384 192' run 2. the size and number of spawned bubbles is control by file bspawn.dat, so to have three sources with larger bubbles, try

0.015 0.002 0.015 0.002
0.02 0.002 0.015 0.002
0.025 0.002 0.015 0.002

3. the boundary conditions in x and z are already periodic, as set in std.conf

set int hypre_periodic_x 1
set int hypre_periodic_z 1

4. non-newtonian would require implementing additional force terms or non-uniform viscosity, which in principle can be done using the hook mechanism as in examples/210_step_hook
5. the velocity field and volume fraction field can be written to hdf+xdmf files by setting in add.conf

set string dumplist vx vy vz vf

6. this can also be done using hooks examples/210_step_hook

[SaumiliJana]

Thanks a lot for your response! It really helped a lot.

[SaumiliJana]

Could you please explain how set extent works? I am running the example on my local machine before trying it on HPC. On my local machine, I was using 4 processors and m = 48 32 48 for the original case. To increase the box size by three times I did set double extent 0.06 and make 'm=48 96 48' run. However, I am getting strange output with 2 boxes one on top of the other and one having almost 3 times more height than the other, with bubbles originating and clustering only in the smaller box. Could you please help to resolve this issue? Thanks!

[pkarnakov]

seems that you also need to change b.dat, which initializes the volume fraction field with a free surface (as one huge bubble) the default values are 0 0.02 0 1e5 0.005 1e5, corresponding to x,y,z,rx,ry,rz for an ellipsoid. so this bubble is supposed to be centered on the domain top, with extent 0.06, you would need to change b.dat to

0 0.06 0 1e5 0.005 1e5

to center the bubble at the top

[SaumiliJana]

Yeah, it worked perfectly. Thanks a lot!