Particles not conserved?

bendudson / hermes-3

Multifluid drift reduced fluid model

GNU General Public License v3.0

37 stars 17 forks source link

Particles not conserved? #247

Closed GF-Abetharan-Antony closed 2 months ago

GF-Abetharan-Antony commented 2 months ago

Hello,

I am running a non-conventional setup where along y variation is not of importance and only the radial is.

I run with fixed velocity of 0 and evolution of all other equations.
bndry_xout = free_o3 for all density terms
bndry_xout = dirichlet on the deuterium/electron pressures
Recycling happening at bndry_xout for impurity and main species.

Appears to be generating impurities, uncertain how. I have attached the BOUT.inp as a txt file. A simple sum of ionised impurities see orders of magnitude increase in density from t=0 to t=-1 (no initialized background neutral impurity)

Tested not using neutral_mixed and appears to preserve particles as expected.

I am guessing I am doing something strange, though not completely certain.

BOUT.txt

bendudson commented 2 months ago

Hi @GF-Abetharan-Antony ! This is a slightly unusual setup, but we should be able to get this working. What is the model intended to be in the parallel direction? Is there no flow of energy or particles in Y? Should ions recycle from the outer walls (in X)? Thanks! Ben

GF-Abetharan-Antony commented 2 months ago

Indeed! Thanks for the help!

The big picture we trying to simulate is a core flux-tube in contact with a wall with all particle and power loss is due to the outer radial wall.

Answers to your questions:

The Y-axis should have no loss of energy/particles.
Ions should recycle from the outer wall

GF-Abetharan-Antony commented 2 months ago

Hi Ben, Happy to take a look at this myself, if you have any suggestions where to look, I will do a dive.

Abi

mikekryjak commented 2 months ago

Hi @GF-Abetharan-Antony, what's your email address? I can add you to our Slack channel, where it's a bit easier to discuss and troubleshoot these things.

GF-Abetharan-Antony commented 2 months ago

abetharan.antony@generalfusion.com

GF-Abetharan-Antony commented 2 months ago

After Discussions with @mikekryjak, no bug, using free_o3 boundary condition choices lend itself to influx of impurities, some solutions discussed but preliminarily will simply work with a decay length BC instead.

bendudson commented 2 months ago

Thanks @mikekryjak and @GF-Abetharan-Antony !