[sparta-users] Doubt on how accelerations of external field must be computed

[Max A. <max...@gm...>]

Dear all,

I have been doubting myself on the implementation of external force fields
when dividing by particle mass to compute acceleration.
The way I'm doing it is by doing variable ax  equal "v_Force_x / mass", and
this mass is the sparta variable of the mass of each real particle. My
(maybe silly) doubt is that I have been second guessing whether the
accelerations must be computed per real particle, or per simulated
particle, so basically if I need to use mass or mass*fnum.
The source of this doubt is that using real particle mass I get very high
accelerations and then I need to reduce the timestep a lot to track
particle movement, and with very small timestep then I reached a point
where the fix in/emit command was injecting 0 particles per timestep.

Maybe what I'm trying to do is not very well suited for Sparta, I think it
is complicated and I'm not sure I understand everything correctly, but
hopefully I explained myself.

Does anybody have any insights that may help my case?

Thank you in advance,
Kind regards

P.S. I wrote an email a few weeks ago on how to compute density in a gas
region inside the simulation to be used in the stats printing, if somebody
has any advice on that it would be great too!

Max