Spurious heating when running Hybrid solver

[JohanKoelsenDeWit]

Hi everyone,

I am looking into running 1d simulations with the hybrid solver such that the electrons are not modelled as particles in the simulation. However, when I use this setting I see an unphysical large temperature increment of the ions.

For an initial test, I wish to simulate a simple low-temperature and homogeneous plasma. Attached is my parameter file for the simulation:
IBW_input.txt.txt

The ions are modeled as protons, the temperature of both electrons and ions are 5 eV, the grid spacing is L_{debye}/sqrt(2) where L_{debye} is the debye length. I use 1000 particles per cell and have a constant background magnetic field using the vector potential.
The boundaries for the ions and fields are periodic, and my time step in the simulation is the Courant condition, dt = 0.95 dx/c. I use 1024 grid cells, giving me a domain size of about 5 cm.

Below is the ion temperature as a function of space and time. It starts correctly with with T = 5 eV, but then increases drastically.

I have tried various settings with PML and thermal boundaries too, but I always see this temperature increase.
I am clearly doing something wrong, and hope you can help me spot my error!

Thank you very much.
Best, Johan

[roelof-groenewald]

Hi @JohanKoelsenDeWit. Thanks for reaching out. I haven't had yet taken a careful look at your input script but the setup you describe is very similar to the example described here. Would you mind running that example input script to confirm that you do not see the spurious ion heating with those parameters? Once that is confirmed it might be helpful to slowly change the parameters in that input to your desired parameters to see when things start going wrong.