collisional sheath with secondary electron emission

[Hanquist]

Hi, we are interested in using WarpX to investigate plasma sheaths in the presence of secondary electron emission. Is this something that WarpX could help us with?

We have done this before without the presence of collisions (Sec. 2 of https://doi.org/10.1088/1361-6595/aac6b9) and Fig. 1 of https://doi.org/10.2514/6.2022-3361 shows a nice setup of this.

[roelof-groenewald]

Hi @Hanquist! Welcome to WarpX!

Broadly speaking I do believe WarpX should be able to capture the physics you want. There is an example of secondary ion emission that you could start with and modify for secondary electron emission at https://github.com/ECP-WarpX/WarpX/blob/development/Examples/Tests/secondary_ion_emission/inputs_test_rz_secondary_ion_emission_picmi.py.

What collisions would you need? WarpX includes a MCC module for colliding particles with a static background (commonly used in capacitive discharge simulations, for example) as well as a DSMC module for kinetic-kinetic collisions.

[roelof-groenewald]

Based on the Backtrace I don't think the MPI warning is causing any problems. It looks like the failure is happening in a callback function executing afterstep. Do you get a standard error file as part of the output?

[Hanquist]

Yes, we do get an error file well:
error.txt

[roelof-groenewald]

It looks like the actual reason for the simulation crashing is

IndexError: index 0 is out of bounds for axis 0 with size 0

I'm guessing this is due to the buffer that collects the particles that have hit the wall being empty (possibly since no particles have hit the wall yet). Could you try adding a check around the line 211 (/xdisk/hanquist/hanquist/warpx/SEE_v3/inputs_test_rz_secondary_ion_emission_picmi.py(211): secondary_emission) to skip if the array being accessed is empty?

[shorth13]

Hello, we have just executed the laser‑acceleration example provided in the repository. The run completes without any messages in the .err file, so I do not believe a runtime error occurred. However, I am not certain where the simulation writes its diagnostic or field‑dump files, nor how best to inspect them. Before I modify the input deck in line with the most recent instructions, could you please clarify the default location and naming convention for the output files generated by this example, as well as the recommended method for viewing or post‑processing them?

[roelof-groenewald]

WarpX supports two main output formats - so-called plotfile and openPMD formats. Please take a look at the "Data analysis" section in the documentation for some guides on how to visualize the output data - https://warpx.readthedocs.io/en/latest/dataanalysis/formats.html.

Personally, I like to use Paraview - see https://warpx.readthedocs.io/en/latest/dataanalysis/paraview.html.

Assuming you executed the 3d example, you can also use this script as a minimal starting example of how to visualize the data - https://github.com/BLAST-WarpX/warpx/blob/development/Examples/Physics_applications/laser_acceleration/plot_3d.py.

[Hanquist]

Thanks, @roelof-groenewald!

We are diverging from the purpose of the original thread and are trying to run a simpler case to make sure we have the code running properly before trying SEE (learning to walk before running!).

I will respond with some details in #5721 to make sure Warpx is working okay with our HPC before trying SEE again.

[jlvay]

Hi @Hanquist and @roelof-groenewald. If implementing secondary electron emission using the Python interface only, we could possibly leverage the SEE Python modules from the CERN code PyECLOUD:
https://github.com/PyCOMPLETE/PyECLOUD/tree/master, for example from https://github.com/PyCOMPLETE/PyECLOUD/blob/master/sec_emission_model_ECLOUD.py or https://github.com/PyCOMPLETE/PyECLOUD/blob/master/sec_emission_model_furman_pivi.py.
Note that I also have a self-contained version of the later (the Furman-Pivi model) of the original code in Fortran (with Python wrapping) that we could also either use as is or translate in (AMReX) C++ or Python.