Constant macroparticle weight injection at a point

[binahn91]

Dear developers,

I am having trouble with constant MW(Macroparticle Weight) continuous particle injection at a single point(or node).

I would like to perform both ES and EM PIC for a RZ system,
with initially empty space,
a point source(invisible cathode) at the center,
which is emitting energetic electrons,
isotropically at a constant rate.

I tried NFluxPerCell,
but it is hard to figure out
how to adjust <species_name>.flux for constant rate emission, isotropy and constant MW.
(No varying MW)

My goal is to investigate the confinement of single charged species (non-neutral plasma) in various E B field, in a lab-scale chamber.

I am trying things by referring to examples such as inputs_test_rz_flux_injection, but to no avail.

Maybe, should I use the '<species_name>.do_continuous_injection with warpx.do_moving_window=1' approach
like issues5131 ?

Thank you for your help!
Sincerely,
Bin Ahn

[roelof-groenewald]

Hi @binahn91.

I think you should be able to do what you are interested in with the NFluxPerCell injector. The one issue is that you won't be able to inject from a single point, but will need to inject over a small area - this can be controlled with the flux_profile input. For example, setting:

<species_name>.flux_profile = "(x*x + y*y < 0.1)
<species_name>.surface_flux_pos = 0
<species_name>.momentum_distribution_type = 'gaussian'

should continuously inject particles over the disc r < 0.1 with a Gaussian velocity distribution.
Will you give that a try and let us know if you encounter any further issues?

[binahn91]

Thank you @roelof-groenewald
I actually partially solved this issue earlier, like the following, similar to your suggestion.

############################################################
################################## [Particle initialization]

particles.species_names = elec
elec.species_type = electron

elec.xmin =  0    
elec.xmax =  0.001  
elec.zmin =  0.049
elec.zmax =  0.051

elec.injection_sources = inject0 inject1

elec.inject0.injection_style = NFluxPerCell
elec.inject0.flux_normal_axis = r
elec.inject0.surface_flux_pos = 0.0001
elec.inject0.flux_direction = +1
elec.inject0.num_particles_per_cell = 1.e1
elec.inject0.flux_profile = constant
elec.inject0.flux = 1e20
elec.inject0.momentum_distribution_type = constant
elec.inject0.ux = 0.186
elec.inject0.uz = 0.186

This makes the emission constant in time and monoenergetic, as I wanted.
However, making it isotropic was not very simple.
For now, I just assigned many injection_sources and gave different directions for pseudo-isotropy.

I will try @roelof-groenewald 's suggestion.

I am also hoping to easily adjust constant particle weight,
like the MultipleParticles method, but with certain rate of injection per time.
(not determined by num_particles_per_cell, flux, etc.)

So, I may try to implement this myself.

I will share the result here in the future.