No evidence of beam evolution

[cemitch99]

I tried modeling a positron bunch propagating through a plasma over a longitudinal distance of 0.1 m, using 50 time steps, and I don't see evidence of the expected bunch evolution. Attached are screenshots of projections (z,uz). Given =7.3e3, I would have expected a change in <z> of 0.1 m (unless z is evaluated in a moving frame). Likewise, I don't see evolution in (x,ux) or (y,uy), so there must be some mistake in the input. Below are a few of the most relevant inputs:

hipace.max_time = 0.1/clight
hipace.dt = hipace.max_time/50

beams.names = bunch
bunch.element = positron
bunch.position_mean = 0. 0. 0.
bunch.position_std = 2.357965224510e-4 1.7291616466e-5 6.5e-3
bunch.injection_type = fixed_weight
bunch.profile = gaussian
bunch.num_particles = 1000000
bunch.total_charge = 3.20435326e-9
bunch.u_mean = 0. 0. 7827.804659792008635
bunch.u_std = 0.184576911720494 0.013535539594625 6.262243727833607

[MaxThevenet]

Thanks for this issue @cemitch99!
Indeed HiPACE++ uses a moving frame, so you should not expect much motion in z. As for x and y, I am not sure, the input parameters you share look reasonable. Could you share the full input script you used?

[AlexanderSinn]

Assuming you have no or thin plasma, for testing you could try with hipace.max_time = 10/clight and look at x/ux, that should give visible results.

[cemitch99]

Thanks @MaxThevenet and @AlexanderSinn! I attached the complete input script here, and I will try increasing the runtime until some evolution is visible. In the meantime, can you point me to an example of postprocessing the in-situ diagnostic output using read_insitu_diagnostics.py?

inputs.txt

[AlexanderSinn]

I have been using something like this:

import sys
# change path as needed here
sys.path.insert(1, "/home/hipace/tools/")
import read_insitu_diagnostics as diag
import numpy as np
import matplotlib.pyplot as plt
import scipy.constants as scp


def multiplot(all_data, title):
    ni = 3
    nj = 3
    fig, axs = plt.subplots(ni, nj, figsize=(16,12), dpi=150)

    axs[0][0].plot(all_data["time"]*scp.c, diag.energy_mean_eV(all_data))
    axs[0][0].set_xlabel("Distance in m")
    axs[0][0].set_ylabel( "Energy" )

    axs[0][1].plot(all_data["time"]*scp.c, all_data["average"]["[x]"] )
    axs[0][1].plot(all_data["time"]*scp.c, all_data["average"]["[y]"], ":" )
    axs[0][1].set_xlabel("Distance in m")
    axs[0][1].set_ylabel( "Pos mean" )

    axs[0][2].plot(all_data["time"]*scp.c,diag.position_std(all_data["average"], direction='x') )
    axs[0][2].plot(all_data["time"]*scp.c,diag.position_std(all_data["average"], direction='y'), ":" )
    axs[0][2].set_xlabel("Distance in m")
    axs[0][2].set_ylabel( "Pos std" )

    axs[1][0].plot(all_data["time"]*scp.c, 100 * diag.energy_spread_eV(all_data) / diag.energy_mean_eV(all_data))
    axs[1][0].set_xlabel("Distance in m")
    axs[1][0].set_ylabel( "rel. Energy spread %" )

    axs[1][1].plot(all_data["time"]*scp.c, all_data["average"]["[ux]"] )
    axs[1][1].plot(all_data["time"]*scp.c, all_data["average"]["[uy]"], ":" )
    axs[1][1].set_xlabel("Distance in m")
    axs[1][1].set_ylabel( "u mean" )

    axs[1][2].plot(all_data["time"]*scp.c,diag.normalized_momentum_std(all_data["average"], direction='x') )
    axs[1][2].plot(all_data["time"]*scp.c,diag.normalized_momentum_std(all_data["average"], direction='y'), ":" )
    axs[1][2].set_xlabel("Distance in m")
    axs[1][2].set_ylabel( "u std" )

    axs[2][0].plot(all_data["time"]*scp.c, diag.total_charge(all_data))
    axs[2][0].set_xlabel("Distance in m")
    axs[2][0].set_ylabel( "Charge" )

    axs[2][1].plot(all_data["time"]*scp.c, diag.emittance_x(all_data["average"]))
    axs[2][1].set_xlabel("Distance in m")
    axs[2][1].set_ylabel( "Emittance in x" )

    axs[2][2].plot(all_data["time"]*scp.c, diag.emittance_y(all_data["average"]))
    axs[2][2].set_xlabel("Distance in m")
    axs[2][2].set_ylabel( "Emittance in y" )


    plt.suptitle(title)

    plt.tight_layout()
    plt.show()


# change path as needed here
path_bunch = "/diags/insitu/reduced_bunch*"

all_data = diag.read_file(path_bunch)

multiplot(all_data, "Bunch")

[cemitch99]

This was very helpful, thanks. There is indeed visible evolution over a distance of 10 m.

[AlexanderSinn]

Great!

[cemitch99]

A final question: Regarding postprocessing the in-situ diagnostic output, do you have something similar for the plasma output?

[AlexanderSinn]

Insitu diagnostic is also available for the plasma, its main use is to compute its temperature. It would use largely the same python script for post-processing, but with diag.temperature_in_eV.

[cemitch99]

Thanks--we're interested in the evolution of the electron plasma density. Is this information available in either the plasma in-situ diagnostics or the openPMD output (e.g. as a field)? The available field output visible is: ['Bx', 'By', 'Bz', 'ExmBy', 'EypBx', 'Ez', 'Psi', 'Sx', 'Sy', 'chi', 'jx', 'jx_beam', 'jy', 'jy_beam', 'jz_beam', 'rhomjz'].

[AlexanderSinn]

You can set hipace.deposit_rho_individual = true in the input script and then you get the electron charge density in the openPMD field output.