Test current smoothing fix

[steindev]

I will document the tests for the current smoothing/current interpolation fix here.

3D reference run without current smoothing. Standard LWFA example using 4 GPUs.

@t=7000DT

The contour is a boundary of the region marking the frequencies which propagate slower than electrons in the bunch (e_energy 0.5MeV)

[steindev]

3D run with new current Interpolation. Standard LWFA example using 4 GPUs.

Comparing to reference run

In this run with filter:

• charge distribution peaks at lower energy
• charge distribution seems wider over energy
• bin to bin fluctuations of charge distribution are larger
• dip of charge distribution at E~400keV is more pronounced
• third peak of charge distribution, being originally located at E~600keV, is gone
• high energy region of charge distribution (E>600keV) looks similar
• reals space images of force field distribution look similar
• Fourier space image of force field distribution clearly shows damping of high frequency/short wavelengths components
• low frequency peaks in Fourier space distribution of force field have more pronounced mirror images along the transverse direction
• These transverse mirror images are not damped in contrast to all other force field components

[ax3l]

Careful, don't test such things in under-resolved simulations. Simulations barely resolving the plasma physics, such as the default LWFA, will rather smear around the artifacts they have already. Increase resolution and density gradients to something reliable first!