Dear PIConGPU-Users,


in case you are using the Gaussian-Beam implementation and performing
2D3V simulations, you will be affected by the following bug (all
releases including `0.1.0`):

In the Gaussian Laser pulse *underestimates* the laser amplitude in 2D3V
by square root of the ratio of the focal spot size over the spot size at
the beginning of the simulation (FOCUS_POS_SI in laserConfig.param).

For short focal positions, this does not cause dramatic differences (few
percent and below) but 3D3V moving window simulations are highly
encouraged to cross-check!

The full description of the issue is provided here:
  https://github.com/ComputationalRadiationPhysics/picongpu/issues/1052

A patch will follow at the end of this day.

Note: The 2D3V implementation is an inherited specialization from our
3D3V default geometry. Unfortunately, some parts such as analytical
formulas for initializations are not fully cross-checked in 2D3V yet.
This bug is basically a missing implementation for 2D3V and if you
should spot similar problems feel free to report them in our GitHub
issue tracker!

  https://github.com/ComputationalRadiationPhysics/picongpu/issues

Thanks to Rémi Lehe (LBNL) for pointing me to the discrepancy and to
Klaus Steiniger (HZDR) quickly pulling out this Diploma thesis
containing a great description!


Best regards,
Axel Huebl
--

Axel Huebl
Phone +49 351 260 3582
https://www.hzdr.de/crp
Computational Radiation Physics
Laser Particle Acceleration Division
Helmholtz-Zentrum Dresden - Rossendorf e.V.

Bautzner Landstrasse 400, 01328 Dresden
POB 510119, D-01314 Dresden
Vorstand: Prof. Dr.Dr.h.c. R. Sauerbrey
          Prof. Dr.Dr.h.c. P. Joehnk
VR 1693 beim Amtsgericht Dresden