Picongpu: Laser: oblique incidence

[ax3l]

Question of @dengshuan from here:

It seems that PIConGPU can't deal with oblique incidence of laser or maybe I just don't understand PIConGPU well enough. And if not, can you give me some guidance on implementing this.

[ax3l]

if you want to shoot a laser under an angle > 0 on a target you got two options:

first, simply tilt the target instead of the laser. depending on the field solver, that might even be beneficial to keep the laser propagating on-axis.

alternatively, one could implement laser initialization via an antenna. background-fields can be used for that (via currents).

[dengshuan]

Thanks, @ax3l . I' ll try it as soon as possible.

[ax3l]

all right, feel free to come back and comment on the issue. I will just close it for now, but we can reopen it if you need more details.

[WenyinWei]

Hello @ax3l , I think the oblique incidence is a necessity for the multiple incident laser array setting, which involves various interfering phenomena. Is it possible to encapsulate the laser input file one more layer? For example, use a vector to include every laser info.

[sbastrakov]

Hello @WenyinWei . I am, unfortunately, out of the context of this issue. Just to clarify, do you want to have a new type of laser which would act like a number of existing lasers combined? If so, I believe this could be achieved by creating a new laser profile compatible with the existing ones and using it in your laser.param. I could provide more details if this is indeed the case.

[ax3l]

I think one should (re-)implement lasers in PIConGPU via an antenna for exactly that purpose.

[ax3l]

Well, let's open a fresh issue, because that one is from 2015.

[WenyinWei]

Hello @sbastrakov , thx for your prompt reply. The combination is exactly what I want indeed as long as I can adjust the oblique incident angle. The issue is really old and we can tirn to the new issue to discuss.

[PrometheusPi]

@WenyinWei Just to clarify your issue before going to the full-fledged antenna concept (#3159).

1.) You require an angle between your plasma target and your laser other than 90° - is this correct?
2.) Do you need multiple laser pulses in your setup?
3.) Do these multiple laser pulses have different propagation directions, or are they just delayed (train of pulses)?

[WenyinWei]

Hello @PrometheusPi , here is my request.
1) Yes, not only 90 degree.
2) Yes, multiple lasers. Each with an incident angle. Because the lasers cross glass to reach vacuum, they have maxiun intensity that the glass can undertake.
3) In fact, I want them to interfere with each other on a specified sphere range. Of course that needs different propagation directions. A delay setup is also needed to make the laser pulses work on the specified range at specified time, but this has been achieved.

[PrometheusPi]

@WenyinWei Thanks for clarifying this.
(1+2 would be easy to implement with a tilted target, but 3 not thus:)

The only way to currently implement such a setup would eb to use background field inialization. For that you need to define the electric and magnetic field for each position and time in fieldBackgroud.param. In order to let these fields acct on the macro-particles in the PIC code (via the pusher) you need to switch the InfluenceParticlePusher bool to true (for both electric and magnetic fields).

An alternative would be the antenna approach #3159 which is however not yet implemented.

[PrometheusPi]

@WenyinWei: @sbastrakov and I had a brief offline discussion: If your (different) propagation direction (of the various laser pulses) are relatively close together, then adjusting the current side initialization might be the fasted way to get to what you want to simulated.