search

Warwick-Plasma / epoch

Particle-in-cell code for plasma physics simulations
https://epochpic.github.io
GNU General Public License v3.0
184 stars 58 forks source link

3D Gaussian laser #683

Closed crazy-gif closed 3 months ago

crazy-gif commented 4 months ago

Dear All, I am new to EPOCH and have finished the example of 2D Gaussian laser. Now I am trying to transfer 2D laser to 3D laser. I add the grid nz and t_profile, then replace the y in profile with r=sqrt(y^2+z^2). However, the code can not run with the error that "no compile time options". I don't know how to tackle this question. Can anyone give me some advice?

Thank you. Regards, crazy_gif input.deck.txt

Status-Mirror commented 4 months ago

Hey @crazy-gif,

I'm not sure if this is your problem, but are you aware your t_profile can go negative? At $t=2.5$ fs, your t_profile is -1. This key is supposed to describe the envelope of the laser, not the oscillation term.

Have a look at our laser pulse example to see how you might want to implement a temporal envelope. If you still have issues, let me know.

Cheers, Stuart

crazy-gif commented 4 months ago

Dear @Status-Mirror， Thanks for your advice. Acording to it, I have correct the t_profile to make it be positive. However, the code still can not run.If there are other error? Thank you, crazy_gif 3Dlaser.txt

TomGoffrey commented 4 months ago

no compile time options is not an error. It's a message printed if EPOCH is compiled with default options. What makes you think the code can't run?

crazy-gif commented 4 months ago

Hey @TomGoffrey, Thanks for your help. When I try to run it on the super computing platform, there is no the sdf file that I usually get. Then, I think the code have some errors so that it can not run.

Status-Mirror commented 4 months ago

Hi @crazy-gif,

A 3D EPOCH simulation like this will be slower than a 2D equivalent. Maybe you just need to run the simulation for longer?

You can make the code output its progress on every step by adding the line:

    stdout_frequency = 1

to your control block. As a quick warning, most groups find 3D simulations too computationally demanding to actually run, even on supercomputers. Most of the papers using PIC simulation keep to 2D.

Hope this helps, Stuart

TomGoffrey commented 4 months ago

Might also be worth adding dump_first=T to an output block to check you have no I/O issues. The files are potentially very large.

hoaithanhbk commented 3 months ago

Dear All, I am new to EPOCH and have finished the example of 2D Gaussian laser. Now I am trying to transfer 2D laser to 3D laser. I add the grid nz and t_profile, then replace the y in profile with r=sqrt(y^2+z^2). However, the code can not run with the error that "no compile time options". I don't know how to tackle this question. Can anyone give me some advice?

Thank you. Regards, crazy_gif input.deck.txt

can you contact with me via email "hoaithanhbk281113@gmail.com" thank you

crazy-gif commented 3 months ago

Dear @Status-Mirror, Thanks for your advice again.I have solved the proplem and gotten the sdf files quikly，after I change nx=500 to nx=200 and recompile the DevelopmentenEnvironment. thanks again crazy_gif

crazy-gif commented 3 months ago

hey @TomGoffrey ， thanks for your advice. Your guess is right. The output files are so large that I have to change it to nx=200 to get the sdf files quickly. At the same time，I use more computer nodes to run it.Now，I have solved this proplem. thank you again carzy_gif