Re: [ibsimu-forum] Initial beam&plasma questions

[Kalvas, T. <tan...@jy...>]

Dear Zhen,

The convenience functions cover most of typical situations, but there are always particle distributions that are not included in any program. For these situations there is the add_particle() function in the particle database. With it you can generate custom beam distributions. In ICIS 2017 I showed some 3D ECR-simulations with triangular beam distributions being extracted. These distributions were based on single-electron tracking plasma simulations. You can see pictures from these simulations on the last two slides of my presentation, which is available at https://indico.cern.ch/event/628126/contributions/2673125/attachments/1541594/2417725/kalvas_extraction.pdf

The space charge averaging factor "coef" is a damping factor similar to the one often known as under-relaxation factor in the literature of iterative methods. As you can see in the formula, the larger the factor the larger the contribution of the new space charge calculation (in scharge) is compared to the previous space charge (in scharge_ave). The coefficient should always be larger than zero. If coef is large, the simulation generally converges faster if it converges. Using a small coef helps in convergence by suppressing overshoot in the iteration but on the other hand it also makes the convergence slower, requiring more iteration rounds. Sometimes you have simulation cases that do not converge at all of coef is too large. You can see my thesis (available in "Publications" section on the ibsimu website) for detailed discussion on the convergence.

You should always use some diagnostics for observing the convergence of your simulation. The example 2 in the JSS material (http://ibsimu.sourceforge.net/jss2015/02/plasma.cpp) produces a file emit.txt, which has emittance and Twiss parameter from each iteration round. This is one way of doing it.

Taneli

--
Taneli Kalvas
Ph.D., Senior researcher
Department of Physics, room FL114
P.O. Box 35 (YFL)
40014 University of Jyväskylä, Finland
Mobile: +358-44-314-1602
Fax:    +358-14-617-411
Email:  tan...@jy...

________________________________________
From: 申圳 <she...@im...>
Sent: Friday, April 26, 2019 12:41:14 PM
To: ibs...@li...; tan...@jy...
Subject: Initial beam&plasma questions

Dear all,


I'm using IBSimu (version 1.0.6) to simulate the extraction of multi-charged Ar beams from a 28GHz ECR ion source.

I'm using 3D cylindrical model with ParticleDataBase3D.add_cylindrical_beam_with_energy function to set initial beam conditions.

I've got some simulated beam profiles to compare with the experiment results of a flourescent at the extraction region.

Now I want to change the initial beam distributions of different ions, and I have some questions I can't deal with:

  1. I know from the Reference Manual that we can add beam with uniform, gaussian, waterbag,etc. distributions, but is there a chance to introduce distributions from other programs?

  2. If 1. is too complicated to do, then how can I add a hollow beam (for low charge state ions)?

  3. In space charge averaging:   if( i == 0 ) {
                                      scharge_ave = scharge;
                                  } else {
                                       double coef = 0.3;
                                       scharge *= coef;
                                       scharge_ave += scharge;
                                       scharge_ave *= (1.0/(1.0+coef));
                                  }

     what does this "coef" means?


Thanks a lot!

Zhen.