Hi Richard,

1. Thanks for a hint. The average  for the last column gives a values closer to the analytical one -- see the plot attached. However, it looks like the problem is still there...

2. I have checked my version of PoG (0.2.1), and it looks like the fix is there...
danila@GPU: picongpu$ git log --grep 1625
commit aa4f203104bf04b0aceb6e00d92312f2706bc514
Author: Axel Huebl <axel.huebl@plasma.ninja>
Date:   Thu Nov 3 18:36:38 2016 +0100

    Fix #1625: HDF5 openPMD positionOffset
   
    Fix the offset in particlePatches and the positionOffset
    to use the totalCellIdx of a particle instead of the globalCellIdx. 

I also checked my version of openPMD viewer. It is quite old, however I didn't find any significant changes in the core of if recently...

Regards,
Danila.

---

From: picongpu-users@hzdr.de [picongpu-users@hzdr.de] on behalf of Richard Pausch [r.pausch@hzdr.de]
Sent: Monday, January 16, 2017 11:04 AM
To: picongpu-users@hzdr.de
Subject: Re: [PIConGPU-Users] [PIConGPU-Users] Gas profile

Hi Danila,

the gas setup you are using looks correct to me. You should get the gas profile you are expecting.

1): However, your python code only uses one density value from the zy slice. Thus, due to the random particle initialization, or an ionization method, or the used particle shape, ... the resulting density value varies statically as you can see in your `sim_gas_profile.png`. If you use `print '%.4g\t%.4g' % (ex[1], np.mean(ne[:,-1]))` instead, you should get an average value of the `-1` slice which should more closely resemble your analytical gas profile.

2): A am not an expert in openPMD, but your code looks correct. Your position ex[1]  is for your setup your maximum y position as intended. Are the following bug fixes already included in your PIConGPU version? (https://github.com/ComputationalRadiationPhysics/picongpu/pull/1626, https://github.com/ComputationalRadiationPhysics/picongpu/pull/1627)

Best regard,
Richard

Am 16.01.2017 um 10:00 schrieb Khikhlukha Danila:

Hi Axel,
thanks for you replay. If this is correct, then I observe some strange behavior which I can't explain. Most likely I did some mistake in the settings, however at the moment I can't find it...Maybe you can advice me, what settings should I check...

1. So I'm trying to calculate a LWFA in a 4mm plasma channel. To specify it I modified gasConfig.param as follows:
(PMACC_C_VALUE(float_64, gasCenterLeft_SI, 1.0e-4))
(PMACC_C_VALUE(float_64, gasCenterRight_SI, 3.9e-3))
(PMACC_C_VALUE(float_64, gasSigmaLeft_SI, 0.5e-4))
(PMACC_C_VALUE(float_64, gasSigmaRight_SI, 0.5e-4))
I left gasFactor and gasPower unchanged. Please find a plot to verify my settings of the initial gas distribution attached.

2. Fro a given time step
danila@GPU: run_0021$ grep -ni DELTA_T_SI picongpu/include/simulation_defines/param/gridConfig.param 
32:        BOOST_CONSTEXPR_OR_CONST float_64 DELTA_T_SI = 1.7639e-16;
I expect my simulation to finish after 4e-3/2.99792458e8/1.7639e-16 ~ 76000 steps. In fact I specified 100k steps.

3. To verify the gas density which was used during the simulation I did the following. For every simulation dump I took YZ cross section of my computational domain. I used openPMD viewer do it.  My idea was to get a density of undisturbed plasma just before the laser pulse. So I took a middle value in the last column of the electron density matrix. I also record a position of the right edge of a moving box. I hope attached python script makes it a bit more clear :) 

4. I plot collected data against the expected gas profile and found them to be very different... 

Could you please advise me if there is any setting I need to double check? Please let me know if you need more information about my run...

Best regards,
Danila.

________________________________________
From: picongpu-users@hzdr.de [picongpu-users@hzdr.de] on behalf of Axel Huebl [a.huebl@hzdr.de]
Sent: Friday, January 13, 2017 4:19 PM
To: picongpu-users@hzdr.de
Subject: Re:  [PIConGPU-Users] [PIConGPU-Users] Gas profile

Hi,

yes that is exactly the profile you are looking for and the right way to
init it. It only varies along y (longitudinally) as you described.

Otherwise: just use FreeFormula.

Cheers,
Axel

On 13.01.2017 16:10, Khikhlukha Danila wrote:

Dear all,
this question is a bit basic, however I would like to double check if my
understanding is correct.

Using PoG 0.2.1 I want to simulate LWFA in a plasma channel of let say 3
mm long. I want my gas profile to be uniform in transverse direction. In
the longitudinal direction I want my gas profile to have a plateau and
super-gaussain up/down ramps -- exactly how it is specified in
gasConfig.param file. Am I correct that to achieve it the only thing I
need to do is to fix numerical values for a length of ramps and plateau
to fit my overall plasma length?
Please note that in the speciesIntizialation.param file this profile is
already pointed a source of particles.

Thanks a lot for you support,
Danila.

--

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

#############################################################
This message is sent to you because you are subscribed to
  the mailing list <picongpu-users@hzdr.de>.
To unsubscribe, E-mail to: <picongpu-users-off@hzdr.de>
To switch to the DIGEST mode, E-mail to <picongpu-users-digest@hzdr.de>
To switch to the INDEX mode, E-mail to <picongpu-users-index@hzdr.de>
Send administrative queries to  <picongpu-users-request@hzdr.de>

#############################################################
This message is sent to you because you are subscribed to
  the mailing list <picongpu-users@hzdr.de>.
To unsubscribe, E-mail to: <picongpu-users-off@hzdr.de>
To switch to the DIGEST mode, E-mail to <picongpu-users-digest@hzdr.de>
To switch to the INDEX mode, E-mail to <picongpu-users-index@hzdr.de>
Send administrative queries to  <picongpu-users-request@hzdr.de>

#############################################################
This message is sent to you because you are subscribed to
  the mailing list <picongpu-users@hzdr.de>.
To unsubscribe, E-mail to: <picongpu-users-off@hzdr.de>
To switch to the DIGEST mode, E-mail to <picongpu-users-digest@hzdr.de>
To switch to the INDEX mode, E-mail to <picongpu-users-index@hzdr.de>
Send administrative queries to  <picongpu-users-request@hzdr.de>