Message: Re: Electric field and "local time goes back" issue Not Logged In (login)
 Next-in-Thread Next-in-Thread
 Next-in-Forum Next-in-Forum

None Re: Electric field and "local time goes back" issue 

Forum: Fields: Magnetic and Otherwise
Re: None Mixed Electromagnetic field and ChordFinder (Francois Mauger)
Re: None Electric field and "local time goes back" issue (Francois Mauger)
Re: None Re: Electric field and "local time goes back" issue (John Apostolakis)
Date: 25 Apr, 2015
From: Francois Mauger <Francois Mauger>

Dear John,

Finally I was able to install G4.9.6.p04 and make it work with my SW 
framework.
The geometry setup is rather simple: a box of STP air with various simple
daughter volumes of solid materials (lead copper, plastic...) in it.

I put a E_y electric field of 3e5 V/m in the world volume (filled with 
STP air)
and I ran again my G4 simulation with 1 MeV electrons shot in 4pi from a 
thin
film representing a radioactive source in front of what represents an 
"optical module" (scintillator+light guide+PMT).
Then let G4 track the electrons through the setup.

Unfortunately, the bug is still here, despite the patch. So I made some 
investigations.
First I have changed the intensity of the field to see if the mean 
number of "invalid steps"
(those with backward times) is changing.

The results are in the table below, where E_y is the intensity of the 
electric field
and P(path#) is the proportion of events that have at a "bad step":
<pre>
------------------------------------------------------
E_y (V/m)| 3e5  | 3e4  | 3e3 | 3e2 |  30 |  3  |  0  |
------------------------------------------------------
P(p01)   |  ~2% | idem | idem| idem| idem| idem| idem|
------------------------------------------------------
P(p04)   |  ~2% | idem | idem| idem| idem| idem| idem|
------------------------------------------------------
</pre>
Conclusion:
- no difference between p01 and p04
- no dependance with the intensity of the E field
More, even if the E field is strictly zero (i.e. no field) the bug occurs
with the same intensity!

Then, using the "G4ParticleChange Information" dumped at each bad step, 
I have built
the histogram of the kinetic energy of the electrons that are involved 
in the "bug".
The conclusion is that most of them are rather low energy, below 50 keV. 
So I suspect
these are delta-rays, randomly scattered along the primary tracks.

Next I have plotted the positions of the "bad steps". They can occur 
everywhere
in the geometry, even in the gas where no strong stepping algorithm 
constraints exist (MS at gas->solid transition),
but with a larger probability in high density materials, particularly 
near the surfaces of solid volumes
where particles leave the gas to enter a solid detector. So again, this 
is not a surprise because, I expect
a rather big amount of generated delta-rays as soon as a primary 
electron enters a solid volume, so near its surface
(a few mm). I consider this does not give ant valuable hint to 
understand the nature of the problem.

This is all for now. I envisage to investigate the role of the tuning 
parameters provided by G4.
My current setup uses the following (G4 default) values:
<pre>
double delta_chord = 0.25 * CLHEP::mm;
double min_step = 0.01 * CLHEP::mm;
double delta_one_step = 0.01 * CLHEP::mm;
double delta_intersection = 1.0 * CLHEP::micrometer;
double eps_max = 0.001;
double eps_min = 5e-5;
</pre>
but I have no idea of what is the best direction to go further and I 
cannot make my FW works
with 4.10 within a reasonnable short delay.

Regards

frc
--

>Le 24/04/2015 11:22, John Apostolakis a écrit :
> Dear Francois,
>
> Thank you for your clear message.
>
> Could you try to run your program with most recent versions of Geant4 ?  There is a possibility that there was an underlying error which has already been found and fixed - but the fix was not available in Geant4 9.6 patch 1.
>
> It would be ideal if you can use the latest Geant4 release and patch - Geant4 10.1 patch 1.
>
> In case you are not able to use your program with this, can you please upgrade to 9.6 patch 4, as this contains many fixes which were identified since the creation of 9.6 patch1.  ( It is difficult for us to justify trying to find an error in patch 1, if it has already been fixed in patch 4.)
>
> Let me provide some background about the issue.
>
> In setups in which there is an electric field, the field propagation classes integrate time during the motion along the curved trajectory.  This is needed because the velocity changes (along with the momentum) - which does not happen in the case of a pure magnetic field.
> So the code which is exercised in this case is not used in the pure B field case.  Different behaviour is expected - although an error is not.
>
> I hope that these suggestions lead you to obtain a version which already works.  Please let us know if the issue is corrected (or not), and which version of Geant4 you have used.
>
> In case it does not resolve the issue, we will need some more information to help us to diagnose it.  We encourage you to provide it by filing a problem report in the Geant4 problem reporting system (you can find the link to it in the Support pages on the Geant4 web site.)  This will also allow you and us to better track the progress of the issue.
>
> Best regards,
> John Apostolakis
> ===================================================
> John Apostolakis,  PH Department, CERN
>
>
>> On 23 Apr 2015, at 10:42, Francois Mauger <mauger@lpccaen.in2p3.fr> wrote:
>>
>>
>> *** Discussion title: Fields: Magnetic and Otherwise
>>
>> Hi all,
>>
>> I have a Geant4 setup that implements a field with both magnetic and
>> electric components.
>> I can activate on demand the electric field. Basically, my program compiles
>> and runs. I fire electrons of 1 MeV in STP air with a Bz magfield and Ey
>> electric field.
>> However with the electric field activated I obtain random step errors
>> like the following:
>>
>> <pre>
>> ...
>> *************************************************************
>>   Geant4 version Name: geant4-09-06-patch-01    (1-February-2013)
>>                        Copyright : Geant4 Collaboration
>>                        Reference : NIM A 506 (2003), 250-303
>>                              WWW : http://cern.ch/geant4
>> *************************************************************
>> ...
>> [notice:virtual void mctools::g4::event_action::BeginOfEventAction(const
>> G4Event*):171] Event #1
>> [notice:virtual void mctools::g4::event_action::BeginOfEventAction(const
>> G4Event*):171] Event #2
>> [notice:virtual void mctools::g4::event_action::BeginOfEventAction(const
>> G4Event*):171] Event #3
>> [notice:virtual void mctools::g4::event_action::BeginOfEventAction(const
>> G4Event*):171] Event #4
>> [notice:virtual void mctools::g4::event_action::BeginOfEventAction(const
>> G4Event*):171] Event #5
>> [notice:virtual void mctools::g4::event_action::BeginOfEventAction(const
>> G4Event*):171] Event #6
>> [notice:virtual void mctools::g4::event_action::BeginOfEventAction(const
>> G4Event*):171] Event #7
>>    G4ParticleChange::CheckIt    : the local time goes back  !!
>> Difference:  0.14314337090688[ns]
>> e- E=0.0012850040011642 pos=0.011471222165611, -0.070632466275141,
>> -0.097956932327808 global time=0.14314337090688 local time=0 proper time=0
>>    G4ParticleChange::CheckIt    : the proper time goes back  !!
>> Difference:  0.14278431258136[ns]
>> e- E=0.0012850040011642 pos=0.011471222165611, -0.070632466275141,
>> -0.097956932327808 global time=0.14314337090688 local time=0 proper time=0
>>        -----------------------------------------------
>>          G4ParticleChange Information
>>        -----------------------------------------------
>>          # of 2ndaries       :                    0
>>        -----------------------------------------------
>>          Energy Deposit (MeV):                    0
>>          Non-ionizing Energy Deposit (MeV):                    0
>>          Track Status        :                Alive
>>          True Path Length (mm) :               0.0313
>>          Stepping Control      :                    0
>>          Mass (GeV)   :                    0
>>          Charge (eplus)   :                    0
>>          MagneticMoment   :                    0
>>                  :  =                    0*[e hbar]/[2 m]
>>          Position - x (mm)   :                 11.5
>>          Position - y (mm)   :                -70.6
>>          Position - z (mm)   :                -97.9
>>          Time (ns)           :               -0.143
>>          Proper Time (ns)    :               -0.143
>>          Momentum Direct - x :               -0.428
>>          Momentum Direct - y :                0.721
>>          Momentum Direct - z :                0.545
>>          Kinetic Energy (MeV):              0.00129
>>          Velocity  (/c):               0.0708
>>          Polarization - x    :                    0
>>          Polarization - y    :                    0
>>          Polarization - z    :                    0
>>          Touchable (pointer) :            0x2511f90
>>
>> -------- EEEE ------- G4Exception-START -------- EEEE -------
>> *** G4Exception : TRACK003
>>        issued by : G4ParticleChange::CheckIt
>> momentum, energy, and/or time was illegal
>> *** Event Must Be Aborted ***
>> -------- EEEE -------- G4Exception-END --------- EEEE -------
>>
>> [notice:virtual void mctools::g4::event_action::BeginOfEventAction(const
>> G4Event*):171] Event #8
>> [notice:virtual void mctools::g4::event_action::BeginOfEventAction(const
>> G4Event*):171] Event #9
>> 10 events have been kept for refreshing and/or reviewing.
>> Idle> exit
>> ...
>> </pre>
>>
>> I suspect something is wrong with the parameters passed to the G4
>> field/chord/equation/stepper stuff
>> but I have no hint to understand the problem.
>> This problem does not occur with a pure magnetic field, only when an
>> electric component is activated.
>>
>> Thanks for help.
>>
>> my system: G4.9.6 on Ubuntu 14.04LTS 64bits
>>
>> François Mauger
>> --
>>
>> -------------------------------------------------------------
>> Visit this GEANT4 at hypernews.slac.stanford.edu message (to reply or unsubscribe) at:
>> http://hypernews.slac.stanford.edu/HyperNews/geant4/get/emfields/225/1.html
>

-- 
François Mauger
   Groupe "Interactions Fondamentales et Nature du Neutrino"
   NEMO-3/SuperNEMO Collaboration
LPC Caen-CNRS/IN2P3-UCBN-ENSICAEN
Département de Physique -- Université de Caen Basse-Normandie
Adresse/address:
   Laboratoire de Physique Corpusculaire de Caen (UMR 6534)
   ENSICAEN
   6, Boulevard du Marechal Juin
   14050 CAEN Cedex
   FRANCE
Courriel/e-mail: mauger@lpccaen.in2p3.fr
Tél./phone:      02 31 45 25 12 / (+33) 2 31 45 25 12
Fax:             02 31 45 25 49 / (+33) 2 31 45 25 49

 Add Message Add Message
to: "Re: Electric field and "local time goes back" issue"

 Subscribe Subscribe

This site runs SLAC HyperNews version 1.11-slac-98, derived from the original HyperNews


[ Geant 4 Home | Geant 4 HyperNews | Search | Request New Forum | Feedback ]