Problem 1671

Summary:	Compton scattering is not computed isotropic in azimuthal angle
Product:	Geant4	Reporter:	martin.stieben
Component:	processes/electromagnetic	Assignee:	Vladimir.Ivantchenko
Status:	RESOLVED INVALID
Severity:	major
Priority:	P5
Version:	10.0
Hardware:	All
OS:	All
Attachments:	azimuthal angle distribution with strange modulation

Description martin.stieben 2014-09-02 16:51:19 CEST

I've encountered a strange behavior of the G4ComptonScattering implementation. 
For all specific Compton models (Livermore, Penelope, polarized Livermore) there is some kind of modulation in the probability of scattering in azimuthal angle. 

I've simulated some non-polarized 150 keV photons propagating in (0,0,1) direction that are scattering in some water and plotted the azimuthal angle distribution of scattered photons (please have a look at the attached file).
The only physics process activated was Compton scattering.  
You can see there is a modulation in the event counts with a repetition period of about 72 degrees.
However, I expected the distribution to be completely flat, since non-polarized Compton scattering must be independent (hence isotropic) of the azimuthal angle.

Is this a physics bug or am I missing something here?

Martin

Comment 1 martin.stieben 2014-09-02 16:52:33 CEST

Created attachment 281 [details]
azimuthal angle distribution with strange modulation

Comment 2 martin.stieben 2014-09-15 10:25:13 CEST

Finally I found out that the reason of this strange behavior was not a bug in the G4ComptonScattering implementation but in the user code.
It appeared due to a non-obvious but wrong conversion between degree and rad which was unfortunately hidden quite well.

Martin