[gate-users] source of germanium68

[Dirk Kruecker]

Hello everybody,

Let me summarise the present status on the beta decay module in GEANT4:

There is a beta decay module in GEANT4 that works quiet nice. It allows you to 
simulate the decay of all kinds of nuclei using the ENSDF 
(http://www.nndc.bnl.gov/ensdf/).
It simulates any possible decay, taking into account the latest knowledge on 
branching ratios and creates all possible gammas from the de-excitation of 
the daughter-nucleus.
So you get the right ratio between beta+ and EC, and for each possible decay 
channel the appropriate gamma spectrum in coincidence with the positron.

In GATE the commands to implement e.g. Y86 looks like:  
/gate/source/addSource ytt86
/gate/source/ytt86/setActivity 500000000. becquerel
/gate/source/ytt86/gps/particle ion
/gate/source/ytt86/gps/ion 39 86 0 0
/gate/source/ytt86/gps/energytype Mono
/gate/source/ytt86/gps/monoenergy 0. MeV
/gate/source/ytt86/gps/angtype iso
/gate/source/ytt86/gps/number 1
/gate/source/ytt86/gps/centre 0. 0. 0. cm
/gate/source/ytt86/gps/type Volume
/gate/source/ytt86/gps/shape Sphere
/gate/source/ytt86/gps/radius 0.1 mm
/gate/source/ytt86/gps/mintheta    0. deg
/gate/source/ytt86/gps/maxtheta  180. deg
/gate/source/ytt86/gps/minphi      0. deg
/gate/source/ytt86/gps/maxphi    360. deg

The important part here is the /gate/source/ytt86/gps/ion 39 86 0 0 command
where Z and A are defined. You can define all kinds of isotops here.

Unfortunately there had been several bugs in the GEANT4 code for the beta 
decay and even the latest version  does not run absolutely stable. In 
addition the original implementation is astonishing inefficient (i.e. slow) 
but there is a fast mode sufficient for our purposes.

For GEANT4 version 4.52-4.62 you need patches in GateRadioactiveDecay5.cc
and G4NuclearDecayChannel.cc which I can provide but I would rather use gate 
1.2 with GEANT 4.7. For this version of GEANT I have provided an improved 
algorithm for simulating the beta energy spectra. 

To my knowledge the version GEANT 4.7 will give you correct physical results 
but it still contains some bugs which affect the stability of the program. 
There are at least 2 memory leaks. One I already fixed
http://geant4.cern.ch/asdcgi/geant4/problemreport/show_bug.cgi?id=711
but this is not yet part of any GEANT4 patch. Presently you have to include it 
by hand.

I am still hunting for the second memory leak. Usually it only show up after 
simulating several millions of decays. On average my jobs run for 24h/12 
million events and about 1 of 8 jobs crashes. If you can live with this I 
would give it a try.

I hope the helps. Let me know if you really want to work with the older GEANT4 
version.

Best Regards,
Dirk