Problem 2414

Summary: Incorrect range of alpha particles in NIST water
Product: Geant4 Reporter: db001
Component: processes/electromagneticAssignee: Vladimir.Ivantchenko
Status: RESOLVED FIXED    
Severity: major CC: daren.sawkey, jemcfee
Priority: P4    
Version: 10.7   
Hardware: All   
OS: All   
Attachments: TestEM7 example with changes to DetectorConstruction.cc and alpha.mac

Description db001 2021-08-18 13:14:29 CEST
Created attachment 708 [details]
TestEM7 example with changes to DetectorConstruction.cc and alpha.mac

In Geant4 10.7 - 10.7p2, when firing helium ions in G4_WATER with EM standard option 3 or 4 (options 0-2 don't seem to be effected) the range is reduced from its expected value. 

This can be replicated using the TestEM7 example by replacing the fAbsorMaterial from its default material (manually defined water) to just the default NIST water (G4NistManager::Instance()->FindOrBuildMaterial("G4_WATER")). And running with the supplied alpha.mac macro with either EM option 3 or 4 enabled.
There doesn’t appear to be any change in the range of helium ions when running in water which is manually constructed, only when using NIST/G4_WATER.

With this configuration the range of the beam decreases by ~15% for a 600 MeV 
 (150 MeV/u) beam.

I've attached the TestEM7 example with the changes to the DetectorConstruction.cc and alpha.mac files as well as pictures of the h1 histogram using versions 10.6 and 10.7.
Comment 1 Vladimir.Ivantchenko 2021-08-31 18:04:51 CEST
Hello,

thank you very much for the alarm. The problem is real. Concerning TestEm7: it is not needed modifying example code (I mean DetectorConstruction.cc). Geant4 electromagnetic examples have built in UI commands allowing change many things, for example:

/testem/det/setMat G4_WATER

provide NIST water as a material for the target for this example.

The Problem is understood and the fix will be available with the next public release 11.0 and in the patches to 10.7. For the time being you may use UI command which will recover stopping:

/process/eLoss/ICRU90 false

which disable buggy method of computation of the alpha stopping power.

VI
Comment 2 Vladimir.Ivantchenko 2021-09-29 22:36:53 CEST
Hello,

the bug is fixed in the development branch and will be available with the next public release 11.0 (December, 2022). Patch to Geant4 10.7 may be also provided but the recommendation of today is to disable ICRU90 flag.

VI
Comment 3 Vladimir.Ivantchenko 2021-09-29 22:41:10 CEST
The next public release will be in December 2021.
Comment 4 Gabriele Cosmo 2021-10-25 10:37:26 CEST
*** Problem 2439 has been marked as a duplicate of this problem. ***