I’m working on Geant4 to estimate the full energy peak efficiency for an Ge detector and determine the dead layer starting by absolute efficiency calculus.
The simulated absolute efficiency was calculated by dividing the number of gamma detected by the initial emitted gamma.
Upon comparing them, I noticed a significant difference. The simulation is higher than the measurement efficiency
Question ; which physical process should I use, should I add as folder.CC in src
Thank’s lot for helping
Are you using an existing physics list (if so, which one?) or having you implemented your own?
I’m using Penelope proposed by Geant4
But I think I Didn’t implemented well
I added in main.cc
// Physics list
G4VModularPhysicsList* physicsList = new QBBC;
physicsList->SetVerboseLevel(1) ;
physicsList->ReplacePhysics (new G4EmPenelopePhysics) ;
runManager-›SetUserInitialization (physicsList);
- can I implement directly instead QBBC ??
I defer to the EM experts here (@civanch and @pandola among others) for recommendations for a suitable physics list. Are you applying any correction to the number of gammas detected in the simulation (solid angle, thresholds etc)?
I have done many efficiency comparisons over the years, for both NaI and Ge detectors. The agreements I usually got were in the low percent (~ a few percent), for both the photo peak and the overall spectrum shape. At the time I used the low energy Electromagnetic physics. But the geometry of the source and the detector has to be very well described. Ge detectors have a finger that goes into the detector to cool it, so you really have to take into account ALL the geometry. I would trust the low energy physics in geant to ~1-3%.