I am trying to model the response from a CZT detector for gamma radiation. For the present case, I have used a point gamma source of energy 0.6617 MeV (Cs-137). I did the same simulation both with FLUKA and MCNP5. For scoring the response in FLUKA, I have used DETECT card and the routine, usreou.f, provided by David @horvathd for implementing Gaussian energy broadening. Also, to increase simulation efficiency, I have biased the source direction (using SOURCE routine) and normalized the results accordingly. The results are shown in the figure below.
I am not able to understand why there are differences between MCNP5 and FLUKA results near the Compton edge and near the photo-peak (~ 0.63 - 0.64 MeV). Please let me know whether this is expected or if I am doing some mistake in the simulation. I have attached the input files used in this simulation.
It would then be a matter for you to compare with what MCNP does.
Please triple check that 1) the geometry and 2) the source spectrum were identical in both simulations.
Regarding the bump you witness a bit below the photopeak in the MCNP simulation, it might well be an escape peak (the occasional low-energy fluorescence photon managing to escape carrying out some 20-30 keV, and thus making the deposited energy occasionally fall short by 20-30 keV, hence the small bump).
I see you use the DEFAULTS card with SDUM=PRECISIOn. If you don’t specify otherwise, the photon transport (and production) threshold is 33 keV, i.e. not giving the chance of any photon below that to be transported out of the detector. Could you repeat the simulation lowering the photon production/transport threshold? See the EMFCUT card. For good measure lower also the electron production/transport thresholds from their otherwise default value of 100 keV.
Let us know in case the escape-peak problem persists. We’d take a more detailed look.
