I think I agree with your conclusion, code is here as a reference. Thanks!
Ciao,
indeed this is the case. The binding energy is initially entirely ascribed to local energy deposit (L. 546). Then, in the if() block of L. 563 this binding energy can be partially attributed to fluorescence x-rays and Auger electrons, generated by the AtomicDeexcitation module. This block is never visited for He, so the 24.59 eV binding energy of the K-shell (which is subtracted by the proposed electron kinetic energy, see L. 545) is taken by the model as local energy deposit.
Still, I expect than in a normal simulation, this local energy deposit in a ārealā detector is summed with some energy release from the electron, so the 24.59 eV peak, which is an artifact, is not actually visible. I guess this depends on what you are exactly scoringā¦
Ciao,
Luciano
Ciao,
thanks, very helpful. I am scoring for each step in a ādetectorā volume using GetTotalEnergyDeposit().
Ok, when you sum up on many steps within the ādetectorā volume, I would expect this unphysical peak to me smeared out by the energy deposit of the accompanying electron. I donāt see why it is not the caseā¦
well the detector is 1cm3 of helium, so probably not many steps happens there
Indeed, not so many step, but I would expect at least one step of the accompanying electron, which should be enough to ādisruptā the unphysical peak from the He3 binding energy
The peak is always slightly smaller than the nominal binding value, but still on a single bin.
is this a feature or a bug? can somehow be fixed?
Ciao,
no, it is not a bug. This was driven by the fact that the declared range of validity for the low-energy (Livermore) models gets down to 250 eV, so it makes little sense to generate secondaries with energy below that. Basically, the total binding energy (24 eV or so, in your case), is deposited locally instead of going into one or more low-energy x/Auger secondaries (which would have a very short range, though).
Ciao
Luciano
Thanks, makes sense to me.