Message: Re: Running IAEA phase space files as particle generator  Not Logged In (login) 
Dear Miguel,
I have a question regarding your message. About the recycling, you wrote "As for recycling, you must know that *all* copies of the same particle are produced *at the same event*. The reason of doing this is because it is the only way that one can keep statistical correlations between particles produced by the same original history (at it applies for recycling). Simultaneously, the statistical weight of the track is divided by the number of repetitions set, so that dosimetry quantities can be normalized accordingly." Could you explain me please how to divide the statistical weight of the track by the number of repetitions set? If I recycle particles I also get an increased dose value (and as far as I understood from your message, I should get the same value as without any recycling). Thank you very much, Juliana (On Wed, 08 Jun 2016 21:20:10 GMT, Miguel CortesGiraldo wrote: > Dear Brett, > > What you have observed is the expected behaviour of our IAEAphsp tool > for Geant4. Let me explain you why you observe those apparently wrong > results: > > 1) When you write an IAEAphsp file, it keeps track of what happened at > each single event of the original simulation  this is the reason why > you need to simulate 10^7 events in order to read *all* the phsp file > (since it corresponds to 10^7 original histories). A short example: > imagine you have generated a phsp from 100 original histories, and you > have only scored a photon at events #44 and #76. Then, if you read back > that phsp file in another Geant4 application, what you will get when > applying /run/beamOn 100 (in this particular case), you will get a bunch > of 'null' events, except for #44 and #76, which correspond to the events > when the photons were scored. > > The reason of doing this was simply to always keep track of the number > of original histories you are actually shooting in a simulation (so that > eventually normalization to dose per primary particle can always be > done). The drawback is that you will get many 'blank' events for the > sake of reproducing exactly what happened when the phsp was written. I > know that other codes (such as EGSnrc) did not followed this approach in > the past (I do not know if they do it now), so I understand your > question. > > 2) As for recycling, you must know that *all* copies of the same > particle are produced *at the same event*. The reason of doing this is > because it is the only way that one can keep statistical correlations > between particles produced by the same original history (at it applies > for recycling). Simultaneously, the statistical weight of the track is > divided by the number of repetitions set, so that dosimetry quantities > can be normalized accordingly. > > To explain your problem, I assume you are scoring the energy deposited > at a certain volume, but you are not multiplying by the track's > statistical weight. If you do so, you should get again 17 keV. If you > are scoring a different quantity, then I will need to know more details > about it. > > I hope it helps, > > Miguel >

Inline Depth:  Outline Depth:  Add message: 
to: 