Dear all,
I wish to simulate a commercially available X-ray tube with a transmissive 2.5ìm Au target, a 500ìm Be window and a 60keV electron rectangular beam that is normal to the target. I will be dumping in a collision file anything that is ejected from the Be window and enters a narrow cylinder a few mm away. Find my input file attached.
Given the low energy and the thin geometry, I would like to confirm that my reasons for the choice of physics settings are correct. My defaults are EM-CASCA.
1.
EMFCUT: Transport cuts for electrons and photons are set to the minimum values because the lack of any filtration shall allow low energy x-rays to emerge from the window and there is no reason to cut them off artificially.
2. EMFCUT: Production cuts for electrons and photons are set to the minimum values because they can't be higher than transport cuts. Nevertheless I am interested in the full spectrum of this source.
3. EMFRAY: Rayleigh scattering and Compton binding and profile corrections are on because I expect them to be significant at such low energies.
4. EMFFLUO: I keep fluorescence on because I want to get the characteristic X-rays of gold.
5. IONFLUCT: Ionization fluctuations with SDUM=BLANK are set to default EM-CASCA settings. Accuracy is set to maximum because I do not know what the other levels would or would not produce.
6. IONFLUCT: Primary ionization in the Au target is on because I do not know why I would leave it disabled. However, I cant understand which model is appropriate either.
7. MULSOPT is not used because manual recommends EMFFIX. Max.Frac. in Au is set to 0.02 because this is the smallest value I have seen in the literature. I am using a low value because I would expect it to be a safe choice for thin targets against the default EMFFIX=20% for the following reason: Stopping power of 60keV primary electrons = 2.768MeV*cm2/g = 5.34keV/ìm and 20% of 60 = 12keV. After 2.5ìm of Au the electrons will loose approximately(in reality a bit higher) 2.5*5.34 = 13keV. Therefore, the 20% maximum step is almost equal to the thickness of the target and looks unsafe to leave it at 20%. But then again, 2% seems like an overkill.
I would assume that production and transport cuts for electrons could be a little higher to save a little CPU time but I do not have solid excuse for it.
Also, I would assume that MCS with single scattering called only for E<Molliere and on boundary crossings would give the same results and faster but yet I do not have solid excuse for this either, nor do I know the optimum #scatterings or optimal step.
Can anyone please give feedback on the above?
Best regards,
Thomas Primidis
__________________________________________________________________________
You can manage unsubscription from this mailing list at
https://www.fluka.org/fluka.php?id=acc_info
Received on Wed Jul 03 2019 - 18:02:17 CEST