Physical fidelity of secondary field generation in ion radiotherapy

Physics, Transport and Magnetic Fields
1

Versions

Please provide the used software versions.

FLUKA: 4.5.0
Flair: n/a

Description

I am trying to determine optimal combination of physics cards for my MC simulation. Basically, I am trying to score secondary neutron field in ion radiotherapy (so, yield as-is just beyond the patient, I’m not interested in shielding). This would entail C ion beam with energy up to 350 MeV/A impinging on Calcium target (heaviest scenario). I’m interested in reasonable physical fidelity (i.e. PHOTONUC is overkill for my purposes), and I am trying to determine if DEFAULTS HADROTHE or DEFAULTS PRECISIO + EVAPORAT + COALESCE is appropriate for the purpose (or something third?).

Another question: since there is explicit need for specific libraries (RQMD and DPMJET), I used ldpmqmd to compile my custom executable (specific beam shaping). If I did anything wrong with compiling, simulation will stop with error, not silently fail and move to some inadequate libs? The sim is working flawlessly, btw.

Input files

.inp for reference, but not really needed?

head-C.inp (7.5 KB)

2

Ok, just a quick follow-up on the topic. It seems that, for exactly same input, Fluka runs about 3x faster with DEFAULTS HADROTHE against DEFAULTS PRECISIO + EVAPORAT + COALESCE, so I chose latter for production run. I’ll report on statistically relevant differences between spectra, if any.

4

Dear @sxg,

Compared to other Monte Carlo codes, FLUKA has just one physics settings, e.g. the lower your particle thresholds, the more accurate your simulation, but you pay the computing time, as you already hinted. Similarly, some processes which are not relevant for most scenarios are deactivated.

The HADROTHE, mostly used for protons, requires the additional EVAPORAT and COALESCE cards to correctly estimate residual nucleus production, including fragments from ion beams.

In addition to the CPU time, you can compare your physics figure of merit (in this case your secondary neutron yield) for the different cards you are using.

Let me know how I could be of further assistance.

Best,
Daniel

5

Thanks a bunch. I experimented with some cards and found PRECISIO+EVAP+COAL with aggressive cut for e-,e+,γ and δ is a very nice compromise for my purpose. Statistically indistinguishable neutron fields (in angle and energy!) with 5-10x execution speed compared with PRECISIO+EVAP+COAL with default thresholds. Which is physically expected, as low energy particles from EM sector hardly produce neutrons downstream.

-sg