I have three simple alpha source simulations where I have modelled a point source, a cylindrical source with 0 cm height (I will call this second source the ‘disc’ source), and the same cylindrical source but against a block of iron (I will call this third source the ‘contam’ source). These are all in a box of air. The energy of the alpha particles deposited in the air is then used to generate optiphots. I have attached the three simulations to the bottom of this topic. I measure the optiphot fluence after 4.8 cm from the source in order to avoid the bragg peak. This is to see if the 1/r^2 drop off appears as I’d expect.
Indeed, for the three sources I do observe a 1/r^2 drop; however I also notice that for the contam source the normalisation factor is less than half than that of the disc source. Thus, I took a simulation with the iron block on one side of the disc, then on the other, and added the results together and indeed the signal did not add up to that of the disc source. This is posted below along with the point source and the disc source.
Could it be that you have secondaries created in the piece of iron which travel to the block of air? Then these secondaries are absorbed in air faster than 1/z^2.
I suggest you look at the fluence of charged secondaries.
Apologies for restarting this forum thread. I had initially thought this was the issue as, indeed, there were charged ions being absorbed in the air faster than the 1/z^2 but when replacing the iron material with a black hole material the graph does not change and still get a normalisation factor of a = 0.033.
From your simulation I understand you have a source of alpha particle which are emitted at the border between a region of air and a blackhole region. The primary alpha particles have their initial momentum towards the blackhole region.
This is a very ambiguous situation for FLUKA to work with and will have unpredictable results. If the first step shall in the black hole, you should score nothing at all. If the first step is taked in air, you might have a first interaction. Instead of wondering what is FLUKA’s prefered behaviour, you should avoid such ambiguous situations.
Just one side comment, I noticed you use a time constant of 5 in your opt-prod card:
Note FLUKA uses seconds for units of time. This is rather large as a scintillation time constant and you might want to use a more realistic value.
Thank you for your reply. I have moved the source 0.00001 m away from the border and changed the time constant to 200 ns as recommended but unfortunately still not seen any change.
It’s difficult for me to give a straightforward answer because what you are describing is not a technical problem or question with the software but rather a matter of understanding the physics of your model.
In general, you should perhaps backtrack a bit and ask yourself whether you should actually expect a 1/z^2 dependence. I would expect such dependence only when having a point source of optical photons without absorption. In any other situation this is merely an approximation.
Concerning the normalization factor, this highly depends on the dose deposition in your scintillator. Perhaps the material at the back of your source has an influence on your charged secondaries which influence the spread of optical photon primary vertices. To check that you can perhaps score both optical photons and energy deposition with USRBIN and see the influence of the material.
If you wish to check whether the code works as you would expect (which I encourage you to do in case of doubt) you should perhaps use a much simpler model for which you know exactly what to expect analytically. In the current model, I am affraid you cannot do that.
That said, if you placed your USRBIN detector much farther from the source you should see this 1/r^2 dependence since your source of optical photons would become quasi-point-like. To properly guess the normalization factor is a much more complicated matter. Finally, did you check the errors on the fitted parameters? Perhaps they are larger than you think and that all your results were statistically consistent.
Once again thank you for your response.
I should firstly apologise as I think I was unclear in my previous posts.
My question isn’t to do with the 1/z^2 dependancy, rather what I’m unsure about is why I’m seeing a reduction in the signal by specifically more than half when I expect to see only a reduction by half when introducing a back wall. To clarify, I do believe that the code is working properly and actually the original graphs do contain the statistical errors, they’re just too small to see! I do think this is a physical effect that is causing this I’m just unsure as to what the root of this physical effect is and was hoping I’d be able to find an answer on the forums.
With regards to your idea about the material behind the source having an influence on the charged secondaries, I would have thought that by changing the material to black hole I would expect to see a change in that case? But I’ve not observed any change at all.
As always, thank you very much for being patient with me through this and for your kind help!
Let’s consider only the simulations with the disk source.
In a first attached simulation, there is no other material than air, and the alpha particle are fully absorbed in the scintillating air. You scored at the position 4.8cm<z<50cm, downstream of the beam.
In a second attached simulation, the alpha particles start in a slab of iron in which they are fully absorbed. You score upstream at the positions -50cm<z<-4.8cm. The slab of iron is not a scintillating material and is very absorbative. Therefore, all optical photons which are scored originate from secondary particles escaping from the iron slab. For that simulation, you should score less optical photons. How much less is not clear, it highly depends on the materials and physics settings.
Then you mention that, in another simulation which is not attached, the ion slab is placed on the other side of the source. This would mean that optical photons are again absorbed in air. For that simulation, however, I don’t know where you scored. If downstream, I would expect similar results to the first simulation. If you scored upstream, you probably scored almost nothing.
It is still unclear to me what exactly you attempted to simulate and score.
Apologies again for the lack of clarity from my part, indeed as you suspect I scored downstream for the simulation that was not attached and I see similar results as the first simulation as you expect.
You mention that the secondaries from the slab of iron are the only thing that would generate optical photons, but from my understanding the alpha source is emitting in a 4pi sr solid angle and so it should still mainly be the alpha source generating the optical photons.
Further, as I mentioned prior, when replacing the iron slab with a slab of blackhole I see similar results, implying that there was no significant amount of optical photons generated by the iron slab. As such, the amount of optical photons scored should be exactly half of the first simulation.
Unless I’m misunderstanding a key part of your explanation?
My apologies, I missed the fact that you used an isotropic source.
When having only air, say you score in the negative direction, some optical photons originating from alpha particle propagating in the positive directive will contribute to the signal.
When adding the slab of iron, these optical photon are stopped in the slab, so they do not contribute to the scoring.
