MCNP6: Energy and angular distribution of neutrons after moderation

[louisdpt04]

Hello,

I'm currently working on a project that studies the influence of a PE moderator on the neutron capture rate of a He3 detector. So far, I've only experimented with the thickness of the PE moderator. The results I have obtained are interesting and I'm trying to justify them. In order to do so, I need to obtain three things:
1) The mean energy of the transmitted neutrons
2) Their energy distribution
3) Their angular distribution

I've already obtained (1) by using a very large half sphere full of He3 and the *F8 tally. However for (2), I struggle to understand what tally I should be using. I would either need the energy of the neutrons (as a spectrum) when absorbed by He3 (assuming there is little energy lost due to scattering) or the energy deposited by the neutrons before absorption as a spectrum as well.

For (3) I'm planning to sort of stack He3 cells on top of each other but is there a better way to approach the problem ?

Thank you in advance,

Louis

 

[Alex A]

Welcome to physicsforums Louis,

I don't quite follow your solution to (1). (2) can be solved using a tally with energy bins. For example f4 for an empty cell in the right place, put e4 for the next line and add a list of boundary energies. (3) can be solved a few ways, you can put a sphere round it and chop it up with, say cones, you could probably use a mesh.

If you can share your input file, rename to add .txt and then attach to a post.

 

[louisdpt04]

Thanks for your response Alex.

For (1), I've basically created a half sphere containing He3 that is sufficiently large to make sure all the neutrons are absorbed. Then I used the *f8:n tally in that half sphere to get the energy deposited by the neutrons. This method works well as, when removing the PE moderator from the simulation, the energy deposition is equal to the energy I've given my neutrons through the SDEF card.

That seems like a good idea for (2), and one quite simple to implement. If I understand correctly, MCNP will score for every neutron that passes through the cell and class them depending on their energy ?

For point (3), chopping a sphere into cones as you suggested seems like the best solution to me. I've started reading the documentation about FMESH & TMESH but I'm still not sure to understand how to implement it, could you perhaps provide me some insights ?

I've attached my input file, it's a bit messy so don't hesitate if you get lost.

 

[Alex A]

There is some advanced stuff in that, yikes. I am mystified by "NOT USING F6 tally because unreliable since CPE (charged particles equilibrium) not holding".

The neutron source has a weirdly low energy, 12.3kev. Not saying it's wrong, I don't know what you are modeling, but it's just strange.

Right now you are simulating a sheet of moderator, is the detector arrangement a sheet or a ball or a cylinder?

 

[louisdpt04]

I initially started using the F6 tally to get the average energy of the neutrons after moderation but the results it would give me were wrong. For instance, it would say that the energy deposited by each neutron was higher than the energy I gave them through the sdef card. Using the *F8 tally instead solved the issue. I think that has to do with the fact I'm creating charged particles (H3 & protons) from neutral ones (neutrons).

I'm working in the context of the 7Li(p,n) reaction near threshold and the neutrons near that have that sort of energy. I've calculated the neutrons energy based on the protons' and it's in accord with what I found in the literature so no problem on that front.

In the file I gave you, the cell of He3 (the detector sort of) is basically half a sphere which is large enough to make sure the neutrons are absorbed, allowing me to get reliable value from the *F8 tally. Otherwise in normal time I'm using a cylinder as detector (that has 1.2cm radius for a length of 20.3cm). But if we go back to my "to do" list, at (2) with the option you suggested, I don't think a cell of He3 will even be needed. No idea for (3) though.

 

[Alex A]

The more I think about this the more I wonder why you are doing (3). I wouldn't have thought abstract experiments would help. Why not simulate the detector and move the source around to work out directional sensitivity?

 

[louisdpt04]

The source is initially a perfect beam that opens up as it enters the moderator. I have already moved the detector closer to the moderator to see if the capture rate of the detector changes. It does indeed but it's not a massive change. I believe the angle the detector covers is quite small despite being very close to the moderator hence the relatively minor change.

I have experimented with the detector radius to make it cover almost half the solid angle and here there is very a very significant change. So the only thing left would be to obtain some sort of angular distribution. Chopping up a sphere into cones using meshes (as you suggested earlier) is what I'd basically like to achieve but I don't know whether that's possible or not.