Best Point Gamma Source in MCNP simulation?

AI Thread Summary
Simulating a point isotropic gamma source in MCNP often results in a low number of particle histories in the tally region, prompting the consideration of converting to a cone source to increase particle counts. A cone or parallel beam can effectively enhance the tallying efficiency while still accurately modeling Compton scattering. The discussion highlights the importance of adjusting the distance between the source and detector, suggesting that a closer setup would yield better results if feasible. However, if the experimental setup requires a 30 cm distance, using a conical source is a viable alternative. Additionally, options like restricting or biasing the source can optimize simulation efficiency while maintaining accurate tally results.
Salman Khan
Messages
35
Reaction score
2
TL;DR Summary
Point gamma source in mcnp
When I simulate a point isotropic gamma source in mcnp usually number of particle histories to the tally region (let say a detector placed at 30 cm from point source) is much low (few hundreds only) as compared to actually particle histories (1.0E7), Will it be correct if convert isotropic point source to a cone source so that I can increase number of particle histories towards tally region?.
Thanks every one in advance
 
Engineering news on Phys.org
This is 1/r^2 for radiation, and you will get a flux of around 885 per square cm per 1e7 source particles. (Think surface of a sphere).

A cone shaped beam can work, a parallel beam could also work, set a VEC and DIR=1. The Compton scattering would still be modeled well but it might be worth checking if the background from the rest of the flux affects your result. Can you change what you are modeling to be more efficient or are you replicating a real experiment that has already been done?
 
Alex A said:
This is 1/r^2 for radiation, and you will get a flux of around 885 per square cm per 1e7 source particles. (Think surface of a sphere).

A cone shaped beam can work, a parallel beam could also work, set a VEC and DIR=1. The Compton scattering would still be modeled well but it might be worth checking if the background from the rest of the flux affects your result. Can you change what you are modeling to be more efficient or are you replicating a real experiment that has already been done?
I just want to model my problem more efficiently by entering more possible particles to the tally region to reduce statistical error. As we know mcnp gives answer per particle that is why I want to enter more particle to the region of interest so that I can get more accurate result. M I right ?.if not please suggest thanks
 
Yes, you are right.

But why 30cm? Why not closer?
 
  • Like
Likes Salman Khan
Alex A said:
Yes, you are right.

But why 30cm? Why not closer?
Yes you are right I can decrease distance to further improve results. But what will I do if distance of source and detector is 30 cm in an experimental setup, here the only choice seem to me is to convert point isotropic source to a conical source.
 
I am late in answering this question, but maybe it will help somebody else.

There are two options when you are simulating a situation like this with a point source.
You can "restrict" the source to a cone or you can "bias" the source in the direction of the cone.

The difference between the two cases is that in the restrict case, the tally, which can be per emitted photon, will be different from the isotropic case by some factor depending upon how small the solid angle of the cone is compared to the isotropic source.

In the biasing case, you will still get the run time efficiency of the restrict case, but the tally will still be as if you were simulating an isotropic case.

I am the author of a Python library for generating MCNP input decks which can make this very simple.
https://github.com/pnnl/CardSharpForMCNP
Look for a function called: "insertSource_PointWithAngularAndEnergyDistrib"
In particular look for the parameter "dirDistrib" which can take one of three options: None, Bias, Restrict.

Hope that helps!
 
  • Like
Likes Salman Khan and berkeman
Hello, I'm currently trying to compare theoretical results with an MCNP simulation. I'm using two discrete sets of data, intensity (probability) and linear attenuation coefficient, both functions of energy, to produce an attenuated energy spectrum after x-rays have passed through a thin layer of lead. I've been running through the calculations and I'm getting a higher average attenuated energy (~74 keV) than initial average energy (~33 keV). My guess is I'm doing something wrong somewhere...
Back
Top