Angular distribution of sourse.

User Support
1

I study neutron protection calculations and solve problems with very large absorption.
In geometry, there is a large tokamak hill and relatively small thalli. As a result, more than 90% (or maybe 99) of neutrons even from the inner side of the wall do not play a role (and accordingly do not affect the answer, consuming time and memory).
To solve that issue, I want to calculate the parameters of neutron flux in different areas before the protective wall, and use this data in another simulation to create a simplified geometric model, which consists only from a one wall (and sometimes the channels in it). As far as I understand this is a typical approach, but I have encountered problems in technical execution.
I have already made such a model with a precalculated neutron energy spectrum. The next step is to use not an isotropic source but a precalculated angular distribution (and use only a limited solid angle less than 2Pi).

Using a tally with an angle filter is not a problem for me (similar to an energetic filter). But how do I create a box-shaped source with the obtained angular distribution?

In the documentation I found a few functions for example openmc.stats.PolarAzimuthal.

But the combination of my level in programming, knowledge of openmc.stats and extremely laconic explanation of this and similar functions did not allow me to use it even after a long study of the question.
Maybe someone has an example of a tabulated task of angular distribution or another one, which can be adapted for the described task?

Thank you for having read my long letter :slight_smile:

2

I figured I’d have to be a little more specific.
I understand how to make the directed radiation using the openmc.stats.polarAzimuthal function. How to “cut out” a certain solid angle from the uniform radiation. But I don’t know how to make weighting coefficients for different angles by the same way as with energetic distribution.

3

Hi Illia,

Are you saying that you want to apply a biasing on the angular distribution itself? Currently we don’t have a way of performing any source biasing (modifying the weight of the particles produced). If you are just looking for an example of how to use PolarAzimuthal, here’s a quick snippet:

Create uniform azimuthal angle

phi = openmc.stats.Uniform(0.0, 2*pi)

Create forward directed polar angle

mu = openmc.stats.Uniform(0.95, 1.0)

Combine into distribution and set forward direction as positive x

angle = openmc.stats.PolarAzimuthal(mu=mu, phi=phi, reference_uvw=(1.0, 0.0, 0.0))

Best,
Paul