Weight windows to improve the accuracy of the flux spectrum at low energies

Dear Community,

I am trying to study a neutron flux spectrum and to get a better accuracy in the low energy ranges to carry out some sensitivity analyses.
The domain of the simulation is an infinite lattice, so I’m expecting a neutron flux that is almost spatially uniform.

Since I have successfully used the weight window method to speed up some deep penetration problems, I was wondering if it would work in this case:

I tried to set a flux tally with two filters:

  • A MeshFilter, with dimension (1,1,1);
  • An EnergyFilter, with several energy bins.

I tried to generate the weight windows using: “openmc.lib.WeightWindows.from_tally(tally, particle=‘neutron’)”.
What I expect is a number of lower-upper bound pairs equal to the number of energy bins.
Running the simulation with a few MAGIC iterations, I do not see any improvement.

Here comes the question: in this scenario, in which the lower/upper bounds change only energetically and not spatially, do the weight windows work?

Thank you and compliments for the great work you are doing!

1 Like