Hi everyone,
I’m working on verifying and validating OpenMC for nuclear well logging applications. Amongst the applications are so called Pulsed Neutron spectroscopy measurements. As the name suggested, the tools use a DT source of neutrons capable of pulsing. We detect gammas in various types of scintillators in two-time domains – during pulse (mostly sensitive to inelastic reaction) and during pulse off when we are more sensitive to capture reactions. The gammas originate from the tool itself, but more importantly from the surrounding formation.
The example I’m showing here is a spectrum during pulse on (up to 40 micro seconds). The tool was placed in a block of dolomite (CaMg(CO3)2) .
If we first look at just the gamma flux spectrum, we see an almost perfect match between MCNP and OpenMC. This, in my opinion, verifies we have an exact replica of the MCNP model, including source, material, tally definition and nuclear data, in OpenMC. The plots show the spectrum in the first pane, relative uncertainty of the simulation in the middle, and the ratio of the calculations in the bottom pane.
However, if we look at the pulse height tally (PHT) responses (F8 in MCNP nomenclature) we see significant deviations between OpenMC and MCNP. The MCNP results (when Gaussian energy broadened to account for the detector resolution) match experimental values pretty well below 7 MeV (within 10 %). The deviation of OpenMC from MCNP and the experiment is especially clear below the 6.13 MeV Oxygen peaks (main, single and double escape).
I’ve tried a bunch of different settings in OpenMC (electron treatment, weight cut-offs etc.) without being able to match the MCNP results.
Do you have any ideas what might be the cause? Clearly the pulse height treatment is different in mixed neutron-gamma fields in MCNP and OpenMC. We have verified and validated OpenMC PHT in a gamma only cases and it matches with MCNP and experiments perfectly.
p.s. Unfortunately I cannot share the experimental data in this forum.