[Feature request] - C/E refinement

[milocco]

JADE currently calculates C for the C/E using differential spectra (by_energy option before condense_groups in yaml); calculation of C (and E) in coarse binnning should be carried out as sum of the bin integral flux for every bin included in the limits of the coarse binning of the condense_groups, which is though read as a string and needs numerical conversion.

The C/E evaluation would hold for:

• direct energy spectra;
• TOF spectra, after converting the energy bins into time bins and before converting differential spectra from time domain to energy domain;
• Differential reaction rates, that is the tally in FM card.

The JADE calculation may be validated via MCNP runs with fine and coarse binning, as in following tallies provided for energy spectra of RCR SS exp and TOF spectra of the IPPE DT Fe exp and made available through the F4E sharepoint accessible to Davide/Marco.

RCR SS
IPPE DT

[mcampos16]

@milocco Can you please confirm if I am understanding the issue correctly?
These are my takes from the description above:

• The issue described affects only the benchmarks in JADE which have a coarse binning result and, therefore, the ones for which the condense_groups tally modifier is applied.
• The issue is that for the coarse binning results we are normalizing by energy (or by lethargy) before summing (so, as you mentioned, summing differential spectra). Instead, we should be directly summing bin integral flux values (non normalized by energy) and then, if needed, apply the normalization by energy (or lethargy). The normalization by energy or lethargy after condensing would require a small modification, as you pointed out, because the energy bins are read as string instead of numerical values.
• The issue described in the previous point should be considered both for computational and experimental data.
• No further action should be taken on the rest of results that are not using the condense_groups tally modifier but are normalized by energy or lethargy.

Thank you for your support!

[milocco]

Hi Marta, agree on any point, this is what I meant when opening the issue. Would like to top up with a minor one, that is plotting C/E in coarse bins: it should be the same if integral or differential (normalized by energy bin), but for plotting only could that make any difference? Indeed, the chance to plot the differential spectra in fine binning as usual and the C/E ratio underneath the main plot in coarse binning could be an option.