Search

The nuclear de-excitation process of fission fragments (FF) provides fundamental information for the understanding of nuclear fission and nuclear structure in neutron-rich isotopes. The variation of the prompt-neutron multiplicity, ν(A), as a function of the incident neutron energy (En) is one of many open questions. It leads to significantly different treatments in various fission models and implies that experimental data are analyzed based on contradicting assumptions. One critical question is whether the additional excitation energy (Eexc) is manifested through an increase of ν(A) for all fragments or for the heavy ones only. A systematic investigation of ν(A) as a function of En has been initiated. Correlations between prompt-fission neutrons and fission fragments are obtained by using liquid scintillators in conjunction with a Frisch-grid ionization chamber. The proof-of-principle has been achieved on the reaction 235U(nth,f) at the Van De Graff (VdG) accelerator of the JRC-Geel using a fully digital data acquisition system. Neutrons from 252Cf(sf) were measured separately to quantify the neutron-scattering component due to surrounding shielding material and to determine the intrinsic detector efficiency. Prelimenary results on ν(A) and spectrum in correlation with FF properties are presented.

In this paper, we discuss the development of a nuclear data evaluation pipeline, based around the TALYS code system. The pipeline focuses on the evaluation of the fast neutron energy range, above the resolved resonances. A strong focus in development lies on automation and reproducibility, as well as the efficient use of large-scale computational infrastructure, to enable rapid testing of new algorithms and modified assumptions. Several novel concepts for nuclear data evaluation methodology are implemented. A particular problem in evaluating the neutron-induced reaction cross-section using TALYS, relates to the intermediate energy range. While TALYS only predicts the smooth energy-averaged cross-section, experiments reveal unresolved resonance-like structures. In this paper, we explore ways to treat this type of model defect using heteroscedastic Gaussian processes to automatically determine the distribution of experimental data around an energy-averaged cross-section curve.

The Total Monte Carlo (TMC) technique has proven to be a powerful tool to propagate uncertainties in nuclear data to the uncertainty in macroscopic quantities, such as neutron fluxes at detector positions and the criticality of reactor cores. Nuclear data uncertainties can be used to create self-consistent sets of cross-sections. Each set contains files generated by variations of nuclear model parameters to properly fit the model to the nuclear data, accounting for their uncertainty. These files are called random files. The random files reflect the covariances of the nuclear data due to the uncertainties of the nuclear physics model parameters. TMC uses particle transport codes, such as MCNP, to transport particles through arbitrarily complex geometries. Each set of random files is used in a separate transport code run. This allows for the propagation of uncertainties in nuclear data, which otherwise could be hard to account for in the transport codes. However, particle transport techniques are well-known to be computationally expensive. The Half Monte Carlo (HMC) technique uses the random files of the TMC technique but does not rely on transport codes to propagate the uncertainties of nuclear data to the uncertainty of the sought macroscopic quantity. Instead, it uses pre-calculated sensitivity matrices to calculate the difference in a macroscopic quantity, given the difference of the random files relative to the best estimate of the nuclear data evaluation. In this work, we demonstrate how to use the HMC technique to calculate the uncertainty of macroscopic quantities in integral experiments for a set of random files relative to the best nuclear data evaluation. In this paper, we demonstrate how HMC can be used to incorporate integral experiments into an automated nuclear data evaluation. After applying the Bayesian Monte Carlo method in conjunction with the HMC technique and random files of uranium-235 from the TENDL library on the Godiva experiment, we conclude that the HMC technique gives similar results to that of the TMC technique: the mean value and the standard deviation of ∆keff is -6.30 pcm and 1220 pcm, respectively.