Engineering Papers⌕ Search

Engineering topics

Haeck, Wim

Publications and source records attributed to Haeck, Wim.

71 records · Page 4

Enhancing nuclear data validation analysis by using machine learning

In this work, we demonstrate how machine learning (ML) techniques can be used as an auxiliary tool for nuclear data validation analysis. The output of the ML analysis can inform evaluators and validators on the quality, or lack thereof, of specific nuclear data and benchmark experiments with respect to simulating these experimental benchmark values. To this end, measured and simulated effective neutron multiplication factors, k eff , of 875 selected ICSBEP critical assemblies and the latter's sensitivities with respect to nuclear data as well as benchmarks' features (e.g., material nuclide, core geometry) are used as input for a random forest (RF) regression model. The RF is used to encode the complex inter-dependencies between thousands of nuclear data features (i.e., sensitivity profiles and aspects describing the measurements) and 875 simulated and experimental k eff values in order to understand what nuclear data features are most informative for predicting bias. The complexity of relationships and high-dimensional space is difficult-to-impossible to search fully using simply expert judgment. As a first proof-of-concept—step, it is demonstrated that this technique is able to correctly trace large discrepancies between simulated and experimental k eff back to fabricated shortcomings in nuclear data that were perturbed to simulated k eff values. In a second, real-case scenario, step, the RF algorithm is used to validate the ENDF/B-VIII.0 library in comparison to ENDF/B-VII.1 nuclear data. One case is showcased where the chosen ML algorithms highlighted nuclear data (the 19 F(n,inl) cross section from 0.4–0.9 MeV) that are shown to be problematic by comparing them to associated differential experimental data and nuclear data from other libraries. In addition to that, it is shown that the RF results point towards poor benchmark experiments and associated underestimated uncertainties (e.g., the PU-SOL-THERM-028 series). However, using the RF algorithm for validating nuclear data with respect to k eff is currently limited to pinpointing groups of questionable nuclear data due to the inherent correlations between features introduced by the nuclear data themselves and how k eff is simulated. Due to this, we recommend that the ML methods presented be used to augment—rather than replace—the expert knowledge of evaluators and validators.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

FAUST - Benchmark and Validation Framework [Slides]

This presentation discusses Python packages for nuclear data applications and for benchmarking. It also discusses the goals and objectives for FAUST which include providing input and output processing for different calculation codes, allowing for exchanging results between different applications and codes, running benchmarks and processing the results, automating and simplify plot and report generation, and providing a basis for developing applications useful for nuclear data evaluators. The presentation also discusses how the packages can be used both inside and outside LANL.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

A simple json format for storing and exchanging code calculation and experimental results

Calculation codes and code systems give their results in a plethora of different formats and/or files. For instance, MCNP gives a user the MCNP specific output file, the mctal file, the ptrac file, etc. while PARTISN has a PARTISN specific output file and a number of binary files with various kinds of information about the calculation. However, the results we are interested in are always the same. These can be single values such as the effective multiplication factor k eff or the effective delayed neutron fraction β eff . It can be histogram data (such as particle spectra, reaction rates, sensitivity profiles, etc.) or even pointwise data (for example the nuclide composition as a function of time, etc.). It is also worthwhile to note that this would also apply to experimental data that we may wish to compare with as well. Calculation and experimental results can basically be split into two components: attributes (or metadata) and the actual result itself. Attributes represent information about the result that are not necessarily required for its’ interpretation but that are still good to have. Often, these attributes are what we will want to filter results on. This means that these attributes are often the answer to specific questions we may have about the result. The following is a non exhaustive list of such basic questions: what type of result is it: e.g. the effective multiplication factor k eff , the effective delayed neutron fraction β eff , etc.; which case does it relate to: e.g. a specific ICSBEP case such as PU-MET-FAST-001-001; which code and version produced this result: e.g. MCNP 6.2, etc.; which nuclear data library was used to obtain the result: e.g. ENDF/B-VIII.0 or JEFF 3.3; which nuclide and reaction does it apply to (for reaction rates and sensitivity profiles); which volume in the geometry was it calculated in? The second component of the calculation or experimental result is the actual numerical data of the result. It is this data that we want to compare, exchange, plot, interpret, etc. It consists of values with optional uncertainties and units. It also includes the underlying structure (if any structure is required to interpret the result). Depending on the type of result, it may be a single value (e.g. the effective multiplication factor), an array of values (e.g. a reaction rate or particle spectrum) or even a multidimensional array. In order to exchange results between different organisations, for instance for benchmarking, people often resort to spreadsheets or custom text files that need to be filled out. While this approach works well if little information is to be exchanged, it can become cumbersome when dealing with large amounts of data (like for instance particle spectra for thousands of cases). In addition, such an approach does not always capture the metadata we may be interested in to allow for automated processing of these files. The purpose of this report is to propose a standardised json structure to store any type of result and present an associated python interface to interact with these results directly. For such a structure to be useful, it must fulfill a specific set of functional requirements that we try to meet with this proposal. The format must allow for the calculation result to stand by itself. This means that we should not have to look in multiple places to be able to understand it. For example, if the result relates to a particle spectrum, the group structure in which the result is given should be stored with it. The format should be code agnostic. This means that we do not need to know where it came from to understand it. The format should be result type agnostic. This means that the format must be generic enough so that we can even use it for results we have not thought of yet. The format should be relatively easy to interact with through different scripting or programming languages (i.e. the machine interface) while still providing a clean structure that can be used directly by any human.

97 MATHEMATICS AND COMPUTING↗

Updating NJOY's HEATR module for ENDF/B-VIII.0 - 25493

With ENDF/B-VIII.0, new formats were added for allowing tabulated fission energy release components in addition to the existing thermal point and polynomial representations. Within ENDF/B-VIII.0, this new format was adopted for U{sup 235}, U{sup 238} and Pu{sup 239}. The fission energy release is a major component in nuclear heating for fissionable nuclides. Within NJOY, all heating and damage related calculations are done in the HEATR module. The introduction of the tabulated fission energy release components requires changes in HEATR so that these new data can be used. At the same time, the calculation of heating from fission using the other available formats was reviewed so that fission heating can now be calculated in a consistent way if the same data is used, independent of the format. For thermal point evaluations, this leads to differences between the old and new implementation of the order of 0.1 % up to incident energies of 100 keV, which increases gradually up to a value of the order of 1 % above 10 MeV (it should be noted that thermal point evaluations are no longer used in ENDF/BVIII.0). Polynomial evaluations still yield the same results with respect to older versions of NJOY. (author)

61 RADIATION PROTECTION AND DOSIMETRY↗

Nuclear Data Specifications, Utilities, and Libraries

Nuclear data provides descriptions of physical processes used by numerical simulation codes throughout the world. The data is provided in specified formats. Knowledge of the formats and what it represents is important for those that produce and read the data. Nuclear data is typically processed and made available to simulation codes at various temperatures applicable to the general class of problems that the simulation code is designed to model. These libraries are historically pre-processed and made available to the world. The libraries are based on fundamental physical data and are used by scientists and researchers.

Conlin, Jeremy↗

NJOY2016 updates for ENDF/B-VIII.0

Each release of the ENDF/B library often results in the introduction of new data formats in the ENDF-6 format. With ENDF/B-VIII.0 [1], which was released in February 2018, new formats were added for allowing tabulated fission energy release components in the MF1/MT458 section, to allow for sub-actinide fission and non-neutron induced fission in the MF8/MT18 and MF10/MT18 sections, and to add fission neutron and gamma emission probabilities in the MF6/MT18 section [2]. Processing codes must be adapted to ensure that they are capable of using these new data, or at the very least understand the new formats. NJOY2016, the production version of the nuclear data processing code developed at Los Alamos National Laboratory is no exception to this.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Release of ENDF/B-VIII.0-Based ACE Data Files

In February 2018, the National Nuclear Data Center released ENDF/B-VIII.0 in the standard Evaluated Nuclear Data Format (ENDF). This represents the advances made in nuclear data since the release of ENDF/B-VII.1 in 2011.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

New fission heating capabilities in the NJOY2016 HEATR module

Each release of the ENDF/B library often results in the introduction of new data formats in the ENDF-6 format. With ENDF/B-VIII.0, which has been released in February 2018, new formats were added for fission neutron emission probabilities in MF6/MT18 and for allowing tabulate fission energy release components in MF1/MT458. This last option has been used for U235, U238 and Pu239 to incorporate recent experimental data for the kinetic energy of the fission fragments.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

A Comparison of Monte Carlo and Deterministic Solvers for keff and Sensitivity Calculations

Verification and validation of our solutions for calculating the neutron reactivity for nuclear materials is a key issue to address for many applications, including criticality safety, research reactors, power reactors, and nuclear security. Neutronics codes solve variations of the Boltzmann transport equation. The two main variants are Monte Carlo versus deterministic solutions, e.g. the MCNP [1] versus PARTISN [2] codes, respectively. There have been many studies over the decades that examined the accuracy of such solvers and the general conclusion is that when the problems are well-posed, either solver can produce accurate results. However, the devil is always in the details. The current study examines the issue of self-shielding and the stress it puts on deterministic solvers. Most Monte Carlo neutronics codes use continuous-energy descriptions of the neutron interaction data that are not subject to this effect. The issue of self-shielding occurs because of the discretisation of data used by the deterministic solutions. Multigroup data used in these solvers are the average cross section and scattering parameters over an energy range. Resonances in cross sections can occur that change the likelihood of interaction by one to three orders of magnitude over a small energy range. Self-shielding is the numerical effect that the average cross section in groups with strong resonances can be strongly affected as neutrons within that material are preferentially absorbed or scattered out of the resonance energies. This affects both the average cross section and the scattering matrix.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Continuous-Energy Data Checks

The purpose of this report is to provide an overview of all Quality Assurance tests that have to be performed on a nuclear data set to be transformed into an ACE formatted nuclear data file. The ACE file is capable of containing different types of data such as continuous energy neutron data, thermal scattering data, etc. Within this report, we will limit ourselves to continuous energy neutron data.

99 GENERAL AND MISCELLANEOUS↗

Simulation of image detectors in radiology for determination of scatter-to-primary ratios using Monte Carlo radiation transport code MCNP/MCNPX

Purpose: The purpose of this study was to compare and validate three methods to simulate radiographic image detectors with the Monte Carlo software MCNP/MCNPX in a time efficient way. Methods: The first detector model was the standard semideterministic radiography tally, which has been used in previous image simulation studies. Next to the radiography tally two alternative stochastic detector models were developed: A perfect energy integrating detector and a detector based on the energy absorbed in the detector material. Validation of three image detector models was performed by comparing calculated scatter-to-primary ratios (SPRs) with the published and experimentally acquired SPR values. Results: For mammographic applications, SPRs computed with the radiography tally were up to 44% larger than the published results, while the SPRs computed with the perfect energy integrating detectors and the blur-free absorbed energy detector model were, on the average, 0.3% (ranging from -3% to 3%) and 0.4% (ranging from -5% to 5%) lower, respectively. For general radiography applications, the radiography tally overestimated the measured SPR by as much as 46%. The SPRs calculated with the perfect energy integrating detectors were, on the average, 4.7% (ranging from -5.3% to -4%) lower than the measured SPRs, whereas for the blur-free absorbed energy detector model, the calculated SPRs were, on the average, 1.3% (ranging from -0.1% to 2.4%) larger than the measured SPRs. Conclusions: For mammographic applications, both the perfect energy integrating detector model and the blur-free energy absorbing detector model can be used to simulate image detectors, whereas for conventional x-ray imaging using higher energies, the blur-free energy absorbing detector model is the most appropriate image detector model. The radiography tally overestimates the scattered part and should therefore not be used to simulate radiographic image detectors.

62 RADIOLOGY AND NUCLEAR MEDICINE↗