Giovanni Nicodemo , Giovanni Zullo , Antonio Cammi , Lelio Luzzi , Davide Pizzocri
{"title":"Data assimilation for accelerated development of fuel performance codes: Towards automatic implementation of material properties correlations","authors":"Giovanni Nicodemo , Giovanni Zullo , Antonio Cammi , Lelio Luzzi , Davide Pizzocri","doi":"10.1016/j.jnucmat.2025.155954","DOIUrl":null,"url":null,"abstract":"<div><div>Developing fuel performance codes requires lengthy technical and legal verification procedures. This work proposes an approach to automate and facilitate these processes by using a data assimilation method. The method considered is Gaussian process regression, which allows for exploiting experimental data to correct/update material property correlations in fuel performance codes. As a demonstration of this method, the correlation for the single-atom xenon diffusivity is updated in the SCIANTIX code with recent lower-length scale results, without the need to modify the SCIANTIX code itself. The goal is to demonstrate the validity of this approach for data-driven developments of physics-based fuel performance codes.</div></div>","PeriodicalId":373,"journal":{"name":"Journal of Nuclear Materials","volume":"615 ","pages":"Article 155954"},"PeriodicalIF":2.8000,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Nuclear Materials","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0022311525003484","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"MATERIALS SCIENCE, MULTIDISCIPLINARY","Score":null,"Total":0}
引用次数: 0
Abstract
Developing fuel performance codes requires lengthy technical and legal verification procedures. This work proposes an approach to automate and facilitate these processes by using a data assimilation method. The method considered is Gaussian process regression, which allows for exploiting experimental data to correct/update material property correlations in fuel performance codes. As a demonstration of this method, the correlation for the single-atom xenon diffusivity is updated in the SCIANTIX code with recent lower-length scale results, without the need to modify the SCIANTIX code itself. The goal is to demonstrate the validity of this approach for data-driven developments of physics-based fuel performance codes.
期刊介绍:
The Journal of Nuclear Materials publishes high quality papers in materials research for nuclear applications, primarily fission reactors, fusion reactors, and similar environments including radiation areas of charged particle accelerators. Both original research and critical review papers covering experimental, theoretical, and computational aspects of either fundamental or applied nature are welcome.
The breadth of the field is such that a wide range of processes and properties in the field of materials science and engineering is of interest to the readership, spanning atom-scale processes, microstructures, thermodynamics, mechanical properties, physical properties, and corrosion, for example.
Topics covered by JNM
Fission reactor materials, including fuels, cladding, core structures, pressure vessels, coolant interactions with materials, moderator and control components, fission product behavior.
Materials aspects of the entire fuel cycle.
Materials aspects of the actinides and their compounds.
Performance of nuclear waste materials; materials aspects of the immobilization of wastes.
Fusion reactor materials, including first walls, blankets, insulators and magnets.
Neutron and charged particle radiation effects in materials, including defects, transmutations, microstructures, phase changes and macroscopic properties.
Interaction of plasmas, ion beams, electron beams and electromagnetic radiation with materials relevant to nuclear systems.