A prototype multi-instrument quality assurance system for responsive nuclear fuel manufacturing

A. Parker, M. Bandala, P. Chard, N. Cockbain, D. Dunphy, D. Eaves, D. Hutchinson, X. Ma, S. Marshall, P. Murray, P. Stirzaker, J. Taylor, J. Zabalza, M. Joyce

Research output: Chapter in Book/Report/Conference proceedingConference contribution book

Abstract

A prototype apparatus for in line quality control assessment of sintered uranium dioxide fuel pellets has been demonstrated. In this work we combine γ-ray spectrometer, laser profilometer, and hyperspectral and RGB cameras in a single system. The combination of instruments provides analytical data on uranium enrichment, pellet geometry, physical defects, and potential chemical contamination. The images and spectra are analyzed using a convolutional neural network and an automated quality control decision taken. The prototype was deployed at the National Nuclear Laboratory, Preston site to analyze Advanced Gas Cooled sintered UO 2 pellets. Results from the pellet analysis are added automatically to a pellet data library for future nuclear security and safeguards auditing and model-training to improve quality assurance decision-making.
Original languageEnglish
Title of host publication2024 IEEE Nuclear Science Symposium (NSS), Medical Imaging Conference (MIC) and Room Temperature Semiconductor Detector Conference (RTSD)
Place of PublicationPiscataway, NJ
PublisherIEEE
Number of pages2
ISBN (Electronic)9798350388152
ISBN (Print)9798350388169
DOIs
Publication statusPublished - 25 Sept 2024
Event2024 IEEE Nuclear Science Symposium, Medical Imaging Conference, and Room-Temperature Semiconductor Detectors Symposium - Tampa, United States
Duration: 26 Oct 20242 Nov 2024

Publication series

NameIEEE Symposium on Nuclear Science (NSS/MIC)
ISSN (Print)1082-3654
ISSN (Electronic)2577-0829

Conference

Conference2024 IEEE Nuclear Science Symposium, Medical Imaging Conference, and Room-Temperature Semiconductor Detectors Symposium
Abbreviated title2024 NSS MIC RTSD
Country/TerritoryUnited States
CityTampa
Period26/10/242/11/24

Keywords

  • convolutional neural network
  • hyperspectral
  • combination of instruments
  • physical defects
  • uranium enrichment
  • nuclear security

Fingerprint

Dive into the research topics of 'A prototype multi-instrument quality assurance system for responsive nuclear fuel manufacturing'. Together they form a unique fingerprint.

Cite this