Advanced data processing of ATF claddings

Digital image processing (DIP), neural networks, and artificial intelligence (AI) are revolutionizing materials science, enabling precise and efficient analysis of microscopic features. From nuclear fuel inspections to advanced microscopy studies, DIP has become a cornerstone of material analysis to obtain relevant data quality. At CVR, integrating DIP and AI has streamlined processes, enhanced data reliability, and provided valuable insights in areas such as various microscopy studies (SEM, TEM), reactor shielding evaluations and nuclear fuel inspections based on image data processing with different resolution A critical application of DIP is the detection of precipitates—microscopic features that affect material properties. We can recognize the nature of precipitates, e.g. the secondary phase particles (SPP)-precipitates which are specifically present in the microstructure from the manufacturing processes/heat treatment, or the radiation-induced precipitates (RIP) formed in the microstructure by diffusion processes caused by fast neutron irradiation. Size, shape, type and distribution of both types of precipitates influence material behaviour under stress, precipitates increase hardness, make the material more brittle and can create a stress field. By controlling of SPPs formation through alloy composition and heat treatment, researchers can optimize material properties. The RIPs distribution is very important part of Post Irradiation Examination (PIE). In SEM/TEM, DIP is essential for segmenting precipitates in construction materials of nuclear power plants and conducting statistical analysis before and after irradiation. DIP minimizes repetitive tasks, reduces human error, and ensures consistent results, making SEM data more reliable and reproducible. DIP also plays a crucial role in analysing biological shielding concrete, which endures thermal, gamma, and neutron flux in nuclear reactors. Over time, these exposures cause cracking, especially at the interface between coarse aggregate and hardened cement paste, compromising properties like strength and stiffness. Monitoring crack formation is key to understanding degradation mechanisms. At CVR, DIP provides precise crack tracking and volumetric damage analysis, which can validate other methods as ultrasonic testing (UT), offering a non-invasive way to assess structural integrity. In nuclear fuel inspections, DIP processes video data to reconstruct high-resolution images of fuel assemblies, verifying key parameters such as bow, twist, and growth with 0.2 mm accuracy. These measurements ensure nuclear safety and operational reliability. Long-term inspections using DIP enable core behaviour verification across different fuel designs, with terabytes of data generated annually. DIP optimizes workload and ensures consistency in results over time. Nuclear cladding materials, structural components, shielding concrete, and nuclear fuel are vital for the long-term operation (LTO) of nuclear power plants. DIP technologies developed at CVR support LTO policies by enhancing material analysis and quality assurance. Beyond nuclear research, DIP has broad applications in non-nuclear fields. Industries requiring microscopic analysis, such as crack tracking in concrete or defect detection in alloys, can benefit from DIP's precision and efficiency. DIP provides versatile solutions for challenges across a wide range of materials. At CVR, advancing DIP technologies remains a priority in the nuclear research and development, where DIP is transforming material science, optimizing workflows, improving data quality, and driving innovation in both nuclear and non-nuclear industries. We acknowledge the state support of the Technology Agency of the Czech Republic within the National Competence Centre Programme II, project TN02000012 „Center of Advanced Nuclear Technology II“.