A new PCI analysis methodology based on artificial neural network. Issue 3 (4th March 2021)

• Record Type: Journal Article
• Authors: Wang, Yang; Wei, Jun; Wang, Jie; Wang, Kai-Yuan; Deng, Yong-Jun
• Is Part Of: Journal of nuclear science and technology. Volume 58:Issue 3(2021)
• Page Start: 333
• Page End: 346
• Publication Date: 2021-03-04
• Subjects: Fuel rod -- Pellet-Cladding Interaction -- PCI -- Artificial neural network -- ANN; Nuclear engineering -- Periodicals; Nuclear physics -- Periodicals; Nuclear energy -- Periodicals; 621.4805
• DOI: 10.1080/00223131.2020.1831998
• Languages: English
• ISSNs: 0022-3131

Abstract

ABSTRACT: Pellet-Cladding Interaction (PCI) is one of the most important issues in fuel rods of a Pressure Water Reactor (PWR) which may cause the failure of fuel rods, and it should be analyzed under operating transients to ensure the fuel rods' integrity. The traditional PCI analysis method simulates the detailed physical phenomena that occurred in fuel rods and it considers the combined effects of irradiation, thermodynamics, and mechanics, which consumes a large amount of time and computational resources. A two-stage artificial neural network model has been developed by China General Nuclear Power Corporation (CGN), which provided a new solution for the PCI analysis. This approach employs two data-driven models basing on artificial neural networks. It takes a very small amount of time and computational resources, while the accuracy is comparable to the traditional method.