Comparison of sensitivity calculation between mathematical and physical adjoint in the 2-D/1-D transport solver KYADJ. (August 2019)

• Record Type: Journal Article
• Authors: Wu, Qu; Xu, Fei; Peng, Xingjie; Yu, Yingrui; Li, Qing; Wang, Kan
• Is Part Of: Annals of nuclear energy. Volume 130(2019)
• Page Start: 357
• Page End: 364
• Publication Date: 2019-08
• Subjects: 2-D/1-D transport solver -- Mathematical adjoint flux -- Physical adjoint flux -- Sensitivity and uncertainty analysis; Nuclear energy -- Periodicals; Nuclear engineering -- Periodicals; 621.4805
• DOI: 10.1016/j.anucene.2019.03.004
• Languages: English
• ISSNs: 0306-4549

Abstract

Highlights: The physics adjoint and the mathematical adjoint are compared based on the 2-D/1-D solver. Different adjoint flux functions are used to perform S&U analysis for comparison. The SF96 problem and the PB-2 assembly problem are selected for verification. Abstract: Sensitivity and Uncertainty analysis (S&U) based on the Classical Perturbation Theory (CPT) requires the solution of the adjoint flux. The adjoint flux is used as a weight function for sensitivity calculation. In the previous study, the MOC-based adjoint flux in the 2-D/1-D transport solver was given and applied to S&U analysis in the UAM benchmarks. The eigenvalue k obtained by the forward calculation is different from that obtained by the adjoint calculation. In order to illustrate the phenomenon, three different adjoint flux functions, the MOC physics adjoint, the CMFD physics adjoint, and the CMFD mathematical adjoint are solved based on the 2-D/1-D solver. Next, three different adjoint flux functions are used to perform S&U analysis. The SF96 problem and the PB-2 assembly problem are selected for verification. Recommendation is made for the proper application of the CMFD mathematical adjoint flux to sensitivity calculation due to less computation time.