2023/10/12 by Peter P. Toma, Md Ali Muntaha, Toma, Peter +5 · 1 citation
Engineering · Materials Science · Physics and Astronomy · #Disordered Systems and Neural Networks (cond-mat.dis-nn) #FOS: Computer and information sciences #FOS: Physical sciences #Machine Learning (cs.LG) #Mesoscale and Nanoscale Physics (cond-mat.mes-hall) #Nuclear Materials and Properties #Nuclear Physics and Applications #Nuclear reactor physics and engineering
paper · pdf · doi:10.48550/arxiv.2310.08767
openalex publication_date 2023/10/12 · openalex created_date 2023/10/17 · openalex updated_date 2026/07/28
Mesoscale simulations of fission gas release (FGR) in nuclear fuel provide a powerful tool for understanding how microstructure evolution impacts FGR, but they are computationally intensive. In this study, we present an alternate, data-driven approach, using deep learning to predict instantaneous FGR flux from 2D nuclear fuel microstructure images. Four convolutional neural network (CNN) architectures with multiscale regression are trained and evaluated on simulated FGR data generated using a hybrid phase field/cluster dynamics model. All four networks show high predictive power, with R2 values above 98%. The best performing network combine a Convolutional Block Attention Module (CBAM) and InceptionNet mechanisms to provide superior accuracy (mean absolute percentage error of 4.4%), training stability, and robustness on very low instantaneous FGR flux values.