Neural Operator-Based Surrogate Model for CFD: Helical Coil Steam Generator in Small Modular Reactor

This paper proposes a neural operator-based surrogate model for real-time thermal-hydraulic simulation of helical coil steam generators in small modular reactors (SMRs). Traditional CFD provides high-fidelity flow analysis but incurs prohibitive computational cost for digital twin applications. The neural operator approach aims to learn the mapping between input parameters and flow fields, enabling rapid surrogate predictions while maintaining CFD-level accuracy. The research addresses the need for efficient simulation tools that support safe and efficient SMR operation within digital twin frameworks.