Conference Agenda

The high-temperature gas-cooled reactor (HTGR), combined with a helium turbine cycle, represents a cutting-edge application of advanced nuclear energy technology. A critical component in this system is the printed-circuit heat exchanger (PCHE), which features microchannels (1–2 mm wide) for efficient heat transfer. However, these systems face challenges, particularly from dust particles generated within the reactor. These particles tend to deposit on PCHE surfaces due to the microchannels' narrow dimensions and frequent turns, potentially degrading heat transfer performance and blocking the channels. Understanding the long-term effects of particle deposition on PCHE performance is essential for the sustainable operation of HTGR systems. This study investigates this issue using a near-wall drag model to incorporate shear flow effects and an EA rebound model to simulate particle deposition behavior. A dynamic mesh method was employed to track the evolution of deposition morphology over time. Key findings reveal a non-linear relationship between particle size and deposition fraction, with deposition increasing and then decreasing as particle size grows. Smaller particles deposit primarily at upstream bends, where flow dynamics promote adherence, while larger particles settle downstream after sufficient energy dissipation. Additionally, upstream deposition significantly reduces downstream particle accumulation, influencing deposition distribution patterns.

The shell-and-tube heat exchanger (STHE) is crucial for industrial safety and efficiency. Refined simulations of its operational characteristics often use three-dimensional (3D) and one-dimensional (1D) models to calculate outlet temperatures.3D models are complex and time-consuming meanwhile 1D models are simple but less accurate, and data-driven models lack interpretability and have limited application. To overcome these limitations, this paper proposes a hybrid "mechanism-data" driven model for STHE.Firstly, based on heat transfer and fluid mechanics, a 1D simulation model is built using MWORKS/Modelica, analyzing outlet temperature and pressure changes under steady, transient, and fault conditions. Secondly, operational data from STHE test benches, including flow rates, temperatures, pressures, and outlet temperature simulations from mechanism-driven models, are collected to build a data-driven model using deep learning algorithms. This model captures nonlinear relationships and dynamic characteristics, addressing the mechanism model's inability to observe and describe factors like corrosion and baffles.Combining both models, a hybrid "mechanism-data" driven model is established, offering physical interpretability and high accuracy. By simulating test bench operations in real-time, it detects potential faults, identifying their types and severity, supporting maintenance and management.Experimental validation shows the hybrid model outperforms single models in simulation accuracy and fault diagnosis, accurately reflecting STHE operational status and fault characteristics. Future work will optimize and enhance this model for broader applicability and accuracy across different conditions and STHE types.

The supercritical carbon dioxide (sCO₂) Brayton cycle, as a core component in the design of next-generation nuclear energy systems, emphasizes safety, operational efficiency, and non-proliferation characteristics. Within this framework, the Printed Circuit Heat Exchanger (PCHE) plays a key role in optimizing the heat transfer process under high temperature and high pressure conditions. This paper proposes a numerical design framework for the PCHE, focusing on the reduction of flow non-uniformity through a combination of secondary heads and porous baffles. Computational Fluid Dynamics (CFD) methods are employed to simulate the thermal-hydraulic performance of the heat exchanger, assessing the impact of different geometric parameters on flow distribution and heat transfer efficiency. The results demonstrate that the combination of secondary heads and optimized porous baffles significantly improves flow uniformity, thereby enhancing heat transfer efficiency and reducing pressure drop. This study provides valuable insights for optimizing the thermal-hydraulic performance of heat exchangers in supercritical CO₂ Brayton cycles.

Surrogate-based optimization (SBO) is a powerful approach for the design of the airfoil fin printed circuit heat exchanger (PCHE), which maximizes heat transfer and simultaneously minimizes pressure loss. Existing optimization studies of the airfoil fin PCHE commonly focus on the channel configuration and the fin arrangement. However, a meticulous optimization of the airfoil PCHE may consider the parameterization of the shape of the airfoil fin, while existing optimization methods are insufficient in such cases. To address this issue, the free form deformation (FFD) method is applied to parameterize the airfoil fin shapes and design variables are extracted to control the deformation of the shapes. The airfoil fins are divided into two groups according to the upstream and downstream of the channel. Fins within the same group deform synchronously. The heat transfer rate and pressure drop are employed as the objective functions of the optimization. To improve the optimization efficiency, the Kriging surrogate model is adopted to approximate the relations between the design variables and objective functions. Then, a multi-objective optimization using Non-dominated Sorting Genetic Algorithm-II (NSGA-II) is conducted and the Pareto solutions are obtained. Comprehensive optimal designs are selected on the Pareto front, and the thermal and hydraulic characteristics of the optimized designs have the advantage over those of the PCHE with original airfoil fin shape.

The water-cooled SMR is generally designed for flexible operation. In the performance analysis of nuclear power plants, system codes are used to evaluate plant performance under various Performance-related Design Basis Event (PRDBE) conditions. The load-following operation is one of the major PRDBE conditions. Many water-cooled SMRs employ a helical type once-through steam generator (SG), which produces superheated steam on the secondary side. Unlike conventional steam generators, the helical SG has no concept of water level, making superheated steam pressure a key control parameter in balancing plant pressure during load-following operation.

The helical SG comprises thousands of helical coils surrounding the riser, each with a different helical geometry. These geometric differences affect the heat transfer characteristics, potentially altering the outlet conditions. However, modeling every tube’s unique geometry would be inefficient. Instead, the SG is typically divided into multiple units, and system code calculations are performed on this segmented model.

This study explores how the fineness of the nodalization, or the number of divisions of the SG, affects performance analysis results. Using MARS-KS code, the reference helical steam generator is divided into 5, 10, 15, and 20 units, and steady-state calculations are performed for each case. The focus is on comparing the steam condition at the secondary outlet. It is expected that the outlet conditions of superheated steam are similar across different nodalizations, suggesting that coarse nodalization does not significantly impact analysis results. This is expected to allow for more efficient calculations in large-scale performance scenarios.