Kyung Mo Kim

Post-critical heat flux (post-CHF) heat transfer corresponds to a highly complex two-phase flow regime with important implications for the safety assessments of light water reactors. However, the available data for the dispersed flow film boiling (DFFB) regime remain limited, both in terms of the range of flow conditions studied and the availability of measurements for critical parameters, such as void fraction, vapor superheat, and droplet temperature. Furthermore, the complexity of the phenomenon, particularly due to the presence of entrained droplets, contributes to significant uncertainties in system-level safety analyses for loss of coolant accidents (LOCAs). To improve our understanding of the DFFB regime and reduce code and model uncertainties in predicting DFFB heat transfer, a sequence of experiments was conducted to achieve quasi-steady-state DFFB conditions. The experiments covered a mass flux range of 60–150 kg/m2-s and a system pressure range of 0.138–0.414 MPa. The measured wall heat transfer exhibits a clear dependence on the thermodynamic equilibrium quality, indicating a transition in the prevailing mechanism. Specifically, this transition reflects a shift from droplet–vapor interfacial heat transfer to convection-dominated heat transfer within the vapor phase. To further investigate this shift in heat transfer behavior, local flow parameters, including the vapor Reynolds number and void fraction, were analyzed using a one-dimensional two-fluid film boiling model. The model predictions were validated through comparison with void fraction measurements obtained using an X-ray radiography system. Furthermore, existing DFFB heat transfer correlations were subsequently benchmarked against the acquired experimental data to evaluate their predictive capacity and identify any inherent limitations. The benchmark results indicate that these correlations exhibit limited predictive performance, primarily due to lack of the regression robustness and insufficient incorporation of the physical effects associated with liquid droplets. To address these deficiencies, a new wall heat transfer correlation was developed based on the experimental database and the Reynolds analogy, incorporating the variation of the interfacial area concentration as a function of the droplet size. The proposed model demonstrates improved predictive performance for wall heat transfer in the DFFB regime, successfully capturing variations in the Nusselt number with respect to the vapor Reynolds number, system pressure, and mass flux.

The design of the compact fusion device, called K-CPD, is under preliminary design in Korea. One of potential options for the superconducting magnet system in K-CPD is the utilization of high-temperature superconductors (HTS) based on its ability generating higher magnetic fields compared to low-temperature superconductors (LTS). As a preliminary research preceding the K-CPD project, an agreement has been made to develop HTS conductors and an HTS model coil. Among the promising HTS materials, REBCO (Rare-Earth Barium Copper Oxide) tape is widely used in high-field magnet applications, including fusion reactors. However, cabling and twisting with REBCO tape is significantly challenging due to its structural properties. A potential approach to address these challenges is the Stacked in Conduit Conductor (SICC) design, which consists of non-transposed and stacked REBCO tapes; thereby the design and jacketing process of the SICC are currently under evaluation in Korea. To preliminary ensure reliability in the winding process of the SICC for coil fabrication, extensive structural analyses have been conducted with improving our understanding on deformation behaviors. Therefore, this study aims to understanding the deformation behaviors during the winding process of the SICC design and effect of bending radius on the stress-strain distribution and structural variations to provide key insights for improving HTS application in K-CPD coil fabrication.

Nanofluid (NF) pool boiling experiments have been conducted widely in the past two decades to study and understand how nanoparticles (NP) affect boiling heat transfer and critical heat flux (CHF). However, the physical mechanisms related to the improvements in CHF in NF pool boiling are still not conclusive due to the coupling effects of the surface characteristics and the complexity of the experimental data. In addition, the current models for pool boiling CHF prediction, which consider surface microstructure characteristics, show limited agreement with the experimental data and do not represent NF pool boiling CHF. In this scenario, artificial intelligence tools, such as machine learning (ML) regressor models, are a very promising means of solving this nonlinear problem. This study focuses on creating a new model to provide more accurate NF pool boiling CHF predictions based on pressure, substrate thermal effusivity, and NP size, concentration, and effusivity. Three ML models (supporting vector regressor—SVR, multi-layer perceptron—MLP, and random forest—RF) were constructed and showed good agreement with an experimental database built from the literature, with MLP presenting the highest mean R² score and the lowest variability. A systematic methodology for optimizing the ML models is proposed in this work.