Heat transfer coefficient for upward forced convective flows of heated supercritical carbon dioxide in vertical tubes

Accurate heat transfer prediction is crucial for optimizing supercritical power cycles. This study presents new Nusselt number correlations for forced convection heat transfer of supercritical carbon dioxide flowing upward in heated tubes. Existing correlations often suffer from reduced accuracy near the pseudocritical point. The study addresses this challenge by employing a systematic correlation modelling framework to develop region-specific correlations tailored to distinct fluid regions, namely liquid-like, near-pseudocritical, and gas-like regions. A novel interpolation methodology utilizing sigmoid functions is implemented to ensure smooth transitions between these regions. Furthermore, stability functions based on kinematic viscosity are introduced to enhance the stability of the correlations during iterative processes. The resulting three-variable correlation, incorporating the Reynolds number , Prandtl number , and a stability function, demonstrates significantly improved accuracy relative to existing correlations, achieving a maximum percentage error of 52 % and a mean absolute percentage error of 11 %. This work provides valuable tools for the design and optimization of supercritical power cycles, particularly during transient events in which precise heat transfer predictions are essential. • Region-specific correlations address accuracy issues near the pseudocritical point. • Sigmoid interpolation ensures smooth transitions between fluid regions. • Kinematic viscosity-based stability functions enhance correlation stability. • Interpolated three-variable correlation achieves an 11 % MAPE. • Challenges associated with zero-root and multiple-root issues are addressed.