Abstract Spray cooling is an effective method for achieving rapid heat removal in high-temperature applications such as metal quenching, nuclear reactor safety systems, and electronic thermal management. In this study, transient spray cooling experiments were conducted on a large-diameter horizontal cylinder to analyze heat transfer behavior above the Leidenfrost temperature. The surface temperature during quenching was recorded at high frequency and used to determine the local heat transfer coefficient using an energy balance approach. Experiments were performed for different nozzle-to-surface distance ratios (Z/D), and empirical correlations were developed through regression analysis. The results showed that the heat transfer coefficient strongly depends on Z/D but exhibits minimal variation in the temperature exponent. The proposed correlation accurately predicts heat transfer coefficients within ±25 % of experimental data. The developed correlation extends the applicability to large horizontal surfaces under transient spray cooling conditions, offering practical relevance for industrial quenching and thermal management systems.
Abstract The Fukushima-Daiichi accident renewed interest in accident-tolerant fuel (ATF) systems that offer extended safety margins during loss-of-coolant events. This paper presents a Monte Carlo eigenvalue assessment of four fuel-cladding configurations in a BEAVRS 17 × 17 PWR-type assembly at beginning-of-life, using OpenMC 0.15.3 with the ENDF/B-VIII.0 nuclear data library. The assembly model uses a BEAVRS-based 17 × 17 PWR assembly model with light-water coolant; soluble-boron-free SMR designs require separate spectral and control analysis. The configurations are: (A) UO 2 /Zircaloy-4, (B) U 3 Si 2 /Zircaloy-4, (C) UO 2 /SiC, and (D) U 3 Si 2 /SiC. Forty eigenvalue calculations characterize k∞ across 3.5–5.0 wt% 235 U, the Doppler temperature coefficient (αD), the moderator temperature coefficient (αM), neutron spectrum, four-factor decomposition, and pin-power distribution at nominal conditions (4.5 wt%, T fuel = 900 K, T cool = 573 K). SiC cladding provides a reactivity gain of +332 to +344 pcm through reduced parasitic absorption and spectral softening, partially offsetting spectral hardening from U 3 Si 2 fuel. U 3 Si 2 incurs a reactivity penalty of −635 to −704 pcm from spectral hardening and carries a lower melting point (∼1,665 °C versus ∼2,865 °C for UO 2 ), reducing thermal safety margins during transients. All four configurations maintain strongly negative αD (−1.543 to −1.590 pcm K −1 ) and αM (−26.6 to −30.3 pcm K −1 ), confirming inherent neutronic safety feedback. The pin-power peaking factor for Config D is F q = 1.062. These results establish a quantitative neutronic baseline for ATF deployment in PWR-type SMR assemblies; a complete safety evaluation requires thermal-hydraulic and depletion assessments.
Abstract Accurate prediction of two-phase pressure drop is essential for the design of the tubular heat-exchange equipment used throughout energy-conversion and thermal-hydraulic systems, yet classical empirical correlations generalize poorly across fluids and operating conditions. This study develops a metaheuristic-driven neural architecture search for predicting the pressure drop due to evaporation and condensation of refrigerants in smooth and micro-fin tubes. Eight state-of-the-art metaheuristic optimizers, each enhanced with the Chaotic Sech–Tanh Dynamic Opposition-based Learning strategy developed in a companion study, were used to optimize the architecture and hyperparameters of a deep neural network trained on an experimental database, comprising 946 measurements (626 evaporation and 320 condensation points). The optimized networks were benchmarked against nine established machine learning regression models. A network with three hidden layers of 50 neurons achieved a mean absolute error (MAE) of 0.157 and a coefficient of determination ( R 2 ) of 0.99446 on the evaporation data, with 88.6 % of predictions falling within the 5 % error band. The opposition-enhanced optimizers consistently outperformed their standard counterparts: the improved Marine Predators variant was the most accurate for evaporation (MAE 0.224, R 2 0.98977), and the improved Mantis Search variant was the most accurate for condensation (MAE 0.957, R 2 0.98347). Among the machine learning models, extreme gradient boosting attained the highest training accuracy ( R 2 up to 0.99826) but showed signs of overfitting, whereas random forests generalized best.
Abstract Microreactors are distinguished by their compact configuration, cost-effectiveness, and adaptable deployment, which has prompted extensive research attention worldwide. In the realm of reactivity control methodologies, the utilization of drums has been identified as a more efficacious approach when it comes to the compactness of microreactors. The employment of burnable poisons in microreactors has been demonstrated to be an effective strategy for reducing excess reactivity, thereby ensuring compliance with the reactivity control requirements of control drums. In this paper, neutronics calculations and performance analysis of the effect of burnable poisons on a 3 MW yttrium hydride-moderated micro-reactor are carried out using RMC with ENDF/B-VIII database. The neutronic performance of single-layer and double-layer burnable poisons were analyzed, and the results showed that 2 μm-thick ZrB 2 and 4 μm-thick Gd 2 O 3 restrict the maximum excess reactivity to 3,888.2 pcm. Based on the optimized burnable poison, further reactivity control analysis shows that the sixteen core control drums can be divided into two independent reactivity control systems. It is therefore concluded that both systems satisfy the lifetime reactivity control requirements. It is demonstrated that this provides a technically feasible pathway for the reactivity control of in microreactors.
Abstract The liquid breeding blanket is a critical component of magnetically confined fusion reactors, with corrosion fatigue induced by pulsed operation and the high-temperature lead-lithium environment constituting a primary failure mechanism. This study proposes a stress-coupled corrosion-fatigue life prediction and reliability evaluation framework applicable to fusion liquid breeding blankets. A unified corrosion-fatigue damage model was developed for reduced activation ferritic-martensitic (RAFM) steel in high-temperature lead-lithium environments, explicitly incorporating the effects of temperature, flow velocity, stress, and damage evolution. Bayesian statistical methods were employed to estimate the posterior distributions of key model parameters from experimental data, thereby quantifying parameter uncertainty. For a typical outboard equatorial module of the China Fusion Engineering Test Reactor (CFETR) supercritical CO 2 -cooled lead-lithium (COOL) blanket, the life distribution and reliability metrics of the blanket were evaluated by integrating the corrosion-fatigue damage model, COOL blanket operational loads, and Monte Carlo simulations. Computational results indicate an average failure time of 8.25 years for the blanket, with a 90 % confidence interval ranging from 4.94 to 12.72 years. The influence of material-environmental parameters was assessed through sensitivity analysis. The results can provide essential input parameters for Reliability, Availability, Maintainability, and Inspectability (RAMI) studies of blankets, offering insights to potentially improve CFETR duty cycle.
