Conference Agenda

In a specific model of nuclear power plant, the cooling water source for the Steam Generator Blowdown System's (TTB) regenerative heat exchanger is characterized by low undersaturation. This condition can easily lead to the occurrence of Two-Phase flow upon encountering disturbances. The resulting Two-Phase water hammer can cause vibrations that lead to pipe system failure and equipment damage. Severe vibrations induced by condensation-induced water hammer can occur in the return line of the cooling water to the deaerator inlet, potentially leading to conventional island faults. This paper first establishes a TTB system model using FLOMASTER software and conducts simulation calculations to identify the fluid medium conditions that are prone to condensation-induced water hammer. Subsequently, a mathematical model for condensation induction is established based on fluid characteristics, and local Two-Phase water hammer simulation analysis is conducted using FLUENT software. This study investigates the mechanism behind the original design scheme's condensation-induced water hammer and characterizes the pressure oscillation conditions and mass transfer characteristics associated with the destructive intensity of the water hammer. Sensitivity analyses are performed on parameters such as the pipe diameter, inlet flow velocity, temperature, and pressure of the cooling water pipeline. Finally, based on the aforementioned research findings, several feasible solutions are proposed to improve the original design and prevent the occurrence of Two-Phase water hammer. This study offers a universal approach to preliminarily assess the potential for condensation phase change water hammer in nuclear power plant fluid systems, enhancing the overall safety of the plant.

Accurate modelling of natural convection is essential for nuclear safety applications, particularly in passive cooling systems of Small Modular Reactors (SMRs). Reynolds-Averaged Navier- Stokes (RANS) models, widely used in industry, often fail to capture key buoyancy effects, limiting their accuracy. This study presents a new transport equation model for temperature variance dissipation, εθ, specifically designed for imposed temperature boundary conditions. The model aims to improve the estimation of εθ and the time scale ratio, which is crucial for thermal turbulence modelling. The proposed model is first evaluated a priori using Direct Numerical Simulation (DNS) data from Flageul et al. and then validated a posteriori for forced and mixed convection cases using DNS data from Abe et al. and Kasagi and Nishimura, respectively. Results show that the model provides a significantly better prediction of εθ near the wall and in the buffer layer, while also improving the estimation of temperature variance throughout the flow. In mixed convection, it effectively captures buoyancy effects and reduces errors compared to existing algebraic models.

Boiling heat transfer is an effective cooling technique that utilizes the high latent heat from phase change. The most critical parameter in boiling heat transfer is the Critical Heat Flux (CHF). CHF represents the maximum heat flux through nucleate boiling, at which a vapor film forms on the surface, impeding heat transfer and causing a rapid increase in surface temperature, potentially leading to surface damage. In previous studies, infrared thermometry technique was used to analyze the temperature distribution at the CHF on surfaces, measuring the 2D temperature distribution on surfaces. However, infrared thermometry technique has limitations: the substrate must be opaque to infrared, which restricts material choice, and the infrared camera must observe from below the substrate, complicating experimental setup and limiting use under nuclear reactor conditions. This study aims to overcome limitations of infrared visualization techniques by using fiber optic sensors to measure the 2D temperature distribution on surfaces. Fiber optic temperature sensors are easy to install, either by embedding them in surface grooves or inserting them into capillary tubes, can withstand high-temperatures of 600-700°C, and can measure thousands of temperature points simultaneously with high resolution (1mm between points, 100 Hz measurement speed), making them a promising alternative to infrared thermometry technique. In this study, high-resolution fiber optic sensors were used to measure the 2D temperature distribution on surfaces in real-time during boiling heat transfer. Measurements were conducted on micro-pillar surfaces and flat plates, with results showing a significant improvement in CHF on micro-pillar surfaces compared to flat plate.

Bubble growth and departure are critical phenomena in applications such as boiling heat transfer on heated surfaces and gas generation during water electrolysis. These processes play a significant role in engineering systems requiring efficient phase-change heat transfer and fluid management. While bubbles under static conditions have been widely studied, bubble behavior in dynamic environments involving acceleration remains less understood. Such conditions frequently occur in real-world scenarios, including mechanical vibrations in industrial equipment, seismic events, and rapid accelerations during emergency maneuvers. Understanding bubble growth and detachment under these circumstances is essential for improving system performance and ensuring reliability.

This study experimentally investigates bubble growth and departure on a plate subjected to controlled linear acceleration. High-speed imaging captures the dynamic evolution of bubbles from nucleation to detachment. The applied acceleration alters the force balance on the bubble, influencing its growth rate, interface deformation, and departure characteristics. Key parameters such as detachment radius and frequency are measured and analyzed under various acceleration conditions. The results show that acceleration-induced forces lead to asymmetric bubble shapes and earlier detachment, which could significantly impact heat transfer and fluid transport performance.

This research provides insights into bubble dynamics in accelerated environments, enhancing the understanding of multiphase flow behavior. The findings contribute to the development of more efficient and reliable thermal management systems in dynamic conditions, addressing critical challenges in engineering applications involving non-static settings.

The boron concentration stability of the Reactor Coolant System (RCS) is crucial for the safe operation of nuclear power plants. During the dilution operation, some equipment has a strong fluid retention effect on the replenished deionized water. This effect leads to the RCS boron concentration fluctuations, making it difficult to stabilize the RCS boron concentration. This paper utilizes a coupled model based on the lumped parameter method and computational fluid dynamics (CFD) method to study the dynamic response of the RCS boron concentration during the dilution process. The spatiotemporal distribution of boron concentration is first analyzed. Then, the effects of flow rate ratio, initial concentration difference, and initial temperature difference on the dynamic response of RCS boron concentration are analyzed. The results show that significant concentration stratification appears in the Volume Control Tank (VCT). A large amount of water is retained in VCT and then slowly released into RCS. Due to this phenomenon, the boron concentration of RCS takes a long time to reach the target value. Besides, the flow ratio and initial temperature difference should be increased to shorten the completion time in practice. The effect of the initial concentration difference is almost negligible.

Many applications involving in-line check valves in nuclear power plants require customized designs that deviate from standard offerings. These customizations require tradeoffs in performance, such as fast closure to mitigate water hammer effects, versus low flows to open the valve, reducing pressure losses. One lesser discussed phenomenon faced in the customization process is the reverse force on the valve obturator, pulling it into the flow direction, rather than the intuitive push in the direction of opening the valve. This force can be encountered when certain geometric and flow parameters coincide. To better understand the correlation, a select number of geometric parameters were varied using a sphere-in-pipe configuration, in a design-of-experiments (DOE) consisting of 81 configurations, intended to replicate the actuation forces on a check valve obturator, where the reverse force may be encountered. CFD simulations determined the force on the obturator, while maintaining a constant Reynolds number across the DOE cases. A mesh convergence study was performed, and steady state results were validated against transient results. The normalized force was mapped in response surfaces as a function of the normalized geometric parameters. Results showed a strong correlation between the obturator diameter and the orifice diameter creating zones of reverse force, suggesting what design configurations to avoid, while the correlation of the distance between the obturator and the orifice to the reverse force was weaker.

Accurate and efficient calculation of the fuel rod resonance cross section (XS) in the case of non-uniform temperature distribution is an important challenge in numerical reactor physics calculation when considering the neutronics and thermal hydraulics coupling. This paper studies the deep learning based global-local coupling resonance calculation method, which can accurately and efficiently calculate the effective self-shielding XS under the condition of non-uniform temperature distribution in the fuel rod. Through theoretical and data analysis, the input parameters and output parameters required for deep learning are obtained and the deep learning model is generated by training. The deep learning model is applied to the calculation of the local effective self-shielding XS in the global-local coupling resonance calculation method for considering the non-uniform temperature distribution in the fuel rod. Combined with the 2D/1D coupled transport method and the thermal hydraulics code, high-fidelity core neutronics and thermal hydraulics calculations are realized. The numerical results show that, compared with the calculation results directly using the ultra-fine group method, the calculation efficiency is increased by more than 20 times while maintaining the calculation accuracy.

The PAFS is passive cooling system to replace the conventional auxiliary feedwater and this system was adpoted as one of the cooling systems of the APR1000 and iSMR. This system removes decay heat from the reactor core by cooling down the secondary system of steam generator using a PCHX installed in the PCCT which has the role of ultimate heat sink. Due to the design characteristic, the cooling performance of PAFS is determined by boiling heat transfer outside PCHX and condensation heat transfer inside. In this paper, the prediction capability of SPACE and MARS-KS codes which currently used for thermal-hydraulic analysis in Korea were evaluated for cooling performance of PAFS. For this study, one of the IETs for PAFS performed by KAERI was selected. To evaluate the prediction capability of system codes, sensitivity calculations were performed using various boiling and condensation model options embedded within codes. As a result, it was confirmed that the system codes conservatively under-predict the heat transfer derived from the PAFS experiment, and it was concluded that further studies were needed to increase the accuracy in the future.

Boiling models are typically validated on greater than 5 mm hydraulic diameter channels as found in shell and tube boilers, or on microchannels with diameters less than 0.5 mm for CPU cooling. However, for the application of the Compact Steam Generator (CSG), which utilizes Printed Circuit Heat Exchanger (PCHE), the flow channels in PCHE are in the range of 1-2 millimeters, categorized as mini-channels. There is limited research on boiling in channels of this size. Due to the rapid advancement of simulation tools, particularly Computational Multiphase Fluid Dynamics (CMFD), the Volume of Fluid (VOF) interface tracking method has the potential to become a powerful tool for studying two-phase flow regimes in non-conventional flow diameters, such as mini-channels. In this study, both the experiment and simulation are performed and cross-validated against each other. R-134a is selected due to its similarity in density ratio to water under CSG conditions. A test loop has been built to conduct the experiment using a single mini-rectangular channel in a vertical configuration. The experiment measures pressure drop, and captures high-speed video of the flow patterns. The results show that the VOF simulation is capable of reproducing good bubble structures and void fraction distribution similar to those observed in the experiment. However, finer mesh sizes may be necessary to resolve small bubbles or droplets in flow regimes such as low-void bubbly flow or mist flow. The validated simulation results are then used to develop models applicable to boiling in mini-channels.

Small Lead-based cooled fast reactors have advantages of large natural circulation ability, great inherent safety and compact reactor structure, making it one of the main candidates of small modular reactors to produce electricity for communities or islands. However, most of present system analysis codes used for Small Lead-based cooled fast reactors are developed by modifying system analysis codes for pressurized water reactor or sodium cooled fast reactor. In order to improve the applicability and extensibility of system analysis code for small Lead-based cooled fast reactors, a multi-domain system analysis code is self-developed and validated in this paper. First, the physical models including the properties models, the flow and heat transfer models, power calculation models, steam generator models, main pump and so on are established. Then, a system analysis code named SASLFR aiming for the system performance analysis and safety evaluation of small Lead-based cooled fast reactors has been developed. Using the object-oriented programming language Modelica, SASLFR has characteristics of supporting modularized, multi-field physical model unified modeling, which greatly improves the system modeling efficiency, applicability and extensibility. Furthermore, SASLFR has been validated with variety of test problem covering fundamental models test and integration effect test. The simulation results agree well with the available analytical results and experimental data, which demonstrate that SASLFR has the capacity of simulating the basic physical process and complex system characteristics of small Lead-based cooled fast reactor. Hence, SASLFR can be used for the design optimization and safety evaluation of small liquid metal cooled fast reactors.

KAERI (Korea Atomic Energy Research Institute) has been operating an IET (Integal Effect Test) facility, which ATLAS (Advanced Thermal-Hydraulic Test Loop for Accident Simulation) with reference to the APR 1400 (Advanced Power Reactor 1400) for experiments for transient and DBAs (design basis accidents). An experiment for MSGTR (Multiple Steam Generator Tube Rupture) with failire of AFWS (Auxiliary FeedWater System) had been conducted at the ATLAS. The purpose of the experiment was to resolve a safety issue that multiple failure accident shus as loss of AFWS during MSGTR could lead to the damage of the core. Thus, the experiment aims at evaluating the importance of the PAFS(Passive Auxiliary Feedwater System) operation during postulated accident. The PAFS is adopted as one of the cooling systems i-SMR(innovative Small Modular Reactor) as well as passive cooling system to replace the conventional active AFWS. In this study, the experiment has been analyzed by Korea thermal hydraulic system analysis codes, which SPACE and MARS-KS the comparison with experimental and calculation results have been performed. In addition, the evaluation of PAFS cooling capability prediction will be disscussed. The main objective of this study is the investigation of the thermal-hydraulic phenomena during a MSGTR accident as well as the predict ability of the system analysis codes at the PAFS operation.

Passive safety systems, with their reduced dependence on operator action and external power supply, are considered to have higher reliability and safety than traditional active safety systems. However, concerns remain regarding whether these systems can adequately perform safety functions across diverse scenarios due to their lower driving force. Consequently, Korea has developed a robustness assessment methodology to identify potential degradation factors affecting the performance of passive safety systems and, ultimately, reactor accident mitigation characteristics. This methodology’s applicability was evaluated through the passive safety system of SMART100.

The developed methodology has not been validated against the experiments. Therefore, this study aims to validate the methodology through the PERSEO experiment, designed to test passive heat removal systems (PHRS). PERSEO was selected as a benchmark due to its application in OECD/NEA/CSNI/WGAMA’s international benchmark problem, which has validated its utility for such evaluations. Following the robustness assessment methodology, the experiment review includes a detailed examination of the geometry, operational conditions, experimental procedures, and thermal-hydraulic phenomena. For robustness assessment, the regulatory system analysis code of Korea, which is MARS-KS 2.0 was used, as it has demonstrated capability in simulating passive heat removal systems.

The conservative estimation model analysis predicted heat exchanger performance below experimental results, similar to previous studies. After modifications to the heat exchanger model, the best-estimation model more accurately reflected experimental heat removal rates. Moving forward, the optimized PERSEO model will facilitate evaluating various degradation scenarios, advancing experimental validation of the robustness assessment methodology.

As an effective heat transfer mechanism, nucleate boiling phenomenon has been widely studied in both experiments and numerical simulations. In water-cooled nuclear reactors, it is critical to obtain localized information of the surface temperature and heat flux at the fuel cladding, which requires high-fidelity numerical simulation. One of key issues for the high-fidelity simulation of nucleate boiling is microlayer model. Currently, many studies utilize the interface tracking technique to realistically simulate the effects of microlayer evaporation heat transfer to the growth of boiling bubbles. However, the approach requires very fine meshes and thus extremely high computational costs, which deter practical exploitation of the approach. In this study, to achieve the precise simulation of microlayer evaporation heat transfer with reduced computational cost comparable to conventional CFD, an implicit microlayer model along with a new conjugate heat transfer solution scheme is proposed for the region where the microlayer depletes. The transition region with the maximum thickness about 50 is treated by the mesh inflation for the near wall cell. To improve the simulation accuracy, the mech refinement is also adapted in the bubble interface. Obtained results show good agreements with experiment data including the coupling relation of wall temperature, wall heat flux, microlayer thickness and bubble geometry. However, this preliminary study is limited to a single bubble, which can be seen as a benchmark test. Later, further numerical simulations of multiple bubbles will be performed to implement the proposed model to the practically useable mechanistic wall boiling model.

During the operation of a nuclear power plant, fuel assemblies in the core need to be replaced periodically. In-containment storage racks in the refueling pool are installed for temporary storage of fuel assemblies during refueling shutdowns. It is important to ensure that the temperature of the water and the fuel cladding are within a reasonable range. Therefore, we study the flow and heat transfer characteristics in the in-containment storage racks under different operating conditions. In this study, the natural circulation phenomenon at the fuel storage racks in the containment is numerically investigated. The thermal-hydraulic characteristics of the refueling pool under different loading conditions are simulated based on the porous media method. The local peak temperatures of water and fuel cladding are obtained. The results show that the structure and arrangement of the in-containment fuel storage racks can satisfy the cooling requirements under various operating conditions. There is no boiling phenomenon occurring in the refueling pool. The local peak temperatures in the refueling pool are almost the same for both the full and single-load conditions. The results of this research can provide important guidance for optimizing the design of fuel storage racks in the subsequent work.

This study investigates the core inlet flow distribution in a 1/5 scale model of the APR1000 reactor using Computational Fluid Dynamics (CFD). The APR1000 is an advanced reactor design integrating proven and innovative technologies, developed for projects such as the Czech Republics's nuclear new-build initiative. The objective is to ensure uniform flow and pressure distributions across fuel assemblies, critical for maintaining thermal and mechanical integrity. A CFD model was developed using ANSYS CFX, incorporating a geometry and grid strucuture that replicates the reactor's key components. Porous media modeling was employed for complex regions like perforated plates of the core simulator, balancing computational efficiency with fidelity. Comparative analysis of CFD results, based on RANS-based Turbulence model, against experimental data revealed discrepancies in flow mixing phenomena, particularly in the core's outer regions. While the CFD simulations showed a 10% margin of error for overall flow distribution, limitations in capturing large eddy dynamics led to deviations in mixing performance. These findings emphasize the challenges of simulating turbulent flow in complex geometries. Based on these findings, the aplication of a Large Eddy Simulation (LES) model to the CFD analysis enabled a more realistic simulation of flow mixing characteristics. Compared to the results from the RANS-based turbulence models, the LES approach yielded analysis results that were more consistent with experimental data.

The safety relief valve is used in nuclear power plants to prevent the over-pressurization of the primary system. The valve is located at the top of the pressurizer, which can cause a temperature gradient between the top and bottom, which may cause different thermal expansion of the internal components of the valve. In this study, we intend to analytically compare the temperature differences due to changes in the surface heat transfer coefficient of the valve. The analysis was performed using ANSYS CFX, and the analysis was performed assuming three cases. The first case is when there is no heat loss in the valve body, the second case is when there is heat loss in the body, and the third case is when there is heat loss in the upper parts of the valve body. The CFD results demonstrate that insulation type and steam volume fraction significantly impact temperature distribution inside the valve. When heat transfer through the valve body was minimized, the temperature difference between the inlet and the upper valve region was reduced. However, in cases where heat loss occurred in the valve body or upper pilot valves, the temperature difference increased.

The hypervapotron is a water-cooled device extensively used in thermonuclear fusion reactors for managing ultra-high heat fluxes (20–30 MW/m²). Its effectiveness stems from boiling heat transfer mechanisms and fin structures that promote efficient thermal regulation. This study employs multiphase computational fluid dynamics (CFD) simulations using OpenFOAM to investigate the effects of geometric variations such as fin height and side slot on cooling performance and internal flow dynamics in hypervapotron cooling channels.

Results revealed that side slots significantly enhance thermal management by enabling bidirectional fluid ingress, maintaining consistent liquid inflow into fin slots, and optimizing bubble removal. The side slots further contributed to the formation of diagonal flow patterns, which improved heat dissipation even under intense heat flux conditions. Conversely, channels without side slots exhibited higher wall temperatures and reduced cooling performance due to limited liquid ingress.

Adjustments in fin height were found to critically influence the cooling performance. Taller fins improved vortex formation and enhanced liquid inflow, resulting in superior heat exchange. In contrast, shorter fins led to incomplete fluid exchange and vapor accumulation within the fin slots, reducing efficiency.

This work bridges the relationship between hypervapotron design parameters and thermal performance. It provides actionable insights for optimizing the cooling channels to manage extreme thermal loads in fusion reactor divertors and similar applications. By integrating these findings, future designs can achieve enhanced thermal management and structural integrity under high-heat-flux environments, contributing to the advancement of hypervapotron technologies.

The natural circulation flap valve is a critical structural component of the poot-type low-temperature heating reactor. Under accident conditions, the two-phase natural circulation formed between the system and the reactor pool effectively cools the core to achieve safe shutdown. The flow space enclosed by the core riser, natural circulation flap valve, and top space of the reactor core forms a transverse T-shaped tube configuration. The characteristics of internal bubble-phase separation are intrinsically linked to the natural circulation capability, which plays a pivotal role in reactor safety analysis. Therefore, this study conducts an in-depth investigation on the characteristics of bubbly flow separation in the transverse T-shaped tube using numerical simulation methods, exploring the effects of branch height, main/branch tube diameter, void fraction, liquid flow rate, etc. on the separation characteristics. The results indicate that there is significant phase separation behavior in the transverse T-shaped tube. Specifically, the void fraction of the branch decreases with the increase of the branch height, while the stable rate of the branch pressure drop fluctuation increases with the increase of the branch height. These findings provide critical support for understanding and analyzing the characteristics of the two-phase natural circulation through the valve under accident conditions.

The integration of several codes has been recently developed to analyze multi-physical and multi-dimensional phenomena in a nuclear power plant. SPACE, CUPID, and MASTER codes are coupled. SPACE code is a one-dimensional system analysis code. CUPID code is a component scale three-dimensional thermal hydraulic analysis code. And MASTER is a three-dimensional neutronic code. SPACE and CUPID have similar governing equations for thermal-hydraulics. Thus, SPACE and CUPID are coupled to formulate a single pressure matrix for the coupled domain. This coupling scheme is well-validated with various theoretical problems and separate effect tests. CUPID and MASTER codes are coupled with sharing major variables, such as heat flux and temperature. To verify this multi-dimensional and multi-physic coupled code at a plant level, the main steam line break accident in the APR1400-kind plant is analyzed using this couple code. The core and connected parts of hot-leg and cold-leg are modeled with CUPID and the neutronic part is modeled with MASTER. The rest of the parts including the secondary loop are modeled with SPACE. In addition, the same scenario is modeled with only SPACE code. Comparing transient results of the coupled code and SPACE code, it is found that the newly developed coupled code is well verified for plant-level transient.

The potential of small modular reactors to solve the increasing international need for carbon-free electricity can only be realized with the concurrence of national nuclear regulators, such as the United States Nuclear Regulatory Commission (NRC). A major component of NRC approval of each unique design is the evaluation model (EM) development and assessment process, which produces an EM to demonstrate that nuclear safety is maintained within design basis accidents. The model is supported by experimental data, which may necessitate the creation of new test facilities or the design and acquisition of new loops and equipment capable of increasing the state of knowledge for impactful phenomena. The need for new data is determined by the phenomena identification and ranking table (PIRT) committee and depending on the extent of experimental needs for unique design elements, may include integral effects test and/or separate effects test facilities to be constructed for the verification and validation effort. Once the need for supplementary experimental data is decided, cost and schedule often necessitate scaling of test loops. A rigorous scaling effort is then employed to determine the most important physical parameters and dimensions that best represent the targeted prototype facility, while ensuring the key phenomenon of interest are preserved. These test facilities may also require improvements to existing infrastructure which could introduce new challenges while expanding future capabilities.

The McMaster Nuclear Reactor (MNR) is a research reactor located in Hamilton on the McMaster campus. The main purpose is to supply the medical industry with isotopes for a wide range of applications, most importantly for cancer treatments. In order to ensure the safety and performance of the MNR it is crucial to provide a solid database for a new thermal-hydraulic Safe Operation Envelope (SOE). With the McMaster hydraulic loop, there is a perfect tool to support the new SOE. The aim of the current work is threefold: First, gain a better understanding of the thermal-hydraulic characteristics of MNR fuel assemblies. Second, validate and refine computational tools such as system codes and CFD. Third, develop an artificial flow resistance that mimics the pressure drop of actual fuel assemblies, enabling cost-effective scaling of experiments without the use of original fuel elements consists the third goal. To complete these tasks, a single assembly is placed in a closed water circuit and subjected to varying mass flow rates. Parameters such as pressure drop, pressure distribution, temperature, mass flow, and density are precisely measured. The insights gained from this research contribute significantly to the optimization of reactor design and safety analyses, ultimately enhancing the accuracy of thermohydraulic prediction models for nuclear reactors. This comprehensive approach bridges the gap between theoretical models and practical applications, advancing the field of nuclear engineering. The results of different setups are compared and discussed in this work.

Recently, in the thermal hydraulic safety research area of nuclear power plant in Korea, the safety analysis methodology development with improvement of safety analysis code, SPACE, is now promoting to establish an IBLOCA, which has a smaller break size compared to large break loss of coolant accident (LBLOCA), as design basis accidents. In order to apply the SPACE code to the system analysis on the IBLOCA transients, appropriate SPACE code improvement along with the development of a new safety analysis methodology is necessary. And, of course, improved SPACE code should be evaluated and verified to confirm its capability.

In this study, an integral effect test database was established by utilizing ATLAS test facility which was constructed and operated by KAERI to verify the improved SPACE code. Three kinds of IBLOCA was simulated under the operating conditions of OPR1000 nuclear power plant. During the transient simulation, the system showed very general thermal hydraulic behavior that can occur in the IBLOCA transient, including a loop seal clearing phenomenon.

From the test results, the major thermal hydraulic phenomena were investigated and evaluated for the system cooling capability with an operation of safety systems. In addition, the difference of system behavior during the transient simulation according to the different break simulation will be investigated. The present test data can be utilized to verify and evaluate the improved SPACE code for application to an IBLOCA, as originally intended for the purpose of this study.

Major challenge for lead cooled fast reactors (LFRs) is the performance of structural materials. Specifically, flow-accelerated corrosion and erosion (FACE) phenomena may lead to deterioration of reactor internal structures. A dedicated facility, the Separate Effect Test Facility for Flow-Accelerated Corrosion and Erosion (SEFACE) is under design at KTH Royal Institute of Technology, to obtain experimental data in liquid lead conditions with high relative velocities and high temperatures. The facility is supposed to provide data for a wide range of flow and thermal conditions to study material degradation over months long periods of time in autonomous operation. A reliable system therefore is needed to maintain and control the required liquid metal temperatures at various flow velocity conditions. The flow velocity conditions are achieved in SEFACE by rotating disks accommodating experimental specimens in the main cylindrical vessel. The thermal effect of the disk rotation is the heat dissipation in the liquid lead. To keep the required temperature conditions, simultaneous active heating and cooling system is needed. To minimize vessel penetration and interference with vessel internal structures, a heating cooling jacket design is being developed. The active heating and cooling are achieved by heaters and water cooling tubes attached to thin fins of the jacket. Various configurations of fins, heaters, and water cooling tubes and material choices are discussed with numerical simulations and modular experimental tests. The optimal design is then selected and used for the temperature control system of the SEFACE facility.

The OECD/NEA ATLAS (Phase 3) project, spanning from 2021 to 2024, is a collaborative international project focused on addressing thermal-hydraulic safety and accident management challenges associated with water reactors utilizing the ATLAS test facility.
As part of this project, the C2.3 test was conducted to simulate a steam line break (SLB) scenario, incorporating the operation of a passive auxiliary feedwater system (PAFS). The primary goal was to analyze thermal-hydraulic transients within the reactor coolant system (RCS) during an SLB event with PAFS activation. Key parameters such as system pressures, temperatures, collapsed water levels, and natural circulation flows were measured and analyzed. The test outcomes highlighted the effectiveness of the PAFS operational system in cooling down the RCS during an SLB. Additionally, asymmetric thermal-hydraulic behavior was observed under asymmetric cooling conditions induced by the SLB. Effective cooling of the primary system was achieved during the operation of the main steam safety valves (MSSVs) of Steam Generator 2. Following PAFS actuation, a continuous decrease in core temperature and primary pressure was noted. This study further involved comparing the simulation results of MARS-KS with the OECD/NEA ATLAS C2.3 test data to evaluate the predictive capabilities of MARS-KS.

For the licensing and safety criteria of nuclear power plants, the standards for nuclear fuel are undergoing changes due to the adoption of DEC (Design Extension Conditions) and safety concerns related to high burn-up fuel. As a result, multi-physics coupled safety analysis has become a significant issue with the introduction of new LOCA (Loss Of Coolant Accident) criteria. In this study, the behavior of fuel cladding under simulated LOCA conditions was investigated using a thermo-mechanical and thermal–hydraulic coupled experimental facility known as ICARUS. The oxidized cladding was utilized to study high burn-up fuel, with an oxidized cladding being produced through an oxidation process. To assess the impact of oxidation, two reflood experiments were conducted—one with fresh cladding and the other with oxidized cladding—and the results were compared.

In this study, a safety analysis on a Pressurizer Level Control System Malfunction (PLCSMF) in the Advanced Power Reactor 1000 (APR1000) was conducted to investigate the effects of the pressurizer spray and heater operation. The Safety and Performance Analysis CodE for nuclear power plants version 3.3 (SPACE 3.3) was used for the transient calculation of the PLCSMF event. Based on the APR1000 design data, a nodalization of the nuclear steam supply system (NSSS) was modeled, and the steady-state calculations conducted for the combinations of the initial operating conditions, such as the RCS pressure, RCS temperature, and RCS flow rate. The PLCSMF event with 12 sets of the initial operating condition combinations were simulated as assuming the maximum charging pump flow rate and the minimum letdown flow rate with several conservative assumptions. In addition, the effects of pressurizer spray and heater operation were analyzed for each PLCSMF event, assuming both operation and non-operation of these systems. The pressurizer spray increased the RCS peak pressure and delayed its reaching time, resulting in a conservative outcome. The pressurizer heater had a negligible effect on the RCS peak pressure, but accelerated its reaching time. In conclusion, the present study highlights the importance of assuming pressurizer spray operation in the safety analysis of the ARP1000 PLCSMF event is necessary to obtain conservative. On the other hand, all APR1000 PLCSMF events analyzed in this study revealed that the RCS peak pressure were below the acceptance criteria.

In a severe accident in a nuclear power plant, hydrogen is generated within the reactor core and released into the containment building. It could accumulate and mix with air, posing a risk of explosion with low ignition energy. Passive autocatalytic recombiners (PARs) aim to reduce the hydrogen concentration inside the containment building below the flammable limits to prevent such explosions. The hydrogen-air mixture enters the bottom of the PAR by buoyancy. It moves upward between vertical plates coated with platinum (Pt) or palladium (Pd), where hydrogen recombines with oxygen through a catalytic reaction. The exothermic reaction and heating of product gases add additional driving force to the natural convection, and the resulting steam-air mixture is discharged at the top. In this study, we modeled the heat and mass transfer, fluid flow, and chemical reactions between two vertical catalytic plates using computational fluid dynamics (CFD) simulations. While most previous PAR studies applied a constant inlet velocity condition, we introduced natural convection conditions at the inlet to investigate how the flow rate of the hydrogen-air mixture entering the PAR correlates with the hydrogen concentration. As the inlet hydrogen concentration varied from 1 to 6 vol.%, the flow velocity entering the PAR reached a maximum at 4 vol.% and showed a decreasing trend beyond this concentration. We analyzed these phenomena by considering the buoyancy of the hydrogen-air mixture, the natural convection resulting from the exothermic chemical reaction, and the molar reduction due to the generation of steam.

By eliminating the graphite sleeve, a dual-separate direct cooling of fuel compacts is achieved, significantly improving both fuel cooling efficiency and the thermal output of high temperature gas-cooled reactors (HTGRs). In this sleeveless fuel design, however, fuel spacers hold crucial aspects in supporting the fuel compacts and maintaining open flow channels during normal operation in a helium coolant environment. The design of these fuel spacers has been an urgent issue in the related research field. To minimize pressure drops in the helium coolant flow, this study introduces a ring type fuel spacers that stabilizes each fuel compact from four directions. Real-scale prototypes of the fuel components were fabricated using 3D printing. To examine the flow characteristics around the fuel spacers, pressure drops measurements were conducted in an inert gas flow. As computational fluid dynamics (CFD) simulations of the spacers, this work performed CFD modeling using the commercial software, Star CCM+. Coupled with experimental results, this work has developed a validated fluid model around the spacers. This CFD model is a key tool for future V&V studies in thermal-hydraulics analysis of the reactor core.

Molten salt reactors (MSRs) are emerging as a next-generation nuclear technology due to their high operating temperatures and enhanced safety. The high melting point of molten salt leads to rapid solidification upon release from the reactor, which helps limit its spread. However, additional safety measures are necessary to minimize the potential dispersion of molten salt, as its spread and solidification significantly influence the release of radioactive materials. Argonne National Laboratory (ANL) previously used the MELTSPREAD code, originally developed for corium behavior analysis, to simulate the spread of molten salt on flat surfaces, using FLiNaK as the working fluid. However, MELTSPREAD tends to underestimate the spread radius because it does not consider for remelting. In this study, computational fluid dynamics (CFD) is used to improve the simulation accuracy of molten salt spread and solidification by incorporating remelting. NaCl-KCl-UCl3, in which 10 kg is poured for 1500 s, was chosen as the working fluid, and the simulations utilized models including the VOF multiphase model, Solidification and Melting model, RANS model, and DO radiation model. The sensitivity of these models to various parameters affecting molten salt behavior was evaluated. The simulation results showed an increase in both the spread radius and heat transfer over time, with the leading edge of the molten salt solidifying first, which limited further spreading. To enhance the accuracy of molten salt leakage assessments, model selection and parameters such as initial temperature and flow rate were crucial. Future research will explore the inclusion of decay heat models.

Heat pipe reactors are pivotal for next-generation nuclear technologies due to their inherent safety and flexibility. However, temperature redistribution under operational variations poses challenges to system safety and efficiency. Existing data-driven methods for temperature prediction often neglect physical principles, raising concerns about reliability. This study proposes a hybrid data-physics framework integrating neural networks with heat transfer physics to predict peak fuel rod temperatures in heat pipe reactors. A 1/6 centre-symmetric component area was analyzed using Computational Fluid Dynamics simulations to generate temperature datasets. The framework combines a backpropagation neural network (data-driven) for rapid feature-point prediction with a physics-driven Virtual Circle Model. Validation against Fluent simulations demonstrated maximum reconstruction errors of 3.8 K (0.4% relative error) under normal conditions. A novel diagnostic protocol for single heat pipe failure localization was developed, leveraging measuring through-holes temperature differences (ΔT) to identify failure modes. The method also characterized dual-failure temperature escalation patterns, revealing peak fuel temperatures exceeding 1000 K in severe cases. This hybrid approach enhances interpretability, accuracy, and reliability in thermal monitoring, offering a robust tool for safety evaluation and design optimization in advanced nuclear systems.

The increasing demand for safer and more efficient nuclear power has driven the development of innovative small modular reactors (i-SMRs) incorporating fully passive safety systems. These systems eliminate the need for active safety components, ensuring reactor safety under accident conditions without external power or operator intervention. This study presents the i-SMR’s passive safety system, which integrates the Passive Emergency Core Cooling System (PECCS), Passive Auxiliary Feedwater System (PAFS), Passive Containment Cooling System (PCCS), and Containment Isolation System (CIS). These systems leverage natural forces such as gravity and thermal differentials to sustain core cooling and maintain containment integrity. The PECCS ensures core cooling through a depressurization and recirculation mechanism, while the PAFS facilitates long-term decay heat removal. The PCCS supports containment cooling using a dry cooling mechanism, and the CIS prevents radioactive material release by automatically isolating the containment vessel during accidents. The i-SMR’s passive safety design enables extended accident mitigation, maintaining core cooling and structural integrity for at least 72 hours without operator action. Compared to conventional large reactors, the i-SMR achieves a significantly lower core damage frequency (CDF), demonstrating enhanced safety and reliability. This research provides insights into system functionality, regulatory considerations, and the next steps for validation and deployment of this advanced passive safety technology.

Severe Accident simulation codes like ASTEC (Accident Source Term Evaluation Code) are essential for predicting nuclear reactor behavior under Severe Accident conditions. However, their high computational demands, particularly in thermal-hydraulic simulations (e.g., the CESAR module in ASTEC), can constrain their effectiveness. A significant portion of CESAR’s computational load arises from solving non-linear partial differential equations at each timestep using the Newton-Raphson iteration method. While Newton-Raphson convergence depends on having initial guesses close to the final solution, the current ASTEC implementation relies solely on values from previous converged states, without predictive insights.

This paper introduces a hybrid approach aimed at enhancing CESAR’s Newton-Raphson solver through machine learning (ML) models, which offer more refined initial guesses. Drawing on recent advancements, this approach explores using ML-based surrogate models to learn the intricate, non-linear relationships within transient conditions, aiming to reduce the number of iterations needed for convergence and potentially allow longer timestep intervals without sacrificing accuracy. The choice of surrogate model remains adaptable, seeking to balance predictive accuracy and computational efficiency within CESAR’s frequent initialization routines.

Preliminary results show that an ML-augmented approach can significantly reduce ASTEC's computation time without altering the convergence criteria, suggesting promising implications for broader applications in thermal-hydraulic simulations in nuclear safety assessments. Future work may also involve developing an ensemble of surrogate models, complemented by a classifier that dynamically selects the most suitable initialization based on the reactor state and accident phase, optimizing the solver’s performance across varying scenarios.

This study investigates advanced Physics-Informed Neural Networks (PINNs)—a-PINN, n- PINN, and can-PINN—applied to benchmark fluid dynamics problems, including lid-driven cavity, Taylor-Green vortex, and natural convection in a square cavity. Transfer learning and targeted modifications to governing equations significantly enhance PINN accuracy and stability at higher Reynolds and Rayleigh numbers. Despite computational complexity, results highlight PINNs’ flexibility and their potential as effective alternatives to traditional CFD methods.

This study proposes a new method for predicting Critical Heat Flux (CHF) and its position in fuel rod bundles, enhancing the safety analysis of heat transfer systems. By leveraging the attention mechanism of Transformers, CHF prediction is achieved using sequential features from continuous physical fields, improving accuracy in complex rod bundle subchannels while increasing model interpretability. This approach identifies key physical quantities and control volumes that influence CHF. Initially, typical CHF experimental data were collected, including inlet conditions, CHF locations, and corresponding CHF values for individual fuel rod bundles. Then, numerical studies were used to generate upstream physical field data such as flow velocity, pressure, and qualities, which are critical for CHF predictions. These quantities were organized sequentially along the flow direction and served as inputs for the Transformer. The attention mechanism processed these sequential features to achieve high-precision CHF predictions. Finally, interpretability analysis evaluated feature importance in the model's outputs, revealing the contributions of different physical quantities and control volumes. Results demonstrated that the Transformer-based model significantly improved CHF prediction accuracy within rod bundle subchannels. The interpretability analysis identified the most influential physical factors, validating the model's reliability and providing insights into the CHF occurrence mechanism. In conclusion, the Transformer-based attention mechanism enhances CHF prediction accuracy in reactor rod bundles and identifies critical influencing factors through interpretability analysis. These findings advance the understanding of CHF behavior under complex multi-subchannel conditions in rod bundles with Mixing Vane Grids(MVGs) and provide a valuable tool for safety analysis and thermal system optimization.

Heat pipes have long been recognized as an efficient and versatile way of transferring heat, outperforming traditional mechanisms. In recent years, the interest towards this technology has risen, driven by its potential applications in emerging technologies and energy systems. One notable field driving the implementation of Liquid Metal Heat Pipes (LMHP) is their application in nuclear reactors for space exploration. Due to their versatility, simple mechanisms, and reliance on physical processes that function independently of gravity, heat pipes are expected to play a crucial role in the future of nuclear reactors for space missions. In this context, accurately predicting the behavior of HPs under a broad range of conditions becomes essential. The frozen startup of LMHPs, in particular, poses a significant challenge due to the variety of interconnected physical phenomena involved.

Among the different modeling techniques present in literature, conduction-based models represent a relatively simple yet effective method of capturing the transient behavior of LMHPs.

This study examines the conduction-based model proposed by Yoo et al., with particular focus on the ability of the latter to reproduce the temperature evolution over the heat pipe during the frozen startup phase. The model has been validated against experiments conducted by Ponnappan and by Faghri et al. to ensure that the results align with those proposed by the original authors. Additionally, the model has been validated over novel experiments performed at MISOH1 facility at the University of Michigan.

Thereafter, the impact of selecting appropriate correlations for the material properties is systematically analyzed.

Under the US High-Performance Research Reactor (USHPRR) project, new fuel has been developed for several high-performance research reactors, in order to transition the reactors from highly enriched uranium (HEU) to low-enriched uranium (LEU). Fuel plates utilized in some of these reactors have a thin cylindrical shape containing uranium-molybdenum alloy fuel core in aluminum alloy cladding. The changes in the design of the LEU fuel element (e.g., thinner fuel plates, different coolant channel gaps) compared to the HEU fuel element necessitate the evaluation of the hydraulic performance of the LEU fuel plates.

Previously developed analytical models by Donald R. Miller and Wade R. Marcum both predict a critical flow velocity at which a cylindrical laminate plate would collapse, but the data available for validation is sparse. The flow velocity in all USHPRR fuel elements is significantly lower than the calculated critical velocity. This research, in addition to supporting the displacement sensor testing for the USHPRR hydraulic performance evaluation, aims at providing high fidelity data on the relationship between the maximum deformation at the leading edge of a plate, the pressure differential between channels of a single plate, and the critical flow velocity. The outcome of this research can help to further validate the existing analytical models

The emergence of Small Modular Reactors (SMR) has focused attention on small nuclear power plant systems. Minimizing the heat exchanger volume is crucial for developing such systems. The Printed Circuit Heat Exchanger (PCHE) is a key candidate for achieving this goal.

PCHE is manufactured through multiple processes. First, micro-channels are created on metal plates using chemical etching. These plates are then stacked and joined through diffusion bonding, forming a single module. With its micro-channel structure, PCHE achieves high heat transfer efficiency per unit volume and exhibits excellent mechanical strength.

The thermal-hydraulic characteristics of PCHE vary depending on the shape of the micro-channels. For example, zigzag channels induce high pressure drops due to their complex geometric structure but enhance fluid mixing, leading to superior heat transfer performance. Existing studies have primarily focused on analyzing turbulent flow conditions using supercritical CO₂ (SCO₂) and helium as working fluids, while studies on laminar and transitional flow regimes using water remain limited. Moreover, previous research has mainly emphasized heat transfer performance and pressure drop, lacking detailed flow pattern analysis.

This study focuses on the flow pattern analysis of zigzag channels. Flow visualization experiments were conducted under various Reynolds number conditions using Particle Image Velocimetry (PIV), and the results were compared and validated with Computational Fluid Dynamics (CFD) simulations. Through this approach, changes in flow patterns due to variations in the bending angle were analyzed, and a correlation for the friction factor, a key parameter in PCHE design, was derived.

One of the defining features of integrated energy systems is multi-energy coordination, where renewable energy sources, baseload energy sources, and storage technologies are flexibly combined to overcome the limitations of single energy systems. This highlights the importance of studying the characteristics and reliability of hybrid energy systems, for delivering stable and efficient energy. In this paper, we propose and analyze an innovative hybrid energy system consisting of the small modular reactor, concentrated solar power and a packed-bed thermal energy storage, with a closed Brayton cycle as its energy conversion system. The system's configuration and operational control strategies are detailed. A one-dimensional dynamic model of the entire system, as well as its main equipment, has been proposed to facilitate performance evaluation. Through simulations and analyses, we explore the behavior of key parameters under varying operating conditions. Simulation results demonstrate that the proposed hybrid energy system effectively meets critical performance requirements. First, the system exhibits excellent power flexibility, dynamically adjusting output power to match varying energy demands by operational control strategies. Second, it achieves strong reliability, maintaining stable operation under fluctuating conditions such as varying solar irradiance. The study proves the viability of nuclear-solar-thermal storage hybrid systems as a promising solution for achieving stable and reliable energy supply. It provides a foundation for understanding the dynamic behavior of such systems and offer guidance for their application in diverse energy scenarios.