Pub Date : 2025-12-30DOI: 10.1016/j.anucene.2025.112072
Matthew Nyberg , Joseph Eickman , Una Baker , Patrick Shriwise , Ben Lindley
Two challenges were identified in modeling externally driven systems (EDS) for transuranic (TRU) burning with existing workflows: complex, spatially varying sources and flexible open-source production of cross-sections. A workflow is presented coupling the source and multi-group cross-section generation within OpenMC to the GeN-Foam multiphysics solver. First, OpenMC generated nuclear data was exported for transuranic-based molten salt reactor GeN-Foam model, and neutronics characteristics were verified against an equivalent OpenMC model. This TRU-based fueled salt was compared to uranium-thorium molten salt at steady state and over a range of potential accident scenarios. Secondly, a method utilizing the OpenMC C++ API sampled a fusion source onto a mesh used within GeN-Foam simulations, enabling closely coupled EDS multiphysics analysis. This spatially accurate source definition was verified against OpenMC models, and then was demonstrated for transient scenarios. This workflow enables closely coupled multiphysics analysis of complex critical or subcritical systems using open-source tools.
在利用现有工作流程为超铀（TRU）燃烧建模外部驱动系统（EDS）时，确定了两个挑战：复杂的、空间变化的来源和灵活的开源截面生产。提出了将OpenMC中的源和多组截面生成与GeN-Foam多物理场求解器相耦合的工作流程。首先，导出OpenMC生成的超铀熔盐堆GeN-Foam模型的核数据，并与等效OpenMC模型进行中子特性验证。在稳定状态和一系列潜在事故情景下，将这种基于trur的燃料盐与铀钍熔盐进行了比较。其次，利用OpenMC c++ API将融合源采样到GeN-Foam模拟中使用的网格上，从而实现紧密耦合的EDS多物理场分析。在OpenMC模型中验证了这种空间精确的源定义，然后在瞬态场景中进行了演示。该工作流可以使用开源工具对复杂的关键或次关键系统进行紧密耦合的多物理场分析。
{"title":"An open source multiphysics workflow for the analysis of subcritical transmutation systems","authors":"Matthew Nyberg , Joseph Eickman , Una Baker , Patrick Shriwise , Ben Lindley","doi":"10.1016/j.anucene.2025.112072","DOIUrl":"10.1016/j.anucene.2025.112072","url":null,"abstract":"<div><div>Two challenges were identified in modeling externally driven systems (EDS) for transuranic (TRU) burning with existing workflows: complex, spatially varying sources and flexible open-source production of cross-sections. A workflow is presented coupling the source and multi-group cross-section generation within OpenMC to the GeN-Foam multiphysics solver. First, OpenMC generated nuclear data was exported for transuranic-based molten salt reactor GeN-Foam model, and neutronics characteristics were verified against an equivalent OpenMC model. This TRU-based fueled salt was compared to uranium-thorium molten salt at steady state and over a range of potential accident scenarios. Secondly, a method utilizing the OpenMC C++ API sampled a fusion source onto a mesh used within GeN-Foam simulations, enabling closely coupled EDS multiphysics analysis. This spatially accurate source definition was verified against OpenMC models, and then was demonstrated for transient scenarios. This workflow enables closely coupled multiphysics analysis of complex critical or subcritical systems using open-source tools.</div></div>","PeriodicalId":8006,"journal":{"name":"Annals of Nuclear Energy","volume":"229 ","pages":"Article 112072"},"PeriodicalIF":2.3,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145881413","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-27DOI: 10.1016/j.anucene.2025.112097
Xingyu Zhao , Xiaoyu Guo , Guodong Liu , Yuntao Zheng , Yaru Li , Lanyu Zhou , Shuliang Huang , Shanfang Huang , Qiaoyan Chen , Kan Wang
This paper presents a high-fidelity neutronics/thermal-hydraulics transient coupling approach using the Monte Carlo code RMC and CFD code ANSYS Fluent, applied to control rod ejection accidents in a prismatic high-temperature gas-cooled reactor (HTGR) core. The Picard-iteration-based scheme uses RMC time-space dynamics calculation to update both power amplitude and shape. Data transfer and mesh mapping are realized through hierarchical data format (hdf) and a multi-superposition mesh mapping strategy. Verification through pressurized water reactor (PWR) mini-core cases shows good agreement with reference results. Various control rod ejection scenarios with different locations and reactivity insertions are simulated in the prismatic HTGR. The time-dependent results demonstrate thermal-hydraulics feedback amplified by larger reactivity, confirming favorable passive safety. Compared with conventional methods, the high-fidelity approach yields less fluctuating results and reduces redundant conservatism, thus enhancing the overall efficiency. The high-fidelity approach also has a certain capability to simulate prompt supercritical transients.
{"title":"High-fidelity transient neutronics/thermal-hydraulics coupling analyses of control rod ejection accident in a prismatic gas-cooled reactor core","authors":"Xingyu Zhao , Xiaoyu Guo , Guodong Liu , Yuntao Zheng , Yaru Li , Lanyu Zhou , Shuliang Huang , Shanfang Huang , Qiaoyan Chen , Kan Wang","doi":"10.1016/j.anucene.2025.112097","DOIUrl":"10.1016/j.anucene.2025.112097","url":null,"abstract":"<div><div>This paper presents a high-fidelity neutronics/thermal-hydraulics transient coupling approach using the Monte Carlo code RMC and CFD code ANSYS Fluent, applied to control rod ejection accidents in a prismatic high-temperature gas-cooled reactor (HTGR) core. The Picard-iteration-based scheme uses RMC time-space dynamics calculation to update both power amplitude and shape. Data transfer and mesh mapping are realized through hierarchical data format (hdf) and a multi-superposition mesh mapping strategy. Verification through pressurized water reactor (PWR) mini-core cases shows good agreement with reference results. Various control rod ejection scenarios with different locations and reactivity insertions are simulated in the prismatic HTGR. The time-dependent results demonstrate thermal-hydraulics feedback amplified by larger reactivity, confirming favorable passive safety. Compared with conventional methods, the high-fidelity approach yields less fluctuating results and reduces redundant conservatism, thus enhancing the overall efficiency. The high-fidelity approach also has a certain capability to simulate prompt supercritical transients.</div></div>","PeriodicalId":8006,"journal":{"name":"Annals of Nuclear Energy","volume":"229 ","pages":"Article 112097"},"PeriodicalIF":2.3,"publicationDate":"2025-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838560","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-27DOI: 10.1016/j.anucene.2025.112098
Mohsin Raza , Ikram UI Haq , Waqar ul Hassan , Jun-Liang Guo , Hong-Na Zhang , Xiao-Bin Li , Yue Wang , Wei-Hua Cai , Shu-Qi Meng , Fang Chen , Yu-Long Mao , Feng-Chen Li
This review analyzes advanced predictive methodologies related to thermal striping in nuclear reactors, which involves the interaction of hot and cold fluid streams leading to temperature fluctuations, thermal stresses, and potential structural vulnerabilities. This study emphasizes high-fidelity simulation techniques, such as direct numerical simulations (DNS) and large eddy simulations (LES), which effectively capture transient flow dynamics with high fidelity. The review also explores data-driven innovations like the machine learning (ML), which exhibits significant potential for improving predictive accuracy by integrating physical principles with high-quality datasets. Primary failure mechanisms, such as thermal fatigue, stress corrosion cracking, and thermal embrittlement, are thoroughly examined. Well-established reduced-order modeling (ROM) approaches, such as proper orthogonal decomposition (POD) and dynamic mode decomposition (DMD)-based reduced models—reduce the dimensionality and can substantially lower computational cost; with stable reduced integration, near real-time predictions are achievable within calibrated operating ranges. This review highlights the significant impact of multi-scale hybrid modeling for rapid and accurate prediction of thermal striping. This review identifies key limitations of current modeling approaches, particularly in balancing computational cost, accuracy, and speed. A detailed comparison shows that while traditional models offer precision, they are often too slow or expensive for real-time use.
On the other hand, ROM and ML enable faster predictions but may sacrifice accuracy in complex scenarios. Based on this trade-off, this study highlights hybrid modeling approaches as a promising solution for balancing accuracy, speed, and computational cost prediction of thermal striping. Finally, this study outlines critical research gaps and suggests future directions that may guide the development of smarter and more reliable prediction tools for thermal fluid systems and advance reactor technology.
{"title":"A state-of-the-art review of accurate and rapid prediction methods for thermal striping phenomenon in nuclear reactors","authors":"Mohsin Raza , Ikram UI Haq , Waqar ul Hassan , Jun-Liang Guo , Hong-Na Zhang , Xiao-Bin Li , Yue Wang , Wei-Hua Cai , Shu-Qi Meng , Fang Chen , Yu-Long Mao , Feng-Chen Li","doi":"10.1016/j.anucene.2025.112098","DOIUrl":"10.1016/j.anucene.2025.112098","url":null,"abstract":"<div><div>This review analyzes advanced predictive methodologies related to thermal striping in nuclear reactors, which involves the interaction of hot and cold fluid streams leading to temperature fluctuations, thermal stresses, and potential structural vulnerabilities. This study emphasizes high-fidelity simulation techniques, such as direct numerical simulations (DNS) and large eddy simulations (LES), which effectively capture transient flow dynamics with high fidelity. The review also explores data-driven innovations like the machine learning (ML), which exhibits significant potential for improving predictive accuracy by integrating physical principles with high-quality datasets. Primary failure mechanisms, such as thermal fatigue, stress corrosion cracking, and thermal embrittlement, are thoroughly examined. Well-established reduced-order modeling (ROM) approaches, such as proper orthogonal decomposition (POD) and dynamic mode decomposition (DMD)-based reduced models—reduce the dimensionality and can substantially lower computational cost; with stable reduced integration, near real-time predictions are achievable within calibrated operating ranges. This review highlights the significant impact of multi-scale hybrid modeling for rapid and accurate prediction of thermal striping. This review identifies key limitations of current modeling approaches, particularly in balancing computational cost, accuracy, and speed. A detailed comparison shows that while traditional models offer precision, they are often too slow or expensive for real-time use.</div><div>On the other hand, ROM and ML enable faster predictions but may sacrifice accuracy in complex scenarios. Based on this trade-off, this study highlights hybrid modeling approaches as a promising solution for balancing accuracy, speed, and computational cost prediction of thermal striping. Finally, this study outlines critical research gaps and suggests future directions that may guide the development of smarter and more reliable prediction tools for thermal fluid systems and advance reactor technology.</div></div>","PeriodicalId":8006,"journal":{"name":"Annals of Nuclear Energy","volume":"229 ","pages":"Article 112098"},"PeriodicalIF":2.3,"publicationDate":"2025-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145881412","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The release of fission gas in nuclear fuel significantly impacts fuel performance. Currently, many engineering models for fission gas release (FGR) rely on empirical corrections of simplified processes, introducing considerable uncertainty. Therefore, implementing mechanism-based FGR models grounded in physical behavior is crucial for improving the reliability of fuel performance codes. In this study, an established mechanism-based FGR model (incorporating atomic diffusion, intra-granular bubble re-solution, grain-boundary sweeping, and inter-granular bubble dynamics) was integrated into the fuel performance analysis code FROBA, along with a non-thermal release model. The implementation couples grain-boundary gas release with swelling equations. Model validation against literature benchmarks under steady-state conditions demonstrates excellent agreement with experimental data and other codes for both FGR fraction and swelling rate. Uncertainty analysis confirms the model’s effectiveness within the implemented scope.
{"title":"Development and application of a mechanism-based fission gas release model in FROBA fuel performance code","authors":"Kou Minghai , Xiao Xinkun , Yu Songjiao , Chen Ronghua , Jiang Pinting , Dai Mingliang , Zhang Kui , Wu Yingwei , Tian Wenxi , Qiu Suizheng","doi":"10.1016/j.anucene.2025.112094","DOIUrl":"10.1016/j.anucene.2025.112094","url":null,"abstract":"<div><div>The release of fission gas in nuclear fuel significantly impacts fuel performance. Currently, many engineering models for fission gas release (FGR) rely on empirical corrections of simplified processes, introducing considerable uncertainty. Therefore, implementing mechanism-based FGR models grounded in physical behavior is crucial for improving the reliability of fuel performance codes. In this study, an established mechanism-based FGR model (incorporating atomic diffusion, intra-granular bubble re-solution, grain-boundary sweeping, and inter-granular bubble dynamics) was integrated into the fuel performance analysis code FROBA, along with a non-thermal release model. The implementation couples grain-boundary gas release with swelling equations. Model validation against literature benchmarks under steady-state conditions demonstrates excellent agreement with experimental data and other codes for both FGR fraction and swelling rate. Uncertainty analysis confirms the model’s effectiveness within the implemented scope.</div></div>","PeriodicalId":8006,"journal":{"name":"Annals of Nuclear Energy","volume":"229 ","pages":"Article 112094"},"PeriodicalIF":2.3,"publicationDate":"2025-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838533","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-26DOI: 10.1016/j.anucene.2025.112093
Hanwen Xu , Rui Pan , Shuai Chen , Qiusun Zeng , Yanan Ma , Zhen Wang
In the context of rising global energy demand and the shift towards low-carbon energy, lead–bismuth cooled fast reactors (LFRs) have emerged as a key technology in fourth-generation nuclear reactor development. The advantages of LFRs—low neutron absorption cross-section, atmospheric-pressure operation, excellent heat transfer performance, and strong chemical inertness—endow them with significant potential in specialized energy supply scenarios. However, the compact structure of LFRs and complex system parameter changes during failures present challenges, and existing prediction methods fail to meet practical requirements. This study proposes a novel model integrating the attention mechanism, convolutional neural networks (CNN), and long short-term memory networks (LSTM). CNNs extract local spatiotemporal features, the attention mechanism highlights critical information, and LSTMs capture both long- and short-term dependencies. Combined with a multi-input, multi-output (MIMO) prediction strategy, the model enables multi-step prediction of LFR safety–critical parameters under fault conditions. Experimental results based on simulation data from various operating scenarios of China’s Lead-based Research Reactor (CLEAR-I) demonstrate that the proposed model outperforms advanced alternative models. Compared with RNN, Attention-GRU, and TCN, it reduces RMSE by an average of 35%, RRMSE by 32%, MAE by 29%, and MAPE by 33% across 6-step, 12-step, and 24-step predictions. Its single-sample inference time ranges from 1.99 to 2.12 ms, with a 95% confidence interval coverage rate of 96.84%. This model effectively predicts safety-related parameters under LFR fault conditions, providing crucial support for reactor safety and stability, and demonstrating significant application value in fault parameter prediction for lead–bismuth cooled reactors.
{"title":"Prediction method for Safety-Related parameters of Lead-Bismuth cooled fast reactor using Attention-CNN-LSTM fusion model","authors":"Hanwen Xu , Rui Pan , Shuai Chen , Qiusun Zeng , Yanan Ma , Zhen Wang","doi":"10.1016/j.anucene.2025.112093","DOIUrl":"10.1016/j.anucene.2025.112093","url":null,"abstract":"<div><div>In the context of rising global energy demand and the shift towards low-carbon energy, lead–bismuth cooled fast reactors (LFRs) have emerged as a key technology in fourth-generation nuclear reactor development. The advantages of LFRs—low neutron absorption cross-section, atmospheric-pressure operation, excellent heat transfer performance, and strong chemical inertness—endow them with significant potential in specialized energy supply scenarios. However, the compact structure of LFRs and complex system parameter changes during failures present challenges, and existing prediction methods fail to meet practical requirements. This study proposes a novel model integrating the attention mechanism, convolutional neural networks (CNN), and long short-term memory networks (LSTM). CNNs extract local spatiotemporal features, the attention mechanism highlights critical information, and LSTMs capture both long- and short-term dependencies. Combined with a multi-input, multi-output (MIMO) prediction strategy, the model enables multi-step prediction of LFR safety–critical parameters under fault conditions. Experimental results based on simulation data from various operating scenarios of China’s Lead-based Research Reactor (CLEAR-I) demonstrate that the proposed model outperforms advanced alternative models. Compared with RNN, Attention-GRU, and TCN, it reduces RMSE by an average of 35%, RRMSE by 32%, MAE by 29%, and MAPE by 33% across 6-step, 12-step, and 24-step predictions. Its single-sample inference time ranges from 1.99 to 2.12 ms, with a 95% confidence interval coverage rate of 96.84%. This model effectively predicts safety-related parameters under LFR fault conditions, providing crucial support for reactor safety and stability, and demonstrating significant application value in fault parameter prediction for lead–bismuth cooled reactors.</div></div>","PeriodicalId":8006,"journal":{"name":"Annals of Nuclear Energy","volume":"229 ","pages":"Article 112093"},"PeriodicalIF":2.3,"publicationDate":"2025-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838559","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-26DOI: 10.1016/j.anucene.2025.112071
Furqan Arshad , Minjun Peng , Wasiq Ali , Zikang Li , Fazle Haseeb , Awais Khan
This study proposes a framework which integrates the machine learning based fault diagnosis with the efficiency monitoring of a pressurized water reactor (PWR) nuclear power plant. The purpose of the efficiency monitoring is to detect the operational deviations from the optimum conditions, while the fault diagnosis part identifies the faulty equipment along with the extent estimation. The fault diagnosis has been performed through the use of feed forward back propagation (FFBP) and long short term memory (LSTM) neural networks, and its performance has further been improved through the incorporation of physics augmented feature space. In total, thirty three fault conditions related to the internal leakages in steam generators and feed water heaters have been studied in this work. It has been demonstrated that through the augmentation of physics-based features, the overall performance of the fault diagnosis is significantly improved. This improved performance has further been verified through the application of SHapley Additive exPlanations (SHAP) analysis, and also the model robustness has been demonstrated through testing against the noisy data.
{"title":"Physics-informed fault diagnosis through online efficiency monitoring of PWR type nuclear power plants","authors":"Furqan Arshad , Minjun Peng , Wasiq Ali , Zikang Li , Fazle Haseeb , Awais Khan","doi":"10.1016/j.anucene.2025.112071","DOIUrl":"10.1016/j.anucene.2025.112071","url":null,"abstract":"<div><div>This study proposes a framework which integrates the machine learning based fault diagnosis with the efficiency monitoring of a pressurized water reactor (PWR) nuclear power plant. The purpose of the efficiency monitoring is to detect the operational deviations from the optimum conditions, while the fault diagnosis part identifies the faulty equipment along with the extent estimation. The fault diagnosis has been performed through the use of feed forward back propagation (FFBP) and long short term memory (LSTM) neural networks, and its performance has further been improved through the incorporation of physics augmented feature space. In total, thirty three fault conditions related to the internal leakages in steam generators and feed water heaters have been studied in this work. It has been demonstrated that through the augmentation of physics-based features, the overall performance of the fault diagnosis is significantly improved. This improved performance has further been verified through the application of SHapley Additive exPlanations (SHAP) analysis, and also the model robustness has been demonstrated through testing against the noisy data.</div></div>","PeriodicalId":8006,"journal":{"name":"Annals of Nuclear Energy","volume":"229 ","pages":"Article 112071"},"PeriodicalIF":2.3,"publicationDate":"2025-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838534","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-24DOI: 10.1016/j.anucene.2025.112091
Fikri A. Furqan, Deni Mustika, Saga Octadamailah, Mu’nisatun Sholikhah, G.K. Suryaman, Ade Saputra, Supardjo
This study presents an initial neutronic analysis of the new nuclear fuel U-7Mo-xTi/Al (x = 1, 2, 3) proposed as a replacement for U3Si2/Al in the GA Siwabessy Multipurpose Reactor (RSG-GAS). Ti is added to the U-7Mo alloy to stabilise the γ-U phase, improving powder fabrication, and contributes to enhanced corrosion resistance of the fuel. Evaluations were conducted on enrichment optimisation, burnup, and temperature coefficients using OpenMC with the RSG-GAS fuel element geometry model without considering neutron leakage. Simulation results show that the optimal enrichment for each composition is 13.715 % (U-7Mo-1Ti/Al), 14.140 % (U-7Mo-2Ti/Al), and 14.5 % (U-7Mo-3Ti/Al) to achieve a k-infinity value comparable to U3Si2/Al at 19.75 % enrichment. Burnup behaviour indicates an extension of fuel lifetimes from 25 days to 45.909 days (U-7Mo-1Ti/Al); 45.723 days (U-7Mo-2Ti/Al); 45.572 days (U-7Mo-2Ti/Al), indicating improved fuel cycle efficiency. Safety margins are strengthened by strongly negative temperature coefficients: FTC (−1.94700 to −2.72296 pcm/K) and MTC (−0.64651 to −4.18274 pcm/K), which support the inherent safety characteristics of the reactor. Overall, U-7Mo-xTi/Al has higher fuel efficiency and safety margins than U3Si2/Al.
{"title":"Neutronic analysis of U-7Mo-xTi/Al fuel elements as replacement candidates for Indonesia’s RSG-GAS research reactor fuel: Enrichment optimisation, burnup behaviour, and temperature coefficients","authors":"Fikri A. Furqan, Deni Mustika, Saga Octadamailah, Mu’nisatun Sholikhah, G.K. Suryaman, Ade Saputra, Supardjo","doi":"10.1016/j.anucene.2025.112091","DOIUrl":"10.1016/j.anucene.2025.112091","url":null,"abstract":"<div><div>This study presents an initial neutronic analysis of the new nuclear fuel U-7Mo-xTi/Al (x = 1, 2, 3) proposed as a replacement for U<sub>3</sub>Si<sub>2</sub>/Al in the GA Siwabessy Multipurpose Reactor (RSG-GAS). Ti is added to the U-7Mo alloy to stabilise the γ-U phase, improving powder fabrication, and contributes to enhanced corrosion resistance of the fuel. Evaluations were conducted on enrichment optimisation, burnup, and temperature coefficients using OpenMC with the RSG-GAS fuel element geometry model without considering neutron leakage. Simulation results show that the optimal enrichment for each composition is 13.715 % (U-7Mo-1Ti/Al), 14.140 % (U-7Mo-2Ti/Al), and 14.5 % (U-7Mo-3Ti/Al) to achieve a k-infinity value comparable to U<sub>3</sub>Si<sub>2</sub>/Al at 19.75 % enrichment. Burnup behaviour indicates an extension of fuel lifetimes from 25 days to 45.909 days (U-7Mo-1Ti/Al); 45.723 days (U-7Mo-2Ti/Al); 45.572 days (U-7Mo-2Ti/Al), indicating improved fuel cycle efficiency. Safety margins are strengthened by strongly negative temperature coefficients: FTC (−1.94700 to −2.72296 pcm/K) and MTC (−0.64651 to −4.18274 pcm/K), which support the inherent safety characteristics of the reactor. Overall, U-7Mo-xTi/Al has higher fuel efficiency and safety margins than U<sub>3</sub>Si<sub>2</sub>/Al.</div></div>","PeriodicalId":8006,"journal":{"name":"Annals of Nuclear Energy","volume":"229 ","pages":"Article 112091"},"PeriodicalIF":2.3,"publicationDate":"2025-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145839023","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The paper presents the models used in the severe accident module SAFR of the EUCLID/V2 integral code for analyzing the destruction of fuel rods with nitride fuel in a fast reactor cooled by liquid metal (lead, sodium). The study focuses on key fuel degradation mechanisms, including dissociation at the fuel-liquid coolant (sodium/lead), fuel-liquid melt (e.g., cladding or uranium melt), and fuel-sodium vapor interfaces, as well as subsequent eutectic interactions between the dissociation products and the cladding steel. The paper also presents validation results of the models. The melting and relocation models for the fuel cladding were validated against experimental data from the Institute of Thermophysics of the Siberian Branch of the Russian Academy of Sciences (IT SB RAS), while the models for nitride fuel dissociation were validated using experiments from the National Research Nuclear University MEPhI. The models describing nitride fuel behavior under accident conditions were validated based on experiments conducted at the Impulse Graphite Reactor. Data from the Argonne National Laboratory was used to validate models of uranium relocation and eutectic interactions with stainless steel.
{"title":"Models and validation results of the SAFR module of the integral code EUCLID/V2 for calculating thermal destruction of fuel pins with nitride fuel","authors":"E.V. Usov, V.D. Ozrin, V.I. Chukhno, I.A. Klimonov, A.A. Butov, I.G. Kudashov, M.G. Kozlov, N.A. Mosunova, V.F. Strizhov, N.A. Pribaturin","doi":"10.1016/j.anucene.2025.112095","DOIUrl":"10.1016/j.anucene.2025.112095","url":null,"abstract":"<div><div>The paper presents the models used in the severe accident module SAFR of the EUCLID/V2 integral code for analyzing the destruction of fuel rods with nitride fuel in a fast reactor cooled by liquid metal (lead, sodium). The study focuses on key fuel degradation mechanisms, including dissociation at the fuel-liquid coolant (sodium/lead), fuel-liquid melt (e.g., cladding or uranium melt), and fuel-sodium vapor interfaces, as well as subsequent eutectic interactions between the dissociation products and the cladding steel. The paper also presents validation results of the models. The melting and relocation models for the fuel cladding were validated against experimental data from the Institute of Thermophysics of the Siberian Branch of the Russian Academy of Sciences (IT SB RAS), while the models for nitride fuel dissociation were validated using experiments from the National Research Nuclear University MEPhI. The models describing nitride fuel behavior under accident conditions were validated based on experiments conducted at the Impulse Graphite Reactor. Data from the Argonne National Laboratory was used to validate models of uranium relocation and eutectic interactions with stainless steel.</div></div>","PeriodicalId":8006,"journal":{"name":"Annals of Nuclear Energy","volume":"229 ","pages":"Article 112095"},"PeriodicalIF":2.3,"publicationDate":"2025-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838532","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-24DOI: 10.1016/j.anucene.2025.112092
Baptiste Bouchon, Koen Smits, Giorgio Valocchi, Jean Tommasi
In this work, we revisit the modeling of two MUSE-4 Pulsed Neutrons Source (PNS) experiments using the CEA Monte Carlo code TRIPOLI-4® and JEFF-3.1.1 nuclear data library. The MUSE-4 experiments were carried out in the MASURCA zero-power facility at CEA Cadarache between 2000 and 2004, in order to characterize ADS (Accelerator Driven Systems) and ways of monitoring their reactivity. We investigate two subcritical configurations, with respective multiplication factors around 0.97 and 0.957 and detectors placed at several positions across the core, reflector and shield. For each configuration we model the prompt neutron kinetic response after a single pulse of neutrons, the floor plateau resulting from precursor build-up after a long series of pulses using the area-method and the intrinsic source from the plutonium-based (MOx) fuel. The transient, which represents the response of the subcritical core to a burst of neutrons, is measured with a set of fission chambers. Each of these fission chambers is individually modeled in the Monte Carlo code TRIPOLI-4®. Despite the uncertainties in the experimental protocol, such as the exact loading maps of the reactor core and detector efficiencies, the time series of the transient that we obtain match the experimental data in both the timing and shape of the peak, and overall reproduce the key behaviors observed during the MUSE-4 PNS experiments. However, for the lowest reactivity and detectors far from the core, some discrepancies are observed in the shape of the decreasing part of the prompt neutron population. The origin of these discrepancies is likely multifactorial: they may arise from experimental uncertainties or from biases in steel cross-sections at low energies, as neutrons reaching these detectors have traveled long paths. Another possible factor is the hydrogen in the concrete surrounding the core, which could reflect slowed-down neutrons, as observed during the FREYA experiment. The methodology to estimate the plateau value and some tentative explanations on the discrepancies observed are provided in the article.
{"title":"MUSE-4 Pulsed Neutrons Source (PNS) experiments modeling using the Monte Carlo transport code TRIPOLI-4®","authors":"Baptiste Bouchon, Koen Smits, Giorgio Valocchi, Jean Tommasi","doi":"10.1016/j.anucene.2025.112092","DOIUrl":"10.1016/j.anucene.2025.112092","url":null,"abstract":"<div><div>In this work, we revisit the modeling of two MUSE-4 Pulsed Neutrons Source (PNS) experiments using the CEA Monte Carlo code TRIPOLI-4® and JEFF-3.1.1 nuclear data library. The MUSE-4 experiments were carried out in the MASURCA zero-power facility at CEA Cadarache between 2000 and 2004, in order to characterize ADS (Accelerator Driven Systems) and ways of monitoring their reactivity. We investigate two subcritical configurations, with respective multiplication factors around 0.97 and 0.957 and detectors placed at several positions across the core, reflector and shield. For each configuration we model the prompt neutron kinetic response after a single pulse of neutrons, the floor plateau resulting from precursor build-up after a long series of pulses using the area-method and the intrinsic source from the plutonium-based (MOx) fuel. The transient, which represents the response of the subcritical core to a burst of neutrons, is measured with a set of fission chambers. Each of these fission chambers is individually modeled in the Monte Carlo code TRIPOLI-4®. Despite the uncertainties in the experimental protocol, such as the exact loading maps of the reactor core and detector efficiencies, the time series of the transient that we obtain match the experimental data in both the timing and shape of the peak, and overall reproduce the key behaviors observed during the MUSE-4 PNS experiments. However, for the lowest reactivity and detectors far from the core, some discrepancies are observed in the shape of the decreasing part of the prompt neutron population. The origin of these discrepancies is likely multifactorial: they may arise from experimental uncertainties or from biases in steel cross-sections at low energies, as neutrons reaching these detectors have traveled long paths. Another possible factor is the hydrogen in the concrete surrounding the core, which could reflect slowed-down neutrons, as observed during the FREYA experiment. The methodology to estimate the plateau value and some tentative explanations on the discrepancies observed are provided in the article.</div></div>","PeriodicalId":8006,"journal":{"name":"Annals of Nuclear Energy","volume":"229 ","pages":"Article 112092"},"PeriodicalIF":2.3,"publicationDate":"2025-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838558","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-23DOI: 10.1016/j.anucene.2025.112090
Xuewei Miao, Zhonghao Li, Qingyue You, Dingping Peng, Bo Cao
This study proposes a source term inversion method for nuclear accidents based on the Harris Hawks Optimization (HHO) algorithm and a Gaussian plume model, enabling accurate estimation of radionuclide release rates and the two-dimensional location of release points using off-site monitoring data under accident scenarios. To evaluate model performance, validation was conducted through simulated experiments under two accident scenarios with known and unknown release locations and tracer experiments involving seven different release scenarios. The simulation results demonstrate that, compared with two other swarm intelligence algorithms, Particle Swarm Optimization (PSO) and Genetic Algorithm (GA), the HHO-based inversion model achieves higher estimation accuracy, faster convergence speed, and greater stability during iterative inversion. The convergence rate and accuracy of the model are somewhat dependent on the initialization range of the population and the boundary constraints of the target parameters. The tracer experiment validation shows that the HHO model performs well in most cases, with an average relative error of 0.0341 in release rate inversion and an average positional deviation of 133 m across the seven experiments. Sensitivity analysis indicates that the HHO inversion model exhibits certain robustness in estimating release rates, while the two-dimensional location of the release point is more susceptible to interference from noise in off-site monitoring data.
{"title":"Source term inversion method for nuclear accidents based on Harris Hawks Optimization","authors":"Xuewei Miao, Zhonghao Li, Qingyue You, Dingping Peng, Bo Cao","doi":"10.1016/j.anucene.2025.112090","DOIUrl":"10.1016/j.anucene.2025.112090","url":null,"abstract":"<div><div>This study proposes a source term inversion method for nuclear accidents based on the Harris Hawks Optimization (HHO) algorithm and a Gaussian plume model, enabling accurate estimation of radionuclide release rates and the two-dimensional location of release points using off-site monitoring data under accident scenarios. To evaluate model performance, validation was conducted through simulated experiments under two accident scenarios with known and unknown release locations and tracer experiments involving seven different release scenarios. The simulation results demonstrate that, compared with two other swarm intelligence algorithms, Particle Swarm Optimization (PSO) and Genetic Algorithm (GA), the HHO-based inversion model achieves higher estimation accuracy, faster convergence speed, and greater stability during iterative inversion. The convergence rate and accuracy of the model are somewhat dependent on the initialization range of the population and the boundary constraints of the target parameters. The tracer experiment validation shows that the HHO model performs well in most cases, with an average relative error of 0.0341 in release rate inversion and an average positional deviation of 133 m across the seven experiments. Sensitivity analysis indicates that the HHO inversion model exhibits certain robustness in estimating release rates, while the two-dimensional location of the release point is more susceptible to interference from noise in off-site monitoring data.</div></div>","PeriodicalId":8006,"journal":{"name":"Annals of Nuclear Energy","volume":"229 ","pages":"Article 112090"},"PeriodicalIF":2.3,"publicationDate":"2025-12-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145838528","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号