Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Tada, Kenichi; Kondo, Ryoichi; Endo, Tomohiro*; Yamamoto, Akio*
Journal of Nuclear Science and Technology, 60(6), p.624 - 631, 2023/06
Times Cited Count：2 Percentile：53.91(Nuclear Science & Technology)The sensitivity analysis and the uncertainty quantification have an important role in improving the evaluated nuclear data library. The current computational performance enables us to the sensitivity analysis and uncertainty quantification using the continuous energy Monte Carlo calculation code. The ACE file perturbation tool was developed for these calculations using modules of FRENDY. This tool perturbs the microscopic cross section, the number of neutrons per fission, and the fission spectrum. The uncertainty quantification using the random sampling method is also available if the user prepares the covariance matrix. The uncertainty of the k-effective using the perturbation tool was compared to the current sensitivity analysis codes SCALE/TSUNAMI and MCNP/KSEN. The comparison results indicated that the random sampling method using this tool accurately estimates the uncertainty of k-effective.
Narukawa, Takafumi
Kaku Nenryo, (54-2), P. 3, 2019/07
no abstracts in English
Tanaka, Masaaki
Proceedings of 12th International Topical Meeting on Nuclear Reactor Thermal-Hydraulics, Operation and Safety (NUTHOS-12) (USB Flash Drive), 14 Pages, 2018/10
A numerical simulation code MUGTHES has been developed to estimate high cycle thermal fatigue in SFRs. In development of numerical simulation code, verification, validation, and uncertainty quantification (VVUQ) are indispensable. In this study, numerical simulation at impinging jet condition in the WATLON experiment which was the water experiment of a T-junction piping system was performed for the fundamental validation. Based on the previous studies, the simplified least square version GCI method and the area validation metrics were employed as reference methods to quantify uncertainty and to measure the degree of difference between the numerical and the experimental results, respectively. Through the examinations, the potential applicability of the MUGTHES to the thermal striping phenomena was indicated and requirements of modification in the simulation was suggested in accordance with the uncertainty values.
Narukawa, Takafumi
Kaku Nenryo, (53-2), P. 5, 2018/08
no abstracts in English
Tanaka, Masaaki; Miyake, Yasuhiro*
Proceedings of 22nd International Conference on Nuclear Engineering (ICONE-22) (DVD-ROM), 13 Pages, 2014/07
In this study, numerical simulation for the WATLON experiment which was the water experiment of a T-junction piping system (T-pipe) was carried out to validate the MUGTHES and to investigate the relation between the mechanism of temperature fluctuation generation and the unsteady motion of large eddy structures. In the numerical simulation, the large eddy simulation (LES) approach with standard Smagorinsky model was employed as eddy viscosity model to simulate large scale eddy motion in the T-pipe. As for uncertainty quantification in the validation process, the modified method of the Grid Convergence Index (GCI) estimation based on the least squire version could successfully quantify uncertainty. Through the numerical simulations, it was indicated that the fine mesh arrangement could improve the temperature distribution in the wake. It could be found that the thermal mixing phenomena in the T-pipe were caused by the mutual interaction of the necklace-shaped vortex around the wake from the front of the branch jet, the horseshoe-shaped vortex and the Karman's vortex motions in the wake.
Tanaka, Masaaki
no journal, ,
In development of numerical simulation codes and estimation methods for plant design and safety assessment, implementation of verification and validation (V&V) and uncertainty quantification is required. To establish the practical procedures of the uncertainty quantification, the SLS-GCI (Simplified Least Square version GCI) method modified from the Eca's least-square version GCI method was established. Through the several examinations, the applicability of the SLS-GCI method was confirmed.
Tanaka, Masaaki
no journal, ,
A numerical simulation code named MUGTHES has been developed to estimate thermal fatigue issue in sodium-cooled fast reactors (SFRs). Additionally, author has been developed a practical procedure named V2UP for verification and validation (V&V) process and numerical prediction with uncertainty quantification in order to ensure credibility of the numerical estimation results. In the V2UP, uncertainty quantification is required. Therefore, the area validation metric (AVM) and the modified AVM (MAVM) methods were examined to measure degree of agreement (difference) between the numerical results by MUGTHES and experimental results of sodium experiments (PLAJEST) for the triple parallel jets thermal mixing phenomena. Through the examinations of the AVM and MAVM methods, values of the degree of agreement between them were successfully estimated and it was indicated that the MAVM method could be a reference method in the validation process of the V2UP.
Ohgama, Kazuya; Ikeda, Kazumi*; Ishikawa, Makoto; Kan, Taro*; Oki, Shigeo
no journal, ,
no abstracts in English
Tanaka, Masaaki
no journal, ,
A procedure called as V2UP (Verification and Validation plus Uncertainty quantification and Prediction) was made by referring to the existing guidelines on V&V and the methodologies of the safety assessment (CSAU, ISTIR, EMDAP). The V2UP consisted of five components as follows: (1) phenomena analysis with the Phenomena Identification and Ranking Table (PIRT) method, (2) implementation of the V&V, (3) design and rearrangement of experiments for the V&V, (4) uncertainty quantification in each problem and integration of uncertainties and (5) numerical prediction (estimation) for the target issue. Although the complete application of the procedure has not been performed at this moment, a flow chart of the V2UP procedure was described in this paper with recent results of the examinations.
Ohgama, Kazuya
no journal, ,
no abstracts in English
Sugino, Kazuteru; Numata, Kazuyuki; Ishikawa, Makoto; Takeda, Toshikazu*
no journal, ,
no abstracts in English
Ohgama, Kazuya
no journal, ,
no abstracts in English
Terada, Hiroaki; Nagai, Haruyasu
no journal, ,
For the assessment of the radiological doses to the public due to the atmospheric discharge of radioactive materials during the Fukushima Daiichi Nuclear Power Station accident, the spatiotemporal distribution of radioactive materials in the environment are reconstructed by atmospheric dispersion simulation with the improved WSPEEDI. In this study, the influence of chemical form of 
I (particle, inorganic and organic gas) in source term on the simulated results was analyzed to assess the uncertainty of the simulation. From the sensitivity test with the chemical form composition, surface deposition distribution of I was influenced according to the property of the chemical forms to deposition processes, whereas time-integrated air concentration was less influenced. From the comparison with the measurement, the surface deposition simulated with the source term in Katata et al. (2015) had a good reproducibility compared with those with extreme gas to particle ratios.
Narukawa, Takafumi; Yamaguchi, Akira*; Jang, S.*; Amaya, Masaki
no journal, ,
no abstracts in English
Narukawa, Takafumi
no journal, ,
no abstracts in English
Takamizawa, Hisashi; Nishiyama, Yutaka
no journal, ,
no abstracts in English
Maruyama, Shuhei; Endo, Tomohiro*; Yamamoto, Akio*
no journal, ,
no abstracts in English
Zheng, X.; Tamaki, Hitoshi; Shibamoto, Yasuteru; Maruyama, Yu
no journal, ,
The nuclear industry is expressing a growing interest in the research and use of artificial intelligence and machine learning (AI/ML) technology to improve plant operational performance and reduce the risks associated with nuclear power generation. JAEA is applying the AI/ML technology to advancing researches on severe accidents and probabilistic risk assessment (PRA). To efficiently perform dynamic PRA and uncertainty quantification of source terms, both simulation-based, we are introducing surrogate models trained via machine learning to estimate core damage frequency (conditional core damage probability), to obtain information about the probability distribution of source terms and importance ranking of parameters. AI/ML can be expected to efficiently provide risk and uncertainty information to make rational decisions for the continuous improvement of nuclear safety.
