R&D status of digital technology on inverse estimation of radioactive source distributions and related source countermeasures; Fast Digital Twin Tech. in Decommissioning Field: 3D-ADRES-Indoor FrontEnd

Machida, Masahiko; Yamada, Susumu; Kim, M.; Tanaka, Satoshi*; Tobita, Yasuhiro*; Iwata, Ayako*; Aoki, Yuto; Aoki, Kazuhisa; Yanagisawa, Kenichi*; Yamaguchi, Takashi; et al.

RIST News, (70), p.3 - 22, 2024/09

Inside the Fukushima Daiichi Nuclear Power Plant (1F), there are many locations with high radiation levels due to contamination by radioactive materials that leaked from the reactor. These pose a significant obstacle to the smooth progress of decommissioning work. To help solve this issue, the Japan Atomic Energy Agency (JAEA), under a subsidy from the Ministry of Economy, Trade, and Industry's decommissioning and contaminated water management project, is conducting research and development on digital technologies to improve the radiation environment inside the decommissioning site. This project, titled "Development of Technology to Improve the Environment Inside Reactor Buildings (Enhancing Digital Technology for Environment and Source Distribution to Reduce Radiation Exposure)," began in April of FY 2023. In this project, the aim is to develop three interconnected systems: FrontEnd, Pro, and BackEnd. The FrontEnd system, based on the previously developed 3D-ADRES-Indoor (prototype) from FY 2021-2022, will be upgraded to a high-speed digital twin technology usable on-site. The Pro system will carry out detailed analysis in rooms such as the new office building at 1F, while the BackEnd system will serve as a database to centrally manage the collected and analyzed data. This report focuses on the FrontEnd system, which will be used on-site. After point cloud measurement, the system will quickly create a 3D mesh model, estimate the radiation source from dose rate measurements, and refine the position and intensity of the estimated source using recalculation techniques (re-observation instructions and re-estimation). The results of verification tests conducted on Unit 5 are also presented. Furthermore, the report briefly discusses the future research and development plans for this project.

Application of machine learning to spattering phenomena in laser cutting

Kusumoto, Toshiyuki*; Saruta, Koichi; Naoe, Takashi; Teshigawara, Makoto; Futakawa, Masatoshi; Hasegawa, Kazuo*; Tsuboi, Akihiko

Jikken Rikigaku, 23(4), p.310 - 315, 2023/12

https://doi.org/10.11395/jjsem.23.310

Reducing spatter, i.e., melt droplets flown out of the melt pool, is one of the critical issues when laser cutting is employed as a machining tool for radioactive wastes because the ejected droplets can lead to radioactive contamination with potential human exposure. The spattering phenomena are complicated processes that involve multiple physical phenomena, causing difficulty in the determination of laser parameters to minimize the amount of spatter. Here we observe the spatter ejected from 316L stainless steel plates using a high-speed camera and apply a machine learning technique to these captured images on the basis of three distinctive behaviors appeared at specific time intervals of the process of spattering phenomena: (I) a vapor, (II) a liquid film and breakup into droplets, and (III) a liquid capillary. The numerical model established through the machine learning technique predicts the spattering phenomena with an accuracy of 89% and can be used to determine the laser power and beam diameter that reduce the spatter eruption during laser irradiation.

R&D of digital technology on inverse estimation of radioactive source distributions and related source countermeasures; R&D status of digital platform including 3D-ADRES-indoor

Machida, Masahiko; Yamada, Susumu; Kim, M.; Okumura, Masahiko; Miyamura, Hiroko; Shikaze, Yoshiaki; Sato, Tomoki*; Numata, Yoshiaki*; Tobita, Yasuhiro*; Yamaguchi, Takashi; et al.

RIST News, (69), p.2 - 18, 2023/09

https://www.rist.or.jp/rnews/69/69s2.pdf

The contamination of radioactive materials leaked from the reactor has resulted in numerous hot spots in the Fukushima Daiichi Nuclear Power Station (1F) building, posing obstacles to its decommissioning. In order to solve this problem, JAEA has conducted research and development of the digital technique for inverse estimation of radiation source distribution and countermeasures against the estimated source in virtual space for two years from 2021 based on the subsidy program "Project of Decommissioning and Contaminated Water Management" performed by the funds from the Ministry of Economy, Trade and Industry. In this article, we introduce the results of the project and the plan of the renewal project started in April 2023. For the former project, we report the derivative method for LASSO method considering the complex structure inside the building and the character of the source and show the result of the inverse estimation using the method in the real reactor building. Moreover, we explain the platform software "3D-ADRES-Indoor" which integrates these achievements. Finally, we introduce the plan of the latter project.

Microstructural evolution in tungsten binary alloys under proton and self-ion irradiations at 800C

Miyazawa, Takeshi; Kikuchi, Yuta*; Ando, Masami*; Yu, J.-H.*; Yabuuchi, Kiyohiro*; Nozawa, Takashi*; Tanigawa, Hiroyasu*; Nogami, Shuhei*; Hasegawa, Akira*

Journal of Nuclear Materials, 575, p.154239_1 - 154239_11, 2023/03

https://doi.org/10.1016/j.jnucmat.2023.154239

Estimation of children's thyroid equivalent doses in 16 municipalities after the Fukushima Daiichi Nuclear Power Station accident

Suzuki, Gen*; Ishikawa, Tetsuo*; Oba, Takashi*; Hasegawa, Arifumi*; Nagai, Haruyasu; Miyatake, Hirokazu*; Yoshizawa, Nobuaki*

Journal of Radiation Research (Internet), 63(6), p.796 - 804, 2022/11

https://doi.org/10.1093/jrr/rrac058

To elucidate the association between radiation dose and thyroid cancer after the 2011 Fukushima Daiichi Nuclear Power Station (FDNPS) accident, it is essential to estimate individual thyroid equivalent doses (TEDs) to children. In a previous study, we reported a methodology for reconstructing TEDs from inhalation based on individual behavioural survey sheets combined with a spatiotemporal radionuclides database constructed by an atmospheric transport, diffusion, and deposition model (ATDM). In the present study, we further refined our methodology and estimated the combined TEDs from inhalation and ingestion among children in 16 municipalities around the nuclear power station utilizing 3,256 individual whereabouts questionnaire survey sheets. Distributions of estimated TEDs were similar to estimates based on direct thyroid measurements in 1080 children in Iwaki City, Kawamata Town, Iitate Village, and Minamisoma City. Mean TEDs in 1-year-old children ranged from 1.3 mSv in Date City to 14.9 mSv in Odaka Ward in Minamisoma City, and the 95th percentiles varied from 2.3 mSv in Date City to 28.8 mSv in Namie Town. In the future, this methodology can be useful for a case-control study of thyroid cancer after the FDNPS accident.

Performance measurement of an urban wind simulation code with the Locally Mesh-Refined Lattice Boltzmann Method over NVIDIA and AMD GPUs

Asahi, Yuichi; Onodera, Naoyuki; Hasegawa, Yuta; Shimokawabe, Takashi*; Shiba, Hayato*; Idomura, Yasuhiro

Keisan Kogaku Koenkai Rombunshu (CD-ROM), 27, 5 Pages, 2022/06

We have ported the GPU accelerated Lattice Boltzmann Method code "CityLBM" to AMD MI100 GPU. We present the performance of CityLBM achieved on NVIDIA P100, V100, A100 GPUs and AMDMI100 GPU. Using the host to host MPI communications, the performance on MI100 GPU is around 20% better than on V100 GPU. It has turned out that most of the kernels are successfully accelerated except for interpolation kernels for Adaptive Mesh Refinement (AMR) method.

AMR-Net: Convolutional neural networks for multi-resolution steady flow prediction

Asahi, Yuichi; Hatayama, Sora*; Shimokawabe, Takashi*; Onodera, Naoyuki; Hasegawa, Yuta; Idomura, Yasuhiro

Proceedings of 2021 IEEE International Conference on Cluster Computing (IEEE Cluster 2021) (Internet), p.686 - 691, 2021/10

https://doi.org/10.1109/Cluster48925.2021.00102

We develop a convolutional neural network model to predict the multi-resolution steady flow. Based on the state-of-the-art image-to-image translation model pix2pixHD, our model can predict the high resolution flow field from the set of patched signed distance functions. By patching the high resolution data, the memory requirements in our model is suppressed compared to pix2pixHD.

Acceleration of locally mesh allocated Poisson solver using mixed precision

Onodera, Naoyuki; Idomura, Yasuhiro; Hasegawa, Yuta; Shimokawabe, Takashi*; Aoki, Takayuki*

Keisan Kogaku Koenkai Rombunshu (CD-ROM), 26, 3 Pages, 2021/05

We develop a mixed-precision preconditioner for the pressure Poisson equation in a two-phase flow CFD code JUPITER-AMR. The multi-grid (MG) preconditioner is constructed based on the geometric MG method with a three- stage V-cycle, and a cache-reuse SOR (CR-SOR) method at each stage. The numerical experiments are conducted for two-phase flows in a fuel bundle of a nuclear reactor. The MG-CG solver in single-precision shows the same convergence histories as double-precision, which is about 75% of the computational time in double-precision. In the strong scaling test, the MG-CG solver in single-precision is accelerated by 1.88 times between 32 and 96 GPUs.

Development of long pulse arc driven ion source for iBNCT

Shibata, Takanori*; Sugimura, Takashi*; Ikegami, Kiyoshi*; Takagi, Akira*; Sato, Masaharu*; Naito, Fujio*; Okoshi, Kiyonori; Hasegawa, Kazuo

JPS Conference Proceedings (Internet), 33, p.011009_1 - 011009_6, 2021/03

https://doi.org/10.7566/JPSCP.33.011009

Upgrade of beam current in the Linac of Ibaraki Boron Neutron Capture Therapy (iBNCT) is one of the most important requirements to realize clinical trial. By 2018, the measurement of the produced neutrons characteristics and the neutron irradiation experiment for living cells have been done by producing 8-MeV proton beam current at the beryllium target with average current up to 2 mA. In order to satisfy the original clinical trial conditions, 5 mA average beam current is required at the target. For this goal, peak beam current extracted from the ion source should be increased to 60 mA from the present 30 mA with duty factor up to more than 10% (pulse width up to 1 ms and repetition rate up to more than 100 Hz). Stability of the peak current in the macro pulse is also important for the clinical application.

GPU acceleration of multigrid preconditioned conjugate gradient solver on block-structured Cartesian grid

Onodera, Naoyuki; Idomura, Yasuhiro; Hasegawa, Yuta; Yamashita, Susumu; Shimokawabe, Takashi*; Aoki, Takayuki*

Proceedings of International Conference on High Performance Computing in Asia-Pacific Region (HPC Asia 2021) (Internet), p.120 - 128, 2021/01

https://doi.org/10.1145/3432261.3432273

We develop a multigrid preconditioned conjugate gradient (MG-CG) solver for the pressure Poisson equation in a two-phase flow CFD code JUPITER. The MG preconditioner is constructed based on the geometric MG method with a three-stage V-cycle, and a RB-SOR smoother and its variant with cache-reuse optimization (CR-SOR) are applied at each stage. The numerical experiments are conducted for two-phase flows in a fuel bundle of a nuclear reactor. The MG-CG solvers with the RB-SOR and CR-SOR smoothers reduce the number of iterations to less than 15% and 9% of the original preconditioned CG method, leading to 3.1- and 5.9-times speedups, respectively. The obtained performance indicates that the MG-CG solver designed for the block-structured grid is highly efficient and enables large-scale simulations of two-phase flows on GPU based supercomputers.