3D visualization in complicated flow channel using deep learning-based bubble detection

Uesawa, Shinichiro; Ono, Ayako; Yoshida, Hiroyuki

Gazo Rabo, p.1 - 5, 2025/08

This paper introduces a new measurement technique for visualizing the three-dimensional distribution of bubbles in a complex channel such as a nuclear reactor fuel assembly. Bubbly flow is important in many engineering fields, and especially in nuclear engineering, where bubble behavior significantly affects the performance and safety of nuclear reactors, and thus requires detailed understanding. Conventional rule-based image recognition has difficulty identifying bubbles overlapping in the line-of-sight direction, but in this study, deep learning (Mask R-CNN and Swin Transformer) is used to achieve highly accurate bubble detection with a small amount of training data. Furthermore, the tracking technique using ByteTrack made it possible to track many bubbles with complex motions, and by combining images taken from different viewpoints using two high-speed cameras and reconstructing the 3D shape of the bubbles using the ellipsoid approximation, 3D instantaneous local information such as bubble position, diameter, and velocity was obtained. To eliminate the effects of refraction and obstruction of vision by structures in the channel, a simulated fuel rod was made of a transparent material (PFA tube) with a refractive index similar to that of water, enabling distortion-free imaging and measurement even in channels with complex structures. This enabled 3D visualization of bubble behavior in complex channels, which had been difficult to achieve in the past. Since this technology enables highly accurate 3D visualization with a small number of cameras and a small amount of learning, it is expected to be applied to objects other than bubbles.

Numerical investigation of the accuracy of a conductance-type wire-mesh sensor for a single spherical bubble and bubbly flow

Uesawa, Shinichiro; Ono, Ayako; Nagatake, Taku; Yamashita, Susumu; Yoshida, Hiroyuki

Journal of Nuclear Science and Technology, 62(5), p.432 - 456, 2025/05

https://doi.org/10.1080/00223131.2024.2447366

We performed electrostatic simulations of a wire-mesh sensor (WMS) for a single spherical bubble and bubbly flow to clarify the accuracy of the WMS. The electrostatic simulation for the single bubble showed the electric current density distribution and the electric current path from the excited transmitter to receivers for various bubble locations. It indicated systematic errors based on the nonuniform current density distribution around the WMS. The electrostatic simulation for the bubbly flow calculated by the computational fluid dynamics code, JAEA Utility Program for Interdisciplinary Thermal-hydraulics Engineering and Research (JUPITER), indicated that the WMS had difficulty in quantitatively measuring the intermediate values of the instantaneous void fraction between 0 and 1 because they cannot be estimated by previous transformation methods from the WMS signal to the void fraction, such as linear approximation or Maxwell's equation, and have a significant deviation of the void fraction of 0.2 for the WMS signal. However, the electrostatic simulation indicated that the time-averaged void fractions around the center of the flow channel can be estimated using linear approximation, and the time-averaged void fraction near the wall of the flow channel can be estimated using Maxwell's equation.

3D visualization in rod bundle flow channel using deep learning-based bubble detection

Uesawa, Shinichiro; Ono, Ayako; Yoshida, Hiroyuki

Konsoryu, 39(1), p.61 - 71, 2025/03

https://doi.org/10.3811/jjmf.2025.004

Bubble visualization using a high-speed video-camera has been used as a measurement technique of bubble diameters and velocities. However, the bubble detection was difficult under the condition of the high void fraction because the overlapping bubbles for the sight direction of the camera increased with the void fraction. Additionally, the visualization for a system with objects, such as rod bundle flow channels, becomes more difficult. In this study, we applied a deep learning-based bubble detection technique with Shifted Window Transformer to bubble images shoot from two directions to identify the bubble size, three-dimensional (3D) positions of bubbles, 3D bubble trajectories in the rod bundle flow channel. Furthermore, we used perfluoroalkoxy alkane tubes with almost the same reflection as water in the channel to visualize the bubbly flow in the whole of the flow channel. We confirmed that the detection technique can segment individual bubbles in overlapping bubbles and bubbles behind the rod. By using the detection results, we estimated the diameter and velocity of each bubble and cross-sectional void fraction.

Numerical investigation of accuracy of conductance-typed wire-mesh sensor using CFD and electrostatic simulations

Uesawa, Shinichiro; Ono, Ayako; Yamashita, Susumu; Yoshida, Hiroyuki

Proceedings of 13th Korea-Japan Symposium on Nuclear Thermal Hydraulics and Safety (NTHAS13) (Internet), 7 Pages, 2024/11

A conductance-typed wire-mesh sensor (WMS), utilizing the difference in conductivity between gas and liquid phases between the electrodes, is one of the practical measurement techniques of a cross-sectional void fraction distribution in a flow path. In this study, we performed two-phase computational fluid dynamics (CFD) and electrostatic simulations around a WMS for a single spherical bubble and bubbly flow to clarify the systematic error in the WMS. The results for the single bubble indicated that there were systematic errors based on the non-uniform current density distribution around the WMS. The correlation between instantaneous void fractions and WMS signals is not uniquely determined for positions of the single bubble moving across the WMS, even for the same bubble. Moreover, the correlation between the instantaneous void fractions and the WMS signals did not fit in a linear approximation and Maxwell's equation, which traditionally used transformation methods from the WMS signal to the void fraction. The results for the bubbly flow indicated that the WMS had difficulty in quantitative measurements of the instantaneous void fraction because the values had a significant deviation of the void fraction of approximately 0.2. On the other hand, time-averaged void fraction values had relatively small deviation. Thus, we concluded that the WMS, using existing transformation methods, can measure time-averaged void fractions, but it is difficult to measure quantitatively instantaneous void fractions.

Bubble visualization in rod bundle flow channel using deep learning-based bubble detection

Uesawa, Shinichiro; Ono, Ayako; Yoshida, Hiroyuki

Konsoryu Shimpojiumu 2024 Koen Rombunshu (Internet), 2 Pages, 2024/09

Bubble visualization using a high-speed video-camera has been used as measurement techniques of bubble diameters, interfacial area concentrations, and void fractions in dispersed bubbly flow. However, the bubble detection was difficult under the condition of the high void fraction because the overlapping bubbles for the sight direction of the camera increased with the increase in the void fraction. In this study, we developed the deep learning-based bubble detector with Shifted window Transformer (Swin Transformer) to overcome the issue. To verify the performance, we used the synthetic bubble images obtained by Generative Adversarial Networks (GAN) and obtained average precisions (APs) for the number of the training dataset. The result showed that the AP was large enough for 50 datasets, and bubble detection was possible even with a small number of the training data. Additionally, we confirmed that the detector can detect and segment individual bubbles in overlapping bubbles obtained in the visualization experiments of pipe and bundle flows. By using the detection results, we estimated the interfacial area concentrations and void fractions. In comparison with commonly used relations, the results were in good agreement with the relations. Thus, the detector can measure not only bubble diameters but also interfacial area concentrations and void fractions.

Stability enhancement by hydrophobic anchoring and a cross-linked structure of a phospholipid copolymer film for medical devices

Uchida, Kazuto*; Masuda, Tsukuru*; Hara, Shintaro*; Matsuo, Yoichi*; Liu, Y.*; Aoki, Hiroyuki; Asano, Yoshihiko*; Miyata, Kazuki*; Fukuma, Takeshi*; Ono, Toshiya*; et al.

ACS Applied Materials & Interfaces, 16(30), p.39104 - 39116, 2024/07

https://doi.org/10.1021/acsami.4c07752

Neutron reflectivity study on the adsorption layer of polyethylene grown on Si substrate

Shimokita, Keisuke*; Yamamoto, Katsuhiro*; Miyata, Noboru*; Shibata, Motoki*; Nakanishi, Yohei*; Arakawa, Masato*; Takenaka, Mikihito*; Kida, Takumitsu*; Tokumitsu, Katsuhisa*; Tanaka, Ryo*; et al.

Langmuir, 40(30), p.15758 - 15766, 2024/07

https://doi.org/10.1021/acs.langmuir.4c01584

Development of measurement method for gas-liquid two-phase flow inside a fuel bundle to obtain code validation data

Ono, Ayako; Okamoto, Kaoru*; Makino, Yasushi*; Hosokawa, Shigeo*; Yoshida, Hiroyuki

Proceedings of Specialist Workshop on Advanced Instrumentation and Measurement Techniques for Nuclear Reactor Thermal-Hydraulics and Severe Accidents (SWINTH-2024) (USB Flash Drive), 13 Pages, 2024/06

JAEA has been developing an advanced neutronic/thermal-hydraulics coupling simulation system. In the coupling simulation system, the detailed thermal-hydraulics codes based on an interface-capturing method (JUPITER or TPFIT) will be adopted to simulate thermal-hydraulics behavior in a fuel bundle. The experimental data and findings relating to the gas-liquid two-phase flow in a fuel bundle are especially required to validate JUPITER/TPFIT. In this study, we therefore develop a measurement method by combining Laser-Doppler Velocimetry (LDV) and photodiodes, which can access to a small flow channel such as a subchannel of a fuel bundle. The developed measurement method is validated by comparison with the measument by a electrical conductance probe. Finally, we obtain experimental data on local flow structures and interactions between gas and liquid phases. The developed measurement method is actually applied to an air-water dispersed bubbly flow to confirm its capability.

Development of high-fidelity multi-physics platform JAMPAN

Tada, Kenichi; Kondo, Ryoichi; Kamiya, Tomohiro; Nagatake, Taku; Ono, Ayako; Nagaya, Yasunobu; Yoshida, Hiroyuki

Proceedings of International Conference on Physics of Reactors (PHYSOR 2024) (Internet), p.1488 - 1497, 2024/04

JAEA has developed a new high-fidelity multi-physics platform JAMPAN for connecting single-physics codes such as a neutronics code and a thermal-hydraulics code. It consists of the HDF5 formatted data container and input and output data handler modules to generate the input file and read the output file of the single-physics code. Users can easily add or exchange the code by implementing input and output data handler modules for this code. The first target of JAMPAN is the coupling of neutronics and thermal-hydraulics calculations to provide reference results of core analysis codes. The current version of JAMPAN couples the neutronics code MVP and the thermal-hydraulics codes JUPITER, ACE-3D, and NASCA. Users can select the thermal-hydraulics code depending on the scale of problems to be solved, computational performance, and so on. This presentation explains the overview of JAMPAN and shows the results of the neutronics and thermal-hydraulics coupling calculation.

Summary report in FY2022 of subsidy program for "the Project of Decommissioning and Contaminated Water Management (Development of Analysis and Estimation Technology for Characterization of Fuel Debris (Development of Technologies for Enhanced Analysis Accuracy, Thermal behavior Estimation, and Simplified Analysis of Fuel Debris)" started in FY2011

Koyama, Shinichi; Ikeuchi, Hirotomo; Mitsugi, Takeshi; Maeda, Koji; Sasaki, Shinji; Onishi, Takashi; Tsai, T.-H.; Takano, Masahide; Fukaya, Hiroyuki; Nakamura, Satoshi; et al.

Hairo, Osensui, Shorisui Taisaku Jigyo Jimukyoku Homu Peji (Internet), 216 Pages, 2023/11

https://en.dccc-program.jp/wp-content/uploads/20231117_JAEA.pdf

In FY 2021 and 2022, JAEA perfomed the subsidy program for "the Project of Decommissioning and Contaminated Water Management (Development of Analysis and Estimation Technology for Characterization of Fuel Debris (Development of Technologies for Enhanced Analysis Accuracy, Thermal Bahavior Estimation, and Simplified Analysis of Fuel Debris)" started in FY 2021. This presentation material summarized the results of the project, which will be available shortly on the website of Management Office for the Project of Decommissioning, Contaminated Water and Treated Water Management.