@inproceedings{oai:jaxa.repo.nii.ac.jp:00002933, author = {大道, 勇哉 and 小林, 憲司 and 金崎, 雅博 and Omichi, Yuya and Kobayashi, Kenji and Kanazaki, Masahiro}, book = {宇宙航空研究開発機構特別資料: 第50回流体力学講演会/第36回航空宇宙数値シミュレーション技術シンポジウム論文集, JAXA Special Publication: Proceedings of the 50th Fluid Dynamics Conference / the 36th Aerospace Numerical Simulation Symposium}, month = {Feb}, note = {第50回流体力学講演会/第36回航空宇宙数値シミュレーション技術シンポジウム (2018年7月4日-6日. 宮崎市民プラザ), 宮崎市, 宮崎, 50th Fluid Dynamics Conference /the 36th Aerospace Numerical Simulation Symposium (July 4-6, 2018. Miyazaki Citizen's Plaza), Miyazaki, Japan, The feature extraction technique based on dynamic mode decomposition (DMD) and mode selection methods for large CFD datasets was proposed. The incremental proper orthogonal decomposition (POD) was introduced as a preconditioning step. By performing the preconditioning step, the DMD and mode selection can be applied to large datasets with low memory consumption. The proposed algorithms were applied to a subsonic flowfield around a re-entry capsule. We found that the flowfield has four dominant fluid phenomena and they have the frequency of St approx. 0.2 and St = 0.0159. Furthermore, the contribution of these fluid phenomena on the aerodynamic coefficient fluctuations of the capsule was clarified. The result showed that the lift and drag coefficient fluctuations are dominated by vortex shedding phenomena (of St approx. 0.2) and pressure oscillation phenomena in the recirculation region (of St = 0.0159), respectively. This pressure oscillation phenomenon of St = 0.0159 has not been reported so far and seems to be related to dynamic instability phenomena of the capsule because its time scale is close to that of the capsule's motion reported previously., 形態: カラー図版あり, Physical characteristics: Original contains color illustrations, 資料番号: AA1830029004, レポート番号: JAXA-SP-18-005}, pages = {83--90}, publisher = {宇宙航空研究開発機構(JAXA), Japan Aerospace Exploration Agency (JAXA)}, title = {大規模非定常データに対する特徴構造抽出法の開発と大気突入カプセル後流解析への適用}, volume = {JAXA-SP-18-005}, year = {2019} }