TY - GEN
T1 - Parallel Dynamic Mode Decomposition for Rayleigh–Taylor Instability Flows
AU - Tan, Weiwei
AU - Bai, Junqiang
AU - Tian, Zengdong
AU - Li, Li
N1 - Publisher Copyright:
© 2019, Springer Nature Singapore Pte Ltd.
PY - 2019
Y1 - 2019
N2 - Many fluid flows of engineering interest, though very complex in appearance, can be approximated by low-order models governed by a few modes, and Dynamic Mode Decomposition (DMD) has been proved effective in analyzing the coherent structures of complex flows. In this article, we present the formulation and design progress of a parallel dynamic mode decomposition program, especially the parallel I/O strategy, as a significant supplementation of parallel dynamic mode decomposition algorithm presented in others literatures. Parallel I/O performance with different data block size and processor number is demonstrated with a 6.9 GB file generated by 1001 snapshots of Rayleigh-Taylor instability flow. Analysis of flow structure and spatio-temporal coherent structure are performed by Fast Fourier Transformation (FFT) and dynamic mode decomposition for flow field resulted from high order weighted essentially non-oscillatory (WENO) schemes. For test case of Rayleigh-Taylor instability flow with Atwood number A = 0.5, we find a significant phenomenon that WENO9 with very fine grid (h = 1/1920) exhibits the characteristic of large unsymmetrical bubble-like plumes, but the others take the form of symmetric bubble-like plumes.
AB - Many fluid flows of engineering interest, though very complex in appearance, can be approximated by low-order models governed by a few modes, and Dynamic Mode Decomposition (DMD) has been proved effective in analyzing the coherent structures of complex flows. In this article, we present the formulation and design progress of a parallel dynamic mode decomposition program, especially the parallel I/O strategy, as a significant supplementation of parallel dynamic mode decomposition algorithm presented in others literatures. Parallel I/O performance with different data block size and processor number is demonstrated with a 6.9 GB file generated by 1001 snapshots of Rayleigh-Taylor instability flow. Analysis of flow structure and spatio-temporal coherent structure are performed by Fast Fourier Transformation (FFT) and dynamic mode decomposition for flow field resulted from high order weighted essentially non-oscillatory (WENO) schemes. For test case of Rayleigh-Taylor instability flow with Atwood number A = 0.5, we find a significant phenomenon that WENO9 with very fine grid (h = 1/1920) exhibits the characteristic of large unsymmetrical bubble-like plumes, but the others take the form of symmetric bubble-like plumes.
KW - Dynamic mode decomposition
KW - Parallel I/O
KW - Rayleigh-Taylor instability
KW - WENO scheme
UR - https://www.scopus.com/pages/publications/85070772173
U2 - 10.1007/978-981-13-3305-7_63
DO - 10.1007/978-981-13-3305-7_63
M3 - 会议稿件
AN - SCOPUS:85070772173
SN - 9789811333040
T3 - Lecture Notes in Electrical Engineering
SP - 800
EP - 815
BT - The Proceedings of the Asia-Pacific International Symposium on Aerospace Technology, APISAT 2018
A2 - Zhang, Xinguo
PB - Springer Verlag
T2 - Asia-Pacific International Symposium on Aerospace Technology, APISAT 2018
Y2 - 16 October 2018 through 18 October 2018
ER -