Hybrid POD-FFT analysis of nonlinear evolving coherent structures of DNS wavepacket in laminar-turbulent transition
- Submitting institution
-
Abertay University
- Unit of assessment
- 12 - Engineering
- Output identifier
- 17661718
- Type
- D - Journal article
- DOI
-
10.1063/1.4999348
- Title of journal
- Physics of Fluids
- Article number
- 084105
- First page
- -
- Volume
- 29
- Issue
- 8
- ISSN
- 1070-6631
- Open access status
- Deposit exception
- Month of publication
- August
- Year of publication
- 2017
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
1
- Research group(s)
-
D - Modelling & Simulation
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This output represents the application of a machine learning technique (proper orthogonal decomposition, POD) to this area of fluid mechanics concerning the laminar-turbulent transition of wavepackets in a boundary layer. It has shown that an unsupervised feature recognition algorithm can independently re-discover (within a few minutes) classical theories of transition to turbulence derived by experts over several decades. Furthermore, it also represents an effort towards explainable artificial intelligence (XAI) by providing a physical interpretation connecting the results of a well-known Fourier transform with the new and unfamiliar output of machine learning on a wavepacket.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
