READ ME File For 'Data-driven sparse reconstruction of flow over a stalled aerofoil using experimental data'

Dataset DOI: 10.5258/SOTON/D1752

ReadMe Author: D. W. Carter, University of Southampton [0000-0001-8675-1849]

This dataset supports the publication:
AUTHORS: D. W. Carter, F. de Voogt, R. Soares, B. Ganapathisubramani
TITLE: Data-driven sparse reconstruction of flow over a stalled aerofoil using experimental data
JOURNAL: Data-Centric Engineering
DOI: [TBD]


This dataset contains: Code for the methodology presented in the article applied to a simple flow over a clyinder and figure files corresponding to the figures in the article in MATLAB's .fig file format.
To run the code, execute 'Run_Example.m'.

The figures are as follows:

Fig. 1   Instantaneous PIV snapshots of separated flow and singular values of global basis
Fig. 2   Graphic of experimental setup and water flume facility at the University of Southampton
Fig. 3   Conceptual illustration of reconstruction method
Fig. 4   Block diagram for reconstruction metholodogies
Fig. 5   Horizontal velocity modes for global and global probes basis (500 probes, QDEIM placement)
Fig. 6   Loss functions and number of iterations before stopping during shallow neural network training
Fig. 7   Reconstruction accuracy for Q-DEIM placement
Fig. 8   Reconstruction accuracy for random placement
Fig. 9   Reconstruction accuracy comparison with laminar cylinder flow
Fig. 10   Reconstruction accuracy for Q-DEIM placement after shallow neural network refinement
Fig. 11  Reconstruction accuracy for random placement after shallow neural network refinement
Fig. 12  Extracted singular values of reconstructions for QDEIM/Random placements
Fig. 13  Comparison of instantaneous PIV field with reconstructions using 5 probes and method 1


Date of data collection: November, 2019

Licence: CC BY

Related projects:
LAMINAR CYLINDER DATASET -> If reporting results using this dataset, please cite:
AUTHORS: S. L. Brunton., J. N. Kutz
TITLE: Data-driven science and engineering: Machine learning, dynamical systems, and control
PUBLISHER: Cambridge University Press.
URL: http://www.databookuw.com


Date that the file was created: February, 2021