A database of CFD-computed flow fields around airfoils for machine-learning applications (Q6704655)

This dataset is designed to test Machine-Learning techniques on Computational Fluid Dynamics (CFD) data. It contains two-dimensional RANS simulations of the turbulent flow around NACA 4-digits airfoils, at fixed angle of attack (10 degrees) and at a fixed Reynolds number (3x10^6). The whole NACA family is spawned. The present dataset contains 2600 geometries, and 425 further geometries are published in an accompanying repository (10.5281/zenodo.4638071). For further information refer to: Schillaci, A., Quadrio, M., Pipolo, C., Restelli, M., Boracchi, G. Inferring Functional Properties from Fluid Dynamics Features 2020 25th International Conference on Pattern Recognition (ICPR) Milan, Italy, Jan 10-15, 2021