Georgia Tech's Turbulence Study Published in JFM
The study aimed to refine turbulence models and improve weather prediction and vehicle design, as stated by P. K. Yeung.
Simulation used a 32,768³ periodic cube, 35 trillion grid points, and Reynolds number 2,500.
Visualizations included 5x and 25x zooms showing tornado-like negative fluctuations.
Data are publicly available at the Johns Hopkins Turbulence Database, with Charles Meneveau as principal investigator.
Team used multiresolution protocol and INCITE allocation for simulations.
P. K. Yeung noted the scale exceeds any other machine's capacity.
2 weeks ago