Forced isotropic turbulence at Taylor-scale Reynolds number 2500: structure functions and acceleration statistics

ORAL

Abstract

Turbulence simulations have been advancing steadily in size, with the most prominent drivers being to reach high Reynolds number and to achieve better resolution of the small scales. In this talk, we continue from a recent study of small-scale properties including the scaling of single-point statistics of dissipation and enstrophy fluctuations (Yeung et al. J. Fluid Mech. (in press)). In particular, we examine the scaling of velocity structure functions in the inertial range, with consideration for a protocol that replaces prohibitively long simulations at extreme problem sizes by ensemble averaging over several short simulations based on grid refinement from modest resolutions. We also address statistics of the acceleration in both Eulerian and Lagrangian frames. A single snapshot of a 32768-cubed velocity field is publicly available at the Johns Hopkins Turbulence Database.

*Supported by INCITE 2025 award of supercomputer resources at Oak Ridge Leadership Computing Facility, and NSF subcontract via The Johns Hopkins Univ. (Grant 2103874).

Publication: Small-scale properties from exascale computations of turbulence on a 32768-cubed periodic cube. Yeung P.K etal. J. Fluid Mech. (Vol 1019, R2, 2025)

Presenters

  • Pui-Kuen Yeung

    • Georgia Institute of Technology

Authors

  • Pui-Kuen Yeung

    • Georgia Institute of Technology
  • Kiran Ravikumar

    • Analytical Mechanics Associates
  • Rohini Uma-Vaideswaran

    • Georgia Institute of Technology
  • Daniel L Dotson

    • Georgia Institute of Technology
  • K.R. Sreenivasan

    • New York University
  • Stephen B Pope

    • Cornell University
  • Charles Meneveau

    • Johns Hopkins University
  • Stephen Nichols

    • Oak Ridge National Laboratory