Broadband Reconstruction of Inhomogeneous Turbulence Using a Combination of Data and Physics Driven Modeling
Speaker: Aditya Ghate, Science and Technology Corporation/NASA Ames Research Center
Presentation
Abstract
A new methodology to construct three-dimensional, temporally stationary but spatially inhomogeneous, incompressible turbulence is presented. The method combines the use of data-driven identification of a global basis for the fluctuation field to identify and isolate large-scale coherent motions of the flow, together with a physics-based enrichment algorithm using spatiotemporally localized wavepackets that capture the inertial subrange turbulence. These wavepackets have special properties in that they can represent large range of scales in a compressed manner and accurately capture the inter-scale physics of the energy cascade via a specific variant of the Rapid Distortion Theory (RDT) that uses WKB-asymptotics. To demonstrate the approach, we consider two distinct problems: a) wake turbulence of a dragging actuator disk impinged by homogeneous isotropic turbulence, and b) inhomogeneous wake of a realistic landing gear (LAGOON) geometry. The reconstructed flow field has single- and two-point correlations that are consistent with the reference high-resolution simulation data. This fusion of data-driven and physics-based methods enables a statistically correct reconstruction of broadband turbulent flows at a fraction of computational cost compared to a conventional scale resolving simulation method such as Large Eddy Simulation, as well as in other applications such as generation of statistically consistent inflow boundary conditions.
Biography