Visiometrics of complex physical processes: diagnosing vortex-dominated flows

Described is an iterative refinement procedure called visiometrics that can be useful for gaining insight into data sets that result from the study of complex physical processes. These data sets are generated from numerical simulations. Visiometrics, which stands for 'visualization incorporating metrics for quantification,' consists of two steps: identifying and visualizing important features and then computing diagnostics over these features. The principal tools of visiometrics are described. The use of this procedure is then illustrated in the analysis of turbulence simulations and the vorticity-amplification process observed in turbulence simulations is characterized. This phenomenon is present in both a reduced-model (Biot-Savart) simulation, which can run on a workstation, and the more complex Navier-Stokes simulation, which runs on a supercomputer. The goal is to understand and identify vortex collapse and reconnection and, in particular, to study reduced simulations and their handling of this phenomenon. In this article, visiometrics is applied to the reduced model and the more complex simulation. In addition, the use of both a supercomputer and a graphics workstation is explored for the visualization and quantification processes.