Turbulence is regarded as the main unsolved problem in "classical physics". Due to its importance from both practical and fundamental viewpoints, it is a research area of interest to engineers (mechanical, environmental, chemical, civil, etc.), applied mechanicians, physicists, mathematicians, oceanographers, atmospheric scientists. This interdisciplinary flavor makes this an exciting and fruitful field for research. One of our main research goals is to achieve better understanding of how turbulent small-scale motion is related to the dynamics of large-scale motion, with the specific purpose of developing improved modeling tools. One of our main interests has been on testing, understanding, and developing improved subgrid-scale models for Large-Eddy-Simulation. This is an increasingly popular numerical approach to turbulent flow prediction, in which the large-scale vortices are directly simulated, while small-scale motions must be judiciously parametrized.

A research collaboration with Prof. Greg Eyink of Applied Mathematics, Prof. Shiyi Chen in Mechanical Engineering, Alex Szalay of Physics & Astronomy Department and Prof. Randal Burns of Computer Sciences centers on basic understanding of the turbulent energy cascade and on the use of new-generation database management tools to analyze very large data bases of turbulence. Various parts of this new research thrust are funded by the National Science Foundation through an ITR grant, a MRI grant, as well as by the Keck Foundation. Working on this project were, among others, former postdocs Laurent Chevillard and graduate students Yi Li and Yunke Yang. In a recent development that has received significant attention (see e.g. report in "Mechanical Engineering Magazine, December Issue, 2005"), we have found a new system of simple equations (the "delta-vee" system) that provides simple explanations for well-known trends of intermittent turbulence. For more details, click here. A new model for the full velocity gradient tensor has been proposed with Dr. Chevillard (see the paper), and a multiscale generalization has been recently explored in a collaboration with Prof. L. Biferale (Univ. Rome) and Dr. F. Toschi. With current postdoc Marco Martins-Afonso, several versions of models of the velocity gradient tensor are being studied using various expansions, forcing and coupling schemes.

In terms of tools, this project is developing novel ways to do turbulence research using databases. So far, a 1024^4 space-time history of a direct numerical simulation of isotropic turbulent flow, in incompressible fluid in 3D has been recently stored in the JHU turbulence public database cluster. The simulation was performed using 1024 grid points in each direction using a pseudo-spectral method, and forcing at large scales. The database allows access to 1024 time steps covering about one integral turn-over time-scale of the turbulence. The dataset comprises 27 Terabytes of data. The database allows users anywhere in the world to access the turbulence database for their research.

We continue developments, applications, and testing of a new technique (Renormalized Numerical Simulation - RNS) to simulate flow over multiscale objects. With Associate Research Scientist Dr. Hyung-Suk Kang, a new set of wind-tunnel experiments is being designed to study drag forces over fractal grids. This project is supported by the National Science Foundation, CBET. Recently, a new set of measurements in a shear-free mixing layer flow has been completed. In collaboration with Prof. J. Katz, a new experiment to measure drag forces over fractal trees placed in the index-matched facility for full optical access is being designed. The goal is to measure flow and drag forces and compare with RNS calculations. Furthermore, in NSF and ONR-funded projects, holographic and detailed PIV measurements of smooth and rough-wall turbulent boundary layers are performed.

In an exploratory project with postdoctoral fellow Dr. Raul Cal and in collaboration with RPI Professor Prof. Luciano Castillo, the properties of high-Reynolds number turbulent boundary layers sujected to non-standard perturbations are studied. Of particular interest is to develop improved understanding of interactions between boundary layers and arrays of wind turbines. Model experiments are being planned to be performed in the Corrsin wind-tunnel. Visiting graduate student Marc Calaf (visiting from EPFL, and in collaboration with Prof. Marc Parlange) is studying the effects of periodic helical vortices in wind-turbine wakes and how to incorporate such effects into LES of wind-farm flows. In order to perfect the numerical simulation of developing turbulent boundary layers, postdoc Dr. Guillermo Araya (in collaboration with Prof. Luciano Castillo) is developing a dynamic recycle-plane approach.