This research involves integration of Optimization, Control and Estimation Theory with Computational Fluid Dynamics into a unified framework. In our work we focus on adjoint based methods for optimization problems and Riccati based methods for feedback control and estimation problems. Such problems are often ill-posed, especially when formulated in the context of multiscale fluid systems such as high Reynolds number turbulence. Therefore, numerical solution of such problems usually requires some form of regularization.

My computational studies involve an array of model systems with a varying degree of computational complexity ranging from the periodic 1D Kuramoto Sivashinsky equation to the 3D Navier Stokes equation in wall-bounded domains. The primary applications of my research are in industrial flow control (e.g., in the aerospace industry) and in numerical weather prediction.

Dr. B. Protas

• Formulated a computational framework for adjoint based optimization of multiscale PDE systems

• Formulated a retrograde strategy for adjoint based state estimation

• Formulation and implementation of an adjoint based strategy of wake control for drag reduction

• Investigation of open loop strategies for rotary wake control