Collaborative Research: Physically-Based Models and System Analysis Tools for Feedback Fluid Flow Control

Northeastern University, Boston MA

Investigators

Abstract

The research objective is to develop a generic theory for model based feedback flow control. Accessible models, design and analysis methods and, indeed, clearer understanding of fluidic temporal actuation and system parameters influence of flow system behavior, as characterized by the onset or suppress of various instabilities, are the key components. This integrated theoretical and numerical research program combines unique fluid flow modeling expertise at Rensselaer Polytechnic Institute (RPI) and system analysis and control theory capabilities at Northeastern University (NEU). We envisage a closelycoupled, well-coordinated synergy between the two institutions and with an already existing collaborative team in industry and academe. A set of case studies of increased complexity will be used to isolate key stumbling blocks and develop critical enablers. Well beyond these benchmarks, effective feedback flow control will provide critical enablers for substantial improvements in a wide range of applications, ranging from aircraft, automotive, semiconductor, imaging, air conditioning, the chemical process industry, bio-reactors, furnaces, to aeroengines & ground-based gas turbines for power generation. The defining challenge in this nascent field arises from the deep chasm between the intrinsic complexity of unsteady and transient actuated fluid dynamics and the simplicity and robustness required of models useful for feedback design and analysis, and architecture (i.e., sensors and actuators placement and specifications) optimization. Addressing this challenge, the proposal is characterized by an analytic, physics-based approach and aims at the fundamentals of fluid dynamics in relation to actuation. Essential new modeling tools will be developed to efficiently characterized targeted dynamic manifolds, matched by controller and observer design and architecture optimization tools, geared to respect - and exploit - the exceptionally limited validity envelopes of such models. The selected benchmarks, characterized by significant unsteady motion and transient dynamics, include separation and stall of airfoils, the stability of the backward-facing step flow, the separated vortical wake flow behind a bluff body, the stability of shock waves over airfoils, and the transition to breakdown states in a swirling flow in a pipe. Each is important in its own right for both fundamental science and critical applications. More significantly, they highlight and isolated key technical issues in a framework which is both tractable and sufficiently rich. Intellectual Merit. Dominant techniques in the emerging field of feedback flow control such as transfer and describing functions and empirical (POD) Galerkin modes are often limited to nearly linear systems or tend to exhibit very poor capability to capture transient dynamics and actuation, and to sustain parametric changes in the operating point. The need for new tools, fresh and innovative approaches in this field, and indeed the very evolution of feedback flow control as a new and well defined discipline, are an outgrowth of these gaps. The main thrust of the research will be to develop means to address these shortcomings by an appeal to the fundamental physics of the problems, combined with a system theoretical approach. We look to develop reduced-order models which are derived from the Navier-Stokes equation and phenomenological physical characteristics, such as energy production, absorption and transfer, and natural instabilities, and account for realistic boundary conditions and meaningful actuation techniques. The research will be characterized by tight coupling of the PIs knowledge bases in system theory and fluid dynamics and leverage prolific ongoing interactions with a cross disciplinary group of experts in numerical flow simulations and imaging, experimentalists, and industrial researchers; this range of expertise is imperative, and external collaborations will serve as a source of initial insight, an impetus for new directions, and an ongoing resource and for validation of developed tools and understanding. Broad Impact. The proposed approach is unique in its synergistic use of classical and modern mathematical tools, fluid dynamics and control techniques to provide a comprehensive theory of feedback flow control, led by a collaboration between two experts in these areas. The project will lead to an improved understanding of the mechanisms that provoke instabilities and transition phenomena in various flow systems, the ways to actively control them for improved performance, and will provide rigorous means to explore and define the yet unclear feasibility scope of feedback control in fluid flow systems. These advances will impact both on the basic science and on a wide range of industrial applications involving fluid flow / aerodynamics, shear layer mixing and combustion. It will also contribute to nonlinear control theory, developing new design and analysis methods for a challenging class of systems.