MME Seminar: “Feather-inspired flaps for unsteady aerodynamic flow control: physics and reinforcement learning-based control” Presented by Dr. Andres Goza

About the event

Title:
Feather-inspired flaps for unsteady aerodynamic flow control: physics and reinforcement learning-based control

Presented by:
Andres Goza, Department of Aerospace Engineering, University of Illinois Urbana-Champaign

Abstract:
Birds have deployable covert feathers that are hypothesized to aid in aerodynamic flow control. Inspired by this biological control solution, we use high-fidelity simulations to explore a simplified configuration in which a flat plate is mounted via a torsional spring on an airfoil. We first characterize what dynamical regimes this system can undergo for various spring and inertia values, explaining some of the underlying fluid-structure interaction mechanism and their implications on performance changes. Then, we discuss the use of a reinforcement learning-based control law, in which the controller is allowed to actively tune the hinge stiffness to induce different fluid-structure interaction dynamics. This control paradigm can be viewed as a hybrid active-passive approach, in the sense that the spring stiffness is being actively tuned but the aerodynamic changes are indirectly obtained by the coupled interaction between the flap and the flow. We discuss the benefits obtainable through this control approach, and what changes to the flow field enable them.

Biography:
Andres Goza is an Assistant Professor in the Aerospace Engineering Department at the University of Illinois at Urbana-Champaign. He received his PhD from Caltech in 2017 and was a postdoctoral researcher at Princeton University from 2017-2018. His research focuses on the use of computational fluid dynamics and modeling techniques to gain fundamental physical insights into fluid-structure interaction (FSI) systems. Andres’s interest in these systems ranges from harnessing flow-induced vibrations for robust energy harvesting to utilizing FSI for passive control and/or estimation of unsteady aerodynamic flows. His group’s research has been funded through NSF, AFOSR, and Sandia National Labs. He was awarded an NSF Graduate Research Fellowship to perform his doctoral work, and his thesis work led to his selection as a Caltech Everhart Lecturer. He was also a “Teacher Ranked As Excellent” at UIUC in 2020-2021, and was awarded the 2021 “AIAA Teacher of the Year” award by the Aerospace Engineering Department at UIUC.