Resolvent-based estimation and control of a laminar airfoil wake

Junoh Jung, Rutvij Bhagwat, Aaron Towne · 2024

Read Resolvent-based estimation and control of a laminar airfoil wake by Junoh Jung, Rutvij Bhagwat, Aaron Towne (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-19386.txt.

Introduction

Junoh Jung, Rutvij Bhagwat, and Aaron Towne built in 2024 a lean way to watch and calm the wake behind a wing. Vortex shedding off a NACA 0012 wing at low speed was their test case. Their resolvent-based tools match the classic Kalman filter and its control kin under like assumptions, at far less compute. Unlike those tools, theirs can fold in colored push statistics from the flow's own nonlinear terms.

Four skin sensors let the watcher call downstream flow within about three percent error in clean air and thirty in noisy air. In the noisy case, four actuators cut the wake's churn energy by ninety-eight percent. The solver is built for big machines, ready for harder flows.

Catalog record

CREATORJunoh Jung, Rutvij Bhagwat, Aaron Towne
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE115 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-19386

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Resolvent-based estimation and control of a laminar airfoil wake Junoh Jung, Rutvij Bhagwat, Aaron Towne Abstract We develop an optimal resolvent-based estimator and controller to predict and attenuate unsteady vortex shedding fluctuations in the laminar wake of a NACA 0012 airfoil at an angle of attack of 6.5 degrees, chord-based Reynolds number of 5000, and Mach number of 0.3. The resolvent-based estimation and control framework offers several advantages over standard methods. Under equivalent assumptions, the resolvent-based estimator and controller reproduce the Kalman filter and LQG controller, respectively, but at substantially lower computational cost using either an operator-based or data-driven implementation. Unlike these methods, the resolvent-based approach can naturally accommodate forcing terms (nonlinear terms from Navier-Stokes) with colored-in-time statistics, significantly improving estimation accuracy and control efficacy. Causality is optimally enforced using a Wiener-Hopf formalism. We integrate these tools into a high-performance-computing-ready compressible flow solver and demonstrate their effectiveness for estimating and controlling velocity fluctuations in the wake of the airfoil immersed in clean and noisy freestreams, the latter of which prevents the flow from falling into a periodic limit cycle. Using four shear-stress sensors on the surface of the airfoil, the resolvent-based estimator predicts a series of downstream targets with approximately 3 % 3\% and 30 % 30\% error for the clean and noisy freestream conditions, respectively. For the latter case, using four actuators on the airfoil surface, the resolvent-based controller reduces the turbulent kinetic energy in the wake by 98 % 98\% . 1 Introduction The laminar flow over an airfoil is a canonical problem in fluid mechanics due to its role in aerodynamics and status as a prototypical problem for studying wakes. Unsteady perturbations in the wake are of particular interest for several reasons. First, these perturbations are intimately tied to the separation bubble that forms over the suction side of the airfoil, which in turn increases drag ( Alam et al., 2010 ; Chang et al., 2022 ) . Second, unsteady fluctuations can degrade aerodynamic performance in many aircraft flight control scenarios, such as maneuvers at high angles of attack, landing, takeoff, or encountering atmospheric turbulence. Third, wake perturbations significantly contribute to aerodynamic noise, which is a concern for wind turbines ( Wagner et al., 1996 ; Agrawal et al., 2015 ) and rotorcrafts, including drones. In all these scenarios, accurate estimation and effective closed-loop control of wake perturbations are critical to improve engineering performance, i.e., to reduce drag, enhance flight control, and mitigate aerodynamic noise. A dominant feature of the laminar flow over an airfoil is vortex shedding in the wake. Vortices form on each side of the airfoil and shed periodically, creating downstream flow patterns such as the Karman vortex street. Factors including the shape of the object, the angle of attack, and the Reynolds number influence this vortex shedding. In NACA0012 airfoils at moderate angles of attack (6-10 degrees), an anticlockwise vortex is generated at the trailing edge, and a clockwise vortex at the front suction side separates and is entrained downstream ( Chang et al., 2022 ) – accordingly, both the front section of the suction side and the trailing edge impact vortex-shedding. Many studies have been conducted to suppress vortex shedding in the wakes behind cylinders ( Jin et al., 2020 ; Déda et al., 2023 ; Lin and Tsai, 2024 ) and airfoils ( Colonius and Williams, 2011 ; Broglia et al., 2018 ) . An accurate estimator is essential for successful closed-loop control ( Stengel, 1994 ; Brunton and Noack, 2015 ) . Standard estimation and control methods such as the Kalman filter ( Kalman, 1960 ) and the linear-quadratic-Gaussian (LQG) controller have been applied to fluid mechanics problems over the past two decades, e.g., Kalman filter ( Rafiee et al., 2009 ; Colburn et al., 2011 ; An et al., 2021 ) and LQG control ( Bagheri et al., 2009 ; Fabbiane et al., 2014 ; Fabbiane et al., 2015 ; Fabbiane et al., 2017 ; Sasaki et al., 2018a ) . However, these standard methods have two significant limitations when applied to flow-control problems. First, solving the Riccati equations required to obtain the Kalman and LQG gains scales poorly with problem size and becomes computationally expensive, or in some cases prohibitive, for the large systems typically obtained when discretizing the Navier-Stokes equations. This issue can be partially mitigated by reducing the size of the system a priori via some model-reduction method ( Pasquale et al., 2017 ; Gomez et al., 2019 ) , but this potentially degrades the performance of the controller. Second, standard methods model the nonlinear terms of the Navier-Stokes equations as a white noise forcing on the linear dynamics. These terms are not white ( Zare et al., 2017 ; Towne et al., 2017 ; Morra et al., 2021 ) , and treating them as such deteriorates estimation and control performance ( Martini et al., 2020 ; Amaral et al., 2021 ) . In this paper, we utilize a recently developed class of estimation and control methods based on resolvent analysis ( Towne et al., 2020 ; Martini et al., 2020 ; Martini et al., 2022 ) . Resolvent analysis, or input-output analysis, is a powerful methodology based on a linear mapping between forcing (input) and response (output) modes and the gain between them. These modes and gains are obtained from a singular value decomposition of the resolvent operator obtained from the linearized Navier-Stokes equations. Early studies viewed the input as an external forcing on the Navier-Stokes equations linearized about some laminar fixed point ( Jovanović and Bamieh, 2005 ; Sharma et al., 2006 ; Sipp et al., 2010 ) . McKeon and Sharma (2010) extended resolvent analysis to turbulent flows by interpreting the nonlinear terms from the Navier-Stokes equations as a forcing on the linear dynamics. Towne et al. (2018) showed that resolvent modes and spectral proper orthogonal decomposition (SPOD) modes are identical when the forcing is white noise, directly linking resolvent modes to coherent structures observed in flow data. However, resolvent analysis poses computational challenges for high-dimensional problems. To address these difficulties, Ribeiro et al. (2020) applied randomized SVD. Martini et al. (2020) used a time-stepping approach to obtain the action of the resolvent operator on a specified forcing, eliminating the need to compute the inverse of large matrices. Farghadan et al. (2023) ; Farghadan et al. (2024) further extended the range of problems amenable to resolvent analysis by combining randomized SVD with an approach to minimize the cost of the time-stepping method and successfully applied the method to several three-dimensional problems. Resolvent analysis has been extensively used to study the flow over airfoils at different Reynolds numbers and angles of attack. Thomareis and Papadakis (2018) investigated the physics of separated and attached flows around a NACA 0012 airfoil at \emph{Re}_{L_{c}}=50,000 and angle of attack 5^{\circ} . Symon et al. (2019) investigated two angles of attack, 0^{\circ} and 10^{\circ} , for a NACA 0018 airfoil and showed that these two cases behave as an oscillator and amplifier ( Huerre and Monkewitz, 1990 ) , respectively. Kojima et al. (2020) identified the origin of the two-dimensional transonic buffet over a NACA 0012 airfoil at \emph{Re}_{L_{c}}=2,000 , Ma_{\infty}=0.85 and \alpha=3^{\circ} . Marquet et al. (2022) studied

Read the complete text