Bulletin of the American Physical Society
72nd Annual Meeting of the APS Division of Fluid Dynamics
Volume 64, Number 13
Saturday–Tuesday, November 23–26, 2019; Seattle, Washington
Session Q33: Flow Instability: Control |
Hide Abstracts |
Chair: Mathew Juniper, University of Cambridge Room: 615 |
Tuesday, November 26, 2019 7:45AM - 7:58AM |
Q33.00001: Shape Optimization for Stability of a Cyclone Separator Matthew Juniper, Jack Brewster A cyclone separator uses a swirling flow to remove particles from a particle-laden fluid. Cyclone separators are frequently used in domestic appliances and in industry to clean contaminants from gases. At practical Reynolds and swirl numbers, the steady axisymmetric flow through a cyclone becomes linearly unstable. It develops a precessing vortex core (PVC) which is responsible for increased pressure loss and unwanted acoustic noise. Unsteady mixing caused by the precession also leads to the re-entrainment of separated particles. We perform shape optimization of a cyclone separator in order to weaken this instability. The onset of the PVC appears as an unstable global mode with azimuthal wavenumber $m = −1$. We calculate the shape gradient of the growth rate of this global mode. We identify the boundary regions that most influence the growth-rate and then use a gradient-based method to update and optimize the geometry. We interpret this physically and also present a family of orthogonal geometry changes that cause the greatest changes in the cyclone’s base-flow. Theis process highlights the geometry changes that a parametrisation must be able to reproduce in order to effectively optimize a cyclone. [Preview Abstract] |
Tuesday, November 26, 2019 7:58AM - 8:11AM |
Q33.00002: Control of Laminar-turbulent transition on a natural laminar flow airfoil John Wylie, Keith Taylor, Michael Amitay Tollmien-Schlichting (T-S) waves are the principal mechanism for laminar- turbulent transition over external surfaces. The present work aims to identify T-S waves on a natural laminar flow airfoil in the presence of an adverse pressure gradient, and subsequently reduce or eliminate the amplitude of identified T-S waves, delaying the transition to turbulence in the observed flow field. Three Piezoelectric-Driven Oscillating Surface (PDOS) actuators were used at three corresponding streamwise locations. These actuators act as dynamic surface modification, which both introduce artificial T-S waves, and subsequently cancel these waves. In the study, two upstream actuators were used to excite and phase-lock the T-S waves through systematic oscillation. The downstream actuator was used to cancel the T-S waves by applying an anti-phase disturbance at the proper amplitude. Particle Image Velocimetry was used to observe the development of the T-S waves and to characterize the three-dimensional development of the T-S waves along the airfoil model. The results indicate a dampening of the T-S wave magnitude, resulting in a delay of the transition of laminar to turbulent flow in the presence of a known adverse pressure gradient. [Preview Abstract] |
Tuesday, November 26, 2019 8:11AM - 8:24AM |
Q33.00003: Nonlinear Performance of Linear Sensor-Based Output Feedback Control of Transitional Channel Flow Yiyang Sun, Huaijin Yao, Maziar S. Hemati A large transient energy growth (TEG) of small flow perturbations can lead to laminar-to-turbulent transition in channel flow at sub-critical Reynolds number. Full-state feedback control, such as linear quadratic regulation (LQR), has been demonstrated to suppress TEG and prevent transition; however, access to full-state information is rarely possible in practice. In this study, we investigate two sensor-based output feedback control strategies: (1) static-output-feedback LQR and (2) linear quadratic Gaussian (LQG) control. These controllers use a few sensor measurements at the walls to determine the control input. We study the nonlinear effects on TEG behavior and transition control effectiveness by performing direct numerical simulations (DNS). We find that the nonlinearity serves to saturate TEG, with amplification being reduced as the initial disturbance amplitude is increased. Accordingly, the reduction of TEG in the controlled case also decreases, indicating a degradation of control performance when nonlinear effects become substantial. For LQG control, the nonlinear effects on the relation between physical and estimated states are examined. We investigate the estimator performance in DNS and further assess its influence on the control results. [Preview Abstract] |
Tuesday, November 26, 2019 8:24AM - 8:37AM |
Q33.00004: Resolvent-based Modelling, Estimation and Control of the Cylinder Wake Bo Jin, Richard Sandberg, Simon Illingworth We use a resolvent based approach for the modelling, estimation and control of the cylinder wake at Reynolds numbers between $60$ and $120$. The work has three parts. First, we consider the optimal estimation problem in which a single sensor is used to estimate the whole flow field. Second, we consider full-information control in which a single actuator uses knowledge of the entire flow field for control. Third, we consider feedback control with a single sensor for measurement and a single actuator for control. A range of Reynolds numbers is considered and the trends of optimal sensor and actuator placements for estimation and control are presented. [Preview Abstract] |
Tuesday, November 26, 2019 8:37AM - 8:50AM |
Q33.00005: Experimental Study of Laminar Separation Bubbles with Active Flow Control David Borgmann, Jesse Little Laminar separation bubbles (LSB) that form under the influence of an adverse pressure gradient are examined in a low turbulence wind tunnel. The LSBs are formed along the surface of a flat plate using a displacement body that is mounted from above. Rapid formation of spanwise vortical structures due to the primary shear layer instability promotes transition to turbulence and determines the extent of the separated region. Direct numerical simulations have shown the effectiveness of techniques that exploit the shear layer instability for active flow control (AFC). AFC in the form of 2D disturbance waves leads to the formation of coherent spanwise vortical structures that can reduce or completely eliminate the LSB. With a properly chosen forcing frequency and amplitude, the secondary absolute instability can be suppressed thus delaying transition and even relaminarizing the flow downstream of reattachment. The success of this AFC strategy is strongly affected by freestream turbulence where even modest levels can reduce the control authority due to excitation of Klebanoff modes. Therefore, the question arises whether the observed transition delay and relaminarization is possible in a real environment. This motivates a study of LSB AFC on a flat plate using DBD plasma actuators. [Preview Abstract] |
Tuesday, November 26, 2019 8:50AM - 9:03AM |
Q33.00006: Suppressing flow separation over a flat plate using machine learning Amirkhosro Kazemi, Paul Rousseau, Daniel Gomez, Aishwarya Sureshkumar Nair, Luciano Castillo, Siddhartha Verma, Oscar M. Curet Suppressing flow separation in wall-bounded flows is essential for drag reduction and consequently decreases fuel consumption, pollutant emissions, the noise of aerial and marine vehicles and affects seawall erosion. There are many mechanisms to delay flow separation, however, achieving it with minimal energy expenditure remains a challenge. Thus, it is paramount to understand how to delay the onset of transition in the presence of small disturbances. In a series of experiments, we implemented multiple cylindrical roughnesses on a flat plate to induce flow separation. In addition, we examined the effects of the roughness on the development of the boundary layer using particle image velocimetry, soap film setup as well as numerical simulations. A linear actuator is used to modulate the near-wall structure and measure the resulting change in the friction coefficient. This data is incorporated with a reinforcement-learning algorithm in order to reduce drag by delaying flow separation. [Preview Abstract] |
Tuesday, November 26, 2019 9:03AM - 9:16AM |
Q33.00007: Supersonic cavity flow control using resolvent analysis Qiong Liu, Yiyang Sun, Chi-An Yeh, Kunihiko Taira We examine supersonic flows over a 3D cavity at Mach number of 1.4 with cavity-depth-based Reynolds number of 10,000 using LES and propose a resolvent analysis based flow control technique to suppress the pressure fluctuations over the cavity. The resolvent analysis identifies the range of the forcing frequency and spanwise wavenumber for the turbulent cavity flow where energy amplifies significantly. Given these insights from resolvent analysis, we perform a series of controlled flow simulations by prescribing the unsteady actuation with various combinations of the spanwise wavelength and actuation frequency over the supersonic turbulent cavity flow. The controlled achieves over 28{\%} of reduction in pressure fluctuations. By combining the insights from resolvent analysis with DMD of the 3D cavity flows, we uncover two main mechanisms responsible for effective suppression: (1) the unsteady actuation thickens the shear layer near the leading edge and modifies its spreading rate to reduce the receptivity to acoustic disturbances; (2) the actuation disrupts the formation of large-scale spanwise vortices which mitigates the obstruction of incoming flow and alleviates the trailing-edge impingement. [Preview Abstract] |
Tuesday, November 26, 2019 9:16AM - 9:29AM |
Q33.00008: Pressure Sensor-based Flow Estimation in a Wake-flow Model: Influence of Sensor Location. Andre Popinhak, Robert Martinuzzi, Chris Morton Estimation of an unsteady velocity field with remote sensors is an important tool for many technological applications and an integral part of flow control strategies. Many works in the current literature have implemented sensor-based flow estimation via pressure-velocity correlations in wake-flows (e.g. Hosseini et al. 2015). Using the theoretical model of von K\'{a}rm\'{a}n Vortex Street, the aim of this work is to (i) understand whether sensor placement impact in the performance of the estimators. (ii) What is the benefit of using Quadratic Stochastic Estimation (QSE) relative to Linear Stochastic Estimation? (iii) What is the benefit of formulating QSE into a set of orthogonal regressors? It will be shown that the discrete pressure information location relaxes the training method requirement and a proper selection of an orthogonal basis for the estimator reduces overfitting. [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2025 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700