King Active Flow Control II

Papers Contributed to the Conference "Active Flow Control II 2010", Berlin, Germany, May 26 to 28, 2010
2010
ISBN: 978-3-642-11735-0
Verlag: Springer
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, Band 108, 480 Seiten

Reihe: Notes on Numerical Fluid Mechanics and Multidisciplinary Design

The interest in the field of active flow control (AFC) is steadily increasing. In - cent years the number of conferences and special sessions devoted to AFC org- ized by various institutions around the world continuously rises. New advanced courses for AFC are offered by the American Institute of Aeronautics and Ast- nautics (AIAA), the European Research Community on Flow, Turbulence and Combustion (ERCOFTAC), the International Centre for Mechanical Sciences (CISM), the von Karman Institute for Fluid Dynamics (VKI), to name just a few. New books on AFC are published by prominent colleagues of our field and even a new periodical, the 'International Journal of Flow Control', appeared. Despite these many activities in AFC it was felt that a follow-up of the highly successful 'ACTIVE FLOW CONTROL' Conference held in Berlin in 2006 was appropriate. As in 2006, 'ACTIVE FLOW CONTROL II' consisted only of invited lectures. To sti- late multidisciplinary discussions between experimental, theoretical and numerical fluid dynamics, aerodynamics, turbomachinary, mathematics, control engineering, metrology and computer science parallel sessions were excluded. Unfortunately, not all of the presented papers made it into this volume. As the preparation and printing of a book takes time and as this volume should be available at the conf- ence, the Local Organizing Committee had to set up a very ambitious time sch- ule which could not be met by all contributors.

Autoren/Hrsg.

King, Rudibert

Inhaltsverzeichnis

1;Preface;6
2;Table of Contents;9
3;Part I Airfoils;13
3.1;Transitory Control of Dynamic Stall on a Pitching Airfoil;14
3.1.1;Introduction;14
3.1.2;Experimental Setup and Procedures;16
3.1.3;Results;17
3.1.3.1;Circulation and Pressure Recovery by Successively Pulsed Actuation;17
3.1.3.2;Transitory Flow Control on a Dynamically Oscillating Airfoil;21
3.1.4;Conclusions;27
3.1.5;References;28
3.2;Unsteady Lift Suppression with a Robust Closed Loop Controller;30
3.2.1;Introduction;30
3.2.2;Flow Configuration;32
3.2.2.1;Experimental Setup;32
3.2.3;Open-Loop Control;33
3.2.3.1;Baseline Flow with and without Actuation;33
3.2.3.2;Description of Actuated Flow – Static Map;34
3.2.4;Feedback Control;35
3.2.4.1;Robust Control;35
3.2.4.2;Controller Performance in Experiments;38
3.2.5;Conclusions;40
3.2.6;References;41
3.3;Active Flow Control on a S10 Glider Configuration;42
3.3.1;Introduction;43
3.3.2;Experimental Set-Up;43
3.3.2.1;Wind Tunnel Model;44
3.3.2.2;Excitation System;45
3.3.3;Results;47
3.3.3.1;Unexcited Flow;47
3.3.3.2;Excited Flow;48
3.3.4;Conclusion;52
3.3.5;References;53
3.4;Numerical Investigation of Active Flow Control Applied to an Airfoil with a Camber Flap;55
3.4.1;Introduction;55
3.4.2;Test Configuration;56
3.4.3;Computational Method;57
3.4.4;UnexcitedFlow;60
3.4.5;ExcitedFlow;61
3.4.6;Conclusion;69
3.4.7;References;70
3.5;On Amplitude Scaling of Active Separation Control;72
3.5.1;Introduction;72
3.5.2;Scaling Analysis;73
3.5.2.1;The Velocity Ratio (VR );73
3.5.2.2;Frequency Corrected Velocity Ratio;74
3.5.2.3;Untitled;74
3.5.2.4;Reynolds Corrected Momentum Coefficient;74
3.5.2.5;Vorticity-Flux Coefficient;75
3.5.2.6;Considering the Effect of the Airfoil Incidence;76
3.5.3;Experimental Set-Up;76
3.5.4;Scaling of Lift Data;78
3.5.4.1;Baseline Airfoil Performance;78
3.5.4.2;Control Modes;80
3.5.4.3;Controlled Data;82
3.5.5;Conclusions;87
3.5.6;References;88
3.6;Lock-On to a High-Lift State with Oscillatory Forcing in a Three-Dimensional Wake Flow;90
3.6.1;Introduction;90
3.6.2;Simulation Method;91
3.6.3;Vortex Dynamics;92
3.6.4;Active Flow Control;93
3.6.4.1;Actuator Model;94
3.6.4.2;Steady Forcing;95
3.6.4.3;Oscillatory Forcing;97
3.6.4.4;Towards Feedback Stabilization;99
3.6.5;Summary;100
3.6.6;References;100
3.7;Active Flow Control on an Industry-Relevant Civil Aircraft Half Model;103
3.7.1;Introduction;104
3.7.2;Experimental Setup;104
3.7.2.1;Wind Tunnel Model and Test Facility;104
3.7.2.2;Actuator System;105
3.7.2.3;Actuator Performance Parameters;107
3.7.3;Results and Discussion;109
3.7.3.1;Definition of Relevant Quantities;109
3.7.3.2;Impact of Actuation Intensity on Lift Gain and Efficiency;109
3.7.3.3;Impact of Duty Cycle on Lift Gain;110
3.7.3.4;Impact of Phasing on Lift Gain;111
3.7.3.5;Superposition of Lift Gain;112
3.7.3.6;Comparison of Passive and Active Lift Maximization;113
3.7.4;Conclusion;114
3.7.5;References;114
3.8;Numerical Investigation of Spatially Distributed Actuation on a Three-Element High-Lift Configuration;116
3.8.1;Introduction;116
3.8.2;Swept Constant Chord Half Model;118
3.8.3;Numerical Method;118
3.8.4;Results;120
3.8.4.1;Unexcited Flow;120
3.8.4.2;Excited Flow;121
3.8.5;Conclusion and Outlook;129
3.8.6;References;129
3.9;Robust Closed-Loop Lift Control on an Industry-Relevant Civil Aircraft Half Model;131
3.9.1;Introduction;131
3.9.2;Experimental Set-Up;133
3.9.3;Plant Modeling;133
3.9.4;Controller Design and Robustness Analysis;137
3.9.5;Experimental Validation of Closed-Loop-Control;141
3.9.6;Conclusion;144
3.9.7;References;145
4;Part II Turbomachines;146
4.1;Closed Loop Blade Tone Control in Axial Turbomachines by Flow Induced Secondary Sources in the Blade Tip Regime;147
4.1.1;Introduction;147
4.1.2;Theoretical Background;149
4.1.3;Experimental Facility;150
4.1.4;Experimental Results;151
4.1.5;Closed Loop Control;152
4.1.5.1;Modeling the Effect of Air Jet Actuation;152
4.1.5.2;Model Identification and Controller Design;155
4.1.5.3;Experimental Results;156
4.1.6;Conclusions;158
4.1.7;References;160
4.2;Turbofan Tone Noise Reduction by Flow-Induced Unsteady Blade Forces;161
4.2.1;Introduction;162
4.2.2;Testcase Description;163
4.2.3;Numerical Approach;164
4.2.4;Results;167
4.2.4.1;Mode Content;167
4.2.4.2;Numerical Results for the Unsteady Pressure Field;167
4.2.4.3;Sound Radiation of the Baseline Case;170
4.2.4.4;Influence of ANC on Radiated Sound Power;172
4.2.5;Conclusion;173
4.2.6;References;174
4.3;Experimental AFC Approaches on a Highly Loaded Compressor Cascade;175
4.3.1;Introduction;175
4.3.2;Experimental Set-Up;176
4.3.3;Actuator Concepts;177
4.3.4;Measurement Systems;178
4.3.5;Results;179
4.3.5.1;Base Flow;179
4.3.5.2;Excited Flow: Sidewall Actuator;182
4.3.5.3;Excited Flow: Blade Actuator;184
4.3.5.4;Excited Flow: Combination of Both Actuator Concepts;187
4.3.6;Conclusion and Outlook;188
4.3.7;References;189
4.4;Robust Control in Turbomachinery Configurations;191
4.4.1;Introduction;191
4.4.2;Experimental Facilities;193
4.4.2.1;Control Setups;193
4.4.3;Robust H$\infty$- and Slope-Seeking Control;196
4.4.3.1;H$\infty$-Control of the Axial Fan;197
4.4.3.2;Accelerated Slope-Seeking Control of the Axial Fan;197
4.4.4;Results;198
4.4.5;Conclusion and Outlook;201
4.4.6;References;203
4.5;URANS Simulations of Active Flow Control on Highly Loaded Turbomachinery Blades;206
4.5.1;Introduction;206
4.5.2;Compressor Cascade;207
4.5.2.1;Numerical Setup;208
4.5.2.2;Base Flow Simulation;211
4.5.2.3;Active Flow Control at the Side Walls;212
4.5.3;Axial Fan;213
4.5.4;Conclusion and Outlook;220
4.5.5;References;220
5;Part III Bluff Bodies;223
5.1;Application of Active Flow Control on Generic 3D Car Models;224
5.1.1;Introduction;224
5.1.2;First Configuration: Ahmed Body with 25$\degree$ Slant Angle;226
5.1.2.1;Experimental Set-Up;226
5.1.2.2;Results for the Forced Flow;227
5.1.2.3;Closed-Loop Control;230
5.1.2.4;Combination of Actuator Systems;232
5.1.3;Second Configuration: Ahmed Body with 90$\degree$ Slant Angle;233
5.1.4;Experimental Set-up;233
5.1.5;Results;235
5.1.6;Conclusion;238
5.1.7;References;239
5.2;Simulation of Active Drag Reduction for a Square-Back Vehicle;241
5.2.1;Introduction;241
5.2.2;Numerical Setup;243
5.2.3;Reference Flow;244
5.2.4;Control Approach;246
5.2.5;Flow Control Results;247
5.2.5.1;Drag Reduction;247
5.2.5.2;Wake Structure;248
5.2.5.3;Alternative Control Approach;250
5.2.5.4;Efficiency;252
5.2.6;Conclusion;253
5.2.7;References;254
5.3;Model Predictive Control for a 2D Bluff Body Under Disturbed Flow Conditions;256
5.3.1;Introduction;256
5.3.2;Experimental Setup and Flow Characteristics;257
5.3.2.1;Experimental Setup;257
5.3.2.2;Characteristics of Natural Flow;259
5.3.2.3;Characteristics of the Flow with Pulsed Suction;260
5.3.3;Model Identification;262
5.3.4;Controller Design;264
5.3.4.1;Model Predictive Control;264
5.3.4.2;Robust Model Predictive Control;266
5.3.5;Results;267
5.3.6;Conclusions;269
5.3.7;References;270
6;Part IV Burner and Cavities;272
6.1;Closed-Loop Control of an Unstable Open Cavity;273
6.1.1;Introduction;273
6.1.2;Configuration and Reduced-Order Model for the Unstable Subspace;275
6.1.3;Modeling the Stable Subspace;278
6.1.4;Closed Loop Control: Analysis of the Compensated System;283
6.1.5;References;286
6.2;A Zero-Mach Solver and Reduced Order Acoustic Representations for Modeling and Control of Combustion Instabilities;288
6.2.1;Introduction;288
6.2.1.1;Background and Motivation for the Present Work;289
6.2.1.2;Rationale for Coupled Zero-Mach–Network Approach;290
6.2.2;Zero-Mach Solver;291
6.2.3;Linear and Nonlinear Flame Response;293
6.2.4;Acoustic Model and Coupling;295
6.2.5;Linear Stability and Limit Cycle Amplitude;296
6.2.6;Coupled Simulation and Control;298
6.2.6.1;Uncontrolled Instability;298
6.2.6.2;Control by Equivalence Ratio Modulation;300
6.2.6.3;Control of the Acoustic Boundary Conditions;300
6.2.7;Summary;301
6.2.8;References;302
6.3;Modeling the Fuel/Air Mixing to Control the Pressure Pulsations and NOx Emissions in a Lean Premixed Combustor;304
6.3.1;Introduction;304
6.3.2;Experimental Setups;306
6.3.3;Comparison:Water, Air, and Reacting Flows;307
6.3.4;Mixing Quality Model and NOxEmissions;310
6.3.5;Time Delay Measurements;312
6.3.6;Control of NOx and Pulsations;314
6.3.7;Active Optimization of Fuel Injection;316
6.3.8;Conclusion and Outlook;317
6.3.9;References;318
7;Part V Model Reduction and Feature Extraction;319
7.1;Reduced Order Modeling Using Proper Orthogonal Decomposition (POD) and Wavenet System Identification of a Free Shear Layer;320
7.1.1;Introduction;321
7.1.2;Optical Aberrations in a Free Shear Layer;323
7.1.2.1;3D Simulations;324
7.1.2.2;2D Simulations;325
7.1.3;Dynamic Model Formulation;327
7.1.3.1;Parameter Selection;329
7.1.3.2;WNARX Predictor Validation and Feedback Results;330
7.1.4;Discussion;332
7.1.5;References;333
7.2;Turbulence Control Based on Reduced-Order Models and Nonlinear Control Design;335
7.2.1;Introduction;335
7.2.2;Modeling and Control Strategy;336
7.2.3;Modelling and Control for Aerodynamic Applications;339
7.2.3.1;Generalized Mean-Field Model;339
7.2.3.2;Sliding Mode Control;342
7.2.4;Applications;343
7.2.4.1;High-Lift Configuration;343
7.2.4.2;Bluff Body;345
7.2.5;Conclusions;348
7.2.6;References;349
7.3;A New Discretization Framework for Input/Output Maps and Its Application to Flow Control;351
7.3.1;Introduction;351
7.3.2;I/O Maps of Linear Time-Invariant Systems;353
7.3.2.1;I/O Maps of $\infty$-Dimensional State Space Systems;354
7.3.2.2;I/O Map of Linearized Navier-Stokes Systems;355
7.3.3;Discretization of Signals;358
7.3.4;Approximation of System Dynamics;359
7.3.4.1;Kernel Function Approximation;359
7.3.4.2;Dynamics Approximation Error;361
7.3.4.3;Error Estimation for the Homogeneous PDE;361
7.3.5;Total Error Estimates;362
7.3.6;Application to Flow Control;362
7.3.6.1;Tests of Convergence in Signal Approximation;363
7.3.6.2;Application to Optimal Flow Control;364
7.3.7;Conclusion;365
7.3.8;References;365
7.4;Extraction of Coherent Structures from Natural and Actuated Flows;367
7.4.1;Introduction;367
7.4.2;Feature Extraction;368
7.4.2.1;Fundamental Concepts;368
7.4.2.2;Algorithms for Steady Data;369
7.4.2.3;Algorithms for Unsteady Data;370
7.4.2.4;Discrete Feature Extraction;374
7.4.3;Features of Actuated Flows – Results;377
7.4.4;Discussion and Conclusion;379
7.4.5;References;380
8;Part VI Optimal Flow Control;382
8.1;OptimizedWaveforms for Feedback Control of Vortex Shedding;383
8.1.1;Introduction;383
8.1.2;Governing Equations and Numerical Method;384
8.1.3;Sinusoidal Forcing;386
8.1.4;Optimization;387
8.1.5;Feedback;389
8.1.6;Results of Optimized Feedback Control;392
8.1.7;Conclusion;395
8.1.8;References;396
8.2;Optimal Boundary Control Problems Related to High-Lift Configurations;397
8.2.1;Introduction;398
8.2.2;A Stationary Optimal Boundary Control Problem;398
8.2.2.1;Definition of the Problem;398
8.2.2.2;Well-Posedness and Optimality Conditions;399
8.2.2.3;Numerical Solution;401
8.2.3;The Nonstationary Case;403
8.2.3.1;Model Reduction;403
8.2.3.2;Reduced-Order Model (ROM);404
8.2.4;Conclusion;409
8.2.5;References;409
9;Author Index;412

Leseproben

"OptimizedWaveforms for Feedback Control of Vortex Shedding (p. 391-392)

Won Tae Joe, Tim Colonius, and Douglas G. MacMynowski

Abstract. Optimal control theory is combined with the numerical simulation of an incompressible viscous flow to control vortex shedding in order to maximize lift. A two-dimensional flat plate model is considered at a high angle of attack and a Reynolds number of 300. Actuation is provided by unsteady mass injection near the trailing edge and is modeled by a compact body force.

The adjoint of the linearized perturbed equations is solved backwards in time to obtain the gradient of the lift to changes in actuation (the jet velocity), and this information is used to iteratively improve the controls. The optimized control waveform is nearly periodic and locked to vortex shedding. We compare the results with sinusoidal open- and closed-loop control and observe that the optimized control is able to achieve higher lift than the sinusoidal forcing with more than 50% lower momentum coefficients.

The optimized waveform is also implemented in a simple closed-loop controller where the control signal is shifted or deformed periodically to adjust to the (instantaneous) frequency of the lift fluctuations. The feedback utilizes a narrowband filter and an Extended Kalman Filter to robustly estimate the phase of vortex shedding and achieve phase-locked, high lift flow states.

1 Introduction

Previous work on flow control over an airfoil has used periodic excitation, such as unsteady mass injection or synthetic jets, to show that the oscillatory addition of momentum can delay boundary layer separation and reattach the separated flow [7, 8], or delay dynamic stall on a rapidly pitching airfoil [10]. Unsteady actuation was also shown to change the global dynamics of vortex shedding of post-stall flow, leading to higher unsteady lift than the natural shedding [13, 17].

In this paper, we investigate a simple model of a purely translating flat plate at high angle of attack at a Reynolds number of 300, where strong, periodic vortex shedding occurs. A small amplitude body force intended to mimic oscillatory mass injection is applied near the trailing edge in order to modulate the vortex shedding. We first consider open-loop control utilizing sinusoidal waveforms. It is observed that open-loop forcing can significantly amplify the lift, but feedback is required to tune the phase of actuation to a particular phase of the measured lift in order to lock the forcing with the phase shift associated with the highest period-averaged lift.

Rather than optimizing the phase of the control relative to the lift using only sinusoidal waveform, we investigate the possibility of optimizing the lift using more general (non-sinusoidal) actuation waveforms.We utilize a gradient-based approach that has been used previously in simulations to reduce the turbulent kinetic energy and drag of a turbulent flow in a plane channel [4], or to reduce free-shear flow noise[16]. Given the DNS for a particular actuator waveform, we solve the adjoint of the perturbed linearized equations backward in time to determine the sensitivity of the lift to the actuator input, and subsequently use this information to iteratively improve control."