SLIDE 1 Florida State University Florida Center for Advanced Aero-Propulsion
Some Challenging Problems in (Active) Flow and Noise Control A (Limited) Experimental Perspective
Future Directions in CFD Research Aug 6-8, 2012, Hampton Roads, VA
SLIDE 2
- R. Kumar, A. Uzun, M. Y. Hussaini, A. Krothapalli,
- J. Solomon, J. Gustavsson, W. Oates, Y. Ali, C. Foster,
- E. Thomas, A. Wiley, H. Lou, V. Kumar, N. Zhuang
…. many, many more
- S. Haack, B. Cybyk (JHU/APL)
Collaborators
Acknowledgements
Research Sponsors AFOSR, AFRL, DARPA, BOEING NASA, ONR, State of Florida and others
SLIDE 3 Outline
– Supersonic Cavity Flows – Supersonic Impinging Jets – Separated Flows
- Experimental & Computational Results
– Base Flowfield – Response to Active Control
– Pulsed Microjet Actuators
SLIDE 4 acoustic waves/pressure disturbances
k M M r m
L fU St
1 2 1 1
5 . 2
m – Mode no. r – Phase delay k – Uconv/U∞
Modified Rossiter’s Model (Heller & Bliss, 1975)
structures scale
TE LE es disturbanc pressure / waves Acoustic Waves y Instabilit
Receptivity
Cavity Flows
Flow –Acoustic Resonance
Resonance observed for Low subsonic to high Supersonic Flows
SLIDE 5
Phase-locked Schlieren Images, L/D = 2, M = 0.5 (Kegerise, et al., 1999)
Cavity Flow Visualizations
Subsonic Flow Subsonic to Transonic
Krishnamurti, 1955
SLIDE 6
Cavity
Flo w
Supersonic Cavity Flow M = 2
SLIDE 7 “Quantitative” Measurements
M∞= 2, L/D =5
- High unsteady pressures throughout the cavity
- Maximum loads near the aft
SLIDE 8 Large- scale structures
L/D ~ 5
Velocity Field Phase-Conditioned
Phase-conditioned component Ensemble averaged component Periodic component
SLIDE 9
Effect of Microjet Control
Mach 2 Cavity Flow
baseline with control
SLIDE 10 Unsteady Pressure Spectra
Effect of Microjet Control
SPL dB
5 10 15 20 110 120 130 140 150 160
Control Off Control On
Cavity front
Frequency KHz SPL dB
5 10 15 20 130 140 150 160 170 180
Control Off Control On
Cavity aft
9 dB OASPL reduction 20+ dB Tonal reduction
Zhuang, Alvi, Shih AIAA J, 2006
SLIDE 11
Microjet Control
Effect of Microjet Control Unsteady Velocity, Vrms
Baseline
LES (courtesy CRAFT Tech) Ensemble-Averaged (using PIV)
SLIDE 12 Velocity Iso-surfaces
Microjets & Slot Jets Simulations*
Microjets Slotjets
* S. Arunajatesan et al. AIAA Jrnl, 2009
SLIDE 13 Effect of Microjets : Simulations*
Velocity Iso-surfaces
*Courtesy CRAFT Tech
Microjet Control Baseline
SLIDE 14 Complex Cavity Flows
* S. Arunajatesan et al., AIAA Jrnl, 2009 Microjet Control (40 psi)
SLIDE 15 Unsteady Flow Field - Simulations
Baseline Microjets Slotjets
* S. Arunajatesan et al., AIAA Jrnl, 2009
Simulations provide insight into the 3-D nature of the flow
SLIDE 16 Frequency(Hz) Prms
5000 10000 15000 20000 25000 30000 35000
110 120 130 140 150 160 170 180 190
Ground Plane Lift Plate Microphone
37-s-nc-h4-GPLPMic.lay
NPR 3.7 h = 4.0 No Control
Unsteady Pressure Loads: Ground Plane ~ 185-195 dB Lift Plate ~ 165-175 dB
Pressure Spectra
Supersonic Impinging Jets
Instantaneous Shadowgraph
SLIDE 17 Microjets (dm =400m)
de =27.5 mm
Lift plate Kulite
400 m Microjets
Control of Impinging Jets Using Steady Microjets
Frequency(Hz) Prms
5000 10000 15000 20000 25000 30000 35000
130 140 150 160 170 180 190
No Control 100 psi
37-s-nc10-h4-GP.lay
NPR 3.7 h = 4.0 Ground Plane
To date, control demonstrated for cold and hot impinging Jets Alvi et al. - AIAA J, 2003, 2006; JFM:2008; Kumar et al. AIAA J 2009
Unsteady Pressures and Noise Reduced by 4-12+ dB
SLIDE 18 Heated Mach 1.5 Jet Mean Axial Velocity (U/Uj)
Experiments & Simulations*
- Near-ideally expanded isothermal and
heated impinging jet matching experimental cases
- Re ≈ 0.9 × 106 to 1.3 × 106 h/d = 5
- Laminar nozzle inflow conditions
- Fully 3-D LES using 200 million grid
points
*Uzun, A. Hussaini, M. Y. et al, 2010 Iso-thermal Jet Total velocity magnitude iso-surfaces
SLIDE 19
Heated Mach 1.5 Jet
Experiments & Simulations
SLIDE 20 Pressure Iso-Surfaces Associated with Vortex Ring Structures at the Most Amplified Frequency using DMD
Identification of Coherent Structures
Uzun, A., Hussaini, M. Y. et al, 2012
SLIDE 21
Pulsed (Micro)Actuators
SLIDE 22 hm
(h/d)m=1-2 L/dm=1-5
L
D=1.6 mm
Cylindrical cavity
H=hm+L
dm=1mm
Po
Four micro nozzles at
the bottom of the cavity(400 μm dia) Unsteady Microjets
Solomon et al. (AIAA J 2010, 2012) NPR= Po/Pamb
Under expanded source jet
Pamb
Resonance Enhanced Microjet (REM) Actuator Schematic
Dime Nozzle (d =1mm) Cavity (L=1-5 mm) Micro nozzles h/d (1-2)
Actuator model-Gen 1
SLIDE 23 L/d= 1 fmin = 42 kHz f max= 58 kHz L/d= 2 fmin = 24 kHz f max= 36 kHz
- ΔL (1mm) produce Δf = 22 kHz
Actuator Frequency Effect of L/d & NPR
L/d= 5 fmin = 6 kHz fmax= 10 kHz
- ΔL (4mm) produce Δf = 52 kHz
Configuration
Po
Source jet
h
Impinging Cavity
L
Δ NPR ~ 1 changes the frequency from 6-10 kHz
SLIDE 24 Actuator Performance Summary
ideal m ideal
U fd St /
45 . 1
) / ( 4 .
m ideal
d H St
dm : diameter of source jet Uideal : ideally expanded velocity corresponds to the source jet NPR
Δh=0.6 mm =>Δf= 20 kHz ΔNPR=1 =>Δf= 5 kHz Δh=0.6 mm =>Δf= 13 kHz ΔNPR=1 =>Δf= 11 kHz Δh=0.6 mm =>Δf= 12 kHz ΔNPR=1.1 =>Δf= 7 kHz Δh=0.6 mm =>Δf= 6 kHz ΔNPR=1.1 =>Δf= 5 kHz
Cylindrical geometry
45 . 1 1 2 / 1 1
) / ( ) ( 1 ) ( 1 2 4 .
d H NPR RT NPR d f
SLIDE 25
Phase-Conditioned Images REM: Pulsed Actuator
Phase Averaged Instantaneous 30° - 210° ‘filling’* 240° - 0° ‘spilling’* 100 images averaged at each phase
Foster, Alvi et al. AIAA 2011
SLIDE 26
Uzun, A., Hussaini, M. Y. et al,
REM Actuator Simulations & Experiments
SLIDE 27
REM Actuator Simulations & Experiments
SLIDE 28
REM Actuator:Simulations
SLIDE 29 SmartREM Actuator
8/20/2012
Source Jet Movable Walls Controlled by Piezo-Stack Actuators Impingement Cavity Cavity Spreader
V
NPR = P0/Pamb P0 h d = 1 mm
3.43kHz 3.52 kHz
SLIDE 30 Separated Flows (& Control of)*
- Separated flow past an airfoil is
characterized by frequencies associated with the – wake – shear layer (SL) – separation bubble (SB) – actuation (if applied)
29
- “Lock-on” describes when these components are the same or are harmonics
– Harmonics could be evidence of non-linear interactions
- Effectiveness of a control strategy may be related to the presence of lock-
- n [Kotapati et al. 2001]
* Courtesy: Cattafesta, Mittal & Rowley
SLIDE 31 Experiments
- Increasingly sophisticated, providing high-fidelity data
- 2-D/Stereo/Tomo-PIV /(Plenoptic)
=> 2 component/3component/volumetric measurements
- High-speed/time resolved and Phase conditioned
- Synchronous: P, V, ρ…
- They provide significant physical insight into flow physics.
- Difficult and expensive to run; limited conditions
Actuators
- A wider array of actuators with a range of control authority and bandwidth
- Plasma Based (LAFPA, SJA, DBD); ZNMF (Synthtetic Jets); COMPAct, REM and more
- REM: Simple, robust, scale-adapt/able, appropriate complexity & capability
- Smart REM: ‘on-the-fly’ frequency control BUT more complex
- Still unclear: “which actuator and under what conditions”
Parting Thoughts
SLIDE 32 Simulations
- Increasingly sophisticated, provide high-fidelity data for increasingly complex flows
- More rigorous validation using better experimental data
- Good to excellent agreement (for some cases?)
- Provide physical insight into flow physics, provide properties not easily measured.
- Difficult and expensive to run; limited conditions
- Rarely go beyond experimental conditions?
As we Progress
- Simulations+Theory used to better explore flow dynamics;l arger range of conditions
- Provide guidance for active-adaptive control (that is realistic/feasible):
- Temporal and Spatial requirements (freq., wavelength) for actuation and
- Location (where to best place them)
- Type of actuation: momentum, body force, thermal…
- Provide guidance for (sparse/minimal, practical) sensing requirements
- Help develop, simpler/low-order/…, practical models for closed-loop control
- Plan experiments and computations together from the start
- We need to improve our communication skills
Parting Thoughts
