Effect of wall heating on turbulent boundary layers with temperature-dependent viscosity
Skin-friction reduction by viscosity stratification
- To investigate the effects of wall heating on turbulent skin friction in turbulent boundary layer
¡¤ Direct numerical simulation (DNS)
¡¤ Fractional step algorithm with a local volume-flux formulation (Rosenfeld, Kwak & Vinokur 1991)
¡¤ The Arrhenius-type viscosity model for water (White 2006)
¡¤ Inflow simulation from the by-pass transition due to isotropic free-stream turbulence (Jacobs & Durbin 2001)
Schematic diagram of computational domain
- The effects of the wall temperature for high Prandtl-number fluid — 30, 70 and 99¡ÆC
Decreased skin friction due to the decreased viscosity near the heated wall,
¡¤ Weakened turbulent energy in the outer region
¡¤ Increased near-wall energy which leads to enhanced dissipation
¡¤ Decreased growth rate of the boundary layer
Skin-friction coefficient (Cf) and drag reduction rate (DR). T¡Ä=30¡ÆC.
Vortical structures with the lambda-2 criteria
(Lee et al., Journal of Fluid Mechanics, Vol. 726, pp.196-225, 2013)
Flow Control
Inflow pulsation on a turbulent coaxial jet
- To investigate dynamic and mixing properties in a confined coaxial jet flow with pulsating turbulent inflow
¡¤ Dynamic SGS model (Germano et al. 1991)
¡¤ Fully-implicit fractional step method (Kim et al. 2002)
¡¤ Fully conservative finite difference scheme (Morinish et al. 2004)
¡¤ Pressure Poisson equation solver: V-shape multigrid method
¡¤ Bounded QUICK scheme for scalar advection (Herrmann et al. 2006)
- Inflow pulsation
- St effects
¡¤ 0.1<St<0.9
¡¤ Optimal pulsation frequency for the reduction of reattachment length: St=0.327
¡¤ Optimal pulsation frequency for the mixing enhancement: St=0.180
( Jang & Sung, International Journal of Heat and Fluid Flow, Vol.31, pp.351-367, 2010)
Drag reduction
Ultrasonic forcing for skin friction reduction
- Objective
¡¤ Reducing skin friction in a turbulent boundary layer
- Schematic diagram of Experiment
- Ultrasonic cleaner mechanism
- Forcing effect (mean velocity)
- Forcing effect (turbulent intensities)
¡¤ Streamwise direction turbulent intensity is decreased (reduced skin friction)
¡¤ Wall normal direction turbulent intensity is increased (forcing effect)
Turbulent boundary layer control using periodic suction/blowing
- Schematic diagram of Experiment
¡¤ Periodic suction and blowing through a spanwise slot
¡¤ Effect of forcing frequency and forcing angle ¥á
- Control mechanism
- Vortex spatial fraction
¡¤ A spanwise vortex which is generated frequency is more effective in reducing skin friction.
- Experimental results
¡¤ Periodic suction/blowing reduces the wall region velocity so the velocity profile is upward shifted.
¡¤ At that time, effect of higher frequency forcing and reverse forcing angle persists farther downstream.
- Wall region ¥øz and velocity behavior with forcing frequencies
- Wall region ¥øz and velocity behavior with forcing angles
- Experimental results
¡¤ High frequency and ¥á=120¡Æ are more effective to reduce skin friction
¡¤ Spanwise vortex retards wall region velocity
DNS of turbulent boundary layer with local forcing
- Objective
¡¤ Effects of local forcing on skin friction in a turbulent boundary layer
- Flow configuration
¡¤ Spatially developing turbulent boundary layer
- Numerical method
¡¤ DNS by fully-implicit decoupling method
- Vorticity field after the slot
- Convectional forcing parameter ¥ò=vW/U¡Ä¥è
DNS of TBL with periodic blowing through a spanwise slot
- Objective
¡¤ To investigate the effects of unsteady blowing on turbulent boundary layer
- Insensitivity of long time mean velocity for unsteady forcing
- Convection of lower wall-shear stress region
- Unsteady forcing enhances energy redistribution
- Spanwise vortical structure inducing a reverse flow near the wall reducing wall shear stress periodically
- Pressure strain correlation tensor
Experiment with local forcing in turbulent boundary layer
- Understanding the flow structure behind the local blowing/suction in a flat plate boundary
- Skin friction was reduced up to 40% compared with the undisturbed boundary layer
Control of TBL by local spanwise oscillating Lorentz force
- Objective
¡¤ Drag reduction by Lorentz force
¡¤ Effects of local forcing on turbulent near wall coherent structures
- Numerical method
¡¤ DNS by fully-implicit decoupling method
Spanwise wall oscillation
- Objective
¡¤ Drag reduction
- Flow configuration
¡¤ Periodic channel / pipe flows
- Spanwise wall oscillation
- Scale similarity on drag reduction
- Conditionally-averaged flow fields (Streamwise Vorticity ¥øx and streaks u¡¯)
Control of turbulent channel flow
- Objective
¡¤ Drag reduction
- Flow configuration
¡¤ Periodic channel flows
- Suboptimal control algorithm
¡¤ Spatial optimum
- Actuators: suction / blowing (¥õ2), spanwise sliding velocity (¥õ3)
- Control mechanism and instantaneous vorticity field
- Assessment of suboptimal control
Micro flow control system
Particle Image Velocimetry
- Velocity field measurement technique using particle images in fluid flow
- Displacement of particle images between the Nd:Yag laser pulses is determined through spatial correlation
¡¤ Statistical processing using correlation between two interrogation windows
¡¤ Cross-correlation method covers linear shifts only (interrogation window size should be sufficiently small)
Influence of unsteady wake on a backward-facing step flow
- Control of separated flows over a backward-facing step
- Flow configuration
¡¤ Backward-facing step flow at ReH=560
¡¤ Wake disturbance within 10% of U¡Ä
- Experimental results
¡¤ Effective flow control at StH=fH/U¡Ä=0.2
Influence of unsteady wake
- Objective
¡¤ Flow control
- Flow configuration
¡¤ Turbulent separation bubble
- Phase averaging technique
¡¤ Effective means of investigation
- Velocity: Hot-wire anemometry
- Pressure: Microphone array
- xR measurement
¡¤ Further reduction of xR on the ¡®pressure¡¯ side
(¡®wrapping¡¯ regime, ¡®cutting¡¯ regime)
- Spreading rate
¡¤ Larger spreading rate at StH=0.2
¡¤ Further enhancement on the ¡®pressure¡¯ side (CCW)
- Turbulence intensity
¡¤ Larger turbulent intensity at StH=0.2
¡¤ Further increase for CCW
¡¤ Active enhancement of rolled-up vortices
¡¤ Reduction as StH increases
¡¤ Further reduction for CCW
- Pressure spectrum for CCW
¡¤ Significant increase at x/H=0.8
¡¤ Fast decay after xR/H=2.4
- Coherence (St=0 and x0/H=0.8)
¡¤ Flapping motion at St=0.02 : 0<¥î/H<3
- Coherence (St=0 and x0/H=8.0)
¡¤ Weakened flapping motion : -8<¥î/H<-3
¡¤ Shedding motion St ¥î/H=Const. after reattachment (white dotted curve)
- Coherence (StH=0.02 and x0/H=0.8, CW)
¡¤ Coherent motion of St=0.02 and its harmonics
¡¤ Broad coherent region of St=0.37 observed
- Coherence (StH=0.02 and x0/H=0.8, CCW)
¡¤ Weakened coherent motion of St=0.02
¡¤ Significant broad coherent region of St=0.3
¡¤ Possible cause of different behavior between suction and pressure sides
- Coherence (StH=0.20 and x0/H=0.8, CW)
¡¤ Coherent motion of St=0.2 and first harmonic at St=0.4
¡¤ Broad coherent region of St=0.14
- Coherence (StH=0.20 and x0/H=0.8, CCW)
¡¤ Weakened coherent motion of St=0.2
¡¤ Significant repetitive motion (spatial version of unsteady wake)
¡¤ High coherence means xR reduction
Large-scale vertical structure of turbulent separation bubble
- Objective
¡¤ Quantitative visualization of large-scale vertical structure of turbulent separation bubble utilizing pressure-velocity correlation
- Flow configuration
¡¤ Flows over a thick blunt body with unsteady wake by a wake generator
¡¤ ReH=560 (comparison purpose with the results by Hwang et al..2001)
¡¤ Glycerol StH=0, 0.2 (clockwise and counter-clockwise rotation of a wake generator)
- Method : Spatial box filtering to smooth out the 3-dimensional irregularities on spanwise large-scale vortices ( Lee and Sung, 2002)
- Formation of large-scale vertical structure and its convection
- Control effect of unsteady wake by reducing vertical structure
- CCW more effective than CW for the more active interaction between turbulent separation bubble and unsteady wake
- Dipole noise calculation with Curle¡¯s extension formula
- Noise source region reduced by interacting with wake
- Noise source strengthened by active interaction between separation bubble and unsteady wake
Local forcing over a backward-facing step
- Objective
¡¤ Separation control
- Flow configuration
¡¤ Turbulent backward-facing step
¡¤ On the separation edge at 45¡Æ
¡¤ Effective means of flow control
- Hot-wire anemometry
¡¤ Definition of local focing
¡¤ Time-mean flow field
¡¤ Velocity spectrum
- Local forcing
¡¤ Effective forcing frequency: StH=0.27
¡¤ Decrease of xR with increasing A0
¡¤ Two minima
StH=0.27 : shedding type instability
StH=0.40 : shear layer instability
- Velocity spectrum
¡¤ Effective forcing frequency: StH=0.27
¡¤ Vortex merging: StH/2, StH/4
¡¤ Initial region (x/H=1)
Enhancement of vortices
Increased turbulence level
¡¤ Reattachment region
Earlier reattachment
No forcing: xR/H=7.8
Effective forcing: xR/H=4.9
Control of turbulent separated and reattaching flow by spanwise local forcing
- Objective
¡¤ Mixing enhancement of turbulent separated flow
¡¤ Control of coherent structure of turbulent separated flow
¡¤ Control of noise radiation from wall pressure fluctuations
- Control strategy
¡¤ Local forcing by using acoustic excitation
¡¤ Spanwise variation(d,t)
¡¤ Swirling effect(¥è)
Control of separated flows by local variations at separation edge
- Objective
¡¤ Mixing enhancement at separation edge (passive control)
- Velocity measurement by PIV
- Wall pressure fluctuations by microphones
DNS/LES of backward-facing step flow
- Numerical method
¡¤ Implicit velocity decoupling procedure of Kim et al.(2002)
¡¤ Multi-grid method to solve pressure Poisson¡¯s equation
¡¤ Inlet boundary condition : Lund et al.(1998)
¡¤ Parallelization using MPI (Message-Passing interface)
- Physics
¡¤ Flow control with local forcing
Unsteady RANS and CFD
- Turbulence heat transfer model
¡¤ Low-Reynolds-number 4-equation model(IJHMT 1996, IJHFF 1997)
¡¤ Algebraic heat transfer model (IJHMT 2000)
- Computation Fluid Dynamics
¡¤ Inflow condition for RANS closure (AIAA J.2000)
¡¤ New decoupling algorithm for unsteady RANS and LES(on going)
- Unsteady RANS computation
¡¤ Locally forced turbulent boundary layer with heat transfer( on going)
¡¤ Locally forced backward-facing step( FDR 2000)
¡¤ Enhancement of heat transfer in backward-facing step ( FDR 2000)
¡¤ Impinging Jet with periodic inflow oscillation (on going)
- Locally-forced backstep flow
- Enhancement of heat transfer
Control of vortex shedding from a cylinder
- Phase relations & Control result
- System identification in lock-on range
- Feedback control system
Lock-on behavior of the cylinder wake
- Alternate shedding of vortices in the near wake leads to large fluctuating pressure forces in a direction transverse to the flow
¡¤ natural shedding frequency
- Rotary oscillation of a cylinder is applied
- Maximum value of CL and time of negative maximum CL
¡¤ Phase change of the vortex shedding, relative to the body oscillation
Large-scale vertical structure of turbulent separation bubble
- In the forced wake flows, vortex shedding is entrained by the cylinder motion.
- Self-excited vortex shedding frequency synchronizes with the forcing frequency provided that the two frequencies are not too different
- If the difference in the two frequencies is large enough, a quasi-periodic or chaotic oscillation may occur.
- Power spectra on the cylinder wake & Frequency selection
Response to superharmonic excitation
- The lock-on of frequency still occurs when the ratio between the natural frequency of the self-excited oscillation and forcing frequency is in the neighborhood of an integer (different from unity) or a fraction(Hayashi 1985). => superharmonic lock-on and subharmonic lock-on
- Schematic diagram of the vortex formation patterns
Control of separated TBL by local Lorentz force
- Objective
¡¤ Separation control by local forcing
¡¤ Understanding the characteristics of the separated turbulent boundary layer
- Flow configuration
¡¤ Spatially developing turbulent boundary layer with separation bubble
- Numerical method
¡¤ Direct Numerical Simulation