Experimental and numerical investigations of particle clustering in isotropic turbulence

1
Experimental and numerical investigations of
particle clustering in isotropic turbulence
Workshop on Stirring and Mixing The Lagrangian
Approach Lorentz Center Leiden, The
Netherlands August 21-30, 2006
Cornell University SUNY Buffalo Max Planck Institute
Dr. Lance R. Collins Dr. Hui Meng Dr. Eberhard Bodenschatz
Juan Salazar Scott Woodward
Dr. Zellman Warhaft Lujie Cao
S. Ayyalasomayajula Jeremy de Jong

International Collaboration for Turbulence
Research (ICTR)
2
Particle Clustering in Turbulence
Vortices
Strain Region

• Maxey (1987) Squires Eaton (1991) Wang
  Maxey (1993)
• Shaw, Reade, Verlinde Collins (1997)
• Falkovich, Fouxon Stepanov (2002) Zaichik
  Alipchenkov (2003) Chun, Koch, Rani, Ahluwalia
  Collins (2005)

3
Turbulence in Clouds
Buoyancy
Cloud Condensation Nuclei (CCN)
4
d2 Law
mass
energy

• Current microphysical models predict
• too slow condensational growth
• too narrow cloud droplet distributions

Shaw (2003)
5
Beard Ochs (1993)
At this rate, we are quite a way off from
being able to predict, on firm micro-physical
grounds, whether it will rain.
0.1 mm
1 mm
10 mm
6
Clouds in Climate Models
Visible Wavelengths
Infra Red
High, cold clouds
Low, warm clouds
Distribution of cloud cover profoundly influences
global energy balance
Raymond Shaw
7
Collision Kernel
Particle clustering impacts the RDF
Sundaram Collins (1997) Wang, Wexler Zhou
(1998)
8
Monodisperse clustering drift
Chun, Koch, Sarma, Ahluwalia Collins, JFM 2005
9
Monodisperse clustering diffusion
Chun, Koch, Sarma, Ahluwalia Collins, JFM 2005
10
Monodisperse clustering RDF
St 0.7
Chun, Koch, Sarma, Ahluwalia Collins, JFM 2005
11
Bidisperse clustering
Chun, Koch, Sarma, Ahluwalia Collins, JFM 2005
12
Bidisperse clustering
Chun, Koch, Sarma, Ahluwalia Collins, JFM 2005
13
Bidisperse clustering stationary
Chun, Koch, Sarma, Ahluwalia Collins, JFM 2005
14
Experiments and Simulations
RDF Measurements
Direct Numerical Simulations
15
Turbulence Chamber
16
Flow Characterization
Conditions at 6 Fan Speeds (MKS)







17
Metal-Coated Hollow Glass Spheres
Mean 6 microns STD 3.8 microns 1-10
particles/cm3 FV 10-7
18
Measurements of RDF
Wind Tunnel
Turbulence Box
Saw, Shaw, Ayyalasomayajula, Chuang Gylfason,
Warhaft (2006)
Wood, Hwang Eaton (2005)
19
Why 3D?
2D Sampling
1D Sampling
Relations
Holtzer Collins (2002)
20
3D Particle Position Measurement Techniques

• Particle Tracking Velocimetry (PTV)
• Advantages Lagrangian particle information
• Disadvantages Limited particle number density.
• Holographic Particle Image Velocimetry (HPIV)
• Advantages Better particle number density than
  PTV, larger 3D volume than Stereo PIV
• Disadvantages Cannot resolve time evolution of
  particles.

21
Hybrid Digital HPIV
NdYag Laser 532 nm
Numerical Reconstruction Intensity-Based Particle
Extraction
Variable Beam Attenuator
Beam Expander
Reference Beam

• n

• F

• a

• a

• F

• n

40 cm

• Z

1k x 1k CCD

• n

• a

• F

(4 cm)3 Volume

• Optical Window

22
Particle Concentration and Phase Averaging
23
Size Distribution Evolution
24
Time Dependence of RDF
25
Direct Numerical Simulations

• 1283 Grid Points
• Rl 80
• 1.2 Million Particles (one way coupling)
• Experimental Particle Size Distribution

Keswani Collins (2004)
26
Filtering by camera
Metal-coated hollow glass spheres
Mean 6 microns STD 3.8 microns
27
Filtering by camera
Metal-coated hollow glass spheres
Mean 6 microns STD 3.8 microns
28
Comparison at Rl 130
29
Comparison at Rl 161
30
Summary

• Clustering results from a competition between
  inward drift and outward diffusion
• Radial Distribution Function (RDF) is the measure
  for collision kernel
• Analysis of RDF involves Lagrangian statistics
  along inertial particle trajectories
• RDF mainly found in direct numerical simulation
• 3D measurements of RDF using holographic imaging
• Reasonable agreement between experiments and DNS
• Challenges for the measurement
• Characterizing flow (dissipation rate, e)
• Particle size distribution (will separate
  particles)
• Increasing resolution of experiment (smaller
  separations)

International Collaboration for Turbulence
Research (ICTR)