You are on page 1of 1

Particle-Wall Specularity Effects

On Turbulent Multiphase Flow


T.N. T.N. Dr. D. J.
Shendruk Abstract Bergstrom
Motivation
Multiphase flows involving a high concentration of solids are encountered in a variety of
applications. Currently computational fluid dynamics (CFD) using Eulerian two-fluid models is
Multiphase Flow
Multiphase flows are encountered in many different industrial and environmental finding increasing application for predicting the velocity and concentration fields in such flows.
applications. Hydro transport systems of the oil sands industry in northern Alberta are In the two-fluid model, both phases are considered to be interacting, interpenetrating fluids. The
an example of liquid solid flow forming a high-concentration slurry. Clean coal constitutive equations for the solids phase are developed using kinetic theory as applied to
technologies (CCT) offer many opportunities for the application of high solids granular flow. Most often the fluid flow is turbulent. The present research project considers the
concentration simulations. For example, pulverized coal combustion relies on application of a two-fluid model to the prediction of turbulent vertical upward flow of air and
pneumatic transport of finely ground powder to the boilers and the removal of sand particles in a rough-wall circular tube. The simulations explore the effect of surface
particulate ash from flue gas. Further, all technologies available to reduce particulate roughness and the associated specularity on the predictions for the solids phase. The wall
emission, such as Electrostatic Precipitators, Fabric Filters, Inertial Collectors and specularity factor measures the average tangential particle momentum transferred to the wall.
Scrubbers that atomize and collect pollutants, can benefit from improved prediction of Numerical simulations were performed for a range of values of the specularity factor. As the
high-concentration, fully-developed, turbulent, gas-solid flow. wall specularity increases, the change in momentum transfer at the wall begins to modify the
flow characteristics of the particle phase.
Wall Boundary Conditions
• Gas phase:
™ No Slip Simulation
CFD
•Decomposes domain into grid of small control volumes (CV) •Modelling follows Bolio et al3

•Reduces differential equations to discrete equations for each •Utilises in-house code
•Solid phase:
CV1. developed by A. Yerrumshetty.
™ Coefficient of restitution to describe normal
momentum loss.
•Flow fields estimated as linear •Fine sand particles (dp = 200μm,
ρp = 2550 kg m-3) are carried
•Linear algebraic equations at each node solved iteratively. vertically upward by air through
a pipe (R = 17.3mm). This is a
standard experimental setup3-5.

Turbulence ™ Specularity Factor, ϕ, to describe average


tangential momentum transfer.
•Specularity is fit until particle
slip matches experimental value:
Velocity

u~ fluc = 0 u~ fluc v~ fluc ≠ 0 ϕ = 0.002.


u
U
in s t

av e
u inst
=U ave
+ u~ fluc
•By varying specularity,
•Averaging Navier-Stokes results in additional term that acts as the turbulence component of the total mean stress, the
T im e

numerical simulations explore


Reynolds stress. Navier-Stokes: Averaged: Reynolds Stress effects of roughness on particle
∂ 1 μ 1 ∂U 1 ∂P 1 ∂ ⎛⎜ ∂U ⎞ ≡ ρ u iinst u inst ave
phase.
u + u • ∇ u = − ∇p + ∇ 2u + F
ave
U =− + μ −ρ u u ⎟ j ave j i inst inst

∂t ρ ρ ρ ∂r ρ ∂r ρ ∂r ⎜⎝ ∂r ⎟

j
j i j j
i j

• Computational models of multiphase flow must bias


•This suggests an eddy viscosity2. their results by fitting the specularity coefficient so
∂U ave that predicted slip matches measured slip.
Total Viscosity = Molecular Viscosity + Eddy Viscosity = μ i − ρ u iinst u inst
∂r j
j

Results Total Wall Shear Stress


1.0
0.7 Fluid Wall Shear Stress
zed
zed

0.35
Solid Phase Velocity

9 Solid Wall Shear Stress


Individu ally Normali
Individually Normali

0.8 0.6
Fluid Veloc it y

0.5
0.00082
0.6
0.07
ture

8 0.4
0.16 0.30
0.06 0.00080
Psue do-Te mpera

Concen tration
0.4 0.3
Solid Phase

0.14
7 0.2 0.05
Fluid Velocity
Fluctu ations

0.2 0.12 0.25


0.1
0.04
0. 00078
0.0 1. 0 0.0
6 0.0 0.8 0.2 1.0 0.10

Wall Shear Stress


0.2 0.6 0.4 0. 8
0.4 0.6 0.03
0.4 r/R 0.6 0.00076
Average Velocity

r/R
0.6
0.2 c to
r
0.8 0.2
0.4
or 0.08 0.20
0.8 Fa 1.0 0.0 Fact
5 0.0 ty ity 0.02
1.0
lari ular
e cu Spec 0.06
Sp
0.01 0.00074
0.04 0.15
4 0.00
0.0 0.0 0.00072
0.2 1.0 0.2
0.4 0.8 0.4
3 0.6 1.0 0.10
r/R 0.6 0.4 0.6 0.8 0.00070
r/R

0.8 0.6 0.0


0.2 or 0.8
Fluid Phase act 0.4
2
1.0 0.0 ty F 1.0 0.2 0.2 1.0
lari or
cu 0.0 Fact 0.8
Solid Phase Spe ularity 0.4 0.05
Spec 0.6
1 r/R 0.6 0.4
0.8 0.2 to r
1.0 Fac 0.00
0.0 rity
Velocity Fluctuations
0
cula
Spe 0.0 0.2 0.4 0.6 0.8 1.0
0.0 0.2 0.4 0.6 0.8 1.0

Specularit y Fact or •Solid phase velocity fluctuations intensify as specularity Specularit y


increases.
Velocity •This increases fluctuations of fluid near the centre. The energy
shifts from the peak towards the centre.
Concentration Distribution Wall Shear Stress
• As specularity increases • In smooth pipes particulate matter accumulates near the
•Fluctuations cause average velocity to decay. Wall and particle wall. • As specularity increases, shear stress on the pipe due to particles increases.
™Fluid velocity profile becomes more sharply curved.
collisions occur more frequently resulting enhanced energy • As the specularity increases the concentration profile
™Solid phase begins to flow more and more as if it obeys a no-slip
dissipation lowering velocity and damping turbulence. rises at the centre.
condition.

Conclusion Future Work


Having explored the matter of specularity, the next step in this research will be to add the hydrodynamic effects of
Most two-fluid simulations have been applied assuming unrealistic smooth wall flow. It is shown that roughness significantly affects the particle phase behaviour. roughness on the fluid phase. Research into non-vertical flow with particle phase sedimentation is the logical extension of
As the specularity and wall roughness increase, the wall shear stress due to the particles enhances causing the solid phase velocity fluctuations to become stronger. this work. This summer project is part of a larger research program which is focused on the development of improved
The increase in fluctuations smoothes the peak concentration toward the centre of the pipe. In a smooth pipe the particle phase velocity profile is flatter but as computational models for turbulent gas-solid and liquid-solid flows.
specularity increases the flow begins to behave as though it were obeying a no-slip condition and a more typical flow profile develops.
[1] S. V, Patinikar,
Patinikar, Numerical Heat Transfer and Fluid Flow. Hemisphere Publishing Corporation,
Corporation, New York, 1980.
Experimental studies on the effects of roughness on turbulent gas-solid flow is problematic but some literature is available on particle-laden flow in channels of [2] D. J. Tritton,
Tritton, Physical Fluid Dynamics. Oxford Science Publications,
Publications, Oxford, 1988.
[3] E. J. Bolio,
Bolio, J. A. Yasuna and J. L. Sinclair. Dilute turbulent gas-
gas-solid flow in risers and particle-
particle-particle interactions. AIChE Journal,
Journal, 41 (6): 1375-
1375-1388, 1995.
different roughness6. It is found that wall roughness decreases the particle phase mean axial velocity and results in a more uniform distribution of velocity [4] S. L. Lee and F. Durst. On the motion of particles in turbulent
turbulent duct flows. Int.J. Multiphase Flow,
Flow, 8 (2): 125-
125-146, 1982.
fluctuations. Increasing roughness further suppresses this result due to stronger turbulence dissipation and momentum loses at the wall. Studies on dimpled pipes5 [5] A. Yu. Varaskin,
Varaskin, M. V. Protasov,
Protasov, T. F. Ivanov,
Ivanov, and A. F. Polyakov.
Polyakov. An experimental investigation of the behaviour of solid particles
particles during their motion in smooth and
show similar trends. dimpled pipes. High Temperature,
Temperature, 45 (2): 221-
221-226, 2007.
[6] J. Kussin and M. Sommerfeld.
Sommerfeld. Experimental studies on particle behaviour and turbulence modificationin horizontal channel flow with different wall roughness.
Fluids, 33: 143–
Experiments in Fluids, 143–159, 2002.

You might also like