International Communications in Heat and Mass Transfer 36 (2009) 1060–1066

Contents lists available at ScienceDirect

International Communications in Heat and Mass Transfer

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / i c h m t

Microchannel heat transfer and dispersion of nanoparticles in slip flow regime with
constant heat flux☆
H. Afshar a, M. Shams a,⁎, S.M.M. Nainian a, G. Ahmadi b
a
Mechanical Eng. Department, K.N. Toosi University of Technology, Pardis St., Mollasadra St., Vanak Sq., Tehran, Iran
b
Mechanical and Aeronautical Engineering Department, Clarkson University, Potsdam NY13699, USA

a r t i c l e i n f o a b s t r a c t

Available online 13 August 2009 If the hydrodynamic diameter of a channel is comparable with the mean free path of the gas molecules
moving inside the channel, the fluid can no longer be considered to be in thermodynamic equilibrium and a
Keywords: variety of non-continuum or rarefaction effects can occur. To avoid enormous complexity and extensive
Heat transfer numerical cost encountered in modeling of nonlinear Boltzmann equations, the Navier–Stokes equations can
Nanoparticle be solved considering the concepts of slip flow regime and applying slip velocity boundary conditions at the
Slip flow solid walls.
Brownian force
In this study, the Navier–Stokes and energy equations for fluid flow in a microchannel in slip flow regime
Saffman lift force
were solved analytically and temperature and velocity profiles were evaluated. The effect of different
parameters on heat transfer and dispersion of nanoparticles was discussed. Dispersion of particles due to
drag force, Saffman lift force, Brownian forces and gravity force was studied using a Lagrangian approach.
The presented results could provide guidelines for heat transfer modeling in nanofluids.
Crown Copyright © 2009 Published by Elsevier Ltd. All rights reserved.

1. Introduction
In recent years, many researchers are interested in small scale
flows and many attempts are made in minimization of scales and
improvement of the performance of instruments. By increasing the
usage of small scale instruments, understanding the behavior of such
flows has become more important. In micro-scale, rarefaction and
interferences between fluid and solid surface that causes the violation
of no slip boundary condition need to be accounted for in the analysis
[1]. Morini et al. (2005) theoretically investigated the conditions for
experimentally evidencing rarefaction effects on the pressure drop. It
was demonstrated that for a fixed geometry of the microchannel
cross-section, it is possible to calculate the minimum value of the
Knudsen number for which the rarefaction effects can be observed
experimentally [2]. Hung and Ru [3] studied the heat transfer
characteristics of fluid flow in microchannel by the lattices-Boltzmann
method. A nine-velocity model and an internal energy distribution
model were used to obtain the mass, momentum and temperature
distributions in micro-channel flow.
Hakak et al. [4] performed a two dimensional numerical
simulation for incompressible and compressible fluid flow through
microchannels in slip flow regime. The wall roughness is simulated
with triangular microelements distributed on wall surfaces to study
☆ Communicated by W.J. Minkowycz.
⁎ Corrresponding author.
E-mail address: shams@kntu.ac.ir (M. Shams).
the effects of roughness on fluid flow. Martin and Boyd [5] modeled
the fluid flow in a laminar boundary layer using a slip boundary
condition. It was shown that the slip condition changes the boundary
layer structure from a self-similar profile to a two-dimensional
structure. Although the slip condition generally leads to decreased
overall drag, two-dimensional assumption leads to the local increases
in skin friction. Other effects include thinner boundary layers, delayed
transition to turbulence, and changes in the heat transfer at the wall.
Without a thermal jump condition, slip will lead to increased heat
transfer. When a thermal jump boundary condition is added to
simulate real gases, the heat transfer decreases to below the no-slip
values. Cooling of electronic microchips is one important application
of microchannels. The heat flux in these cases can reach to 100 w/cm2
[6–8]. In 2007 Bello et al. [9] documented the geometric optimization
of a three-dimensional microchannel heat sink.
The number of studies of fluid flows and particle dispersion in
microchannels is rather limited. In particular, the dispersion and
deposition of nanoparticles in microchannel in slip flow regime is not
well understood. Earlier investigators have discussed the dispersion of
microparticles in turbulent channel flows. Arcen et al. [10] studied the
influence of using wall-corrected drag coefficients and the lift force on
the dispersed phase characteristics of a turbulent dilute gas–solid
channel flow. Tian and Ahmadi [11] conducted a series of numerical
simulations to study the transport and deposition of nano- and
microparticles in turbulent duct flows using different turbulence
models. Barber and Emerson [12] studied low Reynolds number
rarefied gas flow past a confined microsphere within a circular pipe

0735-1933/$ – see front matter. Crown Copyright © 2009 Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.icheatmasstransfer.2009.07.011
 H. Afshar et al. / International Communications in Heat and Mass Transfer 36 (2009) 1060–1066 1061

and particle fluctuating velocities, particle normal mean velocity, and

Nomenclature
particle concentration were determined.
By decreasing the particle diameter, in addition to drag, gravity and
us slip velocity
pressure forces that act on a particle, additional forces that can be
uw wall velocity
neglected for large particles will have significant roles. These forces
uf fluid velocity
include Brownian and Saffman lift forces. When a nanoparticle travels in
up particle horizontal velocity
a fluid, it subjected to the impact of gas or liquid molecules. For ultra fine
vp particle vertical velocity
particles, the instantaneous momentum imparted to the particle varies
σv Tangential Momentum Accommodation Coefficient
randomly, which causes the particle to move on an erotic path now
(TMAC)
known as Brownian motion. Saffman lift force is important for particles
σT Energy Accommodation Coefficient (EAC)
traveling through a strong shear flow.
σ Stefan Boltzmann constant
According to experimental results of different investigators [15],
λ mean free path
continuum assumption of flow in microchannels will lead to
µ dynamic viscosity
inaccurate prediction of mass flow rate. This fact shows that common
v kinematic viscosity
Navier–Stokes equations could not be used for simulation of gas flow
Kn Knudsen number
in microchannels [16]. Arkilic et al. [17] showed that the interaction
x X coordinate
between the fluid and solid surface causes an incomplete momentum
y Y coordinate
accommodation.
xp X position of the particle
Nainian et al. [18] studied the dispersion of nanoparticles in a
yp Y position of the particle
microchannel with constant wall temperature in slip flow regime. The
p pressure
equations of particle motion and energy balance were solved
H half width of the microchannel
numerically and the effect of particle diameter, starting position of
hc convective heat transfer coefficient
particles in microchannel, and velocity slip coefficient on dispersion of
y⁎ non dimensional y coordinate
particles is discussed.
θ⁎ non dimensional temperature
Knudsen number (Kn) is defined as the fluid molecular mean free
Nu Nusselt number
path (λ) over the flow characteristic length (L).
Pr Prandtle number
k thermal conductivity
λ
γ the ratio of specific heat in constant pressure to specific Kn = ð1Þ
L
heat in constant volume
Cp specific heat in constant pressure
In a microchannel, the characteristic length is the hydraulic diameter
ρ density
of the microchannel. For Knudsen number between 0.1 and 0.001, fluid
ū mean velocity
flow will be in slip flow regime. In this regime, the Navier–Stokes
Tw wall temperature
equations with slip boundary condition on solid surfaces will be valid.
Two wall temperature in the entrance of the microchannel
This slip velocity is a function of solid surface property and the molecular
Tf fluid temperature
interaction between the fluid and solid surface. For Knudsen number
Tp particle temperature
between 0.1 and 10, fluid flow is in transitional regime.
q heat flux
In this investigation, gas–solid flow in microchannel is studied
ni(t) Brownian force
according to Eulerian–Lagrangian approach. Microchannel dimen-
mp mass of the particle
sions are comparable to mean free path of the molecules so the carrier
τ particle relaxation time
phase is considered to be in slip flow regime. Dimensions of solid
dp particle diameter
particles are in nano meter scale and so dispersed phase is in
εp particle emissivity coefficient
transitional regime. Dispersion and deposition of nanoparticles in
Ap particle surface
microchannel with heat transfer is studied and the adsorption
dt time step
efficiency of the microchannel in different conditions are presented.
Gi a random number
2. Governing equations

and focused on the estimation of the hydrodynamic drag forces on a In a 2D rectangular microchannel with height equal to 2H, the
stationary sphere. Knudsen numbers covering the continuum and slip lower surface is adiabatic and a constant heat flux (q) enters form the
flow regimes (0 b Kn b 0.1) were studied whilst the Reynolds number upper wall. Eulerian–Lagrangian approach is used for modeling the
was varied between 0.01 b Re b 1. carrier and dispersed phases. The fluid flow is assumed to be steady
Desjardins et al. [13] studied gas-particle flow as dispersed phase state, incompressible and fully developed.
flows using Lagrangian approach. In this method the density function
discretized into parcels in order to reduce the processing procedure. 2.1. Fluid governing equations
Nasr et al. [14] studied the effect of particle–particle collisions and two-
way coupling on the particle deposition velocity in a turbulent channel By applying the slip boundary condition given as
flow. The time history of the instantaneous turbulent velocity vector was !


generated by the two-way coupled direct numerical simulation (DNS) of f 2 − σv Auf
us = u − uw = λ ð2Þ
the Navier–Stokes equation via a pseudospectral method. The particle σv Ay
w
equation of motion included the Stokes drag, the Saffman lift, and the
gravitational forces. The effect of particles on the flow was included in the and solving the momentum equation, the velocity profile becomes
analysis via a feedback force that acted on the fluid on the computational
grid points. Several simulations for different particle relaxation times and  
1 dp 2  y 2 2 − σv
particle mass loading were performed, and the effects of the inter-particle f
u = − H 1− +8 kn ð3Þ
collisions and two-way coupling on the particle deposition velocity, fluid 2μ dx H σv
1062 H. Afshar et al. / International Communications in Heat and Mass Transfer 36 (2009) 1060–1066

σv is the Tangential Momentum Accommodation Coefficient (TMAC). due to the impact of gas molecules on nanoparticles [21]. Brownian
By introducing nondimensional parameters y4 = Hy ; θ4 = TqH−=Tkw = force is a random force that can be expressed in terms of fluid and
4 T − Tw
Nu , and Nu = hckD = 4hkc H and applying the boundary condition particle characteristics [22].
T − Tw
given as, rffiffiffiffiffiffiffiffi
πS0
ni ðt Þ = Gi ð11Þ
h ih i Δt

j 8 2 −σ
4 σT 2γ
dθ γ + 1 Kn
= 0; θ
4
jy =1 =
4 −
T
ð4Þ 216m11 Tf
dy4 y4 = − 1 Pr S0 = ð12Þ
π 2 ρd5 S2 Cc

the energy equation leads to temperature distribution given as ρp

S= ð13Þ

 ρf
1 1 44 3 42   1 4  13 9 4
4 4 4 4
θ =1 + y − y 1 + 2n − y 1 + 3n + + n
1 + 3n4 16 8 2 16 4 k1 = 1:38410 − 23 Kj and T is the fluid temperature in Kelvin.
ð5Þ By solving the particles equation of motion, the velocity and the
position of particles are evaluated. These are,
where n4 = 4Knð2 − σ v Þ = σ v and 14 = 8½ð2 − σ T Þ = σ T ½γ = γ + 1
½Kn = Pr which are related to velocity and temperature jumps because dx
p
p
of incomplete momentum and energy interaction at the wall. =u ð14Þ
q
dt
Note that Tw = Tw0 + 2Hρc x, where ū is the average velocity and
Pu
Tw0 is the wall temperature at the entrance, which is equal to the fluid dyp p
temperature. =v ð15Þ
dt

Energy equation for a particle is given as

ZH
1
u= udy ð6Þ dTp    
2H p 4 4
−H m cp = hc Ap Tf − Tp + ep Ap σ Tf − Tp ð16Þ
dt

Here Cp is the specific heat transfer, Tp is the particle temperature,

The Nusselt number (nondimensional hear transfer coefficient) is
hc is the convection heat transfer coefficient, Tf is the fluid
given as
temperature, εp is the particle emissivity and σ = 5.67 × 10− 8 is the
Stefan Boltzmann constant.
4 4 42
−1 1 26 + 147n + 210n
Nu = +  ð7Þ
4 140 1 + 3n4 2 3. Numerical simulation

In a 2D rectangular microchannel with height equal to 2H = 4 µm

and 1 mm length, the lower surface is adiabatic and a constant heat
2.2 Particulate phase governing equations flux (q = 10 KW/m2) enters form the upper wall. Air flows by various
pressure differences in the microchannel and ferro-nanoparticles are
A particle in a microchannel is transported under the action of released from different point sources in the air flow. The density and
various forces. These include gravity, drag, Saffman lift and Brownian specific heat coefficient of particles are, respectively, equal to
forces. The corresponding particle equation of motion in x and y 7870 kg/m3 and 447 j/kg K. Air and particle inlet temperature is
directions are given as 300 K.
Volume fraction of dispersed phase is under 0.5% and, therefore,
dup 1 f p
 the suspension is quite dilute. An Eulerian–Lagrangian approach is
= u − u + n1 ðt Þ ð8Þ
dt τ used for modeling the carrier and dispersed phases and the modified-
Eulerian method is used to solve the particle momentum equations in
x and y directions. Because of the dilution, particle collision is
1 f 
p
dv p FLðSaffmanÞ
= v − v + n2 ðt Þ + −g ð9Þ neglected and it is assumed that the dispersed phase does not affect
dt τ mp
the velocity and temperature distribution of the carrier phase. The
time step was set to 10− 8 s. The fluid flow is assumed to be steady
The relaxation time (τ = ρ 18μ
p 2
Cc d
) is the characteristic time scale of state, incompressible and fully developed. Bottom and top walls of the
p p
particle. u and v are the corresponding particle velocity in x and y microchannel are assumed to be adsorbent walls and the nanopar-
directions. By increasing the relaxation time, particle reaction to ticles that reach the walls will stick to the wall (Fig. 12).
variations of fluid parameters will decrease.
The Saffman lift force for a particle traveling in a shear flow is given 4. Results
as [19],
Air flow is considered in a microchannel with 4 µm height and
!

j j
  duf 0:5
f
1 mm length under 100 KPa pressure difference. Air inlet temperature
0:5 p2 f p du is 300 k and a heat flux of 10 KW/m2 is assumed at the top surface of
FLðSaffmanÞ = 1:6154ρm d u − u Sgn ð10Þ
dy dy the microchannel while the lower surface is adiabatic. Ferro
nanoparticles with a diameter of 50 nm and a temperature of 300 K
where ρ and v are, respectively, density and kinematic viscosity of are injected form point sources at 1, 2 and 3 µm form the bottom wall
carrier phase. For nanoparticles in a microchannel, based on the value into the microchannel. The Knudsen number under such condition is
of Knudsen number, the particles are in transitional flow regime. 0.01, which is in slip flow regime.
Therefore, corrections (Cc) are needed for the Stokes drag [20] acting Variations of the computed Nusselt number versus Tangential
on the particles. In Eqs. (8) and (9), ni(t) is Brownian force which is Momentum Accommodation Coefficient (TMAC) and Energy
H. Afshar et al. / International Communications in Heat and Mass Transfer 36 (2009) 1060–1066 1063

Fig. 1. Variation of Nusselt number versus TMAC and EAC for different Knudsen numbers.

Fig. 2. Temperature profile in the microchannel (Kn = 0.01).

Fig. 3. Temperature profile in the microchannel for different Knudsen numbers (σv = 1, σT = 1).

Fig. 4. Dispersion of 50 nm particles for pressure drop of 100 KPa.

Fig. 5. Dispersion of 50 nm particles for the pressure drop of 50 KPa.

1064 H. Afshar et al. / International Communications in Heat and Mass Transfer 36 (2009) 1060–1066

Fig. 6. Dispersion of 50 nm particles for the pressure drop of 40 KPa.

Fig. 7. Dispersion of 50 nm particles for the pressure of 30 KPa.

Fig. 8. Dispersion of 50 nm particles for the pressure drop of 20 KPa.

Fig. 9. Dispersion of 50 nm particles for the pressure drop of 10 KPa.

Fig. 10. (a) Dispersion of 50 nm particles for the pressure drop of 10 KPa released at a distance of 10 µm from the upper wall. (b) Dispersion of 50 nm particles for the pressure drop of
10 KPa released from the middle of the microchannel. (c) Dispersion of 50 nm particles for the pressure drop of 10 KPa released at a distance of 10 µm from the bottom wall.
 H. Afshar et al. / International Communications in Heat and Mass Transfer 36 (2009) 1060–1066
Fig. 11. Dispersion of 50 nm particles in 10 KPa released from the middle of the microchannel. (number of particles is equal to Fig. 9).
Accommodation Coefficient (EAC) for different Knudsen numbers are
shown in Fig. 1. As can be seen, Nusselt number is more sensitive on
EAC than TMAC. By increasing the TMAC, Nusselt number decreases,
while the increase in EAC causes an increase in the Nusselt number.
Increasing the Knudsen number causes a reduction in Nusselt number
of the microchannel.
Temperature variation in microchannel is shown in Fig. 2. By
decreasing the EAC, the heat transfer coefficient decreases and the fluid
temperature in the microchannel increases. Decreasing the TMAC,
results only in a minor increase in the temperature at the center of the
microchannel. The rarefaction effect on temperature distribution in the
microchannel is shown in Fig. 3. For Knudsen number about 0.001 that
the flow can be considered as continuum or slip, variation of Knudsen
number causes a minor variation in temperature distribution but near
transitional flow regime, this effect is more significant. It can be
concluded that rarefaction has reverse effect on heat transfer coefficient
and causes an increase in the microchannel temperature.
Particle dispersion for EAC and TMAC equal to unity is shown in
Figs. 4–10.
Trajectory of 50 nm particles in 100 KPa pressure difference in the
microchannel is shown in Fig. 4. Particles travel in different paths
through the microchannel because of the random nature of Brownian
forces that act on tiny particles. The residence time of particles can be
changed by variation of pressure difference between two sides of the
microchannel, so the dispersion of particles and temperature can be
controlled. Trajectory of 50 nm particles in 50 KPa pressure difference
in the microchannel is shown in Fig. 5. The particle temperature at
microchannel outlet for 100 KPa and 50 KPa pressure difference is
about 525 k and 745 k respectively.
By decreasing the pressure difference to 40, 30, 20 and 10 KPa,
nanoparticles will not follow the stream lines anymore and the effect
of Brownian motion will be completely significant (Figs. 6–9). If the
dispersed and carrier phases would be considered as a nanofluid, that
would be considered as a homogeneous fluid except near the walls.
Fig. 12. Adsorption efficiency of the microchannel versus pressure drop.
1065
In Fig. 10 (a, b and c), dispersion of 50 nm particles in 10 KPa
pressure difference that are released from 10, 20 and 30 µm from the
upper wall of the microchannel is shown separately. As can be seen for
this pressure difference, the particles travel in such a way that the
mean path line (red line in Fig. 10) goes toward the center line. In
comparison of Figs. 9 and 11 it is seen that for 10 KPa pressure
difference if we eliminate 30% from the beginning of the micro-
channel, there is no major differences for the particles to be released
from different point sources.
Microchannel adsorption rate varies vice versa with the pressure
difference (Fig. 12). For high pressure differences (over 50 KPa)
adsorption efficiency is under 5%. By decreasing the pressure
difference, it increases rapidly. Microchannel adsorption efficiency
in 10 KPa for 100 nm and 50 nm particles is 54% and 82% respectively.
By increasing the particle diameter, adsorption efficiency decreases
which means that smaller particles can be adsorbed more in a
microchannel. This trend is shown in Fig. 13.
For 50 nm particles in 10 KPa pressure difference, the adsorption rate
is not dependent on the initial position of the particles. In other cases,
the particles that were released near the wall were adsorbed more than
particles that were released from center of the microchannel.
5. Conclusion
Decreasing the characteristic lengths of the flow causes significant
variation in drag and heat transfer coefficients. Generally in rarefied
flows, heat transfer coefficient decreases. The key to improve the heat
transfer is to add nanoparticles so that part of heat is being transported
from the microchannel without any additional pressure drop. For this
purpose the particles should be dispersed homogenously in the flow and
without deposition on the microchannel wall. To achieve to this desire,
pressure difference and entrance location of nanoparticles should be
controlled. By decreasing the particle diameter, the surface to volume
fraction increases and so heat transfer can be increased by using
nanoparticles.
Fig. 13. Adsorption efficiency of the microchannel for 10 KPa pressure drop.
1066 H. Afshar et al. / International Communications in Heat and Mass Transfer 36 (2009) 1060–1066
References
[1] X.F Peng, B.X. Wang, Proc. 10th International heat transfer conference, Brighton,
UK, Aug, 118, 1994, pp. 159–177.
[2] G.L. Morini, M. Lorenzini, M. Spiga, A criterion for experimental validation of slip-
flow models for incompressible rarefied gases through microchannels, Microfluid
Nanofluid 1 (2005) 190–196.
[3] W.Ch. Hung, Y. Ru, A numerical study for slip flow heat transfer, Applied
Mathematics and Computation 173 (2006) 1246–1264.
[4] M. Hakak Khadem, M. Shams, S. Hossainpour, Direct simulation of roughness
effects on rarefied and compressible flow at slip flow regime, International
Communications in Heat and Mass Transfer 36 (2009) 88–95.
[5] M.J. Martin, L.D. Boyd, Momentum and heat transfer in a laminar boundary layer
with slip flow, Journal Of Thermophysics And Heat Transfer 20 (4) (October–
December 2006).
[6] J. Driker, W. Liu, D. Van Wyk, J.P. Meyer, A.G. Malan, Embedded solid state heat
extraction in integrated power electronic modules, IEEE Trans. Power Electron. 20 (30)
(2005) 694–703.
[7] R. Webb, Next generation devices for electronic cooling with heat rejection to the
air, J. Heat Transfer 127 (2005).
[8] G. Hestroni, A. Mosyak, E. Pogrebnyak, L.P. Yarin, Heat transfer in micro-channels:
comparison of experiments with theory and numerical results, Int. J. Heat Mass
Transfer 48 (2005) 5580–5601.
[9] T. Bello-Ochende, L. Liebenberg, J.P. Meyer, Constructal cooling channels for micro-
channel heat sinks, Int. J. Heat Mass Transfer 50 (2007) 4141–4150.
[10] B. Arcen, A. Taniere, B. Oesterle, On the influence of near-wall forces in particle-
laden channel flows, Int. J. Multiph. Flow 32 (2006) 1326–1339.
[11] L. Tian, G. Ahmadi, Particle deposition in turbulent duct flows-comparisons of
different model predictions, Aerosol Sci. 38 (2007) 377–397.
[12] R.W. Barber, D.R. Emerson, Numerical simulation of low Reynolds number slip
flow past a confined microsphere, 23rd International Symposium on Rarefied Gas
Dynamics, Whistler, Canada, 20–25 July, 2002.
[13] O. Desjardins, R.O. Fox, Ph. Villedieu, A quadrature-based moment method for
dilute gas-particle flows, 6th International Conference on Multiphase Flow, ICMF
2007, Leipzig, Germany, July 9–13, 2007.
[14] H. Nasr, G. Ahmadi, J.B. McLaughlin, Effects of inter-particle collisions and two-
way coupling on particle deposition velocity in a turbulent channel flow, 6th
International Conference on Multiphase Flow, ICMF 2007, Leipzig, Germany,
July 9–13, 2007.
[15] E.B. Arkilic, M.A. Schmidt, K.S. Breuer, Gaseous slip flow in long micro-channels,
J. Micro-Electro-Mechanical Systems 6 (2) (1997) 167–178.
[16] M. Gad-el-Hak, Comments on “critical view on new results in micro-fluid
mechanics”, Int. J. Heat Mass Transfer 46 (2003) 3941–3945.
[17] E.B. Arkilic, K.S. Breuer, M.A. Schmidt, Mass flow and tangential momentum
accommodation in silicon micromachined channels, J. Fluid Mech. 437 (2001)
29–43.
[18] S.M.M. Nainian, M. Shams, H. Afshar, G. Ahamadi, Two phase analysis of heat
transfer and dispersion of nano particles in a microchannel, Proceedings of 2008
ASME Summer Heat Transfer Conference (HT2008), August 10–14 2008.
[19] P.G. Saffman, The lift on a small sphere in a slow shear flow, J. Fluid Mech. 2 (1965)
385–400.
[20] G. Karr, M. Owen, Drag of nano-particles in the transitional flow regime, Rarefied
Gas Dynamics: 25-th International Symposium, 2007, pp. 1112–1127.
[21] A. Li, G. Ahmadi, Dispersion and deposition of spherical particles from point
sources in a turbulent channel flow, Aerosol Sci Technol 16 (1992) 209–226.
[22] M. Shams, G. Ahmadi, H. Rahimzadeh, A sublayer model for deposition of nano and
micro-particles in turbulent flows, Chem. Eng. Sci. 55 (2000) 6097–6107.