energies

Article
Numerical Simulation of Turbulent Structure and Particle
Deposition in a Three-Dimensional Heat Transfer Pipe
with Corrugation
Hao Lu 1,2 , Yu Wang 1 , Hongchang Li 3, * and Wenjun Zhao 1,2

1 Laboratory of Energy Carbon Neutrality, School of Electrical Engineering, Xinjiang University,

Urumqi 830047, China; [email protected] (H.L.); [email protected] (Y.W.);
[email protected] (W.Z.)
2 Center of New Energy Research, School of Future Technology, Xinjiang University, Urumqi 830047, China
3 School of Electrical Engineering, Xinjiang University, Urumqi 830047, China
* Correspondence: [email protected]

Abstract: When colloidal particles are deposited in a heat transfer channel, they increase the flow
resistance in the channel, resulting in a substantial decrease in heat transfer efficiency. It is critical to
have a comprehensive understanding of particle properties in heat transfer channels for practical
engineering applications. This study employed the Reynolds stress model (RSM) and the discrete
particle model (DPM) to simulate particle deposition in a 3D corrugated rough-walled channel.
The turbulent diffusion of particles was modeled with the discrete random walk model (DRW). A
user-defined function (UDF) was created for particle–wall contact, and an improved particle bounce
deposition model was implemented. The research focused on investigating secondary flow near the
corrugated wall, Q-value standards, turbulent kinetic energy distribution, and particle deposition
through validation of velocity in the tube and particle deposition modeling. The study analyzed
the impact of airflow velocity, particle size, corrugation height, and corrugation period on particle
deposition efficiency. The findings suggest that the use of corrugated walls can significantly improve
the efficiency of deposition for particles less than 20 µm in size. Specifically, particles with a diameter
of 3 µm showed five times higher efficacy of deposition with a corrugation height of 24 mm compared
to a smooth surface.
Citation: Lu, H.; Wang, Y.; Li, H.;
Zhao, W. Numerical Simulation of
Keywords: particle deposition; corrugated rough wall; discrete particle model; numerical simulation;
Turbulent Structure and Particle
Q criterion
Deposition in a Three-Dimensional
Heat Transfer Pipe with Corrugation.
Energies 2024, 17, 321. https://
doi.org/10.3390/en17020321
1. Introduction
Academic Editor: Nadir Yilmaz
Heat exchangers are widely used in power generation and in heating, ventilation,
Received: 8 November 2023 and air conditioning (HVAC) systems in everyday life and industrial production. After a
Revised: 6 December 2023 long period of use, dirt deposits tend to form on the surface of a heat exchanger, which
Accepted: 9 December 2023 affects the heat transfer efficiency of the heat exchanger. In the case of HVAC systems, such
Published: 9 January 2024 fouling also affects indoor air quality (IAQ), and when the particle size (dp ) is less than
2.5 µm, particulate matter enters the human body through the respiratory tract, posing a
serious threat to human health [1]. Since most people spend most of their time indoors [2],
the study of particle motion characteristics in heat exchangers has received much attention.
Copyright: © 2024 by the authors.
The structure of the heat exchanger is a key factor influencing the heat transfer perfor-
Licensee MDPI, Basel, Switzerland.
mance and particle deposition characteristics. Smooth-walled heat exchangers have been
This article is an open access article
studied more comprehensively [3–5]. Lai et al. [6,7] suggested the addition of low ribs to
distributed under the terms and
enhance particle deposition in the pipe and to control the number of particles in the pipe
conditions of the Creative Commons
Attribution (CC BY) license (https://
wall by periodic cleaning through removable sections. By using the CFD approach, Lu
creativecommons.org/licenses/by/
and Lu [8] investigated how rib shape affected particle deposition in turbulent pipes. The
4.0/).
impacts of square, triangular, and circular fins on particle deposition were examined. It

Energies 2024, 17, 321. https://doi.org/10.3390/en17020321 https://www.mdpi.com/journal/energies

Energies 2024, 17, 321 2 of 27

was found that different rib shapes have a great influence on particle deposition and also
produce different resistance. Lu and Quan [9] discussed particle deposition in 3D ribbed
channels under different conditions. Bi et al. [10] compared and analyzed the influence
of micro-dimpled channels, cylindrical grooved channels, and low-ribbed channels on
fluid pressure loss through the field synergy theory. It was found that the micro-dimpled
channel and the cylindrical grooved channel had less fluid resistance than the low-ribbed
channel. Curved ribs significantly reduce the energy loss of a pipe compared to rectangular
ribs. Han and Lu [11] investigated the deposition of particles on the walls of dimple-like
concave sockets and explored the effects of different dimple depth-to-inner diameter ratios
as well as dimple inner diameter-to-dimple spacing ratios on the deposition of different
particles and found that deposition was facilitated for particles smaller than 10 µm in size
and had little or no effect on particles larger than 10 µm in size. Andaz and Maso [12]
investigated the deposition of particles in a ribbed channel after the addition of a deflector
using the Launder, Reece, and Rodi (LRR) model [13] and the Lagrange model and found
that the deposition rate of particles with a particle size of 50 µm increased by 148.36%.
Another hotspot for the study of fluid–solid coupling in heat exchangers is the study of
nanofluids with heat exchangers. In this area, the research of Ben Hamida’s team is more
comprehensive and influential. Ben Hamida et al. [14] conducted a numerical analysis
of the thermal management of light-emitting diode packages. They simulated the heat
dissipation process from an LED chip to the heat sink. To increase the junction temperature
of a square LED lamp, square and circular holes were drilled in the center of the heat sink.
The researchers found that the junction temperature decreased with an increase in the
internal surface of the hole. Ben Hamida et al. [15] used the Galerkin finite-element method
(GFEM) to simulate the heat transfer process inside a hybrid nanofluid-filled channel under
the action of an electric field. The effect of fin geometry on heat transfer was investigated
using the central composite design (CCD) method. The results showed that each increase
of 0.01 in nanoparticle concentration increased the Nusselt number by 5.19%. In addition,
the effect of using nanofluids in a two-tube heat exchanger with an axial spoiler on the heat
transfer rate and pressure drop in a turbulent flow regime has been investigated [16]. The
problem of magnetohydrodynamic nonconstant natural convection heat transfer in a circu-
lar shell with four heated cylinders in each of the horizontal and vertical midplanes was
investigated using a two-dimensional numerical computational method [17]. The efficiency
of carbon nanotube (CNT)-50% water +50% ethylene glycol nanofluid in a corrugated
finned tilted box radiator under Lorentz force was investigated by numerical simulation
considering the nanofluid radiation effect [18]. Corrugated wall piping is one of the more
common types of piping used in engineering to enhance heat transfer efficiency and save
energy. Russ and Beer [19] discussed the effect of pipes with sinusoidal corrugated walls on
the flow of heat transfer. Their study considered working conditions at different Reynolds
numbers, from laminar to turbulent flow. Using the finite-volume method (FVM), Heidary
and Kermani [20] found that horizontal bellows can improve the transfer rate by more than
50% compared to smooth tubes. There is a large number of studies on the bellows thermal
transfer mechanism [21–25], but there are fewer studies on the deposition of particles on
corrugated walls. Therefore, there is a need for studies on particle deposition in heat
transfer ventilation channels with corrugated wall surfaces.
One of the most important aspects of the study of particle deposition and diffusion in
heat exchangers using numerical modelling is the simulation of turbulence. There are now
three main techniques for modeling turbulent flow: Reynolds-averaged Navier–Stokes
(RANS) simulation [26,27], direct numerical simulation (DNS) [28,29], and large eddy
simulation (LES) [30,31]. Several RANS turbulence models are the most widely used turbu-
lence simulation techniques at present because DNS and LES methods are far from being
applied to practical engineering applications due to the high requirements on computing
resources. The results show that the RSM is a relatively accurate turbulence model [3]. This
really is owing to the fact that other turbulence models disregard mistakes introduced by
anisotropies during flow, while the RSM model takes into account turbulent anisotropy.
Energies 2024, 17, 321 3 of 27

In addition, there are Markov chain models for transient calculations, which are mostly
used for indoor colloidal particulate matter studies due to their better robustness [32–34].
Sajjadi et al. [35] used the lattice Boltzmann method (LBM) in conjunction with the LES as
well as the RANS methods to study particle deposition in turbulent flows. The process of
dust growth was investigated using the CFD dynamic mesh technique of Zheng et al. [36].
The study of deposition–diffusion of particulate matter in heat exchangers is a typical fluid–
solid coupling problem, and in addition to the choice of turbulence methods, the choice
of fluid–solid coupling methods is also very important. The Euler–Euler method and the
Euler–Lagrange method are currently more widely used. Li et al. [37] were the first to use
the Euler–Euler model to calculate the particle deposition problem and found that it had
the same accuracy as the Lagrange method and used fewer computational resources. Han
et al. [38] used an Euler–Euler CFD method to compare it with asphalt scaling experiments
and found good agreement, and used this method to predict particle deposition in pipelines.
The Euler–Euler method is suitable for scenarios with large particle concentrations, and
the Euler–Lagrange method is mostly used for small particle concentrations. Among the
many Euler–Lagrange models, the coupled RSM and DPM approach has been considered
by more studies to have higher accuracy [39–41].
This study builds on existing research by using cosine trigonometric functions to
generate corrugated wall surfaces and to investigate the kinematic behavior of colloidal
particles using heat exchanger ventilation ducts with rectangular cross sections. Numerical
methods were used to study the turbulence near the corrugated wall surfaces, to analyze
the TKE distribution of secondary flows and vortex clusters near the corrugated wall
surfaces, and to investigate the effects of different wind speeds, particle sizes, corrugation
heights, and corrugation periods on the motion of the particles after numerical validation.

2. Numerical Methodology and Experimental Method

2.1. Gas-Phase Model
This study’s input air velocity (Umean ) ranged from 0.6 to 7.0 m/s, and its Re ranged
from 5.476 to 6.3893. The continuous phase in the pipeline was incompressible gas. The
Reynolds stress model (RSM) takes into account the anisotropy of turbulence and is there-
fore considered a more accurate turbulence model [42]. The control equation can be
expressed as:
∂ρui
=0 (1)
∂xi
!
∂ui ∂ui 1 ∂p 1 ∂ ∂ui ′ ′
+ uj =− + µ − ρui u j (2)
∂t ∂x j ρ ∂x j ρ ∂x j ∂x j
 
∂T ∂ ∂ λ ∂T
+ (u T ) = (3)
∂t ∂xi i ∂xi ρc p ∂xi
Following dimension analysis and collation, the Reynolds stress equation is written
as follows:
 
∂u ′ u′  
∂
  
′ u′ + u ∂ u′ u′ =
 ∂ v t i j ′ u′
∂u j ′ u′ ∂ui
∂t u i j k ∂xk i j
  u i k ∂x + u j k ∂x
∂xk σk ∂xk k k
| {z }
Pij = StressProduction
| {z }
   DT,ij = TurbulentDif
 fusion (4)
ε ′ ′ 2 2 2
−C1 ui u j − δij k − C2 Pij − δij P − δ ε
k 3 3 |3{zij}
| {z }
ϕij = Pressurestrain ε ij = Dissipation

where the empirical constants are as follows: σk = 1.0, C1 = 1.8 and C2 = 0.6.
Energies 2024, 17, 321 4 of 27

Here is the equation for calculating the turbulent dissipation rate:

" #
ε2

∂ε ∂ε ∂ vt ∂ε ε ∂u
+ uj = v+ − Cε1 ui′ u′j i − Cε2 (5)
∂t ∂x j ∂x j σε ∂x j k ∂x j k

where σε = 1.3, Cε1 = 1.44 and Cε2 = 1.92.

In this study, as in [43], to deal with the flow in the near-wall area, the two-layer
banded model and the enhanced wall function (EWF) were utilized, since they can provide
accuracy comparable to the normal two-layer method for a fine near-wall mesh [44]. In the
near-wall region affected by viscosity, Wolfstein’s one-equation model was adopted [45].
To ensure that the incoming air was fully developed, a user-defined function (UDF) was
used so that the velocity distribution at the inlet satisfied the one-seventh power law [3].
 1/7
y
U = U f ree for y ≤ H/2 (6)
H/2
1/7
H−y

U = U f ree for y > H/2 (7)
H/2
8
U f ree = Umean (8)
7
where Umean is the average flow velocity.

2.2. Particle-Phase Model

Spherical particles were modelled in this study using the Lagrangian methodology,
DPM. Due to a low particle concentration and an air density-to-particle density ratio of
0.0005, we excluded the impact of particles on fluid or particle interaction forces. Gravity
plays a crucial role in driving particle movement, and in this study, its effect on particle
behavior was noteworthy. Under its force, the particles would sink and ultimately settle on
the wall. Due to their small mass, buoyancy plays an important role in particle motion. A
velocity gradient between the fluid phase and the particle phase generates lift on the low
velocity side when fluid velocities differ on both sides perpendicular to the direction of
particle motion. This lift is known as Saffman lift. We took into account drag force (FD ),
gravity and buoyancy (FG ), Brown force (FB ), thermophoresis force (FT ), and Saffman lift
(FS ) when calculating the force of particles. A more specific part of the formula can be
found in [8]. The following is an explanation of the particle equation in the flow field:

du p
mp = FD + FG + FB + FS + FT (9)
dt

2.3. Particle Deposition Model

When particles move in a pipe, they will bounce off the wall if the kinetic energy is
high; otherwise, they will be attracted to the wall. In this study, a particle bounce model
was constructed using UDF coding. When the normal phase velocity of the particle is
lower than the critical velocity (ucr ), the particle will be captured by the wall. Conversely,
bouncing will occur. And the bouncing process uses fully elastic collision with a recovery
factor of 1.0 [46,47]. The critical velocity model used in this study was [9]:
10/7
ucr = 2K/(d p R2 )

(10)

The effective stiffness parameter K can be described as:

0.4
5π 2 (k s + k p )
K = 0.51[ ] (11)
4ρ1.5
p
Energies 2024, 17, 321 5 of 27

 
k s = 1 − νs2 /πEs (12)
 
k p = 1 − ν2p /πE p (13)
0.718
R = 45.3/(45.3 + Umean ) (14)
where ν stands for Poisson’s ratio; E stands for the Young’s modulus; the subscripts s and
p denote the wall and particle, respectively; νs = 0.28 and νp = 0.13; and Es = 215 GPa and
E p = 192 GPa [48]. The particle deposition velocity is a common metric for assessing the
particle deposition process, which would be defined as [49]:
 
HUmean N
Vd = − ln 1 − d (15)
L Nin

The second representation is [9]:

J N /t /A N /tmax
Vd = = d d = d (16)
C0 Nin /V Nin /H

where L is the length of the pipe, H is the pipe height, and Vd is the maximum deposition
time of particles. The maximum deposition time of particles is defined as:

Vd
Vd+ = (17)
u∗
u∗ was calculated in this study as follows:
q
u∗ =
p
τw /ρ g = Umean f /2 (18)

f was calculated by: "  1.11 #

1 6.9 kd
p = −3.6 log + (19)
f Re 3.7Dh

The flow Reynolds number (Re) was computed by:

Re = Dh Umean /ν (20)

Dh = 4 A /pd (21)
The dimensionless parameters used in this study were calculated as follows:

yu∗
y+ = (22)
v

CC Sd2p u∗ 2
τp+ = (23)
18v2
The particle deposition efficiency (η) can be defined as:

Nd
η= (24)
Nin

2.4. Discrete Random Walk (DRW) Model

According to past research [8,41,50], it is essential to precisely predict the deposition
velocity during turbulent particle dispersion. Using the DRW model, the turbulent velocity
fluctuation for the RSM model was calculated. The DRW model was used to foretell how
 Energies 2024, 17, 321 6 of 27

particles would migrate in channel flows due to diffusion. The time scale (τL ) is expressed
as follows:
( v ′ )2 k
τL = C2 rms = C3 (25)
2ε ε
The following formula was used to determine the eddy’s characteristic lifetime:

τe = −τL log(r ) (26)

According to the DRW model, a varying velocity has a Gaussian probability distribu-
tion. The velocity pulsation in this study was calculated as follow:

u′ = ζurms
′
(27)

v′ = ζvrms
′
(28)
w′ = ζwrms
′
(29)

2.5. Boundary Conditions

The inlet boundary condition during the numerical simulation process was the inlet
velocity, and the inlet gas heat was Tin = 300 K. The pressure outlet specifies the boundary
condition at the outlet. No-slip boundary conditions were adopted at the channel walls. The
temperature of the channel Twall = 350 K. The speed at the entrance (Umean ) was 0.6–7.0 m/s.
In this study, the kinematic viscosity of air was 1.48 × 10−5 m2 ·s−1 , and the air density was
1.225 kg/m3 . In order to ensure full development of the flow, the fully developed velocity
profiles of the one-seventh power law were adopted and imposed in the air flow inlet by
using UDF codes. Zhao et al. [51] investigated how changing particle counts affected the
outcomes of particle deposition simulations. The findings demonstrated that 1000 particles
were a sufficient number to detect statistical stability. In this study, 30,000 particles entered
the computational region using a cone after the turbulent duct flow was computed to
converge and with the same velocity as the air. Particles were incident vertically at the
entrance with cones. A total number of nine different sizes (1, 3, 5, 7, 9, 20, 30, 40, and
Energies 2024, 17, x FOR PEER REVIEW
50 µm) were simulated in this study. 7 of 29

3. Case Description and Solution

Since the actual ventilation ducts were mostly rectangular sections, this study adopted
thea experiment
rectangular section
is shownto study the 3D
in Figure ventilation
1. To reduce theducts. A schematic
influence diagram
of backflow, theoffront
the and
experiment
rear is shown
ends of the in Figure
corrugated 1. To reduce
section the influencethat
were lengthened, of backflow, ‫ܮ‬௙ =front
is, ‫ܮ‬௥ = the 150 and
mm.rearThe pa-
ends of the corrugated section were lengthened, that is, L r = L f = 150 mm. The
rameters of each part of the geometric model are shown in Table 1. According to Table 2, parameters
of each part of the geometric model are shown in Table 1. According to Table 2, a total of
a total of nine cases were simulated for the study.
nine cases were simulated for the study.

Figure 1. Deposition of particles on a three-dimensional corrugated wall.

Figure 1. Deposition of particles on a three-dimensional corrugated wall.

Table 1. Parameters of each part.

Part L H W ࡸࢌ ࡸ࢘
Length (mm) 900 100 200 150 150
Energies 2024, 17, 321 7 of 27

Table 1. Parameters of each part.

Part L H W Lf Lr
Length (mm) 900 100 200 150 150

Table 2. Computational cases.

Air Velocity Air and Particle Wall Corrugated Period of Ripples,

Case No. Surface Type
(m/s) Temperature (K) Temperature (K) Heigh, h (mm) T (mm)
1 3.0 300 350 / / Smooth
2 0.6 300 350 24 6.5 Corrugated
3 1.6 300 350 24 6.5 Corrugated
4 3.0 300 350 24 6.5 Corrugated
5 7.0 300 350 24 6.5 Corrugated
6 3.0 300 350 16 6.5 Corrugated
7 3.0 300 350 10 6.5 Corrugated
8 3.0 300 350 24 5.7 Corrugated
9 3.0 300 350 24 9.7 Corrugated

The contour of the corrugated wall was generated by the following expression:

1 2π
y= cos( x + θ) × h (30)
2 T

3.1. Solution Method

The fluid control problem was solved using the FVM. The discrete method was
employed using second-order upwind. SIMPLE was used for the calculations. A UDF
was used to satisfy the requirement of full air development and to implement the particle
deposition model.

3.2. Computational Grid and Grid Independence Study

Structured meshes were generated using the commercial software ICEM. These are
generally hexahedral grids, which require less computing memory, and the grids are
relatively neat and close to the actual model. Therefore, this paper used a structured grid
for mesh division. Additionally, mesh encryption was performed on the corrugated wall
surface. Figure 2 depicts a typical organized grid system utilized in the pipelines. The
first layer grid requires y+ ≈ 1 and a growth rate of 1.1. To reduce the influence of the
grid on the results, five sets of grids were used to verify the independence of the grid. The
specific verification situation is shown in Figure 3. From 190,000 to 1,200,000, five grid
numbers were employed in the calculation. The sensitivity of the grid was examined by
using the pipe pressure drop and the deposition efficiency of 3 µm particles, and a grid of
820,000 was finally determined as the grid to be used in this study.
 pipe
the pressure
results, drop
five sets and were
of grids the deposition efficiency
used to verify of 3 µm particles,
the independence of the grid.and
Theaspecific
grid of 8
was finally
verification determined
situation as the
is shown grid to3.be
in Figure used190,000
From in thistostudy.
1,200,000, five grid numbers
were employed in the calculation. The sensitivity of the grid was examined by using the
pipe pressure drop and the deposition efficiency of 3 µm particles, and a grid of 820,000
Energies 2024, 17, 321 8 of 27
was finally determined as the grid to be used in this study.

(a)

(a)

(b)

(b)

(c)
(c)
Figure
Figure 2. Computational
2. Computational grid.grid. (a) Corrugated
(a) Corrugated wallmesh.
wall pipe pipe (b)
mesh. (b) YZ(c)plane.
YZ plane. (c) XZ plane.
XZ plane.
Figure 2. Computational grid. (a) Corrugated wall pipe mesh. (b) YZ plane. (c) XZ plane.

Particle efficiency
Particle deposition deposition efficiency
0.800.80 Pressure drop
Pressure drop
Particle deposition efficiency (%)
Particle deposition efficiency (%)

0.780.78 2.0 2.0

Pressure drop (Pa)

0.760.76
Pressure drop (Pa)

1.5
1.5
0.74
0.74

0.72
0.72 1.0
1.0
0.70
0.70
0.5
0.68
0.5
0.68
0.66
200 400 600 800 1000
0.66
200 Mesh
400 number (thousand)
600 800 1000
Mesh number (thousand)

Figure 3. Computational grid sensitivity check.

 Figure 3. Computational grid sensitivity check.

Energies 2024, 17, x FOR PEER REVIEW 9 of 29

4. Results and Analysis
Energies 2024, 17, 321 4.1. Results Verification 9 of 27

4.1.1.
FigureVerification of grid
3. Computational thesensitivity
Turbulent Flow Field
check.

4. DNSand
4. Results
Results data
and were used to ensure the accuracy of this study, and the velo
Analysis
Analysis
4.1. Results
tion in theVerification
smooth pipe was compared with that in the study of Kim et al. [5
4.1. Results Verification
4.1.1. Verification of the Turbulent Flow Field
in Figure
4.1.1.DNS
4. In addition,
Verification
the realizable k-ε model, the SST-k-ω model, and t
data wereof thetoTurbulent
used ensure theFlow Field
accuracy of this study, and the velocity distribution
used, and the
in theDNS
smooth
velocity
pipe wasused
distributions
compared withthe
at the crestofand
thataccuracy
in the study
trough ofasthe bellows we
data were to ensure of thisKim et al.
study, and[52],
the shown indistribu-
velocity
with
Figure
tion inKuzan’s
4.the smooth [53]
In addition, theexperiments,
pipe realizable
was compared aswith
k-ε model, shown
thethat inthe
SST-k-ω
in Figure
model,
study 5.
and The
ofthe
Kim RSM RSM
et were model
used,
al. [52], was
as shown
closer
in theto
andFigure the
4. Inexperimental
velocity distributions
addition, results.
theatrealizable
the The
crest and
k-ε same
trough
model, oftheconclusion
the bellows were
SST-k-ω was
model, obtained
compared
and the with
RSMbywere
Ha
Kuzan’s [53] experiments, as shown in Figure 5. The RSM model was found to be closer to
their
used,study.
and theTherefore, the accuracy
velocity distributions at the of theand
crest results
troughobtained using
of the bellows
the experimental results. The same conclusion was obtained by Han et al. [42] in their study.
thecompared
were RSM m
with
ter Kuzan’s
guaranteed.
Therefore,
[53] experiments, as shown in Figure 5. The RSM model was found to be
the accuracy of the results obtained using the RSM model was better guaranteed.
closer to the experimental results. The same conclusion was obtained by Han et al. [42] in
their study. Therefore, the accuracy of the results obtained using the RSM model was bet-
ter guaranteed.

Figure 4. Validation of turbulent flow velocity profiles using mathematics [52].

Figure 4. Validation of turbulent flow velocity profiles using mathematics [52].
Figure 4. Validation of turbulent flow velocity profiles using mathematics [52].

Figure 5.5.Corrugated
Figure Corrugated pipe speed
pipe verification.
speed (a) Speed
verification. of crest.
(a) Speed (b) Speed
of crest. of trough
(b) Speed of [53].
trough [53].
4.1.2. Verification of Particle Deposition
4.1.2. Verification of Particle Deposition
Figure 6 shows the variation in the dimensionless deposition rate of colloidal parti-
Figure 5. Corrugated
Figure pipe
6 shows the speed in
variation verification. (a) Speed
the dimensionless of crest. rate
deposition (b) Speed of trough
of colloidal parti-
cles in smooth horizontal channels with dimensionless deposition time and compares
cles in smooth
it with horizontal
the results of Woodchannels with
[54] as well asdimensionless deposition time
other studies [3,39,42,55,56]. and compares it
Comparing
4.1.2.
with Verification
the
Equations results
(15) and ofitParticle
of(16),
Wood [54]found
was Deposition
as well asthe
that other studies
results [3,39,42,55,56].
of Equation (15) haveComparing Equations
less error with
Wood’s
(15) andempirical formula.
(16), it was Equation
found (16)results
that the exhibitsofa Equation
distinct “V” shape,
(15) havewhich
less is consistent
error with Wood’s
Figure 6 shows the variation in the dimensionless deposition rate of co
empirical formula. Equation (16) exhibits a distinct “V” shape, which is consistent with
cles in smooth
the results horizontal
of Tian and Ahmadi channels
et al. [3]. with
This isdimensionless deposition
because when the particle sizetime and
is small
with the results
the particles of Wood
are more [54] by
influenced as well as other
the flow studies
field, and [3,39,42,55,56].
the movement Compari
of large particles is
(15) and (16), it was found that the results of Equation (15) have less error
gies 2024, 17, x FOR PEER REVIEW
Energies 2024, 17, 321 10 of 27

with the results of Tian and Ahmadi et al. [3]. This is because when the particle size is small,
mainly
the particles influenced by by
are more influenced inertia
the flow forces. Therefore,
field, and the movement ofthe
largeparticle
particles is depo
first and then increases.
mainly influenced by inertia forces. Therefore, the particle deposition velocity decreases
first and then increases.

Figure 6. Particle deposition in horizontal smooth pipes [3,6,39,42,54–56].

Figure 6. Particle deposition in horizontal smooth pipes [3,6,39,42,54,55,
4.2. Effect of Air Velocity on the Flow Field
4.2.1. Effect of Air Velocity on Turbulence
4.2.InEffect of Air Velocity on the Flow Field
addition to Re, the impact of air velocity on turbulence can also be described in
terms
4.2.1.of turbulent
Effect of kinetic
Airenergy (TKE). TKE
Velocity on isTurbulence
a measure of turbulence strength, and the
larger the TKE, the larger the pulsation scale and the more affected is the particle motion
InAaddition
trajectory. cloud diagram toof Re,
TKEsthe impacttoof
corresponding air velocity
different air speeds ison
shownturbulence
in the
Figure 7. The larger areas of TKE are distributed at the crest and windward side of the
terms ofAsturbulent
corrugation. kinetictheenergy
the velocity increases, TKE also (TKE). TKE isand
increases gradually, a measure
along the flowof turb
larger the
direction, thelargest
TKE, theoflarger
clouds the pulsation
TKE gradually increase, which scale andthat
indicates the more
in the front affect
part of the flow field, the areas with high turbulent kinetic energy at the crest of the ripple
trajectory.
are small, while theA areas
cloud withdiagram
high turbulent ofkinetic
TKEs corresponding
energy in the back part areto different
larger and a
more particles will be obtained to be deposited in these areas.
Figure 7. The larger areas of TKE are distributed at the crest an
The use of corrugated walls will affect the temperature distribution in a pipe, causing
acorrugation. As the
change in the temperature velocity
gradient increases,
at the wall. As the speed the TKEthe
increases, also increases g
temperature
gradient from the low-temperature center region to the corrugated wall surface gradually
flow direction, the largest clouds of TKE gradually increase, wh
increases. The particles are subjected to thermophoretic force from the corrugated wall
fronttopart
surface of the
the upper wallflow
surfacefield,
because the
theareas
force is with high turbulent
in the opposite direction to thekinetic
temperature gradient. And the increase in velocity makes the secondary flow from the
ripple are small, while the areas with high turbulent kinetic ene
bottom of the pipe to the upper wall increase, which will make the particles be rolled and
larger
sucked and
away frommore particles
the corrugated will be
wall surface. Theobtained to be deposited
specific temperature distribution andin thes
secondary flow can be seen in Figure 8.
 trajectory. A cloud diagram of TKEs corresponding to different air speeds is shown in the
Figure 7. The larger areas of TKE are distributed at the crest and windward side of the
corrugation. As the velocity increases, the TKE also increases gradually, and along the
flow direction, the largest clouds of TKE gradually increase, which indicates that in the
front part of the flow field, the areas with high turbulent kinetic energy at the crest of the
Energies 2024, 17, 321 11 of 27
ripple are small, while the areas with high turbulent kinetic energy in the back part are
larger and more particles will be obtained to be deposited in these areas.

Energies 2024, 17, x FOR PEER REVIEW 11 of 29

(a)

Energies 2024, 17, x FOR PEER REVIEW 11 of 29

(c)

(b)

(d)
Figure 7. Cloud images of turbulent kinetic energy (TKE) at different velocities at X = 100 mm sec-
tion. (a) U mean = 0.6 m/s. (b) U mean (c)(c)
= 1.6 m/s. U mean = 3.0 m/s. (d) U mean = 7.0 m/s.

The use of corrugated walls will affect the temperature distribution in a pipe, causing
a change in the temperature gradient at the wall. As the speed increases, the temperature
gradient from the low-temperature center region to the corrugated wall surface gradually
increases. The particles are subjected to thermophoretic force from the corrugated wall
surface to the upper wall surface because the force is in the opposite direction to the tem-
perature gradient. And the increase in velocity (d) makes the secondary flow from the bottom
of the pipe
Figure 7. Figure to the
Cloud images upper wall
of turbulent increase, which will at
make the particles beat Xrolled
= 100 and
mm sucked
7. Cloud images ofkinetic energy
turbulent (TKE)
kinetic energy different
(TKE)velocities
at different velocities sec-
at X = 100 mm section.
away
tion. (a) U
from the
=
corrugated
0.6 m/s. (b) U wall surface.
= 1.6
The
m/s. (c)
specific
U =
temperature
3.0 m/s. (d) U distribution
= 7.0
and
m/s.
second-
(a) U
meanmean = 0.6 m/s. meanUmean = 1.6 m/s.mean
(b)
ary flow can be seen in Figure 8. (c) U mean = 3.0 m/s. (d)
mean U mean = 7.0 m/s.

The use of corrugated walls will affect the temperature distribution in a pipe, causing
a change in the temperature gradient at the wall. As the speed increases, the temperature
gradient from the low-temperature center region to the corrugated wall surface gradually
increases. The particles are subjected to thermophoretic force from the corrugated wall
surface to the upper wall surface because the force is in the opposite direction to the tem-
perature gradient. And the increase in velocity makes the secondary flow from the bottom
of the pipe to the upper wall increase, which will make the particles be rolled and sucked
away from the corrugated wall surface. The specific temperature distribution and second-
ary flow can be seen in Figure 8.

(a)

Figure 8. Cont.
 Energies
Energies 2024, 17, x17,
2024, FOR321PEER REVIEW 12 of 29 12 of 27

(b)

(c)

(d)

Figure 8. Temperature cloud image and flow field vector diagram of Y = 550 mm section at different
velocities. (a) Umean = 0.6 m/s. (b) Umean = 1.6 m/s. (c) Umean = 3.0 m/s. (d) Umean = 7.0 m/s.
Energies 2024, 17, x FOR PEER REVIEW 13 of 29

Figure 8. Temperature cloud image and flow field vector diagram of Y = 550 mm section at different
Energies 2024, 17, 321 13 of 27
velocities. (a) U mean = 0.6 m/s. (b) U mean = 1.6 m/s. (c) U mean = 3.0 m/s. (d) U mean = 7.0 m/s.

4.2.2. Vortex
4.2.2.Identification and Analysis
Vortex Identification and Analysis
Typically, vorticesvortices
Typically, are employed to quantify
are employed the rotating
to quantify motion of
the rotating a flow.
motion ofAnalysis
a flow. Analysis
of the impact of the flow field on particle deposition is facilitated by precise identification
of the impact of the flow field on particle deposition is facilitated by precise identification
of the position and sizeand
of the position of vortices. Setting aSetting
size of vortices. threshold value is necessary
a threshold for the second-
value is necessary for the second-
generation
generation vortex identification method, and different threshold valuesvarious
vortex identification method, and different threshold values have vor- vortex
have various
tex shapes. The influence
shapes. The influence of theofthreshold valuevalue
the threshold on vortex structure
on vortex can be
structure canresolved usingusing the
be resolved
the third-generation
third-generationvortex identification
vortex identificationmethod.
method.InInorder
orderto to catch vortexclusters,
catch vortex clusters,a athreshold
threshold value
value ofof 0.52is isused,
0.52 used,andandmore
moreflimsy
flimsyvortex
vortexstructures
structurescancanbe
beseen
seen[57].
[57]. The specific
specific formula
formula isis described as follows:
‖B‖2F ∥B∥2F
Ω= Ω= (31) (31)
‖A‖2F +‖B‖2F ∥A∥2F + ∥B∥2F
where, respectively, A and B represent the symmetric and antisymmetric components of
where, respectively, A and B represent the symmetric and antisymmetric components of
the velocity gradient tensor (ΔV), while ΔV is:
the velocity gradient tensor (∆V), while ∆V is:
U x U y U z 
 
 
Ux Uy Uz
ΔV =  Vx Vy Vz∆V  =  Vx Vy Vz  (32) (32)
W W W  Wx Wy Wz
 x y z 

A and BA and
are B are calculated
calculated as follows:
as follows:
1 1 1
( ) +B =( ∇V∇−V∇+V∇T V
)
 1
T

∇V = A + B = ∇V
∇V+∇
=VA + T
+ ∇ V − ∇ VT (33) (33)
2 2 2 2
The vortex Theclusters
vortex are depicted
clusters in Figurein9Figure
are depicted at various
9 at speeds.
various The ripple
speeds. Theperiod
ripplewasperiod was
6.5 mm, 6.5andmm,
the height
and theofheight
the ripple
of thewas 12 mm.
ripple was Periodically, vortices vortices
12 mm. Periodically, can be seen
can at
bethe
seen at the
peaks and valleys
peaks andof the waves.
valleys of theThe effectiveness
waves. of particle
The effectiveness of deposition on the corrugated
particle deposition on the corrugated
wall willwall will be impacted
be impacted by the obvious
by the obvious differencesdifferences in theclusters
in the vortex vortex onclusters on sides
the two the twoof sides of
the wall caused by the change in air velocity. As the speed increases, the
the wall caused by the change in air velocity. As the speed increases, the size of the vortex size of the vortex
cluster incluster in the
the tank tank gradually
gradually decreases,decreases, which
which will makewill
themake theofability
ability of thecluster
the vortex vortex cluster
to wind to upwind up the particles
the particles decrease, decrease, thus reducing
thus reducing the particle
the particle depositiondeposition
efficiency. efficiency.
The The
presencepresence of ripples
of ripples causes
causes the the vortex
vortex clustercluster to gather
to gather particles,
particles, whichwhich greatly
greatly increases the
increases
effectiveness
the effectiveness of particle
of particle deposition.
deposition.

(a)

(b)

Figure 9. Cont.
nergies 2024, 17, x FOR PEER REVIEW 14 of 29
rgies 2024, 17, x FOR PEER REVIEW 14 of 29
Energies 2024, 17, 321 14 of 27

(c)
(c)

(d)
(d)
Figure 9. Distribution of vortex clusters at different velocities. (a) U mean = 0.6 m/s. (b) U mean =
Figure
1.6 m/s. U mean 9.= of
(c) Figure
9. Distribution 3.0vortex U mean
Distribution
clusters
m/s. (d) ofat vortex clusters
=different
7.0 m/s. at (a)
velocities. U mean velocities.
different = 0.6 m/s. (b)(a)UUmean= =
mean
0.6 m/s.
(b) Umean = 1.6 m/s. (c) Umean = 3.0 m/s. (d) Umean = 7.0 m/s.
1.6 m/s. (c) U mean = 3.0 m/s. (d) U mean = 7.0 m/s.
The vortexThe cluster distribution
vortex in Figure 9d
cluster distribution differs 9d
in Figure significantly from the other
differs significantly from graphs
the other graphs
in that
Thethere
vortex is an upward
cluster winding
distribution in vortex
Figure cluster,
9d and
differs the upward
significantly
in that there is an upward winding vortex cluster, and the upward velocity velocity
from the gradually
other graphs in-gradually
creases with the
in that thereincreases increase
is an upward in the
winding
with the inlet
increase velocity,
vortexin the which
cluster,
inletand becomes
the upward
velocity, a resistance
which velocity to
becomesgraduallyparticle
a resistancedep-
in- to particle
ositionwith
creases and thusincrease
the
deposition will and
make
inthustheinlet
the particle
will deposition
velocity,
make the which efficiency
particle becomes
deposition adecrease.
resistance
efficiencytodecrease.
particle dep-
osition and thus will make the particle deposition efficiency decrease.
4.3. Effect of Different
4.3. Effect ofWalls on the
Different FlowonField
Walls the Flow Field
4.3. Effect
TheofTKE
Different Walls
distribution on inthe
theFlow
YZ Field
direction
The TKE distribution in the YZ direction for different rough walls
for different rough is walls
shown is in Figure
shown in Figure 10,
10, The
and TKE
the
andmaximum
the maximum
distribution value ofYZ
value
in the TKE increases
ofdirection
TKE forgradually
increases gradually
different as as
rough thethe
height
height
walls ofofthe
is shown therough
rough
in wall
wall increases.
Figure
increases.
10, and theThisThis also proves
also
maximum proves
valuethe of accuracy
TKE increases of thisgradually
study.
study. The
TheasTKE
the maximum
TKE maximum
height of theisisdistributed
rough wallatat the pipe
distributed
the pipe wall
wall in
in a smooth
smooth pipe,
pipe, and
and when
when a a corrugated
corrugated wall
wall
increases. This also proves the accuracy of this study. The TKE maximum is distributed surface
surface is is employed
employed in thein pipe,
the the TKE
at
the pipe wall in a smooth pipe, and when a corrugated wall surface is employed in thedisturbance
pipe, the maximum
TKE maximum emerges
emerges at the
at corrugated
the corrugated wall surface,
wall surface, whichwhichcan increase
can increase fluid
fluid
disturbance
pipe, the TKEandandparticle
maximum deposition.
particle deposition.
emerges atAsthe the
As corrugation
the corrugation
corrugated wallheight increases,
height
surface, the
canarea
increases,
which thewith
increaseareahigh
with
fluid TKE values
gradually
high TKE values
disturbance increases
gradually
and particle and the
increasesAs
deposition. perturbation
andthe thecorrugationof the whole
perturbationheight flow
of the increases,field
whole flowthe becomes
field
area more
with intense. In
becomes
more
high TKE addition,
intense.
values there isincreases
Ingradually
addition, a clear
there is boundary
aandclearthe layer at layer
boundary
perturbationthe smooth
ofatthe tube wall
thewhole
smooth flow surface,
tubefieldwall and the boundary
surface,
becomes
and the layer
boundary is destroyed
layer is at the
destroyed tube
at wall
the tube after using
wall after the
using
more intense. In addition, there is a clear boundary layer at the smooth tube wall surface, corrugated
the wall
corrugated surface,
wall sur-which will
face, which increase
will the heat
increase the transfer
heat efficiency.
transfer efficiency.
and the boundary layer is destroyed at the tube wall after using the corrugated wall sur-
face, which will increase the heat transfer efficiency.

(a)
(a)

(b)
(b)
Figure 10. Cont.
nergies 2024, 17, x FOR PEER REVIEW 15 of 29
Energies 2024, 17, 321 15 of 27

Energies 2024, 17, x FOR PEER REVIEW 15 of 28

(c)

(d)
(c)
Figure 10. Cloud
Figureplot
10. ofCloud
turbulent
plotkinetic energy kinetic
of turbulent (TKE) for different
energy rough
(TKE) for wall surfaces
different at Xwall
rough = 100surfaces at
mm section. (a) Smooth. (b) Corrugated wall height (h) = 5 mm. (c) Corrugated wall height
X = 100 mm section. (a) Smooth. (b) Corrugated wall height (h) = 5 mm. (c) Corrugated (h) =8 wall
mm. (d) Corrugated wall height (h) = 12 mm.
height (h) = 8 mm. (d) Corrugated wall height (h) = 12 mm.

Figure 11 Figure
shows the temperature
11 shows clouds and clouds
the temperature velocityand vector plots under
velocity vector walls
plots of dif- walls of
under
ferent roughnesses, where the temperature distribution is relatively
different roughnesses, where the temperature distribution is relatively uniform uniform in the in the
smooth channel. As the corrugation height increases,
(d) the temperature
smooth channel. As the corrugation height increases, the temperature center is shiftedcenter is shifted up-
ward, making
Figure 10.the
upward, distance
making
Cloud from
the thekinetic
distance
plot of turbulent roughfrom wall
energythe torough
(TKE) the
fortemperature
wall torough
different center
the temperatureof theatcenter
wall surfaces pipe
X = 100in-
of the pipe
crease, mm
which results
increase,
section. (a)whichin
Smooth.a lower
results temperature
in a lower
(b) Corrugated gradient
wall temperature and
height (h) = 5 mm. a lower
gradient thermal
and a wall
(c) Corrugated lower swimming
thermal
height (h) = 8 swimming
force. Vortex
mm. (d)clusters
force.Corrugated
Vortex exist
wallin all exist
height
clusters four
(h) = corners
12
in mm. of the
all four smooth
corners pipe,smooth
of the and in pipe,
the partandnear the part near
in the
bottom, antheupward
bottom, flowan field
upward exists, while
flow fielda downward
exists, while flow field will exist
a downward flowatfield
the bottom
will exist at the
Figure 11 shows the temperature clouds and velocity vector plots under walls of dif-
with the corrugated
bottom with wall
the surface.
corrugated Thewall
deposition
surface.ofThe small particlesofissmall
deposition influenced by is
particles theinfluenced
ferent roughnesses, where the temperature distribution is relatively uniform in the
flow field,by
sothe
theflow
corrugated
field, sowall surface willwall
the corrugated enhance
surfacethewill
deposition
enhanceefficiency of small
the deposition efficiency of
smooth channel. As the corrugation height increases, the temperature center is shifted up-
particles. As
ward, the
small ripple
making theheight
particles. As the
distance increases,
ripple
from the
height
the roughflow field
increases,
wall inthe
to the the bottom
flow fieldregion
temperature becomes
in theof
center bottom more
the piperegion
in- becomes
turbulent. Thewhich
more
crease, vortex
turbulent.cluster
results in area
The avortex
loweralsotemperature
gradually
cluster areaincreases, which
also gradually
gradient and willthermal
increase
increases,
a lower which thewill
depo-
swimming increase the
sition efficiency
force. of small
deposition
Vortex clustersparticles.
efficiency
existofin small
all fourparticles.
corners of the smooth pipe, and in the part near the

(a)

Figure 11. Cont.

Energies 2024, 17, x FOR PEER REVIEW 16 of 29

Energies 2024, 17, 321 16 of 27

(a)

(b)

Energies 2024, 17, x FOR PEER REVIEW 17 of 29

(c)

(d)
Figure 11. Temperature
Figure cloud image
11. Temperature cloud and flow field
image andvector
flow diagram of Y =diagram
field vector 550 mm section
of Y =at 550
X = 100
mm section at
mm section. (a) Smooth. (b) Corrugated wall height (h) = 5 mm. (c) Corrugated wall height (h) = 8
X = 100 mm section. (a) Smooth. (b) Corrugated wall height (h) = 5 mm. (c) Corrugated wall height
mm. (d) Corrugated wall height (h) = 12 mm.
(h) = 8 mm. (d) Corrugated wall height (h) = 12 mm.
4.4. Deposition Characteristics of Micron Particles
4.4.1. Effect of Rebound on Particle Deposition
Figure 12 shows a comparison of the critical particle deposition velocities for different
particle sizes with the results of Han et al. [42]. As particle size grows, the critical velocity
drops. This means that larger-sized particles are more prone to rebound. Figure 13 com-
pares the two different deposition models and shows that small particles are less affected
 4.4. Deposition Characteristics of Micron Particles
4.4.1. Effect of Rebound on Particle Deposition
Energies 2024, 17, 321
Figure 12 shows a comparison of the critical particle17depositio of 27

particle sizes with the results of Han et al. [42]. As particle size gro
4.4. Deposition
drops. This Characteristics
means that of Micron Particles
larger-sized particles are more prone to re
4.4.1. Effect of Rebound on Particle Deposition
pares the12two
Figure showsdifferent
a comparisondeposition models
of the critical particle andvelocities
deposition shows forthat small p
different
by bounce
particle sizes withwhile large
the results particles
of Han are
et al. [42]. As more
particle size affected by bounce,
grows, the critical velocity alo
drops. This means that larger-sized particles are more prone to rebound. Figure 13 compares
rate
the twodecreases for larger
different deposition modelsparticle
and showssizes. These
that small were
particles due
are less to the
affected by highe
larger particles
bounce while and are
large particles themore
smaller
affected bycritical
bounce,velocities.
along with the The same
deposition rateresult
decreases for larger particle sizes. These were due to the higher kinetic energies of the larger
et al. [44].
particles and the smaller critical velocities. The same results were obtained by Sun et al. [44].

gies 2024, 17, x FOR PEER REVIEW

Figure 12. Verification of the critical velocity of particle bounce [42].

Figure 12. Verification of the critical velocity of particle bounce [42].

Figure 13. Effect of rebound on particle dimensionless deposition rate.

Figure 13. Effect of rebound on particle dimensionless deposition rate.

4.4.2. Effect of Air Velocity on Particle Deposition

Figure 14 shows the η values at different air velocities and
 Figure 13. Effect of rebound on particle dimensionless deposition rate.
Energies 2024, 17, 321 18 of 27

4.4.2. Effect of Air Velocity on Particle Deposition

Figure
4.4.2. 14ofshows
Effect the ηonvalues
Air Velocity ParticleatDeposition
different air velocities and that η decreases with
the increase
Figurein 14air velocity.
shows At lower
the η values air velocities,
at different η gradually
air velocities and that η increases withthe
decreases with the in-
crease in ݀in
increase ௣ . air
When the ܷAt
velocity. is high,
lower
௠௘௔௡ η varies
air velocities, η gradually ݀௣ . Thiswith
less withincreases occurs because the
the increase
in d . When the U is high, varies less with d . This occurs because
ܷ௠௘௔௡ is high, the negative velocity in the z-axis direction also gradually increases,
p mean η p the Umean and
is high, the negative velocity in the z-axis direction also gradually increases, and the
the deposition of large particles mainly relies on gravitational sedimentation, so that the
deposition of large particles mainly relies on gravitational sedimentation, so that the η
η of
oflarge
largeparticles
particles decreases. As the ܷ increases,
decreases. As the Umean௠௘௔௡
increases, the wall resistance and pipe pres-
the wall resistance and pipe pressure
sure drop
drop gradually
gradually increase.
increase. Therefore,
Therefore, differentdifferent simultaneous
simultaneous smaller
smaller particle particleand
depositions deposi-
tions and smaller
smaller pressurepressure
drops weredrops were obtained.
obtained.

0.40 30
Pipeline resistance
0.35 Pressure drop
25
0.30

Pipeline resistance(N)

Pressure drop(Pa)
20
0.25

0.20 15

0.15
10
0.10
5
0.05

0.00 0
5476 14604 27382 63893
Reynolds number

(a) (b)
Figure 14.14.
Figure Particle deposition
Particle depositionefficiency and pipe
efficiency and pipepressure
pressure drop
drop and
and resistance
resistance at different
at different speeds.
speeds.
(a) Particle deposition
(a) Particle depositionefficiency.
efficiency. (b) Pipepressure
(b) Pipe pressuredrop
drop and
and resistance.
resistance.

Figure
Figure 1515 showsthe
shows theposition
positionofofparticle
particle deposition
deposition for fortwo
twodifferent
differentparticle
particlesizes
sizes (5
(5 µm and 50 µm) with different air velocities. When dp = 5 µm, the
µm and 50 µm) with different air velocities. When dp = 5 µm, the number of particles number of particles
deposited in the latter part of the corrugated wall was higher than that in the former part,
deposited in the latter part of the corrugated wall was higher than that in the former part,
and the deposition was concentrated in the crest and trough of the corrugation, with the
and the deposition was concentrated in the crest and trough of the corrugation, with the
deposition at the crest being arranged in a straight line. As the dp became larger, the
deposition
deposition at position
the crestchanged
being arranged
more, and inwhen
a straight line.
the air As the
velocity dpsmall,
was becamemore larger, the dep-
particles
osition
of 50 µm were deposited in the front part of the corrugation. Particles were gradually of 50
position changed more, and when the air velocity was small, more particles
µmdeposited
were deposited in the
at the back front
of the part of the
corrugated corrugation.
wall Particles
with an increase in airwere gradually
velocity. deposited
In addition,
at the
largeback of thewere
particles corrugated wall
deposited with
more at an
theincrease
entrance.inWhen
air velocity. In addition,
the air velocity was large
7 m/s,parti-
cles50were
µm particles
deposited were
moredeposited at the front
at the entrance. and rear
When ends
the air of the corrugation
velocity was 7 m/s, 50 and
µmless at
particles
the center part of the corrugation. This is thought to be because tiny particles,
were deposited at the front and rear ends of the corrugation and less at the center part of which have
theacorrugation.
lower mass, can follow fluid flow more effectively. The likelihood of small particles
This is thought to be because tiny particles, which have a lower mass, can
bouncing in the front section of the corrugated wall increases as the velocity and kinetic
follow fluid flow more effectively. The likelihood of small particles bouncing in the front
energy of the particles increases. After a few bounces, the kinetic energy of the particles
reduces on the corrugated wall surface in the back, increasing the number of particles
deposited. When the velocity is low, the kinetic energy of the particles is low and they are
easily deposited in the front portion of the corrugated wall. Inertial forces are the primary
factors that affect the deposition of big particles. As the velocity increases, the deposition
concentration area gradually moves to the rear part of the corrugated wall. Large particles
have a low critical velocity; hence, increasing velocity also improves the likelihood of large
particles rebounding. Therefore, under the impact of the two aforementioned parameters,
the efficiency of large particle deposition rapidly declined as the velocity reached 7 m/s.
 gated wall surface in the back, increasing the number of particles deposited. When the
velocity is low, the kinetic energy of the particles is low and they are easily deposited in
the front portion of the corrugated wall. Inertial forces are the primary factors that affect
the deposition of big particles. As the velocity increases, the deposition concentration area
gradually moves to the rear part of the corrugated wall. Large particles have a low critical
Energies 2024, 17, 321 velocity; hence, increasing velocity also improves the likelihood of large particles re- 19 of 27
bounding. Therefore, under the impact of the two aforementioned parameters, the effi-
ciency of large particle deposition rapidly declined as the velocity reached 7 m/s.

(a)

(b)

Energies 2024, 17, x FOR PEER REVIEW 20 of 29

(c)

(d)

15. Particle deposition locations for different air velocities. (a) U mean = 0.6 m/s. (b) U mean
Figure 15. Figure
Particle deposition locations for different air velocities. (a) Umean = 0.6 m/s.
= 1.6 m/s. (c) U mean = 3.0 m/s. (d) U mean = 7.0 m/s. (I for 5 µm, II for 50 µm).
(b) Umean = 1.6 m/s. (c) Umean = 3.0 m/s. (d) Umean = 7.0 m/s. (I for 5 µm, II for 50 µm).
4.4.3. Effect of Corrugation Height on Particle Deposition
Figure 16 illustrates the particle deposition efficiency and the pipe pressure drop and
pipe resistance at different rough wall heights. As the rough wall height increases, the
pipeline pressure drop and pipeline resistance also increase. When dp < 30 µm, η in-
creases with the increase in pipe resistance. When dp > 30 µm, η and pipe resistance
show negative correlation. This is because when the corrugation height increases, the up-
 (d)
Energies 2024, 17, 321 Figure 15. Particle deposition locations for different air velocities. (a) U mean = 0.6 m/s. (b) U mean20 of 27
= 1.6 m/s. (c) U mean = 3.0 m/s. (d) U mean = 7.0 m/s. (I for 5 µm, II for 50 µm).

4.4.3.
4.4.3.Effect
Effectof
ofCorrugation Heighton
Corrugation Height onParticle
Particle Deposition
Deposition
Figure
Figure16 16illustrates theparticle
illustrates the particledeposition
deposition efficiency
efficiency andand
the the
pipepipe pressure
pressure drop drop
and and
pipe
piperesistance
resistance atat different roughwall
different rough wallheights.
heights. AsAsthethe rough
rough wallwall height
height increases,
increases, the the
pipeline
pipelinepressure
pressuredrop
dropandandpipeline
pipelineresistance
resistance alsoalso increase.
increase. When
When dp dp << 30
30 µm,
µm, ηη increases
in-
creases with the increase in pipe resistance. When dp > 30 µm, η and pipe resistance
with the increase in pipe resistance. When dp > 30 µm, η and pipe resistance show negative
correlation.
show negativeThiscorrelation.
is because This
when is the corrugation
because when theheight increases,
corrugation theincreases,
height upward thevortex
up- mass
gradually
ward vortex mass gradually increases, which makes the large particles be subject to larger thus
increases, which makes the large particles be subject to larger resistance, and
the particle deposition
resistance, and thus theefficiency decreases.efficiency
particle deposition Compared to the effect
decreases. of airtovelocity
Compared the effectvariation
of
onair
particle
velocitydeposition,
variation onit is more costly
particle to useit increased
deposition, rough
is more costly wallincreased
to use height torough
achieve
wallproper
particle
height deposition.
to achieve proper particle deposition.

0.08
50 6
Smooth 0.07 Pipeline resistance
Pressure drop
Particle deposition efficiency (%)

h=10mm
5
40 h=16mm 0.06

Pipeline resistance(N)
h=24mm

Pressure drop (Pa)

0.05 4
30

0.04
3
20
0.03
2
10 0.02

1
0.01
0

0.00 0
1 10 0 10 16 24
dp (μm) h(mm)

(a) (b)

Figure 16. Particle deposition efficiency and pipe pressure drop and resistance at different corrugation
heights. (a) Particle deposition efficiency. (b) Pipe pressure drop and resistance.

Figure 17 illustrates the effect of different ripple heights on the location of particle
deposition, where I indicates a 5 µm particle and II indicates dp = 50 µm. The particles were
deposited less at the entrance of the corrugation, and the number of particles deposited
gradually increased along the flow square. The number of particles deposited at the
corrugation inlet steadily decreased as the corrugation height rose when dp = 50 µm. The
front portion of the corrugation developed a particle-free area even at a corrugation height
of 24 mm. This was mainly because, as the corrugation height increased, the particles with
large particle sizes had a greater chance of bouncing at the front part of the corrugation,
making the particles less likely to be deposited.
In the case of larger particles, particle deposition is mostly focused on the windward
side of the ripple and at the wave’s crest, with less deposition in the trough. This is because
small particles will be deposited at the trough due to the vortex at the trough, while the
main deposition mechanism of large particles is gravitational sedimentation, so it is difficult
for the secondary flow at the trough to absorb large particles, while the particle impact
at the crest will result in a reduction in part of the particles’ kinetic energy and thus their
deposition at the crest.

4.4.4. Effect of Corrugation Period on Particle Deposition

Figure 18 illustrates the particle deposition efficiency, pressure drop, and pipe resis-
tance for different corrugation cycles. When dp < 30 µm, the deposition efficiency gradually
increased as the corrugation period decreased. When dp > 30 µm, η was larger for the
corrugation cycle of T = 6.5 mm, and the corrugation wall of T = 5.7 mm had a higher
particle deposition efficiency than the corrugation wall of T = 9.75 mm at a particle size of
50 µm. As the cycle lengthened, the pipe’s resistance steadily reduced, which was caused
by a reduction in both the number of corrugations and the perimeter of the corrugated
wall surface. However, the pressure drop did not decrease with the increase in the corru-
gation period, and the pressure drop of the pipe was higher when the corrugation period
(T) = 9.75 mm than when T = 6.5 mm.
 side of the ripple and at the wave’s crest, with less deposition in the trough. This is because
small particles will be deposited at the trough due to the vortex at the trough, while the
main deposition mechanism of large particles is gravitational sedimentation, so it is diffi-
cult for the secondary flow at the trough to absorb large particles, while the particle impact
Energies 2024, 17, 321 at the crest will result in a reduction in part of the particles’ kinetic energy and thus21their
of 27
deposition at the crest.

(a)

Energies 2024, 17, x FOR PEER REVIEW 22 of 29

(b)

(c)
Figure 17.
Figure 17. Schematic
Schematicrepresentation
representationofofparticle
particledeposition
deposition locations with
locations different
with ripple
different heights.
ripple (a)
heights.
h = 10 mm. (b) h = 16 mm. (c) h = 24 mm. (I for 5µm, II for 50µm.)
(a) h = 10 mm. (b) h = 16 mm. (c) h = 24 mm. (I for 5 µm, II for 50 µm).

4.4.4. Effect of Corrugation Period on Particle Deposition

Figure 18 illustrates the particle deposition efficiency, pressure drop, and pipe re-
sistance for different corrugation cycles. When dp < 30 µm, the deposition efficiency grad-
ually increased as the corrugation period decreased. When dp > 30 µm, η was larger for
the corrugation cycle of T = 6.5 mm, and the corrugation wall of T = 5.7 mm had a higher
 ually increased as the corrugation period decreased. When dp > 30 µm, η was larger for
the corrugation cycle of T = 6.5 mm, and the corrugation wall of T = 5.7 mm had a higher
particle deposition efficiency than the corrugation wall of T = 9.75 mm at a particle size of
50 µm. As the cycle lengthened, the pipe’s resistance steadily reduced, which was caused
by a reduction in both the number of corrugations and the perimeter of the corrugated
Energies 2024, 17, 321 wall surface. However, the pressure drop did not decrease with the increase in the corru-
22 of 27
gation period, and the pressure drop of the pipe was higher when the corrugation period
(T) = 9.75 mm than when T = 6.5 mm.

35
T=5.7
T=6.5

Particle deposition efficiency (%)

30
T=9.75
25

20

15

10

1 10 10
dp (μm)

(a) (b)
Figure 18. Particle deposition efficiency and pipeline pressure drop and resistance at different cor-
Figure 18. Particlerugation
deposition efficiency and pipeline pressure drop and resistance at different
periods. (a) Particle deposition efficiency. (b) Pipe pressure drop and resistance.
corrugation periods. (a) Particle deposition efficiency. (b) Pipe pressure drop and resistance.
Figure 19 illustrates how the particles are deposited in each cycle of the corrugation
Figure 19 illustrates how the
when the period particles
is short and theare dp deposited in each
= 5 µm. Therefore, for cycle of the
corrugated corrugation
plates with tiny
cycles, η is higher when the particle size is
when the period is short and the dp = 5 µm. Therefore, for corrugated plates withsmaller. As dp increases, when dp = 50 µm,
tiny
the particles are deposited on the corrugated plate and the aggregation of deposition
cycles, η is higher when the particle size is smaller. As dp increases, when dp = 50 µm, the
along the flow square is more obvious and mainly at the crest and windward side. This
particles are deposited on the corrugated plate and the aggregation of deposition along
Energies 2024, 17, x FOR PEER REVIEW 23 of 29
the flow square is more obvious and mainly at the crest and windward side. This makes
the deposition efficiency of the corrugated plate with T = 9.75 mm higher than that of the
corrugatedmakes plate thewith T =efficiency
deposition 5.7 mmof at thedp = 50 µm.
corrugated Therefore,
plate with T = 9.75 mmthe larger
higher the corrugated wall
than that
of the corrugated plate with T = 5.7 mm at dp = 50 µm. Therefore, the larger
surface is for longer periods, the better the particle deposition at the corrugated plate. the corrugated
wall surface is for longer periods, the better the particle deposition at the corrugated plate.

(a)

(b)

Figure 19. Cont.

Energies 2024, 17, 321 Energies 2024, 17, x FOR PEER REVIEW 24 of 29 23 of 27

(c)
Figure 19. Schematic diagram of particle deposition locations for different ripple periods. (a) T =
Figure 19. 5.70Schematic
mm. (b) T = 6.50diagram of mm.
mm. (c) T = 9.75 particle deposition locations for different ripple periods.
(a) T = 5.70 mm. (b) T = 6.50 mm. (c) T = 9.75 mm.
5. Conclusions
Corrugated walls enhance heat transfer and represent a frequently employed struc-
5. Conclusions
ture in conduits designed for heat transfer. While more research has been undertaken on
Corrugated walls
heat transfer enhancewalls,
in corrugated heatstudies
transfer and deposition
on particle represent a frequently
in corrugated employed structure
channels
featuring trigonometric image shapes are relatively scarce. To enhance our comprehen-
in conduitssiondesigned for heat transfer. While more research has been
of particle deposition characteristics in a corrugated-wall ventilation channel with a
undertaken on heat
transfer in trigonometric
corrugatedimage walls, studies
shape, on particle
this investigation deposition
examined the impactsin of corrugated
varying corruga- channels featuring
tion heights, corrugation periods, particle sizes, deposition models, and wind velocities
trigonometric image shapes are relatively scarce. To enhance our comprehension of particle
on particle deposition by coupling the RSM and the DPM. By analyzing the turbulent flow
depositionfield,
characteristics in a corrugated-wall
secondary flow, particle deposition efficiency, and ventilation channel
particle deposition with
location, the a trigonometric
image shape,following
thisconclusions can be drawn:
investigation examined the impacts of varying corrugation heights,
1. The use of corrugated walls enhances the deposition efficiency of particles with par-
corrugation periods, particle
ticle sizes (dp) < 30 µm, sizes,
and when deposition models,
the particle size and
(dp) > 30 µm, wind
particles arevelocities
more on particle
deposition bylikely coupling
to bouncethe RSM
off the and walls,
corrugated the which
DPM. By the
makes analyzing the turbulent
particle deposition effi- flow field,
ciency lower than that for smooth walls. The particle deposition efficiency shows a
secondary flow, particle deposition efficiency, and particle deposition
positive correlation with particle size. When the corrugation height is 24 mm and dp
location, the following
conclusions can = 3 be
µm, drawn:
the particle deposition efficiency on a corrugated wall is five times higher
than that on a smooth wall.
1. The use of velocity
2. Air corrugated wallsfactor
is an important enhances
affecting this the deposition
study. The maximum efficiency
value of TKE of particles with
particle sizes
near the(dp) < 30wall
corrugated µm, and
surface when
occurs the particle
periodically at the crestsize (dp) > side.
and windward 30 µm, particles are
The value of TKE increases gradually with increase in inlet air velocity. Therefore,
more likely to bounce off the corrugated walls, which makes
more secondary flow occurs at the crest and windward side with increasing velocity,
the particle deposition
efficiencywhich
lower thanto that
can lead forinsmooth
changes walls. efficiency.
particle deposition The particle deposition
As the air velocity in- efficiency shows
creases, the rebound probability of large-size particles (dp > 10 µm) increases, so the
a positive correlation with particle size. When the corrugation height is 24 mm and
deposition efficiency of large-size particles decreases. Regarding particle deposition
dp = 3 µm, the particle
location, deposition
the air velocity has a strongefficiency
influence onon the a corrugated
deposition locationwall is five times higher
of large
than that on a smooth wall.
particles; as the air velocity increases, the dense area of particle deposition will grad-
ually move from the inlet to the outlet, and eventually only a small portion of the
2. Air velocity is an
particles will important
be deposited at factor affecting
the inlet and outlet duethis
to thestudy.
rebound.The maximum value of TKE
near the corrugated
3. The shape of the wall surface
corrugated occurs
wall surface is anperiodically
important factor at the crest
affecting and windward side.
this study.
With the increase in the corrugation height, the TKE value at the crest of the corru-
The value of TKE increases gradually with increase in inlet
gated wall will gradually increase, and the secondary flow will gradually move up-
air velocity. Therefore,
more secondary
ward. Whenflow dp < 30occurs at the crest
µm, the deposition and windward
characteristics side with
are mainly determined increasing velocity,
by the
which can flow vortices
lead and mass inertia,
to changes so the deposition
in particle particle deposition efficiency will
efficiency. As gradually
the air velocity increases,
increase with the increase in the corrugation height. The particle deposition efficiency
the rebound probability
gradually increases withofthe
large-size
decrease in particles (dp and
the ripple period, > 10particles
µm) areincreases,
deposited so the deposition
efficiency of large-size particles decreases. Regarding particle
in every ripple period. At dp > 30 µm, the particle deposition efficiency is deposition location,
inversely
correlated with the ripple height due to rebound. The particle deposition efficiency
the air velocity has a strong influence on the deposition location
does not gradually increase with the decrease in the corrugation period. In this study,
of large particles; as
the air velocity
the highestincreases, the dense
particle deposition efficiencyarea of particle
was observed deposition
for the corrugated plate will
with gradually move
from the inlet to the outlet, and eventually only a small portion of the particles will be
deposited at the inlet and outlet due to the rebound.
3. The shape of the corrugated wall surface is an important factor affecting this study.
With the increase in the corrugation height, the TKE value at the crest of the corrugated
wall will gradually increase, and the secondary flow will gradually move upward.
When dp < 30 µm, the deposition characteristics are mainly determined by the flow
vortices and mass inertia, so the particle deposition efficiency will gradually increase
with the increase in the corrugation height. The particle deposition efficiency gradually
increases with the decrease in the ripple period, and particles are deposited in every
ripple period. At dp > 30 µm, the particle deposition efficiency is inversely correlated
with the ripple height due to rebound. The particle deposition efficiency does not
gradually increase with the decrease in the corrugation period. In this study, the
highest particle deposition efficiency was observed for the corrugated plate with
T = 6.5 mm, and with the increase in the corrugation period, a particle-free region
appeared in the front section of the corrugated wall.
Energies 2024, 17, 321 24 of 27

Author Contributions: Conceptualization, H.L. (Hongchang Li) and H.L. (Hao Lu); methodology,
Y.W.; software, Y.W.; validation, Y.W., H.L. (Hao Lu) and H.L. (Hongchang Li); investigation, H.L.
(Hao Lu); resources, H.L. (Hongchang Li); data curation, Y.W. and W.Z.; writing—original draft prepa-
ration, Y.W.; writing—review and editing, H.L. (Hao Lu) and W.Z.; supervision, H.L. (Hongchang Li);
project administration, H.L. (Hongchang Li); funding acquisition, H.L. (Hao Lu) All authors have
read and agreed to the published version of the manuscript.
Funding: The authors appreciate the financial support provided by the National Oversea High-
level Talents Program of China, National Natural Science Foundation of China (no. 52266017),
and the Major Project of the National Social Science Foundation of China (no. 21&ZD133). The
study was also supported by the Xinjiang Natural Science Fund for Distinguished Young Scholars
(no. 2021D01E08), the Xinjiang Regional Coordination Special Project—International Science and
Technology Cooperation Program (no. 2022E01026), the Xinjiang Major Science and Technology
Special Project (nos. 2022401002-2, 2022A01007-1, 2022A01007-4), the Xinjiang Key Research and
Development Project (nos. 2022B03028-2, 2022B01033-2, 2022B01022-1), the Central Guidance on
Local Science and Technology Development Project (no. ZYYD2022C16), the Innovation Team Project
of Xinjiang University (no. 500122006021), and the High-level Talents Project of Xinjiang University
(no. 100521001).
Data Availability Statement: The data that support the findings of this study are available from the
corresponding author upon reasonable request.
Conflicts of Interest: The authors declare no conflicts of interest.

Nomenclature

A Pipe section area, m2

A Symmetric component of the velocity gradient tensor
B Antisymmetric component of the gradient tensor
CD Drag coefficient of particle
dp Particle diameter, µm
Dh Hydraulic diameter
Es Young’s modulus of channel wall, GPa
Ep Young’s modulus of particle, GPa
f Fanning friction factor
FD Drag force, N
FG Gravity and buoyancy, N
FB Brownian force, N
FS Saffman lift, N
FT Thermophoretic force, N
g Gravitational acceleration, m/s2
H Pipe height, m
h Corrugated wall height, mm
J Number of particles deposited per unit time and unit area
k Turbulent kinetic energy (TKE), m2 ·s−1
K Effective stiffness parameter
kd Average roughness height
Nd Deposited particle number
Nin Released particle number
p Time-averaged pressure, Pa
R Kinematic restitution coefficient
Re Reynolds number
S The particle-to-fluid density ratio
tmax The maximum deposition time of particles
Tin Inlet temperature, K
Twall Wall temperature, K
ucr Critical deposition velocity
u∗ Friction velocity
ui The component of the time-averaged velocity
Umean Mean flue gas velocity, m/s
Energies 2024, 17, 321 25 of 27

Vd Particle deposition velocity

Vd+ Dimensionless particle deposition velocity
y+ Dimensionless distance from the wall
Greek symbols
ε Dissipation rate of turbulent kinetic energy
τp+ Nondimensional particle relaxation time
v Kinematic viscosity
νs Poisson’s ratio of the wall
νp Poisson’s ratio of the particle
η Particle deposition efficiency
ρ Density, kg/m3

Abbreviation

RSM Reynolds stress model

DPM Discrete particle model
UDF User-defined function
HVAC Heating, ventilation and air conditioning
IAQ Indoor air quality
GFEM Galerkin finite-element method
CCD Central composite design
FVM Finite volume method
RANS Reynolds-averaged Navier–Stokes
DNS Direct numerical simulation
LES Large eddy simulation
LBM Lattice Boltzmann method
EWF Enhanced wall function
TKE Turbulent kinetic energy

References
1. Dockery, D.W.; Stone, P.H. Cardiovascular risks from fine particulate air pollution. N. Engl. J. Med. 2007, 356, 511–513. [CrossRef]
[PubMed]
2. Marval, J.; Tronville, P. Ultrafine particles: A review about their health effects, presence, generation, and measurement in indoor
environments. Build. Environ. 2022, 216, 108992. [CrossRef]
3. Tian, L.; Ahmadi, G. Particle deposition in turbulent duct flows—Comparisons of different model predictions. J. Aerosol Sci. 2007,
38, 377–397. [CrossRef]
4. Othmane, M.B.; Havet, M.; Gehin, E.; Solliec, C. Mechanisms of Particle Deposition in Ventilation Ducts for a Food Factory.
Aerosol Sci. Technol. 2010, 44, 775–784. [CrossRef]
5. Dehbi, A. A CFD model for particle dispersion in turbulent boundary layer flows. Nucl. Eng. Des. 2008, 238, 707–715. [CrossRef]
6. Lai, A.C.; Byrne, M.A.; Goddard, A.J. Aerosol deposition in turbulent channel flow on a regular array of three-dimensional
roughness elements. J. Aerosol Sci. 2001, 32, 121–137. [CrossRef]
7. Lai, A.C.; Byrne, M.A.; Goddard, A.J. Particle deposition in ventilation duct onto three-dimensional roughness elements. Build.
Environ. 2002, 37, 939–945. [CrossRef]
8. Lu, H.; Lu, L. A numerical study of particle deposition in ribbed duct flow with different rib shapes. Build. Environ. 2015, 94,
43–53. [CrossRef]
9. Lu, H.; Quan, Y. A CFD study of particle deposition in three-dimensional heat exchange channel based on an improved deposition
model. Int. J. Heat Mass Transf. 2021, 178, 121633. [CrossRef]
10. Bi, C.; Tang, G.H.; Tao, W.Q. Heat transfer enhancement in mini-channel heat sinks with dimples and cylindrical grooves. Appl.
Therm. Eng. 2013, 55, 121–132. [CrossRef]
11. Han, Z.; Lu, H. Numerical simulation of turbulent flow and particle deposition in heat transfer channels with concave dimples.
Appl. Therm. Eng. 2023, 230, 120672. [CrossRef]
12. Kooh Andaz, A.; Dal Maso, M. Effect of a deflector on deposition of particles with different diameters in a rib-roughened channel.
Powder Technol. 2023, 428, 118831. [CrossRef]
13. Launder, B.E.; Reece, G.J.; Rodi, W. Progress in the development of a Reynolds-stress turbulence closure. J. Fluid Mech. 1975, 68,
537–566. [CrossRef]
14. Hamida, M.B.B.; Almeshaal, M.A.; Hajlaoui, K.; Rothan, Y.A. A three-dimensional thermal management study for cooling a
square Light Edding Diode. Case Stud. Therm. Eng. 2021, 27, 101223. [CrossRef]
Energies 2024, 17, 321 26 of 27

15. Hamida, M.B.B.; Hatami, M. Investigation of heated fins geometries on the heat transfer of a channel filled by hybrid nanofluids
under the electric field. Case Stud. Therm. Eng. 2021, 28, 101450. [CrossRef]
16. Izadi, M.; Alshehri, H.M.; Hosseinzadeh, F.; Rad, M.S.; Hamida, M.B.B. Numerical study on forced convection heat transfer of
TiO2/water nanofluid flow inside a double-pipe heat exchanger with spindle-shaped turbulators. Eng. Anal. Bound. Elem. 2023,
150, 612–623. [CrossRef]
17. Azzouz, R.; Hamida, M.B.B. Natural Convection in a Circular Enclosure with Four Cylinders under Magnetic Field: Application
to Heat Exchanger. Processes 2023, 11, 2444. [CrossRef]
18. Massoudi, M.D.; Ben Hamida, M.B. Enhancement of MHD radiative CNT-50% water+ 50% ethylene glycol nanoliquid perfor-
mance in cooling an electronic heat sink featuring wavy fins. Waves Random Complex Media 2022, 1–26. [CrossRef]
19. Russ, G.; Beer, H. Heat transfer and flow field in a pipe with sinusoidal wavy surface—I. Numerical investigation. Int. J. Heat
Mass Transf. 1997, 40, 1061–1070. [CrossRef]
20. Heidary, H.; Kermani, M. Effect of nano-particles on forced convection in sinusoidal-wall channel. Int. Commun. Heat Mass Transf.
2010, 37, 1520–1527. [CrossRef]
21. Barba, A.; Rainieri, S.; Spiga, M. Heat Transfer Enhancement in a Corrugated Tube. Int. Commun. Heat Mass Transf. 2002, 29,
313–322. [CrossRef]
22. Andrade, F.; Moita, A.S.; Nikulin, A.; Moreira, A.L.N.; Santos, H. Experimental investigation on heat transfer and pressure drop
of internal flow in corrugated tubes. Int. J. Heat Mass Transf. 2019, 140, 940–955. [CrossRef]
23. Vicente, P.G.; Garcıa, A.; Viedma, A. Mixed convection heat transfer and isothermal pressure drop in corrugated tubes for laminar
and transition flow. Int. Commun. Heat Mass Transf. 2004, 31, 651–662. [CrossRef]
24. Kareem, Z.S.; Abdullah, S.; Lazim, T.M.; Jaafar, M.M.; Wahid, A.F.A. Heat transfer enhancement in three-start spirally corrugated
tube: Experimental and numerical study. Chem. Eng. Sci. 2015, 134, 746–757. [CrossRef]
25. Ağra, Ö.; Demir, H.; Atayılmaz, Ş.Ö.; Kantaş, F.; Dalkılıç, A.S. Numerical investigation of heat transfer and pressure drop in
enhanced tubes. Int. Commun. Heat Mass Transf. 2011, 38, 1384–1391. [CrossRef]
26. Cao, Q.; Liu, M.; Li, X.; Lin, C.H.; Wei, D.; Ji, S.; Zhang, T.T.; Chen, Q. Influencing factors in the simulation of airflow and particle
transportation in aircraft cabins by CFD. Build. Environ. 2022, 207, 108413. [CrossRef] [PubMed]
27. Boulbair, A.; Benabed, A.; Janssens, B.; Limam, K.; Bosschaerts, W. Numerical study of the human walking-induced fine particles
resuspension. Build. Environ. 2022, 216, 109050. [CrossRef]
28. Dehbi, A. Validation against DNS statistics of the normalized Langevin model for particle transport in turbulent channel flows.
Powder Technol. 2010, 200, 60–68. [CrossRef]
29. Ström, H.; Sasic, S.; Andersson, B. A novel multiphase DNS approach for handling solid particles in a rarefied gas. Int. J. Multiph.
Flow 2011, 37, 906–918. [CrossRef]
30. Agnihotri, V.; Ghorbaniasl, G.; Verbanck, S.; Lacor, C. On the multiple LES frozen field approach for the prediction of particle
deposition in the human upper respiratory tract. J. Aerosol Sci. 2014, 68, 58–72. [CrossRef]
31. Wu, P.; Feng, Z.; Cao, S.J. Fast and accurate prediction of airflow and drag force for duct ventilation using wall-modeled
large-eddy simulation. Build. Environ. 2018, 141, 226–235. [CrossRef]
32. Huang, W.; An, Y.; Pan, Y.; Li, J.; Chen, C. Predicting transient particle transport in periodic ventilation using Markov chain
model with pre-stored transition probabilities. Build. Environ. 2022, 211, 108730. [CrossRef]
33. Huang, W.; Chen, C. An improved Markov chain model with modified turbulence diffusion for predicting indoor particle
transport. Build. Environ. 2022, 209, 108682. [CrossRef]
34. Zeng, L.; Gao, J.; Lv, L.; Zhang, R.; Chen, Y.; Zhang, X.; Huang, Z.; Zhang, Z. Markov-chain-based inverse modeling to fast
localize hazardous gaseous pollutant sources in buildings with ventilation systems. Build. Environ. 2020, 169, 106584. [CrossRef]
35. Sajjadi, H.; Salmanzadeh, M.; Ahmadi, G.; Jafari, S. Simulations of indoor airflow and particle dispersion and deposition by the
lattice Boltzmann method using LES and RANS approaches. Build. Environ. 2016, 102, 1–12. [CrossRef]
36. Zheng, Z.; Yang, W.; Yu, P.; Cai, Y.; Zhou, H.; Boon, S.K.; Subbaiah, P. Simulating growth of ash deposit in boiler heat exchanger
tube based on CFD dynamic mesh technique. Fuel 2020, 259, 116083. [CrossRef]
37. Li, X.; Yan, Y.; Shang, Y.; Tu, J. An Eulerian–Eulerian model for particulate matter transport in indoor spaces. Build. Environ. 2015,
86, 191–202. [CrossRef]
38. Han, Z.; Xu, Z.; Yu, X. CFD modeling for prediction of particulate fouling of heat transfer surface in turbulent flow. Int. J. Heat
Mass Transf. 2019, 144, 118428.118421–118428.118429. [CrossRef]
39. Lu, H.; Lu, L. Numerical investigation on particle deposition enhancement in duct air flow by ribbed wall. Build. Environ. 2015,
85, 61–72. [CrossRef]
40. Lu, H.; Lu, L. CFD investigation on particle deposition in aligned and staggered ribbed duct air flows. Appl. Therm. Eng. 2016, 93,
697–706. [CrossRef]
41. Lu, H.; Zhao, W. Numerical study of particle deposition in turbulent duct flow with a forward- or backward-facing step. Fuel
2018, 234, 189–198. [CrossRef]
42. Han, Z.; Xu, Z.; Sun, A.; Yu, X. The deposition characteristics of micron particles in heat exchange pipelines. Appl. Therm. Eng.
2019, 158, 113732. [CrossRef]
43. Chen, H.; Patel, V. Near-wall turbulence models for complex flows including separation. AIAA J. 1988, 26, 641–648. [CrossRef]
Energies 2024, 17, 321 27 of 27

44. Sun, K.; Lu, L.; Jiang, H. Modelling of particle deposition and rebound behaviour on ventilation ducting wall using an improved
wall model. Indoor Built Environ. 2011, 20, 300–312. [CrossRef]
45. Wolfshtein, M. The velocity and temperature distribution in one-dimensional flow with turbulence augmentation and pressure
gradient. Int. J. Heat Mass Transf. 1969, 12, 301–318. [CrossRef]
46. Lu, H.; Ma, T.; Lu, L. Deposition characteristics of particles in inclined heat exchange channel with surface ribs. Int. J. Heat Mass
Transf. 2020, 161, 120289. [CrossRef]
47. Han, Z.; Xu, Z.; Yu, X.; Sun, A.; Li, Y. Numerical simulation of ash particles deposition in rectangular heat exchange channel. Int.
J. Heat Mass Transf. 2019, 136, 767–776. [CrossRef]
48. Wang, F.-L.; He, Y.-L.; Tong, Z.-X.; Tang, S.-Z. Real-time fouling characteristics of a typical heat exchanger used in the waste heat
recovery systems. Int. J. Heat Mass Transf. 2017, 104, 774–786. [CrossRef]
49. Zhang, J.; Li, A. Study on particle deposition in vertical square ventilation duct flows by different models. Energy Convers. Manag.
2008, 49, 1008–1018. [CrossRef]
50. Lo, C.; Bons, J.; Yao, Y.; Capecelatro, J. Assessment of stochastic models for predicting particle transport and deposition in
turbulent pipe flows. J. Aerosol Sci. 2022, 162, 105954. [CrossRef]
51. Zhao, B.; Chen, C.; Yang, X.; Lai, A.C.K. Comparison of Three Approaches to Model Particle Penetration Coefficient through a
Single Straight Crack in a Building Envelope. Aerosol Sci. Technol. 2010, 44, 405–416. [CrossRef]
52. Kim, J.; Moin, P.; Moser, R. Turbulence statistics in fully developed channel flow at low Reynolds number. J. Fluid Mech. 1987, 177,
133–166. [CrossRef]
53. Kuzan, J.D. Velocity Measurements for Turbulent Separated and Near Separated Flow over Solid Waves (Fluid-Mechanics); University of
Illinois at Urbana-Champaign: Champaign, IL, USA, 1986.
54. Wood, N. A simple method for the calculation of turbulent deposition to smooth and rough surfaces. J. Aerosol Sci. 1981, 12,
275–290. [CrossRef]
55. Zhang, Z.; Chen, Q. Prediction of particle deposition onto indoor surfaces by CFD with a modified Lagrangian method. Atmos.
Environ. 2009, 43, 319–328. [CrossRef]
56. Sippola, M.R. Particle Deposition in Ventilation Ducts; University of California: Berkeley, CA, USA, 2002.
57. Liu, C.; Wang, Y.; Yang, Y.; Duan, Z. New omega vortex identification method. Sci. China Phys. Mech. Astron. 2016, 59, 1–9.
[CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.