Journal of Thermal Analysis and Calorimetry (2020) 141:1569–1590

https://doi.org/10.1007/s10973-019-09140-5

Convective drying of a moist porous object under the effects

of a rotating cylinder in a channel
Fatih Selimefendigil1 · Seda Özcan Çoban1 · Hakan F. Öztop2

Received: 9 September 2019 / Accepted: 1 December 2019 / Published online: 11 December 2019
© Akadémiai Kiadó, Budapest, Hungary 2019

Abstract
Numerical study of evaporation in porous medium during convective drying process was examined, and heat and mass transfer
of liquid water and water vapor through porous media was investigated by using the Galerkin weighted residual finite element
method. The porous moist object has a rectangular shape and is assumed to represent a food sample. Two-dimensional laminar
flow of dry hot air was used in the channel with a rotating circular cylinder. Different locations and angular rotational speed
of the rotating cylinder were considered to control the convective heat transfer and mass transportation. The radius of rotating
cylinder and velocity of drying air were also varied with five different values. Results showed that rotational angular speed
and drying air velocity had significant effect on heat transfer and mass transport phenomenon for the porous moist object.

Keywords Convective drying · Heat and mass transfer · Porous matrix · CFD
List of symbols u⃗ Velocity vector
cp Specific isobaric heat capacity (kJ kg−1 K−1) V Volume ­(m3)
c Concentration x x coordinate (mm)
D Diffusivity ­(m2 s−1) y y coordinate (mm)
h Convective heat transfer coefficient (W m−2 K−1)
Greek letters
Hevap Heat of evaporation (J mol−1)
⃗I Identity vector ρ Density (kg m−3)
ε Porosity
k Thermal conductivity (W m−1 K−1)
κ Permeability
L, l Length (mm)
μ Dynamic viscosity
M Molar mass (kg mol−1)
τ Tortuosity
mevap Mass of evaporation (mol m−3 s−1)
n Mass flux (kg m−2 s−1) Subscripts
P, p Pressure (Pa) a Air
R Universal gas constant (J mol−1 K−1) c Capillary
Revap Evaporation rate (kg m−3 s−1) eff Effective
r Radius (mm) evap Evaporation
s Surface area ­(m2) g Gas
S Saturation l Liquid
T Temperature (K) ma Moist air
t Time (min) p Pressure
q Heat flux (kJ m−2 s−1) r Relative
s Solid
sat Saturation
* Fatih Selimefendigil v Vapor
fthsel@yahoo.com
w Water
1
Department of Mechanical Engineering, Celal Bayar
University, Manisa, Turkey
2
Department of Mechanical Engineering, Technology Faculty,
Fırat University, Elazig, Turkey

13
Vol.:(0123456789)
1570 F. Selimefendigil et al.

Introduction the object is close to the equilibrium temperature. Stan-

ish et al. [33] developed a mathematical model of wood
Heat and moisture transport from porous materials such as drying and compared the results with experimental study
foods, bricks and biological membranes is a very common of wood specimens and non-hygroscopic porous ceramic
issue to investigate in the literature. Particularly, there is solid drying. The equations of mass and energy transport
a wide range of foods that the transport mechanism had were one dimensional, and the moisture content assumed
been investigated for the drying [1–9], heating [10–14], to be with three phases: liquid water, bound water and
frying [15–19] and other food heat and moisture transport water vapor. The mathematical results of heat and mass
processes [20–26] and currently, it is a topic of interest. transport were in very good agreement with experimen-
Drying processes are one of the main applications for tal results. Itaya et al. [34] developed a three-dimensional
long-life storage of fresh foods because of large water model of drying to analyze heat and moisture transfer and
removal from products. Convective drying is one of the viscoelastic hygrostress formation in a composite body.
most preferred applications in food drying industries They solved the governing equations by a finite element
because it is the simplest and most economical technique method and validated the solutions with experimental dry-
among several drying processes [27]. The food sample ing processes of triply layered brick-shaped samples made
is considered to be a porous moist object and hot dry air by the hydrates of starch granules and of 3:1 mixture of
flow through the object is a good way of moisture trans- starch granules and sucrose. Both numerical and experi-
portation from the food to air. Mathematical modeling is mental results showed that layer arrangement in forced-air
an efficient way for analyzing the convective drying of drying process had strong effect to moisture transfer and
moist objects with acceptable costs and time of comput- stress formation, while it had weak effect on heat transfer.
ing. Fourier, Fick and Darcy had led the governing equa- Cevoli et al. [35] numerically studied heat and mass trans-
tions of mass, momentum and energy to calculate heat fer phenomena in drying process of edible coating of on a
and mass (water) transfer in porous moist objects [28]. bun bread. In the study, they focused on three main factors
Nasrallah et al. [29] made a detailed study of heat and that affect drying time which are temperature, thickness of
mass transfer phenomenon in forced convection drying of the coating and solvent concentration. 1D and 2D models
two different porous mediums (wood and brick) numeri- were developed in FEM-based computer program COM-
cally. It was observed that the temperature of the surface SOL Multiphysics in order to calculate the drying time and
increased with time, while saturation decreased for both evaluate surface temperature distribution and mass flux,
samples. Kumar et al. [30] made a comprehensive study and the results were validated with experimental studies.
for multiphase convective drying for cylindrical apple Effects of surface temperature distribution on mass flux
slices with FEM method via COMSOL Multiphysics were weak when drying temperature was until 353 K, and
CFD package. The saturation of the fruit was getting lower there is good agreement in results which were validated
from center to the surface of slice, and the convective and with the experimental studies at drying temperatures of
diffusive fluxes were higher on and near the surface of 298 K and 333 K. Onwude et al. [36] developed a model
the sample. Curcio et al. [31] developed a 3D theoreti- for sweet potato drying with combined infrared and hot
cal model of food drying process describing the trans- air drying process numerically. The model geometry was
port phenomena. The problem was about heat and mass 2D axisymmetric section of a cylindrical slice of sweet
transfer and shrinkage of a food sample in simultaneous potato, and simultaneous heat and mass transfer during
turbulent flow conditions in a convective hot air dryer. drying was simulated considering both temperature- and
They used FEM method coupled to arbitrary Lagrange shrinkage-dependent diffusivity. Reduction in drying time
equations for solving the equations and compared with is higher when both of the drying processes were used
their experimental work of convective potato slices drying. and temperature inside the sample was higher than that of
The numerical results were coherent with the experimental the exterior because of the internal heating capability of
slab thickness and shrinkage evolution of potato slices. infrared radiation drying.
Kim et al. [32] developed convective drying of a 2D k–ε In convective heat transfer applications, many active and
turbulent model of a moving moist object numerically. passive heat transfer techniques are used to control the con-
They used a Fluent CFD code to solve the conservation vective heat transfer and fluid flow features [38–39]. In one
equations for convective drying. The results showed that of these methods, obstacles are widely used in cavities or in
the comparison between heat and mass transfer for the channel to control the heat transfer features [40–44]. Station-
convective drying is valid only when the temperature of ary or rotating cylinders are used in many convection studies

13
Convective drying of a moist porous object under the effects of a rotating cylinder in a channel 1571

[46–51]. Size, location and angular rotational speed of the 𝜕pa

cylinder were found to be effective in heat transfer enhance- + ∇ ⋅ 𝜌a u⃗ = 0 (1)
𝜕t
ment [53–54]. In the current study, convective drying of a
moist porous object in a channel under the impact of a rotat- 𝜕u
ing cylinder was examined numerically. The result of the 𝜌a + 𝜌a u⃗ ⋅ ∇⃗u = −∇⋅
present study will be helpful in design and optimization of [ 𝜕t ( )(( )T ) )]
p⃗I + 𝜇a − (2∕3) ∇ ⋅ u⃗ ⃗I
) ( (
∇ ⋅ u⃗ + ∇ ⋅ u⃗
thermal systems related to drying of moist porous products (2)
which have a wide range of application area.
𝜕T
𝜌a cp,a − ∇ ⋅ (ka ∇T) + 𝜌a cp,a u⃗ ⋅ ∇T = 0 (3)
𝜕t
Problem definition

A schematic description of the problem is shown in Fig. 1.

Porous domain equations
Two-dimensional section of food sample is located on the
center of the channel bottom. Laminar two-dimensional flow
The principles of multiphase flow through porous media are
of hot dry air at temperature 333 K through porous domain
valid for the problem except the channel. Porous medium
including liquid water and water vapor is described. Mul-
consists of three phases which are the solid part, pores filled
tiphase transport of water happens by means of evapora-
with gas and pores filled with water. This means the total
tion so saturation of the porous medium changes due to the
volume of the sample is the sum of these three parts. Poros-
transport of water vapor from sample to air.
ity expresses the fraction of void space to the total volume
of the medium as written in Eq. 4:
Governing equations
ΔVg+ ΔVw
Fluid domain equations in channel 𝜀= Δ (4)
ΔV

Flowing air in the channel is a content of hot dry air initially. where ΔV is the total volume of porous medium and
After a period of drying time, because of moisture trans- described as follows:
portation from porous object to air, it becomes a composi-
tion of dry air and water vapor. Conservation of mass and
ΔV = ΔVg + ΔVw + ΔVs (5)
momentum equations for air–water vapor mixture in channel The subscripts are g: gas, w: water or l: liquid and s: solid.
is given as follows [55, 56]: Saturation is relevant to the pore volume filled with water

Fig. 1  Mathematical model of y

problem (a) and mesh distribu- (a)
W
tion (b)

ω
D
U∞, Th
H
Porous
moist object
xc yC

x
l

(b)

13
1572 F. Selimefendigil et al.

and gas, and it can be defined as the fraction of pore volume where 𝜏 is the tortuosity and D is the diffusivity of gas in
of water (or gas) to the total pore volume [30] (Eqs. 6, 7). the bulk and 𝜀 is the porosity. Mixture of water and air is
treated as ideal gas.
ΔVw ΔVw
Sw = = (6) Water in gas phase flows due to pressure gradient, and the
ΔVg+ ΔVw 𝜀ΔV average velocity of moist air through the porous medium is
described as:
ΔVg ΔVg
Sg = = (7) kg 𝜕Pg
ΔVg+ ΔVw 𝜀ΔV ug = − (15)
𝜇g 𝜕s
The relation between the concentration of water (cw)
Permeability is an important effect, and it is related to
and liquid water saturation Sl can be written as:
tortuosity which is a property of solid matrix with pores
cw Mw (Eq. 16). κ is the intrinsic or absolute permeability. κr is the
Sl = (8) relative permeability of gas phase, and in multiphase flows,
𝜌w 𝜀
it depends on the saturation of each phase.
where Mw is the molar mass of water (mol kg−1) and ρw is
the density of water (kg m−3). 1 ∑
𝜅= Δ𝛽i ri (16)
Mass balance equations of liquid water, water vapor 8𝜏 i
and air are described by a non-isothermal formulation of
where Δ𝛽i is the volume fraction of pores in the ith class
strongly distributed evaporation from porous medium [57]:
having radius ri [57]. It can be obviously understood that
𝜕cg mass flux due to pressure is relevant to both fluid and porous
(9)
( )
+ ∇ ⋅ n���⃗g = Revap matrix properties because of the density and viscosity (fluid
𝜕t
properties) and permeability (porous matrix property).
𝜕cl The capillary effects also arise in liquid phase, and a
(10) pressure gradient causes liquid flow. Capillary pressure is
( )
+ ∇ ⋅ n��⃗l = −Revap
𝜕t
described as follows:
𝜕ca pc = pg − pl (17)
(11)
( )
+ ∇ ⋅ n��⃗a = 0
𝜕t
Mass flux of liquid phase is described as follows:
In these equations, Revap is the evaporation rate, and n⃗ is
kl 𝜕P k 𝜕p
the total flux due to pressure gradients described by Dar- p
nl = −𝜌l + 𝜌l l c (18)
cy’s law (Eq. 12) and molecular diffusion of gas species 𝜇l 𝜕s 𝜇l 𝜕s
(vapor) in pores which is described by Fick’s law (Eq. 13): Capillary diffusivity in liquid phase is described as:
kg 𝜕Pg kl 𝜕pc
npg = −𝜌g (12) Dc = −𝜌l (19)
𝜇g 𝜕s 𝜇l 𝜕cl
where g denotes the gas phase, ng is the mass flux, P is the
p
Equation 18 takes the form as follows:
total pressure of gas, 𝜌g is the density, and 𝜇g is the dynamic
viscosity of gas. The permeability of gas phase κg is calcu- p kl 𝜕P 𝜕c
nl = −𝜌l + Dc l (20)
lated as κg = κκr, where k is the intrinsic permeability and κgr 𝜇l 𝜕s 𝜕s
is the relative permeability of the gas phase.
where l denotes the liquid phase. The first term of the right
𝜕cg side of equation is the mass flux due to gas pressure, where
ndg = −Dg (13) P is the total gas pressure, and the second term is the mass
𝜕s
flux due to negative pressure pc caused by the capillary
where Dg is the molecular diffusivity and described as and attractive forces and c­ l is the liquid concentration. κ is
follows: the intrinsic permeability in liquid phase and described as
D κl= κκrl.
Dg = 𝜀 (14) The liquid-phase velocity is small compared to the moist
𝜏
air velocity and described with the equation driven by Dar-
cy’s law:

13
Convective drying of a moist porous object under the effects of a rotating cylinder in a channel 1573

kl 𝜕Pg q̇ = Hevap ⋅ mevap (28)

ul = − (21)
𝜇l 𝜕s
where Hevap is the latent heat of evaporation with the unit
Since the problem is defined as a mechanistic formulation (J mol−1).
for strong evaporation of multiphase domains, mass fluxes are
described for air and vapor as they are composed of convective Effective thermal and physical properties
and diffusive fluxes (Eqs. 22–24) [57]:
Thermal conductivity Effective thermal conductivity is
kg 𝜕Pg c2g (p )
total of the density of water and gas phases multiple with
nv = −𝜌g − Ma Mv Deff ∇ v (22)
𝜇g 𝜕s 𝜌g P the saturation of that particular phase given as follows [58]:

(29)
( )
keff = ks (1 − 𝜀) + kw 𝜀Sl S + kg 𝜀 1 − Sl
kg 𝜕Pg c2g (
P − pv
)
nv = −𝜌g − Ma Mv Deff ∇ (23)
𝜇g 𝜕s 𝜌g P
Density Effective density is similar to effective conduc-
For liquid phase, following equation is written as: tivity referenced from the study of Kumar et al. as given
in Eq. 30 [30]:
kw 𝜕Pw 𝜕S
nv = −𝜌w − Dw 𝜌w 𝜀 w (24) (30)
( )
𝜇w 𝜕s 𝜕s 𝜌eff = 𝜀 Sg 𝜌g + Sw 𝜌w + (1 − 𝜀)𝜌s

When the concentration of water vapor in porous medium is

lower than concentration of saturation (cv,sat), water evaporates Specific heat Specific heat formulation is given as follows:
from porous medium to air. Mass of evaporation is the way
Sg 𝜌ma cp,ma + Sl 𝜌w cp,w
of calculate evaporation rate (Revap) which gives the amount cp,eff = (31)
of water that evaporates to air and the proportion of reduced 𝜌eff
liquid and increased moist air. The saturation concentration
Effective diffusivity can be described as a combination
is determined by the saturation pressure pv,sat (Pa), and it is
of water and vapor transport. It is an empirical equation
described as follows:
because it cannot be directly related only to capillary pres-
pvsat (T) sure or vapor diffusivity and also cannot be derived from
cv,sat = (25) governing equations, so the experimental data helped the
RT
equation to take the form for this model [59]:
Mass of evaporation (mol m−3 s−1) is described as follows:
Deff = Dva 𝜀3∕4 Sg10∕3 (32)
(26)
( )
mevap = Kevap aw csat − c
In the equation, ε is porosity, Sg is gas saturation, Dva
where aw is the water activity which is relevant to the cap- is the vapor–air diffusivity, and Dva = 2.6 × 10−5 m2 s−1.
illary pressure and described as: pv/pv,sat(T), where pv and Velocity field for the water vapor transport equation
pv,sat are the vapor pressure and saturated vapor pressure, takes the form as follows:
respectively. Kevap is the evaporation constant, and c is the
current concentration. ug Ma Deff
Energy balance equation assuming each of the phases is in
ul = − ∇𝜌
Sg Mma 𝜌ma ma (33)
thermal equilibrium is given as follows:
( ) 𝜕T
Average velocity Average velocity of dry air, water vapor
( )
𝜌cp eff + ∇ ⋅ n⃗ g hg + n⃗ w hw = ∇ ⋅ keff ∇T − hfg Revap + q̇
𝜕t and liquid water is given as follows:
(27)
In this equation, the second term on the left side is the na cp, a + nv cp, v + nw cp, w
energy transport due to convection and on the right side, con-
uave = −
𝜌eff cp, eff (34)
duction and evaporation terms are written, respectively. The
quantities hg and hw are the enthalpies of gas and liquid water, where n is the mass flux, cp is the specific heat of each phase
respectively, and described as hw(g)= cpw(g)∆T, where cp is the with relevant subscripts, and ρ is the density.
specific heat. The last term q is the heat source which is the
heat of evaporation in this problem described as follows:

13
1574 F. Selimefendigil et al.

Boundary and initial conditions Table 1  Thermal and physical properties of phases [30, 60]
Property Phase Value
Fluid domain
Thermophysical properties
At the channel inlet, velocity and temperature are uniform Thermal conductivity, k Air 0.025/W mK−1
and constant. Fully developed flow conditions are used at Liquid water 0.59/W mK−1
the outlet, while no-slip conditions on the top and bottom Water vapor 0.026/W mK−1
walls of the channel are valid. For the channel, other initial Specific heat, cp Air 1006/J kg−1 K−1
conditions are given as follows: Liquid water 4182/J kg−1 K−1
Water vapor 2062/J kg−1 K−1
cv (0) = 0, P(0) = Patm , T(0) = 293 K Density, ρ Air 1.205/kg m−3
Liquid water 998.2/kg m−3
Inflow temperature of drying air is T = 333 K
Porous matrix 1528/kg m−3
The cylinder mounted over the porous medium rotates,
Dynamic viscosity, μ Air 1.81.10−5/kg ms−1
and the angular velocity equations are given as follows:
Liquid water 1.002.10−3/kg ms−1
(35) Water vapor 1.8.10−5/kg ms−1
( )
ux = − 𝜔 y − yc
Molar mass Air 0.028/kg mol−1
Liquid water 0.018/kg mol−1
(36)
( )
uy = 𝜔 x − xc Water vapor 0.018/kg mol−1
Porosity, ε Porous matrix 0.8
where ux and uy are the angular velocities of x- and y-axes,
Permeability, κ Porous matrix 1.10−14
xc, yc are the location of the cylinder in the coordinate sys-
Constants
tem, and ω (rad s−1) is the angular rotational speed.
Evaporation constant, K 1000
Universal gas constraint, Rg 8.314/J mol−1 K−1
Porous matrix
Latent heat of evaporation, hfg 2.26e6/J−1 kg−1
Ambient pressure, Pamb 101,325 pa
Initial conditions of porous matrix are given as follows:
Initial concentration of vapor saturation, cv,sat(0) 0.95914/mol m−3
T(0) = 293 K, Sl (0) = 0.5

For the liquid phase of porous medium, initial concentra-

describing κr are defined such that they are always positive
tion of liquid water is described as:
[55]:
S𝐥 (0)𝝆𝐥 𝜺 {
c𝐥 = 1 − 1.1Sl , Sl < 1∕1.1
M𝐥 𝜅r,g =
0, Sl ≥ 1∕1.1 (38)

where Sl is the saturation, Ml is the molar weight, 𝝆𝐥 is the

density of liquid water, and ε is the porosity. {(
Sl −Sli
)3
For the gas phase of water, initial vapor concentration is , Sl > Sli
𝜅r,g = 1−Sli (39)
considered under saturation (cv,sat) and it is calculated with 0, Sl > Sli
the empirical pressure-dependent equation as follows:
The variable Sli is the irreducible liquid phase saturation,
7.5 (T−273.15) describing the saturation of the liquid phase that will remain
610.7 × 10 (T−35.15)
(37)
c𝐯 = inside the porous medium.
R
where T is the temperature with K, and R is the universal
gas constant.
Thermal and physical properties of air, liquid water, vapor Solution methodology
and solid matrix in the problem are given in Table 1.
A commercial computational fluid dynamics code COM-
Permeability SOL was used for the numerical simulations. The govern-
ing equations were solved with finite element method along
Relative permeability of each phase is related to saturation the boundary conditions, and Galerkin weighted residual
and properties of the phases. In this problem, the functions method was used to obtain weak form of the equations. Flow
variables within the domain were approximated by Lagrange

13
Convective drying of a moist porous object under the effects of a rotating cylinder in a channel 1575

(a) 25
cold
hot hot

20

15 Da = 10–3
Da = 10–4
Da = 10–5
10

(b)

101

Da = 10–3
Nux

103 Da = 10–2

10–1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
x/H Fig. 3  Comparison of moisture distribution for convective drying of
potatoes (V = 1 m s−1, AR = 2 and T = 353 K) at t = 100 min
Fig. 2  Local Nusselt number variation along the top wall for a
porous entrapped cavity for fixed values of Ra = 5 × 105, Da = 10−4,
Pr = 0.015. a Basak et al. [61], b current solver Table 3  Comparison of numerical and experimental dimension-
less moisture ratio for convective drying of potato for various time
instances (V = 1 m/s, AR = 2 and T = 353 K)

Table 2  Comparison of average Nusselt number with an inner rotat- Time (min) Dimensionless moisture ratio
ing cylinder for a differentially heated cavity
Current CFD work Experimental
Present Roslan et al. Difference (%) study in Ref.
[54] [63]

Ω = 100, R = 0.2 4.81 4.78 − 0.63 50 0.74 0.71

Ω = 500, R = 0.1 4.74 4.75 − 0.21 100 0.506 0.48
200 0.212 0.20
300 0.0993 0.095

polynomials of different orders. The computational domain

is discretized into triangular elements. Weak form of the
governing equations is established, and residuals are set to comparison is made for the convection in a differentially
be zero in a weighted average sense. The weight function heated cavity with an inner rotating cylinder (Roslan et al.
was chosen from the same set of functions as that of trial [54]). Table 2 shows the average Nusselt numbers consider-
function. The convergence of the solution is assumed when ing two different values of angular rotational speed and size
the relative error for each of the variables satisfies the con- of the cylinder.
vergence criteria of ­10−6. Another validation is performed by using the available
Mesh independence of the solution is assured by utilizing numerical and experimental data for the convective dry-
various grid sizes. The grid is refined especially near the ing of potato as in Refs. [62, 63]. In the numerical study
walls and at the interface to resolve high gradients in those of [62], heat and mass transfer coefficients along with the
locations. Mesh distribution of the computational domain is thermophysical properties are taken from experimental work
shown in Fig. 1b. Code validation is performed by utilizing of [63]. Coupled heat and mass transfer problem with con-
different available studies in the literature. The first valida- vective conditions for heat and mass transport was solved
tion is performed by utilizing the numerical results of Basak numerically. A similar approach is adapted for the current
et al. [61] for entrapped triangular porous cavity. Compari- study for the validation. Comparisons of the moisture dis-
son of the local Nusselt number variation for the top wall tribution for the moist object for convective drying of pota-
is shown for two values of Darcy number (Fig. 2). Another toes at convective conditions for V = 1 m s−1, AR = 2 and

13
1576 F. Selimefendigil et al.

0.4

0.35
0.35

0.3
0.3

0.25
0.25

0.2
0.2

0.15 0.15

0.1 0.1

0.05 0.05

(a) ω = 0, xc = 0.2W (b) ω = 0, xc = 0.4W

0.35

0.3

0.25

0.2

0.15

0.1

0.05

(c) ω = 0, xc = 0.6W

Fig. 4  Streamline distributions for varying cylinder locations for ω = 0 rad s−1 (D = 0.1H, u = 0.2 m s−1)

0.9
1

0.9 0.8

0.8 0.7

0.7 0.6
0.6 0.5
0.5
0.4
0.4
0.3
0.3
0.2
0.2
0.1 0.1

(a) ω = – 200 rad s–1, xc = 0.2W (b) ω = – 200 rad s–1, xc = 0.4W

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

(c) ω = – 200 rad s–1, xc = 0.6W

Fig. 5  Streamline distributions for varying cylinder locations for ω = − 200 rad s−1 (D = 0.1H, u = 0.2 m s−1)

13
Convective drying of a moist porous object under the effects of a rotating cylinder in a channel 1577

1
1
0.9
0.9
0.8
0.8
0.7
0.7
0.6
0.6
0.5
0.5
0.4
0.4
0.3
0.3
0.2
0.2

0.1 0.1

(a) ω = – 200 rad s–1, xc = 0.2W (b) ω = – 200 rad s–1, xc = 0.4W

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

(c) ω = – 200 rad s–1, xc = 0.6W

Fig. 6  Streamline distributions for varying cylinder locations for ω = 200 rad s−1, (D = 0.1H, u = 0.2 m s−1, t = 90 min)

330 330

325 325

320 320

315 315

310 310

305 305

300 300

295 295

(a) ω = 0 rad s–1 (b) ω = – 200 rad s–1

330

325

320

315

310

305

300

(c) ω = – 200 rad s–1 295

Fig. 7  Surface temperature distribution of varying ω values (D = 0.1H, u = 0.2 m s−1, t = 90 min)

13
1578 F. Selimefendigil et al.

0.6 0.8

0.7
0.5
0.6

0.4
0.5

0.3 0.4

0.3
0.2
0.2

0.1
0.1

(a) D = 0.05H (b) D = 0.08H

1
1.4
0.9

0.8 1.2

0.7
1
0.6
0.8
0.5

0.4 0.6

0.3
0.4
0.2
0.2
0.1

(c) D = 0.1H (d) D = 0.15H

Fig. 8  Streamline distribution of varying diameters (ω = − 200 rad s−1, u = 0.2 m s−1, t = 90 min)

330
330

325
325

320 320

315 315

310 310

305 305

300 300

295 295
(a) D = 0.05H (b) D = 0.08H

330 330

325 325

320 320

315 315

310 310

305 305

300 300

(c) D = 0.1H 295

(d) D = 0.15H 295

Fig. 9  Surface temperature distribution of varying diameters (ω = − 200 rad s−1, u = 0.2 m s−1, t = 90 min)

13
Convective drying of a moist porous object under the effects of a rotating cylinder in a channel 1579

1 1.2
0.9

0.8 1

0.7
0.8
0.6

0.5 0.6

0.4
0.4
0.3

0.2
0.2

(a) u = 0.2 m s–1 0.1

(b) u = 0.4 m s–1

1.6
2
1.4
1.8
1.2
1.6

1 1.4

1.2
0.8
1
0.6
0.8
0.4
0.6

0.2 0.4
(c) u = 0.6 m s–1 (d) u = 0.8 m s–1
0.2

Fig. 10  Streamline distribution for different drying air velocities (ω = − 200 rad s−1, D = 0.1H, t = 90 min)

330 330

325 325

320 320

315 315

310
310

305
305

300
(b) u = 0.4 m s–1 300

(a) u = 0.2 m s–1 295

330
330

325
325

320
320

315
315

310
310

305

(c) u = 0.6 m s–1 (d) u = 0.8 m s–1 305

300

Fig. 11  Surface temperature distribution for different drying air velocities (ω = − 200 rad s−1, D = 0.1H, t = 90 min)

13
1580 F. Selimefendigil et al.

1.2
–200 0 200 Results and discussion
1 The convective drying of a rectangular porous medium
in a channel was investigated numerically in this study. A
0.8
rotating cylinder was mounted over the medium, and the
x-axis location, angular velocity and radius of the cylinder
Nuave

0.6
were varied with different values. Temperature of drying
0.4
air fixed at 333 K, while velocity of air had different values
0.2 in laminar and fully developed flow conditions. Heat and
moisture transport from porous sample to air in the chan-
0
0 10 20 30 40 50 60 nel was calculated in those conditions. The problem was a
t/min 2D model in laminar flow conditions, and non-equilibrium
mass conservation equations were used.
Fig. 12  Average heat transfer with time for different ω (rad s−1) val- Figures 4–6 show streamline distribution in channel
ues (D = 0.1H, xc = 0.4W, u = 0.2 m s−1) for varying locations of cylinder for different angular
rotational speeds. Angular direction of rotating cylinder
is clockwise when omega is negative. When the cylinder
1.80
0.2 0.4 0.6 0.8 is stationary (Fig. 4), recirculation is established behind
1.60 the cylinder. In the station where the cylinder is located
1.40 behind the porous object, the vortex established in the rear
1.20 of the object is suppressed for stationary cylinder configu-
1.00 ration (Fig. 2c). Since the gap between the cylinder surface
Nuave

0.80 and the channel and porous medium wall reduces, veloci-
0.60 ties in those regions are increased. For clockwise rotations
0.40 of cylinder in Fig. 5 (ω = − 200 rad s−1), depending on the
0.20 cylinder location, vortices are established below the cyl-
0.00
0 10 20 30 40 50 60
inder, in front of the porous object for location xc = 0.2 W
and behind the object for location xc = 0.6 W. However, for
counter clockwise rotation of cylinder (Fig. 6), more flow
Fig. 13  Average heat transfer with time for different air velocities
(D = 0.1H, xc = 0.4W, ω = − 200 rad s−1) is deflected toward the front side of the porous object, and
no vortex is seen in front of the object for this rotation of
the cylinder.
None 0.05H 0.08H 0.1H 0.15H Figure 7 shows the temperature distribution in the channel
1.4
and within the porous object for different values of ω, when
1.2 the cylinder is located over the middle of the porous object
(xc = 0.4W). For ω =0, boundary layers are developed along
1
the walls of the porous object, and with the rotation of the
0.8 cylinder, they are affected. Within the porous object, thermal
gradients are affected with different ω values.
Nuave

0.6
Figure 8 shows streamline distribution for different diam-
0.4 eter sizes of the cylinder. When the size of cylinder is small,
recirculations are established behind the cylinder in a small
0.2
region, and its effect on the main flow stream in the channel
0 is not that significant. However, when its radius is increased
0 10 20 30 40 50 60
gradually, rotation effect on the main flow stream is seen
obviously. On the upper part of the channel above the cylin-
Fig. 14  Average heat transfer with time for different cylinder diam-
der, recirculations are formed and their sizes increased with
eters (u = 0.2 m s−1, xc = 0.4W, ω = − 200 rad s−1)
increasing cylinder sizes. For higher cylinder sizes due to
more deflection of fluid flow toward the rear of the porous
T = 353 K are shown in Fig. 3 at time of 100 min. Table 3 domain, vortex in this region disappears. But it appears in
shows the comparison of numerical and experimental front of the object. Figure 9 shows surface temperature dis-
dimensionless moisture ratio for convective drying of potato tribution of varying diameter sizes at 90 min drying time.
for various time instances.

13
Convective drying of a moist porous object under the effects of a rotating cylinder in a channel 1581

0.4 0.35

0.35 0.3

0.3
0.25

0.25
0.2
0.2
0.15
0.15

0.1
0.1

(a) ω = 0 rad s–1 , xc = 0.2W 0.05 (b) ω = 0 rad s–1 , xc = 0.4W 0.05

0.35

0.3

0.25

0.2

0.15

0.1

(c) ω = 0 rad s–1 , xc = 0.6W 0.05

Fig. 15  Saturation of liquid water for different cylinder locations for ω = 0 rad s−1 (D = 0.1H, u = 0.2 m s−1, t = 90 min)

0.35

0.4
0.3
0.35

0.25
0.3

0.2
0.25

0.15 0.2

0.1 0.15

0.1

(a) ω = – 200 rad s–1, xc = 0.2W

0.05
(b) ω = – 200 rad s–1, xc = 0.4W

0.4

0.35

0.3

0.25

0.2

0.15

0.1

(c) ω = – 200 rad s–1, xc = 0.6W 0.05

Fig. 16  Saturation of liquid water for different cylinder locations for ω = − 200 rad s−1 (D = 0.1H, u = 0.2 m s−1, t = 90 min)

13
1582 F. Selimefendigil et al.

0.4 0.35

0.35
0.3

0.3
0.25
0.25
0.2
0.2
0.15
0.15

0.1
0.1

(a) ω = – 200 rad s–1, xc = 0.2W (b) ω = – 200 rad s–1, xc = 0.4W
0.05 0.05

0.35

0.3

0.25

0.2

0.15

0.1

(c) ω = – 200 rad s–1, xc = 0.6W 0.05

Fig. 17  Saturation of liquid water for different cylinder locations for ω = 200 rad s−1 (D = 0.1H, u = 0.2 m s−1, t = 90 min)

Thermal gradients in the porous object and clustering of iso- increment in the velocity, heat transfer increases. Figure 14
therms near the interface boundary are significantly affected shows average heat transfer evolution in time for varying
with higher sizes of the cylinder. size of rotating cylinder. The highest value of N ­ uave is seen
Effect of velocity of the drying air in the channel on in the situation without cylinder. When size of the cylinder
streamline and isotherm distributions is given in Figs. 10 increases, heat transfer decreases and stays steady with time
and 11. For lower channel velocity, the rotation of the cyl- except without cylinder situation.
inder is effective, and for higher u values, recirculations are Liquid water saturation distributions on surface of the
established behind the cylinder and porous object, since the porous domain for several variations at 90 min of drying
impact of rotation is reduced in this condition. Even though time are shown in Figs. 15–17. In Fig. 15, Sl for stationary
thermal boundary layers along the interface are disturbed cylinder with different locations is shown. When the cylinder
for lower velocities (increased effect of rotation), within the is in front of the porous object, most of evaporation is seen
porous object, thermal gradients become higher for channel in front corner of the object due to strong effect of cylinder
inlet velocities (Fig. 11). at that place, but a very weak evaporation for other sides of
The transport mechanism of the porous medium in the object is seen. For the center location, saturation decreases
process of drying happens by means of moisture transfer on the upper surface and back corner of the object and for
(concentration gradient) and heat transfer (temperature gra- backward location of cylinder, back corner has more evapo-
dient). Figures 12–14 demonstrate the average heat transfer ration than upper surface. It is clear that most of drying hap-
evolution in time for all the surfaces of the porous material pens in the front corner for all of the locations.
except the bottom surface. In Fig. 12, average heat transfer Figure 16 shows the location effect of cylinder when
is shown for varying angular rotational speeds. Heat transfer ω = − 200. Because of the clockwise rotation, most evapo-
is highest at the value of ω = 200 rad s−1 and at all times, and ration is seen at the location of xc = 0.2 W due to the most
lowest heat transfer is seen for ω = − 200 rad s−1. Also, heat effect of drying air just upon the surface of the object. For
transfer reduces linearly with time at ω = 200 rad s−1 when the counterclockwise rotation of cylinder, evaporation gets
it is nearly steady for other rotational speed values. Spatial deeper within and along the surface of the porous object
average heat transfer versus time for higher dry air velocities especially for the location of xc = 0.6 W.
is shown in Fig. 13. As it can be seen in the figure, with the

13
Convective drying of a moist porous object under the effects of a rotating cylinder in a channel 1583

Fig. 18  Mass fraction of vapor 0.0162 –200

along the y-axis at the middle 0
0.016 200
of porous medium for varying
ω (rad s−1) values at different 0.0158
times (D = 0.1H, u = 0.2 m s−1,

Mass fraction of water vapor

0.0156
xc = 0.4W)
0.0154

0.0152

0.015

0.0148

0.0146

0.0144

0.0142

0.014
0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045
Arc length/m
(a) t = 30 min

–200
0.016 0
200

0.0158

0.0156
Mass fraction of water vapor

0.0154

0.0152

0.015

0.0148

0.0146

0.0144

0.0142

0.014
0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045
Arc length/m
(b) t = 60 min
Line Graph: Mass fraction of water vapour (1)
0.017 –200
0
0.0168 200
0.0166
Mass fraction of water vapor (1)

0.0164
0.0162
0.016
0.0158
0.0156
0.0154
0.0152
0.015
0.0148
0.0146
0.0144
0.0142
0.014
0.0138
0.0136
0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045
Arc length/m
(c) t = 90 min

13
1584 F. Selimefendigil et al.

Fig. 19  Mass fraction of vapor 0.017

u0 = 0.2 m/s
along the y-axis at the middle 0.0168 u0 = 0.4 m/s
of porous medium for varying 0.0166 u0 = 0.6 m/s
u0 = 0.8 m/s
air velocities at different times 0.0164
(ω = − 200 rad s−1, D = 0.1H, 0.0162

mass fraction of vapor

xc = 0.4W) 0.016
0.0158
0.0156
0.0154
0.0152
0.015
0.0148
0.0146
0.0144
0.0142
0.014
0.0138
0.0136
0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045
Arc length/m
(a) t = 30 min

0.02 u0 = 0.2 m/s

u0 = 0.4 m/s
u0 = 0.6 m/s
0.019 u0 = 0.8 m/s

0.018
mass fraction of vapor

0.017

0.016

0.015

0.014

0.013

0.012

0.011

0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045

Arc length/m
(b) t = 60 min
0.026
0.025 u0 = 0.2 m/s
u0 = 0.4 m/s
0.024 u0 = 0.6 m/s
0.023 u0 = 0.8 m/s

0.022
0.021
mass fraction of vapor

0.02
0.019
0.018
0.017
0.016
0.015
0.014
0.013
0.012
0.011
0.01
0.009
0.008
0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045
Arc length/m
(c) t = 90 min

13
Convective drying of a moist porous object under the effects of a rotating cylinder in a channel 1585

Fig. 20  Liquid water saturation Line Graph: Liquid water saturation (1)
along the y-axis at the middle 0.48
of porous medium for varying –200
0
ω (rad s−1) values at different 0.47 200

times (u = 0.2 m s−1, D = 0.1H, 0.46

xc = 0.4W)

Liquid water saturation (1)

0.45

0.44

0.43

0.42

0.41

0.4

0.39

0.38
0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045
Arc length/m
(a) t = 30 min

Line Graph: Liquid water saturation (1)

0.45 –200
0
0.44 200
0.43
0.42
Liquid water saturation (1)

0.41
0.4
0.39
0.38
0.37
0.36
0.35
0.34
0.33
0.32
0.31
0.3
0.29
0.28
0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045
Arc length/m
(b) t = 60 min

Line Graph: Liquid water saturation (1)

0.42 –200
0
0.4 200
0.38
0.36
0.34
Liquid water saturation (1)

0.32
0.3
0.28
0.26
0.24
0.22
0.2
0.18
0.16
0.14
0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045
Arc length/m
(c) t = 90 min

13
1586 F. Selimefendigil et al.

Fig. 21  Liquid water saturation Line Graph: Liquid water saturation (1)
along the y-axis at the middle
of porous medium for varying 0.48 None
H*D.05
cylinder diameters at differ- 0.47 H*D.08
H*D.10
ent times (ω = − 200 rad s−1, 0.46 H*D.15

Liquid water saturation (1)

u = 0.2 m s−1, xc = 0.4W) 0.45
0.44
0.43
0.42
0.41
0.4
0.39
0.38
0.37
0.36

0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045

Arc length/m
(a) t = 30 min
Line Graph: Liquid water saturation (1)
0.46 None
H*D.05
0.44 H*D.08
H*D.10
0.42 H*D.15
Liquid water saturation (1)

0.4
0.38
0.36
0.34
0.32
0.3
0.28
0.26
0.24
0.22
0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045
Arc length/m
(b) t = 60 min

Line Graph: Liquid water saturation (1)

0.44
None
0.42 H*D.05
0.4 H*D.08
H*D.10
0.38 H*D.15
0.36
Liquid water saturation (1)

0.34
0.32
0.3
0.28
0.26
0.24
0.22
0.2
0.18
0.16
0.14
0.12
0.1
0.08
0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045
Arc length/m
(c) t = 90 min

13
Convective drying of a moist porous object under the effects of a rotating cylinder in a channel 1587

Fig. 22  Liquid water saturation Line Graph: Liquid water saturation (1)
along the y-axis at the middle 0.48
of porous medium for varying u0 = 0.2 m/s
u0 = 0.4 m/s
0.46
air velocities at different times u0 = 0.6 m/s

(ω = − 200 rad s−1, D = 0.1H,

u0 = 0.8 m/s
0.44

Liquid water saturation (1)

xc = 0.4W)
0.42

0.4

0.38

0.36

0.34

0.32

0.3

0.28

0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045

Arc length/m
(a) t = 30 min

Line Graph: Liquid water saturation (1)

0.44 u0 = 0.2 m/s
0.42 u0 = 0.4 m/s
u0 = 0.6 m/s
0.4 u0 = 0.8 m/s
0.38
0.36
Liquid water saturation (1)

0.34
0.32
0.3
0.28
0.26
0.24
0.22
0.2
0.18
0.16
0.14
0.12
0.1
0.08
0.06
0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045
Arc length/m
(b) t = 60 min

Line Graph: Liquid water saturation (1)

0.42
0.4 u0 = 0.2 m/s
u0 = 0.4 m/s
0.38 u0 = 0.6 m/s
0.36 u0 = 0.8 m/s

0.34
Liquid water saturation (1)

0.32
0.3
0.28
0.26
0.24
0.22
0.2
0.18
0.16
0.14
0.12
0.1
0.08
0.06
0.04
0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045
Arc length/m
(c) t = 90 min

13
1588 F. Selimefendigil et al.

Fig. 23  Liquid water saturation

Line Graph: Liquid water saturation (1)
along the y-axis at the middle 0.5
of porous medium for vary-
ing times (ω = − 200 rad s−1, 0.49
D = 0.2H, xc = 0.4W, 0.48
u = 0.2 m s−1)
0.47

Liquid water saturation (1)

0.46

0.45

0.44

0.43

0.42

0.41

0.4

0.39 10 min
30 min
0.38 45 min
60 min
0.37 90 min

0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045

Arc length/m

In order to analyze the moisture transport phenomenon saturation. Water saturation decreases from center to sur-
inside the porous medium, the amount of liquid and vapor face of the porous domain during drying process (Fig. 22a).
at a cross section vertically at the middle of sample is cal- Liquid saturation decreases with time for all points of the
culated for different conditions. Figure 18 demonstrates section (Fig. 23).
mass fraction of vapor along the middle section at differ-
ent times for different angular rotational speed of cylinder.
As expected, vapor concentration is highest at the time of Conclusions
90 min and it decreases slightly from center to the surface.
On the surface, concentration forms a sharp drop because Convective drying of a porous domain is performed numeri-
of the evaporation to the air for all conditions. Also, cv is cally, and heat and moisture transport for varying conditions
the highest at the counterclockwise rotation of the cylinder, was investigated. Drying air flowed in a rectangular channel
and it is vice versa for the clockwise rotation for all times with porous sample located at the center of the bottom wall.
of drying. In Fig. 19, mass fraction of vapor from center to A rotating cylinder was mounted over the porous object,
surface of the middle of the cylinder is indicated for varying and direction and location of the cylinder changed. Effects
dry air velocities. Concentration increases with increment in of various parameters such as dry air inlet velocity, cylinder
the velocity and drying time. diameter and x-axis location of the cylinder were investi-
Distribution of liquid water saturation along the y-axis gated, and it is understood that each variable had significant
section of the medium for different times and conditions effects on transport mechanism. It was obviously understood
is shown in Figs. 20–22. Figure 20 shows ­Sl for varying ω that velocity of drying air in laminar flow conditions is an
values at three different times of drying. Drying is effective important effect on heat and moisture transport. Also, rota-
in the regions closer to the surface of the object, and rotation tion of the cylinder had noticeable effect especially at coun-
is more effective at the counterclockwise directional rota- terclockwise direction of rotation. Size of the rotating cyl-
tion. Besides, water saturation reduces when time evolves. inder had an important effect on heat and moisture transfer,
In Fig. 21, Sl distribution with the change in cylinder size and increasing the diameter of the cylinder made the dry air
is shown at three different times. When there is no cylin- flow more efficient.
der, moisture content reduces along the vertical axes, and The aim of the study was to make a contribution for
at higher times when cylinder is mounted, increasing the understanding transport phenomena in convective drying
size first reduces the moisture content; then, it increases for processes and find how to make more influence on drying.
higher sizes, but at the highest cylinder size, moisture con- Several variations of the model have been developed in order
tent decreases again. Figure 22 shows water saturation for to understand the behavior of drying with different condi-
different inlet velocities of channel. As it can be clearly seen, tions and find most effective combination of implements in
increasing the velocity is very effective on decreasing water the lack of experimental data. Convective drying of objects

13
Convective drying of a moist porous object under the effects of a rotating cylinder in a channel 1589

of different shapes, three-dimensional effects and moving 17. Sandhu J, Parikh A, Pawan S. Experimental determination of con-
wall effects may be considered in the future studies. vective heat transfer coefficient during controlled frying of potato
discs. LWT Food Sci Technol. 2016;65:180–4.
18. Erdogdu F, Dejmek P. Determination of heat transfer coef-
Acknowledgements This study was supported by the scientific and ficient during high pressure frying of potatoes. J Food Eng.
technological research council of Turkey—TUBITAK under the Grant 2010;96(4):528–32.
No: 119M050 which is gratefully acknowledged. 19. Kaya A, Aydın O, Dincer I. Numerical modeling of heat and mass
transfer during forced convection drying of rectangular moist
objects. Int J Heat Mass Transf. 2006;49:3094–103.
20. Gulati T, Datta AK, Doona CJ, Ruan RR, Feeherry FE. Modeling
moisture migration in a multi-domain food system: application to
References storage of a sandwich system. Food Res Int. 2015;76(3):427–38.
21. Mavrou P, Thorpe R, Frith W, Lian G, Chen T. Mathematical
1. Akpinar EK, Midilli A, Bicer Y. Single layer drying behaviour modelling of moisture migration in confectionery multicomponent
of potato slices in a convective cyclone dryer and mathematical food systems. In: 28th European symposium on computer aided
modeling. Energy Convers Manag. 2003;44:1689–705. process engineering; 2018. p. 1625–30.
2. Hussain MM, Dincer I. Two-dimensional heat and moisture 22. Gulati T, Datta AK. Coupled multiphase transport, large defor-
transfer analysis of a cylindrical moist object subjected to mation and phase transition during rice puffing. Chem Eng Sci.
drying: a finite-difference approach. Int J Heat Mass Transf. 2016;139:75–98.
2003;46:4033–9. 23. Zhang J, Datta AK, Mukherjee S. Transport processes and large
3. Gulati T, Datta AK. Mechanistic understanding of case-hardening deformation during baking of bread. AIChE J. 2005;51:2569–80.
and texture development during drying of food materials. J Food 24. Nicolas V, Salagnac P, Glouannec P, Ploteau J-P, Jury V,
Eng. 2015;166:119–38. Boillereaux L. Modelling heat and mass transfer in deform-
4. Karim MA, Hawlader MNA. Mathematical modelling and experi- able porous media: application to bread baking. J Food Eng.
mental investigation of tropical fruits drying. Int J Heat Mass 2014;130:23–35.
Transf. 2005;48:4914–25. 25. Ousegui A, Moresoli C, Dostie M, Marcos B. Porous multiphase
5. Bai Y, Rahman MS, Perera CO, Smith B, Melton LD. Struc- approach for baking process e explicit formulation of evaporation
tural changes in apple rings during convection air-drying with rate. J Food Eng. 2010;100:535–44.
controlled temperature and humidity. J Agric Food Chem. 26. Feyissa AH, Gernaey KV, Ashokkumar S, Adler-Nissen J. Mod-
2002;50:3179–85. elling of coupled heat and mass transfer during a contact baking
6. Barati E, Esfahani JA. Mathematical modeling of convective dry- process. J Food Eng. 2011;106(3):228–35.
ing: lumped temperature and spatially distributed moisture in slab. 27. McMinn WAM, Magee TRA. Principles, methods and applica-
Energy. 2011;36:2294–301. tions of the convective drying of foodstuffs. Food Bioprod Pro-
7. Kaya A, Aydın O, Dincer I. Experimental and numerical investi- cess. 1999;77(3):175–93.
gation of heat and mass transfer during drying of Hayward kiwi 28. Ramachandran RP, Akbarzadeh M, Paliwal J, Cenkowski S. Com-
fruits (Actinidia Deliciosa Planch). J Food Eng. 2008;88:323–30. putational fluid dynamics in drying process modelling—a techni-
8. Mahiuddin Md, Khan MdIH, Pham ND, Karim MA. Develop- cal review. Food Bioprocess Technol. 2018;11(2):217–92.
ment of fractional viscoelastic model for characterizing viscoelas- 29. Nasrallah SB, Perre P. Detailed study of a model of heat and mass
tic properties of food material during drying. Food Biosci. 2018. transfer during convective drying of porous media. Int J Heat
https​://doi.org/10.1016/j.fbio.2018.03.002. Mass Transf. 1995;31(5):957–67.
9. Diamante LM, Ihns R, Savage GP, Vanhanen L. Short communi- 30. Kumar C, Joardder MUH, Farrell TW, Millar GJ, Karim A. A
cation: a new mathematical model for thin layer drying of fruits. porous media transport model for apple drying. Biosys Eng.
Int J Food Sci Technol. 2010;45:1956–62. 2018;176:12–25.
10. Ni H, Datta AK, Torrance KE. Moisture transport in intensive 31. Curcio S, Aversa M. Transport phenomena and shrinkage mod-
microwave heating of biomaterials: a multiphase porous media eling during convective drying of vegetables. In: Proceedings of
model. Int J Heat Mass Transf. 1999;42:1501–12. the COMSOL conference, Milan; 2009.
11. Chen J, Pitchai K, Birla S, Negahban M, Jones D, Subbiah J. Heat 32. Kim D, Son D, Kim S. Numerical analysis of convective drying
and mass transport during microwave heating of mashed potato in of a moving moist object. Int J Heat Mass Transf. 2016;99:86–94.
domestic oven—model development, validation, and sensitivity 33. Stanish MA, Schajer GS, Kayihan F. A mathematical
analysis. J Food Sci. 2014;79:1991–2004. model of drying for hygroscopic porous media. AIChE J.
12. Dhall A, Halder A, Datta AK. Multiphase and multicomponent 1986;32(8):1301–11.
transport with phase change during meat cooking. J Food Eng. 34. Itaya Y, Kobayashi T, Hayakawa KI. Three-dimensional heat
2012;113:299–309. and moisture transfer with viscoelastic strain–stress forma-
13. Ayappa KG. Modeling transport processes during microwave heat- tion in composite food during drying. Int J Heat Mass Transf.
ing; a review. Rev Chem Eng. 1997;13(2):1–68. 1995;38(7):1173–85.
14. Ni H. Multiphase moisture transport in porous media under inten- 35. Cevoli C, Chakravartula SSN, Rosa MD, Fabbri A. Drying of
sive microwave heating. PhD. Ithaca, New York, United States: coating on bun bread: heat and mass transfer numerical model.
Cornell University; 1997. Biosys Eng. 2019;181:1–10.
15. Farkas BE, Singh RP, Rumsey TR. Modeling heat and mass 36. Onwude DI, Hashim N, Abdan K, Janius R, Chen G, Kumar C.
transfer in immersion frying. I, model development. J Food Eng. Modelling of coupled heat and mass transfer for combined infrared
1996;29:211–26. and hot-air drying of sweet potato. J Food Eng. 2018;228:12–24.
16. Wu H, Tassou SA, Karayiannis TG, Jouhara H. Analysis and sim- 37. Siavashi M, Rostami A. Two-phase simulation of non-New-
ulation of continuous food frying processes. Appl Therm Eng. tonian nanofluid natural convection in a circular annulus par-
2013;53(2):332–9. tially or completely filled with porous media. Int J Mech Sci.
2017;133:689–703.

13
1590 F. Selimefendigil et al.

38. Sheikholeslami M, Gerdroodbary B, Shafee A, Tlili I. Hybrid in a wavy porous cavity heated from below. Int Commun Heat
nanoparticles dispersion into water inside a porous wavy tank Mass Transf. 2018;95:197–209.
involving magnetic force. J Therm Anal Calorim. 2019 (in press). 51. Selimefendigil F, Oztop HF. Mixed convection of nanofluids in a
39. Sheikholeslami M, Sheremet MA, Shafee A, Li Z. CVFEM three dimensional cavity with two adiabatic inner rotating cylin-
approach for EHD flow of nanofluid through porous medium ders. Int J Heat Mass Transf. 2018;117:331–43.
within a wavy chamber under the impacts of radiation and mov- 52. Selimefendigil F, Oztop HF. Numerical analysis and ANFIS mod-
ing walls. J Therm Anal Calorim. 2019;138:573–81. eling for mixed convection of CNT-water nanofluid filled branch-
40. Bozorg MV, Siavashi M. Two-phase mixed convection heat trans- ing channel with an annulus and a rotating inner surface at the
fer and entropy generation analysis of a non-Newtonian nanofluid junction. Int J Heat Mass Transf. 2018;127:83–599.
inside a cavity with internal rotating heater and cooler. Int J Mech 53. Chamkha AJ, Selimefendigil F, Ismael MA. Mixed convection in
Sci. 2019;151:842–57. a partially layered porous cavity with an inner rotating cylinder.
41. Siavashi M, Karimi K, Xiong Q, Doranehgard MH. Numerical Numer Heat Transf Part A Appl. 2016;69:659–75.
analysis of mixed convection of two-phase non-Newtonian nano- 54. Roslan R, Saleh H, Hashim I. Effect of rotating cylinder on heat
fluid flow inside a partially porous square enclosure with a rotat- transfer in a square enclosure filled with nanofluids. Int J Heat
ing cylinder. J Therm Anal Calorim. 2019;137:267–87. Mass Transf. 2012;55:7247–56.
42. Sheikholeslami M, Arabkoohsar A, Jafaryar M. Impact of a hel- 55. Bird RB, Stewart EW, Lightfoot NE. Transport phenomena. Hobo-
ical-twisting device on the thermal–hydraulic performance of a ken: Wiley; 2007.
nanofluid flow through a tube. J Therm Anal Calorim. 2019 (in 56. Verboven P, Scheerlinck N, De Baerdemaeker J, Nicolaı BM. Sen-
press). sitivity of the food centre temperature with respect to the air veloc-
43. Sheikholeslami M, Jafaryar M, Ali JA, Hamad SM, Divsalar A, ity and the turbulence kinetic energy. J Food Eng. 2001;48:53–60.
Shafee A, Nguyen-Thoi T, Li Z. Simulation of turbulent flow of 57. Datta AK. Porous media approaches to studying simultaneous
nanofluid due to existence of new effective turbulator involving heat and mass transfer in food processes. I: problem formulations.
entropy generation. J Mol Liq. 2019;291:111283. J Food Eng. 2007;80:80–95.
44. Sheikholeslami M, Haq R, Shafee A, Li Z. Heat transfer behavior 58. Datta AK. Porous media approaches to studying simultaneous
of nanoparticle enhanced PCM solidification through an enclosure heat and mass transfer in food processes. II: property data and
with V shaped fins. Int J Heat Mass Transf. 2019;130:1322–42. representative results. J Food Eng. 2007;80:96–110.
45. Selimefendigil F, Oztop HF, Abu-Hamdeh NH. Mixed convec- 59. Evaporation in porous media with large evaporation rates, created
tion due to a rotating cylinder in a 3D corrugated cavity filled in COMSOL Multiphysics 5.4.
with single walled CNT-water nanofluid. J Therm Anal Calorim. 60. Çengel YA, Boles MA. Thermodynamics: an engineering
2019;135:341–55. approach. New York City: McGraw-Hill Higher Education; 2006.
46. Selimefendigil F, Oztop HF, Chamkha AJ. Analysis of mixed 61. Basak T, Roy S, Ramakrishna D, Pandey BD. Analysis of
convection and entropy generation of nanofluid filled triangular heat recovery and heat transfer within entrapped porous trian-
enclosure with a flexible sidewall under the influence of a rotating gular cavities via heatline approach. Int J Heat Mass Transf.
cylinder. J Therm Anal Calorim. 2019;135:911–23. 2010;53:3655–69.
47. Costa VAF, Raimundo AM. Steady mixed convection in a differ- 62. Oztop HF, Akpinar EK. Numerical and experimental analysis
entially heated square enclosure with an active rotating circular of moisture transfer for convective drying of some products. Int
cylinder. Int J Heat Mass Transf. 2010;53:1208–19. Commun Heat Mass Transf. 2008;35:169–77.
48. Hussain SH, Hussein AK. Mixed convection heat transfer in a 63. Akpinar EK, Dincer I. Application of moisture transfer mod-
differentially heated square enclosure with a conductive rotating els to solids drying. Proc Inst Mech Eng A J Power Energy.
circular cylinder at different vertical locations. Int Commun Heat 2005;219:235–44.
Mass Transf. 2011;38:263–74.
49. Selimefendigil F, Oztop HF. Conjugate mixed convection of Publisher’s Note Springer Nature remains neutral with regard to
nanofluid in a cubic enclosure separated with a conductive plate jurisdictional claims in published maps and institutional affiliations.
and having an inner rotating cylinder. Int J Heat Mass Transf.
2019;139:1000–17.
50. Alsabery AI, Tayebi T, Chamkha AJ, Hashim I. Effect of rotating
solid cylinder on entropy generation and convective heat transfer

13