International Journal of Heat and Mass Transfer 90 (2015) 769780

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer

journal homepage: www.elsevier.com/locate/ijhmt

Mechanism study on the enhancement of silica gel regeneration

by power ultrasound with eld synergy principle and mass diffusion
theory
Ye Yao , Weiwei Wang, Kun Yang
Institute of Refrigeration and Cryogenics, Shanghai Jiao Tong University, Shanghai 200240, China

a r t i c l e

i n f o

Article history:
Received 13 December 2014
Received in revised form 28 April 2015
Accepted 23 June 2015

Keywords:
Silica gel
Enhancement of regeneration
Power ultrasound
Field synergy principle
Mass diffusivity

a b s t r a c t
The enhancement of silica gel regeneration by power ultrasound has been validated by a series of experiments in our previous studies, but there still lacks theoretical basis for illustrating its mechanism. In this
paper, the mechanism of ultrasound-enhanced regeneration has been explored. Firstly, the benet of
ultrasonic mechanical effect to the enhancement of regeneration has been illustrated by the eld synergy
 , only considering intersection angle between the velocity and
principle. The average synergy degree (j
 ov erall , considering local values of
the temperature gradient) and the average overall synergy degree (j
 ) in terms of the near wall region are suggested for analysis,
velocity and temperature gradient based on j
and they are obtained based on the komega model which is suitable for the near wall free-shear ow
velocity predicting. Results manifest that the ultrasonic mechanical effect can signicantly enlarge the
synergy degrees between the temperature and the velocity eld around the particle, and this can be used
to explain the enhancement of convective heat and mass transfer on the gas side due to the mechanical
effect of ultrasound. Afterwards, a moisture diffusion model is developed to investigate spatial distributions of moisture ratio and temperature in a silica gel particle as well as its surface equilibrium humidity
during the regeneration with and without ultrasound. Results show that ultrasonic heating effect can
lead to an increase in the average temperature and moisture diffusivity in the silica gel particle, and this
conrms the contribution of ultrasonic heating effect to the enhancement of silica gel regeneration.
2015 Elsevier Ltd. All rights reserved.

1. Introduction
Silica gel has been utilized for dehumidication processes in
industrial and residential applications due to its great pore surface
area and good moisture adsorption capacity [13]. Generally, process air ows through the silica gel bed, and the moisture in the air
is absorbed by the silica gel. After the silica gel is saturated with
moisture, it needs to be regenerated (i.e., moisture removal) for
recycling. The thermal heating method is a traditional way for
regenerating the desiccants. But, such regeneration method is of
poor energy efciency, especially for the desiccants with a relatively higher regeneration temperature. This is because, on one
hand, the higher regeneration temperature is not good for the utilization of lower-grade thermal energy (e,g., solar energy, waste
heat), and on the other hand, it will result in more energy dissipation and loss. Thus, we often expect the regeneration temperature
to be as low as possible during the applications of desiccants like
Corresponding author. Tel.: +86 21 13641943577; fax: +86 21 34206814.
E-mail address: yeyao10000@sjtu.edu.cn (Y. Yao).
http://dx.doi.org/10.1016/j.ijheatmasstransfer.2015.06.063
0017-9310/ 2015 Elsevier Ltd. All rights reserved.

silica gel. For such reason, some non-thermal methods have been
developed for the desiccant regeneration, e.g., the use of pulsed
corona plasma [4], pulsed vacuum [5], centrifugal forces [6], and
electrical elds [79]. These non-thermal methods can denitely
improve the kinetics of heat and mass transfer during the regeneration of desiccant, and hence, help to decrease the regeneration
temperature to some extent. Another new regeneration method
by using power ultrasound has been put forward recently. A series
of studies [1013] manifest that the way of applying ultrasound in
silica gel regeneration process can distinctly increase the regeneration rate and become a promising non-thermal regeneration
method. As shown in Fig. 1, the mechanism of enhancement of
regeneration by power ultrasound may be qualitatively illustrated
as below:
(1) The special effect of mechanical vibration induced by the
high-intensity ultrasound helps to intensify the uid turbulence near the solid medium and reduce the thickness of
boundary layer on the gassolid interface, and this will
decrease the external resistance of heat and mass transfer.

770

Y. Yao et al. / International Journal of Heat and Mass Transfer 90 (2015) 769780

Nomenclature
a
cp
dp
D
f
Hads
HF
j
k
K
m
_
m
Nu
p
Pr
q
r
rs
Re
P
Sr
T/t
T
u
!
u
~
u
V
w
w
xi

energy absorption coefcient

specic heat capacity, J/(kg C)
average pore size in silica gel, m
diffusivity, m2/s
acoustic frequency, Hz
adsorption heat of silica gel, J/kg
heat ux, W/m2
imaginary number
turbulent kinetic energy, m2/s
sound wave number,
mass, kg
mass diffusion rate, kg/(m2 s)
Nusselt number
pressure, Pa
Prandtl number, dimensionless
moisture ratio, kg/(kg drysample)
distance, m
radius of silica gel particle, m
Reynolds number, dimensionless
power of ultrasound, W
wave surface area at the distance of r from center of silica gel particle, m2
temperature, K/C
dimensionless temperature
velocity, m/s
uid velocity vector, m/s
dimensionless uid velocity vector
near wall region
humidity, kg water/(kg dryair)
surface equilibrium humidity of silica gel, kg water/(kg
dryair)
cartesian coordinates

d
e

rT
rT
h

q
u
j
j
j ov erall
l
lt
x
s

thermal boundary layer thickness, m

porosity
temperature gradient, K
dimensionless temperature gradient, K
eld synergy angle between velocity vector and temperature gradient, degree
density, kg/m3
surface relative humidity of silica gel, %
specic dissipation, 1/s
local synergy degree, dimensionless
average synergy degree, dimensionless
overall synergy degree, dimensionless
molecular dynamic viscosity, kg/(m s)
turbulent eddy viscosity, kg/(m s)
angular frequency, x = 2pf, rad/s
time, s

Subscripts
a
moist air or on the gas side
atm
standard atmosphere
eff
effective
H2O
water molecule
i
current spatial grid point
N
number of spatial grids in silica gel particle
s
silica gel particle
sv
water vapor in the saturated air
vib
vibration
U
in the presence of ultrasound
NU
in the absence of ultrasound
ult
ultrasound
w
wall surface
1
value at great distance from a body

Greek symbols
am
convective mass transfer coefcient, m/s
k
thermal conductivity, W/(m C)

Superscripts
n
current time point
0
initial condition

(2) The heating effect of ultrasound leads to a temperature rise

of medium, and this will increase the internal moisture diffusion in the solid desiccants.

often used to analyze some key variables (like the uid velocity
and temperature difference) that affect the heat and mass transfer
rate, they can not reveal the mechanism of the heat and mass
transfer enhancement. The theory of boundary layer (known as
the layer of uid in the immediate vicinity of a bounding surface
where the effects of viscosity are signicant) is often employed
to study and explain the phenomena of heat and mass transfer
enhancement [1822]. It has been considered that the resistance
of heat and mass transfer mostly happens within the boundary
layer. However, the mechanism analysis based on the layer boundary theory is only qualitative since the boundary layer thickness is
very difcult to obtain in most cases, particularly for the situations
where the vibration is used for the heat and mass transfer
enhancement.
In 1998, Guo et al. [23] rst proposed a novel concept of optimizing and enhancing convective heat transfer of parabolic ows
which is called eld synergy principle (FSP). The FSP has been successfully employed to study the relationship between local behavior of heat transfer and heat transfer enhancement and analyze
how local behaviors affect the overall thermal performances [24
29]. For example, He et al. [26] employed the eld synergy principle to study the effects of the ve parameters (i.e., Re number, n
pitch, tube row number, spanwise and longitudinal tube pitch) on
the heat transfer performance of the nned tube banks. They found
that the enhancement or deterioration of the convective heat

The above mechanisms have been used to explain the experimental results on the ultrasound-assisted regeneration and
develop corresponding models [1013]. But, the mechanisms still
lack of theoretical basis. In this paper, the mechanisms of the
enhancement of solid desiccant regeneration by power ultrasound
are further discussed in a quantitative way. For the rst mechanism analysis (i.e., due to the mechanical vibration effect of ultrasound), the eld synergy theory is employed; and for the second
(i.e., due to the heating effect of ultrasound), the diffusion theory
is adopted.

2. Enhancement of regeneration by ultrasonic mechanical

effect
2.1. Field synergy principle
Many studies on the heat and mass transfer enhancement
mostly focused on the heat and mass transfer coefcient which
reects the convection intensity between uid and solid wall
[1417]. Although the heat and mass transfer coefcients can be

771

Y. Yao et al. / International Journal of Heat and Mass Transfer 90 (2015) 769780

The energy of mechanical vibration

ultimately converts into heat that results
in a temperature rise in the desiccant.

Micro-vibration effect

Temperature rise in the desiccant will

enhance the moisture diffusivity ; Hence,
solid-side mass transfer resistance will
decrease.

Break moisture film around the solid

external surface; Increase gas turbulence
near the solid surface. Hence, gas-side mass
transfer resistance will decrease.

Ultrasonic
Energy

Part of ultrasonic energy will be absorbed

by the desiccant and converts into heat
that results in a temperature rise in the
desiccant.

Heating effect

Fig. 1. Mechanism of ultrasonic effects on the improvement of solid desiccant regeneration.

transfer across the nned tube banks is inherently related to the

variation of the intersection angle between the velocity and the
uid temperature gradient. Tao et al. [27] made a 3-D numerical
simulation on laminar ow heat transfer of the n-and-tube
surface with vortex generators whose inherent mechanism of heat
transfer enhancement by longitudinal vortex have been well
explained by the eld synergy principle.
According to the FSP, the intensity of convective heat transfer
not only depends on the absolute value of uid velocity and temperature gradient near the wall surface, but also on the intersection
angle between velocity vector and temperature gradient/heat ow,
and the convective heat transfer could be enhanced by reducing
the intersection angle between velocity vector and temperature
gradient.
The heat ux between uid and solid wall, HFw, can be
expressed by Eq. (1).

HF w ka

@t
k
@y w

Z
0

qa cp !
u  rTdy

As seen from Eq. (1), for a certain ow rate and temperature difference between the wall and the incoming ow, the wall heat ux
increases with the decreasing of the included (intersection) angle
between the velocity and temperature gradient/heat ow vectors.
Introducing the following dimensionless variables for the
boundary layer ow,

~
u
~
u
;
~
u1

rT

rT
T 1  T w =d

where, the subscript 1 stands for the value far from the wall.
Eq. (1) can be written in the dimensionless form as below [29]:

Nu Re  Pr

  cos hdy
j~
uj  jrTj

gradient should be as parallel as possible; (b) the local values of the

three scalar elds should all be simultaneously large; (c) the
velocity and temperature proles at each cross section should
be as uniform as possible. A better synergy among such three
scalar elds will bring about a larger value of Nux.
 ) in
For convenience of analysis, the average synergy degree (j
terms of the near wall region (V) is suggested by Eq. (5).

j

jjjdV
V

 ov erall ) considIn addition, the average overall synergy degree (j

ering local values of the velocity and the temperature gradient is
given as below:

j ov erall


jjj  j~
uj  jrTjdV
V

2.2. CFD modeling in the ultrasound eld

Before the eld synergy principle is applied to explain the heat
and mass transfer enhancement by ultrasonic vibration, the eld of
uid velocity, temperature and concentration around the solid surface must be acquired in advance. For the convenience of analysis,
a single particle of silica gel is taken as the example. It is assumed
that the ultrasound waves transmit from solid to air, and the sound
wave front is considered as a spherical surface. The physical model
for the eld synergy analysis is shown in Fig. 2, where, ua,o, ta,o and
wa,o are velocity, temperature and humidity of incoming air.
The governing equations for calculating the eld of air velocity,
temperature and humidity around the silica gel particle in sound
eld are given as follows [30]:
Continuity equation:

rua 0

where, h is the intersection angle between velocity vector and temperature gradient (heat ow vector), and cos h denotes local synergy
degree (j) expressed by Eq. (4).

!
u  rT
j cos h !
j u j  jrTj

Calculation region

Eq. (3) gives us a more general insight on convective heat transfer. From Eq. (3), we can see that increasing Reynolds (Rex) or/and
Prandtl number (Pr) as well as the value of the dimensionless integration will help to increase the Nusselt number (Nux) that reects
the convective heat transfer intensity. The synergy of the three sca and cos h) implies that (a) the intersection angle
lar elds (j~
uj; jrTj
between the velocity and the temperature gradient/heat ow
should be as small as possible i.e., the velocity and the temperature

ua,i

Silica gel particle

ta,i
wa,i

Ultrasound source

Fig. 2. Physical model for the eld synergy analysis.

772

Y. Yao et al. / International Journal of Heat and Mass Transfer 90 (2015) 769780

Momentum equation:

@ua
1
l
ua rua  rpa pultra a r2 ua
@s
qa
qa

Energy conservation equation:

@t a
ka
ua rt a
Dt
@s
qa cp;a a

11

where, r is distance from particle center, m; Sr (Sr = 4pr2) is wave

surface area at the distance of r from center of silica gel particle,
m2.pultra can be expressed as:

pultra

C 1 jxsKr C 2 jxsKr
e
e
r
r

12

where, x is angular frequency, rad/s; K is sound wave number,

K = 2pf/uultra. The rst item on the right side of Eq. (12)
C jxsKr 
1
e
stands for the forward waves, and the second item
Cr2 jxsKr 
for
the back waves. The undetermined constants, C1
e
r
and C2, depend on the boundary conditions. In this case, the back
waves only appear within the solid medium. Thus, C2 = 0.
The inductive vibration speed caused by sound pressure at the
distance of r from the sound source (uvib,r) may be estimated by
[33]:

uv ib r 



@pultra
C1
1
ejxsKr

1
jxqa @r
jKr
r qa cp;a
1

13

Assuming the sound-induced vibration speed at the particle

surface is given as uvib(rs), we have:




 C1
1 

1
juv ib r s j
r q c
jKr s 
s a p;a

14

Thus, the constant, C1, can be got as:

jC 1 j

qa cp;a Kr2s juv ib rs j

q
1 Kr s 2

15

Usually, the sound-induced vibration speed at the particle surface, uvib(rs), can be calculated by sound frequency (f) and vibration
amplitude at the particle surface (Am) as below:

uv ib r s Am sin2pf s

16

The sound pressure at the distance of r from the sound source

(pultra(r)) can be written as:

jpultra rj

qa cp;a Kr2s juv ib rs j

r

q
1 Kr s 2

17

The komega model is one of the most commonly used turbulence models. It is a two-equation model, which includes two extra
transport equations to represent the turbulent properties of the
ow. This allows a two equation model to account for history
effects like convection and diffusion of turbulent energy. It has
been reported that the komega model is suitable for the near wall
free-shear ow velocity predicting, such as wake ow, detour ow

18

xi 2 x; y; z
19

10

@
The arithmetic operator r @r
, and D @r@2 .
In Eq. (8), the pressure due to the sound vibration, pultra, can be
obtained through solving the wave equation as below [31,32]:

@ 2 pultra @pultra @ ln Sr
1 @ 2 pultra

2
2
cp;a @ s2
@r
@r
@r



@
@
@
@k
Gk  Y k Sk xi 2 x; y; z
qk
qkui
Ck
@s
@xi
@xi
@xi


@
@
@
@ G-  Y - Sq-
q-ui
C@s
@xi
@xi
@xi

Mass conservation equation:

@wa
ua rwa DDwa
@s

and jet ow [3438]. Therefore, the komega model is employed

for this study where the ow around a spherical particle under
the ultrasonic vibration is considered.
The standard komega model can be expressed as below [39]

The rst transported variable (k) is turbulent kinetic energy

(m2/s), which determines the energy in the turbulence (Eq. (18));
the second one (-) is the specic dissipation rate (1/s), and it
determines the scale of the turbulence (Eq. (19)). Ck and C- are
diffusion rate of k and -, respectively. Gk and G- are turbulent
kinetic energy produced by k and -, respectively. Yk and Y- are
turbulence due to diffusivity of k and -, respectively. Sk and Sare custom parameters. The key model constants in the standard
komega model have been suggested in Ref. [39].
The unsteady segregate solution is adopted in the calculations.
The SIMPEC algorithm is used to solve the pressurevelocity coupled equations. The implicit rst-order scheme is used for the
unsteady terms. The third-order upwind difference scheme
QUICK is applied to the convection terms in the momentum equations and a second-order scheme for the k and - transport equations. For the diffusion terms, the central difference scheme is
applied. These numerical algorithms are quite standard now.
A silica gel particle with the diameter of 4 mm is taken as the
case study. The ultrasonic frequency (f) is 23 kHz, and the vibration
amplitude at the particle surface (Am) due to the sound vibration is
assumed to be 20 lm. The Integrated Computer Engineering and
Manufacturing Code for Computational Fluid Dynamics
(ICEMCFD) is used for the meshing. The O-Grid scheme is adopted
to reduce mesh distortion near the surface of particle. Starting
from the surface of particle, the grid heights near the surface of
particle grow from 0.02 mm with the growth ratio of 1.05 until
to 0.5 mm. A ne grid is created near the particle in the wake
and in the far eld, as shown in Fig. 3. The other calculation conditions for this study are as follows: 300 K for the surface temperature of silica gel particle; 0.025 (kg/kg dryair) for the air
humidity near the silica gel particle; 323 K and 0.01 (kg/kg dryair)
for the incoming air temperature and humidity, respectively. Three
incoming air velocities, i.e., 0.6, 1.0 and 1.4 m/s, are investigated.
2.3. Results
The software of CFD-Post is used to manipulate the CFD results
of velocity, temperature and humidity eld around the silica gel
particle and obtain the synergy degree. A computer with 16-gabit
memory and 8-core CPU clocked at 3.2 GHz is used for the CFD calculation. The time step is set as 1/(80f), and it takes about 70 h to
obtain the convergence of result.
Fig. 3 gives the air velocity eld around the particle in terms of
different sound vibration periods. As seen from Fig. 4, the sound
vibration will produce great inuence on the patterns of uid ow
near the particle. Obvious difference of uid velocity eld can be
found among different sound vibration periods.
The temperature and humidity elds around the particle with
and without ultrasound (time = 1.5 s) are presented in Fig. 5 and
Fig. 6, respectively. It can be seen that in the wake region, there
is a signicant distinction in the temperature and humidity elds
between the case with and without ultrasound. The isothermal
and isohume curves in the wake zone bend less in the presence
of ultrasound than that without ultrasound.

Y. Yao et al. / International Journal of Heat and Mass Transfer 90 (2015) 769780

773

0.6 m/s. Obviously, the higher vibration amplitude will be conduc and the j
 ov erall : It indicates that the
tive to increasing both the j
increase of acoustic vibration amplitude will help to increase the
convective heat and mass transfer rate, and this can be directly
reected by the overall heat transfer coefcients under different
vibration amplitudes of ultrasound as shown in Fig. 11. The overall
heat transfer coefcients are estimated by using the heat transfer
quantity and temperature gradient based on the CFD results.
3. Enhancement of regeneration by ultrasonic heating effect
During the transmission of ultrasound in medium, part of
acoustic energy will convert into heat and cause the medium to
have a visible temperature rise due to energy dissipation and
acoustic adsorption. This is called heating effect of ultrasound.
Since the temperature affects greatly the mass diffusion in solid
medium, the ultrasonic heating effect can surely improve the
regeneration rate of silica gel particles. In the following section,
the mass diffusion model in silica gel is built with the mass diffusion theory. The model can be used to investigate the moisture
change in silica gel particles during the regeneration with or without ultrasonic irradiation and illustrate the positive inuence of
ultrasonic heating effect on the regeneration of silica gel.
3.1. Moisture diffusion modeling
Silica gel has very high moisture adsorption capacity because of
its microporous structure of internal interlocking cavities that is of

Fig. 3. Computational grid on the symmetry plane.

The synergy degrees between the temperature and the velocity

eld are used for the analysis of convective heat and mass transfer
enhancement by ultrasound vibration. The local synergy degrees
between the temperature and the velocity eld have been obtained
and shown in Fig. 7. As seen from Fig. 7, the eld synergy degrees
between the temperature and the velocity eld around the particle
in the presence of ultrasound are generally higher than those without ultrasonic irradiation.
 ) and average overThe variations of average synergy degree (j
 ov erall ) between the temperature and the velocity eld
all one (j
against the period of sound vibration under different incoming
air velocities (0.6, 1.0 and 1.4 m/s) are presented in Fig. 8. The symbol U and NU stand for the situation with and without ultrasound, respectively; 0.6, 1.0 and 1.4 for the incoming air
velocity of 0.6, 1.0 and 1.4 m/s, respectively. The computation zone
for this case is within the temperature gradient of 5000 C/m from
the surface of particle. Obviously, the synergy degrees in the presence of ultrasound change with the sound vibration period in a
periodic way, and they are much larger than those without
ultrasound.
Fig. 9 shows the average synergy degrees in terms of a vibration
period under different incoming air ow velocities. It can be seen
 ) decreases with the increase
that the average synergy degree (j
of incoming air ow velocity, while it is opposite for the average
 ov erall ). This can be used to explain the
overall synergy degree (j
phenomenon that the increase slope of convective heat and mass
transfer will tend to drop with the increase of air ow velocity.
Fig. 10 gives variations of synergy degrees with the vibration
amplitude of ultrasound under the incoming air ow velocity of

1/8 period of sound vibration

1/2 period of sound vibration

Fig. 4. Air velocity eld around the particle for different periods of sound vibration.

774

Y. Yao et al. / International Journal of Heat and Mass Transfer 90 (2015) 769780

diffusion are considered on the solid side) [40]. Compared with the
mass diffusion, the water molecule desorption is much faster. So, it
can be assumed that there should always exist phase equilibrium
within the solid, based on which the moisture mass equation of silica gel can be written as:

es

@qa wa
@q
1 @ 2
_ s
qs s  2
r m
@s
r @r
@s

21

_ s (kg/s) is diffusion rate of water

where, es is porosity of silica gel; m
molecules in silica gel, which can be expressed by Eq. (21).

_sm
_ Knudsen m
_ surface
m

22

_ Knudsen and m
_ surface stand for the Knudsen and the surface
where, m
diffusion rate, respectively.

_ Knudsen qa DKnudsen

m
_ surface qs Dsurface
m

(a) In the presence of ultrasound

@wa
@r

@qs
@r

23

24

Thus, Eq. (20) can be rewritten as:

es

 

@qa wa
@q
1 @
@qs qs
@qa wa
r 2 Dsurface
qs s 2
DKnudsen
@s
@r
@r
@ s r @r
25

(b) In the absence of ultrasound

Fig. 5. Comparison of temperature eld between with and without ultrasound
(time = 1.5 s).

a high internal surface area up to 800 m2/g. During the regeneration, the water vapor pressure in the ambient air is normally lower
than that near the silica gel particles. Such pressure difference
makes the water molecules move from the silica gel to the surrounding environment. Thus, the moisture transfer during the
regeneration of silica gel consists of two basic processes: one is
moisture diffusion in silica gel (i.e., solid-side mass diffusion);
another is convective mass transfer between silica gel and air
(i.e., gas-side mass transfer). The schematic diagram for moisture
diffusion modeling in a silica gel particle is presented in Fig. 12
where qs stands for moisture ratio, ts for temperature of silica
gel, rs for radius, hult for energy gain from ultrasound, wa for surface equilibrium humidity of silica gel and wa,1 for ambient air
humidity.
_ H2 O;a )
The moisture diffusion rate on the gas side of particle (m
can be expressed as below:

_ H2 O;a am;a qa wa  wa;1

m

(a) In the presence of ultrasound

20

The mass transport mechanisms on solid side mainly include

the Fick, the Knudsen and the surface diffusion. The Fick diffusion
occurs when the molecules of the gas collide with each other more
often than with the pore walls of a porous medium, and it can be
ignored in silica gel particles (i.e., only the Knudsen and the surface

(b) In the absence of ultrasound

Fig. 6. Comparison of humidity eld between with and without ultrasound
(time = 1.5 s).

775

Y. Yao et al. / International Journal of Heat and Mass Transfer 90 (2015) 769780

(a) In the presence of ultrasound

(b )In the absence of ultrasound

Fig. 7. Comparisons of local eld synergy degrees between the temperature and the velocity eld with and without ultrasound.

In Eqs. (22)(24), DKnudsen and Dsurface are molecular diffusivity

(m2/s), and they can be calculated by Eqs. (25) and (26), respectively [41,42].
1.4_U
1.0_U
0.6_U

Average synergy degree

1.4_NU
1.0_NU
0.6_NU

DKnudsen 22:86  T 0:5

s dp
Dsurface

1:6  106

1surface

exp

26


0:974Hads
Ts

27

where, Ts is temperature of silica gel, K; dp is average pore size in

silica gel, m; 1surface is surface tortuosity factor that accounts for
the increase in diffusion length due to tortuous paths of real pores,
whose recommended value is about 2.8 [41]; Hads is the heat of
adsorption (J/kg), which can be considered as a function of moisture
ratio (q : kg water/(kg dry sample)) in silica gel as given by Eq. (27)
[42].

(
Hads
Period of sound vibration

1:240qs 0:350  107

qs 6 0:05

qs > 0:05

0:140qs 0:295  10

28

The surface equilibrium humidity of silica gel, wa , can be calculated by

Average overall synergy degree

1.4_NU
1.0_NU
0.6_NU

1.4_U
1.0_U
0.6_U

wa

0:622ua psv
patm  ua psv

29

where, the water vapor pressure (psv) and the surface relative
humidity (ua ) may be calculated by Eq. (29) [43] and Eq. (30)
[44], respectively.

psv 611:21  exp



18:678t s  t2s =234:5
257:14 t s

ua s1 ts q2s s2 ts qs s3 q4s s4 q3s s5 q2s s6 qs s7 Lnqs

Period of sound vibration

Fig. 8. Synergy degrees against the period of sound vibration under different
incoming air velocities.

30
31

In Eq. (30), the parameters, s1, s2, s3, s4, s5, s6, s7, have been
obtained, respectively, as 82.9, 4.64, 252148.6, 125044.6,
16001.7, 173.4, and 0.863 in Ref. [12]. From Eqs. (28)(30), we
can see that the surface equilibrium humidity of silica gel, wa ;
can be expressed as a function dependent of surface temperature
(ts) and moisture ratio of silica gel (qs), e.g., wa Ft s ; qs .

776

Y. Yao et al. / International Journal of Heat and Mass Transfer 90 (2015) 769780

Average synergy degree_NU

Average synergy degree_U

Average overall synergy degree_NU

Average overall synergy degree_U

Fig. 9. Comparisons of average synergy degrees in terms of a vibration period between the cases with and without ultrasound under different incoming air ow velocities.

regarded as uniform distribution. Therefore, we have the following

equations:

Average synergy degree

Average overall synergy degree_U

Average overall synergy degree

Average synergy degree_U

@qa wa

@s
@qa wa

@r





@qa wa @qs
@q
g 0 qs s
@qs
@
s
@s
t

32


@qa wa @qs
@q
g 0 qs s
@qs
@r
@r
t

33

where, g 0 qs

@qa wa
@qs

Combining Eqs. (24), (31) and (32), we have

Vibration amplitude of ultrasound (m)

Fig. 10. Variations of synergy degrees with the vibration amplitude of ultrasound
under the incoming air ow velocity of 0.6 m/s.

34

For the silica gel particle, 0 < g0 (qs) < 0.4 kg/m3, es < 1, qs > 500 kg/m3, which results in a very small value in esg 0 (qs)/qs (<8  104).
Thus, Eq. (33) can be simplied as



@qs
1 @
@q
2
r 2 Deff s
@ s r @r
@r

35

Overall heat transfer coefficient (W/m K)

The Biot number for a silica gel particle is normally no bigger

than 0.15, i.e., the resistance of heat transfer is much smaller than
that of mass transfer, and the temperature in silica gel can be

@qs
1
1

@ s es g 0 qs =qs 1 r 2
 


@
g 0 qs @qs
r 2 Dsurface DKnudsen

@r
qs
@r

Vibration amplitude of ultrasound (m)

Fig. 11. Overall heat transfer coefcients for different vibration amplitudes of
ultrasound (Incoming air ow velocity: 0.6 m/s).

Fig. 12. Schematic diagram for moisture diffusion modeling in a silica gel particle.

777

Y. Yao et al. / International Journal of Heat and Mass Transfer 90 (2015) 769780

where, Deff is effective diffusivity (m2/s), Deff = Dsurface +

DKnudseng 0 (qs)/qs .
The energy equation for silica gel particle is written as

cp;s ms

@t s
4ah;a pr2s t s  t a;1  4am;a qa pr2s
@s
wa  wa;1 Hads ault Pult

Equations. (34)(39) can be used to calculate the temperature and

humidity ratio distribution in a silica gel particle. Through adopting
central-difference scheme in space and fully implicit difference format in time, Eqs. (34) and (35) can be converted into discrete forms
as below:

36

qns;j  qn1
s;j

Ds

where, cp,s is specic heat capacity of silica gel, J/(kg C); ault is ultrasonic absorption coefcient; Pult is ultrasonic power, W.
Boundary conditions for solving Eqs. (34) and (35) are given as
below:

qs r; s 0 q0

37

t s s 0 ts;0

38


@qs 
0
@r r0

39

qs Deff


@qs 
qa am;a wa  wa;1
@r rrs

cp;s ms

Deff

!
 n

n
qns;j1  2qns;j qns;j1
2 qs;j1  qs;j1

rj
2 Dr
Dr2

t ns  tn1
s
4ah;a pr 2s t ns  t a;1  4am;a qa pr 2s
Ds
wn
a  wa;1 H ads ault P ult

Start

Input the necessary parameters including properties of silica gel particle (like density, thermal
capacity, particle size and so on), the ambient air temperature and humidity, ultrasonic energy
0

and frequency (in the case of ultrasonic irradiation). Meanwhile, give initial moisture ratio ( qs )
0

and initial temperature ( ts ) of silica gel (assuming uniform distribution); n=1,

Assuming qs , N +1 = qs

n-1

, ts = ts

n-1

*n

Using Eqs. (28)-(30) to calculate wa with qs , N +1 and ts

Using Eq. (43) to obtain qs , N

Using Eq. (40) to obtain the moisture ratio distribution in silica gel particle
at current time qs , j , j [1, N ]
n

No

42

where, the superscript n and n  1 denote the current and the

previous time point of calculation, respectively; the subscript
j  1, j and j + 1 stand for the adjacent three spatial grids,
j e [1, N] in which N is the number of spatial grids in silica gel particle. j = 1, j = 2,. . ., j = N  1 and j = N stand for the spatial grids
from the center to the external surface of the particle.
Meanwhile, the boundary conditions Eqs. (38) and (39) can be
expressed by:

40

Update qs , N +1

41

qsn,1 qsn,2 qsn,1 < 0.01

Yes
n*

Using Eq. (41) to obtain ts (here, * denotes new value of ts )

No

t sn* t sn tsn* < 0.01

Update ts

Yes
No

n < N

n=n+1

Yes
Output final results of qs , j , n [1, N ] , j [1, N r ] , N is total
n

calculation time, and N r is spatial grid number.

Fig. 13. Flow chart for calculating moisture ratio distribution in silica gel particle during the regeneration process.

Y. Yao et al. / International Journal of Heat and Mass Transfer 90 (2015) 769780

Moisture ratio

778

Distribution of moisture ratio in

silica gel at 400 s in time (NU)

Moisture ratio

r/r s
Distribution of moisture ratio in
silica gel at 400 s in time (U)

r/r s

Fig. 14. Spatial distributions of moisture ratio in silica gel particle during regeneration with or without ultrasonic irradiation.

Fig. 15. Comparisons of average moisture ratio in silica gel during the regeneration
with and without ultrasound.

qns;1 qns;2

qs Deff

 n

qN1  qnN
qa am;a wn
a  wa;1
Dr

43

Fig. 16. Comparisons of temperature in silica gel during the regeneration with and
without ultrasound.

ratio in the gas phase and the solid phase is uniform. Then, Eqs.
(40)(43) are solved in a sequential manner using an overall iterative loop. The ow chart for the calculation is given in Fig. 13.

44

The calculation process is illustrated as follows: To begin with,

the necessary parameters including properties of silica gel particle
(e.g., density, thermal capacity, particle size and so on), the ambient air temperature and humidity, ultrasonic energy and frequency
(in the case of ultrasonic irradiation) as well the initial conditions
are given in advance. Initially (s = 0), temperature and moisture

3.2. Results
A silica gel particle with the diameter of 5 mm is taken as a case
study. The basic physical parameters of the silica gel are given as
below: 750 m3/kg in density, 921 J/(kg C) in specic heat,
0.2 kg/(kg dry) in moisture ratio. For the case of ultrasonic irradiation, the acoustic frequency and power density are 25 kHz and

Y. Yao et al. / International Journal of Heat and Mass Transfer 90 (2015) 769780

779

Fig. 17. Comparisons of moisture diffusivity in silica gel during the regeneration
with and without ultrasound.

0.1 W/cm3. The air ow velocity is 0.5 m/s, the initial temperature
of silica gel is 20 C, and the ambient air temperature and humidity
are 50 C and 0.01 kg/(kg dry air), respectively.
The spatial distributions of moisture ratio in silica gel particle
during regeneration with or without ultrasonic irradiation are presented in Fig. 14. Basically, the moisture ratio decreases gradually
with the time, and it has a small decrease from center (r/rs = 0) to
surface (r/rs = 1). Fig. 15 compares the average moisture ratios in
silica gel particle during the regeneration with and without ultrasound. It shows that the average moisture ratio in the presence
of ultrasonic irradiation is always lower than that without ultrasound, and the gap between them increases rst and ultimately
becomes fairly regular with the regeneration time going on.
The variations of temperature in silica gel during the regeneration with or without ultrasound are shown in Fig. 16. It can be seen
that the temperature of silica gel in the presence of ultrasonic irradiation is obviously higher than that without ultrasonic irradiation,
and the gap between them increases with the time. Besides, the
temperature in silica gel decreases rst and then increases with
the time. This is possible because in the beginning the water desorption may take away more heat from silica gel particle than
the heat gain from ambient air due to the heat transfer. Since temperature determines moisture diffusivity to a large degree, the
change patterns of moisture diffusivity in silica gel with the time
are similar to that of the temperature, as shown in Fig. 17.
Fig. 18 compares the surface equilibrium humidity of silica gel
during the regeneration with and without ultrasound. The results
show that the surface equilibrium humidity of silica gel in the

Surface equilibrium humidity (kg/kg DA)

ZVB18












presence of ultrasonic irradiation is higher than that without ultrasonic irradiation, and the gap between them increases rst and
then decreases with the time. For the case of regeneration, the
higher surface equilibrium humidity means the larger potential
difference of mass transfer on the gas side, and this is conducive
to enhancing the regeneration rate.
The contribution of heating effect of ultrasound to the moisture diffusivity in silica gel and the surface equilibrium humidity
will be affected by some parameters like ambient air temperature.
Fig. 19 gives contributions of ultrasonic heating effect to the average effective diffusivity in terms of 600 s under different ambient
air temperatures. The results indicate that the heating effect of
ultrasound contributes more to the moisture diffusivity in silica
gel and less to the surface equilibrium humidity under the lower
ambient air temperatures.

4. Conclusions

ZVB8



Fig. 19. Contributions of ultrasonic heating effect to the average effective diffusivity in terms of 600 s under different ambient air temperatures.







Time (s)
Fig. 18. Comparisons of surface equilibrium humidity of silica gel during the
regeneration with and without ultrasound.

The mechanisms of enhancement of solid desiccant regeneration by power ultrasound have been studied with the eld synergy
principle and mass diffusion theory. Firstly, the komega model
(which has been demonstrated to be suitable for this situation) is
employed to compute the airow velocity eld and temperature
eld around a silica gel particle in the presence of ultrasonic irradiation, and the eld synergy principle is used to illustrate the
enhancement of convective heat and mass transfer by ultrasonic
vibration. Results show that ultrasonic vibration can increase the
synergy degree between the temperature and the velocity eld
around the particle, and this can be used to reveal the mechanism
of enhancement of silica gel regeneration by ultrasonic vibration.
Then, a moisture diffusion model has been built to study another
mechanism of enhancement of regeneration by ultrasonic heating
effect. The model can be used to investigate the temperature and

780

Y. Yao et al. / International Journal of Heat and Mass Transfer 90 (2015) 769780

moisture ratio distribution and the effective moisture diffusivity in

a silica gel particle under any specic conditions with or without
ultrasonic irradiation. The results arrived at by this study conrm
that the ultrasonic mechanical effect and heating effect can
indeed improve the regeneration rate of solid desiccants.
Conict of interest
None declared.
Acknowledgements
This work was supported by Shanghai Pujiang Program (2012)
and a grant from the National Natural Science Foundation of
China (No.51076100).
References
[1] K.S. Chang, M.T. Chen, T.W. Chung, Effects of the thickness and particle size of
silica gel on the heat and mass transfer performance of a silica gel-coated bed
for air-conditioning adsorption systems, Appl. Therm. Eng. 25 (1415) (2005)
23302340.
[2] A.K. Ramzy, R. Kadoli, T.P.A. Babu, Experimental and theoretical investigations
on the cyclic operation of TSA cycle for air dehumidication using packed beds
of silica gel particles, Energy 56 (7) (2013) 824.
[3] C.Y. Tso, Christopher Y.H. Chao, Activated carbon, silica-gel and calcium
chloride composite adsorbents for energy efcient solar adsorption cooling
and dehumidication systems, Int. J. Refrig 35 (6) (2012) 16261638.
[4] T. Yamamoto, G. Tanioka, M. Okubo, T. Kuroki, Water vapor desorption and
adsorbent regeneration for air conditioning unit using pulsed corona plasma, J.
Electrostat. 65 (4) (2007) 221227.
[5] I. Escriche, R. Garefa-Pinchi, A. Andres, P. Fito, Osmotic dehydration of kiwi
fruit (Actinidia chinensis): uxes and mass transfer kinetics, J. Food Process
Eng 23 (3) (2002) 191205.
[6] E. Azuara, H.S. Garca, C.I. Beristain, Effect of the centrifugal force on osmotic
dehydration of potatoes and apples, Food Res. Int. 29 (2) (1996) 195200.
[7] N.K. Rastogi, M.N. Eshtiaghi, D. Knorr, Accelerated mass transfer during
osmotic dehydration of high intensity electrical eld pulse pretreated carrots,
J. Food Sci. 64 (6) (1999) 10201023.
[8] R.H. Qi, C.Q. Tian, S.Q. Shao, M.S. Tang, L. Lu, Experimental investigation on
performance improvement of electro-osmotic regeneration for solid desiccant,
Appl. Energy 88 (8) (2011) 28162823.
[9] R.H. Qi, C.Q. Tian, S.Q. Shao, Experimental investigation on possibility of
electro-osmotic regeneration for solid desiccant, Appl. Energy 87 (7) (2010)
22662272.
[10] Y. Yao, Shiqing Liu. UltrasonicA New Regeneration Technology For
Dehumidizer. in: The Fourth International Conference on Cryogenics and
Refrigeration, Shanghai, China, April 5th9th, (2008) 984990.
[11] Y. Yao, Shiqing Liu, Weijiang Zhang, Regeneration of silica gel using ultrasonic
under low temperatures, Energy Fuels 23 (1) (2009) 457463.
[12] Y. Yao, Weijiang Zhang, Kun Yang, Shiqing Liu, Beixing He, Theoretical model
on the heat and mass transfer in silica gel packed beds during the regeneration
assisted by high-intensity ultrasound, Int. J. Heat Mass Transfer 55 (2324)
(2012) 71337143.
[13] Y. Yao, Kun Yang, Weijiang Zhang, Shiqing Liu, Parametric Study on Silica Gel
Regeneration by Hot Air Combined with Ultrasonic Field Based on a Semitheoretic Model, Int. J. Therm. Sci. 84 (10) (2014) 89103.
[14] F.D. Natale, R. Nigro, A critical comparison between local heat and mass
transfer coefcients of horizontal cylinders immersed in bubbling uidised
beds, Int. J. Heat Mass Transfer 55 (2526) (2012) 81788183.
[15] T. Defraeye, B. Blocken, J. Carmeliet, Analysis of convective heat and mass
transfer coefcients for convective drying of a porous at plate by conjugate
modelling, Int. J. Heat Mass Transfer 55 (13) (2012) 112124.
[16] J. Kaew-On, S. Wongwises, New proposed two-phase multiplier and
evaporation heat transfer coefcient correlations for R134a owing at low
mass ux in a multiport minichannel, Int. Commun. Heat Mass Transfer 39 (6)
(2012) 853860.

[17] I. Fujita, E. Hihara, Heat and mass transfer coefcients of falling-lm

absorption process, Int. J. Heat Mass Transfer 48 (13) (2005) 27792786.
[18] E. Ghasemi, D.M. McEligot, K.P. Nolan, J. Crepeau, A. Siahpush, R.S. Budwig, A.
Tokuhiro, Effects of adverse and favorable pressure gradients on entropy
generation in a transitional boundary layer region under the inuence of
freestream turbulence, Int. J. Heat Mass Transfer 77 (10) (2014) 475488.
[19] H. Rosali, A. Ishak, R. Nazar, J.H. Merkin, I. Pop, The effect of unsteadiness on
mixed convection boundary-layer stagnation-point ow over a vertical at
surface embedded in a porous medium, Int. J. Heat Mass Transfer 77 (10)
(2014) 147156.
[20] J.B.R. Loureiro, A.P. Silva Freire, Transient thermal boundary layers over rough
surfaces, Int. J. Heat Mass Transfer 71 (4) (2014) 217227.
[21] F. Houshmand, Y. Peles, Heat transfer enhancement with liquidgas ow in
microchannels and the effect of thermal boundary layer, Int. J. Heat Mass
Transfer 70 (3) (2014) 725733.
[22] A. Adrover, A. Pedacchia, Mass transfer through laminar boundary layer in
microchannels with nonuniform cross section: the effect of wall shape and
curvature, Int. J. Heat Mass Transfer 60 (5) (2013) 624631.
[23] Z.Y. Guo, D.Y. Li, B.X. Wang, A novel concept for convective heat transfer
enhancement, Int. J. Heat Mass Transfer 41 (14) (1998) 22212225.
[24] W.Q. Tao, Z.Y. Guo, B.X. Wang, Field synergy principle for enhancing
convective heat transferits extension and numerical verications, Int. J.
Heat Mass Transfer 45 (18) (2002) 38493856.
[25] Z.Y. Guo, W.Q. Tao, R.K. Shah, The eld synergy (coordination) principle and its
applications in enhancing single phase convective heat transfer, Int. J. Heat
Mass Transfer 48 (9) (2005) 17971807.
[26] Y.L. He, W.Q. Tao, F.Q. Song, W. Zhang, Three-dimensional numerical study of
heat transfer characteristics of plain plate n-and-tube heat exchangers from
view point of eld synergy principle, Int. J. Heat Fluid Flow 26 (3) (2005) 459
473.
[27] J.M. Wu, W.Q. Tao, Investigation on laminar convection heat transfer in nand-tube heat exchanger in aligned arrangement with longitudinal vortex
generator from the viewpoint of eld synergy principle, Appl. Therm. Eng. 27
(1415) (2007) 26092617.
[28] Z.X. Li, Z.Y. Guo, The eld synergy principle of convection heat transfer
optimization, Science Press, Beijing, 2010 [in Chinese].
[29] G.M. Chen, C.P. Tso, Field synergy principle analysis on convective heat
transfer in porous medium with uniform heat generation for thermally
developing ow, Int. J. Heat Mass Transfer 55 (1516) (2012) 41394147.
[30] G.Z. Gershuni, Alexander Lyubimov (Thermal Vibrational Convection), John
Wiley & Sons, New York, 1998.
[31] D.E. Crighton, A.P. Dowling, J.E.F fowcs Williams, M. Heckl, F.G. Leppington
(Modern Methods in Analytical Acoustics), Springer, New York, N.Y., 1992.
[32] T. Sai Kumar, M.K. Patle, R.I. Sujith (Characteristics of acoustic standing waves
in packed bed columns), AIChE J. 53 (2) (2007) 297304.
[33] C. Zwikker, C.W. Kosten, Sound Absorbing Materials, Elsevier Pub. Co., New
York, 1949.
[34] D.C. Wilcox, Turbulence Modeling for CFD, third ed., DCW Industries Inc., La
Canada CA, 2006.
[35] D.C. Wilcox, Formulation of the komega Turbulence Model Revisited, AIAA J.
46 (11) (2008) 28232838.
[36] Z.Y. Pan, W.C. Cui, Q.M. Miao, Numerical simulation of vortex-induced
vibration of a circular cylinder at low mass-damping using RANS code, J.
Fluids Struct. 23 (1) (2007) 2337.
[37] G.S. Constantinescu, W.F. Krajewski, C.E. Ozdemir, Simulation of airow
around rain gauges: Comparison of LES with RANS models, Adv. Water Resour.
30 (1) (2007) 4358.
[38] P.A. Costa Rocha, H.H. Barbosa Rocha, F.O. Moura Carneiro, M.E. Vieira da Silva,
A. Valente Bueno (kx SST (shear stress transport) turbulence model
calibration: A case study on a small scale horizontal axis wind turbine),
Energy 65 (2) (2014) 412418.
[39] <http://www.cfd-online.com/Wiki/Standard_k-epsilon_model>.
[40] A.A. Pesaran, A.F. Mills, Moisture transport in silica gel packed bedsI.
Theorectical study, Int. J. Heat Mass Transfer 30 (6) (1987) 10371049.
[41] K.J. Sladek, E.R. Gilliland, R.F. Baddour, Diffusion on surface. II. Correlation of
diffusivities of physically and chemically adsorbed species, Ind. Eng. Chem.
Fundam. 13 (2) (1974) 100105.
[42] A.A. Pesaran, A.F. Mills, Moisture transport in silica gel packed bedsII.
Experimental study, Int. J. Heat Mass Transfer 30 (6) (1987) 10511060.
[43] ASHRAE, 2009 ASHRAE Handbook, in Psychrometrics. 2009. pp. 2021.
[44] M.M. Awad, A.K. Ramzy, A.M. Hamed, M.M. Bekheit, Theoretical and
experimental investigation on the radial ow desiccant dehumidication
bed, Appl. Therm. Eng. 28 (1) (2008) 7585.