This article was downloaded by:

On: 21 January 2011

Access details: Access Details: Free Access
Publisher Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-
41 Mortimer Street, London W1T 3JH, UK

Drying Technology
Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t713597247

Numerical Study of Two-Stage Horizontal Spray Dryers Using

Computational Fluid Dynamics
Lixin Huanga; A. S. Mujumdarb
a
Research Institute of Chemical Industry of Forest Products, Nanjing, Jiangsu Province, P.R. China b
Mechanical Engineering Department, National University of Singapore, Singapore

To cite this Article Huang, Lixin and Mujumdar, A. S.(2006) 'Numerical Study of Two-Stage Horizontal Spray Dryers
Using Computational Fluid Dynamics', Drying Technology, 24: 6, 727 — 733
To link to this Article: DOI: 10.1080/07373930600685046
URL: http://dx.doi.org/10.1080/07373930600685046

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.
 Drying Technology, 24: 727–733, 2006
Copyright # 2006 Taylor & Francis Group, LLC
ISSN: 0737-3937 print/1532-2300 online
DOI: 10.1080/07373930600685046

Numerical Study of Two-Stage Horizontal Spray Dryers

Using Computational Fluid Dynamics
Lixin Huang1 and A. S. Mujumdar2
1
Research Institute of Chemical Industry of Forest Products, Nanjing, Jiangsu Province, P.R. China
2
Mechanical Engineering Department, National University of Singapore, Singapore

chamber as well as overall chamber volume effectiveness

This article presents the findings of a numerical simulation model and drying performance. Another possible choice for the
of the spray-drying process in a two-stage horizontal chamber drying chamber may be horizontal configuration.
design with the aid of a computational fluid dynamic (CFD) model. A two-dimensional horizontal spray dryer configuration
The model describes heat, mass, and momentum transfer between was first suggested as an alternative to the conventional
two phases; namely, a continuous gas phase and a discrete phase
of droplets (or particles), using the finite volume method. In this vertical design[3] to reduce difficulties in scale-up. Kwamya
study, a new two-dimensional horizontal spray dryer (HSD) geometry and Mujumdar[4] first presented a mathematical model to
is considered as a pilot study into the spray-drying process in this evaluate and demonstrate the feasibility of this novel con-
Downloaded At: 03:18 21 January 2011

novel chamber configuration. The tested model is able to predict cept for hot air and superheated steam drying operations.
some important features of the spray-drying process, such as air However, their model considered a uniform flow and single
flow patterns indicating recirculation zones and particle trajectory
plots. Some performance parameters for spray drying, such as the particle motion and drying.
rate of evaporation, average volumetric heat and mass transfer Later, Cakaloz et al.[5] presented a simplified model for
rates, etc., are calculated and discussed. This two-stage drying pro- a horizontal spray dryer (6.0 m
3.0 m
3.0 m). Their
cess especially applicable for the horizontal spray dryer (HSD) simulation results on moisture history of an a-amylase
model is investigated and modeled. The bottom wall of the HSD droplet agreed reasonably well with their experiment
is assumed to be a shallow fluid bed used for second stage drying.
In this article, the fluid bed drying conditions are changed and com- results. When the computational fluid dynamic (CFD)
pared. The drying within the fluid bed itself is not modeled in this model described in this thesis was used to simulate such a
study, however. It is shown that the particle residence time is higher horizontal spray dryer, it was observed that the flow pat-
when the fluid bed is included. The drying performance of this two- tern in Cakaloz et al’s.[5] design is not optimal for spray
stage horizontal spray dryer is expected to be better than that of drying since the main air inlet is located in a corner of
a single-stage dryer.
the chamber. With this arrangement, the spray is more
likely to hit the top wall.
Keywords Air flow pattern; Heat and mass transfer; Horizontal
spray dryer; Particle trajectories At least two U.S. companies appear to market horizontal
spray dryers designed mainly for heat-sensitive food and
pharmaceutical products.[6,7] Both manufacturers use inte-
INTRODUCTION gral bag filters located within the chamber for product
Cylinder-on-cone or simple cylindrical chamber geome- collection. Any fine powder present in the exhaust air
tries with co-current flow of hot air and spray are the most stream is returned to the dryer. Both designs use multi-jet
popular spray dryer chamber geometries in use today. atomizers with indirect-fired burners. In one of the systems,
Huang et al.[1,2] investigated other possible chamber the product is collected at the bottom of the horizontal
geometries, i.e., pure conical, hour-glass geometry, and lan- chamber. The vendor claims substantial energy savings
tern, as the spray-drying chamber. It was shown that the relative to the traditional vertical spray dryer due to lower
cylinder-on-cone is not the only geometry that could be electrical load for a given capacity. The manufacturers offer
used for the drying chamber from the perspective of drying the dryer with stainless steel construction. Commercial
performance and particle residence times in the drying applications listed by the vendors include dryers for egg,
albumin, whole egg powder, cheese powder, skim milk,
Correspondence: Lixin Huang, Research Institute of Chemical whey protein, etc. Among the advantages claimed are mini-
Industry of Forest Products, No. 16 Suojin 5th village, Nanjing, mum floor space and building height, low installation and
Jiangsu Province, P.R. China 210042; E-mail: l_x_huang@163.
com maintenance cost, and ability to automatically control the

727
 728 HUANG AND MUJUMDAR

(droplets) in a spray dryer occupies a very small fraction of

the total volume, the Eulerian model is not the most suit-
able method to simulate flows in spray dryer. In the
Lagrangian model, only the carrier phase (air) is treated
as a continuum. The dispersed phase (droplets or particles)
is subject to the Lagrangian equation. In this model, two-
way couplings of heat, mass, and momentum transfer are
assumed. In order to take into account the effect of the dis-
persed phase on the continuum phase by transfer of heat,
mass, and momentum, the Particle Source in Cell (PSI-
Cell) method[9] is commonly used.
In order to simplify the CFD computation, the main
FIG. 1. Schematic flow chart of the horizontal spray dryer. assumptions in our CFD model are made as follows:
 The dilute two-phase flow, i.e., compared with the
dryer operation. Unfortunately, details of the structure of volume of the drying medium, the particle volume
the drying chamber, the atomizer layout, droplet size distri- occupies a very small fraction. In this study, the frac-
butions, and product quality, are not available in the public tion of particles volume in air volume is less than 5%.
 No agglomeration or collision and break-up
domain. The common flow chart is shown in Figure 1.
The objective of this article is to use a computational fluid among the droplets.
 No temperature gradient within the droplet or
dynamic model that has been validated with data for conven-
tional vertical spray dyers, to arrive at several alternative con- particle.
ceptual designs for a horizontal spray dryer (HSD).[8] It is
Downloaded At: 03:18 21 January 2011

The CFD model describes heat, mass, and momentum

shown that the use of a mathematical model can speed up transfer between two phases, namely a continuous gas
the development process for an innovative product or process phase and a discrete phase of droplets (or particles), using
with very substantial savings in time, effort, and cost. Their the finite volume method. The main governing equations
works also showed that although a reasonable design of a are simply discussed in the following part. Readers may
horizontal spray drying chamber for a single-stage spray dryer refer to the literature on modeling of the turbulence and
is one in which the air outlet is located at the end and side of the CFD analysis of spray dryers for more details.[1,2,10,11]
chamber with no bag filter installed in situ, a single-stage
horizontal spray dryer studied still has to be improved. Hence, Governing Equations for Continuous Phase (Air)
a new two-stage horizontal, i.e., combination of spray dryer
and fluid bed dryer, is considered for further investigation. Continuity equation
It is expected to have a longer particle residence time and @qui
¼ Mm ð1Þ
better drying performance than that of a single-stage horizon- @xi
tal spray dryer. Momentum equations
Results presented here could be useful to spray dryer
@ðqui uj Þ @P @
vendors or users interested in developing HSD as a viable ¼ þ
@xi @xj @xi
alternative to the popular vertical designs for specific appli- 
 
cations. It is not suggested as a replacement for the conven- @ui @uj 0 0

l þ  qui uj þ MF ð2Þ
tional vertical designs, which have a smaller foot print. @xj @xi
Energy equation
METHODOLOGY  
The flow in a spray dryer is turbulent and two-phase @ðqcp ui TÞ @ @T
¼ k  qu0i T 0 þ Mh ð3Þ
(gas and droplets or gas and particles). Hence, we @xi @xi @xi
employed a flow model in which one of the phases (dro-
plets or particles) is dispersed throughout the other phase Governing Equations for Dispersed Phase (Droplets
(gas or air). Two-way coupling is assumed between the or Particles)
two phases, i.e., there is finite transfer of momentum, heat, The Euler-Lagrangian approach is used to obtain the
and mass between the two phases. particle trajectories by solving the force balance for the
There are two distinct methods commonly used to particles as follows
model two-phase flows. In the Eulerian method, each
dupi 18l Re qg  q
phase is treated as a continuum. This is appropriate when ¼ CD ðui  upi Þ þ gi þ Fxi ð4Þ
the two phases are at similar fractions. Because one phase dt qp dp 24 qg
 NUMERICAL STUDY OF TWO-STAGE HORIZONTAL SPRAY DRYERS 729

Mass and Heat Transfer Between Two Phases

The rate of vaporization is governed by gradient
diffusion; i.e.,
Ni ¼ kc ðCi;s  Ci;1 Þ ð5Þ
The heat transfer between the droplet and the hot gas is
updated according to the heat balance; i.e.,
dTp dmp
m p cp ¼ hAp ðT1  Tp Þ þ hfg ð6Þ
dt dt
The second period, called droplet boiling, is applied to
predict the convective boiling of a discrete phase droplet
when the temperature of the droplet has reached the boil-
FIG. 2. Two-stage horizontal spray dryer chamber (all dimensions
ing point and while the mass of the droplet exceeds the in mm).
non-volatile fraction. The boiling rate equation is
applied:[12]
that this base case is a better conceptual design for a hori-
dðdp Þ 4k1  pffiffiffiffiffiffiffiffi zontal spray dryer. Therefore, the proposed design with the
¼ 1 þ 0:23 Red
dt qp cp;1 dp relevant dimensions is similar to the base one.
  In order to carry out the investigation of two-stage hori-
cp;1 ðT1  Tp Þ

ln 1 þ ð7Þ zontal spray dryer, two new cases, i.e., Cases B and C, were
hfg
constructed. Here, the air outlet is located at the side of one
Downloaded At: 03:18 21 January 2011

end of the chamber. The product collector is located exter-

Turbulence Models nal to the HSD. This two-dimensional horizontal geometry
comprises a main drying air inlet and an outlet as well as a
Because there is no swirling flow in the drying chamber,
fluid bed. The fluid bed is located at the bottom of the hori-
the standard k-e model[13] is expected to be appropriate
zontal chamber and modeled as a series of 32 air inlets of
choice for simulating such a flow.[14] Therefore, the stan-
2.5-mm diameter spaced at regular intervals along the bot-
dard k-e model was used in the simulation. The transport
tom wall. As such, the bottom wall of the chamber can be
equations for the standard k-e model are for transport of
treated as comprising of alternating intervals of chamber
the turbulence kinetic energy, k, and its dissipation rate,
walls and velocity inlets. The rationale for having a fluid
e. They are

  bed is to facilitate the second-stage drying of larger or wet-
@ðqkÞ @ðqkui Þ @ l @k ter particles by providing a gentle influx of drying air along
þ ¼ lþ t
@t @xi @xj rk @xj the bottom wall. The chamber with the important dimen-
þ Gk þ Gb  qe  YM þ Sk ð8Þ sions is shown in Figure 2.
and

  Boundary Conditions
@ðqeÞ @ðqeui Þ @ lt @e For the numerical studies reported here, two-dimen-
þ ¼ lþ
@t @xi @xj re @xj sional models are used to test the two-stage horizontal
e e2 spray drying designs. In these 2D models, triangular grids
þ C1e ðGk þ C3e Gb Þ  C2e q þ Se are used for the CFD simulation of the horizontal drying
k k
ð9Þ chamber. The velocity and temperature profiles obtained
were checked and found to be very similar for different
The turbulence viscosity, lt , is computed as follows: mesh sizes (0.0075, 0.015, and 0.03 m) used in cases A, B,
lt ¼ qCl ðk2 =eÞ. The model constants values are set as: and C. The results are verified to be grid independent.
C1e ¼ 1:44; C2e ¼ 1:92; Cl ¼ 0:09; rk ¼ 1:0; re ¼ 1:3.[15] Thus, a mesh size of 0.015 mm was selected for all tested
geometries.
CHAMBER DESIGN AND BOUNDARY CONDITIONS Each spray dryer has its own specific operating set of
parameters. Since Case A is selected as the base case, most
Design of Horizontal Chamber Geometry of the boundary conditions for Case A are also used for
In order to compare different HSD’s designs, the design Cases B and C. The main operating conditions for Case
of one-stage horizontal spray drying by Huang and A, i.e., inlet air flow rate and temperature, wall boundary
Mujumdar[8] is used as the base case because it is found conditions, nozzle flow rate, etc., are maintained the same
 730 HUANG AND MUJUMDAR

in the numerical simulations. The specified mass flow rate The differences from the base case are described as fol-
of drying air is 0.3 kg=s. Temperature of air at inlet is set at lows:
523 K. The ambient has a relative humidity of 70% and
Case B: The main inlet mass flow rate of air is equal to half
temperature of 25C. The pressure at the air outlet is set
that of in Case A. The remaining half is used as supply
at 20 Pa; i.e., a fan is assumed to draw air out from the
to the fluid bed. This air passes through uniformly dis-
drying chamber. In practice, the drying chamber is usually
tributed holes at the bottom wall of the chamber.
maintained at 0 Pa or at a slight negative pressure.[16]
Case C: In the same way, the mass flow rate used in Case A
A suspension with a 30% solids content is specified as
is separated equally into two parts. One part flows
the wet slurry to be dried. The spray rate is 0.03 kg=s at a
through the main inlet and another is used for the fluid
temperature of 300 K. In order to compute and compare
bed. However, here a fully fluidized region is assumed
the results, only the solids concentration of the material
to exist, implying that the uniform air enters the cham-
is defined. Other physical properties are assumed to be
ber through the entire length of the bottom wall.
the same as those of pure water. Temperature- and
humidity-dependent physical properties of air are used in
the simulation. RESULTS AND DISCUSSION
A pressure atomizer is used for atomization. It is A good spray dryer design should consider, among others,
installed centrally on the left side of the drying chamber, the following factors governed by the fluid dynamics:
as shown in Figures 1 and 2. The spray droplet size distri- 1. Air flow pattern should lead to maximum drying effec-
bution is assumed to be the same as that of the Kieviet tiveness, e.g., good utilization of the chamber volume.
et al.’s[17] measurement. The range of droplet diameters is 2. Enough residence time for particle should be provided
from 10 to 138 mm. The mean droplet diameter is 70 mm. to evaporate all moisture.
In the current simulation, they are modeled using the 3. Deposits along the wall (except the bottom wall) should
Downloaded At: 03:18 21 January 2011

Rosin-Rammler distribution with spread parameter of be avoided or minimized.

2.05. The droplet size distribution is shown in Figure 3. 4. If a bag-house is used, the air flow should be uniform
The spray velocity is 50 m=s and the spray angle is assumed through every bag.
to be 60. 5. No backflow should take place at the air outlet, since it
In the simulations, the interactions among droplets, i.e., will lead to moisture condensation from the cool moist
agglomeration and collision of droplets as well as droplet air in the chamber.
breakup, are neglected. The interaction between droplet
and chamber wall is simply assumed to result in the
Air Flow Patterns
‘‘escape’’ condition. The heat transfer coefficient from the
wall to environment is assumed to be 3.5 W=m2  K, which The air flow patterns for Cases A, B, and C are shown in
is same as that used by Kieviet et al.[17]. Figure 4. It is observed that Case B extends the upper recir-
The two-equation standard k-e turbulence model is used culation zone found in Case A. This recirculation region is
in this work as well since there is no swirl in the flow. moved forward as well. But the below recirculation zone is
According to the FLUENT manual, when the Reynolds almost same as that in Case A.
number is less than 40,000, it is reasonable to assume the On the other hand, Case C shows that both the upper and
turbulent intensity to be less than 4%. In this work, this lower recirculation zones are reduced in size considerably.
value is used as the intensity at the air inlet boundary. The main flow is mostly straight from the inlet to the outlet.
Finally, no recirculation zones are found in the down-
stream flows in all three cases. There is no backflow found
at the outlet, as well.

Temperature Contours
The temperature contours for Cases A, B, and C are
shown in Figure 5. It is seen that most of the chamber vol-
ume in Case B is more or less at the same temperature. It
implies that the evaporation in this region is very small.
Although there is hot air entering at the bottom wall with
the same temperature as the main inlet, it seems that this
does not play a significant role on the evaporation of the
spray. When compared with that in Case A, it also does
not show any advantage from the fluid bed addition at
FIG. 3. Droplet size distribution produced by a pressure nozzle. the bottom.
 NUMERICAL STUDY OF TWO-STAGE HORIZONTAL SPRAY DRYERS 731
Downloaded At: 03:18 21 January 2011

FIG. 4. Streamline patterns for Cases A, B, and C. FIG. 6. Particle trajectories for Cases A, B, and C.

On the other hand, if the fluid bed is assumed to have

been formed already, the temperature contours for Case
C are quite different from those in Cases A and B. It sug-
gests that it may have good evaporation performance.

Particle Deposits and Residence Time Distributions

The particle trajectories for Cases A, B, and C are shown
in Figure 6. It is observed that there are some particles
‘‘sucked’’ by the upper recirculation zone in Case B. They
are recirculated in this region. It is also seen that significant
deposit on the bottom wall occurs even in the presence of

TABLE 1
The particle deposits and residence time for Cases A,
B, and C
Cases A B C
Percentage of Outlet 38.5 31 50.5
droplets hitting Left-bottom-wall 5.5 17 17
the wall (%) Top wall 1 0.5 3.5
Bottom-wall 51.5 41 22.5
Mean particle Outlet 5.7 3.7 9
residence Left-bottom-wall 3.7 3.7 2
time (s) Top wall 2.7 4 4
Bottom-wall 1.9 1.7 3
FIG. 5. Temperature contours for Cases A, B, and C.
 732 HUANG AND MUJUMDAR

TABLE 2
The overall drying performance for Cases A, B, and C
Cases A B C
Total spray rate (kg=s) 0.03 0.03 0.03
Solid content (%) 30% 30% 30%
Suspension temperature (K) 300 300 300
Real evaporation rate (kg=s) 1.71
102 0.63
102 1.86
102
Total heat consumption (W) 43,739 20,028 45,932
Percentage of evaporation rate 81.4 30.5 88.6
Energy consumption per unit evaporation (kJ=kg) 2521 3179 2470
Volume of the chamber (m3) 2.3 2.3 2.3
Evaporation rate per unit volume (kg=m3  s) 7.9
103 2.74
103 8.1
103

the fluid bed. It is probably because the air flow just passes Table 2 also shows that a better horizontal spray dryer
through the small opening and the velocity is too low. design may be obtained by adding a fluid bed at the
Results for Case C show that a number of particles pass bottom horizontal wall, but this fluid bed must be prop-
through the chamber directly. There are a few particles erly designed to make the particles well fluidized in this
deposited in the front section of the bottom wall with region.
almost no particles deposit at the rear. However, the
deposit problem at the main entrance is still serious.
CONCLUSIONS
Table 1 summarizes the particle deposit conditions and
Downloaded At: 03:18 21 January 2011

In this article, conceptual designs of one-stage and two-

relative residence times for Cases A, B and C. It leads to
stage horizontal spray dryers are tested using a computa-
the same conclusion as that found from Figure 4. For
tional fluid dynamic model. Simulation results show that
example, almost half of the particles directly go out of
good chamber design should give good drying perform-
the chamber in Case C compared to only one-third for
ance. A reasonable design of a horizontal spray drying
Case B. Half of the particle deposit on the bottom wall
chamber for a single-stage spray dryer is one in which the
in Case A but only one-fifth for Case C. Since Case C still
air outlet is located at the end and side of the chamber with
has some particles deposits at the bottom wall, it implies
no bag filter installed in situ; i.e., Case A.
that the fluid bed at this condition is still not strong enough
Simulations in a two-stage horizontal spray dryer
to get all particles in the fluidized state at this region.
showed that the effects of incorporating a fluid bed for
Considering the particle residence time for these three test
the purpose of two-stage drying within the chamber were
cases, it is noted that Case C gives a longer particle residence
investigated. The predicted results show that incorporating
time than the other two cases. Although there are some par-
a fluid bed can significantly affect the drying performance
ticles recirculated in Case B; they may recirculate and hit the
of a spray dryer. But a poor design of the fluid bed could
main entrance wall due to the small size of the main entrance.
lead to poor drying results, even worse than that in a sin-
All these findings suggest that Case B may not offer a better
gle-stage spray dryer.
drying performance than that of the other two cases.
Finally, from the findings in Table 1, it may be con-
cluded that good drying performance may not be obtained NOMENCLATURE
if the fluid bed design is not good. For example, Case B is Ap The surface area of the droplet (m2)
not a good design for fluid bed here. CD Drag coefficient
Cg Moisture concentration in the bulk gas
(mol=m3)
Comparison of the Overall Drying Performance for the Cs Moisture concentration at the droplet surface
Three Test Cases (mol=m3)
The overall drying performance for Cases A, B, and C is cp Heat capacity of the droplet (J=kg  K)
summarized in Table 2. It reinforces the findings from the Dm Diffusion coefficient of vapor in the bulk (m2=s)
earlier discussion, i.e., Case B is not a good design. dp Droplet diameter (mm)
Though Case C compares generally well with Case A, it dmp =dt Rate of evaporation (kg=s)
is observed that Case C presents a better drying perform- d Rosin-Rammler mean droplet diameter (mm)
ance, i.e., higher evaporation rate and lower energy con- Gk Production of turbulence kinetic energy due to
sumption per unit evaporation. the mean velocity gradients
 NUMERICAL STUDY OF TWO-STAGE HORIZONTAL SPRAY DRYERS 733

Ge Production of turbulence kinetic energy due to Forestry, for its financial support (Key Laboratary Fund)
buoyancy in this research.
gi Gravity component (m2=s)
h Convective heat transfer coefficient (W=m2  K)
hfg Latent heat (J=kg) REFERENCES
k Turbulent kinetic energy (m2=s2) 1. Huang, L.X.; Kumar, K.; Mujumdar, A.S. Use of computational fluid
kc Mass transfer coefficient (m=s) dynamics to evaluate alternative spray chamber configurations. Dry-
keff Effective thermal conductivity of gas (W=m  K) ing Technology 2003, 21 (3), 385–412.
kg Thermal conductivity of the hot medium 2. Huang, L.X.; Kumar, K.; Mujumdar, A.S. A parametric study of the
gas flow patterns and drying performance of co-current spray dryer:
(W=m  K)
Results of a computational fluid dynamics study. Drying Technology
Mh Source term in energy equation 2003, 21 (6), 957–978.
Mm Source term in continuity equation 3. Mujumdar, A.S. Dryers for particulate solids, slurries and sheet-form
!
MF Source term in momentum equation materials. In Mujumdar’s Practical Guide to Industrial Drying; Deva-
mp Mass of the droplet (kg) hastion, S., Ed.; Exergex Corporation: Montreal, 2000; 37–71.
2 4. Kwamya, K.; and Mujumdar, A.S. M. Eng. Project Report, Depart-
Ni Molar flux of vapor
(mol=m !  s) 
!
qdp j ug  up j
ment of Chemical Engineering, McGill University: Canada, 1984.
Re Reynolds number Re ¼ 5. Cakaloz, T.; Akbaba, H.; Yesugey, E.T.; Periz, A. Drying model for
l
a-amylase in a horizontal spray dryer. Journal of Food Engineering
Sk User-defined source term 1997, 31, 499–510.
Se User-defined source term 6. CERogers Company. http://www.cerogers.com/html/horizontal_
dryer.html, 2004.
T Temperature (K)
7. FES International, Inc. http://www.fesintl.com/htmfil.fld/spray-
Tg Gas temperature (K) dryh.htm, 2004.
t Time (s) 8. Huang, L.X.; Mujumdar, A.S. Development of a new innovative con-
Downloaded At: 03:18 21 January 2011

ui Gas velocity vector (m=s) ceptual design for horizontal spray dryer via mathematical modeling.
up Droplet velocity vector (m=s) Drying Technology 2005, 23 (9–11), 1859–1873.
9. Crowe, C.T. Modeling spray air contact in spray drying systems. In
Advances in Drying, Vol. 1; Mujumdar, A.S., Ed.; Hemisphere:
Greek Letters
New York; 1980, 63–99.
e Energy dissipation rate (m2=s3) 10. Langrish, T.A.G.; Zbicinski, I. The effect of air inlet geometry and
l Viscosity (kg=m  s) spray cone angle on the wall deposition rate in spray dryers. Trans.
lt Turbulent viscosity I Chem. E. 1994, 72 (A), 420–430.
q Gas density (kg=m3) 11. Fluent Manual, Chapter 2, materials; chapter 19, Discrete phase mod-
els; Chapter 20, Turbulence model, www.fluent.com, 2004.
qp Droplet density (kg=m3) 12. Kuo, K.K.Y. Principles of Combustion; John Wiley and Sons:
r Surface tension (N=m) New York, 1896.
rk Turbulent Prandtl number for k 13. Launder, B.E.; Spalding, D.B. The numerical computation of
re Turbulent Prandtl number for e turbulent flows. Computational Methods in Applied Mechanical
Engineering 1974, 3, 269–289.
Subscripts 14. Oakley, D. Scale-up of spray dryers with the aid of computational
fluid dynamics. Drying Technology 1994, 12, 217–233.
C Constant 15. Launder, B.E.; Spalding, D.B. Lectures in Mathematical Models of
g Air Turbulence; Academic Press: London, England, 1972.
p Droplet 16. Masters, K. Spray Drying Handbook, 5th Ed; John Wiley & Sons, Inc.:
New York, 1991, 1–400.
17. Kieviet, F.G.; Raaij, J.V.; Moor, P.P.E.A.D; Kerkhof, P.J.A.M.
ACKNOWLEDGEMENT
Measurement and modelling of the air flow pattern in a pilot-plant
The authors thank Research Institute of Chemical spray dryer. Trans. I. Chem. E. 1997, 75 (A) 321–328.
Industry of Forestry Products, Chinese Academy of