See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/264893081

6 Pneumatic and Flash Drying

Article November 2006

DOI: 10.1201/9781420017618.ch16

CITATIONS READS

2 1,684

2 authors, including:

Avi Levy
Ben-Gurion University of the Negev
158 PUBLICATIONS 1,932 CITATIONS

SEE PROFILE

All content following this page was uploaded by Avi Levy on 15 December 2014.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
 Pneumatic and Flash Drying
16 Irene Borde and Avi Levy

CONTENTS

16.1 Introduction ......................................................................................................................................... 397

16.2 Basic Operation Principle and Applications of Flash Dryers .............................................................. 398
16.3 Design of Flash Dryers ........................................................................................................................ 399
16.4 Materials Dried in Flash Dryers .......................................................................................................... 401
16.5 Modeling and Simulations of Pneumatic and Flash Dryers ................................................................ 402
16.5.1 Hydrodynamic Models............................................................................................................ 403
16.5.1.1 Rocha Models......................................................................................................... 404
16.5.1.2 DryPak Model ........................................................................................................ 404
16.5.2 Two-Fluid Model Balance Equations ..................................................................................... 405
16.5.2.1 The Continuity Equations ...................................................................................... 405
16.5.2.2 The Momentum Equations ..................................................................................... 405
16.5.2.3 The Energy Equations ............................................................................................ 406
16.5.2.4 Heat and Mass Transfer ......................................................................................... 406
16.5.3 Case Study............................................................................................................................... 407
16.6 Expected New Developments in Flash Dryers..................................................................................... 409
References ...................................................................................................................................................... 409

16.1 INTRODUCTION and are also known as convective dryers. Pneumatic

or flash dryers may be classified as gassolid transport
Drying is a separation process that converts a wet solid, systems that are characterized by continuous convect-
semisolid, or liquid feedstock into a solid product by ive heat and mass transfer processes. Hot air pro-
evaporation of the liquid into a vapor phase with the duced by indirect heating or direct firing is the most
application of heat. Essential features of the drying common drying medium in these systems. In direct
process are phase change and production of a solid. flash dryers, the gas stream transports the solid par-
Thermal drying is one of the most important unit ticles through the system, and makes direct contact
operations in most industrial sectors. Indeed, it is with the material to be dried. This gas stream (drying
hard to find a product in daily use that has not medium) also supplies the heat required for drying
undergone drying as a stage of its manufacture. and carries away the evaporated moisture. Super-
Drying is an essential operation in the chemical, heated steam can also be used as drying medium
agricultural, biotechnology, food, polymer, ceramic, yielding sometimes to higher efficiencies and often to
pharmaceutical, pulp and paper, and wood process- higher product quality.
ing industries. Drying is extremely energy-intensive The large surface area for heat and mass transfer
and in many cases has important implications as and the high convective heat and mass transfer coef-
the thermal energy needed for drying is obtained by ficients, which take place at these units, result in high
combustion of fossil fuels, leading to emission of drying rates and as a result, high drying capacity. The
carbon dioxide. Well-designed modern drying eq- size of particulates to be dried is usually in the range
uipment with high thermal efficiencies is becoming of 10500 mm. One of the features of these types of
increasingly important. dryers is the relatively short contact time between
One of the most widely used drying systems is the hot air and the particulate materials (0.510 s)
flash drying and is also known as pneumatic drying. at the drying section. Because of this the material
Flash dryers are most commonly direct drying units temperature stays always low in the drying process.

2006 by Taylor & Francis Group, LLC.

16.2 BASIC OPERATION PRINCIPLE AND of the chemical, pharmaceutical, ceramic, gypsum,
APPLICATIONS OF FLASH DRYERS wood, and mining industries. Flash dryers are simple
in construction and have low capital cost and they are
Figure 16.1 shows a simple pneumatic flash drying almost trouble free.
system in which particulate solids are dried during Vertical type of construction, which facilitates in-
transport in a hot gas stream (usually air or combus- stallation in existing buildings, is advantageous for the
tion gases). The simple flash drying system includes flash drying systems. The tube of most flash dryers is of
six basic components: the gas heater, the wet material circular and uniform cross section. In some cases, the
feeder, the drying duct, the separator, exhaust fan, tube may diverge and converge and may have sudden
and a dried product collector. The wet particles are expansions and contractions. The tube may be heated
fed into the hot gas stream sometimes with special through the wall to keep up the temperature of the gas.
mixing devices. The stream flows up the drying tube. In order to shorten the drying time, recirculation
The gas velocity must be greater than the free fall of the material is used. In this case the number of
velocity of the largest particle to be dried. The gas cycles of different particles may be different and as a
velocity in relation to the particle velocity is high. result the residence time of the particles will not be the
Thermal contact between the conveying air and the same. In conclusion the advantages of flash dryers are
solids as mentioned above is usually very short and the following:
therefore flash dryers are most suitable for removal of
external moisture (surface moisture) and are less suit- . Short contact time and parallel flow make
able for removal of internal moisture. At the end of possible to dry thermolabile materials.
the drying process a dust separation arrangement is . The dryer needs only a very small area and can
installed. It must comply with the regulations for be installed outside a building. However, the gas
pollution control. For this purpose cyclone dust sep- cleaning system should be located inside the
arators, fabric filters, electrostatic precipitators, wet building in order to avoid moisture and dust
scrubbers, and fabric filters are used. deposition.
High rates of evaporation in flash dryers are lead- . The dryer is easy to control. The low mate-
ing to low temperatures of the dried material and rial content in the dryer enables equilibrium
indicate that flash dryers are particularly useful for conditions to be reached very quickly.
drying granular, crystalline, pasty, and powdery prod- . Due to small number of moving parts the main-
ucts, etc. Flash dryers are used in various branches tenance cost is low.
. The capital costs are low in comparison with
other types of dryers.
Bag filter Air outlet
. Simultaneous drying and transportation is useful
for materials handling process.

The disadvantages of flash dryers are as follows:

. High efficiency of gas cleaning system is required.

. Because of powder emission, the dryer cannot
be used for toxic materials.
. In some cases this disadvantage can be avoided
using superheated steam as a drying agent.
. For lumped materials difficult to disperse, drying
Dry is impossible to carry out in this apparatus.
product . There is a risk of fire and explosion, so care
must be taken to avoid flammability limits in
the dryer.
. In general, especially when recirculation is ap-
plied, not all material particles have the same
Wet material inlet residence time in the dryer [1].

Feeder
In order to achieve efficient pneumatic drying pro-
Air inlet
cess, the air velocity should be as low as possible to
Heater
achieve materials transport, the mass flow rate of the
FIGURE 16.1 Simple flash drying system. gas should be the minimum necessary to achieve the

2006 by Taylor & Francis Group, LLC.

specified drying rate, the temperature of the hot gas The feed system has to be carefully chosen and
should be as high as possible without exceeding limits designed in order to supply the wet material into the
imposed by the thermal sensitivity of the solids or dryer at the required rate. Typical feed systems are
safety considerations, and the construction of the shown in Figure 16.2 [3]. Metering and feed elements,
dryer should allow to achieve thermal equilibrium sometimes with mixing devices arranged upstream
between the gas and the solid [2]. lead the wet product into the flash dryer. For free-
flowing powdery solids, a screw feeder or a rotary
valve may be used effectively. Pasty or sticky mater-
16.3 DESIGN OF FLASH DRYERS ials need to be preconditioned by blending them with
dried product using single- or twin-shaft paddle
The materials dried in flash dryers have different blender and then dispersed mechanically using a
properties and each product requires specific design kicker mill or one of the several other designs of
solutions. It depends on the initial and final re- rotating disperser [4].
quired moisture, temperature sensitivity, size and The selection of gassolid separators is based
shape of the particles, etc. Finally, each product to mainly on the material characteristics, required de-
be dried requires an optimum solution of the problems gree of separation, concentration of solids, moisture
involved (efficiency and product quality). content of solids, environmental regulations, and cost.
Design procedure of dryers has to find: In utilization are mainly: gravity separators; different
types of cyclones with different efficiencies; fabric fil-
. Dimensions of flash dryer ters, and wet scrubbers. Usually a combination of
. Choose the type and amount of drying agent separation units is applied.
. Requirement of energy supply There are different possibilities of modifications of
. Inlet and outlet parameters for the drying agent simple flash dryers. For instance in a simple flash
and material to be dried (temperature, moisture dryer a rapid decrease of drying force along the tube
content, velocity, etc.) takes place. The design of a dryer with internal pipe
through which hot drying agent is flowing leads to an
Basically, the design consists of execution of the fol- increase of the driving force. Another possibility to
lowing steps: increase the drying force is to heat up the tube of the
dryer through the wall.
. Heat balance In order to decrease the dryer height and increase
. Mass balance the drying time, the material is recirculated. In Fig-
. Momentum balance ure 16.3 a two-stage system is presented [3]. The solid
. Determination of heat and mass transfer co- particles, after passing through the first stage, which
efficients has the form of a vertical tube, are separated in the
upper part of the dryer, fall down, and are directed to
By design of the whole drying system, care must be the second stage. Wet material is supplied to the first
taken of the gas-heating unit, the material feed sec- stage by a special feeding system. Each drying stage in
tion, the particle separation section, and the product the system is equipped with its own heat generator.
collection system. The drying agent from the second drying stage is fed

1 6 1 6 1 6

3 2 7
2
4

(a) (b) (c)

FIGURE 16.2 Typical feed systems for pneumatic flash dryers. 1, Wet product bin; 2, metering; 3, recirculated product; 4,
mixer; 5, sling; 6, flash dryer tube; 7, disc feeder. (From Flash Dryer, Deutsche Babcock, Babcock-BSH GMBH, 1998. With
permission.)

2006 by Taylor & Francis Group, LLC.

 11

4
5 6 7

10 9

1 8

3 3
2 2
4

FIGURE 16.3 Two-stage pneumatic dryer with vapor utilization. 1, Wet product; 2, primary air inlet; 3, heat generator;
4, fan; 5, dryer tube; 6, cyclones; 7, cyclone separator; 8, predried product discharge; 9, dried product discharge; 10, vapor
return line; 11, waste gas stack. (From Flash Dryer, Deutsche Babcock, Babcock-BSH GMBH, 1998. With permission.)

back to the first drying stage. This system is particu- the material is recirculated until it is dried to the
larly efficient if the drying agent is superheated steam. required humidity. For instance high-temperature
Separation is by means of cyclone separators. short-time ring dryers are used in the food industry
A second drying stage can also be used as a cool- to expand the starch cell structure in potatoes or
ing stage. Two-stage flash dryers can be used for carrots to give a rigid porous structure, which enhances
products difficult to dry, e.g., for methylcellulose. conventional drying rates [5].
The number of cycles of different particles in two- Figure 16.4 presents a spin-flash dryer that can be
stage systems may be different and as a result the utilized for some special applications. As described in
resistance time of the particles will not be the same. Ref. [4] the spin-flash dryer is basically a mechanically
For longer resistance times the duct can be formed agitated fluidized bed device for very short resi-
into a continuous loop (ring dryers). In these systems dence times. Hence it is targeted for surface moisture

Exhaust air

Fan

Bag filter

Drying chamber

Wet feed

Orifice

Feed hopper
and agitator Dried solids

Solids feeder Inlet air

Air heater
Rotor
Hot air plenum
Annular air inlet
FIGURE 16.4 Spin-flash dryer. (From Devahastin, S. (Ed.), Mujumdars Practical Guide To Industrial DryingPrinciples,
Equipment And New Developments, Exergex Corporation, Montreal, Canada, 2000. With permission.)

2006 by Taylor & Francis Group, LLC.

removal. As can be seen in Figure 16.4, a rotor,
placed at the bottom of the chamber, is used to dis-
perse the feed, which falls into the agitated fluidized 4
bed by gravity. Hot drying air enters the chamber
tangentially and spirals upward, carrying and drying
the dispersed particles. The exhaust air containing the
dried powder is entering into a separation device,
which separates the powder from the exhaust air. 2 3
Heavier wet particles remain within the drying cham-
ber for a longer time and are broken up by the rotor.
Thus only dried fine powder can escape to the gas 5
separation system. This type of dryer can be a re-
placement for the more expensive spray dryer (which
needs more thermal energy because the feed is wetter
due to the pumpability requirements and also expen- 1
sive because of the need for an atomizer). It is suited
for drying sludge, pulps, pastes, filter cakes, high
viscosity liquids, without the use of an atomizer. Nu- 6
merous materials have been dried successfully in such 7
units at capacities up to 10 tons/h. The spin-flash
dryer units are more expensive than the conventional
flash or fluidized bed dryers. Care must be taken to FIGURE 16.5 Flash dryer as a drying stage. 1, Wet product
ensure that there is no danger of product accumulation feed; 2, flash dryer; 3, cyclones; 4, fabric filter; 5, indirectly
heated rotary calciner; 6, cooler, final product discharge.
on the walls due to stickiness.
(From Flash Dryer, Deutsche Babcock, Babcock-BSH
Flash dryers using superheated steam as drying GMBH, 1998. With permission.)
medium instead of air have some advantages such as
no fire or explosion risk and higher efficiency (if
exhaust steam is utilized elsewhere in the process). It
is well known that in air-drying units the latent heat than normal like special milk or egg powders and
in exhaust gases is difficult to recover. Sometimes the potato granules [5].
quality of the dried product in superheated steam In Figure 16.5 a flash dryer is used as a drying
dryers is superior in comparison to air-drying units. stage in a complex system with a rotary calcining
The limitations of using superheated steam as drying unit [3]. The drying agent in the flash dryer is indir-
medium are that the system itself and the operation ectly heated in the calcining unit. Such arrangement
of the system are more complex. Leaks are pro- provides sufficient heat utilization and is used by
hibited as noncondensables cause problems with Babcock BSH for instance, for catalyst compounds
energy recovery by condensation or compression of and other products.
the exhaust steam. The feeding and discharge pro-
cess must not allow infiltration of air and start-up 16.4 MATERIALS DRIED IN FLASH DRYERS
and shutdown processes are more complex than for
air dryers. Flash dryers are suitable for drying granular, pow-
As mentioned by Devahastin [4] more recently dery, pasty, and crystalline products. As mentioned
flash dryers consisting of inert media have been above the residence time of the particles in the dryer is
employed at pilot scales to dry slurries and suspen- very short that leads to the fact that only products
sions, which are sprayed onto them. The particles are containing surface moisture can be dried in flash
coated thinly by the slurry and dried rapidly as a thin dryers. Drying of heat-sensitive products in this type
film. Due to particleparticle interactions, particle of dryers is very useful. Flash dryers are successfully
collisions, and shrinkage of the film in the drying used in the chemical, food, pharmaceutical, mining,
process, the powder from the slurry is produced. ceramic, and wood industries. Some of the materials
Flash dryers can be used as a drying stage in more dried in flash dryers as described by Kisakurek [6] are:
complex systems for instance as a predrying stage to a Magnesium sulfate, magnesium carbonate, copper
fluidized bed dryer (batch or continuous) or fluidized sulfate, dicalcium phosphate, ammonium sulfate and
bed cooler, spray dryer, drum dryer, etc. In the food phosphate, calcium carbonate and phosphate, and
industry flash dryers are often used after spray drying boric and adipic acids are common examples of chem-
to produce foods that have a lower moisture content icals and by-products. Antibiotics, salt, blood clot,

2006 by Taylor & Francis Group, LLC.

bonemeal, bread crumbs, cornstarch, corn gluten, et al. [13]. Pan and Wypych [12] employed a modified
casein, gravy powder, soup powder, vegetable pro- version of Equation 16.1 by expressing the solids
tein, spent tea, wheat starch, soybean protein, meat pressure drop as a function of the gas pressure drop
residue, and flour are examples of food products. multiplied by the solid loading ratio and a corrected
Cement, aniline dyes, blowing agents, chlorinated friction factor as follows:
rubber, coal dust, copper oxide, gypsum, iron oxide,
and silica gel catalyst are typical by-products and Dp (1 a)Dpg (16:2)
minerals that can be dried in a very efficient way
in flash dryers. Flash dryers are widely used in the
plastic and polymer industries. L1
Dpg 4f r U2 (16:3)
D2 g g

16.5 MODELING AND SIMULATIONS OF

ls m
_s
PNEUMATIC AND FLASH DRYERS a (16:4)
4f m
_g
Mathematical modeling is a very important aspect in
drying technology, allowing the engineer to choose A similar approach was adopted by Mason et al. [13]
suitable operating conditions for the chosen method and Bradley et al. [15] for estimating the pressure
of drying and if necessary apply scale-up procedures drop caused by bends in the pneumatic transport
[7]. It should be kept in mind that the developed system.
mathematical model should be experimentally valid- In order to estimate the moisture content of the
ated in order to use it as a design tool. Reliable particle at the dryer outlet, two additional common
mathematical modeling for conveying of various assumptions are needed, namely isothermal flow and
powders in a dilute phase pneumatic conveying sys- particles exit temperature similar to the gas tempera-
tems was developed and validated during the last ture. Based on these assumptions, various macro-
three decades [811]. In a dilute phase flow, com- scopic mass and energy balance equations can be
monly referred as a suspension flow, the transport solved [16,17].
velocity is sufficient to ensure that the majority of The second approach is based on theoretical and
the particles are suspended in the conveying gas. mathematical modeling for gasparticle flows. Three
Since the particles in pneumatic and flash dryers are types of theoretical approaches can be used for mod-
conveying in a suspension mode of flow, the various eling the gasparticle flows in the pneumatic dryer,
models, which were developed for pneumatic trans- namely two-fluid theory [18], Eulerian granular [19],
port systems, were extended to model the flow in pneu- and discrete element method [20,21]. Both the two-
matic and flash dryers by including heat and mass fluid theory and the Eulerian granular theory are
transfer between the particles and the conveying gas. based on macroscopic balance equations of mass,
In general, two approaches can be used for mod- momentum, and energy for both the gas and the
eling the flow through pneumatic and flash dryers. solid phases. It is assumed that both phases are
The first approach is based on empirical correla- occupying any point (x, y, z) of the computational
tions for specific dryer and dried products. In this domain with its own volume fraction. The solid phase
approach, a variety of semiempirical correlations is considered as a pseudofluid. The main difference
[1214] for estimating the pressure drop have been between these theories is that the Eulerian granular
proposed for gassolids flow in pipes. Frequently method employs the kinetic theory of rarify gases to
these models consider the total pressure drop as the model the granular phase properties, such as pressure,
sum of gas and solids pressure drop components: temperature, viscosity, etc., whereas the two-fluid the-
ory uses macroscopic correlations to model similar
Dp Dpg Dps (16:1) properties for the solid phase. It should be noted
that traditionally, the two-fluid theory was widely
This type of relationship is usually employed in the used to model dilute phase flow whereas the Eulerian
analysis of experimental data, where the total pres- granular was used to simulate both dense and dilute
sure drop is measured and the gas pressure drop phase flows. Unlike these theories, the discrete elem-
component is evaluated by assuming that only gas is ent method is an EulerianLagrangian approach, in
flowing in the pipe. A correlation may then be derived which the gas phase is assumed as the continuous
for the solids pressure drop component. Examples of phase, which occupies every point in the computa-
this type of approach are the work of Muschelknautz tional domain, and the solid particles are occupying
and Wojahn [9], Pan and Wypych [12], and Mason discrete points in the computational domain. As a

2006 by Taylor & Francis Group, LLC.

consequence, mass, momentum, and energy balance of the numerical calculations were partially compared
equation should be solved for each particle within the with experimental data and the influence of the fric-
computational domain. This method is able to take tion factor on the pressure, gas and particle velocities,
into account various types of particleparticle and voidage and residence time of particle along the axis
wallparticle interactions from the basic dynamic ap- of the apparatus have been presented.
proach and gasparticle interactions from the basic Blasco and Alvarez [26] and Alvarez and Blasco
hydrodynamic models for the flow of a single particle [27] considered the application of flash drying to
through a conveying phase. Hence, there is no need to moisture removal of fish and soya meals. Heat, mo-
develop or to use macroscopic modeling for the trans- mentum, and mass balance equations were formu-
port of heat and mass from the solid phase to the lated. The model was solved numerically with
conveying gas. This modeling needs large amount of appropriate coefficients of convective heat and mass
memory and CPU time in order to solve real prob- transfer. Dilute phase transport of homogeneous ra-
lems, which include millions of particles. As a result, dial monosize particle distribution was considered.
no full-scale three-dimensional problem has been The conveying superheated steam was assumed to
solved yet by using the discrete element method. be an ideal gas. The initial period for heating the
particles, during which condensation takes place,
16.5.1 HYDRODYNAMIC MODELS was neglected. Using the film theory [28], the effect
of the mass transfer on the heat transfer coefficient
Many researches adopted one of the above-mentioned was considered. A variable diffusivity model was util-
approaches and modified it to include various aspects ized for the prediction of the drying rate during the
of the pneumatic drying process. Andrieu and Bressat postcritical drying period. Using a pulse technique
[16] presented a simple model for pneumatic drying of under isothermal conditions, the empirical param-
PVC particles. Their model was based on elementary eters of the variable diffusivity model were experi-
momentum, heat, and mass transfer between the fluid mentally determined. The predictions of the model
and the particles. In order to simplify their model, were then compared with their experimental data
they assumed that the flow is unidirectional, the rela- and good agreement was presented.
tive velocity is a function of the buoyancy and drag Kemp et al. [29] presented a theoretical model for
forces, solid temperature is uniform and equal to the particle motion, heat and mass transfer, and drying
evaporation temperature and that evaporation of free rates in vertical tubular pneumatic conveying dryers.
water occurs in a constant rate period. Based on their The model was one dimensional and it took into
simplifying assumptions, six balance equations were account particlewall interaction, feed agglomeration
written for six unknowns, namely relative velocity, air effects, and the effect of particle shape on the drag
humidity, solid moisture content, equilibrium humid- factor. The flow pattern at the inlet, near the feed point,
ity, and both solid and fluid temperatures. The model was neglected, i.e., fully developed flow. Kemp and
was then solved numerically and satisfactory agree- Oakley [17] extended this model and employed it for
ment with their experimental results was obtained. simulating co- and countercurrent dispersion-type
Similar model was presented by Tanthapanichakoon dryers. Equations for particle motion, heat and mass
and Srivotanai [22]. Their model was solved numeric- transfer, heat and mass balance, and local gas condi-
ally and compared to their experimental data. Their tions were solved simultaneously over a small one-
comparison between the experimental data and their dimensional increment along the dryer. Using the
model predictions showed large scattering for the gas RanzMarshal and modified Weber heat transfer cor-
temperature and absolute humidity. However, their relations, the moisture content of the particles was
comparisons for the solid temperature and the water underestimated. Similar observations were obtained
content were failed. by Baeyens et al. [30] and Levy and Borde [31]. Since
Mindziul and Kmiec [2325] investigated the aero- the heat transfer correlations were obtained for a
dynamics of the gassolid flow in a pneumatic flash single particle, it is not therefore surprising that the
dryer. Their mathematical model was based on the proximity of other particles in the conveying system
continuity equation for both the gas and the solid reduces the heat and mass transfer rates. In order to
phase and momentum equations for the solid phase overcome this problem, Kemp and Oakley [17] ap-
and the solidgas mixture. Heat and mass transfer plied a fitting mode procedure to achieve good agree-
were neglected. Although the drying apparatus was ments between their numerical simulations and the
composed of three elements with varying cross- experimental data.
sectional area, one-dimensional model was solved. Silva and Correa [32] used DryPak for simulating
The effect of various empirical correlations for solid- the drying of sand in a pneumatic dryer. Their pre-
wall friction factor has been investigated. The results dictions were compared with the experimental results

2006 by Taylor & Francis Group, LLC.

and the two models of Rocha [33]. The main differ- A two-stage drying process was implemented. In the
ences between the DryPak and Rocha mathematical first drying stage, heat transfer controls evaporation
models and their assumptions are given below. from the saturated outer surface of the particle to the
surrounding gas. At the second stage, the particles
16.5.1.1 Rocha [33] Models were assumed to have a wet core and a dry outer
crust; the evaporation process of the liquid from a
The basic difference between both models is related to particle assumed to be governed by diffusion through
conservation equation of momentum. In the first the particle crust and by convection into the gas
model (Model a), the momentum conservation equa- medium. As evaporation proceeds, the wet core
tion for the fluid as a mixture of fluid and particles shrinks whereas the particle dries. The drying process
was solved, whereas in the second model (Model b), is assumed to stop when the moisture content of a
conservation of momentum for each phase was particle falls to a predefined value or when the particle
solved. For both models, the following assumptions reaches the exit of the pneumatic dryer. The devel-
were considered: steady-state one-dimensional flow; oped model was solved numerically and two operat-
nonhygroscopic spherical particles; no shrinkage dur- ing conditions, adiabatic and given pneumatic dryer
ing drying; plug flow for both phases; uniform prop- wall temperature, were simulated. Comparison be-
erties and geometry at a pipe cross section; work done tween the prediction of the numerical models of
between the phases was neglected. Based on the Rocha and DryPak [34], which were presented by
above-mentioned assumptions, mass, momentum, Silva and Correa [32], with the prediction of our
and energy balance equations were formulated for numerical simulation revealed better agreements
the mixture and the solid phase. The correlation of with DryPak than with the models of Rocha. The
Ranz and Marshall was used for calculating the heat results of the developed model were also compared
and the mass transfer coefficients. Although Rocha with experimental results of Baeyens et al. [30] and
[33] introduced a heat transfer term from the fluid Rocha [33].
phase to the ambient, no specific model was presented. Rocha and Paixao [36] presented a pseudo two-
dimensional mathematical model for a vertical pneu-
16.5.1.2 DryPak Model [34] matic dryer. Their model was based on the two-fluid
approach. Axial and radial profiles were considered
All the assumptions, which were considered by Rocha for gas and solid velocity, water content, porosity,
[33], were also considered in DryPak balance equa- temperatures, and pressure. The balance equations
tions, with the exception that in DryPak adiabatic were solved numerically using a finite difference
flow conditions were assumed. Other differences method and the distributions of the flow field charac-
were in the way of calculating the area for heat and teristics were presented. This model was not validated
mass transfer and modification of the heat and mass with experimental results.
transfer coefficients. DryPak used the Frossling equa- Silva and Nerba [37] also used the two-fluid ap-
tion for calculating the Nusselt number. Different proach and presented a mathematical model of drying
types of heat and mass transfer analogies were pre- in cyclone. Slip condition of particles on the wall,
sented and Ackermann correction was used to include particlewall heat transfer, and particles shrinkage
the influence of mass transfer on the heat transfer were considered. The mathematical model considered
coefficient. It should be noted that unlike Rocha [33] a steady state, incompressible, two-dimensional, axi-
models, DryPak can take into account particles symmetric, turbulent gassolids flow. The gravity
shrinkage; internal resistance to heat and mass trans- force effect on the particles was neglected. The par-
fer; and moisture content profile inside the particle ticles were assumed to be spherical and distributed in
could be obtained, although it was not used in the a layer of uniform concentration on the cyclone wall
study of Silva and Correa [32]. Silva and Correa [32] and a very small concentration in the central flow.
concluded that predictions of DryPak produced The discretized balance equations were solved by the
better agreements with experimental data than the SIMPLE algorithm [38]. Silva and Nerba [37] com-
numerical results of Rocha [33]. pared the predictions of their numerical simulations
Levy and Borde [35] adopted the two-fluid theory with experimental results and claimed that the most
for modeling the flow of particulate materials through influencing parameters on the predictions are the par-
pneumatic dryer. The model was solved for a one- ticle slip conditions and the material shrinkage during
dimensional steady-state condition and was applied the drying process.
to the drying process of wet PVC particles in a large- Unlike the above-mentioned models, Fyhr and
scale pneumatic dryer and to the drying process of Rasmuson [39,40] and Cartaxo and Rocha [41] used
wet sand in a laboratory-scale pneumatic dryer. an EulerianLagrangian approach, in which the gas

2006 by Taylor & Francis Group, LLC.

phase is assumed as the continuous phase and the velocity vector of the k-phase. The mass source term
solid particles are occupying discrete points in the of the k-phase is Sk and to maintain the conservation
computational domain. As a consequence, mass, mo- of mass Sg  Ss.
mentum, and energy balance equations were solved
for each particle within the computational domain.
16.5.2.2 The Momentum Equations
Fyhr and Rasmuson [39,40] presented a two-
dimensional model for superheated steam drying of The momentum equation for the k-phase is given by
wood chips in a pneumatic conveying dryer. One-
dimensional plug flow was assumed. Steady-state sin- @
gle particle flow and drying models were solved inter- (k rk Vk ) r  [k rk Vk Vk ]
@t
actively. Particleparticle interactions were neglected.
r  [k  k ]r(k Pk ) k rk g Mkj Sk Vs
The irregular movement and the nonsphericity shape
of the wood chips were accounted by measuring (16:6)
drag and heat transfer coefficients. The predictions
of the temperature and the pressure profiles as well as Generally, the variation of the solids density, which
the final moisture content of bark chips agreed well composes the wet particle, is less than few percent. As
with experimental results. Based on the model valid- a consequence, constant solid density may be as-
ation, a parametric study was conducted. The calcu- sumed to simplify the model. Thus by using the mix-
lation showed that the drying rate varies in a very ture theory the density of the dispersed phase can be
complex manner through the dryer. The internal re- expressed as
sistance to mass transfer becomes a dominant factor
in the drying of less permeable wood chips. As the 1 j 1j
(16:7)
particle size was increased, the heat transfer rate de- rs rw rsi
creases and the residence time increases. Hence, they
concluded that less permeable wood species or larger where j is the liquid mass ratio in the particle and rw
chips size leads to longer dryer in order to obtain the and rsi are the densities of the liquid and the solid,
desired final moisture content. which compose the particle. Another common as-
Another two-dimensional, discrete element model sumption is that the conveying air behaves as an
was presented by Cartaxo and Rocha [41]. In this ideal gas. Thus the gas pressuredensity relation is
work, only the dynamic phenomenon was investi- described by
gated (i.e., heat and mass transfer between the phases
were not considered). Thus the influence of the mo- Pg rg R Tg (16:8)
mentum coupling between the discrete particles and
the conveying air on the air radial velocity and the The effective normal stress of the solids phase may be
mass concentration profiles were presented. An ob- written as a sum of the sheared gas pressure and the
ject-oriented numerical model was developed to simu- solids contact stress. Thus the effective normal stress
late the conveying of large spherical particles (3 mm) of the solid phase is described by
through 9.14 m vertical tube with 7.62 cm bore size.
Ps rg R Tg sn0 (s =s0 )1=b (16:9)
16.5.2 TWO-FLUID MODEL BALANCE EQUATIONS
where sn0 is a particular value of the solids contact
In the following, the Eulerian governing equations for stress for solid volume fraction s0 and b is a constant
the pneumatic drying process are presented in their coefficient over a given range of contact pressure
three-dimensional form. These equations are based [42,43].
on the two-fluid approach [18]. The interphase momentum transfer is represented
by
16.5.2.1 The Continuity Equations
Mkj K(Vk  Vj ) Pk rk (16:10)
The continuity equation for the k-phase is given by
The interphase momentum transfer term can be de-
@ rived from correlation developed to model fluidi-
(k rk ) r  (k rk Vk ) Sk (16:5) zation processes, since the range of solids
@t
concentrations experienced in pneumatic transport
where k-phase can be the gas or the solid phase, k, rk, systems is similar. This form has been employed by
and Vk are the volume fraction, the density, and the Patel and Cross [44] for modeling gassolid fluidized

2006 by Taylor & Francis Group, LLC.

beds. For solids concentrations greater than 0.2, the interphase enthalpy (i.e., the enthalpy of the vapor at
interphase friction coefficient, K, may be computed the temperature of the solid particles).
by using the Ergun [45] equation
16.5.2.4 Heat and Mass Transfer
2 m 1
K 150 s 2 1:75s rg jVg  Vs j (16:11) The rate of energy transfer between the phases is
g ds ds
usually expressed as a function of the temperature
For solids concentrations less than 0.2, the inter- difference between that of the conveying gas and
phase friction coefficient is usually based upon the that of the particle surface (i.e., Tg Tss). Hence, the
aerodynamic force on particle as follows: interphase heat exchange between the phases can be
calculated by
 
3s 1
K (CD g2:65 ) g rg jVg  Vs j (16:12) 6s
2ds 2 Qgs hgs (Tg  Tss ) (16:16)
ds
where the single particle drag coefficient, CD, is given
The convective heat transfer coefficient, hgs, is calcu-
by [46]
lated from the Nusselt number, Nu, which is defined as
 
24
CD max (1 0:15Re0:687 ), 0:44 (16:13) hgs ds
Re Nu F (Re, Pr) (16:17)
kg
and is modified to take account of multiparticle ef-
and is often expressed as a function of the Reynolds
fects using the method of Richardson and Zaki [47].
number (Re) and Prandtl number (Pr), which are
The particle Reynolds number is given by
defined as follows:
rg ds (g jVg  Vs j)
Re (16:14) rg jur jds mg cpg
mg Re ; Pr (16:18)
mg kg
The turbulent stresses, tk, in the momentum equa-
Note that kg, mg, and cpg are the thermal conductivity,
tions for the k-phase might be calculated by using
the viscosity, and heat capacity of the gas phase,
the Boussinesq turbulent-viscosity model [8] for both
respectively. Table 16.1 presents common empirical
phases or by applying a model of a Newtonian fluid
correlations that have been used in the literature to
for the gas phase and a granular shear stress for the
calculate the heat transfer coefficient in gasparticle
solid phase [19].
flows.
The friction forces between each phase and the
The mass transfer source term per unit volume can
pipe wall can be modeled by adding a source term
be obtained by multiplying the evaporation rate from
to the phase momentum equation for those control
_ s, by the total number of particles
a single particle, m
volumes adjacent to the pipe wall [11,23,39].
in the control volume:

16.5.2.3 The Energy Equations 6

Sg m
_s (16:19)
The conservation of energy in multiphase application pds3
can be written as an enthalpy equation for each
phase: The drying model for a single wet particle and slurry
droplet is based upon a two-stage drying process
@ [48,49]. In the first drying period, the gas phase resist-
( k r k h k ) r  [ k r k V k h k ] ance controls the evaporation rate. Similar to heat
@t
@ pk transfer, this resistance is between the gas and the
k  k : rVk  r  qk Qk Qkj Sk hkj wet envelope of the particle. This may be expressed by
@t
(16:15)  
Mw pvo Mw pvg
_s
m hm pds2  (16:20)
In this equation, hk is the specific enthalpy of the k- R Tss R Tg
phase, qk is the heat flux, Qk is a heat source term (due
to chemical reaction or radiation), Qkj is the inter- where hm is the convective mass transfer coefficient,
phase heat exchange between the phases, and hkj is the Mw is the molecular weight of the water, R is the

2006 by Taylor & Francis Group, LLC.

TABLE 16.1
Empirical Correlations for Heat Transfer Coefficient in GasParticle Flows
Modified RanzMarshall 2 0:6Re0:5 Pr0:333 Developed for a single wet particle, taking into account
Nu
correlation [48] (1 B)0:7 the resistance of the liquid vapors around the particle
cpv (Tg  Td ) to the heat transfer by Spalding number, B. cpv denotes
B the heat capacity of the liquid vapors in the gas phase
Hfg
and Hfg is the latent heat of evaporation for the fluid
Modified RanzMarshall Nu 2 (0.5 Re0.5 0.06 Re0.8)Pr0.333 Takes into account turbulent boundary layer around
correlation [29] the particle
Gamson correlation [30] Nu 1.06Re0.59Pr0.33 Developed for a fluidized bed dryer
De Brandt correlation [30] Nu 0.16Re1.3Pr0.67 Developed for a pneumatic dryer
Baeyens et al. [30] correlation Nu 0.15Re Developed for a large-scale pneumatic dryer

mg
universal gas constant, and pvo and pvg are the partial Sc (16:23)
pressures of the water vapor at the particle crust and r g Dn
the gas phase, respectively.
The second drying period starts at a critical solid- The correlations for the Nusselt number (see Table
to-liquid mass ratio, jcr, which is obtained from a 16.1) can be used to calculate the Sherwood number,
minimum void fraction, i.e., the porosity of the par- Sh, and the mass transfer coefficient hm by replacing
ticles, (typically varied between 0.05 and 0.25). Dur- the Prandtl number, Pr, with the Schmidt number, Sc.
ing the second period of the drying process, a dry During the first drying stage the diameter of the
crust starts to form, which causes a second resistance particle, ds, shrinks due to evaporation from the outer
to mass and heat transfer. Thus, the wet particle surface to the surrounding gas. Thus the diameter of
consists of a dry crust surrounding a wet core. This the wet particle can be calculated by
resistance is governed by a diffusion process, which
occurs between the outside diameter of the particle, d 2
ds m
_s (16:24)
dso, and the diameter of the wet core, dsi. Assuming dx rw us pds2
that the particle is not shrinking during the second
drying period, the outside diameter of the particle At the second drying stage, the evaporation process
remains constant and the diameter of the wet core of liquid from a particle is assumed to be governed by
decreases. The equation for the evaporation rate diffusion through the particle crust and convection
from a single particle is expressed as a Stephan-type into the gas medium. As evaporation proceeds, the
diffusion rule [50] wet core shrinks as the particle dries. In general, both
0 1 the particle outer diameter and the wet core diameter
can be shrinked, which may deform the particles
dsi  dso 2pDv p B p  psat C shape and size. In order to simplify the model, it
_s
m nB C
dso dsi R Tave @ RTss pvg Tss A was assumed that the particles outer diameter re-
p 2 M
m_s
hm pdso w Tg mains constant during the second drying period.
(16:21) Thus, only the change of the wet core diameter, dsi,
was considered:
where Dv is the diffusion coefficient, psat is the satur-
ation pressure inside the wet core, and Tave is the
d 2
average temperature of the particle. dsi m
_s (16:25)
In analogy to the heat transfer coefficient, the mass dx rw us pdsi2
transfer coefficient hm is calculated from the Sher-
wood number, Sh, which is equivalent to Nusselt
number, Nu 16.5.3 CASE STUDY

hm ds The two-fluid model has been used for modeling the

Sh F (Re, Sc) (16:22) flow of particulate materials through pneumatic
Dn
dryer. The model was solved numerically for a one-
and is often expressed as a function of the Reynolds dimensional steady-state condition and was applied
number, Re, and the Schmidt number, Sc, which is to the drying process of wet sand in a pneumatic
equivalent to Prandtl number, Pr, and is defined by dryer. A two-stage drying process was implemented.

2006 by Taylor & Francis Group, LLC.

The predictions of the numerical simulations were relative error was 0.35 and 0.03%, respectively. When
compared with the experimental results of Rocha adiabatic flow condition was simulated, the gas tem-
[33] (presented by Silva and Correa [32]) that were perature was overestimated and the maximum rela-
obtained in a 4-m high pneumatic dryer with diameter tive error was 5%. The predictions of the numerical
of 5.25 cm. In this study, 380-mm sand particles hav- simulations for the gas humidity (Figure 16.6c) were
ing density of 2622 kg/m3 and mass flow rate of very good for both simulation conditions, i.e., adia-
4.74103 kg/s were dried with 3.947  102 kg/s air batic and known wall temperature. The maximum
mass flow rate. The comparison between the predic- relative errors were 1.2 and 0.70%, respectively. The
tions of the numerical simulations and the experi- predictions of the numerical simulations for the par-
mental data for changes of gas temperature, solid ticle moisture content (Figure 16.6d) were also very
temperature, gas humidity, and particles moisture good for both simulation conditions, i.e., adiabatic
content with length under adiabatic and known wall and known wall temperature, although only two ex-
temperature operating conditions is presented in Fig- perimental data were given. The maximum relative
ure 16.6ad, respectively. When known wall tempera- errors were about 20% at the pipe outlet (i.e., when
ture operating conditions were simulated, it was the particles moisture contents were approxima-
assumed that in average the pipe wall temperature is tely zero). A comparison between the prediction of
just about the outlet air temperature, and it is falling the numerical models of Rocha and DryPak [34],
linearly from 360 K at the inlet to 354 K at the outlet. which were presented by Silva and Correa [32], with
In these figures the circle symbols represent the ex- the prediction of the numerical simulation revealed
perimental data that were published by Silva and better agreements with DryPak than with the models
Correa [32] and the two solid lines represent the pre- of Rocha.
dictions of the numerical simulations for the adiabatic It should be pointed out that the two-fluid ap-
and known wall temperature operating conditions. proach, as described and demonstrated in the previ-
It is clearly seen that the numerical model pre- ous sections, was widely used and validated for
dicted the gas and the solid temperature profiles (Figure various types of pneumatic conveying systems and
16.6a and b) very well when known wall temperature pneumatic flash dryers. Nevertheless, it is only one
operating conditions were simulated. The maximum of various approaches that can be adopted. Section

120 60
110
Solid temperature (8C)
Gas temperature (C)

100 50
90
40
80
70 30
60 Abiabatic flow model Conducting wall model
50 Conducting wall model 20 Abiabatic flow model
Rocha (1988) exp. data Rocha (1988) exp. data
40 DryPak DryPak
Rocha (1988) a
10 Rocha (1988) a
30 Rocha (1988) b Rocha (1988) b
20 0
0 1 2 3 4 5 0 1 2 3 4 5
(a) Dryer length [m] (b) Dryer length (m)

5.6 Abiabatic flow model 6

Moisture content % (kg/kg)

Conducting wall model

5.5 Conducting wall model Abiabatic flow model
Gas humidity % (kg/kg)

Rocha (1988) exp. data 5 Rocha (1988) exp. data

5.4 DryPak DryPak
5.3 Rocha (1988) a Rocha (1988) a
4 Rocha (1988) b
5.2 Rocha (1988) b
5.1 3
5
4.9 2
4.8
1
4.7
4.6 0
0 1 2 3 4 5 0 1 2 3 4 5
(c) Dryer length (m) (d) Dryer length (m)

FIGURE 16.6 Comparison between the predictions of the pneumatic drying model, DryPak model [35], Rocha [33] models
and the experimental data for changes of (a) gas temperature, (b) solid temperature, (c) gas humidity, and (d) particles
moisture content with length under adiabatic and known wall temperature operating conditions.

2006 by Taylor & Francis Group, LLC.

16.5.1 presents a summary of the various studies that 12. Pan, R. and Wypych, P., Bend pressure drop in pneu-
were conducted on pneumatic flash dryers and the matic conveying of fly ash, in Proceedings of the Powder
various mathematical models that have been adopted and Bulk Solids Conference, Chicago, USA, 1992,
by them. pp. 349360.
13. Mason, D.J., Marjanovic, P., and Levy, A., The influence
of bends on the performance of pneumatic conveying
systems, Advanced Powder Technology, 95, 714, 1998.
16.6 EXPECTED NEW DEVELOPMENTS 14. Hyder, L.M., Bradley, M.S.A., Reed, A.R., and Het-
IN FLASH DRYERS tiaratchi, K., An investigation into the effect of particle
size on straight pipe pressure gradients in lean phase
These developments are expected in the fields of: pneumatic conveying, Powder Technology, 112(3), 235
243, 2000.
. Enhancement of product quality 15. Bradley, M.S.A., Farnish, R.J., Hyder, L.M., and
. Increased efficiency by heat recovery Reed, A.R., A novel analytical model for the acceler-
. Applications of heat pumps in drying systems ation of particles following bends in pneumatic convey-
. Efficient combined processes ing systems, in Handbook of Conveying and Handling of
. Development of models for overall systems that Particulate Solids (Levy, A. and Kalman, H. Eds.), Else-
vier, Amsterdam, 2001.
would include mechanical dewatering, drying,
16. Andrieu, J. and Bressat, R., Experimental and theoret-
heat recovery, powder collection, cooling stages, ical study of a pneumatic dryer, in Proceedings of the
etc. in order to study interactions between vari- Third International Drying Symposium, Vol. 2, 1982,
ous parts of the system and develop improved pp. 1019.
design procedures 17. Kemp, I.C. and Oakley, D.E., Simulation and scale-up
of pneumatic conveying and cascading rotary dryers,
Drying Technology, 15(68), 16991710, 1997.
REFERENCES 18. Bowen, R.M., Theory of mixtures, in Continuum Phys-
ics (Eringen A.C. Ed.), Academic Press, New York,
1. Strumio, C. and Kudra, T., Drying: Principles, Appli- 1976, pp. 1127.
cations and Design, Gordon & Breach Science Pub- 19. Gidaspow, D., Multiphase Flow and Fluidization, Aca-
lishers, London, UK, 1986. demic Press, New York, 1997.
2. Thorpe, G.R., Pneumatic Conveying Driers, Chemical 20. Cundall, P.A. and Strack, O.D., A discrete numerical
Industry Development, Incorporating CP&E, 1975, model for granular assemblies, Geotechnique, 29, 4765,
pp. 1319. 1979.
3. Flash Dryer, Deutsche Babcock, Babcock-BSH 21. Tsuji, Y., Tanaka, T., and Ishida, T., Lagrangian nu-
GMBH, 1998. merical simulation of plug flow of cohesionless particle
4. Devahastin, S. (Ed.), Mujumdars Practical Guide To in a horizontal pipe, Powder Technology, 71, 239, 1992.
Industrial DryingPrinciples, Equipment And New De- 22. Tanthapanichakoon, W. and Srivotanai, C., Analysis
velopments, Exergex Corporation, Montreal, Canada, and simulation of an industrial flash dryer in a Thai
2000. Manioc starch plant, in Drying96 Proceedings of the
5. Fellows, P.J., Food Processing TechnologyPrinciples 10th International Drying Symposium, Vol. A, 1996,
and Practice, Woodhead Publishing, Cambridge, UK, pp. 373380.
1997. 23. Mindziul, Z. and Kmiec, A., Modelling gassolid flow
6. Kisakurek, B., Flash drying, in Handbook of Industrial in a pneumatic-flash dryer, in Drying96 Proceedings of
Drying, 2nd ed. (Mujumdar, A.S. Ed.), Marcel Dekker, the 10th International Drying Symposium, Vol. A, 1996,
New York, 1995. pp. 275282.
7. Kemp, I.C., Scale-up of pneumatic dryers, Drying Tech- 24. Mindziul, Z. and Kmiec, A., Modelling gassolid flow
nology, 12(12), 279297, 1994. in a pneumatic-flash dryer, Drying Technology, 15(68),
8. Boothroyd, R.G., Flowing GasSolids Suspensions, 17111720, 1997.
Chapman & Hall, London, 1971. 25. Kmiec, A., Analysis of the gassolid flow in a riser
9. Muschelknautz, E. and Wojahn, H., Auslegung pneu- reactor, Applied Mechanics and Engineering, 2(1), 133
matischer forderanlagen, Chemie-Ing.-Techn., 223235, 152, 1997.
46(6), 1974. 26. Blasco, R. and Alvarez, P.I., Flash drying of fish meals
10. Molerus, O., Overviewpneumatic transport of solids, with superheated steam: isothermal process, Drying
Powder Technology, 88, 309321, 1996. Technology, 17(45), 775790, 1999.
11. Levy, A., Mooney, T., Marjanovic, P., and Mason, 27. Alvarez, P.I. and Blasco, R., Pneumatic drying of
D.J., A comparison of analytical and numerical models meals: application of the variable diffusivity model,
for gassolid flow through straight pipe of different Drying Technology, 17(45), 791808, 1999.
inclinations with experimental data, Powder Technol- 28. Bird, R., Stewart, E., and Lightfoot, N., Transport
ogy, 93, 253260, 1997. Phenomena, John Wiley & Sons, New York, 1960.

2006 by Taylor & Francis Group, LLC.

 29. Kemp, I.C., Bahu, R.E., and Pasley, H.S., Model de- port phenomena, AIChE Journal, 43(11), 28892902,
velopment and experimental studies of vertical pneu- 1997a.
matic conveying dryers, Drying Technology, 12(6), 40. Fyhr, C. and Rasmuson, A., Steam drying of wood
13231340, 1994. chips in pneumatic conveying dryers, Drying Technol-
30. Baeyens, J., van Gauwbergen, D., and Vinckier, I., ogy, 15(68), 17751785, 1997b.
Pneumatic drying: the use of large-scale experimental 41. Cartaxo, S.J.M. and Rocha, S.C.S., Object-oriented
data in a design procedure, Powder Technology, 83, simulation of pneumatic conveyingapplication to a
139148, 1995. turbulent flow, Brazilian Journal of Chemical Engineer-
31. Levy, A. and Borde, I., Steady-state one-dimensional ing, 16(4), 329337, 1999.
flow for a pneumatic dryer, Chemical Engineering and 42. Johanson, J.R., Two-phase-flow effects in solids pro-
Proceedings, 38, 121130, 1999. cessing and handling, Chemical Engineering, January !,
32. Silva, M.A. and Correa, J.L.G., Using DryPak to simu- 7786, 1979.
late drying process, in Drying98 Proceedings of the 11th 43. Johanson, J.R. and Cox, B.D., Practical solutions to
International Drying Symposium, Vol. A, 1998, pp. 303 fine powder handling, Powder Handling and Processing,
310. 1(1), 8387, 1989.
33. Rocha, S.C.S., Contribution to the Study of Pneumatic 44. Patel, M.K. and Cross, M., The modelling of fluidised
Drying: Simulation and Influence of GasParticle Heat beds for ore reduction, in Numerical Methods in Lam-
Transfer Coefficient, Ph.D. thesis, Sao Paulo Univer- inar and Turbulent Flow, Pineridge Press, Swansea, UK,
sity, Sao Paulo, 1988. 1989, p. 2051.
34. Pakowski, Z., DryPak v.3. Program for Psychometric 45. Ergun, S., Fluid flow through packed columns, Chem-
and Drying Computation, 1996. ical Engineering Progress, 48(2), 8994, 1952.
35. Levy, A. and Borde, I., Two-fluids model for pneumatic 46. Clift, R., Grace, J., and Weber, M.E., Bubbles, Drops
drying of particulate materials, in 12th International and Particles, Academic Press, New York, 1987.
Drying Symposium, Nordwijkerhout, Netherlands, 47. Richardson, J.F. and Zaki, W.N., Sedimentation and
2000. fluidization: Part I, Transactions of the Institute of
36. Rocha, S.C.S. and Paixao, A.E.A., Pseudo two-dimen- Chemical Engineering, 32, 3553, 1954.
sional model for a pneumatic dryer, in Drying96 Pro- 48. Levi-Hevroni, D., Levy, A., and Borde, I., Mathemat-
ceedings of the 10th International Drying Symposium, ical modelling of drying of liquid/solid slurries in steady
Vol. A, 1996, pp. 340348. state one dimensional flow, Drying Technology, 13(57),
37. Silva, M.A. and Nerba, S.A., Numerical simulation of 11871201, 1995.
drying in a cyclone, Drying Technology, 15(68), 1731 49. Levy, A., Mason, D.J., Borde, I., and Levi-Hevroni, D.,
1741, 1997. Drying of wet solids particles in a steady-state one-
38. Patankar, S.V., Numerical Heat Transfer and Fluid dimensional flow, Powder Technology, 85, 1523, 1998.
Flow, Hemisphere Publishing, New York, 1980. 50. Abuaf, N. and Staub, F.W., Drying of liquidsolid slurry
39. Fyhr, C. and Rasmuson, A., Mathematical model of a droplets, in Drying86 Proceedings of the Fifth Inter-
pneumatic conveying dryer, fluid mechanics and trans- national Drying Symposium, Vol. 1, 1987, pp. 227248.

2006 by Taylor & Francis Group, LLC.

View publication stats