Professional Documents
Culture Documents
Computational Fluid Dynamics Applications in Food Processing
Computational Fluid Dynamics Applications in Food Processing
AND NUTRITION
C. Anandharamakrishnan
Computational
Fluid Dynamics
Applications in
Food Processing
SpringerBriefs in Food, Health, and Nutrition
SpringerBriefs in Food, Health, and Nutrition present concise summaries of cutting
edge research and practical applications across a wide range of topics related to the
field of food science.
Editor-in-Chief
Richard W. Hartel, University of WisconsinMadison, USA
Associate Editor
J. Peter Clark, Consultant to the Process, Industries, USA
John W. Finley, Louisiana State University, USA
David Rodriguez-Lazaro, ITACyL, Spain
David Topping, CSIRO, Australia
Computational Fluid
Dynamics Applications
in Food Processing
13
C. Anandharamakrishnan
CSIR-Central Food Technological
Research Institute
Mysore
India
vii
Contents
ix
x Contents
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Chapter 1
Computational Fluid Dynamics Applications
in Food Processing
Computational Fluid Dynamics (CFD) has been extensively applied in the food-
processing sector for the design and optimization of equipment such as ovens, spray
dryers, chillers, heat exchangers, etc. Numerous benefits from the implementation
of CFD models have been reported. This book recapitulates the application of CFD
modeling; in particular, design and optimization of spray drying, spray freezing, baking
ovens, high pressure processing, retorts processing and also biological systems. CFD
modeling is often used in spray drying operations, as it is very difficult and expensive
to obtain measurements of airflow, temperature, particle size, and humidity within
the drying chamber. CFD can be a useful tool for predicting the gas flow pattern and
particle histories such as temperature, velocity, residence time, and impact position
during spray drying. CFD modeling of a baking oven provides constructive informa-
tion about temperature and airflow pattern throughout the baking chamber to enhance
heat transfer, and in turn, final product quality. CFD modeling also helps in designing
the ovens for rapid bread baking. CFD modeling can be used in retort processing of
canned solid and liquid foods for understanding and optimization of the heat trans-
fer processes. CFD can be used to numerically model the dynamics of gastrointes-
tinal contents during digestion, based on the motor response of the gastrointestinal
(GI) tract and the physicochemical properties of luminal contents. Advanced com-
putational fluid dynamics programs offer a promising technique to characterize the
mechanisms promoting digestion. Furthermore, this book predominantly focuses on
the recent developments in this field, constraints in CFD modeling approaches, their
strengths and limitations, and future applications in food industries.
velocity, temperature and pressure profiles inside the system. Its powerful graph-
ics can be used to show the flow behaviour of fluid with three dimensional (3D)
images (Anderson 1984; Scott and Richardson 1997).
The history of CFD takes us way back to the 1960s, when the aerospace indus-
try integrated this technique into the design, research and development, and manu-
facture of aircraft and jet engines. Around the 1970s, CFD became an acronym for
a combination of physics, numerical mathematics, and, to some extent, computer
sciences employed to simulate fluid flows. However, the applications were mostly
restricted to the two-dimensional (2D) flow models due to the low speed and storage
capacities of the computers. The beginning of CFD was triggered by the availabil-
ity of more powerful mainframes, and the advances in CFD are still tightly coupled
to the evolution of computer technology. Around the mid-1980s, computer predic-
tions of fluid flow have been used routinely in both science and engineering to pro-
duce results. With the advances of numerical methodologies, particularly of implicit
schemes, the solution of flow problems that require real gas modeling became fea-
sible by the end of 1980s. Toward the 1990s, 3D modeling became possible and led
to an upsurge of interest in a great deal of industrial applications. Nowadays, CFD
methodologies are routinely employed in the fields of aircraft, turbo machinery, car,
and ship design. Furthermore, CFD is also applied in meteorology, oceanography,
astrophysics, and also in architecture (Anderson 1984; Shaw 1992; Versteeg and
Malalasekera 1995; Blazek 2001). For more a detailed historical perspective, the
books by Roache (1976) and Tannehill et al. (1997) are highly recommended. Today,
CFD finds extensive usage in basic and applied research, in design of engineering
equipment and in calculation of environmental and geophysical phenomena.
CFD is a numerical technique for the solution of equations governing the flow of
fluids inside defined flow geometry. The flow of any fluid can be described using
the following transport Eqs.(1.11.4) (Bird et al. 1960; Versteeg and Malalasekera
1995; Marshall and Bakker 2002; Fluent 2006). These equations are derived by
considering mass, momentum and energy balances in an element of fluid, result-
ing in a set of partial differential equations. They are completed by adding other
algebraic equations from thermodynamics, such as the equation of state for density
and a constitutive equation to describe the rheology (Fletcher 2000).
The continuity equation describes the rate of change of density at a fixed point
resulting from the divergence in the mass velocity vector v. Equation(1.1) is the
unsteady, three-dimensional, mass conservation or continuity equation for the sim-
plified case of a constant density fluid (incompressible fluid).
1.2 Theory of CFD Modeling 3
v =0 (1.1)
where has the dimension of reciprocal length:
= i+ j+ k (1.2)
x y z
1.2.2Momentum Equation
1.2.3Energy Equation
The first law of thermodynamics states that the rate of change of internal energy
plus kinetic energy is equal to the rate of heat transfer minus the rate of work done
by system. Fluent solves the energy equation in the following form.
(E) + v(E + p) = keff T hj J j + ( v) (1.4)
t
j
where E is the internal (thermal) energy, keff is the effective conductivity (kta + kt,
where kta is thermal conductivity and kt is turbulent thermal conductivity), T is the tem-
perature, is stress tensor, J j is the diffusion flux of species j, and hj is the enthalpy of
species j. The three terms on the right-hand side of the equation represent energy trans-
fer due to conduction, species diffusion and viscous dissipation, respectively.
1.3Turbulence Model
There are two types of flows; namely, laminar and turbulent. Above a certain Reynolds
number, all flows become unstable and exhibit turbulent behaviour. For laminar flow
problems (low Reynolds number), the flows can be solved by conservation equa-
tions. In the case of turbulent flows (high Reynolds number), the computational effort
involved in solving those for all time and length scales is prohibitive. An engineering
4 1 Computational Fluid Dynamics Applications in Food Processing
approach to calculate time-averaged flow fields for turbulent flows will be developed
for solving turbulent flow problems (Marshall and Bakker 2002).
The turbulence models commonly used for simulations are:
(i) Standard k (kturbulence kinetic energy and turbulence dissipation rate)
(ii) Renormalization Group (RNG) k
(iii) Realizable k
(iv) Reynolds Stress Model (RSM)
Three models (standard, RNG, and realizable k) have similar forms, with
transport equations for k and .
Most commercial CFD codes use turbulence models that are based on the split-
ting up of instantaneous quantities into a time-averaged and a fluctuating part by
a process known as Reynolds decomposition. For turbulent flows, the standard
k model (kturbulence kinetic energy and turbulence dissipation rate) is
the most commonly used, because it converges considerably better than Reynolds
stress model (RSM) (Versteeg and Malalasekera 1995), and is given as follows:
t
(k) + kv = + k + Gk (1.5)
t k
2
t
() + v = + + Cl (Gk ) C2 (1.6)
t k k
Gk is the generation of kinetic energy due to the mean velocity gradients. The quan-
tities k and are the turbulent Prandtl numbers for k and , respectively, and C1, C2,
are constant. The turbulent (or eddy) viscosity t is calculated from k and as follows:
k2
t = C (1.7)
The model constants C1, C2, C, k and took the following values (Launder
and Spalding 1972):
C1 = 1.44, C2 = 1.92, C = 0.09, k = 1.0 and = 1.3
For calculating an approximate solution of fluid flow equations, the equations
have to be made discrete. For this, the flow domain is divided into number of con-
trol volumes. This is called a grid, and at each grid cell, approximate solutions for
the NavierStokes and the continuity equations are calculated. Table1.1, summa-
rises the performance of the turbulence models (Zhang et al. 2007).
1.4Reference Frames
Three different reference frames are widely used: the volume of fluid (VOF),
EulerianEulerian (EE) and EulerianLagrangian (EL) models. The volume of fluid
(VOF) model is designed for two or more immiscible fluids (Fig.1.1a) by solving a
1.4 Reference Frames
single set of momentum equations and tracking the volume fraction of each of the
fluids throughout the domain. Because the fluids do not mix, each computational
cell is filled with purely one fluid, purely another fluid, or the interface between
two (or more) fluids. Typical applications include the prediction of jet breakup, the
motion of large bubbles in a liquid, the motion of liquid after a dam break, and the
steady or transient tracking of any liquidgas interface (Fluent user guide).
The other two-phase modeling frames are the EulerianEulerian and the
EulerianLagrangian methods. In the EulerianEulerian frame (Fig.1.1b), the
dispersed phase (droplets) are treated as a continuous (Eulerian) phase, i.e.
there are two Eulerian phases, one for the gas and another for droplets, which
are interacting and interpenetrating continually (Mostafa and Mongia 1987).
Each computational cell contains certain fractions of gas and droplets, and
the transport equations are written in such a way that the volume fractions of
gas and liquid sum to unity. If the computational cell consists of just a single
phase, the transport equations for the two phases revert to the conventional
single-phase system. The advantages of the EulerianEulerian approach are
usually relatively cheap in terms of computational demands for one additional
set of equations, and turbulence can be modeled fairly simply. However, if a
separate set of transport equations is solved for each particle size (single par-
ticle diameter was used for the dispersed phase), then the Eulerian approach
can be expensive. In addition, there is some uncertainty over the most appro-
priate Eulerian diffusion coefficients and heat transfer coefficients. Hence, the
Eulerian approach is best suited to flows with a narrow range of particle sizes
where a high resolution of the particle properties is not needed (Mostafa and
Mongia 1987; Jakobsen et al. 1997).
In the EulerianLagrangian particle tracking approach (Fig.1.1c), the gas
phase is modeled using the standard Eulerian approach described above and the
spray is represented by a number of discrete computational particles. Individual
particles are tracked through the flow domain from their injection point until
they escape the domain in a Lagrangian frame work (Nijdam et al. 2006). The
EulerianLagrangian model has the advantage of being computationally cheaper
than the EulerianEulerian method for a large range of particle sizes. It can also
provide more details of the behaviour and residence times of individual particles
1.4 Reference Frames 7
and can potentially approximate mass and heat transfer more accurately. On the
other hand, the approach can be expensive if a large number of particles have to be
tracked and it is best when the dispersed phase does not exceed 10% by volume
of the mixture in any region (Marshall and Bakker 2002).
In both the EulerianEulerian and the EulerianLagrangian methods, the
exchange of momentum between particles and gas needs to be modeled. This
exchange can consist of several forces such as drag, lift, virtual mass, and wall
forces. Mostafa and Mongia (1987) concluded that the Eulerian approach per-
forms better than Lagrangian method. In contrast, Nijdam et al. (2006) found that
both Eulerian and Lagrangian modeling approaches gave similar predictions for
turbulent droplet dispersion and agglomeration of sprays for a wide range of drop-
let and gas flows. The two models were found to require similar computing times
for a steady axi-symmetric spray. However, the authors preferred the Lagrangian
models because of their wider range of applicability.
1.5CFD Analysis
CFD analysis involves following three main steps. The first step is pre-processing,
which includes problem definition, geometry, meshing (this can usually be done
with the help of a standard CAD program), and generation of a computational
model. The second step is processing, which uses a computer to solve the mathemat-
ical equations of fluid flow. The final step of post-processing is used to evaluate and
visualize the data generated by the CFD analysis (Xia and Sun 2002) and validate
the simulation results with experimental data. Figure1.2 explains the all three steps
of CFD analysis with the example of egg pasteurization.
Although the origins of CFD can be found in the automotive, aerospace and
nuclear industries with a variety of applications in different processing indus-
tries, it is only in recent years that CFD has been applied to food processing (Scott
and Richardson 1997). The applications of CFD in the food industry have been
reviewed by many researchers (Scott and Richardson 1997; Xia and Sun 2002;
Anandharamakrishnan 2003; Norton and Sun 2006). These reviews envisage the
potential of CFD to be used as a tool in predicting the fluid flow, heat and mass
transfer phenomena in the food processes, leading to better equipment design and
process control for the food industry. Figure1.3 depicts the applications of CFD in
various food processing operations.
SPRAY DRYING
CHILLING AND
FREEZE DRYING
REFRIGERATION
CFD
MIXING
Applications BAKING
HEAT
BIOREACTORS
EXCHANGERS
PASTURIZATION/
STERILIZATION
The main application of CFD includes spray drying processes (Langrish and
Fletcher 2001, 2003), baking process (Therdthai et al. 2003; DeVries et al. 1994;
Mills 19981999), refrigerated display cabinets (Cortella et al. 1998), thermal steri-
lization (Datta and Teixeira 1987; Abdul Ghani et al.1999a, b, 2001), pasteurization
of egg (Denys et al. 2003, 2004, 2005), mixing (Sahu et al. 1999; Scott 1977), refrig-
eration (Hu and Sun 1999, 2000; Davey and Pham 1997, 2000; Moureh and Derens
2000; Mariotti et al. 1995), spray freezing (Anandharamakrishnan et al. 2010b),
heating and cooling processes (Wang and Sun 2003), and humidification of cold
storage (Verboven and Nicolai 2008, 2009). The list given above is non-exhaustive,
and for detailed review of CFD applications to food processing, reader may refer else-
where (Sun 2007). CFD has recently found widespread applications in food processing.
In thermal sterilization processes, CFD has found increased use in analyzing the flow
pattern, temperature distribution, and more importantly, the shape and position of the
slowest heating zone (SHZ), since it is very difficult to estimate these parameters using
experiments.
1.7Nomenclature
Spray drying is a well-established method for converting liquid feed materials into
a dry powder form. It is widely used to produce powdered food, healthcare and
pharmaceutical products. Normally, spray dryer comes at the end-point of the pro-
cessing line, as it is an important step to control the final product quality. It has
some advantages, such as rapid drying rates, a wide range of operating tempera-
tures and short residence times. In spray drying operations, CFD simulation tools
are now often used, because measurements of air flow, temperature, particle size
and humidity within the drying chamber are very difficult and expensive to obtain
in large-scale dryer (Kuriakose and Anandharamakrishnan 2010).
2.1.1Atomization
1
Feed Liquid
Furnace
Air
2
Exhaust gas
4
Spray- drying 3
Cyclone
chamber
separator
Product
2.1.2SprayAir Contact
During sprayair contact, droplets usually meet hot air in the spraying chamber
either in co-current flow or counter-current flow. In co-current flow, the product
and drying medium passes through the dryer in the same direction. In this arrange-
ment, the atomized droplets entering the dryer are in contact with the hot inlet air,
but their temperature is kept low due to a high rate of evaporation taking place,
and is approximately at the wet-bulb temperature. As the droplets pass through
the dryer, the moisture content decreases, the air temperature also decreases, and
so the particle temperature does not rise substantially as the particle dries and the
effect of evaporation cooling diminishes (Mujumdar 1987). The temperature of
the products leaving the dryer is slightly lower than the exhaust air temperature.
This co-current configuration is therefore very suitable for the drying of heat-
sensitive materials. The advantages of the co-current flow process are rapid spray
2.1 Spray Drying Process 13
evaporation, shorter evaporation time and less thermal degradation of the products
(Masters 1991; Anandharamakrishnan et al. 2007).
In contrast, in the counter-current configuration, the product and drying
medium enter at the opposite ends of the drying chamber. Here, the outlet product
temperature is higher than the exhaust air temperature, and is almost at the feed-air
temperature with which it is in contact. This type of arrangement is used for non-
heat sensitive products only. In another type called mixed flow, the dryer design
incorporates both co-current flow and counter-current flow. This type of arrange-
ment is used for drying of coarse free-flowing powder, but the drawback is that the
temperature of the product is high (Masters 1991).
2.1.3Moisture Evaporation
When droplets come in contact with hot air, evaporation of moisture from their
surfaces takes place. The large surface area of the droplets leads to rapid evapo-
ration rates, keeping the temperature of the droplets at the wet-bulb temperature
(Mujumdar 1987). In this period, different products exhibit different characteris-
tics, such as expansion, collapse, disintegration and irregular shape. Methods for
calculating the changes in size, density and studies of droplet drying are described
by Masters (1991).
The dry powder is collected at the base of the dryer and removed by a screw con-
veyor or a pneumatic system with a cyclone separator. Other methods for collect-
ing the dry powder are bag filters and electrostatic precipitators (Fellows 1998).
The selection of equipment depends on the operating conditions, such as particle
size, shape, bulk density, and powder outlet position.
The two main designs of commonly used spray dryers are the short-form and tall-
form driers shown in Fig.2.2.
Tall-form designs are characterized by height-to-diameter aspect ratios of
greater than 5:1. Short-form dryers have height-to-diameter ratios of around 2:1.
The short-form dryers are the most widely used, as they accommodate the com-
paratively flat spray disk from a rotary atomizer (Masters 1991). The flow patterns
observed in short-form dryers are more complex than those in tall-form dryers,
with many dryers having no plug-flow zone and a wide range of gas residence
times (Langrish and Fletcher 2001).
14 2 Computational Fluid Dynamics Applications
Fig.2.2Schematic diagrams
of spray dryer (Langrish and
Fletcher 2001)
2.3Airflow Pattern
During spray drying, the particle behaviour is dependent on the air flow pattern.
Inside the spray chamber, there is presence of significant air flow instabilities due
to the inlet swirl. The various spray dryingair flow studies have been summarized
in Table2.1. Hence, the effect of turbulence inside the spray chamber should be
considered. Huang et al. (2004) showed that RNG k- model prediction was better
for swirling two-phase flow in the spray drying chamber compared to standard k-,
realizable k- and Reynolds stress models.
The air flow patterns in an industrial spray dryer used for milk powder production
have been modeled using the transient Reynolds-averaged NavierStokes equations
with the Shear Stress Transport (SST) turbulence model (Gabites et al. 2010). These
simulations were carried out in the absence of atomized liquid droplets. The simula-
tions showed that the main air jet oscillated and processed about the central axis with
no apparent distinct frequency. In turn, the recirculation zones between the main jet
and the chamber walls fluctuated in size. Good agreement was found between the
movements of the main jet via simulations and from telltale tufts installed in the
plant dryer. The different outlet boundary condition appeared to have little influence
on the overall flow field. In the gas-only simulations, different fluid bed flows within
the range had only a local influence by reducing the length of the main jet. This may
have an effect on the particle capture by the fluid bed.
2.4Atomization
The atomization stage during spray drying is very important, since it affects the
final particle size. A co-current spray dryer fitted with pressure nozzle was investi-
gated both in experiment and CFD simulation by Kieviet et al. (1996), to develop
Table2.1Spray dryingairflow pattern studies (Kuriakose and Anandharamakrishnan 2010)
Problem descriptions Turbulence model Findings Authors
Simulation of airflow pattern Standard k- Non-swirling flow spray chamber; the k- model Oakley and Bahu (1993)
with experimental validation. and RSM gives good predictions of gas velocity profiles,
whereas for swirling flows, RSM model gives
2.4Atomization
a theoretical model for the prediction of final product quality. Good agreement was
obtained between the experimental data and the simulation. An ultrasonic nozzle
spray dryer was studied numerically by Huang et al. (2004). Birchal et al. (2006) sim-
ulated a spray dryer fitted with a rotary atomizer for drying of milk emulsion by using
CFD, and also by a model with simplified particle motion. Authors also discussed
the advantages and limitations of each model in the design and optimization of spray
dryers. Studies on the effects of atomizer types (rotary disc and pressure nozzle) on
droplet behaviour were performed by Huang et al. (2006) using CFD for spray drying
of maltodextrin. They concluded that pressure nozzle may lead to a higher velocity
variation in the center of the chamber than the rotary atomizer. Moreover, large recir-
culation of droplets was also found during pressure nozzle atomization.
2.5Particle Histories
2.6AirParticle Interaction
Momentum
Pressure (Drag) Size
Thermal energy
Temperature Temperature
(Heat transfer)
Gas Droplets
and the heat transfer rate between the droplets and the gas (Crowe et al. 1977). Hence,
all three equations (mass, momentum and energy) are interdependent and should be
included in the gas-droplet interactions (Kuriakose and Anandharamakrishnan 2010).
2.7Particle Tracking
Both the EulerianEulerian and EulerianLagrangian methods have been used in pub-
lished simulations of spray dryings. However, the Eulerian-Lagrangian frame work
was selected most often, because it provides residence time of individual particles with
a large range of particle sizes. Crowe et al. (1977) first proposed the particle source in
the cell (PSI-Cell) model. This is the basis for the discrete phase model (DPM).
In the DPM, the flow field is divided into a grid defining computational cells
around each grid point. Each computational cell is treated as a control volume for the
continuous phase (gas phase). The droplets are treated as source of mass, momentum
and energy inside the each control volume. The gas phase is regarded as a continuum
(Eulerian approach), and is described by first solving the gas flow field, assuming
no droplets are present. Using this continuous phase flow field, droplet trajectories,
together with size and temperature histories along the trajectories, are calculated.
The mass, momentum and energy source terms for each cell throughout the flow
field is then determined. The source terms are evaluated from the droplet equation
and are integrated over the time required to cross the length of the trajectory inside
each control volume. The results are multiplied (scaled up) by the number flow rate
of drops associated with this trajectory (Crowe et al.1977; Papadakis and King 1988;
Fluent 2006). The gas flow field is solved again, incorporating these source terms,
and then new droplet trajectories and temperature histories are calculated. This
approach provides the influence of the droplets on the gas velocity and temperature
fields. The method proceeds iteratively calculating gas and particle velocity fields.
The range of droplet sizes produced by the atomizer is represented by a number
of discrete droplet sizes. Each initial droplet size is associated with one trajectory;
along with the number of drops it is constant, assuming that no coalescence or
shattering occurs. Once the air velocities, temperatures, and humidity are postu-
lated, the transport equations for the droplets of each size are integrated over time
and positioned to yield droplet trajectories, velocities, sizes and temperatures.
Calculations for droplets of each initial size continue until the volatile fractions
(e.g. water) in the droplets evaporate completely, exit the column, or impact the
column wall (Papadakis and King 1988; Fluent 2006).
In the CFD simulation, a combined Eulerian and Lagrangian model is used to
obtain particle trajectories by solving the force balance equation:
dup
18 CD Re p g
= v u p + g (2.1)
dt p dp2 24 p
where v is the fluid phase velocity, up is the particle velocity, g is the density of
the fluid and p is the density of the particle.
20 2 Computational Fluid Dynamics Applications
The particle force balance (equation of motion) includes discrete phase inertia,
aerodynamic drag and gravity. The slip Reynolds number (Re) and drag coefficient
(CD) are given in the following equations:
g dp u v
p (2.2)
Re =
a2 a3
CD = a1 + + (2.3)
Re Re2
where dp is the particle diameter, and a1,a2 and a3 are constants that apply to smooth
spherical particles over several ranges of Reynolds number (Re) given by Morsi and
Alexander (1972).
The velocity of particles relative to air velocity was used in the trajectory calcu-
lations (Eq.2.1). Turbulent particle dispersion was included in this model as dis-
crete eddy concept (Langrish and Zbicinski 1994). In this approach, the turbulent
air flow pattern is assumed to be made up of a collection of randomly directed
eddies, each with its own lifetime and size. Particles are injected into the flow
domain at the nozzle point, and envisaged to pass through these random eddies
until they impact the wall or leave the flow domain through the product outlet.
The heat and mass transfer between the particles and the hot gas is derived fol-
lowing the motion of the particles:
dTp dmp
(2.4)
mp cp = hAp Tg Tp + hfg
dt dt
where mp is the mass of the particle, cp is the particle heat capacity, Tp is the parti-
cle temperature, hfg is the latent heat, Ap is the surface area of the particle, and h is
the heat transfer co-efficient.
The heat transfer coefficient (h) is obtained from the Ranz-Marshall equation.
hdp
Nu = = 2 + 0.6 (Red )1/2 (Pr)1/3 (2.5)
kta
where Prandtl number (Pr) is defined as follows
cp
Pr = (2.6)
kta
where dp is the particle diameter, kta is the thermal conductivity of the fluid, is
the molecular viscosity of the fluid.
The mass transfer rate (for evaporation) between the gas and the particles is cal-
culated from the following equation:
dmp
= kc Ap Ys Yg
dt (2.7)
where Ys is the saturation humidity, Yg is the gas humidity, and kc is the mass
transfer co-efficient and can be obtained from Sherwood number:
2.7 Particle Tracking 21
kc dp
Sh = = 2 + 0.6 (Red )1/2 (Sc)1/3 (2.8)
Di,m
where Di, m is the diffusion coefficient of water vapour in the gas phase and Sc is
the Schmidt number, defined as follows:
Sc = (2.9)
g Di,m
The values of vapour pressure, density, specific heat and diffusion coefficients
can be obtained from Perry (1984).
When the temperatures of the droplet has reached the boiling point and the
mass of the droplet exceeds the non-volatile fraction, then the boiling rate model is
applied (Kuo 1986).
d dp 4kta cg Tg Tp
= 1 + 0.23 Re ln 1 + (2.10)
dt p cg dp hfg
where kta is the thermal conductivity of the gas and cg is the heat capacity of the gas
(Kuriakose and Anandharamakrishnan 2010).
2.8Particle Temperature
The particle temperature is very important in the case of heat sensitive products, since it
influences the aroma retention and thermal stability of heat labile components. Crowe
et al. (1977) predicted that the smaller size particles have higher temperatures than
the larger particles, because the latter have a smaller surface area to volume ratio and
evaporate more slowly. Kieviet (1997) studied the airflow pattern, temperature, humid-
ity, particle trajectories and residence time in a 2D co-current spray dryer fitted with a
pressure nozzle using maltodextrin as feed solution, and concluded that the gradients
in the center region of the drying chamber could be improved. Anandharamakrishnan
et al. (2010a) studied the particle temperature in both short-form and tall-form spray
dryer using CFD simulation for drying of whey proteins. They found that due to
moisture evaporation of droplets, the temperature of droplets was high and was almost
equal to the gas temperatures outside the core region. Moreover, the temperature of gas
in the core spray region and the upper part of the chamber decreased due to the cooling
effects of evaporation. The particle nature was also affected by the outlet air tempera-
ture (Kuriakose and Anandharamakrishnan 2010).
The particle residence time has a great impact on the final powder quality and it
also affects product qualities such as solubility and bulk density. The residence
time (RT) is divided into two parts; namely, primary and secondary residence times.
22 2 Computational Fluid Dynamics Applications
The primary RT is calculated from the time taken for droplets leaving the nozzle
to impact on the wall or leave at the outlet. The secondary residence time can be
defined as the time taken for a particle to slide along the wall from the impact
position to the exit (Kuriakose and Anandharamakrishnan 2010).
Kieviet and Kerkhof (1995) determined the RTD of particles in a co-current
spray dryer during the drying of aqueous maltodextrin solutions. Kieviet (1997)
observed that during spray drying of maltodextrin solution, the larger diameter
particles have longer RTs than smaller particles. He also found enormous differ-
ence between measured and predicted results due to particle wall depositions and
sliding movement. Ducept et al. (2002) performed an experiment to determine the
RTD of particles, and validated with the CFD predictions in a superheated steam
spray dryer. The residence time distribution of different sized particles in a spray
dryer was studied by Huang et al. (2003a), and they found that different droplets
follow different trajectories in the drying chamber.
Anandharamakrishnan et al. (2010a) studied Particle Residence Time Distribution
of whey proteins in both short-form and tall-form dryers and the residence time
(Fig. 2.4). The study indicates that most of the particles have very low RT during
spray drying (short-form). It was observed that a bent outlet pipe inside the chamber
increases gas and particle recirculation (Fig.2.4); consequently, cold gas is mixed
with down-flowing hot inlet gas, and dried particles will be exposed to the high inlet gas
Fig.2.4Particle trajectories
colored by residence time(s)
(Anandharamakrishnan et al.
2010a)
2.9 Residence Time of Particle 23
The knowledge of particle impact positions is important for the design and opera-
tion of spray dryers, as it influences the final product quality. In an earlier numerical
study, Reay (1988) has shown that the most likely areas for wall deposition are an
annular area of the dryer roof and a region below the atomizer, where large particles
are likely to deposit. Later, Kieviet (1997) investigated the interaction of wall deposi-
tion with the residence time, and the effect of wall deposition on the product quality
and yield during spray drying of maltodextrin. Goula and Adamopoulos (2004)
determined the operating conditions that influence the fouling and residue accumula-
tion of the equipment during the drying process. Anandharamakrishnan et al. (2010a)
studied the particle impact position during drying of whey proteins from the simula-
tion data using an in-house post-processor. Figure2.5a, b shows the top and front
cross-sectional views of the simulated results (Anandharamakrishnan et al. 2010a).
Figure2.5a, b indicates that a large fraction of the particles (50%) strike the conical
Fig.2.5Particle impact positions, a top view, b front view (Anandharamakrishnan et al. 2010a)
24 2 Computational Fluid Dynamics Applications
part of the spray dryer chamber (similar with the earlier observation of Langrish and
Zbicinski 1994) and 23% of particles hit the cylindrical part of the wall, but only a
small proportion (25%) of the particles come out of the outlet pipe line (the intended
destination). A very small 2% of particles hit the ceiling despite the large volume of
re-circulated gas, but particles hitting the cone and/or cylindrical wall (73%) should
slide down to the main outlet aided by mechanical hammer operations. They also
found that in a short-form dryer, a large fraction of the particles strike the conical part
of spray dryer chamber, while in tall-form dryer, the particles struck the cylindrical
part of the wall. In both forms of dryer, they found less impact on the ceiling, despite
the recirculation of gas in the zone (Kuriakose and Anandharamakrishnan 2010).
2.11Current Trends
Varying the inlet air temperature from 100 to 180C resulted in contrasting drying
histories. These drying kinetics were then extended to assess the in situ crystalliza-
tion phenomenon. For this spray drying tower, it was found that lower inlet tem-
perature conditions favored a higher degree of crystallinity.
Dropletdroplet interactions during the spray drying were performed by applying
the transient mode of calculations (Mezhericher et al. 2008). The droplet collisions
influenced the temperature and humidity patterns, while their effect on velocity
was less marked. They investigated both insulated and non-insulated spray cham-
bers and reported that the insulation of a spray chamber will affect the airflow pat-
terns, thereby affecting the droplet trajectories.
The modeling of spray dryers using the population balance method is gaining
importance as the model accounts for droplet growth, coalescence and break up dur-
ing the spray drying process. Nijdam et al. (2004) modeled the particle agglomera-
tion within the spray chamber using two different frameworks, namely, Lagrangian
and Eulerian. They validated their prediction using phase doppler anemometry (PDA)
measurement, and found that in terms of ease of implementation and range of appli-
cability, the Lagrangian approach is more suitable for modeling of agglomeration of
particles. The modeling of droplet drying in the spray drying chamber by applying the
unsteady mode of calculations (Mezhericher et al. 2009) showed that among 2D and
3D analyses, the latter predicts asymmetry of flow patterns in the spray chamber. Chen
and Jin (2009b) performed transient 3D simulations in an industrial-scale spray dryer
(15m tall and 10m wide). They observed that the particles make the central jet oscil-
late more non-linearly and that the frequency of oscillation decreases with increasing
feed rate. Woo et al. (2009) have performed unsteady state simulations of spray drying
and investigated the effect of chamber aspect ratio and operating conditions on flow
stability. The authors observed that a large expansion ratio produces a more stable
flow due to the limitations of jet fluctuations by outer geometry constriction.
There remains scope for future research in the area of optimization of the spray
drying process. Further work is needed to refine the turbulence models for the
Lagrangian approach, in order to account for the various particle turbulence phe-
nomena and particleparticle correlations. Modeling of particle agglomeration
(including gasparticle interaction and particleparticle correlations), wall depo-
sition (including nature of the product) and predicting particle residence time
(including sliding movement of particles in the secondary residence time) during
spray drying of food products is currently lacking. Hence, there is also scope for
further study in the area to overcome problems like agglomeration, wall deposi-
tion, particle residence time, thermal degradation of particles and aroma loss
(Kuriakose and Anandharamakrishnan 2010). Langrish (2007) has reported the
same. Thus, the modeling approach may lead to better productivity and high-quality
food products.
Chapter 3
Applications of Computational Fluid
Dynamics in the Thermal Processing of
Canned Foods
Though several food processing technologies have been developed with the aim of
increasing the shelf-life of foods, thermal processing remains the most widely used
food preservation technique. Thermal processing of canned foods can be divided
into two major process methods: in-container sterilization and in-flow sterilization
(Weng 2006). The food is usually packed in metal containers, glass bottles, retort-
able pouches, retortable cartons, etc. A variety of foods, including fruits, vegetables,
meat, poultry, fish and dairy products, are being preserved by this method. In-flow
sterilization process refers to the aseptic processing technique wherein the food
products (mostly liquids) are sterilized prior to packaging (e.g. milk and fruit juices).
This chapter provides insight into the applications of CFD in the thermal processing
of canned foods; analyzing the liquid flow pattern; temperature and velocity profiles;
and shape, size and position of the slowest heating zone (SHZ) and the associated
biochemical changes in various types of canned foods.
3.1Canning of Foods
Recently, CFD modeling studies have been extended to the sterilization of canned
solidliquid food mixtures. An overview of the research work carried out in CFD
applications in thermal processing of canned foods is depicted in Table3.1.
The presence of solids was found to influence the velocity profile and the posi-
tion of SHZ inside the can. Unlike liquid foods, solidliquid food mixtures are
assumed to be heated by both conduction and convection (Ghani and Farid 2006;
Kiziltas et al. 2010; Rabiey et al. 2007). In the case of pineapple slices canned in
liquid sucrose solution, the liquid was assumed to be heated by natural convection
exhibiting the recirculation phenomenon (Ghani and Farid 2006), and the solid
food (pineapple slices) were presumed to be heated by conduction. For the solid
liquid mixtures, solids are randomly dispersed in the liquid phase (canned peas);
Kiziltas et al. (2010) observed slight changes in the velocity profile. This was
attributed to the heat exchange and surface deflections while the flow is slowly
moving through the stack of solid particles. The SHZ of canned solidliquid food
varied in comparison to canned liquid sterilization. Such an effect was expected,
as the presence of the solid reduces the effect of natural convection in the fluid,
causing the SHZ to move slightly upwards. On the other hand, in the case of
asparagus canned in brine, Dimou and Yanniotis (2011) found that the shape of
the solids and the available space between the asparagus does not restrict the flow
of the brine, and thus, the SHZ was found to lie toward the bottom, at a height
of 13.5% of the can height from the bottom. A similar result was observed by
Kiziltas et al. (2010) for canned peas in water, where the SHZ was found toward
the bottom of the can. Furthermore, the sterilization of solidliquid food mixtures
(beef fat in water) under a high-pressure processing unit was studied by Ghani and
Farid (2007). A forced convection phenomenon was observed in the liquid com-
ponent. Solids foods are found to be heated more than the liquid, due to the differ-
ence in their compression heating coefficient.
The following F0 value equation is used to quantify the effects of heat treat-
ment and time with respect to the survival of a microorganism (Holdsworth and
Simpson 2007).
F = 10(T Tref )/Z dt (3.1)
In the canning process, the F0 value is used for the low-acid canned foods and
refers to the sterilization value with a z value of 10K (Holdsworth and Simpson
2007; Weng 2006).
Table3.1Overview of the research work carried out in CFD applications in thermal processing of canned foods
Problem description Findings Authors
Numerical simulation of natural convection The bottom of the can heated up at a slower rate Kumar et al. (1990)
heating of canned, thick, viscous liquid than that predicted by conduction heating.
food products The slowest heating zone (SHZ) was located
at the bottom center of the can
Transient temperature and velocity profiles in The computed particle path predicted that the liquid Kumar and Bhattacharya (1991)
canned non-Newtonian liquid food during just below the top center is exposed to minimum
sterilization in a still-cook retort heat treatment, and should be of concern during
the thermal process calculations
Simulation of natural convection heating of Recirculating flow was observed inside the can. SHZ Ghani et al. (1999a)
3.3 Bacterial Deactivation Kinetics
canned food using CFD kept moving, and eventually stayed at a height of
1015% of the can height from the bottom
Bacterial deactivation of canned liquid food The concentration of the live bacteria depends on Ghani et al. (1999b)
during sterilization using CFD both temperature distribution and flow pattern
Thermal sterilization of canned foods in a 3D The SHZ was found to migrate at 3040% of the Ghani et al. (2001)
pouch using CFD pouch height from the bottom
CFD analysis of transient temperature and velocity For horizontal cans, the SHZ moved downwards and Ghani et al. (2002a)
profiles in a horizontal can during sterilization eventually stayed at a height of 2025% of the can
height from the bottom. However, the vertical cans
showed faster heating, due to enhanced natural
convection by its greater can height
Theoretical and experimental investigation of The measured concentration of Bacillus stearothermophilus Ghani et al. (2002b)
the thermal inactivation of Bacillus spores at different periods of heating showed good
stearothermophilus in food pouches agreement with the CFD simulation results.
Theoretical and experimental investigation of SHZ migrated to 3040% of the pouch height. Ghani et al. (2002c)
the thermal destruction of Vitamin C in The vitamin profile depends on temperature and
food pouches velocity profiles in the pouch. Simulation results
agreed with the measured values
(continued)
29
Table 3.1 (continued)
30
Simulation of heat transfer for solidliquid food CFD simulation of the pasteurization of canned peas in Kiziltas et al. (2010)
mixtures in cans water was validated with experimental measurements
Effect of can orientation on beer pasteurization The package orientation did not result in process improvement Augusto et al. (2010)
evaluated using CFD
Velocity and temperature field characteristics of In the presence of headspace in cans, increased heating Erdogdu and Tutar (2010)
water and air during natural convection rates were observed. Organized velocity motions along
heating in cans the airwater interface resulted in vortex evolution
Evaluation of geometric symmetry condition The axial symmetry of the bottles allows the use of Augusto and Cristianini (2010a)
of thermal process of packed liquid food smaller models, which saves computational effort
using CFD modeling.
CFD analysis of viscosity influence on Thermal The sterilization process values became lower when Augusto and Cristianini (2010b)
In-Package liquid food process the viscosity was increased up to a critical value
Investigation of asparagus sterilization process Heating of the asparagus was not uniform. The top of Dimou and Yanniotis (2011)
in a still can, using CFD modeling the asparagus receives higher heat treatment than
the bottom, due to faster heating of the top of the can
Quantifying enhancement in heat transfer due The transition to convection-augmented heat transfer Koribilli et al. (2011)
to natural convection during canned food from that of solely conduction mode occurred at a
thermal sterilization in a still retort critical Fourier number
31
32 3 Applications of Computational Fluid Dynamics
During the natural convection heating of canned liquid food products, the fluid
movement accelerates the sterilization process by enhancing the rate of heat trans-
fer. Hence, heat transfer by convection is governed by both fluid motion and tem-
perature difference (Rao and Anantheswaran 1988). A common application of
CFD and numerical simulation is the prediction of flow patterns and temperatures
during the thermal processing of foods (Scott and Richardson 1997). During natu-
ral convection heating, the velocity in the momentum equations is coupled with
temperature in the energy equation, because the movement of fluid is solely due to
buoyancy force.
In the thermal sterilization of canned liquid foods, initially the liquid near the
wall is at rest, due to the application of no-slip boundary condition. Later on, as
the steam comes in contact with the outer walls of the can, the liquid adjacent to
the wall heats up to the wall temperature by conduction. However, liquid away
from the wall is still at the initial temperature. Because of gravity and the varia-
tion of liquid density throughout the can, buoyancy forces are created. Throughout
the heating period, the buoyancy forces are opposed by the liquid viscous forces.
The velocity of the convective current is dependent upon the strength of the buoy-
ancy forces, and the magnitude of the liquid viscositys resistance to flow. As the
heating progresses, temperature-dependent viscosity decreases, resulting in an
increased velocity. This results in faster heating of the product. Later on, as the
temperature inside the container becomes uniform, the buoyancy forces decrease,
leading to a significant reduction in velocity and cessation of recirculation (Kumar
and Bhattacharya 1991; Ghani et al. 2001). For canned milk, the conduction pat-
tern, followed by convection of temperature and velocity, has been observed with
progression of time, as shown in Fig.3.1 (Anandpaul et al. 2011).
(a) (b)
(c) (d)
Temperature C
Fig.3.1Temperature (C) profiles of the pasteurization (85C) process of milk in stationary can
position at a 60s, b 120s, c 240s, and d 360s (Anandpaul et al. 2011)
Widespread applications of CFD have been found in accurately predicting not only
the location, but also the movement of SHZ as the thermal processing proceeds. The
SHZ location also depends on the mode of heat transfer and the nature of food prod-
uct. However, in the case of canned liquid foods, as the heating proceeds, the mode of
heat transfer changes from conduction to convection. This pushes the SHZ toward the
bottom of the can. Recently, Padmavathi and Anandharamakrishnan (2012) reported
that determination of the SHZ is important for analyzing the effectiveness of thermal
processing of canned products. SHZ was tracked inside canned pineapple slices dur-
ing different processing time periods (60, 300, 600, 900, 1500, and 2100s), as shown
in Fig.3.2. The authors reported that heat is initially transferred to the liquid sugar
solution adjacent to the heated wall by conduction. Further, sugar solution under the
influence of buoyancy moves upward, and leads to the downward movement of
the SHZ of liquid region to about 10% of the can height (Fig.3.2). The difference
in the temperature due to natural convection creates a density gradient that makes
the hot sugar solution move upwards, and also carries the colder fluid by a viscous
drag. The top surface of the can deflects the movement of the fluid flowing in the
upward direction to the radial direction. However, this whole process of recircula-
tion is observed until uniformity of the temperature of fluid is achieved (Kumar
et al. 1990; Kiziltas et al. 2010; Holdsworth and Simpson 2007). The difference in
the location of SHZ is observed in the case of solidliquid food mixtures heated by a
combination of conduction and convection.
60 s 300 s 600 s
Fig.3.2Temperature contours (K) of canned pineapple slices in sugar solution (Padmavathi and
Anandharamakrishnan 2012)
3.5 Thermal Processing of Canned Fruits 35
60 s 300 s 600 s
In this regard, the SHZ of the canned pineapple slices is located radially away
from the cans center, as shown in Fig.3.2. Figure3.3 shows the temperature pro-
file of the pineapple slices during thermal processing at three zones: (1) center; (2)
top, at a distance of 3.5cm from the center; and (3) bottom, at a distance of 3.5cm
from the center. It clearly shows the sequence of the region getting heated. The
heating of solidliquid processes was often based on the assumption that the solid
liquid heats up by pure conduction, ignoring the effect of natural convection and
leading to much longer than needed heating times, which results in over-processed
products. Additionally, the effect of natural convection on the heating process of
pineapple slices in a model liquid was verified by Ghani and Farid (2006), indicat-
ing that natural convection in the liquid has significant effects in the evolution of
temperature and heat transfer rate.
1
P1
P2
0.8 P3
0.6
F-Value
0.2
0
0 5 10 15 20 25 30 35 40
Time (min)
The survival of the microorganism can be quantified from process value F0 (in
minutes). The F0 values were calculated for the canned pineapple slices at three
different positions (Fig.3.4) for every 1 min interval of the thermal process.
Figure 3.4 shows that the pineapple slices attained a maximum F0 value (higher
inactivation) at 40min of heating. Further studies in the area of thermal processing
of canned solidliquid foods are recommended, as the literature reports on such
works are limited.
The significance of CFD in analyzing the thermal processing techniques is dis-
cussed in this chapter. Numerous investigations have been carried out to analyze
the velocity profile, flow pattern, temperature profile, location and movement of
the SHZ in canned foods. Furthermore, F0 value predictions indicate that the time
required for inactivating microorganisms was drastically reduced during the process-
ing of pineapple tidbits compared to pineapple slices. Apart from the canned liquid
foods heated primarily by convection, in recent years, research has been extended to
canned solidliquid food mixtures exhibiting combined conductionconvection heat
transfer. The effect of can orientation at different inclinations on the temperature dis-
tribution inside the can was also analyzed. Geometry modifications have also been
included in the survey, exposing the enhanced heat transfer exhibited by upright
conical cans. CFD has been valuable in predicting the inactivation kinetics of bac-
teria and vitamin degradation, which are difficult to tag experimentally during the
sterilization process. On the whole, CFD is a precious tool in scrutinizing the various
problems associated with the thermal processing industries, and promises to solve all
kinds of difficulties that could be encountered in the future.
Chapter 4
Computational Fluid Dynamics Modeling
for Bread Baking Process
Computational fluid dynamics (CFD) modeling of the entire bread baking pro-
cess is very complicated, due to the involvement of simultaneous physiochemical
and biological transformations. Bread baking is a complex process, where com-
position, structure and physical properties of bread change along the process,
and bread transforms from dough to a product containing soft crumb and crispy
crust. CFD can be applied in the modeling of such a complex process. This chap-
ter provides the insight of different radiation models used for modeling of heating
in electrical heating ovens; and modeling of the bread baking process, along with
the predictions of bread temperature, starch gelatinization and browning index.
Moreover, current limitations, recent developments and future applications in CFD
modeling of bread baking process are also discussed in detail.
4.1Introduction
Bread is closely related to peoples daily life, and bread baking is a food process-
ing technique in which a series of complex physical, chemical and biochemical
changes simultaneously take place in the given product. Under the influence of
heat, raw dough piece is transformed into a light, porous, readily digestible and fla-
vorful bakery product (Chhanwal et al. 2012). Although bread baking had been in
practice for a very long time, comprehensive understanding of the physical process
in baking is still lacking due to its complexity (Cauvain 2003). Bread quality pre-
dominantly depends on four parameters, i.e. texture, moisture content, bread sur-
face colour and structure (shape and size) of the bread. These four parameters vary
during the baking process, due to variations in operating parameters. However, all
the above parameters are temperature dependent. Therefore, bread quality can be
controlled through optimization of the oven temperature profile (Therdthai and
Zhou 2003). Moisture content and temperature are responsible for physiochemical
and biological transformations in the bread, such as evaporation of water, gelati-
nization of starch, volume expansion, crust formation, denaturation of protein
and browning reactions etc.; the make the baking of bread a complex process
(Therdthai and Zhou 2003). Variations in bread quality can be minimized by proper
design of the oven, as well as maintaining proper processing conditions such as air
temperature, heating power, baking time and bread size (Zhou and Therdthai 2007;
Wong et al. 2007a, b; Chhanwal et al. 2012). Due to the increase in computational
power, trained technical people, and availability in the last decade of user friendly
and robust software such as Fluent, CFX, and COMSOL Multiphysics etc., use of
CFD in the baking industry has increased (Williamson and Wilson 2009).
Baking ovens play an important role in deciding the final product quality of any
baking process. Ovens are an integral part of any baking process, and serve as an
energy source that leads to heat and mass transfer in the product. The success or
failure of any baking venture is determined by the oven and its operating condi-
tions (Chhanwal et al. 2012). Baking ovens can be categorized based upon scale
(size and capacity), product being baked (bread or biscuit, etc.), physical arrange-
ments (batch or continuous), heating source (electrical or hot air heating) and
mode of heating (conductive or convective). Electrical heating ovens are widely
used in the batch baking process, due to their adaptability for different bakery
products. The most commonly used continuous bread baking ovens are the trav-
elling tray oven and the rotary rack oven. Location of heating source, air flow,
product load, vent position, placement of the bread, and baking time are the major
factors affecting heat distribution in the oven chamber. The product quality varies
during the initial period of the baking process, due to a high temperature gradient
between hot oven and product. Temperature rise occurs due to initial heat absorp-
tion, and initiates a number of physical mechanisms (Zhou and Therdthai 2007;
Chhanwal et al. 2012).
During the baking process, radiation is the most dominant mode in an electri-
cal heating oven among radiation, conduction and convection heat transfer modes.
Radiative heat transfer occurs from red hot heating coils and hot metal surfaces
in the form of electromagnetic waves to the surroundings. It occurs by photons,
which are emitted by the respective surface and travel in straight lines without
attenuation. Heating by radiation depends on emissivity of surface, and the higher
the emissivity, the higher the heating rate is (Zhou and Therdthai 2007). Air inside
the oven is heated when it comes in contact with the heat source (coils), as well
as hot metal parts of oven walls. Heat transfer from hot air to the product surface
occurs by convection, and heat from a metal pan is transferred to the bread by
conduction. Sparrow and Abraham (2003) extensively studied heat transfer in an
electrical heating oven with a variety of geometries, radiative sources and operat-
ing conditions of the oven and also placement of thermal load.
Simultaneous heat and mass transfer during bread baking is governed by a
mechanism of evaporationcondensation. In this mechanism, water evaporates
4.2 Bread Baking Process 39
on the warmer side of the gas cell, absorbing latent heat of vaporization. Water
vapor migrates through the gas phase, due to a vapor concentration gradient inside
the cell. Further, water vapor condenses, setting free its latent heat on the colder
side of the gas cell and becoming liquid. Finally, heat transfer occurs by conduc-
tion and moisture is transported by diffusion, through a dough membrane to the
warmer region of the next gas cell, where all the above processes reiterate. When
the gas phase becomes continuous, the last step no longer exists. This evapora-
tioncondensation mechanism takes place until the temperature gradient remains
and the temperature of the whole crumb reaches 100C (Therdthdai and Zhou
2003; Wagner et al. 2007; Purlis and Salvadori 2009; Chhanwal et al. 2012).
The baking process occurs as follows: an increase in temperature initiates rapid
evaporation of water and release of carbon dioxide, which produces oven spring in
the first stage of baking. The top crust is thus pushed up, crumb development follows,
and finally, gradual color development occurs. Structural change also occurs dur-
ing the bread baking process and is comprised of solidification and expansion. Due
to the starch gelatinization and protein denaturation, network-like structure forms in
the breads crumb. During gelatinization, starch granules hydrate and swell, and an
increase in viscosity, as well as clarity of the medium, is observed. Starch gelatini-
zation is normally used as the minimum baking index. The non-enzymatic brown-
ing reaction is induced by increase in temperature and lower moisture content, which
results in crust formation at the surface of the bread. Moisture loss due to evapora-
tion of water varies with the properties of dough and baking conditions. Evaporation
condensation of water keeps the moisture of the breads core region constant, while
maximum moisture loss occurs at the surface of the bread. However, the completion
of baking results in two different zones within the bread, each with its own unique tex-
ture, crumb and crust (Zhou and Therdthai 2007; Chhanwal et al. 2012).
In recent years, CFD has been applied more and more to the design and develop-
ment of baking ovens, and also to investigate the baking process (Wong et al. 2006;
Zhou and Therdthai 2007). The published research works on CFD simulation of
bread baking process, along with their salient findings, are summarized in Table4.1.
CFD simulation studies were initiated for the improvement of oven design and opti-
mizing the baking conditions. In an early study, Therdthai et al. (2003) developed
a 2D model for an industrial continuous baking oven, to study the temperature and
airflow pattern throughout the baking chamber under different operating conditions.
Subsequently, the model was extended to a 3D model with moving grid and tran-
sient state assumption, to more accurately predict temperature and velocity inside the
oven (Therdthai et al. 2004b). The impact of the dough/breads physical properties
on the accuracy of the predictions was analyzed by Wong et al. (2006), using a com-
bined experimental and CFD modeling approach. They found that physical proper-
ties have a significant impact, and that density and specific heat capacities tend to be
40
Table4.1Overview of the research work carried out in CFD simulation of the bread baking process (Chhanwal et al. 2012)
Description of CFD model Turbulent/radiation models Salient findings Reference
3D Industrial electrical forced convection RNG k- turbulence model Airflow pattern inside the oven was Verboven et al.
oven obtained. Relationship between fan (2000)
head capacity, fan swirl and oven
geometry was studied
2D geometry of industrial continuous bread Turbulence model with convective Effect of different energy supply and Therdthai et al.
baking oven. Heat supplied with gas heating and steady state simulation air volume on oven temperature (2003)
burner and circulated by airflow profile and airflow pattern at
various cross sections inside the
oven was studied
3D model with moving grid mesh; U-turn Turbulence model for transient state Temperature profile and airflow Therdthai et al.
movement simplified by putting into simulation pattern inside the oven was (2004b)
two parts, i.e. front and back part obtained with tin surface tempera-
ture for entire baking process
3D geometry of electric heating oven with S2S radiation model Temperature profile of baking oven Mistry et al.
natural convection flow was obtained for broil and baking (2006)
cycle
2D geometry of industrial continuous bread DO Radiation with k- turbulence Effect of thermo physical properties Wong et al.
baking oven was used with sliding mesh. model on prediction of bread temperature (2006)
U-turn zone ignored, geometry simpli- profile was investigated. Density
fied by flipping lower part and aligning and specific heat capacities were
along upper part more dominant factors for
accuracy of the model predictions
2D geometry with sliding mesh. Grid DO Radiation model with k- Temperature profile of the bread was Wong et al.
resolution study was also carried out turbulence model obtained for entire baking process (2007a)
Incorporation of a UDF for feedback DO Radiation with k- turbulence Feedback controller tuned fine and Wong et al.
controller to optimize oven temperature model preheating time reduced (2007b)
(continued)
4 Computational Fluid Dynamics Modeling
Table4.1(continued)
Description of CFD model Turbulent/radiation models Salient findings Reference
3D design of gas-fired radiant burner was Radiation model with shear stress A burner developed with more Williamson and Wilson
used in a chamber of industrial tunnel transport k- based turbulence uniform radiative flux in oven (2009)
baking oven model chamber. Effect of oven chamber
humidity and surface emissivity on
radiative heat transfer was studied
3D geometry of industrial bakery pilot oven S2S radiation model with k- realizable Temperature and radiative heat flux Boulet et al.
was used with controlled air tempera- turbulence model were obtained for top and bottom (2010)
ture and velocity plates
3D domestic electrical heating oven was Three radiation models, i.e. DO, S2S The temperature profiles of three Chhanwal et al.
used with bread. Kinetic model for and DTRM, were used for transient radiation models were compared. (2010)
starch gelatinization was incorporated state simulation Starch gelatinization profile for
bread slice was visualized
3D pilot-scale electrical heating oven DO radiation model Different temperature profiles of top Anishaparvin
containing with nine heating coils was and bottom chamber were et al. (2010)
4.3 CFD Modeling of the Bread Baking Process
more dominant factors. Wong et al. (2007a) developed a 2D CFD model, extending
the earlier works of Therdthai et al. (2003, 2004b) for continuous movement of a
traveling tray using sliding mesh technique and employing a discrete ordinate (DO)
radiation model. Wong et al. (2007b) proposed the elimination of preheating of the
oven before baking, and also incorporated a feedback controller for temperature in
the CFD model by using user-defined function (UDF) for the industrial baking oven.
In the literature, different types of radiation models such as discrete transfer
radiation model (DTRM), surface to surface (S2S) and discrete ordinate (DO) have
been used for analyzing radiative heat transfer inside a baking oven. Mistry et al.
(2006) studied CFD modeling of heat transfer in a batch electric heating oven using
the S2S radiation model with bake and broil cycle, without the product. Wong et al.
(2007a) used the DO radiation model to simulate a continuous baking oven involv-
ing u-turn movement with sliding mesh. Dhall et al. (2009) modeled a near-infrared
oven using the DO model to predict heating of a cubical food sample. Boulet et al.
(2010) used the S2S model to predict radiative heat transfer inside a pilot oven.
Recently, Chhanwal et al. (2010) compared all three radiation models (i.e. DO,
S2S and DTRM) for a domestic electrical heating oven, and concluded that predic-
tions by all these radiation models were alike. However, the DO model needs more
computational power and time compared to the S2S and DTRM. A combination of
convective and radiative heat transfer was considered for both glass door and steel
walls of the oven, and heating coils were modeled as a volumetric heat source. The
authors also studied the temperature profile inside a domestic electrical heating
oven and found that a low temperature zone exists near the oven walls.
Anishaparvin et al. (2010) studied the temperature distribution inside a pilot
scale electrical heating oven, as shown in Fig.4.1. A significant difference of
temperature profile was observed between the bottom and top tray due to the
heating source, inlet vent and product placement. Furthermore, Anishaparvin et al.
(2010) developed a CFD model for the process of baking bread in a pilot-scale
baking oven, to find out the effect of hot air distribution and placement of bread
on temperature and starch gelatinization index of bread. Their study indicated
that vent position and placement of bread are the most important factors in
influencing the air temperature profile inside the oven cavity. Due to the airflow
(a) (b)
Fig.4.1Oven with bread: (a) geometry (b) meshed oven (Chhanwal et al. 2011)
4.3 CFD Modeling of the Bread Baking Process 43
pattern, bread placed on the top tray bakes quickly compared to bread placed on
the bottom tray. Thus, this study indicates that apart from temperature and time,
the placement of bread influences the final bread quality during a batch baking
process in an electrical heating oven. Chhanwal et al. (2010) studied the tem-
perature profile of bread during the entire baking process, with preheating of a
domestic electrical heating oven.
Chhanwal et al. (2011) used an evaporationcondensation mechanism to imitate
the real bread baking process, where bread-crumb temperature remains below 100C
for the entire baking process. The evaporationcondensation mechanism keeps the
bread-crumb temperature in the range of 9899C. However, earlier CFD mode-
ling studies were unable to predict this phenomenon. The evaporationcondensation
mechanism was included by defining the specific heat of bread as a function of tem-
perature, including enthalpy jump at the phase change and thermal conductivity of
bread as a function of temperature, where thermal conductivity increases until bread
reaches 100C and is made constant (0.2W/mK) once condensation starts (Purlis
and Salvadori 2009). In this study, temperature predictions for crumb were identical to
the real bread baking process. This model prediction mimics the experimental bread
center temperature. This model is very useful for predicting the bread temperature
more accurately than the earlier published CFD models. Moreover, this model
temperature prediction can also be used to predict physiochemical changes of bread
during baking, such as starch gelatinization and browning index.
Recently, Tank et al. (2012) studied the effect of a partially loaded and a fully
loaded oven on the temperature profile of a bun, and concluded that placement of
the bun inside the oven affects the temperature of the bun and consequently, the
quality of the product. Velocity and temperature profiles differ in a partially loaded
and a fully loaded oven.
Figure4.2 shows temperature profiles of a partially loaded oven, in which the
top portion of the oven shows a higher temperature as compared to the bottom,
due to the natural convection and presence of a vent at the bottom. Higher tem-
perature was noted at the extreme left, close to the oven wall, and in the right
central portion, compared to the rest of the oven, which showed non-uniform
heat distribution. This non-uniform heat distribution inside the oven was due to
the arrangement of the heating source (heating coils), which covered an uneven
area within the oven cavity, as well as location of vent. In this study, a 2530C
temperature difference was observed in the top and bottom zones of the oven cavity.
Due to uneven temperature distribution in the oven, a bun placed on the bottom
rack showed slower baking, while a bun placed on the top rack showed rapid
baking.
The degree of starch gelatinization can be used as a minimum baking index
in the industrial baking process, which decides the baking time of bread.
Gelatinization properties depend on the type and origin of the starch. Differential
scanning calorimetric (DSC) and X-ray diffraction methods are mainly used for the
measurement of the degree of starch gelatinization during the bread baking process.
Starch gelatinization is a function of time/temperature and follows first order kinetics
(Zanoni et al. 1995a). Zanoni et al. (1995a) studied starch gelatinization using
44 4 Computational Fluid Dynamics Modeling
(a)
(b)
Fig.4.2Comparison of temperature profiles of an oven: (a) partially (one tray) loaded (b) par-
tially (two trays) loaded, and (c) fully loaded oven at 300s of the baking process (Tank et al. 2012)
the differential scanning calorimetric (DSC) method, and calculated the model
parameters such as k0 and Ea, as shown in the following equations.
I = exp (kt) (4.1)
where is degree of starch gelatinization, k the reaction rate constant, and t the
time in seconds. The reaction rate constant (k) in turn can be calculated using
Arrhenius type equation:
Ea
k = k0 exp (4.2)
RT
(a) (e)
(b) (f)
(c) (g)
(d) (h)
Temperature (C) Starch Gelatinization
Fig.4.3Comparison of temperature [(a) 60s (b) 300s (c) 600s, and (d) 900s] and starch
gelatinization [(e) 60s (f) 300s (g) 600s, and (h) 900s] profiles in bread during the bread bak-
ing process, taken at 7.5cm to the right and left from center of the bread, in radial direction
Temperature and starch gelatinization profiles of crumb across the bread at (a)
60s (b) 300s (c) 600s, and (d) 900s at 7.5cm (left and right) from the center
of the bread are depicted in Fig.4.3ah. The temperature on and near to the sur-
face of the bread was very high compared to the center, which is responsible for
the crust formation and browning of the bread, and thus decides the quality of the
bread (Fig.4.3ad). Bread core temperature remains below 100C, due to the
incorporation of evaporation-condensation mechanism. Starch gelatinization pro-
gresses from crust to crumb as temperature increases in the crumb region. From
Fig.4.3, it can be observed that the starch gelatinization progressed from crust to
the crumb, with the formation of different, corresponding gelatinization zones.
The extent of gelatinization at the top and bottom was rapid as compared to the
side portion of bread, which is reflected in the temperature profile. Gelatinization
was found to be complete within 300s on outer surface as compared to 900s at
the center of the bread (Fig.4.3eh). Starch gelatinization completes within 900s
46 4 Computational Fluid Dynamics Modeling
2011) 160
120
80
40
0
0 300 600 900 1200 1500
Time (s)
4.3 CFD Modeling of the Bread Baking Process 47
Browning Index
process (Chhanwal et al.
2011) 40
30
20
10
0
0 300 600 900 1200 1500
Time(s)
the right side of the bread was higher than on the left side, due to the oven air tem-
perature and airflow pattern inside the oven.
Browning index across bread surface was obtained by integrating a kinetic
model with CFD simulation predictions of temperature at various points, and is
presented in Fig.4.5. As the surface temperature of the bread increases, browning
also increases. The browning index reached the saturation value (i.e. constant
value) toward the end of baking process. The degree of browning at the corner
(points 17) was rapid compared to the surface of bread (points 8 and 9); this was
proportional to the temperatures at those points (see Fig.4.4).
Zanoni et al. (1995b) determined the maximum browning index (dark
brown charred bread) and minimum browning index as 52 and 30, respectively.
Figure4.5 illustrates that browning index of all points fall within this range during
the complete baking process. Thus, 25min of baking was ideal for producing high
quality bread under the above baking conditions.
It is just the beginning of an era of CFD modeling in the process of baking bread,
and there is much more to achieve for optimizing different oven conditions and
configurations. Although prediction of CFD models match with the experimental
observations to some extent, they have some limitations due to assumptions, such as
exclusion of moisture transfer and volume expansion. Moisture transfer and volume
expansion are critical phenomena in the bread baking process, but due to limitations of
commercial CFD software codes, these are not considered in any of the reports pub-
lished so far. CFD modeling of different heating modes, such as infrared and microwave
with or without conventional heating, needs to be further studied for a better under-
standing of the effects that these heating methods have on the bread baking process.
48 4 Computational Fluid Dynamics Modeling
This chapter highlights the importance of CFD in the modeling of the bread
baking process. There is considerable growth in the development and application
of CFD simulation in the areas of baking ovens and the bread baking process,
to predict the complex flow patterns in the ovens as well as bread temperatures.
However, more computational modeling work, including mass transport and volume
expansion during bread baking process, needs to be undertaken. Other baking
process stages, such as mixing, fermentation, and proofing, need to be thoroughly
studied to optimize product quality. On retrospection, this chapter clearly high-
lights CFD as a valuable tool for the design of baking ovens, and prediction of
bread temperature along with various physiological changes such as browning
index and starch gelatinization during the bread baking process.
Chapter 5
CFD Modeling of Biological Systems
with Human Interface
The process of food digestion begins in the mouth, where food is broken down
by mastication and lubrication with saliva to form a consistent mass known as a
food bolus (Hiiemae and Palmer 1999; Smith 2004). Later on, disintegration of the
ingested food bolus takes place inside the stomach, due to the flow of gastric fluid
shearing the food surface (Abrahamsson et al. 2005; Kong and Singh 2008). The
gastric juice secreted along the walls of the stomach brings about the enzymatic
breakdown of food material, while the peristaltic movements generated along the
distal part of stomach cause mechanical disintegration. The flow of gastric juice is
strongly influenced by peristaltic movements of the stomach wall and rheological
properties of the food material. Finally, the discharge of digested materials into the
small intestine is regulated by the pylorus valve, and the food materials are broken
down to particles 12mm in size (Kong and Singh 2010).
5.2.1Stomach Geometry
The stomach is divided into two compartments: the upper fundus region (reservoir
to accommodate incoming ingesta), which occupies one-third of the total stomach,
and the remaining lower part, which contains the antrum (grinding and shearing)
and pylorus (regulator valve for stomach emptying).
Ferrua and Singh (2010) developed a simplified 3D model capable of reproduc-
ing the average shape and capacity of a human stomach (Einhorn 1898; Geliebter
et al. 1992; Keet 1993; Kim et al. 2001; Schulze et al. 1998). The model represents
the stomach as a J shaped organ with a greater curvature of 34cm, a maximum
transverse diameter of 10cm, a pylorus sphincter diameter of 1.2cm and a capacity
of 0.9l, as shown in Fig.5.1.
The movement of the stomach wall known as peristalsis induces mechanical break-
down of food particles. As soon as food material enters the stomach via the esopha-
gus, a series of powerful peristaltic antral contraction waves (ACW) develops within
the stomach, propagating toward the pylorus (Ferrua et al. 2011). These antral con-
traction waves are responsible for the breakdown of food materials by generating
fluid motions and mixing gastric juice with food. The patterns for this ACW forma-
tion have been described in numerical models developed by Pal et al. (2004) and
Ferrua and Singh (2010). ACW initiated every 20s at a distance of 15cm from the
pylorus, and their velocity of propagation is about 2.5mm/s (Fig.5.2).
The total lifespan of an ACW is 60s. The motility pattern is periodic after 40s
and repeats every 20s (Xue et al. 2012). The depth of this contraction has been
5.2 Modeling of Food Digestion Inside the Human Stomach 51
Fig.5.1Three dimensional
geometry of the human
stomach (Ferrua et al. 2011)
Fig.5.2ACW motility
pattern in the stomach (Xue
et al. 2012)
defined based on an occlusion ratio (), which is the ratio of change in diameter of
section before and after contraction (Pal et al. 2004; Ferrua and Singh 2010). The
length of propagation, amplitude of contractions, and the time taken by the stom-
ach to generate the contraction waves are affected by the stomach volume, as well
52 5 CFD Modeling of Biological Systems with Human Interface
as by chemical and physical properties of the meal (Camilleri and Prather 1993;
Mayer 1994). A typical ACW pattern is defined in Eq.5.1 (Pal et al. 2004):
0.25 0.25(17.5 t)/17.5 0 t 17.5
D
0.25 17.5 t 33.5
(5.1)
=1 =
D0
0.5 0.25(57 t)/23.5 33.5 t 57
0.5 0.5(t 57)/3 57 t 60
The flow of gastric juice inside the human stomach is predominantly character-
ized by two flow patterns; namely, retropulsive jet-like motion and flow recir-
culation, i.e. formation of eddies (Ferrua et al. 2011). These two flow patterns
are mainly responsible for the mixing and grinding of food particles inside the
human stomach (Pallotta et al. 1998; Pal et al. 2004; Schulze et al. 1998). The
contraction force of the wall directly elevates the pressure of the stomach in a
closed system. The pressure difference between the two compartmentsthe
body and the antro-pyloric region inside the stomachcreates the vortex flow.
As the contraction waves propagate toward the pylorus, the pressure builds up in
the distal part of stomach, i.e. the antro-pyloric region. This increase in pressure
at the pylorus region causes the food material to be forced back into the mix-
ing region (i.e. the antrum). This particular flow of gastric juice that causes food
material to be forced back to the antrum region is called retropulsion (Ferrua and
Singh 2010).
There are two different modeling approaches used to model the multiphase flow
(involving more than two phases) dynamics occurring in the stomach. Firstly, the
EulerLagrange approach is used to track the movement of each particle by taking
particleparticle interactions and the forces acting on them into account. Secondly,
the EulerEuler approach considers all phases to be continuous and interpenetrating.
As the former method is expensive and not suitable for dense particulate suspension,
the Eulerian model was found suitable for gastric fluid modeling (Xue et al. 2012).
The flow field that develops within the stomach was modeled with a laminar and
incompressible fluid flow of a continuous liquid phase (Pal et al. 2004, 2007). Under
these flow conditions, the conservation of mass and momentum within the system is
given by Eqs.5.2 and 5.3, respectively (Ferrua and Singh 2010).
5.2 Modeling of Food Digestion Inside the Human Stomach 53
u (5.2)
=0
x
u 1 (5.3)
+ u u = p + 2 u
t
where u is the velocity component (m/s), p is the pressure (Pa), / is the kin-
ematic viscosity of the fluid (cP), and is the density of the fluid (kg/m3).
The rheological properties of the fluid can have a significant effect on the dynamics
of gastric flow during digestion. Efficient digestion of ingested materials can be
achieved through appropriate alterations in the physical structure and reduction in
the content of non-nutrient components that influence the viscosity of the liquid
phase (Ferrua et al. 2011).
Ferrua and Singh (2010) studied the effects of viscosity changes on the flow
behavior of gastric juice based on the two characteristic flow patterns discussed
earlier. Figure5.3ab shows the prediction of a characteristic flow field for a
Newtonian fluid at two different viscosities.
Fig.5.3Gastric fluid velocity profile (cm/s) associated with (a) 1cP viscous fluid and (b)
1,000cP viscous fluid (Ferrua and Singh 2010)
54 5 CFD Modeling of Biological Systems with Human Interface
It can be observed from Fig.5.3 that gastric fluid with 1cP viscosity attained a
maximum flow velocity of 2.8cm/s (Fig.5.3a), and that when viscosity increased to
1,000cP, the maximum velocity attained was 4cm/s (Fig.5.3b). The results were vali-
dated with previously published experimental results, and suggested that an increase in
fluid viscosity diminished the formation of a retropulsive jet (as seen in Fig.5.3b) and
resulted in poor mixing of food particles with gastric juice (Ferrua and Singh 2010).
The density difference between two phases may impact particle distribution within
the mixture through buoyancy effects (Xue et al. 2012).
From Fig.5.4ab, it can be seen that when the particle density was increased
to 5% greater than the gastric fluid, a retropulsive jet with a very high maximum
velocity of 6.2cm/s (Fig.5.4b) was observed near the antro-pyloric region. This
may be due to sedimentation of heavier particles at the bottom of the stomach,
resulting in narrower passage for fluids to flow through. This leads to the creation
of a higher retropulsive velocity jet when compared to single phase.
Xue et al. (2012) studied the effect of particle loading ratio on the mixing of food
particles using the multiphase EulerEuler modeling approach. The particle vol-
ume loading that defines the average distance between individual particles has
been found at various levels, as shown in Fig.5.5ad.
Fig.5.4Particle volume fraction (%) contour in the stomach with mild ACWs at time t = t0
+10s with (a) =1,000kg/m 3 and (b) =1,050kg/m3 (Xue et al. 2012)
5.5 Effect of Particle Loading on Mixing 55
Fig.5.5Velocity magnitude for gastric juice (m/s) with 2mm particle diameter and 10% par-
ticle volume fraction at (a) particle density, =1,000kg/m3, and (b) =1,050kg/m3; when
particle volume fraction increased to 30% with 1mm diameter at (c) =1,050kg/m3 and (d)
=1,050kg/m3, respectively (Xue et al. 2012)
The velocity profile for gastric juice at 10% (Fig.5.5ab) and 30% (Fig.5.5cd)
particle loading levels were analyzed after 10s of processing time. The authors
found that there was a significant reduction in mixing capability of ACW at a
higher particle loading ratio of 5% difference between solid and liquid phases.
From Fig.5.5, it can be observed that a higher particle volume fraction can block
the passage to form a retropulsive jet, resulting in a reduced degree of mixing.
The small intestine is a complex and highly regulated system. It is the longest part
of digestive tract system, measuring about 26m in length. The total surface area
of the human small intestine is 2106cm2 (Mahler et al. 2012), and the duodenum
is the first section of the small intestine, followed by the jejunum and ileum
(Fig.5.6), comprising 5, 50, and 45% of the small intestinal length, respectively
56 5 CFD Modeling of Biological Systems with Human Interface
(Balimane and Chong 2005). The diameter of gut varies along each section of the
small intestine and is considered to be an average of 2.5cm, varying from 5cm in
the upper duodenum to less than 2cm in the distal ileum (Stoll et al. 2000).
Modeling of the small intestine is to obtain efficient conversion and increased
bioavailability of nutrients; there should be perfect mixing of gastric contents
inside intestinal lumens to enable faster diffusion of nutrients through the wall.
Development of such a model has been a great challenge due to the complexity
involved in mass balance around the small intestine (Stoll et al. 2000; Spratt et al.
2005; Dokoumetzidis et al. 2007). It is very complex to understand the mecha-
nisms of digestion and absorption inside the small intestine, since there is very
little information available from an engineering perspective (Norton et al. 2006;
Tharakan et al. 2007). Computational modeling can be used for understanding the
absorption pattern in the small intestine with the effects of physiological variables,
physicochemical properties and their formulation (Parrott and Lave 2010). Using
the Computational Fluid Dynamics (CFD) and/or Finite Element Modeling (FEM)
technique, it is possible to theoretically evaluate fluid dynamic parameters such
as intestinal fluid velocities and shear rates during the segmentation motion of the
small intestine.
5.6 Modeling of Absorption Process in the Small Intestine 57
(a)
0.5
s
1s
2s
3.2
s
5.5
s
(b)
(c)
Fig.5.8Horizontal velocity profiles (cm/s) for flow of Newtonian fluids with (a) 103 m2 s1
(b) 104 m2 s1, and (c) 105 m2 s1 viscosity (Love et al. 2012)
The movement of the small intestinal wall is responsible for the flow behavior of
intestinal fluids, and can be divided into segmentation contractions and propulsive
58 5 CFD Modeling of Biological Systems with Human Interface
Love et al. (2012) used a finite element model to simulate flow through the small
intestine as induced by peristalsis and segmentation motion. The segment of intes-
tine was represented by a two dimensional mesh, and equations were formulated
to describe the motion of the mesh boundary during peristalsis.
Figure 5.8ac shows the horizontal velocity profiles for flow of fluids with
various viscosities through a tube simulating the small intestine. The model was
solved using a finite element method to determine the consequent deformation of
the mesh, as a function of time. It was assumed that the flow of digesta could be
adequately represented by a 2D cross-section of intestine with a length of 10cm
and a diameter of 0.5cm.
CFD modeling of biological systems allows the food industry to design and
provide unique food structure, as both mixing and food formulation have signifi-
cant effects on mass transfer phenomena. At the same time, as mixing enhances
mass transfer processes, increase in viscosity of the food results in a reduction of
the absorption across the membrane of small intestine.
Chapter 6
Computational Fluid Dynamics Modeling
for High Pressure Processing
High pressures processing (HPP) has shown its importance in food preservation,
due to its microbial inactivation ability at lower temperature, i.e. non-thermal steri-
lization, and high pressure in the range of 100900MPa. HPP keeps a products
freshness intact, along with nutrients such as vitamins. HPP is used for pressure
shift freezing, high pressure thawing, and high pressure thermal (HPT) processing.
HPP technology has shown rapid development in the preservation and processing
of liquid and high moisture foods, such as fruit juices, ready-to eat meals, meat
products, sea-food products, etc. Combined with high temperature, high pressures
thermal (HPT) processing has shown promise in increasing shelf-life of foods at
room temperature for months to years. HPT helps to achieve commercial sterile
low acid shelf-stable food products. However, the temperature profile of a food
during the high pressure thermal process has a significant effect on the final
products quality and preservation. Therefore, it is very important from the food
safety point of view to achieve uniform temperature inside the processing vessel
(Koutchma 2012; Ghani and Farid 2007a).
Product composition plays an important role in optimizing processing conditions
for the HPT process, as temperature increase per 100MPa varies with product
composition. Earlier, researchers observed that during the HPP process, the rate
of microbial inactivation decreased with processing time. Therefore, for product
safety and quality, microbial inactivation should be ensured by considering com-
plicated intrinsic and extrinsic factors. Computational Fluid Dynamics (CFD)
modeling can help in optimization of processing conditions and system structure
of the HPP process (Koutchma 2012; Ghani and Farid 2007a, b). In the last dec-
ade, CFD modeling has shown promise in predicting the temperature of canned
product during thermal sterilization and pasteurization processes. Similarly,
CFD can be used to predict the temperature profile across a vessel during HPT
processing.
Generally, high pressure processing is carried out in batch mode, which con-
sists of compression phase, pressure-holding phase, and pressure release phase. In
compression phase, pressure is increased to the desired level. This is followed by
the pressure-holding phase, where the desired elevated pressure is kept constant
for a few minutes. Lastly, the pressure is reduced to ambient pressure. In the HPP
system, water is usually used as a pressure-transmitting medium, i.e. carrier material.
During HPT, heat is generated volumetrically within the food due to the rapid
pressurization. HPP has several advantages over conventional processes, such as
retention of freshness, texture, flavor, and color, along with reduced processing time.
Commercially available HPP products include oysters, fruit preserves, meats, and
milk (Ghani and Farid 2007a; Khurana and Karwe 2009).
Earlier, Knoerzer et al. (2007) developed a model for a pilot scale 35L ves-
sel for the HPT process using polytetrafluoroethylene (PTFE) carrier. PTFE
carrier helps to produce thermal uniformity in a vessel, and is thus able to pro-
duce 12D reduction during the process, whereas an empty vessel fails to achieve
spore reduction. Further, Juliano et al. (2009) used a computational thermal fluid
dynamic (CTFD) model to predict sterilization uniformity in a 35L HPT vessel
by predicting inactivation of C. botulinum. A CFD model for high pressure com-
pression of beef fat and water mixture was developed to obtain the temperature
distribution along pressure and velocity profiles. It was observed that solid pieces
of food were heated more rapidly than liquid food, due to a different compression-
heating coefficient (Ghani and Farid 2007b).
Khurana and Karwe (2009) developed a transient model of high hydrostatic
pressure processing to predict the temperature distribution of a pressurizing
medium (water). Water was pressurized in a 10 L vessel of a high hydrostatic
Fig.6.1Configuration of
model geometry and mesh
in a vertical cross section
of the high-pressure vessel
steel structure containing
the PTFE carrier. (a) HP
system (b) HP system with
carrier dimensions indicated
(Knoerzer et al. 2010)
6 Computational Fluid Dynamics Modeling for High Pressure Processing 61
pressure processing unit, from 0.1 to 586MPa in 180s, with an initial tempera-
ture of 25C. Pressure was maintained for 600s in the pressure hold up period.
It was observed that adiabatic heat generation in water and cooling at the walls
result in non-uniform temperature distribution at the end of the process. It was
also observed that the increasing initial temperature of the pressurizing medium
showed an increase in temperature non-uniformity. Vessel wall temperature distri-
bution was affected, mainly due to natural convection and conjugate conduction
heat transfer. However, due to the unavailability of water properties (such as density,
heat capacity, thermal conductivity, viscosity, and thermal expansion coefficient)
at elevated pressure, all these properties were assumed to be constant. Due to this
assumption, the model predictions were not very accurate.
Knoerzer et al. (2010) developed a model for carrier optimization in a pilot-
scale high-pressure sterilization plant using finite element methods. The geometry
of the model of the high-pressure chamber is shown in Fig.6.1a, and meshed
geometry is depicted in Fig.6.1b. In Fig.6.1a, the outer cylinder shows the high-
pressure chamber, the inner cylinder shows the carrier, and all shaded areas represent
the axis-symmetrical computational domains (Knoerzer et al. 2010). In this model,
the k- turbulence model was used for the water domain, with the convection and
conduction model applied to both liquid and solid regions. Variations in the physi-
cal properties of water (expansion coefficient, density, specific heat capacity, thermal
conductivity and viscosity) as a function of temperature and pressure were obtained
from NIST/ASME database. It was found that the physical properties follow a first
order polynomial equation, with temperature and pressure as variables (Knoerzer
et al. 2010).
Figure 6.2 depicts temperature distribution for three different wall thick-
nesses (0, 5, and 70mm) at the end of pressure hold time (430s). In the case of
no carrier wall thickness, temperature does not cross 105C, due to lack of insu-
lation (Fig.6.2a). For the 5mm wall thickness, the average temperature reached
118C, with uniform temperature distribution at the end of pressure holding
time (Fig.6.2b). Therefore, insulation thickness has a major role in HP process-
ing (Knoerzer et al. 2010). For the 70mm wall thickness, the effective volume of
vessel reduced to 1.3L. Carrier walls insulate the contents from the cooler vessel
and act as heat sinks. This resulted in a temperature gradient and also heat loss
(Fig.6.2c). With the help of CFD modeling, in the near future it will be possible to
optimize different carrier materials and their thicknesses, which will enhance per-
formance of the HPT processing system. Recently, Knoerzer and Chapman (2011)
highlighted the importance of accurate input data, such as thermophysical prop-
erties of a material and their correlation to pressure and temperature and process
conditions, for precise model predictions for the HPT process.
All the above studies clearly indicate that CFD can be effectively used for the
modeling of high-pressure processing. CFD modeling helps in designing and opti-
mizing the carrier wall thickness for optimum uniform temperature distribution
and also maximum usable volume. However, lack of thermo-physical properties
data for different material and processing conditions is a major hurdle in the pre-
cise modeling of high-pressure processing.
Chapter 7
Applications of Computational Fluid
Dynamics in Other Food Processing
Operations
0.12 m
0.1 m
1.25 m
1.225 m
0.8 m
0.5 m
0.15 m
outlet
Fig.7.1Spray freezing chamber (a) photograph and (b) dimensions; (c) computational domain
and meshing (Anandharamakrishnan et al. 2010b)
each other and the gas is determined by the fluid mechanics of the spray; (2) the
heat transfer between the gas and the droplets depends on local conditions, e.g. gas
temperature, droplet temperature and droplet-gas slip velocity; and (3) the freezing
and ice crystallization within the drops.
The experimental temperature profiles of the gas are plotted in Fig.7.2ac at axial
positions of z=0.58, 0.8 and 1.23m below the nozzle. When the CFD model of
cooling gas is compared to the experiment, the cooling gas flow pattern appears to
show some difficulty in penetration through the spray region. In the experiment,
however, there may be droplets of liquid nitrogen entrained in the inlet gas flow,
which is able to penetrate into the core of the spray and provide significant cooling,
the effect of which is not included in the CFD model. Further down the chamber
at z=0.8 and 1.23m (where the conical section begins), the temperature profiles
flatten and the core region of higher temperature broadens as the spray fans out due
to (1) evaporation of liquid nitrogen and (2) super cooling/recalescence. The tem-
perature of gas set outside the core is almost uniform, and it appears that most of
the droplets do not penetrate into this zone. This trend was also observed by Kieviet
(1997) and Huang et al. (2006) in their spray drying studies.
66 7 Applications of Computational Fluid Dynamics
Temperature (K)
predictions for Case A (solid
230
cone spray) at (a) z=0.58m
(b)0.8m, and (c) 1.23m 210
from the nozzle spray point
190 Simulation-X axis
(Anandharamakrishnan et al. Simulation-Y axis
2010b) 170 Experimental data
150
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5
Radial distance (m)
(b) 270
250
Temperature (K)
230
210
Simulation-X axis
190
Simulation-Y axis
170 Experimental data
150
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5
Radial distance (m)
(c)
270
Temperature (K)
250
230
210
Simulation-X axis
190
Simulation-Y axis
170
Experimental data
150
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5
Radial distance (m)
Particle trajectories are shown in Fig.7.3, and clearly reveal that some particles
are re-circulated by the gas phase and have upward velocities close to the walls.
Knowledge about the positions where particles are having impact is important
for designing and operating spray freezing equipment. Comparisons of the simu-
lated and experimental results for particle impacts on the chamber walls are shown
in Fig.7.4ab (top views) and cd (front views). These figures indicate that a
large fraction of the particles (65%) strike the conical part of the spray freezing
7.1 CFD Simulation of Spray Freezing Operations 67
Fig.7.3CFD simulated
particle trajectories
(Anandharamakrishnan et al.
2010b)
chamber; 11% of the particles hit the cylindrical part of the wall, and only a small
proportion (22%) of the particles come directly out of the chamber. The inter-
esting observation is that no particles hit the ceiling due to gas re-circulation,
which happens only at the bottom (see Fig.7.3). These results were in reasonably
good agreement with the experimental observations shown in the photographs of
Fig. 7.4bd (Anandharamakrishnan et al. 2010b). An important point to note is
that in these experiments, a significant number of particles stick to the walls.
In spray drying simulations, it is generally assumed that (non-sticky) particles
slide down the walls toward the main product outlet. However, in spray freezing,
a little more attention is required, as when a frozen particle hit the wall during
the operation, it tended to stick and build an icy layer, as shown in Fig.7.4b. This
may be because either the ice particles (being crystalline) are rougher, or because
incomplete freezing has occurred (which is likely, considering the results for large
particles). Hence, in order to maximize the freezing efficiency, the amount of
product conveyed to the outlet, as well as maintaining a sufficiently cold wall tem-
perature during the process, needs a little more consideration.
Success and growth of any industry depends on good quality product and energy
efficiency of the process. In the food industry, heating is one of the main pro-
cess techniques used for enhancement of product shelf-life, flavor, texture, palat-
ability etc. Conventional heating or cooling methods give good quality product.
68 7 Applications of Computational Fluid Dynamics
Jet impingement is one of the techniques that utilizes energy efficiently and
reduces processing time.
In the jet impingement technique, high velocity jets of fluid are injected
through a nozzle onto a particular surface for the purpose of rapid heat and
mass transfer. When a food product is heated or cooled using a fluid medium
like air or water, a boundary layer forms. This boundary layer formation offers
7.2 CFD Modeling for Jet Impingement Oven 69
high resistance to heat transfer and makes the heating or cooling process slow.
Therefore, the heating process becomes more time and energy consuming.
However, this boundary layer problem can be solved using jet impingement, as
jet impingements have both high velocity and turbulent flow. The high velocity of
jets reduces the boundary layer between the heating surfaces and heating medium;
hence, turbulent flow increases the heat transfer in food products (Jambunathan et
al. 1992). Jet impingement heating is superior to conventional heating due to its
advantages, such as higher efficiency, high heat and mass transfer rate, rapid mois-
ture removal and uniform heating (Ovadia and Walker 1998; Li and Walker 1996;
Nitin 2009).
For optimization of the jet impingement process, it is very important to under-
stand its flow pattern, jet velocity profile, temperature distribution and effect of
different parameters on heat transfer rate. The key parameters for effective heat
transfer in jet impingement are nozzle geometry, nozzle to surface distance, veloc-
ity and temperature of jet (Garimella and Nenaydykh 1996). Some of the stud-
ies carried out by various researchers focused primarily on the effect of nozzle
geometry on heat transfer (Garimella and Nenaydykh 1996, Zhao et al. 2004),
the effects of single and multiplicity jets on heat transfer (Lou et al. 2005), the
effects of pulsating jets on heat transfer (Zulkifli et al. 2009; Kurnia et al. 2012),
etc. However, there is less understanding about the flow pattern of a jet. The effect
of surrounding fluid, jet velocity and product position on the flow pattern of a jet
needs to be explored. Moreover, it is very difficult to experimentally calculate the
heat transfer coefficient at different positions (Kocer et al. 2007). Computational
Fluid Dynamics (CFD) can be used to improve, understand and optimize the jet
impingement process. CFD modeling of the jet impingement effect of nozzle
geometry or position on heat transfer rate would be helpful in the analysis of effect
of flow pattern on heat transfer, velocity and temperature profile, etc.
Jet flow is classified into three regions: a free jet region, a stagnation region and a
wall jet region. The free jet region can be further classified into three sub-regions,
such as the potential core region, the developing flow region and the developed
flow region, as shown in Fig.7.5. As the fluid exits from nozzle, the potential core
region starts; it is a very short region, and the velocity of jet is almost similar to
nozzle exit velocity and has no influence of surrounding air. Further, surrounding
air starts mixing with jet air, resulting in the formation of turbulent peaks. As the
downstream distance of the jet increases, velocity decreases. The potential core,
developing and developed regions altogether constitute a free jet region. Further,
when fluid jet reaches near the wall, the radial distance of the jet increases rapidly
and velocity decreases, and this region called the wall jet region. Finally, the stag-
nant region is formed due to decrease in jet velocity to zero (Nitin 2009; Gardon
and Akfirat 1965).
70 7 Applications of Computational Fluid Dynamics
Selecting the jet configuration is very important, since nozzle shape and its geom-
etry directly affect the length of the potential core region and the rate of energy
dissipation (Jambunathan et al. 1992). Gariella and Nenaydykh (1996) reported
that as nozzle to heating surface spacing increases, the effect of nozzle aspect ratio
on the heat transfer coefficient is almost negligible. Compared to the aspect ratio,
axial length to diameter of nozzle (X/D) and radial length to diameter of nozzle
ratios (H/D) show much effect on the heat transfer coefficient (Sibulkin 1952).
Angioletti et al. (2005) studied CFD modeling of jet impingement at differ-
ent Reynolds number values using three different turbulent models (Reynolds
Normalized Group k-, k- Shear Stress Transport and Reynolds Stress
Model), and validated with experimental data. Reynolds number values were
between 1,000 and 4,000, and three models were compared with Particle Image
Velocimetry (PIV) flow field at Re-1000 and h/d-4.5. Similarities between the
results of the CFD (k) model and PIV can be observed.
Recently, Ma et al. (2012) developed a CFD model for a counter-current reactor
and a confined jet mixer of a continuous hydrothermal flow synthesis (CHFS) sys-
tem. They studied the effect of different operating conditions and reactor geometry
on the flow field, temperature and concentration. They used the species transport
equation coupled with the turbulent model k-. Velocity and temperature distribu-
tion were analyzed for the supercritical exit region.
Jafari and Alavi (2008) investigated the effect of various parameters on freezing
time of slab-shaped food product, using air jet impingement with the help of CFD
modeling. The RSM turbulent model was used for the modeling of turbulent flow of
jet. In this study, the authors analyzed the effect of jet air temperature on temperature
distribution, and jet air velocity on the flow and velocity field of the system. Reduction
in freezing time was observed with the lowering of air temperature. However, as noz-
zle-to-surface spacing decreases, freezing time increases. Recently, Kurnia et al. (2012)
analyzed the effect of pulsating impinging-jet on drying performance. Velocity vec-
tors in the drying chamber for steady laminar and turbulent jets are shown in Fig.7.6
(Kurnia et al. 2012). The RSM turbulent model was used for simulating turbulent flow
7.2 CFD Modeling for Jet Impingement Oven 71
Fig.7.6Velocity vectors in the drying chamber for steady laminar (a) and turbulent (b) jets
(Kurnia et al. 2012)
of the impinging-jet dryer. The authors studied moisture content distribution, and their
results showed uniform moisture distribution for a thin slab after 30min of drying.
CFD modeling has a huge potential to address the problems faced by large scale
grain processing industries. It is possible to model unit operations in the flour mill-
ing industry, where problems in handling and operating the equipment are encoun-
tered. CFD is also used in optimizing the processing conditions, so that losses in
time, raw material and energy can be minimized. This increases the efficiency of the
machine and decreases the economy (cost involved) in producing the desired prod-
uct. However, the amount of research employed in this perspective is very limited.
A spouted bed dryer (an alternative to fluidized bed dryer) is used in drying coarse
particles (such as grain) with narrower particle size distribution. Generally, three types
of spouted beds shapes are used: (1) cylindrical (2) conical cylindrical, and (3) slot-
rectangular (Cui and Grace 2008). A spouted bed dryer consists of injecting a high-
speed gasfluid that moves through a bed of solids, pushing and forcing the solids to
the center of the container until reaching the upper level of the bed, where the solids
later fallthrough the effects of gravityas a rain of particles forming a spout. One of
the drawbacks with conventional spouted beds is the need for control of particle trajec-
tory (Claflin and Fane 1984). However, CFD models can be used to solve this prob-
lem, and CFD models for spouted bed dryers can provide important information on the
flow field within the spouted beds. This information can be utilized for process design,
scale-up, optimization, and importantly, for reducing the need for experimentations.
A CFD model for the cylindrical spouted bed with spherical particles was
developed using the EulerianEulerian two-fluid modeling approach to study the
72 7 Applications of Computational Fluid Dynamics
Fig.7.7Volume fraction distribution of grain for different air inlet velocities (Sobieski 2008)
gas-particle flow behavior (Zhonghua and Mujumdar 2008). Heat and mass trans-
fer during grain drying in a spouted bed dryer was studied using a CFD (Szafran
and Kmiec 2004). The EulerianEulerian multiphase approach was used to predict
gassolid flow behavior. This model predicts mass transfer very well, but under-
predicts heat transfer. Further, this CFD model was extended to study spouted-bed
dryer hydrodynamics (Szafran et al. 2005). This model explains the interactions
between the jet, fountain, and the annulus region and two phase flow of gassolid
very well, while traditional techniques fail to explain these interactions.
Flow behavior of grain, phase circulation and cluster formation in the load-
ing region were predicted to be analogous to experimental observations. Sobieski
(2008) used the Eulerian multiphase approach to model grain drying in a spouted
bed dryer. CFD model predictions of volume fraction distribution of grain for vari-
ous inlet air velocities are shown in Fig.7.7 (Sobieski 2008). In this study, Sobieski
found that inlet air velocities play a crucial role in the behaviour of particles inside
the spouted bed dryer. Figure7.8 shows the distribution of volume fraction for vari-
ous turbulence models (Sobieski 2008). Each turbulence model predicts a different
fountain height. The effect of grain density, diameter, change in volume and packing
coefficient on the distribution of the grain height was studied by Sobieski (2008).
Recently, Sobieski (2010) discussed the flow analysis inside the spouted bed
dryer in terms of creation of geometry; selection of computational space and grid;
the Eulerian multiphase model; and the turbulence model and its sensitivity to
phase changes and flow parameters. Further, spout, fountain and annular regions
were well predicted by the above model (Fig.7.9). This study was performed
using an Eulerian multiphase model.
7.3 Application of CFD Modeling in the Flour Milling Industry 73
Fig.7.9Volume fraction occupied by the granular phase for differently sized structural grids
(from the left: 2,160; 4,284; 6,615; 11,732; 17,136; 22,347; 27,090; 32,900; 40,144; and 47,880)
(Sobieski 2010)
74 7 Applications of Computational Fluid Dynamics
Flour mills normally use fumigation to kill insects that can create damage to the
grain. The fumigant used was methyl bromide. This process can be structurally
modeled using CFD, where the model will have two phases, such as external flow
model and internal flow model. The external flow model includes the flour mill
and surrounding structures to predict stagnation pressures on the mills walls as a
function of wind speed and direction data. The internal flow model includes inte-
rior details of the mill, such as building plans, and locations of major equipment,
partitions, piping and ducting. The cracks are represented as an effective leakage
zone (Chayaprasert et al. 2008). The locations of fumigant introduction sites and
circulation fans are also accounted for in the model. Chayaprasert et al. (2008)
concluded that the CFD models developed in their study are valid, and that the
established methodology can be utilized for any type of structure for modeling of
fumigation process.
Thus, CFD is a promising tool that can be used for troubleshooting unit oper-
ations. It can also help in optimizing conditions of the equipment used in roller
flour mill industries. For optimizing operating conditions, experiments are usually
performed on the equipment itself; this is time-consuming, laborious and expen-
sive. However, these problems can be solved by using CFD to model the equip-
ment (or system, rather than performing experiments on the system itself). Hence,
CFD is a powerful and pervasive tool for process and product development in the
food-processing sector.
References
Birchal VS, Huang L, Mujumdar AS, Passos ML (2006) Spray dryers- modeling and simulation.
Dry Technol 24:359371
Bird B, Stewart WE, Lightfoot EN (1960) Transport phenomena. Wiley, New York
Boulet M, Marcos B, Dostie M, Moresoli C (2010) CFD modeling of heat transfer and flow field
in a bakery pilot oven. J Food Eng 97:393402
Camilleri M, Prather CM (1993) Gastric motor physiology and motor disorders. In: Feldmann M,
Scharschmidt BF, Sleisenger MH (eds) Gastrointestinal and liver disease: pathophysiology/
diagnosis/management, 6th edn. WB Saunders Co., Philadelphia, pp 572586
Cauvain SP (2003) Bread making-improving quality. CRC Press, New York
Charm SE (1971) The fundamentals of food engineering. The AVI Publishing Company,
Westport
Chayaprasert W, Maier DE, Ileleji KE, Murthy JY (2008) Development and validation of com-
putational fluid dynamics models for precision structural fumigation. J Stored Prod Res
44:1120
Chen XD, Xie GZ (1997) Fingerprints of the drying behavior of particulate or thin layer food
materials established using a reaction engineering model. Trans Inst Chem Eng, Part C
75:213222
Chen XD, Jin Y (2009) Numerical study of the drying process of different sized particles in an
industrial-scale spray dryer. Dry Technol 27:371381
Chen XD, Jin Y (2009) A three-dimensional numerical study of the gas/particle interactions in an
industrial-scale spray dryer for milk powder production. Dry Technol 27:10181027
Chen XD, Pirini W, Ozilgen M (2001) The reaction engineering approach to modeling drying of
thin layer of pulped kiwi fruit flesh under conditions of small Biot numbers. Chem Eng Prog
40:165181
Chhanwal N, Anishaparvin A, Indrani D, Raghavarao KSMS, Anandharamakrishnan C (2010)
Computational fluid dynamics (CFD) modeling of an electrical heating oven for bread-baking
process. J Food Eng 100:452460
Chhanwal N, Indrani D, Raghavarao KSMS, Anandharamakrishnan C (2011) Computational
fluid dynamics modeling of bread baking process. Food Res Int 44:978983
Chhanwal N, Tank A, Raghavarao KSMS, Anandharamakrishnan C (2012) Computational
fluid dynamics (CFD) modeling for bread baking process-a review. Food Bioprocess Tech
5:11571172
Claflin JK, Fane AG (1984) Fluid mechanics, heat transfer in spouted beds with draft tubes. Dry
137141
Claussen C, Ustad TS, Strommen I, Walde PM (2007) Atmospheric freeze dryinga review. Dry
Technol 25:957967
Cortella G, Manzan M, Comini G (1998) Computation of air velocity and temperature distribu-
tions in open display cabinets. In: Advances in the refrigeration systems, food technologies
and cold chain. International Institute of Refrigeration, Paris
Crowe CT, Sharam MP, Stock DE (1977) The particle source in cell (PSI-Cell) model for gas-
droplet flows. J Fluid Eng 9:325332
Cui H, Grace JR (2008) Spouting of biomass particles: a review. Bioresource Technol
99:40084020
Datta AK, Teixeira AA (1987) Numerical modelling of natural convection heating in canned liquid
foods. Trans ASAE 30(5):15421551
Davey LM, Pham QT (1997) Predicting the dynamic product heat load and weight loss during
beef chilling using a multi-region finite difference approach. Int J Refrig 20(7):470482
Davey LM, Pham QT (2000) A multi-layered two-dimensional finite element model to calculate
dynamic product heat load and weight loss during beef chilling. Int J Refrig 23(6):444456
Denys S, Pieters JG, Dewettinck K (2003) Combined CFD and experimental approach for deter-
mination of the surface heat transfer coefficient during thermal processing of eggs. J Food Sc
68:943951
Denys S, Pieters JG, Dewettinck K (2004) Computational fluid dynamics analysis of combined
conductive and convective heat transfer in model eggs. J Food Eng 63:281290
References 77
Denys S, Dewettinck K, Pieters JG (2005) CFD analysis for process impact assessment during
thermal pasteurization of intact eggs. J Food Protect 68:366374
DeSesso JM, Jacobson CF (2001) Anatomical and physiological parameters affecting gastroin-
testinal absorption in humans and rats. Food Chem Toxicol 39:209228
DeVries U, Velthuis H, Koster K (1994) Baking ovens and product quality: a computer model.
Food Sci Technol Today 9(4):232234
Dhall A, Datta AK, Torrance KE, Almeida MF (2009) Radiative heat exchange modeling inside
an oven. AIChE J 55:24482460
Dimou A, Yanniotis S (2011) 3D numerical simulation of asparagus sterilization in a still-can
using computational fluid dynamics. J Food Eng 104:394403
Dokoumetzidis A, Valsami G, Macheras P (2007) Modelling and simulation in drug absorption
processes. Xenobiotica 37(1011):10521065
Ducept F, Sionneau M, Vasseur J (2002) Superheated steam dryersimulations and experiments
on product drying. Chem Eng J 86:7583
Einhorn M (1898) Anatomy and physiology. In: Diseases of the stomach: a text-book for practi-
tioners and students. William Wood and Company, New York
Erdogdu F, Tutar M (2010) Velocity and temperature field characteristics of water and air during
natural convection heating in cans. J Food Sci 76(1):119129
Farid M, Ghani AG (2004) A new computational technique for the estimation of sterilization time
in canned food. Chem Eng Process 43:523531
Fellows PJ (1998) Food processing technology, 2nd edn. CRC Press, New York
Ferrua MJ, Singh RP (2010) Modeling the fluid dynamics in a human stomach to gain insight of
food digestion. J Food Sci 75(7):151161
Ferrua MJ, Kong F, Singh RP (2011) Computational modelling of gastric digestion and the role
of food material properties. Trends Food Sci Tech 112
Fletcher AJ (2000) Computational techniques for fluid dynamics, 2nd edn. Springer, New York
Fluent (2006) Fluent users guide. Ansys Inc, USA
Gabites JR, Abrahamson J, Winchester JA (2010) Air flow patterns in an industrial milk powder
spray dryer. Chem Eng Res Des 88(7):899910
Gardon R, Akfirat JC (1965) Role of turbulence in determining the heat transfer characteristics of
jet impingement. Int J Heat Mass Tran 8:12611272
Garimella SV, Nenaydykh B (1996) Nozzle-geometry effects in liquid jet impingement heat
transfer. Int J Heat Mass Tran 39:29152923
Geliebter A, Mellon PM, McCray RS, Gallagher DR, Gage D, Hashim SA (1992) Gastric capacity,
gastric emptying, and test-meal intake in normal and bulimic women. Am J Clin Nutr
56(4):656661
Ghani AG, Farid MM, Chen XD, Richards P (1999) Numerical simulation of natural convection
heating of canned food by computational fluid dynamics. J Food Eng 41(1):5564
Ghani AG, Farid MM, Chen XD, Richards P (1999) An investigation of deactivation of bacteria
in a canned liquid food during sterilisation using computational fluid dynamics (CFD). J Food
Eng 42(4):207214
Ghani AG, Farid MM, Che XD, Richards P (2001) Thermal sterilization of canned food in a 3-D
pouch using computational fluid dynamics. J Food Eng 48:147156
Ghani AG, Farid MM, Chen XD (2002) Numerical simulation of transient temperature and
velocity profiles in a horizontal can during sterilization using computational fluid dynamics. J
Food Eng 51:7783
Ghani AG, Farid MM, Chen XD (2002) Theoretical and experimental investigation of the thermal
inactivation of Bacillus stearothermophilus in food pouches. J Food Eng 51:221228
Ghani AG, Farid MM, Chen XD (2002) Theoretical and experimental investigation of the thermal
destruction of Vitamin C in food pouches. Comput Electron Agric 34:129143
Ghani AG, Farid MM, Chen XD (2003) The effect of can rotation on the thermal sterilization of
liquid food using computational fluid dynamics. J Food Eng 57:916
Ghani AG, Farid MM (2006) Using the computational fluid dynamics to analyze the thermal ster-
ilization of solid-liquid food mixture in cans. Innov Food Sci Emerg 7:5561
78 References
Ghani AGA, Farid MM (2007a) Modeling of high pressure food processing using CFD. In: Sun
DW (ed) Computational fluid dynamics in food processing. CRC Press, Boca Raton
Ghani AGA, Farid MM (2007) Numerical simulation of solid-liquid food mixture in a high pres-
sure processing unit using computational fluid dynamics. J Food Eng 80:10311042
Goula AM, Adamopoulos KG (2004) Influence of spray drying conditions on residue accumula-
tion-simulation using CFD. Dry Technol 22:11071128
Guyton AC, Hall JE (2006) Textbook of medical physiology. Elsevier, Philadelphia
Heldman DR (1974) An analysis of atmospheric freeze drying. J Food Sci 39:147155
Hiiemae K, Palmer JB (1999) Food transport and bolus formation during complete feeding
sequences on foods of different initial consistency. Dysphagia 14:3142
Hindmarsh JP, Russell AB, Chen XD (2003) Experimental and numerical analysis of the tem-
perature transition of a suspended freezing water droplet. Int J Heat Mass Tran 46:11991213
Holdsworth D, Simpson R (2007) Thermal processing of packaged foods. Food engineering
series, 2nd edn. Springer, New York
Hu Z, Sun DW (1999) The temperature distribution of cooked meat joints in an air-blast chiller
during cooling process: CFD simulation and experimental verification. Paper presented at the
20th international congress of refrigeration, Sydney
Hu Z, Sun DW (2000) Simulation of heat and mass transfer for vacuum cooling of cooked meats
by using computational fluid dynamics code. Paper presented at the 8th international congress
on engineering and food, Puebla
Huang LX, Kumar K, Mujumdar AS (2003) Use of computational fluid dynamics to evaluate
alternative spray chamber configurations. Dry Technol 21:385412
Huang LX, Kumar K, Mujumdar AS (2003) A parametric study of the gas flow patterns and drying
performance of co-current spray dryer: results of a computational fluid dynamics study. Dry
Technol 21(6):957978
Huang LX, Kumar K, Mujumdar AS (2004) Simulation of a spray dryer fitted with a rotary
disk atomizer using a three-dimensional computational fluid dynamic model. Dry Technol
22(6):14891515
Huang LX, Passos ML, Kumar K, Mujumdar AS (2005) A three dimensional simulation of a
spray dryer fitted with a rotary atomizer. Dry Technol 23:18591873
Huang LX, Kumar K, Mujumdar AS (2006) A comparative study of a spray dryer with rotary
disc atomizer and pressure nozzle using computational fluid dynamic simulations. Chem Eng
Process 45:461470
Huang LX, Mujumdar AS (2007) Simulation of an industrial spray dryer and prediction of off-
design performance. Dry Technol 25:703714
Jafari M, Alavi P (2008) Analysis of food freezing by slot jet impingement. J Appl Sci
8(7):11881196
Jakobsen HA, Sannaes BH, Grevskott S, Svendsen HF (1997) Modelling of vertical bubble-
driven flows. Ind Eng Chem Res 36:40524074
Jambunathan K, Lai E, Moss MA, Button BL (1992) A review of heat transfer data for single cir-
cular jet impingement. Int J Heat Fluid Fl 13(2):10615
Juliano P, Knoerzer K, Fryer P, Versteeg C (2009) C. botulinum inactivation kinetics implemented
in a computational model of a high pressure sterilization process. Biotechnol Prog 25:163175
Kannan A, Sandaka PCHG (2008) Heat transfer analysis of canned food sterilization in still
retort. J Food Eng 88:213228
Keet AD (1993) Infantile hypertrophic pyloric stenosis. In: The pyloric sphincteric cylinder in
health and disease. Springer, Berlin
Khurana M, Karwe MV (2009) Numerical prediction of temperature distribution and measure-
ment of temperature in a high hydrostatic pressure food processor. Food Bioprocess Tech
2:279290
Kieviet FG, Kerkhof PJAM (1995) Measurements of particle residence time distributions in a co-
current spray dryer. Dry Tech 13(57):12411248
Kieviet, FG, Kerkhof PJAM (1996) In: Mujumdar AS (ed) Using computational fluid dynamics
to model product quality in spray drying: air flow temperature and humidity patterns. In
Drying96, vol A, Krakow, pp 259266
References 79
Kieviet FG (1997) Modeling quality in spray drying. Ph.D. thesis, Endinhoven University of
Technology, Netherlands
Kim DY, Camilleri M, Murray JA, Stephens DA, Levine JA, Burton DD (2001) Is there a
role for gastric accommodation and satiety in asymptomatic obese people? Obesity Res
9(11):655661
Kiziltas S, Erdogdu F, Palazoglu TK (2010) Simulation of heat transfer for solidliquid food mix-
tures in cans and model validation under pasteurization conditions. J Food Eng 97:449456
Knoerzer K, Juliano P, Gladman S, Versteeg C, Fryer PJ (2007) A computational model for tem-
perature and sterility distributions in a pilot-scale high-pressure high-temperature process.
AIChE J 53:29963010
Knoerzer K, Buckow R, Juliano P, Chapman B, Versteeg C (2010) Carrier optimisation in a pilot-
scale high pressure sterilization plantan iterative CFD approach. J Food Eng 97:199207
Knoerzer K, Chapman B (2011) Effect of material properties and processing conditions on the
prediction accuracy of a CFD model for simulating high pressure thermal (HPT) processing. J
Food Eng 104:404413
Kocer D, Nitin N, Karwe MV (2007) Application of CFD in jet impingement oven. In: Sun DW
(ed) Computational fluid dynamics in food processing. CRC Press, Boca Raton, pp 469485
Kong F, Singh RP (2008) A model stomach system to investigate disintegration kinetics of solid
foods during gastric digestion. J Food Sci 73(5):210
Kong F, Singh RP (2010) A human gastric simulator (hgs) to study food digestion in human
stomach. J Food Sci 75(9):E627E635
Koribilli N, Aravamudan K, Varadhan A (2011) Quantifying enhancement in heat transfer due to
natural convection during canned food thermal sterilization in a still retort. Food Bioprocess
Tech 4:429450
Koutchma T (2012) Design for high-pressure processing. In: Ahmed J, Rahman MS (eds)
Handbook of food process design. Wiley-Blackwell, Oxford, pp 9981030
Kumar A, Bhattacharya M, Blaylock J (1990) Numerical simulation of natural convection heat-
ing of canned thick viscous liquid food products. J Food Sci 55(5):1403
Kumar A, Bhattacharya M (1991) Transient temperature and velocity profiles in a canned
Non-Newtonian liquid food during sterilization in a still-cook retort. Int J Heat Mass Tran
34(45):10831096
Kuo KKY (1986) Principles of combustion. Wiley, New York
Kuriakose R, Anandharamakrishnan C (2010) Computational fluid dynamics applications in
spray drying of food products. Trends Food Sci Tech 21:383398
Kurnia JC, Sasmito AP, Tong W, Mujumdar AS (2012) Energy- efficient thermal drying
using impinging-jets with time-varying heat inputa computational study. J Food Eng
114:269277
Langrish TAG, Oakley DE, Keey RB, Bahu RE, Hutchinson CA (1993) Time-dependent flow
patterns in spray dryers. Trans Inst Chem Eng Part A 71:355360
Langrish TAG, Zbicinski I (1994) The effects of air inlet geometry and spray cone angle on the
wall deposition rates in spray dryer. Trans Inst Chem Eng Part A 72:420430
Langrish TAG, Fletcher DF (2001) Spray drying of food ingredients and applications of CFD in
spray drying. Chem Eng Process 40:345354
Langrish TAG, Fletcher DF (2003) Prospects for modelling and design of spray dryers in the 21st
century. Dry Technol 21:197215
Langrish TAG (2007) New engineered particles from spray dryers: research needs in spray drying.
Dry Technol 25:981993
Langrish TAG, Williams J, Fletcher DF (2004) Simulation of the effects of swirl on gas flow pat-
terns in a pilot-scale spray dryer. Trans Inst Chem Eng Part A 82(A7):821833
Launder BE, Spalding DB (1972) Lectures in mathematical models of turbulence. Academic
Press, London
Leuenberger L, Plitzko M, Puchkov M (2006) Spray freeze drying in fluidized bed at normal and
low pressure. Dry Technol 7:11719
Li A, Walker CE (1996) Cake baking in conventional, impingement and hybrid ovens. J Food Sci
61(1):188191
80 References
Liu L, Fishman ML, Kost J, Hicks KB (2003) Pectin-based systems for colon-specific drug deliv-
ery via oral route. Biomaterials 24:33333343
Lou ZQ, Mujumdar AS, Yap C (2005) Effects of geometric parameters on confined impinging jet
heat transfer. Appl Therm Eng 25:26872697
Love RJ, Lentle RG, Asvarujanon P, Hemar Y, Stafford KJ (2012) An expanded finite element
model of the intestinal mixing of digesta. Food Digest 110. doi: 10.1007/s13228-012-0017-x
Ma CY, Wang XZ, Tighe CJ Darr JA (2012) Modelling and simulation of counter-current and
confined jet reactors for continuous hydrothermal flow synthesis of nano-materials. 8th IFAC
symposium on advanced control of chemical processes Furama Riverfront, Singapore
MacLeod CS, McKittrick JA, Hindmarsh JP, Johns ML, Wilson DI (2006) Fundamentals of spray
freezing of instant coffee. J Food Eng 74:451461
Mahler GJ, Esch MB, Tako E, Southard TL, Archer SD, Glahn RP, Shuler ML (2012) Oral expo-
sure to polystyrene nanoparticles affects iron absorption. Nat Nanotechnol 7(4):164271
Malecki GJ, Shinde P, Morgan AI, Farkas DF (1970) Atmospheric fluidized bed freeze drying.
Food Technol 24:601603
Mariotti M, Rech G, Romagnoni P (1995) Numerical study of air distribution in a refriger-
ated room. Proceedings of international congress of refrigeration, vol 2. Den Hague, The
Netherlands, pp 98105
Marshall EM, Bakker A (2002) Computational fluid mixing. Fluent Inc, Lebanon
Masters K (1991) Spray drying. Wiley, Essex
Mayer EA (1994) The physiology of gastric storage and emptying. In: Johnson L (ed)
Physiology of the gastrointestinal tract, 3rd edn. Raven Press, New York, pp 92976
Mezhericher M, Levy A, Borde I (2008) Droplet-droplet interactions in spray drying by using 2D
computational fluid dynamics. Dry Technol 26:265282
Mezhericher M, Levy A, Borde I (2009) Modeling of droplet drying in spray chambers using 2D
and 3D computational fluid dynamics. Dry Technol 27:359370
Mills D (19981999) Development and validation of a preliminary model for optimisation of
baking ovens. The Food and Packaging Cooperative Research Centre Annual Report (1998-
1999), Australia
Mistry H, Ganapathi-subbu S, Dey S, Bishnoi P, Castillo JL (2006) Modeling of transient natural
convection heat transfer in electric ovens. Appl Therm Eng 26:24482456
Morsi SA, Alexander AJ (1972) An investigation of particle trajectories in two-phase flow sys-
tems. J Fluid Mech 55(2):193208
Mostafa AA, Mongia HC (1987) On the modeling of turbulent evaporating sprays: Eulerian ver-
sus Lagrangian approach. Int J Heat Mass Tran 30(12):25832593
Moureh J, Derens E (2000) Numerical modelling of the temperature increase in frozen food
packaged in pallets in the distribution chain. Int J Refrig 23(7):540552
Nijdam JJ, Guo B, Fletcher DF, Langrish TAG (2004) Challenges of simulating droplet coales-
cence within a spray. Dry Technol 22(6):14631488
Nijdam JJ, Guo B, Fletcher DF, Langrish TAG (2006) Lagrangian and Eulerian models for sim-
ulating turbulent dispersion and coalescence of droplets within a spray. Appl Math Model
30:11961211
Nitin N (2009) A dissertation on transport phenomenon in jet impingement baking. Ph.D Thesis,
Graduate School-New Brunswick Rutgers, The State University of New Jersey
Norton IT, Fryer PJ, Moore S (2006) Product/process integration in food manufacture: engineer-
ing sustained health. AIChE J 52(5):16321640
Norton T, Sun DW (2006) Computational fluid dynamics (CFD)an effective and efficient
design and analysis tool for the food industry: a review. Trends Food Sci Tech 17:600620
Oakley DE, Bahu RE (1993) Computational modelling of spray dryers. Comp Chem Eng
17:493498
Ovadia DZ, Walker CE (1998) Impingement in food processing. Food Technol 52(4):4650
Padmavathi R, Anandharamakrishnan C (2012) Computational fluid dynamics modeling of the
thermal processing of canned pineapple slices and titbits. Food Bioprocess Tech 6(4):882895
References 81
Spratt P, Nicolella C, Pyle DL (2005) An Engineering Model of the Human Colon. Trans
IChemE C 83:147157
Stoll BR, Richard PB, Leipold HR, Milstein S, Edwards DA (2000) A theory of molecular
absorption from the small intestine. Chem Eng Sci 55:473489
Straatsma J, Van Houwelingen G, Steenbergen AE, De Jong P (1999) Spray drying of food products:
1. Simulation model. J Food Eng 42(2):6772
Sun DW (2007) Computational fluid dynamics in food processing. CRC Press, Taylor and
Francis group, Boca Raton
Szafran G, Kmiec A (2004) CFD modeling of heat and mass transfer in a spouted bed dryer. Ind
Eng Chem Res 43:11131124
Szafran G, Kmiec A, Ludwig W (2005) CFD modeling of a spouted-bed dryer hydrodynamics.
Dry Technol 23:17231736
Tank A, Chhanwal N, Indrani D, Anandharamakrishnan C (2012) Computational fluid dynam-
ics modeling of bun baking process under different oven load conditions. J Food Sci Tech.
doi:10.1007/s13197-012-0736-6
Tannehill JC, Anderson DA, Pletcher RH (1997) Computational fluid mechanics and heat transfer,
2nd edn. Taylor & Francis, Philadelphia
Tattiyakul J, Rao MA, Datta AK (2002) Heat transfer to a canned corn starch dispersion under
intermittent agitation. J Food Eng 54:321329
Tharakan A, Rayment P, Fryer PJ, Norton IT (2007) Modelling of physical and chemical pro-
cesses in the small intestine. Proceedings of European congress of chemical engineering
(ECCE-6), Copenhagen, pp 1620
Therdthai N, Zhou W (2003) Recent advances in the studies of bread baking process and their
impact on the bread baking technology. Food Sci Tech Res 9(3):219226
Therdthai N, Zhou W, Adamczak T (2003) Two dimensional CFD modeling and simulation of an
industrial continuous bread baking oven. J Food Eng 60:211217
Therdthai N, Zhou W, Adamczak T (2004) Simulation of starch gelatinization during baking in
a travelingtray oven by integrating a three dimensional CFD model with a kinetic model. J
Food Eng 65:543550
Therdthai N, Zhou W, Adamezak T (2004) Three-dimensional CFD modeling and simulation of
the temperature profiles and airflow patterns during a continuous industrial baking process. J
Food Eng 65:599608
Ullum T (2006) Simulation of a spray dryer with a rotary atomizer: the appearance of vortex
breakdown. In: Farkas I, Mujumdar AS (eds) Drying 2006proceedings of the 15th interna-
tional drying symposium. Budapest, Hungary, pp 251257
Vanina FM, Lucasa T, Trystram G (2009) Crust formation and its role during bread baking.
Trends Food Sci Tech 20:333343
Varma MN, Kannan A (2005) Enhanced food sterilization through inclination of the container
walls and geometry modifications. Int J Heat Mass Tran 48(18):37533762
Varma MN, Kannan A (2006) CFD studies on natural convective heating of canned food in coni-
cal and cylindrical containers. J Food Eng 77:10241036
Verboven P, Nicolai B, Delele M, Schenk A, Ramon H (2009) Evaluation of a chicory root cold
store humidification system using computational fluid dynamics. J Food Eng 94(1):110121
Versteeg HK, Malalasekera W (1995) An introduction to computational fluid dynamics. Pearson
Education Ltd, Essex, England
Verboven P, Scheerlinck N, de Baerdemaeker J, Nicolai BM (2000) Computational fluid dynamics
modelling and validation of the isothermal airflow in a forced convection oven. J Food
Eng 43:4153
Verboven P, Nicolai B, Delele M, Tijskens E, Atalay Y, Ho Q, Ramon H (2008) Combined dis-
crete element and CFD modelling of airflow through random stacking of horticultural prod-
ucts in vented boxes. J Food Eng 89(1):3341
Wagner MJ, Lucas T, Le Ray D, Trystram G (2007) Water transport in bread during baking. J
Food Eng 78:11671173
References 83
Wang L, Sun DW (2003) Recent developments in numerical modelling of heating and cooling
processes in the food industrya review. Trends Food Sci Tech 14:408423
Weng JZ (2006) Thermal processing of canned foods. In: Sun DW (ed) Thermal food processing,
CRC Publications, Boca Raton, p 335
Williamson ME, Wilson DI (2009) Development of an improved heating system for industrial
tunnel baking ovens. J Food Eng 91:6471
Wong SY, Zhou W, Hua J (2006) Robustness analysis of CFD model to the uncertainties in its
physical properties for a bread baking process. J Food Eng 77:784791
Wong SY, Zhou W, Hua J (2007) CFD modeling of an industrial continuous bread-baking process
involving U-movement. J Food Eng 78:888896
Wong SY, Zhou W, Hua J (2007) Designing process controller for a continuous bread baking
process based on CFD modeling. J Food Eng 81:523534
Woo MW, Daud WRW, Mujumdar AS, Talib MZM, Wu ZH, Tasirin SM (2008) CFD evalu-
ation of droplet drying models in a spray dryer fitted with a rotary atomizer. Dry Technol
26:11801198
Woo MW, Daud WRW, Mujumdar AS, Talib MZM, Wu ZH, Tasirin SM (2009) Non-swirling
steady and transient flow simulations in short-form spray dryers. Chem Prod Process Model
4(1):34, Article 20
Xia B, Sun DW (2002) The application of computational fluid dynamics (CFD) in the food
industry: a review. Comput Electron Agric 34:524
Xue Z, Ferrua MJ, Singh RP (2012) Computational fluid dynamics modeling of granular flow in
human stomach. Alimentos Hoy 21(27):314
Zanoni B, Peri C, Bruno D (1995) Modeling of starch gelatinization kinetics of bread crumb during
baking. LWT-Food Sci Technol 28:314318
Zanoni B, Peri C, Bruno D (1995) Modeling of browning kinetics of bread crust during baking.
LWT-Food Sci Technol 28:604609
Zbicinski I (1995) Development and experimental verification of momentum, heat and mass
transfer model in spray drying. Chem Eng J 58:123133
Zbicinski I, Strumillo C, Delag A (2002) Drying kinetics and particle residence time in spray
drying. Dry Technol 20(9):17511768
Zhang J, Datta AK (2006) Mathematical modeling of bread baking process. J Food Eng 75:7889
Zhang Z, Zhang W, Zhai Z, Chen QY (2007) Evaluation of various turbulence models in predict-
ing airflow and turbulence in enclosed environments by CFD: Part 2-comparison with experi-
mental data from literature. HVAC&R Res 13(6):871886
Zhao W, Kumar K, Mujumdar AS (2004) Flow and heat transfer under confined noncircular tur-
bulent impinging jets. Dry Technol 22(9):20272049
Zhonghua W, Mujumdar AS (2008) CFD modeling of the gas-particle flow behavior in spouted beds.
Powder Technol 183:260272
Zhou W, Therdthai N (2007) Three-dimensional modeling of a continuous industrial baking process.
In: Sun DW (ed) Computational fluid dynamics in food processing. CRC Press, Boca Raton,
pp 287312
Zulkifli R, Sopian K, Abdullah S, Takriff MS (2009) Comparison of local Nusselt number for
steady and pulsating circular jet at Reynolds number of 16000. Eur J Sci Res 29:369378
Index
P T
Particle impact positions, 23, 66 Temperature distribution, 9, 29, 33, 36, 42, 43,
Particle residence time distribution, 21, 22 60, 62, 70
Particle tracking, 6, 19 The freezing and ice crystallization, 64
Population balance method, 25 Thermal processing
pulsating impinging-jet, 70 pasteurization, 9, 32, 33, 59
Pylorus valve, 50 sterilization methods (in-container and
in-flow), 27
heat resistance
R Bacillus stearothermophilus, 29
Radiation models, 40, 42 Clostridium Botulinum, 27, 60
DTRM (discrete transfer radiation model), canning
42 slow heating zone, 9, 2731, 34
S2S (surface to surface) and, 40, 42 movement of, 36
DO (discrete ordinate), 40, 42 viscous drag, 34
Radiative heat transfer, 38, 42 Thermally sensitive materials, 63
Rapid moisture removal, 69 Transport equations
Reference Frames conservation of mass, 13, 52
volume of fluid (VOF), 4 momentum equation, 3, 32
Eulerian-Eulerian (EE) and, 4 energy equation, 3, 4, 32
Eulerian-Lagrangian (EL) models, 6, 19 Turbulence model
Rheological properties, 49, 53 standard k- (k- kinetic energy and - dis-
sipation rate), 4, 14, 64, 73
RNG k-, 4, 14, 16, 73
S realizable k-
Saliva, 49 Reynolds Stress Model (RSM), 4, 15, 70
Simultaneous heat and mass transfer, 38 Types of spray dryer, 13
Small intestine
movement of small intestinal wall, 50, 57