(Ebooks PDF) Download Transport Phenomena in Food Processing Food Preservation Technology 1st Edition Jorge Welti-Chanes Full Chapters
(Ebooks PDF) Download Transport Phenomena in Food Processing Food Preservation Technology 1st Edition Jorge Welti-Chanes Full Chapters
(Ebooks PDF) Download Transport Phenomena in Food Processing Food Preservation Technology 1st Edition Jorge Welti-Chanes Full Chapters
com
https://ebookname.com/product/transport-phenomena-in-food-
processing-food-preservation-technology-1st-edition-jorge-
welti-chanes/
OR CLICK BUTTON
DOWLOAD NOW
https://ebookname.com/product/advances-in-fresh-cut-fruits-and-
vegetables-processing-food-preservation-technology-1st-edition-
olga-martin-belloso/
https://ebookname.com/product/handbook-of-vegetable-preservation-
and-processing-food-science-and-technology-1st-edition-y-h-hui/
https://ebookname.com/product/functional-polymers-in-food-
science-from-technology-to-biology-volume-2-food-processing-1st-
edition-giuseppe-cirillo/
https://ebookname.com/product/food-preservation-techniques-p-
zeuthen/
Food Processing Technology Principles and Practice
Third Edition P. J. Fellows
https://ebookname.com/product/food-processing-technology-
principles-and-practice-third-edition-p-j-fellows/
https://ebookname.com/product/handbook-of-food-preservation-2nd-
ed-edition-rahman/
https://ebookname.com/product/thermal-technologies-in-food-
processing-1st-edition-philip-richardson/
https://ebookname.com/product/handbook-of-antioxidants-for-food-
preservation-1st-edition-fereidoon-shahidi/
https://ebookname.com/product/organic-acids-and-food-
preservation-1st-edition-maria-m-theron/
Transport Phenomena
in Food Processing
TP371.2.T73 2002
664—dc21 2002073736
This book contains information obtained from authentic and highly regarded sources. Reprinted material
is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable
efforts have been made to publish reliable data and information, but the author and the publisher cannot
assume responsibility for the validity of all materials or for the consequences of their use.
Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic
or mechanical, including photocopying, microfilming, and recording, or by any information storage or
retrieval system, without prior permission in writing from the publisher.
All rights reserved. Authorization to photocopy items for internal or personal use, or the personal or
internal use of specific clients, may be granted by CRC Press LLC, provided that $.50 per page
photocopied is paid directly to Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923
USA. The fee code for users of the Transactional Reporting Service is ISBN 0-8493-0458-0/02/$0.00+$1.50.
The fee is subject to change without notice. For organizations that have been granted a photocopy license
by the CCC, a separate system of payment has been arranged.
The consent of CRC Press LLC does not extend to copying for general distribution, for promotion, for
creating new works, or for resale. Specific permission must be obtained in writing from CRC Press LLC
for such copying.
Direct all inquiries to CRC Press LLC, 2000 N.W. Corporate Blvd., Boca Raton, Florida 33431.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation, without intent to infringe.
Gustavo V. Barbosa-Cánovas
Jorge Welti-Chanes
Jorge Vélez-Ruiz
Gustavo V. Barbosa-Cánovas
J. Cárcel P. Coronel
Department of Food Technology Department of Food Science
Universidad Politécnica de Valencia North Carolina State University
Valencia, Spain Raleigh, North Carolina
F. Castaigne
J.G. Crespo
Department of Food Science and
Department of Chemistry
Nutrition and Horticulture Research
Faculdade de Ciências e Tecnología
Center
Universidade Nova de Lisboa
Laval University
Caparica, Portugal
Sainte-Foy, Quebec, Canada
P. Fito
B. Heyd
Department of Food Technology
Joint Research Unit Food Process
Universidad Politécnica de Valencia
Engineering
Valencia, Spain
Cemagref, ENSIA, INAPG, INRA
Massy, France
A.L. Gabas
Departamento de Engenharia de
Alimentos M.V. Karwe
Universidade Estadual de Campinas Department of Food Science
Campinas, São Paulo, Brazil Rutgers University
New Brunswick, New Jersey
M.A. Garcia-Alvarado
Departamento de Ingeniería Química y
D. Knorr
Bioquímica
Department of Food Biotechnology and
Instituto Tecnológico de Veracruz
Food Process Engineering
Veracruz, México
Berlin University of Technology
Berlin, Germany
C. González-Martínez
Department of Food Technology
Universidad Politécnica de Valencia H. Krishnamurthy
Valencia, Spain Department of Food Science and
Agricultural Chemistry
S. Grabowski McGill University, Macdonald
Food Research and Development Campus
Centre Ste. Anne de Bellevue, Quebec,
Agriculture and Agri-Food Canada Canada
St. Hyacinthe, Quebec, Canada
R. León-Cruz
G.F. Gutiérrez-López
Departamento de Ingeniería Química y
Departamento de Graduados e
Alimentos
Investigación en Alimentos
Universidad de las Américas-Puebla
Escuela Nacional de Ciencias
Santa Catarina Mártir
Biológicas — I.P.N.
Cholula, Puebla, México
México, México
F. Hamouz M. Marcotte
Department of Nutritional Science and Food Research and Development
Dietetics Centre
University of Nebraska-Lincoln Agriculture and Agri-Food Canada
Lincoln, Nebraska St. Hyacinthe, Quebec, Canada
CONTENTS
1.1 Introduction
1.2 Mass Transfer Variables
1.2.1 Concentration
1.2.2 Velocity
1.2.3 Flux
1.2.4 Flux Relations for Binary Systems
1.3 Mass Transfer by Diffusion
1.3.1 Steady State Diffusion
1.3.2 Molecular Diffusion in Gases, Liquids, and Solids
1.3.2.1 Molecular Diffusion in Gases
1.3.2.2 Molecular Diffusion in Liquids
1.3.2.3 Molecular Diffusion in Solids
1.3.3 Unsteady State Diffusion
1.3.3.1 Solutions of Fundamental Equations
1.4 Mass Transfer by Convection
1.4.1 Film Theory and Mass Transfer Coefficient
1.4.2 Two-Film Theory and Mass Transfer Coefficient
1.4.3 Dimensionless Numbers for Mass Transfer
1.4.4 Transport Analogies
1.4.5 Mass Transfer Coefficients and Correlations
1.4.6 Mass Transfer Units
Nomenclature
References
1.1 INTRODUCTION
Mass transfer can be defined as the migration of a substance through a mixture under
the influence of a concentration gradient in order to reach chemical equilibrium.
Biochemical and chemical engineering operations, such as absorption, humidifi-
cation, distillation, crystallization, and aeration, involve mass transfer principles.
1.2.1 CONCENTRATION
The concentration of a mixture and its components may be expressed in terms of
3
mass and mol. In terms of mass, the mass concentration of the mixture (ρ, kg/m ),
3
the mass concentration of a component i (ρi, kg/m ), and the mass fraction of
component i (wi) are given by:
ρ= m/V (1.1)
ρi = m i / V (1.2)
w i = m i / m = ρi / ρ (1.3)
where m and mi are the mass flux of the mixture and component i, respectively.
3
The bulk molar concentration (C, kg mol/m ), the molar concentration of com-
3
ponent i (Ci, kg mol/m ), and the mole fraction of component i (xi) are defined by:
C = n/V (1.4)
Ci = n i / V (1.5)
x i = n i / n = Ci / C (1.6)
where n and ni are the mol of the mixture and component i, respectively.
n n
m= ∑
i =1
m i and ρ = ∑ρ
i =1
i (1.7)
n n
n= ∑
i =1
n i and C = ∑C
i =1
i (1.8)
n n
w= ∑
i =1
w i = 1 and x = ∑x =1
i =1
i (1.9)
where xi is the mole fraction of component i, and ρi and Ci are related through the
molecular weight of constituent i (Mi, kg/kg mol):
ρi = M i C i (1.10)
1.2.2 VELOCITY
In mass transfer phenomena, the velocity of a bulk mixture and of its components
can be measured with respect to fixed coordinates. In addition, the velocity of the
components can also be measured relative to the bulk velocity. Figure 1.1 illustrates
these velocities in a binary system of components A and B in the z direction.
fixed coordinate
vB – v = diffusion velocity of B = UB
vB
v= wA vA + wB vB
vA
z
vA – v = diffusion velocity of A = UA
∑x V = ∑ ∑ CV
ni Ci
V= i i V = i (1.12)
i =1 i =1
n i i =1
The velocity of the constituent i relative to the bulk velocity of the mixture is:
u i = vi − v (1.13)
Ui = Vi − V (1.14)
where ui (m/sec) and Ui (m/sec) are the mass and molar diffusion velocities, respectively.
1.2.3 FLUX
2 2
The mass bulk flux (n, kg/m sec) and the molar bulk flux ( N , kg mol/m sec) of a
mixture relative to fixed coordinates are:
n
n = ρv = ∑ρ v
i =1
i i (1.15)
N = CV = ∑C V
i =1
i i (1.16)
The flux of the components of a mixture can also be expressed relative either
to fixed coordinates or to the bulk average velocity. The flux of the component i
relative to stationary coordinates is:
n i = ρi v i (1.17)
N i = Ci Vi (1.18)
The diffusion fluxes of the constituents i of the mixture with respect to the
2 2
average bulk velocity are ji (kg/m sec) for the mass flux and Ji (kg mol/m sec) for
the molar flux.
ji = ρi u i = ρi ( vi − v) (1.19)
J i = C i U i = C i (V i − V) (1.20)
N A = CA VA (1.21)
J A = C A U A = C A (V A − V) (1.22)
CA
J A = CA V A − CA V = CA VA − (C A V A + C B V B ) (1.23)
C
CA
NA = JA + (N A + N B ) (1.24)
C
CB
NB = JB + (N A + N B ) (1.25)
C
Equations (1.24) and (1.25) show that the absolute molar flux (NΑ or NB) results
from a concentration gradient contribution or a molar diffusion flux (JA or JB) and
a convective contribution ( C A V or C B V). The molar diffusion flux is described by
Fick’s law, which for component A is written as:
dC A
J A = − D AB (1.26)
dz
where DAB is the diffusion coefficient of A through B and dCA/dz is the change of
the concentration A with respect to the position z.
In terms of mass, the mass fluxes for components A and B are, respectively:
ρA
n A = jA + (nA + nB ) (1.27)
ρ˜
nB
n B = jB + (n + n B ) (1.28)
ρ̃ A
JA D dC A
NA = = − AB (1.29)
1 − xA 1 − x A dz
2. Equimolar counter-diffusion NA + NB = 0,
dC A
N A = J A = − D AB (1.30)
dz
where xA is the mole fraction of component A in the case of dilute systems xA <<
1, and Equations (1.29) and (1.30) become the same.
∂Ci
J i = − Dim (1.31)
∂z
2
where Ji is the molar diffusion flux of component i (kg mol/sec m ) in the z direction
3
(m), Ci is the concentration of component i (kg mol/m ), and Dim is the diffusion
2
coefficient of i with respect to the mixture (m /sec). This equation is widely used
in problems related to the processing and preservation of foods.
The mass flux (ji) of the component i is expressed as:
∂ρi
ji = − Dim (1.32)
∂z
2 3 2
where the units of ji, ρi, and Dim are, respectively, kg/sec m , kg/m , and m /sec.
(1 / M A + 1 / M B )1/ 2
D AB = 0.001858T 3/ 2 (1.33)
Pσ 2ABΩ D
where P is the absolute pressure (atm), T is the absolute temperature (K), σAB is the
collision diameter (°A), ΩD is the collision integral for molecular diffusion, and MA
and MB are the molecular weights of A and B, respectively (Sherwood et al., 1975).
T
D AB = 7.4 × 10 −8 (ϕ BM B )1/ 2 (1.34)
µ BVA0.6
where ϕB, µB, and MB are an association parameter, the viscosity (centipoises), and
the molecular weight of solvent B, respectively, T is the absolute temperature (K),
3
and VA is the molal volume of A (cm /g mol).
The diffusion of gases and liquids through porous solid materials may occur by a
combination of Fick diffusion and Knudsen diffusion.
If the pores are large, the mass transfer within the gas or liquid contained in the
pores will be by Fick diffusion. Nevertheless, the diffusivity in the solid is reduced
below what it would be in a fluid, due to the tortuous nature of the path that a
molecule must travel to advance a given distance in the solid and to the restricted
free cross-sectional area (Sherwood et al., 1975). In such a case, the flux must be
described in terms of an “effective” diffusion coefficient, defined as:
ε
Deff = D (1.35)
τ AB
where DAB is the diffusion coefficient in a binary system, τ is the tortuosity, and ε is
the fractional void space. Values of ε, τ, and Deff must be determined experimentally.
Therefore, when the system is a porous solid that has interconnected voids that
affect the diffusion, for a binary system the molar flow is:
ε dC
N A = − D AB A (1.36)
τ dz
19400ε 2 T
DK = (1.37)
τS′ ρ M
where ρ is the density of the solid material, M is the molecular weight of the diffusing
gas, and T is the temperature (K).
The molar flux for gas and liquids through a porous solid can be described in
terms of DK by:
(C A1 − C A 2 ) (p − pA2 )
N A = DK = D K A1 (1.38)
z RTz
where CA1 and CA2 are the bulk concentrations in the gas and liquid phases, respectively,
and pA1 and pA2 are the partial pressures of the gas and liquid phases, respectively.
When Fick and Knudsen diffusions are important, the effective diffusion coef-
ficient is defined by:
1 1 1
= + (1.39)
Deff D AB D K
∂C A d 2CA
= D AB (1.40)
∂t dz 2
Analogous equations can be written for diffusion in spherical or cylindrical
shapes and two or three dimensions. These equations are used to find the concen-
tration of a solute as a function of time and position and are mainly applicable to
diffusion in solids and to limited situations in fluids. The analysis of unsteady state
systems, however, is frequently simplified to a one-directional flow.
1. Infinite slab:
n =∞
C − C∞ D(2n + 1)2 π 2 t
∑ 2n− + 1 cos 2n 2+δ1
4 ( 1)n ( )πx
= exp − (1.41)
C0 − C∞ π n=0
4δ 2
2. Infinite cylinder:
n =∞
C − C∞
∑ b J (b , r) exp(−Db t)
2 J0 (bn , r)
= 2
(1.42)
C0 − C∞ R n =1 n 1 n
n
3. Sphere:
n =∞
C − C∞ (−1)n +1 1 sin nπr exp Dn 2 π 2 t
∑
2R
= (1.43)
C0 − C∞ π n r R R
2
n =1
X, Relative time αt D AB t D AB t
x12 x12 x12
m, Relative resistance k D AB D AB
hx1 k m x1 Kk m x1
n, Relative position x x x
x1 x1 x1
Source: Adapted from Welty, J.R., Wicks, C.E., and Wilson, R.E., Momentum, Heat, and Mass
Transfer, John Wiley & Sons, New York, 1984 and Geankoplis, C.J., Transport Processes and Unit
Operations, 3rd ed., Prentice-Hall, London, 1993. With permission.
solid
CA1
bulk fluid
CA2
x
FIGURE 1.2 Fluid-solid interfacial region: z
the film theory.
Thus, the molar flux and the concentration profile of species A (JA) are found from
D AB
JA = (C A1 − C A 2 ) = − k m (C A1 − C A 2 ) (1.44)
x
where DAB/z is the mass transfer coefficient, km, CA1, and CA2 are the bulk concen-
trations in the fluid and solid phase, respectively, and
z(C A1 − C A 2 )
C A = C A1 − (1.45)
x
jA = k ρ (ρA1 − ρA 2 ) (1.46)
where CA is the concentration of the components A, and ρA1 and ρA2 are the mass
concentration of the fluid and solid, respectively.
2
The mass transfer coefficient kC or kρ has units of velocity (m /sec). It can be
determined either from experimental data and empirical formulas derived from them
or with the aid of methods of similitude theory. A rough estimation of km can be
attained by assuming km = D/x, provided the effective film thickness and the diffusion
coefficient are known (Sherwood, 1974).
x1 x2
CA1
CA1i
CA2i
CA2
layer layer
CA 2
C*A 2 = mCA1 C*A1 =
m
FIGURE 1.3 Mass transfer at a gas liquid interphase: the two-film theory.
the liquid phase. The two-film theory assumes that the resistance to mass transfer
lies in each adjacent phase to the interphase and that no resistance is offered to the
transfer of the solute across the interphase (Welty et al., 1984).
At steady state, the fluxes in gas and liquid (J1) phases must be equal:
J1 = k mG (C A1 − C A1i ) = k mL − (C A 2 i − C A 2 ) (1.47)
where kmG and kmL are the mass transfer coefficients in the gas and liquid phases,
respectively, CA1i and CA2i are the concentrations of component A at the interface,
and CA1 and CA2 are the bulk concentrations in the gas and liquid phases, respectively.
Since the concentrations at the interphase are not easily measurable, it is con-
venient to calculate an overall mass transfer coefficient based on an overall potential
gradient between the bulk compositions. The overall driving force is not, however,
CA1 – CA2, since at the interphase discontinuity of the concentrations exists, and the
solubility in the liquid is not necessarily the same as in the gas. Moreover, the film
thickness x1 and x2 and the diffusivity of the solute may be different in the two phases.
The solubility relationship that governs the equilibrium concentration between
phases is of the form:
CG = mC L (1.48)
where m is the solubility constant between the two phases, and CG and CL are the
concentrations of the gas and liquid, respectively.
C
J = K mG (C A1 − C*A1 ) = K mG C A1 − A 2 (1.49)
m
*
where CA2/m = C A1, which is the concentration in the gas phase that would exist in
equilibrium with CA2, the concentration of species A in the liquid.
A similar equation can be obtained for the overall coefficient if the driving force
is based on the concentration in the liquid phase:
J = K mL (C*A1 − C A 2 ) = K mL ( mC A1 − C A 2 ) (1.50)
where mCA1 is the concentration in the liquid that would exist in equilibrium with
*
C A1. KmG and KmL represent the overall mass transfer coefficient based on the gas
phase and the liquid concentration driving force.
Coefficients KmG and KmL are related to the individual mass transfer coefficients
and the equilibrium constant m of a gas–liquid (kmG) or vapor–liquid (kmL) system
as follows:
1 1 m
= + (1.51)
K mG k mG k mL
1 1 1
= + (1.52)
K mL mk mG k mL
These last two equations show the relationship among the coefficients for the
individual phases and the overall transfer coefficients, expressed as a global resis-
tance (1/KmG or 1/KmL) to the transfer of the diffusing component. In studying
performance separation processes, it is important to determine which individual
resistance is the limiting factor.
gd 3β∆Tρ2
Grashof Gr = Buoyancy forces/viscous forces
ν
α
Lewis Le = Heat diffusivity/mass diffusivity
D
vd
Peclet Pe = Convection/diffusion
D
vd
Reynolds Re = Inertial force/viscous force
ν
ν Momentum diffusivity/mass
Schmidt Sc =
D diffusivity
k md Measure of boundary layer
Sherwood Sh =
D thickness
km Wall mass transfer/mass transfer
Stanton St =
v by convection
The Graetz number is defined as the product of the Reynolds and Schmidt
numbers. It includes the effects of forced and free convection.
The Grashof number is a measure of free convection, which will be enhanced
by buoyancy forces (ρ) and decreased by viscous forces (ν). The Grashof number
is found in correlations of free convection mass transfer.
The Lewis number, Le = α/D, is the ratio of heat to mass diffusivity. It plays
an important role in processes where simultaneous convective transfer of energy and
mass occurs.
The Peclet number, defined as Pe = vd/D, is the ratio of bulk mass transport to
diffusive mass transport.
The Reynolds number, Re = vd/ν, characterizes the nature of the motion of a
flowing gas or liquid, and is interpreted as the ratio of inertial forces to the viscous
forces in the flow. When the Re value is below a certain critical value, the flow is
laminar; when it is larger, the flow is turbulent.
In the Schmidt number, Sc = ν/D, the two molecular transport coefficients (ν
and D) are the physical properties of the medium in which the transfer of mass takes
place. The Schmidt number, which represents the ratio of momentum to mass
diffusivities, is of great importance in convective mass transfer (Welty et al., 1984).
µ d d
τ= (ρv) = v (ρv) (1.53)
ρ dz dz
k d d
q=− (ρC p T) = −α (ρC p T) (1.54)
ρC p dz dz
dC
Ji = −D (1.55)
dz
All three processes are quite different from one another at a molecular level.
However, certain analogies exist among them. In effect, molecular diffusivities
kinematic viscosity (ν), thermal diffusivity (α), and diffusion (D) have the same
2
dimensions (L /t). Also, in Fick’s law, the molar flux varies with the gradient in mol
per volume; in the rewritten Fourier’s law, the energy flux is proportional to the
gradient of energy per volume (ρCpT); and the momentum flux, given by the rewritten
Newton’s law, varies with the gradient of the momentum per volume (ρv) (Cussler,
1984). These analogies are shown in Table 1.3.
The corresponding equations for momentum, heat and mass flux in convective
motion are:
τ = f ρv2 = (ρv − 0)
1 fv
(1.56)
2 2
h
q = h∆T = ∆(ρC p T) (1.57)
ρC p
N i = k m ∆Ci (1.58)
Variable ρv ρC p T C
(momentum/volume) (energy/volume) (mol/volume)
Molecular diffusivity ν α D
(kinematic viscosity) (thermal diffusivity) (diffusion coefficient)
Transfer coefficient f h km
(friction factor) (heat transfer coefficient) (mass transfer
coefficient)
Dimensionless ν ν
number Pr = Sc =
α D
h / ρC p km
St = St =
v v
α
Le =
D
where the transfer coefficient fv/2 is like h/ρCp and k. Note that the driving forces in
the momentum, heat, and mass flux are volume concentrations: (ρv − 0) is expressed
in momentum per volume, ∆(ρC p T) in energy per volume, and ∆Ci in mol per volume
(Table 1.3).
Since the molecular diffusivities have the same dimensions, a ratio of any of
two of these leads to dimensionless numbers: Pr number for heat transfer, Sc and
Le numbers for mass transfer. Likewise, the ratio transfer coefficients to the flow
velocity lead to an [St] number for heat transfer and an [St] number for mass transfer
(Tables 1.2 and 1.3).
A useful and simple analogy relating all three types of transport simultaneously
is the Chilton–Colburn analogy, which is written as
k m 2/3 h / ρC p 2/3 f
Sc = Pr = (1.59)
v v 2
The group
k m 2/3
Sc
v
is called the jD factor for mass transfer, and
h / ρC p
Pr 2 / 3 = St Pr 2 / 3
v
defines the jH factor for heat transfer.
TABLE 1.4
Typical Mass Transfer Correlations
Situation Equation Remarks
Sh = 0.332 Re 1/ 2
Sc 1/ 3 Used for laminar flow over a plate
Used for evaporation drops in spray
Evaporation drops Sh = 2 + 0.6Sc1/ 3 Re1/ 2 drying
1/ 4
d 4 ( P / V) ν
1/ 3 Used in mixing liquid solutions; P/V
Stirred drops Sh = 0.13 D is important in scale-up
ρν
3
1/ 2 1/ 3
Solid spheres Sh = 2 + 0.6
dv ν Used for forced convection
ν D
d 3 [ ∆ρ]g
1/ 4 1/ 3
Sh = 2 + 0.6 ν Used for natural convection
D
ρν
2
Bubbles Sh = 2 + 0.31(Gr )1/ 3 (Sc)1/ 2 Used for rising bubbles, d > 2.5 mm
where HTUc is the height of a transfer unit and NTUD is the number of mass transfer
units given by:
C2
∫ C −C
dC
NTU D = (1.61)
i b
C1
Subscripts 1 and 2 refer to inlet and outlet values. The integral is evaluated from
considerations of bulk (b) and interphase (i) conditions at any cross-section of the
drier, absorption tower, etc., the former being determined by mass balances and the
latter by equilibrium data.
The height of a transfer unit is given by:
G
HTU = (1.62)
k m aA tρg
where the quantity NEU is the number of energy units, and the Le Goff number
(Lf) represents the deviation of the Colburn analogy from unity, the Lf number being
in the range 1–0.02.
∫ T −T
dT
NTU H = (1.64)
i b
T1
P1 − P2
NEU = (1.65)
ρv2
The number of transfer units is a basic parameter that appears frequently in the
design of heat and mass exchangers (Van den Bulck et al., 1985; Khodaparast, 1992;
Treybal, 1981).
NOMENCLATURE
2 2
a Specific area; cm or m
2
At Cross-sectional area of drier; m
bn Roots of Bessel function
3
C, c Molar concentration; kg mol/m
* 3
C Molar concentration at equilibrium; kg mol/m
Cp Specific heat capacity; kJ/kg K
d Characteristic dimension of a solid body; m or cm
2 2
D Diffusion coefficient; m /sec or cm /sec
f Friction factor; dimensionless
2
g Acceleration constant for gravity, m/sec
2
G Flow rate; kg/m sec
2
h Heat transfer coefficient; W/m K
HTUD Height of a mass transfer unit, m
j The mass flux of the mixture with respect to the average bulk velocity;
2
kg/m sec
J Molar diffusion flux of the mixture with respect to the average bulk velocity;
2
kg mol/m sec
Jo Bessel function of first kind and order zero
J1 Bessel function of first kind and order one
2
km Mass transfer coefficient; m /sec
k Thermal conductivity; W/m K
Km Overall mass transfer coefficient based on the gas system concentration
driving force
K Partition coefficient; dimensionless
m Mass of the mixture; kg in Equation (1.1)
m Solubility constant between the two phases in Equation (1.48)
2
m Mass flux of the mixture relative to fixed coordinates, kg/m sec in Equation
(1.11)
M Molecular weight; kg/kg mol
n Mole of the mixture; kg mol in Equation (1.4)
n Relative position in Equations (1.7) and (1.12)
GREEK SYMBOLS
2
α Thermal diffusivity; cm /sec
β Coefficient of thermal expansion; 1/K
δ Thickness of a hypothetical stagnant film; cm or m
ε Porosity
µ Viscosity; centipoises
2
ν Kinematic viscosity; m /sec
3
ρ Mass concentration of the mixture or fluid density; kg/m
σAB Collision diameter; °A
τ Tortuosity
ΩD Collision integral for molecular diffusion; dimensionless
ϕ Association parameter
REFERENCES
Crank, J., The Mathematics of Diffusion, 2nd ed., Oxford University Press, London, 1975.
Cussler, E.L., How we make mass transfer seem difficult, Chem. Eng. Educ., 18(3), 124–127,
149–152, 1984.
CONTENTS
2.1 Introduction
2.2 General Background
2.2.1 Thermal Properties of Foods
2.2.1.1 Specific Heat
2.2.1.2 Thermal Conductivity .
2.2.1.3 Thermal Diffusivity
2.2.1.4 Surface Heat Transfer Coefficient
2.2.2 Heat Transfer by Conduction
2.2.2.1 Steady State
2.2.2.2 Nonsteady State
2.2.3 Heat Transfer by Convection
2.2.3.1 Natural Convection
2.2.3.2 Forced Convection
2.2.4 Heat Transfer by Radiation
2.3 Conclusions
References
2.1 INTRODUCTION
Heating and cooling are common activities in food processing. Several operations
involving heating of raw foods are performed for different purposes, such as reduc-
tion of the microbial population, inactivation of enzymes, reduction of the amount
of water, modification of the functionality of certain compounds, and of course,
cooking. On the other hand, heat is removed from foods to reduce the rate of its
deteriorative chemical and enzymatic reactions and to inhibit microbial growth,
extending shelf life by cooling and freezing. Heat transfer plays a central role in all
of these operations; therefore, food engineers need to understand it in order to achieve
better control and avoid under- or over-processing, which often results in detrimental
effects on food characteristics. In practice, heat transfer to or from foods can be
attained either by indirect or direct methods. Indirect methods involve the use of
heat exchangers that isolate the product from the medium used as a source or sink
Direct Frying, boiling, baking, and grilling solid Fluidized bed individual quick freezing
foods (IQF)
Indirect Fluid food pasteurization; canned products Fluid food cooling; ice production
sterilization
of heat. Direct methods allow contact between the food and the heating/cooling
medium. Examples of these methods can be found in Table 2.1.
Indirect heating methods use gases and vapors, such as steam or air, and liquids,
such as water or organic compounds, as a source of thermal energy. Chilling by
indirect methods involves the use of coolant gases, such as ammonia or Suva , or
the use of coolant liquids, such as water or ethylene glycol. Direct heating can be
attained by means of hot air, oil, infrared energy, and dielectric or microwave
methods, among others. Direct chilling can be achieved by the use of cold air or by
the application of the Peltier effect.
Cp = ∑ (C Cp )
i
i i Choi and Okos, 1986a (2.7)
where Ci is the mass concentration of each constituent i and Cpi its corresponding Cp.
k= ∑ (k X )
i
i
v
i (2.10)
th v
where ki is the thermal conductivity of the i component and Xi its volume fraction.
When heat is flowing through wrappings or compound layer materials, a total thermal
conductivity coefficient is needed. This expression can be calculated if the individual
thermal conductivity coefficients are known.
For heat flowing parallel to two layers, the total thermal conductivity coefficient
is given by
kT = k1 (1 – c) + k2c Hallstrom et al., 1988 (2.11)
where c is the volume fraction of material two. On the other hand, if the heat flow
is perpendicular to the material’s layer orientation:
k1k 2
kT = Hallstrom et al., 1988 (2.12)
ck1 + (1 − c)k 2
In the case of material mixtures with random size and orientation particles, the
value for the total thermal conductivity value will be found between the value for
parallel flow and the value for perpendicular flow. For mixtures of more than two
components, the same method is followed, taking two materials at a time.
α= ∑ (α X )
i
i i Choi and Okos, 1986b (2.16)
th
where αi is the thermal diffusivity of the i component and Xi its mass fraction. As
thermal diffusivity can be calculated from other thermal properties, data are not
frequently found in the literature.
The heat transfer coefficient (h) is not a property of materials themselves, but rather
a property of convective heat transfer systems involving a solid surface and a fluid.
This coefficient is used as a proportionality factor in Newton’s law of cooling,
adjusting for the characteristics of the system under study.
To define the value of this factor, it is necessary to characterize the convective
medium and the surface involved in the convective heat transfer process. Some of
the characteristics involved in the calculation of the heat transfer coefficient are the
fluid’s velocity, viscosity, density, thermal conductivity, and specific heat. The form
and surface texture of the solid involved are also important. As can be determined
from a dimensional analysis of Newton’s cooling law, the units for the heat transfer
2
coefficient are W/m °C. Since the heat transfer coefficient is a property of the system
rather than of the material, its measurement is difficult, and several empirical expres-
sions have been developed to overcome this problem. Some of these expressions
will be reviewed below in the section dealing with heat transfer by convection.
A compilation of surface heat transfer coefficient data and empirical expressions
to calculate this coefficient can be found in the 1985 edition of the ASHRAE
Handbook of Fundamentals. Further information regarding measurement of the heat
transfer coefficient can be found in the literature (Arce and Sweat, 1980).
Now that we have shown methods for determining the most relevant thermal
properties of foods, we will find out how to use them to model heat transfer processes.
The particulars of the different heat transfer modes will be dealt with in the following
sections.
The study of heat conduction in the steady state is useful in modeling the perfor-
mance of insulators, heat exchangers, and other equipment used to transfer heat from
one point to another, such as containers, pans, and walls. In these, the initial varia-
tions in internal temperature dependent on time have settled, and the temperature
profile inside the material is stationary. The important issue here is to determine the
amount of heat a particular material of a given thickness will allow to flow through it.
This mechanism can be modeled using Fourier’s law for heat conduction
(Equation 2.17), which establishes that the heat flux Qx transmitted through a solid
in the direction x is inversely proportional to the thickness x and directly proportional
to both the perpendicular transmission area A and to the temperature difference
between its two opposite faces ∆T. The proportionality constant needed by this model
is the thermal conductivity (k), one of the physical characteristics previously
described. In some materials, thermal conductivity may vary with temperature. In
these situations, the value corresponding to the average temperature should be used.
∂T(x)
Q x = − kA (2.17)
∂x
The negative sign represents the heat flow from the hottest to the coolest surface,
thereby rendering a positive value for the heat flux. As can be seen, this equation
describes heat flow only in one direction. For a complete mathematical description, we
would need to write similar equations for the other two directions in a three-dimensional
system and to integrate over the entire volume. However, most steady state processing
applications involve heat conduction in only one direction, as when heat flows across
walls or heat exchangers. For practical purposes, these materials can be considered as
infinite slabs, so we do not need to be concerned with a general mathematical solution.
A∆T
Qx = − k (2.18)
x
Equation (2.18) is the integrated form of Fourier’s law for unidirectional steady
state heat conduction over a path of constant cross-sectional area in a parallelepiped.
It can be used directly to calculate the heat flux through a body. Besides heat flowing
through flat surfaces, another common situation in food engineering is the use of
The study of heat conduction in the nonsteady state is pertinent when calculating
processes in which the focus is to heat or cool a body instead of using it as a heat
conduction medium. Such processes include freezing, cooking, and thermal sterilization
∂T ∂2 T ∂2 T ∂2 T
= α 2 + 2 + 2 (2.24)
∂t ∂x ∂y ∂z
where T stands for temperature, t for time, x, y, and z for the distance on the x, y
and z axes, respectively, and α for thermal diffusivity, which is a physical character-
istic of the materials, as previously discussed. To arrive at the analytical solution to
this complex expression can be difficult or impracticable. However, the analysis of
some practical situations may be simplified through a couple of useful assumptions.
The first assumption supposes that we are dealing with a semi-infinite body, also
known as a thick solid. This semi-infinite body is defined as one with infinite width,
length, and depth. If a body with these characteristics is immersed in the heating
medium, we can assume that heat will be transferred just from the surface toward
the interior, following a straight trajectory to the center; therefore, Fourier’s law can
be transformed into:
∂T ∂2 T
=α 2 (2.25)
∂t ∂x
where the x-axis corresponds to any one of the dimensions on a parallelepiped or
to the radius on a sphere-like body. The solution to Equation (2.25) becomes simpler
now and may be obtained by applying the boundary conditions, which are:
The solution takes the form of Equation (2.26), where the temperature T at any
point x, measured from the surface, can be determined using a dimensionless tem-
perature ratio:
2
hx h
Tm − T x + αt
k x h αt
= erf +e
k
erf + (2.26)
Tm − T0 4αt 4αt k
where erf is the Gauss’ error function, which can be obtained from Table 2.2, and k
is the thermal conductivity.
Since, in this case, the heat is transferred from a fluid to a solid surface, the
surface heat transfer coefficient h is incorporated into the solution. For situations
where the heat transfer between the fluid and the solid proceeds very efficiently, that
is, if h is infinite, the solid surface takes the medium temperature instantaneously,
and Equation (2.26) can be simplified to:
Tm − T x
= erf (2.27)
Tm − T0 4αt
Now that we know how to handle semi-infinite bodies, a problem arises. The
definition of a semi-infinite body describes a body with infinite dimensions, which
in reality is not a possible situation. When can a body be considered as a semi-infinite
body for practical purposes? The concept of a semi-infinite body in practice is better
related to the system’s behavior than to the body dimensions. The surface heat
transfer coefficient, along with the thermal conductivity, plays a fundamental role
in determining whether a body will exhibit a thick body response. If a finite body
of a given thickness is studied in the early steps of heat penetration, it will exhibit
a semi-infinite body response as heat has only flowed from the outside to the center;
therefore, the ratio of thickness to time plays an important role as well. Schneider
(1973) found a critical Fourier number to define when a body ceases to exhibit a
thick body response:
–0.3
Focritical = 0.00756 Bi + 0.02 for 0.001 ≤ Bi ≥ 1000 (2.28)
αt
Fo = (2.29)
( L )2
hL
Bi = (2.30)
k
where L is half the thickness of the body and t is the time the body has been exposed
to the conditions studied. If the actual Fourier number exceeds the critical value,
the body can no longer be considered to be a semi-infinite body, and a different
approach must be taken.
Another possible approximation used to solve practical problems regarding heat
conduction in a nonsteady state involves the assumption of infinite bodies with
simple geometry. Again, the concept of an infinite body is just theoretical, as in
practice no such bodies occur and, therefore, a practical definition is needed. An
infinite body is a body in which two of its dimensions are much larger than the third.
In this situation, it can be expected that heat will reach the center of the body first
traveling the shortest way; that is, following the axis corresponding to the thinner
dimension. Heat transfer through the remaining axes becomes negligible. This
assumption applies to infinite slabs with thickness 2xo and infinite longitude and
width, infinite cylinders with radius r and infinite length, and spheres.
Analytical solutions for this kind of problem can be developed using dimension-
less ratios. Expressions for temperature Y and position n are:
Tm − T
Y= (2.31)
Tm − T0
x
n= (2.32)
x0
where Tm is the temperature of the cooling or heating fluid surrounding the solid,
T0 is the initial temperature of the body, T is the temperature at a point x, x0 is the
distance from the center or midplane to the surface, depending on the body’s shape,
and x is the distance from the center or midplane to the selected point. Time and
thermal diffusivity are introduced using Fourier module τ, while surface heat transfer
coefficient and thermal conductivity are both included using the Biot module in m:
αt
τ = Fo = (2.33)
x 20
1 k
m= = (2.34)
Bi hx 0
∂T ∂2 Y
=α 2 (2.35)
∂t ∂n
Y = YxYyYz (2.36)
1
m=∞
m=6 x x
0
m
=2
m
=2
n = 1.0
0.8
m
0.6
=
1
0.1 0.4
m
0.2
=
0.0
1
m=
0.5
m=
Y
0.5
n = 1.0
0.8
m=
0.6
m=
0.4
0.2
0
n = 1.0 0.0
0.01 0.8
0.6
0.4
0.2
n = 0.8 0.0
0.6
0.4
0.2
0.0
m=0
n=1
0.001
0 0.5 1 1.5 2 2.5 3 3.5
τ = (Fo)
Habozások.
A botrányos kirándulás.
A bünhödés.