Professional Documents
Culture Documents
To cite this article: B. V. Ramarao , A. Massoquete , S. Lavrykov & S. Ramaswamy (2003) Moisture Diffusion Inside Paper
Materials in the Hygroscopic Range and Characteristics of Diffusivity Parameters, Drying Technology: An International Journal,
21:10, 2007-2056, DOI: 10.1081/DRT-120027044
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
DRYING TECHNOLOGY
Vol. 21, No. 10, pp. 2007–2056, 2003
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
ABSTRACT
2007
through the void space and through the fiber matrix. Diffusivities
in these individual phases are supplemented by a local kinetic
coefficient representing moisture flux interchange between the void
and fiber phases. When the local moisture exchange coefficient takes
on large values, the fibers and the void spaces are at local equilibrium
with no net exchange of moisture. Under such conditions, the model
reduces to the simpler Fickian diffusion model with nonlinear
moisture diffusivity equivalent to earlier models.
1. BACKGROUND
2010
Table 1. Some mechanisms for moisture transport in paper suggested by various authors.
that for average humidities less than about 60%, the dependence of the
diffusivity on humidity was quite weak whereas for higher humidities it
increased strongly. Diffusivity decreased with increased sheet densities.
Nilsson and Stenstrom[13] proposed a periodic array of parallel cylinders
as a model for fibrous media such as paper and nonwoven media. They
obtained analytical solutions to Laplace’s equation in a unit cell and
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
2014
Table 2. Moisture diffusivity, data, and correlations in the literature for diffusion through pore space and through fiber matrix, Dq.
Lee and Hinds 6 Ply sheets 600 Drying, 105 C Dq ¼ 1.4 1010 (q 0.15)2.2
Kobari Dq ¼ exp(3.19q/)" 27.04)
Hagglund et al.a (c) UBKP mats 431 90% 2.95 108
Lin Kraft paper Dq ¼ 2.616 1011 exp(0.5q)
Hashemi et al. BKP 50 Dq ¼ 2.95 106 exp(0.505/q)
Amiri et al. Newsprint 2.1 1010 9.5 1010
Foss et al.a Board 600 2 106
Foss et al.a 1012
Foss et al.b 3.8 1014
ten Donkelaar Corrugated Dq ¼ 1010 exp(15q)
Ramarao and Chatterjee BKP 240 50–90% 2.5 107
Radhakrishnan et al. BKP 240 Dq ¼ 2.616 1011 exp(345q/18)
6
Nilsson and Stenstrom 1.25 10
Massoquete et al. 2.5 106
Moisture Diffusion Inside Paper Materials
a
Representative diffusivity shown. Original paper contains more data at different conditions (q/RH).
b
Representative diffusivity shown. Original paper contains more data at different conditions (q/RH).
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
2015
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
E(t) is the ratio of the moisture gain at a chosen time to the total possible
moisture gain.
MðtÞ Mð0Þ
EðtÞ ¼ ð12Þ
Mð1Þ Mð0Þ
Assuming that the diffusivity is an increasing function of concentra-
tion, Crank has provided the following general conclusions based on
extensive numerical and experimental results. The amount sorbed [E(t)]
is directly proportional to t1/2. A plot of E against t1/2 is referred to as the
sorption (desorption) curve. Crank suggests that this curve should be
linear and the linear behavior can extend well beyond 50% of the final
equilibrium uptake value. If the sorption curve shows an inflection point,
the concentration at the boundary is no longer a constant or it implies
that the diffusivity is a function of other variables (most commonly time)
in addition to the concentration. Then the functional form of Eq. (2) does
not hold.
When the diffusivity (or the boundary concentration) is a function
of time, it is more common to replace the diffusion law by a relaxation
equation drawn from thermodynamics of irreversible processes. One such
formulation represents the moisture flux as the convolution integral of
a diffusivity function and the concentration gradient (Neogi[19,20]).
Zt
j¼ Dðt Þ rcðÞ d ð13Þ
0
in dry kraft paper used for capacitors and in oil soaked samples after
subjecting the sheets to a step increase in humidity. Lin[22] analyzed these
profiles using predictions from the standard unsteady state diffusion
model with a moisture dependent diffusivity. The coefficients in the
moisture dependence of the diffusivity determined by Lin are shown
in Table 2. Lin’s model used Eq. (7) with Eq. (2) substituted for the
diffusivity and the initial and boundary conditions given by Eqs. (8)
through (10).
Roisum[23] and recently, Amiri et al.[8] modeled transient
moisture diffusion through paper rolls using the two-dimensional analog
of Eq. (7). ten Donkelaar and Jaeger[24] analyzed moisture diffusion in a
paper sheet and corrugated board assuming that local equilibrium is
established between the pore and fiber spaces. They showed that diffusion
within the pore and fiber phases can be reduced to a single transient
diffusion equation that can be approximated by Eq. (7). Further, if one
accepts the exponential relation for the diffusivity given by Eq. (2), the
model reduces to that used by Lin. All of these works are based on the
validity of the flux being represented by a single potential gradient.
Thus they can all be classified as generalized Fickian diffusion models,
although nonlinear dependence of the diffusion coefficient is allowed.
@q
¼ c q ð24Þ
@t
In this model, moisture conduction by the fibers in parallel with the
void space is ignored just as Lescanne et al.’s did. The exchange of
moisture between the fibers and the pore space is assumed to be by
adsorption or desorption each of which is represented as a first order
process [Eq. (24)]. Consider the sorption isotherm to be linear with the
constant of proportionality Ks. Then, by equating the kinetic constants,
to Kski and u to ki, this model reduces to Ramarao and Chatterjee’s
model. However, Ramarao and Chatterjee attributed local moisture
exchange between the fibers and the void space to a more generalized
departure from equilibrium of the local moisture concentration (i.e.,
deviation of q from q*).
Hellen et al.[27] conducted random walk simulations of tracer
diffusion through simulated fiber structures in conjunction with the
above ‘‘phenomenological’’ description. The random walk simulations
clearly show that diffusion is nonfickian when fiber and void spaces are
not in local equilibrium.
A more complete discussion of fickian and nonfickian diffusion of
moisture in paper is given by Lavrykov et al.[29]
Paper materials creep under applied stress and their creep rate is
dependent upon the moisture content (or the humidity level).
Furthermore, creep is accelerated significantly when the material is
subjected to transient variations in humidity. The accelerated creep of
paper has been theorized recently to occur due to plasticization of the
fibers in conjunction with the existence of nonuniform moisture fields.
Obviously accurate estimation of the moisture fields is central to any
analysis of accelerated creep. Eusufzai et al.[40] to our best knowledge
were the first to investigate the sorption response of paper sheets to
cyclic variations in humidity with this objective. They used Lin’s
unsteady state diffusion model with the nonlinear diffusivity to describe
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
under which this model reduces to simpler versions especially for the
single dimensional cases. Their physical meaning identifies the parameters
appearing in this model and we correlate them with some papermaking
and fiber variables.
When exposed to air at a certain humidity h, paper attains an
equilibrium moisture content qsat. Obviously, for a given humidity
level, different moisture contents [qsat(c or h)] are obtained depending
upon the prior history of the sheet of paper. For our purpose here it is
sufficient to note that the dependence of the equilibrium moisture content
on humidity and on temperature can be described by the following
equation also known as the GAB isotherm (e.g., Chatterjee et al.[51]).
CG MG kG h
q ðhÞ ¼ ð25Þ
ð1 kG hÞð1 kG h þ kG CG hÞ
@q
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
q ¼ q ðcÞ ð28Þ
@c @q
" þ p ¼ r Dp rc þ p r Dq ðqÞ rq ð29Þ
@t @t
@c @q
kf ðcb cÞ ¼ Dp p Dq ð29aÞ
@z @z
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
porous.
Thus, Eq. (28) can be applied to the moisture content q, and since it is
a function only of the local concentration field c, we obtain
@c dq @c dq
" þ p ¼ r Dp rc þ p r Dq ðq Þ rc ð30Þ
@t dc @t dc
which reduces to
@c 1 dq
¼ r D p þ D
p q ðq Þ rc ð31Þ
@t " þ p ðdq Þ=ðdcÞ dc
This equation is mathematically equivalent to the nonlinear heat
conduction equation with the conductivity defined as
dq
Dcomp ¼ Dp þ p Dq ðq Þ ð32Þ
dc
If the isotherm is linear, i.e., the following approximation holds for
Eq. (28),
q ðcÞ ¼ q0 þ Ks ðc c0 Þ ð33Þ
the slope of the isotherm is independent of c and is given by
dq
¼ Ks ð34Þ
dc
For the steady state case, whether the isotherm is linear or not,
diffusion remains fickian as shown by setting the transient term to
vanish in Eq. (31), resulting in
dq
r Dp þ p Dq ðq Þ rc ¼ 0 ð35Þ
dc
For the uptake of moisture when the sheet is subjected to a step
change in humidity at its surface, the boundary conditions are given by
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
Eqs. (18) through (22). Therefore when local equilibrium can be assumed
within the paper sheet, moisture transport reduces to Fickian diffusion
in the steady state. A single diffusivity for paper sheets can be derived
which incorporates the geometric/structural aspects in addition to the
diffusivity in both these phases.
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
c ¼ cðq Þ ð36Þ
c
¼ Ai þ Bi c þ Ci c2 ð37Þ
q
Hence,
@c dc @q
¼ ð38Þ
@t dq @t
@c dc @q
¼ ð39Þ
@z dq @z
dc @q @q dc
" þ p ¼ r Dp rq þ p r Dq rq ð40Þ
dq @t @t dq
is given by,
Dp ðdc=dq Þ þ p Dq
Dcomp ðqÞ ¼ ð42Þ
"ðdc=dq Þ þ p
A simpler representation is obtained if the sheet density (p) is
large compared to the first term in the denominator, i.e., "(dc/dq*). The
simplification can be represented as
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
Dp dc
Dcomp ðqÞ ¼ þ Dq ð43Þ
p dq
or
Dcomp ðqÞ ¼ Dc0 þ Dql emq ð44Þ
if we assume that the diffusivity of moisture through the fiber matrix can
be represented by an exponential relation as in Eq. (7). We thus note that
the complete transient diffusion model can be simplified to
@q @ @q
¼ D ðqÞ ð45Þ
@t @z comp @z
which is approximately equivalent to the simple nonlinear diffusion
models of earlier workers. The initial and boundary conditions to be
applied for a step change in humidity at the external boundary and
when external surface boundary layers are negligible are given by
Eqs. (18) through (22).
Whenever local equilibrium may be assumed, diffusion simplifies to
a single Fickian process. Local equilibrium within paper sheets
simplifies the treatment substantially for both the steady and unsteady
cases. If the sheet is subjected only to small ranges of humidity or
sufficiently slow transients, it may be useful. Rapid transients or large
swings in humidity are likely to disturb this equilibrium and result in
substantial nonfickian effects due to the interference of transport by
many mechanisms. When the boundary conditions involve convection,
difficulties occur. In the next section, we conduct a perturbation
analysis of the model to delineate the specific conditions under which
this assumption may be valid.
pore space, the fiber matrix and a local uptake of moisture represented by
the algebraic term in the model (Eqs. (28) and (29)). It is our desire to
analyze transient uptake of moisture by a paper sheet subjected to a
change in the external humidity conditions. The analysis we present is
largely that presented by Bandyopadhyay et al.[1]
We first scale the diffusion equations to obtain the following
nondimensional forms. In this, C and Q represent the scaled water
vapor concentration and moisture contents and and represent the
scaled time and z dimension respectively. The definitions of these
variables are provided in the table of symbols.
@C
¼ r Dp rc þ ½Q ðCÞ Q ð46Þ
@
@Q
¼ r FQ ðQÞ rQ ½Q ðCÞ Q ð47Þ
@
Considering transport in the z dimension alone, we obtain[1]
@C @2 C
" ¼ 2 þ ðQ QÞ ð48Þ
@ @
@Q @ @Q
¼ F ðQ QÞ ð49Þ
@ @ Q @
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
@Cð ¼ 1Þ @Qð ¼ 1Þ
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
¼ ¼0 ð52Þ
@ @
Equation (52) indicates that the sheet is initially at equilibrium and is
subjected to a sudden change in humidity at the external boundaries.
Assuming the sheet to be symmetric, the moisture fluxes in condensed
and vapor forms vanish at the centerline denoted by ¼ 1. If the external
convective boundary layer poses a significant resistance, the boundary
condition given by Eq. (53) must be replaced by
@C @Q
Bi ½Cb C ¼ FQ0 ð53Þ
@ @
which is the nondimensional form of Eq. (29a). In order to simplify the
analysis, we will assume that the equilibrium sorption relation is linear
for the materials. In nondimensional form, this can be expressed as
Q ¼ C ð54Þ
3.1. Case 1. 1
In this case, the fibers and the pore space are close to equilibrium
locally throughout the sheet. This will occur if
ki L2
1 ð55Þ
Dp
i.e., when the intra-fiber exchange coefficient is large or the pore diffusiv-
ity is small or the sheet is quite thick. In this case, we see that the model is
nonuniform since it reduces to the simple algebraic condition that
C 0 ¼ Q ¼ Q0 ð56Þ
and the boundary conditions can no longer be adequately satisfied.
However, this shows that when the parameter is large, the sheet
approaches equilibrium. Local equilibrium can be expected whenever the
interchange coefficient ki is large or the pore diffusion process
is substantially slow compared to local diffusion. From dimensional
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
3.2. Case 2. 1
tDp
þ ¼ ¼ ð60Þ
L2
Let
C ¼ Cð , þ Þ ð61Þ
Q ¼ Qð , þ Þ ð62Þ
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
Q ¼ Q0 ð , þ Þ þ Q1 ð , þ Þ þ ð64Þ
Substituting into Eqs. (48) and (49) of the model and collecting terms of
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
( ) " #
@Q1 @ 0 @Q
1
0 @Q0
¼ F ½Q Q ð68Þ
@ þ @ Q @ @
These two problems clearly show that the moisture content changes on a
much longer time scale (*) whereas rapid changes occur within the pore
space. The long time scale is given by Eq. (61).
In the model for moisture diffusion in paper, there are three critical
parameters. These are: diffusivity of water vapor within the void space,
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
Dp, the diffusivity of condensed phase moisture through the fiber matrix,
Dq, and the local moisture interchange coefficient between the fibers
and the pores, ki. If these parameters are known for a paper sheet, the
moisture uptake and distributions in transient and steady states can be
determined by applying the model given by Eqs. (46) and (47), subject of
course to the appropriate initial and boundary conditions. In this article,
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
Dp Z
"p 1
¼Iþ ðnf þ fnÞ dA ð69Þ
Da V 2
Apq
where f is a vector field and n is the unit normal vector at the interface
between the fiber and void interface, Apq is the area of this interface.
The vector field f is harmonic and is defined within the void space.
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
r2 f ¼ 0 ð70Þ
n rf ¼ n ð71Þ
Figure 3. Continued.
(continued)
at Apq with f being periodic. Example unit cells which can be used for the
paper structure are shown in Fig. 3. Note that if we define the tortuosity
as is conventionally used by
¼ "Da D1
p
ð72Þ
the tortuosity is a tensor field representing the pore space. This definition
of the tortuosity is in terms of the effective diffusivity Dp. When Dp is in
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
Figure 3. Continued.
arrays have been widely investigated and numerical and analytical results
are available for these cases. When the structure is an array of spheres or
ellipsoidal particles which are not touching (so-called unconsolidated
systems) analytical solutions are also available. For the case of
consolidated systems solutions have to be obtained numerically.
Random walk simulations of Brownian motion have been shown to be
particularly effective for this purpose. (see Goel et al.[57,58]). An interest-
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
Z
"q 1
D q ¼ D q0 I þ ðng þ gnÞ dA ð74Þ
¼ ¼ ¼ V 2
Aqp
This equation allows us to separate the effect of the geometry of the fiber
space from the diffusion process intrinsic to the fibers. Analogous to the
void space, the tortuosity for the fiber matrix is
ð1 "ÞDq0i
qi ¼ ð75Þ
Dqi
cavities within the cell wall at least some of which survive in the final
paper. Crystalline cellulose is relatively inactive towards water molecules
and it is the amorphous cellulosic regions that are thought to absorb
water in considerable amounts. The state of moisture in the cell wall is
subject to much speculation although it is clear that it should be in the
condensed form. Water condensing inside the cell wall can also create
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
its own pores, which were not present in the dry state (the so-called
‘‘transient pores’’). Condensation of water inside the cell wall causes
stresses resulting in significant expansion of the cell wall in the radial
(transverse) dimension. This will yield better fiber–fiber contact at high
moisture contents and also crowd out the pore space in the sheet.
Individual pores inside the cell wall are considerably anisotropic and in
the transverse plane, their size is of the order of 10 nm.[59,60] The pore
sizes in the axial direction of the fiber seem to be much larger, of the order
of a few micrometers.[61]
As moisture condenses, it penetrates the outer surface and is
absorbed by the amorphous regions inside the cell wall. We can identify
two different models of moisture inside the cell wall. The first view is that
moisture condenses within the voids created in the inter-lamellar spaces
akin to capillary condensation. In the second model (as recently
explained by Foss et al.[31]), moisture binds with the amorphous regions
solvating their hygroscopic components forming a polymer solution.
Based on the shape of the sorption isotherm, Foss et al. proposed that
moisture inside the cell wall can reside in two forms—a solvated polymer
phase and as a free solvent phase. Other workers have also advanced
similar concepts that water exists in the cell wall in multiple phases.
Distinction between the phases is made based on their binding energies.
Bound and free water based on whether it is of the freezing or nonfreez-
ing type has been proposed by Maloney and Paulopuro[62] (see also
Karenlampi et al.[63]) Harding et al.[64] used magnetic resonance imaging
to measure the diffusion of water molecules in the intra-fiber matrix of
paper sheets. The shape of the relaxation of the resonance signal could be
best described by assuming that water in the cell wall existed in bound
and free forms and thus showed two different scales for diffusion.
Harding et al.’s model adds more complexity to our picture of diffusion
inside the cell wall by requiring it to be in two stages. Experimental
evidence from differential scanning calorimetry (DSC) measurements
seem to support this view of further distinctions in the properties of
the condensed water held within the fibers (see Maloney and
Paulopuro[62] and Karenlampi et al.[63]).
Despite which model for water is chosen inside the cell wall, the
diffusivity should be a strong function of moisture content besides
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
pc ¼ ð78Þ
rc
If the pore structure of the cell wall is known, the capillary pressure
relation can be deduced. Note that the surface tension and contact
angle shown here represent the values for condensed water inside the
nanopore of the cell wall. This could conceivably be different from the
equilibrium values of these for water on cellulose. This conceptual model
can be used for determining the dependency of the intrinsic moisture
diffusivity inside the cell wall. We further expect the pore permeability
to be a function of its surface recognizing that the nanopores probably
are bounded by the microfibril sheets. Since the microfibrillar angle could
change between fibers of different origins, we can expect the nanopore
permeability to be effected by the microfibrillar angle. Our model of
diffusion can tie in fiber micro (and nano) structural details including
their chemical composition with moisture migration properties.
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
dDqi ðqÞ
¼ qn ð79Þ
dq
Dqi ðqÞ ¼ þ qnþ1 ¼ þ Dq0 qm ð80Þ
nþ1
where is the diffusivity through the cell wall under completely dry
conditions and Dq0 and m are redefined constants in accordance with
the above relation. The coefficient Dq0 is identified as the diffusivity at
a fixed value of the moisture content. Perhaps a good fixed point for q on
the sorption isotherm would correspond to the moisture content at
saturation, namely at 100% RH when the fiber can be taken to be at
its fiber saturation point. If the moisture diffusivity at this point is known,
Dq0 may be estimated for further purposes easily. Ahlen’’s steady state
diffusion experiments described in Sec. 2 indicated that under dry
conditions i.e., at asymptotically zero relative humidity conditions, the
diffusion of moisture through paper sheets is only in the water vapor
form. That is, the pore diffusion mechanism dominates. This is supported
by the experimental observation that the ratio of the moisture diffusivity
to that of the nonadsorbing gas molecules (Ar or nitrogen) is in the
correct ratio according to Chapman-Enskog theory assuming gaseous
diffusion. Hence, we may conclude that the dry diffusivity contribution
denoted by in Eq. (83) is negligible and it can be simplified to
The above equation can also be supported by the observation that the
dependence of paper’’s electrical conductivity on moisture content
assumes a similar power law form (see e.g., Baum[65]).
m
q
¼ ð82Þ
0 q0
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
Earlier work for the fiber diffusivity Dq was based on Eq. (3) used
by Ahlen and also by Radhakrishnan et al.[16] The intrinsic fiber wall
diffusivity Dq has been modeled by an exponential function as given by
Eq. (2).
A theoretical basis for the exponential model [Eq. (2)] was provided
by Chatterjee and Gupta[66] who postulated that the diffusivity represents
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
dDq dNp
Np ð83Þ
dq dq
resulting in
that for acute microfibrillar angles, moisture diffusion inside the cell
wall is enhanced along the axial direction of the fibers as shown in
Figs. 4 and 5. Diffusion in other orthogonal directions may not be
affected (along the thickness) or adversely effected (along the radial
dimension) by this angle. Furthermore, the water absorbing capability
of the fibers themselves can be expected to be critical. Whether one views
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
Figure 5. Schematic representation of the internal structure of the fiber cell wall
showing microfibrillar lamellas and configuration for water migration.
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
NOMENCLATURE
ACKNOWLEDGMENTS
REFERENCES
3. Nissan. 1975.
4. Hashemi, S.J.; Gomes, V.G.; Crotogino, R.H.; Douglas, W.J.M. In-
plane diffusivity of moisture in paper. Drying Tech. 1997, 15 (2),
265–294.
5. Joy. 1953.
6. Rounsley, R.R. Vapor transport through paper. Tappi 1964, 47 (2),
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
95–98.
7. Ahlen, A.T. Diffusion of sorbed water vapor through paper and
cellulose film. Tappi J. 1970, 53 (7), 1320–1326.
8. Amiri, R.; Hamel, J.; MacDonald, J.D. Moisture distribution in
paper rolls: the effect of wrapping materials. J. Pulp Pap. Sci.
2002, 28 (5), 143–150.
9. Hartley and Richards. 1974.
10. Lee and Hinds. 1982.
11. Kobari. 1985.
12. Nilsson, L.; Wilhelmsson, B.; Stenstrom, S. The diffusion of
water vapor through pulp and paper. Drying Tech. 1993, 11 (6),
1205–1225.
13. Nilsson, L.; Stenstrom, S. Gas diffusion through sheets of fibrous
porous media. Chem. Engng. Sci. 1995, 50 (3), 361–374.
14. Liang, B.; Fields, R.J.; King, J.C. The mechanisms of transport
of water and n-propanol through pulp and paper. Drying Tech.
1990, 8 (4), 641–665.
15. Radhakrishnan, H.; Chatterjee, S.G.; Ramarao, B.V. Moisture
Transport Through a Bleached Kraft Paperboard Stack in a
Diffusion Cup Apparatus. ESPRA Research Report No. 109, 63–74:
Oct; Empire State Paper Research Institute: Syracuse, NY 13210,
USA, 1998.
16. Radhakrishnan, H.; Chatterjee, S.G.; Ramarao, B.V. Steady-state
moisture transport in a bleached kraft paperboard stack. J. Pulp
Pap. Sci. 2000, 26 (4), 140–144.
17. Chatterjee and Gupta. (1998).
18. Gupta, H.; Chatterjee, S.G. Steady State Moisture Transport in
Paper Investigated with a Parallel Diffusion Concept. ESPRA Res.
Report, SUNY ESF: Syracuse NY, 117, 2002; 83 pp.
19. Neogi, 1983a.
20. Neogi. 1983b.
21. Steele. 1966.
22. Lin, S.H. Moisture absorption in cellulosic materials. Int. J. Engng.
Sci. 1990, 28 (11), 1151–1156.
23. Roisum, D.R. Moisture effects on webs and rolls. Tappi 1993, 76 (6),
129–137.
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
24. Ten Donkelaar, J.R.; Jaeger, C.R. Transport and Sorption of Water
Vapor in Corrugated Board, Proc. 4th Intl. Symp. on Moisture and
Creep Effects in Paper, Board and Containers, Serra-Tosio, J-M.,
Vullierme, I., Eds.; EFPG, Grenoble, France, 1999; 245–255.
25. Lescanne, Y.; Moyne, C.; Perre, P. Diffusion mechanisms in a sheet
of paper. Drying 1992, 92, 1017–1026.
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
76. Bernada, P.; Stenström, S.; Månsson, S.; Experimental study of the
moisture distribution inside a pulp sheet using MRI. part II:
drying experiments. Journal of Pulp and Paper Science 1998, 24
(12), J373–J379.
77. Foss, W.R.; Bronkhorst, C.A.; Bennett, K.A. A Model of
Simultaneous Heat and Mass Transport in Paper During Moisture
Downloaded by [University of California Santa Cruz] at 13:42 25 November 2014
Sorption from Air. Proc. 4th Intl. Symp. on Moisture and Creep
Effects on Paper, Board and Containers, 1999; Serra-Tosio, J.-M.,
Vullierme, I., Eds.; 267–279.
78. Holmstad, R.; Ramaswamy, S.; Goel, A.; Gregersen, O.W.;
Ramarao, B.V. Comparison of 3D Structural Characteristics of
High and Low Resolution X-ray Microtomographic Images of
Paper and Board. Proc. Int. Pap. Phys. Conf., Victoria BC, 2003.
79. Ketoja, J.A.; Aurela, B.; Niskanen, K.J. Fiber networks as diffusion
barriers. Proc. Barrier Coating in Packaging, Tampere 2000.
80. Li, T.Q.; Henriksson, U.; Odberg, L. Water diffusion in wood pulp
fibers. Production, Analysis, Structure and Modification 193–198.
81. Neogi, P. Anamolous diffusion of vapors through solid polymers.
Parts I & II. AIChE J. 29 (5), 829–839.
82. Nilsson, L.; Mansson, S.; Stenstrom, S. Measuring moisture gradi-
ents in cellulose fiber networks—an application of the magnetic-
resonance-imaging method. Journal of Pulp and Paper Science
1996, 22 (2), J48–J52.
83. Ochoa, J.A.; Stroeve, P.; Whitaker, S. Diffusion and reaction in
cellular media. Chem. Engng. Sci. 1986, 41 (12), 2999–3013.
84. Ochoa-Tapia, J.A.; Stroeve, P.; Whitaker, S. Diffusive transport in
two-phase media: spatially periodic models and Maxwell’s theory
for isotropic and anisotropic systems. Chem. Engng. Sci. 1994,
49 (5), 709–726.
85. Pietikainen, R. Modeling Heat and Mass Transport in The Paper
Web in Calendering. Proc. Nordic Conf. on Paper Calendering,
June 29–20, 2000; 31–44, 79–92.
86. Ramaswamy, S.; Ramarao, B.V.; Goel, A.; Lee, G.; Choi, D.;
Lavrykov, S. Theoretical and Experimental Investigation on Liquid
Penetration in Porous Media. Proc. Prog. Paper Physics Seminar
2002; Keller, D., Ramarao, B.V., Eds.; SUNY ESF: Syracuse NY,
136–140.
87. Saez, A.E.; Otero, C.J.; Rusinek, I. The effective homogeneous
behavior of heterogeneous porous media. Transport in Porous
Media 1989, 4, 213–238.
88. Stamm, A.J. Wood and Cellulose Science. The Ronald Press
Company: NY, 1964.
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.