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

net/publication/237888442

Thin-Layer Drying of Flax Fiber: I. Analysis of Modeling Using Fick's Second

Law of Diffusion

Article in Drying Technology · December 2006

DOI: 10.1080/07373930601031430

CITATIONS READS

38 922

4 authors:

Ahmad Ghazanfari Shahram Emami

Shahid Bahonar University of Kerman POS Bio-Sciences
40 PUBLICATIONS 632 CITATIONS 39 PUBLICATIONS 678 CITATIONS

SEE PROFILE SEE PROFILE

Lope G Tabil Satyanarayan Panigrahi

University of Saskatchewan Saskatchewan Polytechnic
324 PUBLICATIONS 12,274 CITATIONS 102 PUBLICATIONS 4,800 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Lope G Tabil on 02 December 2015.

The user has requested enhancement of the downloaded file.

This article was downloaded by: [University of Saskatchewan]
On: 6 May 2009
Access details: Access Details: [subscription number 731835059]
Publisher Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

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

Thin-Layer Drying of Flax Fiber: I. Analysis of Modeling Using Fick's Second

Law of Diffusion
A. Ghazanfari a; S. Emami b; L. G. Tabil b; S. Panigrahi b
a
Department of Agricultural Machinery, Shahid Bahonar University, Kerman, Iran b Department of Agricultural
and Bioresource Engineering, University of Saskatchewan, Saskatoon, Saskatchewan, Canada

Online Publication Date: 01 January 2006

To cite this Article Ghazanfari, A., Emami, S., Tabil, L. G. and Panigrahi, S.(2006)'Thin-Layer Drying of Flax Fiber: I. Analysis of
Modeling Using Fick's Second Law of Diffusion',Drying Technology,24:12,1631 — 1635
To link to this Article: DOI: 10.1080/07373930601031430
URL: http://dx.doi.org/10.1080/07373930601031430

PLEASE SCROLL DOWN FOR ARTICLE

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

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

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.
 Drying Technology, 24: 1631–1635, 2006
Copyright # 2006 Taylor & Francis Group, LLC
ISSN: 0737-3937 print/1532-2300 online
DOI: 10.1080/07373930601031430
Thin-Layer Drying of Flax Fiber: I. Analysis of Modeling
Using Fick’s Second Law of Diffusion
A. Ghazanfari,1 S. Emami,2 L. G. Tabil,2 and S. Panigrahi2
1
Department of Agricultural Machinery, Shahid Bahonar University, Kerman, Iran
2
Department of Agricultural and Bioresource Engineering University of Saskatchewan,
Saskatoon, Saskatchewan, Canada
Thin-layer drying of moist flax fiber was performed at four tem-
peratures of 30, 50, 70, and 100
C with a constant absolute humidity
Downloaded By: [University of Saskatchewan] At: 19:52 6 May 2009
of 0.0065 kg water per kg dry air. The coefficients of diffusion of the
fiber at different drying conditions were estimated by modeling the
drying process using the one- to five-term solutions of the second
Fick’s law of diffusion. The models underestimated the drying
process during the initial stages of drying and overestimated this
process during the final stages. The estimated coefficient of diffu-
sions ranged from 5.11 3 10 9 to 1.92 3 10 8 m2/s and linearly
increased with the drying air temperature.
Keywords Coefficient of diffusion; Drying; Flax fiber; Thin-
layer
INTRODUCTION
Flax (Linum usitatissimum L.) fibers have been proven
to be a suitable reinforcement material for composites.
These fibers have good mechanical properties and numer-
ous environmental advantages.[1] Composite applications
of natural fiber are increasing, but still limited due to their
low bonding with their polymer matrix, and due to non-
uniform physical and mechanical behaviors.[2] Flax fibers
are usually treated with various chemical agents to enhance
their physio-mechanical properties.[3,4] For further proces-
sing, i.e., mixing with polymers or grinding, the fiber must
be washed to remove the chemical residues and dried.
Flax fibers, like most other agricultural products, are
heterogonous in nature; thus, their drying involves com-
plex heat and mass transfer between the surrounding
air and the fiber and within the fiber.[5] In drying of
high moisture content material, two distinct periods are
observed: First, the constant rate period, which may occur
during the initial stages of drying when there is excess
moisture on the surface of the drying material and the dry-
ing rate is controlled by amount of evaporation from the
Correspondence: Dr. Ahmad Ghazanfari, College of Agricul-
ture, Shahid Bahonar University, Kerman, Iran; E-mail:
ahg212@mail.usask.ca
1631
surface; the second is the falling drying rate period, which
initiates when the rate of moisture migration to the surface
is less than the rate of evaporation from the surface.[6,7] The
constant rate period compared to the whole drying dur-
ation is very short and for most products it does not even
occur.[7] Thus, in most practical drying simulations, model-
ing of the drying process is done only for the falling rate
period.
Mathematical modeling of any drying process is impor-
tant for optimum management of operating parameters
and predicting the performance of the drying system.
Due to their complexity, drying of agricultural products
or materials cannot be precisely described by mathematical
formulations.[7] In analyzing the drying process of these
materials, often a combination of theoretical and empirical
models are used. Empirical equations in drying simulation
are easy to apply and have been widely used. The main jus-
tification for the empirical approach is the satisfactory fit
to the experimental data.
Thin-layer drying has been considered a suitable process
for drying of numerous agricultural products. Numerous
mathematical models describing thin-layer drying by con-
vection have been published in the literature for agricul-
tural crops.[8–10] Investigators have tried to explain the
drying behavior of various agricultural products by model-
ing their drying process.[11,12] Thin-layer solar drying has
been considered for many agricultural products.[13–15]
Drying of flax fiber, like drying of other agricultural
products, is an energy-consuming process. Currently, flax
fibers are dried by tray or rotary dryers, but no infor-
mation on enhancing the flax fiber drying process is avail-
able. The efficiency of a drying system can be improved by
analysis of the drying process. Thin-layer drying models for
drying flax fiber at different temperatures are needed to
enhance the flax drying process. Simulating flax fiber dry-
ing provides a valuable tool for the prediction and optimi-
zation of the performance of the drying process and
subsequently reducing time and energy consumption.
 1632 GHAZANFARI ET AL.
The objectives of this study were (a) to study the changes
in moisture content of flax fiber during thin-layer drying at
different drying conditions; (b) to investigate the applica-
bility of the series solution of Fick’s second law of diffusion
for modeling thin-layer drying of flax fiber; and (c) to
determine the coefficient of effective diffusion of flax fiber
and its dependence on temperature.
MATERIAL AND METHODS
Thin-Layer Dryer
The thin-layer dryer consisted of an environmental
chamber, a controllable fan, a three-tray drying cabinet,
and a circulating duct. Figure 1 presents the schematic
drawing of the dryer. The required temperature and rela-
tive humidity for a drying process were provided by the
environmental chamber (Model AH-213, Bryant Manufac-
turing Association, Boston, MA). The three-tray cabinet
Downloaded By: [University of Saskatchewan] At: 19:52 6 May 2009
was placed on an electronic scale equipped with a data log-
ger that was interfaced with a personal computer. The
changes in the mass of the samples were monitored using
the LabView software (version 6.0, National Instruments
Corporation, Austin, TX) installed on the computer. The
air flow rate was controlled by controlling the speed of
the fan and it was maintained at 1.0 m=s throughout the
experiments.
Material Preparation
The flax fiber used for these experiments was provided
by Biolin Research, Saskatoon, Saskatchewan, Canada.
The flax fiber was kept at laboratory conditions (24
C
and 45% RH) for 48 h. Then, the moisture content of the
fiber was determined to be 7% wet basis (wb). The fiber
was soaked in tap water for 30 min. Then it was removed
from water and placed in a top-loading commercial spin
dryer and allowed to spin for 20 min to remove excess of
water. The spin-dried fiber was placed in a plastic bag
and kept at room temperature for 6 h for its moisture
FIG. 1. A schematic diagram of the thin-layer dryer, A: environmental
chamber, B: fan, C: drying tunnel, D: trays, E: electronic scale, F: com-
puter, G: circulating air tunnel.
content to reach an equilibrium level. The moisture content
of the flax fiber prior to the drying treatments was determ-
ined to be 54% (wb). For each drying treatment, a sample
of 150 g moist flax fibers was spread on perforated trays
which gave an overall layer thickness of 10 mm. The
trays were placed in the dryer tunnel and exposed to drying
air. Each drying process was repeated two times and the
average of the two runs was used for drying analyses.
Drying Procedure
In a typical drying process, the surrounding air is heated
and used as drying medium. In this heating process, the
absolute humidity remains constant. For our experiments,
the temperature of ambient air was 20
C with 45% relative
humidity (RH). These conditions correspond to an absol-
ute humidity of 0.0065 kg water per kg of dry air. The sur-
rounding air was heated in order to dry the flax fiber at 30,
50, 70, and 100
C. Thus, the resulting drying conditions
were (1) 30
C, 24% RH; (2) 50
C, 8% RH; (3) 70
C,
3.5% RH; and (4) 100
C, 1% RH. Prior to any experi-
ments, the dryer was allowed to run without any fiber until
the desired drying conditions were achieved. During each
experiment, temperature and RH of the drying air was
monitored and recorded. The mass of samples on the dry-
ing trays was recorded every 60 s using the data logger. All
drying processes were allowed to continue until there was
no appreciable change in the moisture content of the fiber
for at least 2 h. The moisture content of the fiber at the end
of each drying process was considered as the equilibrium
moisture content (EMC) for subsequent modeling.
Modeling Procedure
A widely accepted mechanism for moisture transfer dur-
ing drying of agricultural products is Fick’s second law of
for liquid=vapor diffusion[16] described by:

2 
@M @ M @2M @2M
¼D þ þ ð1Þ
@t @x2 @y2 @z2
where M (g) is the moisture, at location of x, y, z within the
object at the time of t (s), and D is the diffusion coefficient
(m2=s). When diffusion coefficient is known, the solution of
Eq. (1) can be used to describe the change in moisture con-
tent as a function of time. The coefficient of diffusion of a
material is not constant throughout a drying process and it
depends on the temperature, moisture content, and the
structural changes of the material. For most drying analy-
ses an average effective diffusion coefficient, De, is used.
With a suitable set of initial and boundary conditions
and considering an average effective diffusion coefficient
for the material throughout the drying process, the analyti-
cal solution to Eq. (1) for slab geometry with a thickness of
L can be expressed in the form of a series of terms:[7]
 DRYING OF FLAX FIBER: I 1633
!
8X1
1 ð2n þ 1Þ2 p2 De t
MR ¼ 2 exp  ð2Þ
p n¼0 ð2n þ 1Þ2 4L2

where MR is called moisture ratio and it is calculated by:

M  Me
MR ¼ ð3Þ
M0  Me
where Me is the equilibrium moisture content (% wb)
and Mo is the initial moisture content (% wb) of the slab.
Taking A ¼ 8=p2 and B ¼ p2=(4L2), then one-, two-, and
three- term forms of Eq. (2) are
MR1 ¼ A expðBDe1 tÞ ð4Þ
1
MR2 ¼ A expðBDe2 tÞ þ Að9BDe2 tÞ ð5Þ
9
1
MR3 ¼ A expðBDe3 tÞ þ A expð9BDe3 tÞ FIG. 2. The actual drying progress of flax fiber as a function of drying
9 temperature.
Downloaded By: [University of Saskatchewan] At: 19:52 6 May 2009

1
þ A expð25BDe3 tÞ ð6Þ
25
achieved at 342, 155, 120, and 72 minutes after initiation
For each treatment, the instantaneous MR values were cal- of drying process, as indicated in Table 2.
culated using Eq. (3). Then each of the models (one- to The moisture ratios calculated from the drying data were
five-term solutions) was fitted to their corresponding plotted versus drying time for different drying conditions.
sequence of time and MR data and the De values were esti- The graphs and their corresponding fitted one- and three-
mated using TableCurve software (Jandel Science, CA). term models are presented in Figs. 3 to 6. In general, the
The goodness of fit for each model was evaluated by com-
paring coefficient of determination, R2, root mean square
error, RMSE,[9] obtained from the predicted and the TABLE 1
experimental MR values. The summary of the modeling results for the drying
The dependency of diffusivity on temperature was inves- treatments of flax fiber
tigated by plotting the obtained coefficient of diffusions
against drying temperatures. The trends were visually Temperature Series De
inspected and appropriate model was fitted to the data (
C) terms (m2=s  10 8) R2 RMSE
using the least square method. 30 1 0.5143 0.91 0.09167
2 0.5116 0.93 0.08319
RESULTS AND DISCUSSIONS 3 0.5112 0.93 0.08190
Figure 2 shows moisture ratio changes versus drying 4 0.5112 0.93 0.08151
time for the four drying conditions and the summary of 5 0.5112 0.93 0.08133
the drying process is presented in Table 1. As it is expected, 50 1 0.8917 0.90 0.09111
the drying in the higher temperature treatments took place 2 0.8899 0.93 0.08131
faster than the cooler treatments. The figure shows the 3 0.8899 0.93 0.07922
trends of drying to follow an exponential decaying func- 4 0.8899 0.93 0.07845
tion. The times for reaching the equilibrium moisture 5 0.8899 0.93 0.07809
content, as shown in Table 2, are significantly different 70 1 1.1528 0.92 0.10942
for all treatments and they decrease as drying temperature 2 1.1450 0.95 0.08534
increases. 3 1.1437 0.96 0.08033
The drying times were significantly different and the 4 1.1430 0.96 0.07863
drying time decreased as the temperature of the drying 5 1.1425 0.96 0.07791
air was increased. The equilibrium moisture content for 100 1 1.9195 0.90 0.07337
the 30, 50, 70, and 100
C treatments were 6.7, 3.1, 0.8, 2 1.9125 0.93 0.06329
and 0.1% wb, respectively. Thus, depending on the pur- 3 1.9124 0.93 0.06121
pose of drying, different drying schemes can be chosen. 4 1.9124 0.93 0.06044
The equilibrium moisture contents were respectively 5 1.9124 0.93 0.06007
 1634 GHAZANFARI ET AL.

TABLE 2
Summary of drying process data for thin-layer drying
of flax fiber
Temperature Relative humidity Me Drying time
(
C) (%) (% wb) (min)
30 24.0 6.7 342
50 8.0 3.1 155
70 3.5 0.8 120
100 1.0 0.1 72

FIG. 5. One- and three-term models fitted to the experimental data at

drying temperature of 70
C.
Downloaded By: [University of Saskatchewan] At: 19:52 6 May 2009

one-term solution underestimates the drying process during

FIG. 3. One- and three-term models fitted to the experimental data at
drying temperature of 30
C.
initial stage of drying. A close look at the graphs in Figs. 3–6
shows that the underestimation is from MR ¼ 1.0 to about
MR ¼ 0.3. This range covers about 1=3 of the drying per-
iod. Basically, the difference between using one-term or
more terms equations is more noticeable during the first
1=5 period of drying, and after this period the curves follow
exactly the same path. Even in the first 1=5 period, there was
no significant improvement in using the four- or five-term
equations over that of the three-term equation.
Table 1 summarizes the results of fitting the five equa-
tions to the drying data. By increasing the number of terms
in the models, the accuracy of fitting increases. This is also
indicated by the R2 values given in Table 1. Of course, the
fitted curve will not produce an exact MR equal 1, unless

FIG. 4. One- and three-term models fitted to the experimental data at FIG. 6. One- and three-term models fitted to the experimental data at
drying temperature of 50
C. drying temperature of 100
C.
 DRYING OF FLAX FIBER: I
FIG. 7. The dependence of the coefficient of diffusion on drying air
temperature.
Downloaded By: [University of Saskatchewan] At: 19:52 6 May 2009
the infinite terms of Eq. (2) are used in a model. However,
looking at the R2 values and RMSE in Table 1 indicates
that using more than three terms of the equation will not
appreciably improve the accuracy of fitting. The highest
MR values estimated by one- to five term equations were
0.81, 0.93, 0.95, and 0.96, respectively.
The minimum estimated De is 5.11  10 9 m2=s for 30
C
treatment when estimated by three-term equation and the
maximum value is 1.92  10 8 m2=s for 100
C when esti-
mated by one-term solution. A closer look at the De values
shows that by increasing the number of terms in the series
solutions the estimated value decreases to a limit. However,
no appreciable change in any estimated De was noticed
with increasing the number of terms above three. All the
above analyses indicate that using a three-term solution is
accurate enough for modeling the drying flax fiber.
To investigate the dependence of De on drying air tem-
perature, the estimated De obtained from the three-term
model were plotted against temperature. The graph is
shown in Fig. 7, which indicates a linear trend with positive
slope between the two variables. Using the least-squares
curve-fitting method, the following linear relation with
R2 ¼ 0.98 was obtained to describe the relation between
De and temperature, T, of drying air:
De ¼ 1:96  1010 T  1:14  109
Thus, the effective coefficient of diffusion of flax fiber lin-
early increases with the temperature of drying air.
CONCLUSIONS
The thin-layer drying of flax fiber were modeled using
one- to five-term series solutions of Fick’s second law of
diffusion and the coefficients of diffusion at different dry-
ing conditions were estimated. Comparison of the RMSE
View publication stats
1635
and R2 values obtained from the experimental data and
the fitted models and also visual inspection of the fitted
curves indicated that, in general, the Fick’s diffusion model
underestimate the drying process during the initial stages
and overestimate it during the final stages of drying. The
accuracy in the estimate of the coefficient of diffusion
increased with increased number of terms in the model,
but using more than a three-term model will not markedly
improve the accuracy of the estimation. The effective dif-
fusion coefficients of the flax fiber were linearly a function
of drying air temperature.
REFERENCES
1. Arbelaiz, A.; Fernandez, B.; Ramos, J.A.; Retegi, A.; Laano-Ponte,
R.; Mondragon, I. Mechanical properties of short flax fiber bundle=
polypropylene composites: Influence of matrix=fiber modification,
fiber content, water uptake and recycling. Composites Science Tech-
nology 2005, 65, 1582–1592.
2. Cantero, G.; Arbelaiz, A.; Llano-Ponte, R.; Mondragon, I. Effects of
fiber treatment on wetability and mechanical behavior of flax=
polypropylene composites. Composites Science and Technology 2003,
63, 1247–1254.
3. Wang, B.; Panigrahi, S.; Crerar, W.; Tabil, L. Application of pre-
treated flax fiber in composites; CSAE=SCGR Paper no. 03–367,
CSAE=SCGR: Mansonville, QC, Canada, 2003.
4. Weyenberg, V.; Ivens, J.; De Coster, A.; Kino, B.; Baetens, E.;
Verpoest, I. Influence of processing and chemical treatment of flax
fibers on their composites. Composite Science and Technology 2003,
63, 1241–1246.
5. Liu, J.Y.; Simpson, W.T. Two-stage moisture diffusion in wood with
constant transport coefficients. Drying Technology 1999, 17 (1 & 2),
257–269.
6. Lopez, A.M.; Pique, T.; Romero, A. Simulation of deep bed drying of
hazelnuts. Drying Technology 1998, 16 (3–5), 651–665.
7. Charm, S.E. Fundamentals of Food Engineering, 3rd Ed; AVI Publish-
ing Co. Inc.: Westport, CT, 1978.
8. Hossain, M.A.; Bala, B.K. Thin-layer drying characteristics for green
chili. Drying Technology 2002, 20 (2), 489–505.
9. Basunia, M.A.; Abe, T. Thin-layer drying characteristics of rough
rice at low and high temperatures. Drying Technology 1998, 16 (3–5),
579–595.
10. Adu, B.; Otten, L.; Brown, R.B. Modeling thin layer microwave
drying of soybeans. Canadian Agricultural Engineering 1994, 36 (3),
135–141.
11. Lopez, A.M.; Pique, T.; Boatella, J.; Ferran, A.; Garcia, J.; Romero,
A. Drying characteristics of the hazelnut. Drying Technology 1998,
16 (3–5), 627–649.
12. Doymaz, I. Drying characteristics and kinetics of okra. Journal of
ð7Þ Food Engineering 2004, 69 (3), 275–279.
13. Arinze, E.A.; Sokhansanj, S.; Schoenau, G.J.; Crerar, B.; Opoku, A.
Design, experimental and economic evaluation of a commercial type
solar dryer for production of high quality hay. Drying Technology
1998, 16 (3–5), 597–626.
14. Togrul, I.T.; Pehlivan, D. Modeling of thin layer drying kinetics of
some fruits under open-air sun-drying process. Journal of Food
Engineering 2004, 65 (3), 413–425.
15. Ghazanfari, A.; Tabil, L.; Sokhansanj, S. Evaluating a solar dryer for
in-shell drying of split pistachio nuts. Drying Technology 2003, 21 (7),
1375–1368.
16. Crank, J. The Mathematics of Diffusion, 2nd Ed; Clarendon Press:
Oxford, UK, 1975.