Journal of Food Engineering 61 (2004) 279286

www.elsevier.com/locate/jfoodeng

Simultaneous heat and mass transfer applied to non-respiring

foods packed in modied atmosphere
R.J. Simpson *, S.F. Almonacid, C.A. Acevedo, C.A. Cort
es
Departamento de Procesos Qu
micos, Biotecnol

ogicos, y Ambientales, Universidad T

ecnica Federico Santa Mar
a, Casilla, 110-V Valpara
so, Chile
Received 5 October 2002; accepted 3 April 2003

Abstract
A mathematical model to predict heat and mass transport phenomena in non-respiring food packed in modied atmosphere
(MAP) was developed and validated. The model incorporates simultaneous gas convection, sorption, diusion, heat convection and
conduction. The model was applied to MAP systems containing CO2 , O2 , N2 and H2 O. Validation test was done with gelatin. The
average errors between experimental and simulated values were low: <0.6 (C) for the temperatures, 3% for relative humidity and
<1.43% for the headspace gas composition. Model predictions during heating and cooling phases indicate that temperature
modication of the packaged product can be quite slow, reecting the relevance of proper chilling in the packaging, transport and
storage processes of MAP products. The model applied to shelf-life studies, for specic products, can be utilized to identify facility
and product handling improvements to generate the greater positive impact on product quality. The development of this or similar
mathematical tools would allow for more technical and informed management decisions.
2003 Elsevier Ltd. All rights reserved.

Keywords: Modied atmospheres packaging; Heat and mass transfer; Non-respiring foods

1. Introduction reactions determine the gas transfer between the plant

material and ambient (Talasila, Chau, & Brecht, 1995;
An increasing demand of natural and minimally Emond, Chau, Bretch, & Ngadi, 1998; Lakakul,
processed food has been observed in the last few de- Beaudry, & Hernandez, 1999). In non-respiring foods,
cades. A large part of this demand can be satised where the gas diusion is the main transfer mechanism,
through the modied atmosphere packaging (MAP) Simpson, Almonacid, and Acevedo (2001) developed a
technique, which together with chilled storage results in mathematical model that describes the gas transfer be-
a signicant shelf life increase of fresh food, without tween the food, headspace and ambient at constant
major eect on its quality (Blakistone, 1999). temperature. Prior to consumption, foods are frequently
Most of technological advances in MAP systems have exposed to temperature uctuations, especially, dur-
been empirically obtained, as such, limiting its applica- ing transportation and storage, these temperature abu-
tion to the conditions at which the experiments were ses can signicantly diminish the useful food shelf life
carried out. (Almonacid & Torres, 1993; Koutsomanis, 2001).
Mathematical models are quantitative tools that An adequate estimation of the variation in gas com-
when applied in food allow for design and process op- position throughout time must consider the variations of
timization, permitting the generation of high quality temperature to which the MAP system was exposed.
products, and diminishing the time and cost of labora- Gas composition proles and temperature over time will
tory tests. On living MAP systems, most of the models nally determine the food shelf life.
are focused in fruits and vegetables, where respiration The objective of this research was to develop and
validate a heat and mass transfer model for non-
respiring food packed in modied atmosphere. The
model describes the temperature variation and gas com-
*
Corresponding author. Tel.: +56-32-654302; fax: +56-32-654478. position in the headspace and interior of non-respiring
E-mail address: ricardo.simpson@pqui.utfsm.cl (R.J. Simpson). food packed in modied atmospheres.
0260-8774/$ - see front matter 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0260-8774(03)00109-2
280 R.J. Simpson et al. / Journal of Food Engineering 61 (2004) 279286

Nomenclature

a Klein constant, dimensionless Sc Schmidt number

aC , aM proportionality coecients dened in Eqs. Si simulated value
(20) and (21), dimensionless Sh Sherwood number
aw water activity, dimensionless S0 constant (kmol/kg Pa)
A area (m2 ) t time (s)
c gas concentration in food (kmol/kg) T temperature (K)
cp specic heat (J/kg K) U overall heat transfer coecient (W/m2 K)
D diusion coecient (m2 /s) v partial volume (m3 )
D0 constant (m2 /s) V volume (m3 )
e thickness (m) W_ work ow (W)
E total energy (J/kg) x distance in the x-axis (m)
Ed activation energy diusivity (J/kmol) X mass fraction (kg/kg)
Ei experimental value y molar volumetric fraction (kmol/kmol)
Ep activation energy permeability (J/kmol) DHS heat of solution (J/kmol)
Gr grashof number q density (kg/m3 )
GrM grashof mass number r deviation between observed and predicted
hC heat convection coecient (W/m2 K) proles (C) or (%)
hM mass convection coecient (kmol/m2 s Pa) l viscosity (kg/m s)
H specic enthalpy (J/kg)
k thermal conductivity (W/m K) Subscripts
M mass (kg) a air
M_ mass ow (kg/s) F food
n number of moles per gas (kmol) Fs food surface
N number of experimental values Fb food bottom
Nu Nusselt number H headspace
p partial pressure (Pa) I initial
pv saturated water vapor pressure (Pa) p package
P permeability (kmol m2 /m3 s Pa) P polymer
P pressure (Pa) S aqueous solution
P0 constant (kmol m2 /m3 s Pa) sl saturated liquid
Pm molecular weight (kg/kmol) sv saturated vapor
Pr Prandtl number t time
Q_ heat ow (W) CO2 carbon dioxide
R ideal gas constant (Pa m3 /kmol K) O2 oxygen
Ra Rayleigh number N2 nitrogen
S solubility (kmol/kg Pa) H2 O water

2. Materials and methods

2.1. Mathematical model

The mathematical model was developed considering:

(a) a low fat food with an innite slab shape, (b) a
exible container, (c) heat transfer between package and
food is carried out by natural convection, and by con-
duction within the food (Fig. 1), (d) a headspace gas
mixture containing CO2 , O2 , N2 and H2 O, (e) changes in
gas composition results from gas sorption by food,
permeability through the packaging and water evapo- Fig. 1. Schematic diagram for heat transfer; food: conduction, head-
ration from food surface (Fig. 2). space: convection.
 R.J. Simpson et al. / Journal of Food Engineering 61 (2004) 279286 281

oTp hCa Ta  Tp  hCH Tp  TH

4
ot cp q ep

2.1.1.4. Food. Heat transfer in the food was obtained

through the second Fouriers law:
oT o2 T
qF cpF kF 2 5
ot ox
Boundary conditions:
TF TI 0 6 x 6 eF t0 6
Fig. 2. Schematic diagram for mass transfer.
TF TFs x 0 t > 0 7


2.1.1. Heat transfer oTFb UAFb Ta  TFb  kF AFb oT
ox
x eF tP0
The following equations describe heat transfer ot M cpFb
mechanisms considered in the model. 8

2.1.1.1. Headspace. Eq. (1) was obtained by applying the The overall heat transfer coecient was mathematically
rst law of thermodynamics considering headspace as described as shown in Zuritz and Sastry (1986).
system:
_ CO H O dM_ O H N dM _ N H H O dM _ H O
2.1.2. Mass transfer
H CO2 dM IN
2 2 2 2 2 2 2
The following equations describe mass transfer
 H CO dM _ CO H O dM _ O H N dM _ N H H O dM _ H O
mechanisms considered in the model.
2 2 2 2 2 2 2 2 OUT
oEH
dQ_  dW_ 1 2.1.2.1. Gas ow through package. The gas ow through
ot
The ideal gas law was applied to each gas, energy terms the lm and tray was estimated as:
   
associated to gases, entrance and exit, were neglected in on PA
P yH  ya
9
comparison with those associated to heat transfer. ot p e p
Considering headspace temperature as uniform, Eq.
(1) can be written as:
oTH 2.1.2.2. Diusion and gas sorption by food. Gas transfer
in the food was obtained from the Flicks second law:
ot
RT hCH Ap Tp  TH  hCH AFs TH  TFs
oc o2 c
D 2 10
P cpPmvCO2 cpPmvO2 cpPmvN2 cpPmvH2 O
ot ox
2 Boundary conditions:
c Spa 0 6 x 6 eF t0 11
2.1.1.2. Food surface. Eq. (3) was obtained by applying c SpH x0 t>0 12
the rst law of thermodynamics in a volume element on
oc
food surface. Neglecting the energy terms associated to 0 x eF tP0 13
gas ow and mechanical work in comparison with those ot
associated to the heat transfer and water evaporation The gas sorption by food as a function of time was ex-
becomes: pressed as:

h i
oTFs H H2 Osv  H H2 Osl PmH2 O hMH2 O AFs pHH2 O  pv aw
hCH AFs TH  TFs  kF AFs oToxF
3
ot q cp V Fs

   
2.1.1.3. Package. Again, the rst law of thermodynamics on oc
was applied to the lm and tray, energy terms associated MF 14
ot F ot
to gases entrance and exit were neglected in comparison
with those associated to heat transfer. Given the absence The mean gas concentration in the food was calculated
of work in the system the energy balance becomes: by
282 R.J. Simpson et al. / Journal of Food Engineering 61 (2004) 279286

R MF mixtures and low mass transfer rates, the rule for ex-
cdM
c R0 MF 15 ploiting the analogy is simple: The Sherwood number is
0
dM obtained by replacing the Prandtl number by the Sch-
midt number in the appropriate heat transfer correlation
2.1.2.3. Water evaporation from food surface. Water (Mills, 1995).
evaporation from food surface toward headspace was Ambient heat transfer coecient (hCa ) for chilling
estimated as: (forced convection) was determined by means of a non-
  linear regression using an aluminum block (Bonacina &
onH2 O Comini, 1973) and a value of 20.7 W/m2 K was calcu-
hMH2 O AFs pHH2 O  pv aw
16
ot F lated. For the heating condition (natural convection)
hCa was determined by correlating Nu with Pr and Gr,
2.1.2.4. Mass balance for headspace. In the development using the aluminum block test as it was used to deter-
of Eq. (17), gas consumption and production due to mine the headspace transfer coecient (hCH ). The fol-
microbial activity was neglected. Microbial growth was lowing relationship was obtained:
minimized due to the refrigeration conditions. A exible Nu 1:33  Pr Gr
1=4
22
package (constant pressure and variable volume) and
ideal gas law were considered to derive the following The pure water vapor pressure was estimated through
equation. the Antoines equation (Perry & Green, 1997). The wa-
2 h i3 ter vapor diusivity was estimated from Marrero and
    o 1
on on P 6 1 oVH TH 7 Mason (1972).
  4 VH 5 17 The gas properties were estimated with valid corre-
ot p ot F R TH ot ot
lations for gases at low pressure (Wassiljewa, 1904;
Headspace gas composition was obtained as: Herming & Zipperer, 1936; cit. by Perry & Green, 1992).
vi Table 1 show data utilized in those correlations.
yi 18 Physical properties of the gelatin were calculated using
VH
the models developed by Murakami and Okos (1980).
Headspace volume change was obtained as: The water activity was measured with an ROTRONIC
VH vCO2 vO2 vN2 vH2 O 19 HYGROMER instrument model AwV C. Table 2 shows
parameters utilized.
The equations were solved using an explicit nite dif-
ference scheme (Crank, 1975).
Table 1
Specic heats, thermal conductivities and gases viscosities at 5 C and
2.2. Model parameters
1 atma

The heat and mass natural convection transfer coef- Property CO2 O2 N2 H2 O
cients on headspace were obtained by correlating Nu cp (J/kg K) 841 919 1040 1877
with Pr and Gr, and Sh with Sc and GrM (Bird, Steward, k (W/m K) 1.49  102 2.48  102 2.43  102 1.85  102
l (kg/m s) 1.39  105 1.95  105 1.69  105 8.17  106
& Lightfoot, 1964). a
Vargaftik (1975).
1=4
Nu aC  Pr Gr 20
1=4
Sh aM  Sc  GrM 21
Table 2
The parameter aC was obtained through tests in isolated Physical properties of the gelatina
containers. Tests consisted of placing an aluminum Property k (W/m K) q (kg/m3 ) cp (J/kg K) aw
block inside the container and registering the surface Gelatin 0.56 1016 4040 0.990
temperature of the aluminum block, the headspace core a
Murakami and Okos (1980).
and upper internal surface of the container in transient
conditions. Then, considering that the heat transfer in
the aluminum block responds to a Biot number <0.1 Table 3
(Welty, Wicks, & Wilson, 1984), an energy balance was Solubility for CO2 , N2 and O2 in water (kg/kg Pa)a
carried out to obtain the relationship of the aluminum Temperature (C) CO2 O2 N2
block and headspace temperatures with the heat transfer 8 10
0 3.09  10 6.88  10 2.90  1010
coecient. Finally, by means of a non-linear regression, 10 2.32  108 5.37  1010 2.30  1010
experimental data was adjusted as described, so that, 20 1.70  108 4.38  1010 1.91  1010
parameter aC was obtained (Bonacina & Comini, 1973). 30 1.30  108 3.69  1010 1.66  1010
The parameter aM was estimated using the analogy 40 1.04  108 3.28  1010 1.48  1010
a
between convective heat and mass transfer. For dilute Geankoplis (1982).
 R.J. Simpson et al. / Journal of Food Engineering 61 (2004) 279286 283

Table 4  
Diusivity for CO2 , N2 and O2 in water (m2 /s)
Ed
D D0 exp 27
RT
Tempera- CO2 O2 N2 Source
ture (C)
Tables 6 and 7 show parameters for Eqs. (26) and (27).
10 1.46  109 Treybal (1980)
18 1.98  109 Geankoplis
(1982) 2.3. Validation experiments
20 1.78  109 2.09  109 Brown (1955)
20 1.64  109 Bolz and Tuve The gas transfer validation was carried out using non-
(1976)
avor gelatin 5 (% w/w) with an average weight of 289 g
25 2.00  109 2.41  109 Geankoplis
(1982) (1.25 cm thickness), packages PP/EVOH/PP and a lm
25 1.90  109 Perry and PET/(PP/EVOH/PP) of 82 lm provided by EMPACK
Green (1992) Co. Santiago, Chile. Forty samples of gelatin and
1030.05 cm3 of gas mixture (43.57 (% v/v) CO2 , 11.71 (%
Gas solubility and diusivity in the gelatin were es- v/v) O2 (rest N2 )) were packed at 25 C in an ILPRA
timated from values in aqueous solution (Brown, 1955; FOOD PACK model Basic V/G and stored at 6 C for 7
Bolz & Tuve, 1976; Treybal, 1980; Geankoplis, 1982; days and at 3 C for 2 days in a REVCO refrigerator
Perry & Green, 1992), Tables 3 and 4, and by means of model REL 5004V14. Sample extraction was done in
the following equations (Robertson, 1992); triplicate. In each sample, CO2 and O2 concentration
was measured on headspace by a PBI DANSENSOR
SP SS 1  XP 23 model CheckMate 9900 O2 /CO2 gas analyzer. Mea-
DP DS expa yP 24 surements were done using destructive assays.
Information related to package characteristics and per- The heat transfer and relative humidity validation
meabilities were provided by EMPACK Co., Santiago, was done exposing the system (food-package) to tem-
Chile (Table 5). peratures ranging from 0 to 25 C. The system was
Temperature dependence on permeability, solubility chilled (forced convection) at 0 C in a HECO modular
and diusivity was estimated from Arrhenius type refrigerator chamber, and heated at a room temperature
models (Robertson, 1992). of 21 C with no forced air circulation (natural con-
  vection). The reasons for selecting these temperatures
EP were to cover the refrigeration storage temperatures
P P0 exp 25
RT used commercially and the common ambient tempera-
 
DHS ture abuses. The relative humidity of the headspace was
S S0 exp 26 measured with a DUCT/OSA humidity transmitter
RT
model HD10K. Ambient, headspace and food (middle
and bottom) temperatures were measured by using
ELLAB T-type copperconstantan thermocouples
model G700TS-16. Temperatures and relative humidity
Table 5
Film permeability for CO2 , O2 , N2 and H2 O (kmol m2 /m3 s Pa)a
were monitored during heating and chilling and the data
were recorded using an OMEGA multi-channel data
Tempera- CO2 O2 N2 H2 O
logger model OM-220.
ture (C)
The deviations between experimental and simulated
5 1.04  1021 4.32  1022 4.32  1022 2.11  1017
values were calculated as:
23 5.23  1021 2.16  1021 2.16  1021 4.74  1017
s
a
Empack (1999). Pn 2
i1 Ei  Si
r 28
Table 6 N
Heat of solution (DHS ) and constant S0 for CO2 , O2 and N2 in gelatin
Additionally, t-tests (t student, p < 0:05) were per-
Parameter CO2 O2 N2
formed to compare experimental and simulated values.
S0 5.01  1012 1.86  1012 1.40  1012
DHS (J/kmol) )1.97  107 )1.32  107 )1.19  107

3. Results and discussion

Table 7
Activation energy (Ed ) and Constant D0 for CO2 , O2 and N2 in gelatin
3.1. Heat and mass convection on headspace
Parameter CO2 O2 N2
D0 6.18  107 7.71  106 1.06  105 For at-shaped non-respiring foods with a tempera-
Ed (J/kmol) 1.45  107 2.03  107 2.14  107
ture lower than headspace temperature (heating):
284 R.J. Simpson et al. / Journal of Food Engineering 61 (2004) 279286

1=4
Nu 1:46  Pr Gr 2:8  105 < Ra < 2:4  107
29
1=4
Sh 1:46  Sc  GrM 2:8  105 < Ra < 2:4  107
30
For at-shaped non-respiring foods with a temperature
higher than headspace temperature (cooling):
1=4
Nu 1:77  Pr Gr 4:4  105 < Ra < 3:6  107
31
1=4
Sh 1:77  Sc  GrM 4:4  10 < Ra < 3:6  107
5

32
Figs. 3 and 4 show the data obtained in headspace
correlations. Fig. 4. Nusselt versus Grashof in pack with 26 cm height on headspace.
In agreement with the literature values the aC coe-
cient for cooling (1.77) was higher than the heating value
characteristic of internal natural convection in enclo-
(1.46). When warm air encloses a cold body, the air just
sures (Mills, 1995). Thus, it would be unreasonable to
around the body will cool down and sink following a
seek a single simple correlation formula valid over a wide
path around the body to the oor. The cold air creates a
range of Rayleigh and Prandtl numbers. Correlations are
tiny insulation layer outside the package and may slow
usually valid for small ranges of Rayleigh number and
down the heating-rate a little. When a warm body is
for particular conguration of enclosures.
placed in cold surroundings, the air around the product
will become a little warmer but evaporates away from
the product (Torstveit, Aune, Nordtvedt, & Magnussen, 3.2. Model validation
1999).
For natural convection on a horizontal plate facing 3.2.1. Heat transfer and relative humidity
up Mills (1995) reported the following correlations: The heating and chilling curves (Fig. 5) show a big
For heating: dierence between headspace and food temperatures,
1=4 probably due to the insulation layer between food and
Nu 0:27  Pr Gr 105 < Ra < 1010 33
upper package. The two dierent product temperatures
For cooling: are however almost similar during the heating and
cooling process. For MAP products, the heat transfer
Nu 0:54  Pr Gr1=4 105 < Ra < 107 34
outside the product is low and the temperature die-
From Eqs. (29)(32) the aC values were higher than the rence between surface and product is small. This is,
literature values. This can be explained by the marked besides the small product thickness, also a contribution
changes in ow pattern with changes in Rayleigh number factor for a small Biot number (Torstveit et al., 1999).

Fig. 5. Experimental data for ambient, headspace and food tempera-

Fig. 3. Nusselt versus Grashof in pack with 9 cm height on headspace. tures.
 R.J. Simpson et al. / Journal of Food Engineering 61 (2004) 279286 285

Fig. 6. Experimental and simulated data for headspace relative

humidity. Fig. 8. Experimental and simulated data for headspace gas composi-
tion.
Fig. 6 shows that after a period of 4 h headspace
reaches an equilibrium between water activity and rela-
tive humidity (saturation). In addition, part of the The mass transfer process can be divided in two dif-
evaporated water condenses and the package surface ferent stages, rst a transient period and then a constant
gets wet with small droplets. period similar to controlled atmosphere. The changes in
The results indicate a good agreement among experi- the ambient temperature do not signicantly aect the
mental and simulated temperatures and relative hu- reached headspace equilibrium gas composition.
midity values. Non-signicant dierences were found From simulations and experimental data, it was ob-
between experimental and simulated values (Figs. 6 and served that an equilibrium in the headspace gas compo-
7). The values of the average medium error were 0.6, 0.3 sition was reached after a short time. This nding is valid
and 0.4 C for headspace and food temperatures (middle for low permeability packagings and when microorga-
and bottom) respectively and 3% for relative humidity. nism gas production and consumption are minimized.
Optimization MAP studies must consider an equi-
librium gas composition that maximizes foods shelf life.
3.2.2. Mass transfer The developed model is only valid for designing rigid
The results indicate a good agreement among experi- packages. However, by modifying Eq. (17), it can be
mental and simulated values obtained for the headspace used for exible packages as well. In exible packages,
gas compositions. Non-signicant dierences were found the free volume of the packages changes with time
between experimental and simulated values (Fig. 8). The whereas the total pressure remains constant.
values of average medium errors were low, 1.43 (%v/v)
for CO2 , 0.51 (%v/v) for O2 .

4. Conclusions

A mathematical model to estimate heat and mass

transfer in low fat non-respiring foods packed in a
modied atmosphere was developed and experimentally
validated. The statistical results showed the mathemati-
cal model validity and their numerical solution. Head-
space gas composition showed that if the lm has low
permeability and gas consumption as a product of
chemical and microbiological reactions is considered
negligible pseudo equilibrium after a short time is es-
tablished on headspace gas composition. Model pre-
dictions during heating and cooling phases indicate that
temperature modication of the packaged product can
be quite slow, reecting the relevance of proper chilling
Fig. 7. Experimental and simulated data for headspace and food during packaging, transport and storage processes of
(middle) temperature. MAP products.
286 R.J. Simpson et al. / Journal of Food Engineering 61 (2004) 279286

This model coupled with shelf-life studies for specic Koutsomanis, K. (2001). Predictive modeling of the shelf life of sh
products, can be used for packaging-alternatives evalu- under nonisothermal conditions. Applied and Environment Micro-
biology, 67(4), 18211829.
ating purposes (size, material and selected thickness) for Lakakul, R., Beaudry, R. M., & Hernandez, R. J. (1999). Modeling
dierent temperature and initial gas concentration sce- respiration of apple slices in modied-atmosphere packages.
narios of MAP products. Journal of Food Science, 64(1), 105110.
Marrero, T., & Mason, E. (1972). Gaseous diusion coecients.
Journal of Physical Chemistry, 1, 3118.
Mills, T. (1995). Convection fundamentals and correlations. In
Acknowledgement T. Mills (Ed.), Basic heat and mass transfer (pp. 208317). Chicago:
Irwin Publishers.
Murakami, E., & Okos, M. (1980). Measurement and prediction of
The research was supported by Universidad T
ecnica thermal properties of foods. Food properties and computer aided.
Federico Santa Mar
a under project # 270124. In Engineering of food processing systems (pp. 348). Dorecht:
Kluwer Academic Publishers.
Perry, J. H., & Green, D. W. (1992). Manual del Ingeniero Qu
mico
Tercera Edici
on. M
exico DF: McGraw-Hill.
References Perry, R. H., & Green, D. W. (1997). Perrys chemical engineers
handbook (7th ed.). New York: McGraw-Hill.
Almonacid, S. F., & Torres, J. A. (1993). Mathematical models to Robertson, G. L. (1992). Permeability of thermoplastic polymers. In
evaluate temperature abuse eects during distribution of refriger- G. L. Robertson (Ed.), Food packaging (pp. 73110). New York:
ated solids foods. Journal of Food Engineering, 20, 223245. Marcel Dekker.
Bird, B., Steward, W., & Lightfoot, E. (1964). Fen
omenos de Simpson, R., Almonacid, S., & Acevedo, C. (2001). Development of a
Transporte. Barcelona: Editorial Revert
e S.A. mathematical model for MAP system applied to nonrespiring
Bolz, R., & Tuve, G. (1976). Handbook of tables for applied engineering foods. Journal of Food Science, 66(4), 561567.
science (2nd ed.). Cleveland: CRC Press. Talasila, P., Chau, K., & Brecht, J. (1995). Modied atmosphere
Blakistone, B. A. (1999). Introduction. In B. A. Blakistone (Ed.), packaging under varying surrounding temperature. American
Principles and applications of modied atmosphere packaging of Society of Agricultural Engineers, 38(3), 869876.
foods (pp. 113). Maryland: Aspen Publishers. Torstveit, A., Aune, E., Nordtvedt, T., & Magnussen, O. (1999).
Bonacina, C., & Comini, G. (1973). On the solution of nonlinear heat Inuence of modied atmosphere packaging on meat products. In
conduction equations by numerical methods. International Journal 20th international congress of refrigeration (pp. 17). Sydney: IIR/
of Heat and Mass Transfer, 16, 581589. IIF.
Brown, G. G. (1955). Operaciones b
asicas de la ingenier
a qu
mica. Treybal, R. E. (1980). Operaciones de transferencia de masa (Segunda
Barcelona: Editorial Mar
n S.A. Edici
on). M
exico DF: McGraw-Hill.
Crank, J. (1975). The mathematics of diusion (2nd ed.). Oxford: Vargaftik, N. B. (1975). Tables on the thermophysical properties of
Clarendon Press. liquids and gases (2nd ed.). Washington: Hemisphere.
Emond, J., Chau, K., Bretch, J., & Ngadi, M. (1998). Mathematical Wassiljewa, A. (1904). Physikalische Zeitschrift, 5, 737.
modeling of gas concentration proles in modied atmosphere bulk Welty, J. R., Wicks, C. E., & Wilson, R. E. (1984). Fundamentals of
packages. American Society of Agricultural Engineers, 41(4), 1075 momentum, heat and mass transfer (3rd ed.). New York: John Wiley
1082. and Sons, pp. 296300.
Empack, Co. (1999). Personal communication. Santiago, Chile. Zuritz, C. A., & Sastry, S. K. (1986). Eect of packaging materials
Geankoplis, C. (1982). Procesos de transporte y operaciones unitarias. on temperature uctuations in frozen foods: mathematical model
Mexico DF: Editorial Continental. and experimental studies. Journal of Food Science, 51, 1050
Herming, F., & Zipperer, L. (1936). Gas Wasserfach., 79, 49. 1056.