J. Dairy Sci.

100:3513–3525
https://doi.org/10.3168/jds.2016-12051
© American Dairy Science Association®, 2017.

Thermal conductivity as influenced by the temperature

and apparent viscosity of dairy products
B. J. Gonçalves, C. G. Pereira, A. M. T. Lago, C. S. Gonçalves, T. M. O. Giarola, L. R. Abreu,
and J. V. Resende1
Department of Food Science, Federal University of Lavras, 37200-000, Lavras, MG, Brazil

ABSTRACT INTRODUCTION

This study aimed to evaluate the rheological be- Technological advances have supported the develop-
havior and thermal conductivity of dairy products, ment of new dairy products while the dairy market has
composed of the same chemical components but with registered sustained and continuous growth (Nagpal et
different formulations, as a function of temperature. al., 2012; Masson et al., 2016). Food engineering within
Subsequently, thermal conductivity was related to the the dairy sector is an expanding field of study, which
apparent viscosity of yogurt, fermented dairy beverage, results in significant improvements in product quality
and fermented milk. Thermal conductivity measures and greater knowledge of ingredients and their influence
and rheological tests were performed at 5, 10, 15, 20, on chemical composition, structure, and rheological
and 25°C using linear probe heating and an oscillatory and sensory properties (Chandrapala and Zisu, 2016).
rheometer with concentric cylinder geometry, respec- Although technological processes such as fermentation
tively. The results were compared with those calculated are traditionally used, the dairy sector has developed
using the parallel, series, and Maxwell-Eucken models techniques to produce a diverse range of milk-based
as a function of temperature, and the discrepancies in products and dairy ingredients (Nagpal et al., 2012).
the results are discussed. Linear equations were fitted Yogurt, fermented dairy beverages, and fermented
to evaluate the influence of temperature on the thermal milk all have the same components (water, protein,
conductivity of the dairy products. The rheological fats, carbohydrate, and ash) but in different contents
behavior, specifically apparent viscosity versus shear in their intrinsic composition (Reddy and Datta, 1994;
rate, was influenced by temperature. Herschel-Bulkley, Minim et al., 2002; Munir et al., 2016). Yogurt is formed
power law, and Newton’s law models were used to fit during the slow lactic fermentation of milk lactose by
the experimental data. The Herschel-Bulkley model thermophilic lactic acid bacteria (Park and Haenlein,
best described the adjustments for yogurt, the power 2013; Shori and Baba, 2014). The resulting lactic acid
law model did so for fermented dairy beverages, and reacts with milk protein, promoting the characteristic
Newton’s law model did so for fermented milk and was texture of this product (Serafeimidou et al., 2013).
then used to determine the rheological parameters. Fer- Fermented dairy beverages are dairy products resulting
mented milk showed a Newtonian trend, whereas yo- from the mixing of milk and whey, vegetable fat, fer-
gurt and fermented dairy beverage were shear thinning. mented milk, lactic acid starter culture, and other dairy
Apparent viscosity was correlated with temperature products. The milk base represents at least 51% (vol/
by the Arrhenius equation. The formulation influenced vol) of the total ingredients of the product, fermented
the effective thermal conductivity. The relationship by a specific microorganism culture (Brazil, 2005).
between the 2 properties was established by fixing the From the technological viewpoint, the main difference
temperature and expressing conductivity as a function between yogurt and fermented dairy beverages is the
of apparent viscosity. Thermal conductivity increased addition of whey to the latter, which results in lower
with viscosity and decreased with increasing tempera- viscosity (Castro et al., 2013). Fermented milk means
ture. a lactic product where fermentation involves the action
Key words: yogurt, fermented milk, thermal of lactic acid bacteria and results in coagulation and a
conductivity, rheology reduction in pH. According to legislation, the cultures
or microorganisms used in the fermentation define the
name of the product to be yogurt or fermented milk
Received September 26, 2016.
Accepted January 8, 2017. (Brazil, 2007; Park and Haenlein, 2013). It is evident
1
Corresponding author: jvresende@dca.ufla.br that the technology for the manufacture of fermented

3513
3514 GONÇALVES ET AL.

Table 1. Chemical composition of the dairy products

Composition (%)

Product Moisture Protein Fat Carbohydrate Fiber Ash

Yogurt 60.48 5.10 4.23 30.00 0.00 0.18
Fermented dairy beverage 68.17 3.53 3.36 24.66 0.00 0.25
Fermented milk 86.82 1.51 0.00 11.48 0.00 0.08

milk is very similar to the production of yogurt; how- for processing, preservation, and production. Given
ever, low viscosity is a desirable and notable feature of the above, the aims of this study were to evaluate the
fermented milk (Tamime, 2006). influence of temperature on thermal conductivity and
In food processing, correct dimensioning of the equip- apparent viscosity of yogurt, fermented dairy bever-
ment used, better understanding of food structural ages, and fermented milk and to develop mathematical
organization, and process optimization all depend on models from experimental measures that correlate their
accurate data of the thermal properties and rheological thermal conductivity and apparent viscosity.
behaviors of the products as well as their behaviors
during the process as a function of temperature to en- MATERIALS AND METHODS
sure the high quality of the final product (Ahmed et al.,
2005; Moura et al., 2005; Mercali et al., 2011; Augusto Sample Acquisition and Chemical Composition
et al., 2012). Another important application of thermal
properties is accurate modeling and simulation of in- Yogurt, fermented dairy beverages, and fermented
dustrial processes, which are useful for the prediction milk were purchased in the local market in the munici-
of process behavior and critical for decision making and pality of Lavras, Brazil. The composition of yogurts, fer-
optimization without putting the real process at risk. mented dairy beverages, and fermented milks in ranged
However, these thermal properties are worked separate- from 2.00 to 3.11% protein, 0 to 2.94% fat, 11.11 to
ly using empirical models derived from experimental 15.90% carbohydrate, 0% fiber, 0.10 to 0.13% ash, and
results and require more computational effort (Munir 77.92 to 85.29% water, as shown in Table 1. According
et al., 2016). Thermal conductivity is considered the to the information from the yogurt manufacturers, the
most important property in thermal processing at high samples tested had modified starch plus gelatin added.
and low temperatures (Muramatsu et al., 2005; Carson, The fermented dairy beverages had modified starch
2006), and there is a correlation between the apparent plus pectin added. The chemical composition of the 3
viscosity and thermal conductivity (Pereira et al., 2014) commercial products were obtained from their labels.
because shear rate is related to thermal conductivity,
and these 2 variables vary with intrinsic composition Experimental Measurements
and temperature. For this reason, it is important to of Thermal Conductivity
characterize the rheological properties and study mod-
els that describe the rheological behavior as a function Thermal conductivity measurements were performed
of the temperature used during processing (Nindo et using a hot wire probe, also known as a transient-state
al., 2007). The viscosity of dairy products is important method. In this technique, the heat is transferred to
in regard to sensory perceptions of food products (So- the material through the probe. The probe is resistively
dini et al., 2004; Walstra et al., 2006). heated at a constant rate using a source that provides
The thermal conductivity of fluid dairy products a continuous electrical current, and then it is possible
increases linearly with increasing temperature and de- to determine the thermal conductivity of a food ex-
creases with increasing total solid contents (Reddy and perimentally (Dawson et al., 2006). The probe was
Datta, 1994; Minim et al., 2002; Munir et al., 2016). constructed from a hypodermic needle with an external
The flow behavior and apparent viscosity also vary diameter of 0.6 mm and a length of 70 mm, into which
with temperature (Penna et al., 2001; Munir et al., was placed a resistance heating wire (nickel-chromium;
2016); an increase in temperature causes a decrease in Omega Engineering Inc., Stamford, CT) with a length
the viscosity of the liquid phase, increasing the move- of 0.36 m and diameter of 0.08 mm, together with a
ment of particles in suspension (Pelegrine et al., 2000). small type-T thermocouple AWG 30 (Omega Engineer-
There is a great need for thermal conductivity values ing Inc.) located at the exact center of the probe. All
and apparent viscosity results of these dairy products components were insulated with epoxy resin. By apply-

Journal of Dairy Science Vol. 100 No. 5, 2017

 THERMAL CONDUCTIVITY AND RHEOLOGY OF DAIRY 3515

ing heat via the resistance of the hot wire probe using probe. From the calibration, we determined a correc-
a current source, a temperature increase was recorded tion factor f, as shown in Eq. [4]:
through a signal conditioning system (National Instru-
ments Mod SCXI, Budapest, Hungary) using LabView ktheoretical
f = . [4]
8.5 software (National Instruments, Newbury, Ireland) kexperimental
with 1.0 s intervals between each recording. A ther-
mostatic bath (model 521/3DENova Ética, Sao Paulo, The values of ktheoretical for sucrose solutions were ob-
Brazil) was used to stabilize sample temperatures at 5 tained from ASHRAE (2002) for 5, 10, 15, 20, and 25°C.
to 25°C. Theoretical Models for Predicting Thermal
Conductivity. In the unfrozen state, effective thermal
Mathematical Analysis of the Hot Wire Probe Method conductivity can be predicted from the intrinsic ther-
mal conductivity combined with a suitable heat trans-
The usual procedure for measuring thermal conduc- fer model and volumetric fraction of each component.
tivity by heating a linear sensor is based on applying a The intrinsic thermal conductivity of each component
constant heat flux from the heat source to the material combined with the respective volume fractions provides
initially in thermal equilibrium, thereby generating a the determination of the thermal conductivity of the
variation in temperature, T0 to T, at some point of the food as a whole, as shown in Eq. [5]:
material (Park et al., 1997). This temperature change
(Δ) can be expressed by Eq. [1]:
ke = f (k1 , k2 , k3 ,..., X1 , X 2 , X 3 ,...) , [5]

q   t 
∆T = T − T0 =  ln   , [1] where ke is the effective thermal conductivity; k1, k2,
4πk  t  k3, … are the values of the intrinsic thermal conduc-
 0 
tivities of the components; and X1, X2, X3, … are the
volumetric fractions (Miyawaki and Pongsawatmanit,
where T = temperature (°C), T0 = initial temperature
1994).
(°C), q = power per meter (W∙m−1), k = thermal con-
The volumetric fractions are determined by Eq. [6]:
ductivity (W∙m−1∙K−1), t = time (s), and t0 = initial
time (s).
(x i ρi )
The power per meter supplied (q) was determined by Xi = , [6]
Eq. [2]: ∑ (x i ρi )
P R ⋅ i2 where xi is the mass fraction of each component, ρi is
q= = , [2] the density of each pure component (protein, ash, fiber,
L L
carbohydrate, fat, water; kg∙m−3), and Σ Xi = 1.
The equations for the determination of the thermal
where P = source power (W), R = electrical resistance
conductivities and densities of the pure components are
(Ω·m−1), i = electrical current (A), and L = length of
obtained from ASHRAE (2002). The theoretical models
the electric resistance wire inside the probe (m).
used to predict the thermal conductivity of the dairy
Thermal conductivity was determined by Eq. [3],
products were parallel, series, and Maxwell-Eucken. In
with the slope (m) obtained from the linear relation
the parallel model, thermal conductivity is calculated
between the displacement (T − T0) and ln(t/t0) (Brock
by multiplying the sum of the thermal conductivity by
et al., 2008):
the volumetric fraction of each component according to
Eq. [7]. In the model in series, Eq. [8], the phases are
q
m= . [3] in series or in the direction perpendicular to heat flow,
4πk which causes high resistance to its passage through the
sample. The Maxwell-Eucken model, represented by
Probe Calibration. Before determining the thermal Eq. [9], is a combination of the series and parallel mod-
conductivity, the probe was calibrated in relation to els, so it has a resistance intermediate between them
temperature, and a linear equation was obtained for its (Miyawaki and Pongsawatmanit, 1994; Pereira et al.,
adjustment. In addition to the temperature calibration, 2013; Giarola et al., 2016):
the probe was calibrated using a 40% sucrose solution
(wt/vol) having a known thermal conductivity (k) in n

the temperature range relating to the operation of the kpa = ∑ X iv ki , [7]

i =1

Journal of Dairy Science Vol. 100 No. 5, 2017

3516 GONÇALVES ET AL.

n v
1 X σ = σ0 + K γ n , [12]
= ∑ i , [8]
kse i =1 ki
 [13]
σ = µγ,
k + 2kc − 2Xdv (kc − kd )
kme = kc  d  , [9] where σ = shear stress (Pa), σ0 = yield stress (Pa), γ =
 k + 2k + X v (k − k ) 
 d d c c d 
shear rate (s−1), K = consistency index (Pa·s), n = flow
behavior index, and µ = Newtonian viscosity (Pa·s).
where kpa, kse, and kme are the conductivities of the par- The experiments were done in 3 repetitions, with each
allel, serial, and Maxwell-Eucken models, respectively; i repetition analyzed in triplicate.
refers to the pure components (protein, ash, fiber, car- Effect of Temperature on Apparent Viscosity.
bohydrate, fat, water); n is the number of components; According to Steffe (1996), the rheological behavior of
kc is the conductivity of the continuous phase; and kd is fluid foods is complex and influenced by numerous fac-
the conductivity of the dispersed phase. tors; Eq. [14] allows the prediction of apparent viscosity
Errors Between Theoretical and Experimental based on shear rate and temperature. The equation is
Conductivity. The effective thermal conductivities of similar to the Arrhenius model, and it was used to as-
the dairy products were compared with the theoretical sess the effect of temperature on the apparent viscosity
values obtained by mathematical models. A percentage of the different products analyzed. Viscosity was deter-
error was calculated using Eq. [10], which allowed us mined at a fixed shear rate of 27.27 s−1:
to determine the model that best fitted the effective
thermal conductivities: E 
ln ηa = B +  a T −1 , [14]
 R 
(kexperimental − ktheoretical )
% Error = ×100. [10]
ktheoretical where ηa is the apparent viscosity (Pa·s); B is the param-
eter adjustment; Ea is the activation energy (J·mol−1);
R is the gas constant (8.314 J·g−1·mol−1·K−1); and T is
Rheological Behavior. The rheological behavior of
the absolute temperature (K).
the dairy products was evaluated at 5, 10, 15, 20, and
25°C using a Haake rheometer ReoStress 6000 (Thermo
Scientific, Karlsruhe, Germany), equipped with a Statistical Analysis
Haake A10 thermostatic bath (Thermo Scientific) and
The fit of the experimental data to the rheological
a Haake UTM Controller universal temperature control
models was performed using SAS software (version
system (Thermo Scientific), coupled to a set of concen-
9.1.2, 2008; SAS Institute Inc., Cary, NC). The influ-
tric cylindrical geometry sensors (CC25 DIN Ti and
ence of temperature on rheological parameters was as-
CCB25 DIN) with a gap of 5.300 mm for all samples.
sessed by ANOVA and linear regression using Origin
The cup had a diameter (D) of 27.201 mm, and the
8.0 (Origin Lab Inc., Northampton, MA). This software
rotor used had the following specifications: length L =
was also used to graph thermal conductivities as a func-
37.603 mm, diameter D = 25.079, cone angle β = 120°,
tion of apparent viscosities using exponential fitting.
and sample volume V = 16.1 mL.
R-squared (coefficient of determination) and P-values
Given the thixotropy, to eliminate the influence of
from ANOVA were used to evaluate the adjustment
time on the flow behavior of products, each sample was
quality.
subjected to a continuous ramp shear rate from 0 to
300 s−1 over 3 min for the up-sweep and 3 min for the
downward curve. After this procedure, a flow curve was RESULTS AND DISCUSSION
generated for the rheological characterization of each
Probe Calibration
sample, which involved applied shear rates from 0 to
300 s−1 for a period of 3 min for each of the tested The probe was calibrated in relation to the tempera-
temperatures. The experimental flow curve data (shear ture by correlation between the temperatures measured
stress vs. shear rate) were adjusted to the power law experimentally and those obtained by thermometer.
(Eq. [11]), Herschel-Bulkley (Eq. [12]), and Newton’s Temperature calibration was carried out using the
law (Eq. [13]) models: linear equation obtained (data not shown), which had
a coefficient of determination (R2) of 0.999. Table 2
σ = K γ n , [11] shows the correction factors and the experimental and

Journal of Dairy Science Vol. 100 No. 5, 2017

 THERMAL CONDUCTIVITY AND RHEOLOGY OF DAIRY 3517
Table 2. Experimental and theoretical values for thermal conductivity of 40% sucrose solutions

Thermal conductivity (W·m−1·K−1)

Temperature Slope
(°C) (m)1 Experimental3 Theoretical4 Factor f2
0 4.606 0.181 (±0.003) 0.443 2.454
20 4.166 0.199 (±0.002) 0.470 2.355
50 3.997 0.208 (±0.006) 0.502 2.413
80 3.826 0.217 (±0.008) 0.525 2.416
Average     2.410
1
m = slope of the straight line (by Eq. [3]).
2
Factor f = correction factor (between the theoretical and experimental thermal conductivity by Eq. [4]).
3
Values are mean ± SD.
4
Theoretical values were obtained from ASHRAE (2002).

theoretical results obtained at each temperature for the the thermal conductivities in the temperature range of
thermal conductivity of the sucrose solution. 278.15 to 298.15 K of yogurt, fermented dairy beverage,
An f factor of 2.410 was used for the correction of and fermented milk, respectively:
the experimental thermal conductivities of the dairy
products analyzed. Yogurt: k = 0.0017T – 0.0258; R2 = 0.9844, [15]

Thermal Conductivity Fermented dairy beverage: k = 0.0010T – 0.2297;

R2 = 0.9589, [16]
Figure 1a shows the experimental results for temper-
ature variation versus time due to the heating provided
Fermented milk: k = 0.0017T – 0.0255;
by the probe inserted into the center of the stabilized
sample of fermented milk at 40°C. Figure 1b shows the R2 = 0.9908. [17]
experimental results of temperature increase versus
the logarithm of time used to obtain the parameter m Thermal Conductivity Obtained Through
(slope) and thermal conductivities. the Theoretical Models
Figure 2 shows the variation of the thermal con-
ductivity of the 3 dairy products for the temperature In addition to the experimental measurements, the
range from 278.15 to 298.15 K. It can be seen that, thermal conductivities of dairy products were evalu-
for the same temperature range, yogurts have the ated using the series, parallel and Maxwell-Eucken
lowest thermal conductivity values. Fermented milks models present in the literature. In these models, be-
and fermented dairy beverages have higher values of cause foods are made of a mixture of carbohydrates,
thermal conductivity, which are related to their water proteins, ash, fat, water, and other components that
contents. Water is the food component that contributes have their intrinsic conductivities, they together give
most to the increase in conductivity. Various studies rise to the thermal conductivity of the food as a whole
(Reddy and Datta, 1994; Tavman and Tavman, 1999; (Miyawaki and Pongsawatmanit, 1994). Figure 3 shows
Minim et al., 2002; Pereira et al., 2013) have confirmed the comparison between the thermal conductivities for
the strong correlation between thermal conductivity yogurt obtained experimentally and by using the theo-
and water contents of dairy products. In these studies, retical models.
it was observed that the thermal conductivity of all In Figure 3, it can be seen that the series model best
dairy products tested increased linearly with increasing fit the experimental data compared with the other 2
temperature and water content. Tavman and Tavman models (parallel and Maxwell-Eucken). The same was
(1999) also reported that in all dairy products, except observed for the fermented milk and fermented dairy
butter, there is a linear reduction in thermal conductiv- beverages (data not shown). Comparisons between the
ity with an increase in the protein content. This trend experimental and predicted results and the errors as-
was also observed in this work as shown in Tables 1 sociated with each temperature are shown in Table 3.
and 3. The series model gave the best fit of the experimental
For all studied products, we detected an increase data for all dairy products analyzed, with a maximum
in thermal conductivity with increasing temperature. error of 3.45% for yogurt, 2.66% for fermented dairy
From Figure 2, it was possible to adjust the linear equa- beverages, and 5.94% for fermented milk. The error for
tions [15], [16], and [17], which describe the behavior of fitting the series model was about 1.72 and 2.23 times
Journal of Dairy Science Vol. 100 No. 5, 2017
3518 GONÇALVES ET AL.

higher in fermented milk than in yogurt and fermented They found small deviations (approximately 10%) be-
dairy beverages, respectively. This can be explained by tween measured and calculated values. They claimed
differences in product viscosities, which influence the that such accuracy is acceptable or useful for most food
heat transfer process between the probe and product engineering heat transfer calculations.
during measurement of the thermal conductivity. The
low viscosity of fermented milk leads to more intense
convective currents in the fluid near the heated probe Rheological Behavior
surface. These currents increase the heat transfer rate,
resulting in higher values of experimental thermal con- Among the models used to describe the rheological
ductivity and higher error values compared with the behavior of dairy products, the Herschel-Bulkley, the
predicted value. Miles et al. (1983) tested the robust- power law, and Newton’s law models provided the best
ness of model equations by comparing the measured statistical parameters for the adjustment of the experi-
and calculated thermal conductivities for 11 foods. mental data for yogurt, fermented dairy beverages, and

Figure 1. (a) Experimental results of temperature variation as a function of time during the insertion of the heating probe in the sample of
fermented milk stabilized at 40°C; (b) linearization of temperature increase data for fermented milk sample initially stabilized at 40°C.

Journal of Dairy Science Vol. 100 No. 5, 2017

 THERMAL CONDUCTIVITY AND RHEOLOGY OF DAIRY 3519

Figure 2. Experimental thermal conductivity for dairy products.

fermented milk, respectively. Karsheva et al. (2013), adequate, as shown by the values of R2 (coefficient of
working with rheological properties of yogurts during determination) and root mean square error (RMSE)
storage, found that the Herschel-Bulkley model gives presented in Table 4.
better accuracy. Regarding fermented dairy beverages, For the rheological parameters obtained for yogurt
Penna et al. (2001) described the power law model as using the Herschel-Bulkley model, we found that the
the most suitable to describe the rheological behavior correlation between yield stress and consistency index
of 5 commercial brands of fermented dairy beverage. versus temperature was significant (P < 0.05). The
As shown in Table 4, the rheological parameters correlation of values of K as a function of tempera-
determined according to each model used were (1) con- ture obtained by the power law model for fermented
sistency index (K), flow behavior index (n), and initial dairy beverages was also significant (P < 0.05), and
yield stress (τ0) from the Herschel-Bulkley model; (2) the different temperatures studied influenced these
K and n from the power law; and (3) Newtonian viscos- parameters. For fermented milk, the viscosity data (μ)
ity (μ) from Newton’s law. Regarding the adjustment determined by Newton’s law was significant (P < 0.05)
of the experimental points, the studied models were for all temperatures tested.

Table 3. Comparison between experimental and predicted results for thermal conductivity (k) of dairy products1

kexperimental kparallel Error kseries Error kMaxwell-Eucken Error

Temperature (°C) (W·m−1·K−1) (W·m−1·K−1) (%) (W·m−1·K−1) (%) (W·m−1·K−1) (%)
Yogurt
 5 0.435 (±0.002) 0.523 20.23 0.450 3.45 0.510 17.24
 10 0.441 (±0.009) 0.530 20.18 0.456 3.40 0.517 17.23
 15 0.453 (±0.005) 0.538 18.76 0.462 1.99 0.524 15.67
 20 0.460 (±0.002) 0.544 18.26 0.467 1.52 0.530 15.22
 25 0.467 (±0.003) 0.551 17.99 0.472 1.07 0.537 14.99
Fermented dairy beverage
 5 0.507 (±0.004) 0.554 9.27 0.512 0.99 0.547 7.89
 10 0.512 (±0.006) 0.561 9.57 0.520 1.56 0.555 8.40
 15 0.520 (±0.005) 0.569 9.42 0.527 1.35 0.562 8.08
 20 0.524 (±0.002) 0.576 9.92 0.534 1.91 0.569 8.59
 25 0.526 (±0.005) 0.583 10.84 0.540 2.66 0.576 9.51
Fermented milk
 5 0.500 (±0.003) 0.559 11.80 0.527 5.40 0.554 10.80
 10 0.505 (±0.006) 0.567 12.28 0.535 5.94 0.562 11.29
 15 0.515 (±0.009) 0.575 11.65 0.544 5.63 0.570 10.68
 20 0.524 (±0.005) 0.583 11.26 0.552 5.34 0.577 10.11
 25 0.533 (±0.005) 0.590 10.69 0.560 5.06 0.585 9.76
1
kexperimental = experimental thermal conductivity; kparallel = thermal conductivity obtained through the parallel model; kseries = thermal conductiv-
ity obtained through the series model; kMaxwell-Eucken = thermal conductivity obtained through the Maxwell-Eucken model. Values are mean ± SD.

Journal of Dairy Science Vol. 100 No. 5, 2017

3520 GONÇALVES ET AL.

Table 4. Rheological parameters1 for dairy products

Herschel-Bulkley (HB)
Temperature
Product   (°C) K n σ0 RMSEHB R2HB
Yogurt 5 1.236 (±0.363) 0.601 (±0.064) 4.781 (±0.151) 0.198 0.999
10 1.201 (±0.446) 0.619 (±0.086) 4.425 (±0.472) 0.134 0.999
15 1.086 (±0.632) 0.585 (±0.113) 4.025 (±0.497) 0.081 0.999
20 1.063 (±0.614) 0.586 (±0.127) 3.492 (±0.510) 0.133 0.999
25 0.806 (±0.057) 0.606 (±0.137) 3.009 (±0.277) 0.105 0.999
R2 0.828 0.670 0.992    
P>F 0.020 0.241 0.000    

Power Law (PL)

K n RMSEPL R2PL
Fermented dairy beverage 5 0.765 (±0.306) 0.605 (±0.029) 0.337 0.995
10 0.650 (±0.168) 0.601 (±0.040) 0.322 0.994
15 0.640 (±0.226) 0.587 (±0.074) 0.321 0.992
20 0.555 (±0.241) 0.584 (±0.095) 0.282 0.992
25 0.423 (±0.186) 0.595 (±0.107) 0.238 0.992
R2 0.921 0.246    
P>F 0.006 0.226    

Newton’s Law

µ RMSE R2
Fermented milk 5 0.007 (±0.002) 0.029 0.999
10 0.006 (±0.001) 0.018 0.999
15 0.005 (±0.001) 0.018 0.998
20 0.005 (±0.001) 0.017 0.998
25 0.004 (±0.001) 0.017 0.997
R2 0.973    
P>F 0.001    
1
K = consistency index (Pa·sn); n = flow behavior index; σ0 = yield stress (Pa); µ = viscosity (Pa·s−1); R2 = coefficient of determination; RMSE
= root mean squared error. Values for rheological parameters are mean ± SD.

The rheological behavior of dairy products and the ent viscosities (Pa·s) and applied shear rate (s−1). The
influence of the different temperatures can be seen in rheological behavior shown in Figure 4a and Figure 4b
Figure 4, which shows the relationship between appar- is typical of non-Newtonian fluids. The non-linearity

Figure 3. Experimental and theoretical (parallel, series, and Maxwell-Eucken models) thermal conductivities for yogurt.

Journal of Dairy Science Vol. 100 No. 5, 2017

 THERMAL CONDUCTIVITY AND RHEOLOGY OF DAIRY 3521

between apparent viscosity and shear rate is evident,

such that the slope of the curve decreases with increas-
ing shear rate. Such behavior, according to Sharma et
al. (2000), is characteristic of pseudoplastic fluids. The
results indicated that at all temperatures tested, the
samples had shear-thinning behaviors (n < 1) as ob-
served in Table 4. The n value indicates the degree of
shear-thinning behavior, and in this respect, the greater
the unit value, the further the fluid moves away from
Newtonian behavior (Conceição et al., 2012). Mathias
et al. (2013) also found shear-thinning behavior by do-
ing a comparative study of the rheology of commercial
yogurts. The value of n was between 0.08 and 0.69 for
the Herschel-Bulkley model, which exhibited the best
suitability for flow and viscosity curves.
In Figure 4c, it can be seen that, of the 3 analyzed
dairy products, fermented milk does not present a sharp
degree of pseudoplasticity, because the apparent viscos-
ity versus shear rate has an almost linear tendency. This
relationship is typical of Newtonian behavior. Viscosity
is one of the most important characteristics of the rhe-
ology of liquid food and is considered the internal resis-
tance of the liquid flow. For a fixed shear rate, apparent
viscosity decreases with increasing temperature for all
dairy products analyzed. In Figure 4, the slopes of the
curves can also be observed to decrease with increasing
shear rate and with increasing temperature, showing a
decrease in apparent viscosity with an increase in these
2 factors.
Ramaswamy and Basak (1991) assessed the rheo-
logical characteristics of 2 brands of yogurts, and the
Herschel-Bulkley model resulted in the best fit to the
experimental data. According to Ramaswamy and Ba-
sak (1991), the samples had shear-thinning and yield
stress characteristics that seem to be an inherent feature
of the yogurt. Yield stress, consistency index (Table
4), and apparent viscosity (Figure 4) decreased with
increasing temperature, whereas flow index showed no
tendency to vary with temperature. Penna et al. (2001)
determined the rheological characteristics of 5 brands of
fermented dairy beverages with measurements taken at
5 and 25°C. The samples showed shear-thinning behav-
ior and the power law model was the most convenient
to describe the behavior of fermented dairy beverages,
which was also observed in our experiments.

Effect of Temperature on Apparent Viscosity Figure 4. Relation between apparent viscosity and shear rate for
(a) yogurt, (b) fermented dairy beverage, and (c) fermented milk at
Figure 5 shows the fit by the Arrhenius model for the different temperatures.
apparent viscosities of yogurt, fermented dairy bever-
ages, and fermented milk as a function of temperature
at a shear rate of 27.27 s−1. The shear rate of 27.27 s−1 tures. At values above a shear rate of 27.27 s−1, major
was chosen because the apparent viscosity values at changes in shear rate yield slight variations in apparent
this point are remarkably different for various tempera- viscosity. In addition, a shear rate of 27.27 s−1 is below
Journal of Dairy Science Vol. 100 No. 5, 2017
3522 GONÇALVES ET AL.

Figure 5. Application of the Arrhenius model for apparent viscosity versus temperature (T) at a shear rate of 27.27 s–1.

the value (50 s−1) at which, according to Akhtar et al. Fermented dairy beverage: ln ηa =
(2006), oral stimuli associated with viscosity perception
−10.5590 + (2,525.6)T−1, R2 = 0.9612, [19]
for low-viscosity products are usually developed.
Equations [18], [19], and [20] relate the effect of tem-
perature on the apparent viscosity of yogurt, fermented Fermented milk: ln ηa = −13.4840 + (2,354)T−1,
dairy beverages, and fermented milk, respectively. The R2 = 0.9967. [20]
R2 values equal to or exceeding 0.8494, for all experi-
mental conditions, indicated good adjustment from the
where ηa is the apparent viscosity (Pa·s) and T is tem-
Arrhenius model:
perature (K).
The use of linear equations ([18] to [20]) with equa-
Yogurt: ln ηa = −7.2992 + (1,825)T−1,
tion 14 it was possible to determine the activation en-
R2 = 0.8494, [18] ergy values (Ea), and the parameter E in the range of

Figure 6. Thermal conductivity and apparent viscosity as a function of temperature; k = thermal conductivity, ηa = apparent viscosity.
Color version available online.

Journal of Dairy Science Vol. 100 No. 5, 2017

 THERMAL CONDUCTIVITY AND RHEOLOGY OF DAIRY 3523
Table 5. Activation energy and parameter B obtained by adjustment
with the Arrhenius model for apparent viscosity at a shear rate of
27.27 s–1

Activation energy
Product (J·mol−1)1 B1,2
Yogurt 15,256.190 (±319.765) −7.299 (±0.121)
Fermented dairy beverage 20,997.838 (±822.404) −10.559 (±0.303)
Fermented milk 19,571.156 (±731.854) −13.484 (±0.307)
1
Values are mean ± SD.
2
B = parameter adjustment.

temperature from 298.15 to 318.15 K. The results are

shown in Table 5.
We observed that yogurt, which has a lower moisture
content and thus higher soluble solids, had lower acti-
vation energy. In the temperature range of 278.15 to
298.15K, fermented dairy beverages, fermented milk,
and yogurt presented a variation in apparent viscosity
of 47.15, 43.28, and 37.89%, respectively. According to
Steffe (1996), higher values of Ea indicate a more rapid
change in viscosity with temperature. Observing the
values in Table 5, the highest value of Ea occurred in
fermented dairy beverage, following by fermented milk,
and then yogurt. Mathias et al. (2013), studying com-
mercial yogurts, found Ea between 8,995.6 and 27,781.7
J·mol−1. Activation energy depends on the shear rate
fixed for determination of the apparent viscosity at each
analysis temperature. Ramaswamy and Basak (1991)
assessed the rheological characteristics of 2 brands of
yogurts at a shear rate range of 0 to 500 s−1, and they
evaluated Ea at 100, 200, 300, and 400 s−1. The val-
ues at various shear rates during the second and third
up-curves and the 3 down-curves varied only slightly.
The range of calculated values was 17,572.8 to 21,756.8
J·mol−1 (24,267.2 to 37,446.8 J·mol−1 in the first up-
curve). We found an activation energy of 15,256.19
J·mol−1 for yogurt at the fixed shear rate of 27.27 s−1.

Relationship Between Thermal Conductivity

and Apparent Viscosity

To evaluate the correlation between thermal con-

ductivity and apparent viscosity, linear equations [15]
(yogurt), [16] (fermented dairy beverage), and [17] (fer-
mented milk) were combined with Eq. [18] (yogurt), [19]
(fermented dairy beverage), and [20] (fermented milk),
respectively. Figures 6 and 7 show the relationship be-
tween the 2 parameters as a function of temperature
for different dairy products, and Table 6 presents the
equations obtained.
Figure 7. Thermal conductivity as a function of apparent viscosity
Analysis of Figure 6 shows that for the same temper- using the equations from Table 6 for (a) yogurt, (b) fermented dairy
ature range, dairy products with higher moisture con- beverage, and (c) fermented milk. Color version available online.
tents exhibited greater values for thermal conductivity

Journal of Dairy Science Vol. 100 No. 5, 2017

3524 GONÇALVES ET AL.

Table 6. Equations of thermal conductivity as a function of the ACKNOWLEDGMENTS

apparent viscosity for yogurt, fermented dairy beverage, and fermented
milk
The authors thank the Fundação de Amparo à Pes-
Product   Equation1 quisa do Estado de Minas Gerais (FAPEMIG, Brazil),
 
the Conselho Nacional de Desenvolvimento Científico
3.1195
Yogurt k =   − 0.0258 e Tecnológico (CNPq, Brazil), and Coordenação de
 ln ηa + 7.2992 
Aperfeiçoamento de Pessoal de Nível Superior (CAPES,
 2.5256 
Fermented dairy beverage k =   + 0.2297 Brazil) for financial support for this research.
 ln ηa + 10.5590 
 4.0018 
Fermented milk k =   + 0.0255 REFERENCES
 ln ηa + 13.4840 
1
k = thermal conductivity (W·m−1·K−1); ηa = apparent viscosity (Pa·s). Ahmed, J., H. S. Ramaswamy, and N. Hiremath. 2005. The effect of
high pressure treatment on rheological characteristics and colour
of mango pulp. Int. J. Food Sci. Technol. 40:885–895.
Akhtar, M., B. S. Murray, and E. Dickinson. 2006. Perception of
and lower values for apparent viscosity. This increase in creaminess of model oil-in-water dairy emulsions: Influence of the
thermal conductivity and decrease in apparent viscosity shear-thinning nature of a viscosity-controlling hydrocolloid. Food
Hydrocoll. 20:839–847.
with decreasing total solids contents is related to the ASHRAE. 2002. Refrigeration Handbook. American Society of Heat-
amount of water present. According to studies by Lee ing, Refrigeration and Air Conditioning Engineers (ASHRAE),
and Irvine (1997), Ikhu-Omoregbe (2009), and Qi Lin Atlanta, GA.
Augusto, P. E. D., A. Ibarz, and M. Cristianini. 2012. Effect of high
et al. (2003), for fluids that are shear thinning, thermal pressure homogenization (HPH) on the rheological properties of
conductivity increases as shear rate increases and ap- tomato juice: time-dependent and steady-state shear. J. Food Eng.
parent viscosity decreases as shear rate increases. In 111:570–579.
Brazil. 2005. SISLEGIS: Sistema de Consulta à Legislação. Instrução
Figure 4a, 4b, and 4c, decreases in apparent viscosity as Normativa no. 16, de 23 de agosto de 2005. (Technical rules of
a function of increasing shear rate were observed for all identity and quality whey-based drinks.) Ministério da Agricul-
of the temperatures in the analysis; therefore, thermal tura, Pecuária e Abastecimento, Brasilia, Brazil.
Brazil. 2007. Instrução Normativa no. 46, de 23 de outubro de 2007.
conductivity increases with decreasing apparent viscos- (Technical regulation of identity and quality of fermented milk.)
ity, as can be observed in Figure 7a, 7b, and 7c for Ministério da Agricultura, Pecuária e Abastecimento, Brasilia,
yogurt, fermented dairy beverage, and fermented milk. Brazil.
Brock, J., M. R. Nogueira, C. Zakrzevski, F. C. Corazza, M. L. Coraz-
The magnitude of the apparent viscosity range applied za, and J. V. Oliveira. 2008. Determinação experimental da vis-
to the models of Table 6 is specific for each product, as cosidade e condutividade térmica de óleos vegetais. Ciênc. Tecnol.
shown in Figure 7. Aliment. 28:564–570.
Carson, J. K. 2006. Review of effective thermal conductivity models
for food. Int. J. Refrig. 29:958–967.
CONCLUSIONS Castro, W. F., A. G. Cruz, M. S. Bisinotto, L. M. R. Guerreiro, J.
A. F. Faria, H. M. A. Bolini, R. L. Cunha, and R. Deliza. 2013.
Development of probiotic dairy beverages: Rheological properties
The experimental data of thermal conductivity of 3 and application of mathematical models in sensory evaluation. J.
dairy products were similar, with thermal conductiv- Dairy Sci. 96:16–25.
ity increasing with increasing temperature. The ex- Chandrapala, J., and B. Zisu. 2016. Novel trends in engineered milk
products. J. Dairy Res. 83:268–280.
perimental data were best fitted to the series model for Conceição, M. C., T. N. Fernandes, M. E. T. Prado, and J. V. Re-
all dairy products analyzed, with maximum errors of sende. 2012. Effect of sucrose and pectin addition on physical,
5.94% for fermented milk, 3.45% for yogurt, and 2.66% chemical, thermal and rheological properties of frozen/thawed
pineapple pulps. Korea-Aust. Rheol. J. 24:229–239.
for fermented dairy beverage. For rheological behavior, Dawson, A., M. Rides, and J. Nottay. 2006. The effect of pressure on
apparent viscosity decreased with increasing shear rate, the thermal conductivity of polymer melts. Polym. Test. 25:268–
and its data were correlated with temperature by the 275.
Giarola, T. M. O., C. G. Pereira, and J. V. Resende. 2016. Thermal
Arrhenius equation. Fermented milk showed a Newto- conductivity of mango pulp (Mangifera indica L.) cultivar “Ubari”
nian trend, whereas behaviors of yogurt and fermented in freezing temperatures. Int. J. Food Prop. 19:385–394.
dairy beverages were best characterized as non-Newto- Ikhu-Omoregbe, D. I. O. 2009. Effect of solid particles on the thermal
conductivity of mango juice in a shear flow field. Int. J. Food Prop.
nian shear thinning (pseudoplastic) and best fit by the 12:885–895.
Herschel-Bulkley and power law models, respectively. Karsheva, M., V. Paskov, R. Tropcheva, R. Georgieva, and S. Danova.
Better understanding and optimization of processes 2013. Physicochemical parameters and rheological properties of
yoghurts during the storage. J. Chem. Technol. Metall. 48:483–488.
in the food industry are possible by establishing the Lee, D. L., and T. F. Irvine Jr. 1997. Shear rate dependent thermal
relationship between apparent viscosity and thermal conductivity measurements of non-Newtonian fluids. Exp. Therm.
conductivity as a function of temperature. Both will be Fluid Sci. 15:16–24.
Masson, M., A. Saint-Eve, J. Delarue, and D. Blumenthal. 2016. Iden-
useful and important for future engineering calculations tifying the ideal profile of French yogurts for different clusters of
involving heat and mass transfer. consumers. J. Dairy Sci. 99:3421–3433.

Journal of Dairy Science Vol. 100 No. 5, 2017

 THERMAL CONDUCTIVITY AND RHEOLOGY OF DAIRY 3525
Mathias, T. R. S., K. C. S. Andrade, C. L. S. Rosa, and B. A. Silva. Penna, A. L. B., K. Sivieri, and M. N. Oliveira. 2001. Relation between
2013. Avaliação do comportamento reológico de diferentes iogurtes quality and rheological properties of lactic beverages. J. Food Eng.
comerciais. Braz. J. Food. Technol. 16:12–20. 49:7–13.
Mercali, G. D., J. R. Sarkis, D. P. Jaeschke, I. C. Tessaro, and L. D. F. Pereira, C. G., J. V. Resende, and T. M. O. Giarola. 2014. Relation-
Marczak. 2011. Physical properties of acerola and blueberry pulps. ship between the thermal conductivity and rheological behavior of
J. Food Eng. 106:283–289. acerola pulp: Effect of concentration and temperature. LWT Food
Miles, C. A., G. van Beek, and C. H. Veerkamp. 1983. Calculation Sci. Technol. 58:446–453.
of thermophysical properties of foods. Pages 269–312 in Physical Pereira, C. G., J. V. Resende, T. M. O. Giarola, S. M. Pinto, and L.
Properties of Foods. R. Jowitt, F. Escher, B. Hallström, H. F. T. R. Abreu. 2013. Thermal conductivity of milk with different levels
Meffert, W. E. L. Spiess, and G. Vos, ed. Applied Science Publish- of moisture and fat: experimental measures and prediction models.
ers, New York, NY. Sem. Ciênc. Agrárias 34:1153–1166.
Minim, L. A., J. S. R. Coimbra, V. P. Minim, and J. Telis-Romero. Qi Lin, S. X., X. D. Chen, Z. D. Chen, and P. Bandoadhayay. 2003.
2002. Influence of temperature and water and fat contents on the Shear rate dependent thermal conductivity measurement of two
thermophysical properties of milk. J. Chem. Eng. Data 47:1488– fruit juice concentrates. J. Food Eng. 57:217–224.
1491. Ramaswamy, H. S., and S. Basak. 1991. Rheology of stirred yogurts. J.
Miyawaki, O., and R. Pongsawatmanit. 1994. Mathematical analysis of Texture Stud. 22:231–241.
the effective thermal conductivity of food materials in the frozen Reddy, Ch. S., and A. K. Datta. 1994. Thermophysical properties of
state. Biosci. Biotechnol. Biochem. 58:1222–1225. concentrated reconstituted milk during processing. J. Food Eng.
Moura, S. C. S. R., V. C. L. França, and A. M. C. B. Leal. 2005. Ther- 21:31–40.
mophysical properties of model solutions similar to juice—Part II. Serafeimidou, A., S. Zlatanos, G. Kritikos, and A. Tourianis. 2013.
Ciênc. Tecnol. Aliment. 25:454–459. Change of fatty acid profile, including conjugated linoleic acid
Munir, M. T., Y. Zhang, W. Yu, D. I. Wilson, and B. R. Young. 2016. (CLA) content, during refrigerated storage of yogurt made of cow
Virtual milk modelling and simulation of dairy processes. J. Dairy and sheep milk. J. Food Compos. Anal. 31:24–30.
Sci. 99:3380–3395. Sharma, S. K., S. J. Mulvaney, and S. S. H. Rizvi. 2000. Food Pro-
Muramatsu, Y., A. Tagawa, and T. Kasai. 2005. Thermal conductivity cessing Engineering: Theory and Laboratory Experiments. Wiley-
of several liquid foods. Food Sci. Technol. Res. 11:288–294. Interscience, Washington, DC.
Nagpal, R., P. V. Behare, M. Kumar, D. Mohania, M. Yadav, S. Jain, Shori, A. B., and A. S. Baba. 2014. Comparative antioxidant activity,
S. Menon, O. Parkash, F. Marotta, E. Minelli, C. J. K. Henry, and proteolysis and in vitro a-amylase and a-glucosidase inhibition of
H. Yadav. 2012. Milk, milk products and disease free health: an Allium sativum yogurts made from cow and camel milk. J. Saudi
updated overview. Crit. Rev. Food Sci. Nutr. 52:321–333. Chem. Soc. 18:456–463.
Nindo, C. I., J. Tang, J. R. Powers, and P. S. Takhar. 2007. Rheologi- Sodini, I., F. Remeuf, S. Haddad, and G. Corrieu. 2004. The relative
cal properties of blueberry puree for processing applications. Food effect of milk base, starter, and process on yogurt texture: A re-
Sci. Technol. (Campinas) 40:292–299. view. Crit. Rev. Food Sci. Nutr. 44:113–137.
Park, K. J., F. E. X. Murr, and M. Salvadego. 1997. Medição da Steffe, J. F. 1996. Rheological Methods in Food Process Engineering.
condutividade térmica de milho triturado pelo método da sonda. Freeman Press, East Lansing, MI.
Ciênc. Tecnol. Aliment. 17:242–247. Tamime, A. Y. 2006. Fermented Milks. Blackwell, Oxford, UK.
Park, Y. W., and G. F. W. Haenlein. 2013. Milk and Dairy Products Tavman, I. H., and S. Tavman. 1999. Measurement of thermal conduc-
in Human Nutrition: Production, Composition and Health. Wiley- tivity of dairy products. J. Food Eng. 41:109–114.
Blackwell, Oxford, UK. Walstra, P., J. T. M. Wouters, and T. J. Geurts. 2006. Dairy Science
Pelegrine, D. H., J. R. M. B. Vidal, and C. A. Gasparetto. 2000. Estu- and Technology. CRC Press, Boca Raton, FL.
do da viscosidade aparente das polpas de manga (Keitt) e abacaxi
(Pérola). Ciênc. Tecnol. Aliment. 20:128–131.

Journal of Dairy Science Vol. 100 No. 5, 2017