You are on page 1of 7





Tamara Dapevi1*, Petar Doki1, Miroslav Hadnaev1,

Milica Poji1

UDC 664.34 : 66.018 1

Institute for Food Technology, Novi Sad, Serbia

Abstract: Measurement of yield stress is important in ensuring the quality of various food products, as
well as in determining the optimal processing and handling conditions. However, due to the complex
nature of food products (three-dimensional network microstructure) there is a difficulty in choosing the
most appropriate test by which the determined yield stress will be the true material physical property.
Five different test methodologies, their advantages and disadvantages were discussed. The ways of
overcoming some disadvantages were proposed. Controlled stress test was chosen as the most
appropriate one in quality control. For qualitative comparison all tests can be used if defined properly.

Key words: yield stress, food dressings, food spreads, vane, serrated plate


A large number of food products are in the ssings etc. has to be handpumped from the
form of soft solids or highly structured li- tube in order to break down its structure
quids, thus showing the properties of the and initiate its flow onto food. Since on the
yield stress materials.The most important food there is no more stress or shear, and
characteristic of these materials is that they the material is left to rest over time, its
can behave as solids under small applied structure rebuilds and it becomes solid
stresses, and as liquids at high stresses again.
(Moller et al., 2006). The minimum stress
required to make a material flow is called Because of the enormous range of appli-
the yield stress (or yield point), and it is a cations the yield stress materials have been
measure of the strength of the material studied intensively in the last two decades.
structure (Trrega et al., 2006). In practice, From 1985 when Barnes and Walters pro-
that means that the yield stress material vocative article "The yield stress myth"
such as mayonnaise, ketchup, salad dre- (Barnes & Walters, 1985) showed up till
*Corresponding author:
Tel: +381 21 485 3811: Fax: +381 21 450725

Dapcevic et al./Food Processing, Quality and Safety 35 (2008) 3, 143-149

1999 when Barnes wrote a review of the The aim of this paper was to compare
yield stress, the phrases yield stress or different tests for yield stress determination,
yield point have been cited nearly 2500 and discuss them by presenting advan-
times in the scientific and engineering lite- tages as well as shortcomings of the each
rature (Barnes, 1999). During that period one. The measurements were performed
the subject of yield stress, its proper de- using different commercial food dressings
finition and even its existence were widely and spreads.
discussed. According to Barnes and Wal-
ters, as a physical property describing a 2. MATERIALS AND METHODS
critical stress below which no flow takes
place, yield stress do not exist. The state- For the yield stress measurements the
ment in some rheological books that eve- following commercial food products were
rything flows if you wait long enough is very used: mayonnaise (Polimark, Serbia), ketc-
well known (Schramm, 2000). In the review hup (Polimark, Serbia), mustard with horse
from 1999 Barnes claimed that the concept radish (Centroproizvod, Serbia) and chi-
of a yield stress, when used and defined cken pt (Argeta, Bosnia and Herzego-
correctly, has proved very useful in a whole vina).
range of applications. According to Barnes The yield stress tests were carried out
the yield stress can be regarded as a cri- using Haake Mars rheometer (Thermo
tical stress for shear-thinning, where below Scientific, Germany) equipped with three
the critical stress the system has a viscosity different geometries: plate/plate (diameter =
plateau (Barnes, 1999). 60mm), serrated plate/plate (diameter =
35mm) and 4-blade vane rotor (diameter =
Although, the concept of yield stress as a 22mm). The samples were stored at room
material physical property has been criti- temperature for 3 hours prior to measu-
cized, it was generally considered and used rement. In order to avoid destruction of the
as important rheological criteria in food sample structure, they were carefully loa-
quality control, for product formulation, food ded, and smoothly brought to measuring
processing design and optimization (Ngu- position using the reduced lift speed of 0.6
yen & Boger, 1992). Yield stress is also a mm/min. All measurements were performed
key product characteristic which determines at 200.1 C using Haake Phoenix P1-
its texture, and thus the consumer sensory C35P temperature circulator.
perception during use and application. It is
very important, especially nowadays, when 3. RESULTS AND DISCUSSION
there are a lot of different producers of the
same product. Generally, while choosing For yield stress determination five different
the product consumers are expressing a tests were employed: indirect (traditional)
preference for the one with increasingly so- test, controlled stress test, controlled defor-
phisticated performance. Everyone would mation test, oscillation stress sweep test
rather choose thick and creamy sauce than and creep test. The mustard served as mo-
the one which is thin and runny (Moonay, del system.
3.1. Indirect (traditional) tests
However, the determination of a yield stress
is a real challenge for the rheologists as the
Flow curve method (mustard)
measured value is dependent on the

measurement apparatus, geometry, expe- 350

rimental protocol and the model used to

Shear stress, (Pa)


evaluate obtained data (Stokes & Telford, 250

2004). Also, the yield stress materials are 200

typically thixotropic, which means that are 150 yield stress

dependent on the shear history of the sam- 100 50s


ple and susceptible to ageing. According to 50 250s


Moller et al. (2006) the yield stress and thi-

0 20 40 60 80 100 120 140 160 180 200 220
xotropy can be understood and modelled as Shear rate, (1/s)

two effects of the same cause.
Fig. 1. Flow curves for mustard sample as a
function of shear time and shear rate range
Dapcevic et al./Food Processing, Quality and Safety 35 (2008) 3, 143-149

The most common way for the yield stress By plotting deformation () versus shear
determination is from a flow curve measu- stress () in a double logarithmic scale, two
red in a CR (controlled rate) mode (Fig. 1). regions with different slopes can be seen
(Fig. 2)
It involves the extrapolation of the shear
stress ()shear rate ( ) data to zero shear The first one (smaller slope) is region of
rate (start method) or calculation using the elastic deformation (solid like behaviour),
mathematical flow models such as Bin- while the second one (larger slop) is the
gham, Casson linear or non-linear form or region of viscous flow (liquid like beha-
Herschel-Bulkley. The models have the fo- viour). The yield stress can be detected as
llowing forms: a breakpoint in the slope of two power low
0 (1)
Shear stress ramps method is sensitive and
0 ( ) (2) provides reproducible results for both low
and high yield stress materials (Kutsc-
n 0n ( ) n (3) hmann, V98-156E). However, as all the
0 K n (4) methods it depends on the measuring geo-
metry used (Fig. 2). Moreover, the wrong
where: 0 is the yield stress or yield point, choice of geometry, results in incorrect yield
is the viscosity and K, n, are the parameters stress determinations (Stokes & Telford,
of flow equations. The selection of the ma- 2004). For example, measuring the yield
thematical equation, which will be used for point with smooth plate resulted in slip bet-
fitting the data, depends on the shape of ween sample and the plate. The resulted
the experimentally obtained flow curve. curve had two yield points. Slip can be avoi-
ded using the geometry with rough surface
By indirect method, the yield stress is (e.g. serrated plate) or the vane rotor. The
relatively easy and rapidly determined, but other advantages of the vane geometry, ex-
thus obtained yield stress value depends on cept the elimination of wall-slip effect, are
the range of shear rate applied and on the its simplicity of fabrication, ease of cleaning
time selected for the ramp (Fig. 1). How- and more than anything else, minimal de-
ever, if a well defined procedure (tempe- struction of the sample structure during
rature, shear rate, shear time, measuring loading (Barnes & Nguyen, 2001). More-
geometry) exists, the indirect method can over, by direct inserting of vane into the ac-
produce very comparable results. One such tual product container the measurements
example is OICCC standard used for tes- can be performed with no disturbance of
ting the liquid chocolate (OICCC, 1973). the material structure, what is especially im-
portant for the materials which exhibit thixo-
3.2. CS (controlled stress) tests or shear tropic behaviour. The food technology
stress ramps groups in the USA headed by Steffe and
For performing shear stress ramp method Daubert (2000) even put an effort to make
the CS rheometer mode has to be em- the vane method a national standard for de-
ployed. termination of the yield stress of food.
serrated plate
CS method (mustard)
The practical problems in using the vane
are that it requires bigger amounts of sam-

ple than the plate geometry and that vane-
in-jar measurement are not always possible


1 yield stress (sometimes the product are packed in tube

or in jar having the unsuitable shape). The-
refore the measurements done in this paper
were recorded using the serrated plate geo-
10 100 1000
Shear stress, (Pa)

The stress ramps measurements can also
Fig. 2. Stress ramps of mustard obtained using be plotted as viscosity versus shear stress
a parallel-plate, serrated plate and vane (Fig. 3). The yield stress is than determined
measuring geometry
Dapcevic et al./Food Processing, Quality and Safety 35 (2008) 3, 143-149

using interpolation between the zero-shear is applied and the magnitude of shear
viscosity and shear-thinning region. stress is recorded for some period of time.

CS method (mustard) plate yield stress CD method (mustard)
serrated plate 200
100000 vane
Apparent viscosity, (Pas)

10000 yield stress 150

Shear stress, (Pa)


100 -1
50 -1
1 0
10 100 1000 0 100 200
Shear stress, (Pa) Time, t (s)

Fig. 3. Viscosity vs. shear stress for mustard Fig. 5. Determination of the yield stress of
obtained using a parallel-plate, serrated plate mustard by controlled deformation test as a
and vane measuring geometry function of the applied shear rate

The effect of time has been reported as the The initial increase in stress represents the
only shortcoming of the test performed in elastic response of the material, while a de-
CS mode (Stokes & Telford, 2004). In this crease of the stress value, indicates the
paper the influence of this effect was exa- gradual structure breakdown (Trrega et
mined by performing the measurements in al., 2006). This transition from viscoelastic
which the time was ranging from 5 to 300s to viscous flow is manifested as a peak in
per applied stress (Fig. 4). shear stress response which corresponds
to the yield stress value. At low shear rate,
5s CS method (mustard)
the material is slow to respond and the
stress increases to a relatively constant va-
lue. Using the higher shear rates, the peak
in the stress occurs followed by stress de-

1 cay towards a relatively constant stress

values. Ideally, a shear rate should be used
where a minimum in peak stress is obser-

0.01 ved, but this requires several tests to est-

blish and the test no longer becomes a
10 100
Shear stress, (Pa)
quick determination of the yield stress
(Stokes & Telford, 2004). The other draw-
back in using this test is that it is not sen-
Fig. 4. Stress ramps of mustard as a function of sitive enough for low yield stress materials
measurement duration time (Kutschmann, V98-156E).
The measured deformation increased with
the time duration of the test, but this in- 3.4. Oscillation stress sweep tests
crease was not very noticeable. As it can
be seen from the Fig. 4 the short duration Oscillation stress sweep measurements
tests (5s) are not favorable, while there is which serve for determination of the linear
no sufficient time for structure to break viscoelastic range of the product can also
down. The test performed using other time be employed for the yield stress deter-
durations overlapped, indicating that the mination. It involves monitoring the loss
time factor has influence only if the time is angle () value with the shear stress at
too short or too long. constant frequency (Fig. 6).
3.3. CD (controlled deformation) tests
At small stress amplitudes the loss angle is
The simplest method for yield point deter- independent of shear stress, while at the
mination is the CD-test (Fig. 5), which can larger stresses, the increase in loss angle
be carried out with a controlled rate rheo- occurs, indicating viscous behaviour. The
meter. In this test constant, low shear rate stress value at the transition from the visco-
Dapcevic et al./Food Processing, Quality and Safety 35 (2008) 3, 143-149

elastic to the viscous region represents the conditions, but for different applied stre-
yield point (Petri, V96-127E). sses, a series of curves are obtained (Fig.
7). The curves are identical to the moment
1Hz Oscillation stress sweep (mustard)
when the yield stress is reached. At the
yield stress the increase of the creep fun-
ction occurs.
Loss angle, tan

150Pa Creep test (mustard)
5 180Pa

Creep function, J (1/Pa)

yield stress 4

0.1 3
1 10 100 1000
Shear stress, (Pa)

Fig. 6. Oscillation stress sweep of mustard as a
function of the applied frequency 0
0 100 200 300
Time, t (s)
From the Fig. 6 it is clearly seen that the

frequency value chosen do not significantly
affect the results. However, it is worth to Fig. 7. Series of creep tests performed with
mention that the results obtained using mustard
stress sweep test, which belongs to the
dynamic measurements, can be different The yield point determined with creep test
from the other tests which are of static is lower that the yield point obtained from
nature (Kutschmann, V98-156E). the other tests (Petri, V96-127E). Also, this
method is very time consuming and its
3.5. Creep tests reproducibility is very hard to achieve beca-
use the sample preparation can affect the
Performing the series of creep tests is results.
supposed to be the most accurate method
for yield stress determination. One creep Finally, the yield stress values of mustard
test involves monitoring of the response to as a model system, determined using the
an applied stress over a certain time period. above described five methods are summa-
Testing always a fresh sample under same rized in Table 1.

Table 1.
Yield stresses of mustard determined using different tests
Yield stress
Test Experimental conditions
value (Pa)
Start method max = 100s-1, t = 500s 220

Indirect test Bingham eq. max = 100s-1, t = 100s 205.5

Casson linear eq. max = 100s-1, t = 100s 131.6
Casson non-lin. eq. max = 100s-1, t = 100s, n = 1/2 158.4
Herschel-Bulkley eq. max = 100s-1, t = 100s, n = 1/2 148.9
CS test t = 150s 200.6
CD test 0.5s 1 201
Oscillation stress sweep test f = 1 Hz 157.6
Creep test / 180

The result obtained using Bingham flow used methods in yield stress determination
equation was similar to the results obtained CS and CD tests produced the same re-
using CS and CD test. The most commonly sults. The results obtained using extrapola-
Dapcevic et al./Food Processing, Quality and Safety 35 (2008) 3, 143-149

tion by Casson or Herschel-Bulkley equa- suming, it is not the appropriate one in qua-
tions and using oscillation stress sweep test lity control measurements.
were similar, although the methodology of CS test seemed to be the most appropriate
these tests is very different. Results of method in quality control measurement, be-
creep test were right in the middle of the cause it is quick, easy to perform and useful
results obtained using other tests. Since for wide range of yield stresses. With the
this test is considered the most accurate, usage of serrated, profiled or vane rotors it
the value obtained using this test can may- became reproducible.Therefore it was used
be be the right one. However, having in for determining yield points of different food
mind that creep test produces lower values products. The obtained values are listed in
of the yield stress and that is very time con- Table 2.

Table 2.
Yield stress values of different food products determined by CS test
Food product ketchup mayonnaise mustard chicken pt

Yield stress value (Pa) 15.3 59.6 200.6 207.8

Ketchup and mayonnaise had lower yield suming and every one creep test in the
stresses which indicate that these materials series required new, fresh sample, which
can be packed in tubes. However, while enlarged the amount of sample required for
ketchup can also be packed in plastic the test performing. As the CS test, CD test
bottles, the mayonnaise has to be packed was also accurate and easy to perform, but
in tubes that are easy to deform by hand- only if the appropriate shear rate was esta-
pumping. On contrary the consumers would blished. However, the operator needs to
not be able to empty container completely, have in mind that CD test can be used only
which does not represent the problem only for the materials having yield stress value
for the consumers, but for the recycling above 10Pa. Oscillation tests were very re-
companies, too. On the other hand, mus- producible, but the obtained values were
tard or pt, having the high value of yield different from static tests. The extrapolation
stress, have to be packed in plastic or metal of the flow curve, although dependent on a
tins. Otherwise, a big part of the product will wide range of conditions, could also be
stay in container, or the consumers will used if the conditions are well defined. Ho-
have much trouble in using it, which will re- wever, for simple qualitative comparison of
sult in rejecting of their further usage. different product or the same product ob-
tained using different conditions, all the me-
4. CONCLUSIONS thods could be used, but it is important to
be consistent when comparing values of the
In this paper different methods for yield yield stress.
stress determination and their advantages
and limitations were widely discussed. The ACKNOWLEDGEMENT
main shortcomings of all the methods were
the results dependence on the test con- This research is a part of the project funded
ditions and sample shear history. There- by Ministry of Science and Technological
fore, the operator has to be well informed Development, Republic of Serbia, project
about the material structure and its rheo- no. 20066.
logical properties. The usage of profiled
geometries or vane rotor helps in obtaining
more reliable results, and eliminates the
mistakes ensued as a consequence of 1. Barnes, H. A., & Walters, K. (1985). The
slippage. The CS test was chosen as the yield stress myth? Rheol. Acta., 24 (4), 323 -
most appropriate yield stress test for food 326.
quality control. It was quick, accurate and 2. Barnes, H. A. (1999). The yield stress: a re-
could be performed easily, with a little view or -everything flows? J. Non-
amount of sample. The creep test, although Newtonian Fluid Mech., 81, 133-178.
the most accurate one, was very time con-
Dapcevic et al./Food Processing, Quality and Safety 35 (2008) 3, 143-149

3. Barnes, H. A. & Nguyen, Q. D. (2001). Ro- 10. Petri, H. Determining the Yield Point of Food
tating vane rheometry a review. J. Non- Products, Haake-Report V96-127E.
Newtonian Fluid Mech., 98, 114.
11. Steffe, J. F. & Daubert, C. R. (2000). The
4. Kutschmann E. Yield Point Determination - a vane method: a standard procedure for yield
Critical Discussion of Different Methods. stress determination. In: Proceedings of the
Haake-Report V98-156E. Institute of Food Technologists Annual Mee-
5. Moller, P. C. F., Mewis, J., & Bonn, D. ting (paper 2). Dallas Texas.
(2006). Yield stress and thixotropy: on the
difficulty of measuring yield stresses in pra- 12. Stokes, J. R. & Telford, J. H. (2004). Mea-
ctice. R. Soc. Chem., 2, 274-283. suring the yield behavior of structures fluids,
J. Non-Newtonian Fluid Mech., 124, 137
6. Moonay, D. (2005). Cost-effective yield 146.
stress test for quality control. LabPlus inter-
national. 13. Trrega, A., Costell, E. & Rao, M.A. (2006).
7. Nguyen, Q. D. & Boger, D. V. (1992). Mea- Vane Yield Stress of Native and Cross-
suring the flow properties of yield stress linked Starch Dispersions in Skimmed Milk:
fluids. Annu. Rev. Fluid. Mech., 24, 4788. Effect of Starch Concentration and -
carrageenan Addition. Food Sci Tech Int.,
8. OICCC. (1973). Rev. Int. Choc., 216. 12(3), 253260.
9. Schramm, G. (2000). A Practical Approach
to Rheology and Rheometry. 2 Ed., Geb-
rueder HAAKE GmbH, Karlsruhe, Germany.