Polymer Testing 34 (2014) 113–119

Contents lists available at ScienceDirect

Polymer Testing
journal homepage: www.elsevier.com/locate/polytest

Test method

Assessment of the stepped isostress method in the prediction

of long term creep of thermoplastics
Mohamed Hadid a, *, Belhi Guerira b, Mebarak Bahri b, Abdelkrim Zouani c
a
Laboratoire de Génie Mécanique, Université de Biskra, Algeria
b
Laboratoire de Génie Electrique, Université de Biskra, Algeria
c
Ford Motor Company, Dearborn, MI, USA

a r t i c l e i n f o a b s t r a c t

Article history: To predict long term creep of thermoplastics, methods based on the time–temperature
Received 1 December 2013 superposition principle (TTSP) or on the time–stress superposition principle (TSSP) are
Accepted 9 January 2014 commonly used. These methods enable the construction of a creep master curve without a
lengthy experimental program. Recently, a new accelerated creep testing method, termed
Keywords: the stepped isostress method (SSM), was proposed and used to predict long term creep of
Creep
technical yarns. This paper focuses on the processing aspects of the SSM test data and its
Master curve
validity in the creep prediction of thick thermoplastic specimens. Excellent correlation is
Stepped isostress method
Thermoplastic
obtained between the master curves constructed by the classical TSSP method and those
Superposition principle constructed by the SSM method. The variation of the SSM testing parameters has no
significant effect on the obtained master curves, which constitutes proof of the SSM
robustness. Further, the trend of the SSM shift factors in terms of the creep stress obeys the
Eyring equation.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction function of material properties, exposure time, exposure

Thermoplastics materials are widely used in structural
components subjected to high load levels. Because of the
weak cohesion between the molecules of the thermoplastic
polymers, those materials allow sliding of the polymer
segments and exhibit significant viscoelastic behavior, even
at ambient temperatures and under moderate stress levels.
Furthermore, an increase in the operating temperature or
in the stress level quickly brings on nonlinear viscoelastic
behavior.
A creep test is conducted in order to characterize the
tendency of the material to deform permanently under
constant loading. The rate of the creep deformation is a
* Corresponding author. Biskra University, Mechanical Department,
07000, Algeria. Tel.: þ213 696 96 1006.
E-mail addresses: mhadid07@yahoo.fr, m.hadid@univ-biskra.dz (M.
Hadid).
temperature and applied loads. The risk over time rises
when the creep deformation becomes large enough to
exceed the design limit for an in-service part. In order to
predict the long-term material creep, the testing needed
may require extensive laboratory time. The application of
time–temperature or stress–temperature superposition
principals provides the capability to predict the long-term
material performance very much beyond the creep test
period.
1.1. Time–temperature superposition principle
Leaderman [1] was among the first to emphasize that a
portion of the creep curve obtained at temperature Tr is
identical to a creep curve obtained at temperature Ti, if all
the time values at Ti are multiplied by a constant factor. This
means that the creep curves plotted versus log time at Ti
temperatures are identical to a corresponding portion of a

0142-9418/$ – see front matter Ó 2014 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.polymertesting.2014.01.003
114 M. Hadid et al. / Polymer Testing 34 (2014) 113–119
creep curve at Tr but with a shift in the origin of the log time
axis. Mathematically, this idea is expressed by the following
equation:
εðTr ; tÞ ¼ εðT; t:aT Þ
where Tr and T are the test temperatures while aT is the
temperature shift factor.
Tobolsky et al. [2] were the pioneers in the use of the
time–temperature superposition principle in the con-
struction of the master curve representing the creep
relaxation modulus over a very long time period. Plazek [3]
analyzed the temperature dependency of creep in term of
free volume. He proposed a relationship of the relaxation
modulus ET at temperature T in term of the relaxation
modulus at the reference temperature Tr, as follows:
T rr
ET ðt:aT Þ ¼ ETr ðtÞ
Tr r
The quantity aT is a function of temperature only and,
from Eq. (1), is defined as equal to unity at Tr. The quantities
rr and r are the material densities at Tr and T respectively.
The ratio rr/r allows one to take into account the variation
of the density of the polymer due to temperature. However
the Tr/T ratio, allows an adjustment regarding the kinetic
aspect of the creep process. It is assumed that the material
does not change its structure with time, so that the time–
temperature superposition (TTS) principle is still valid.
1.2. Time–stress superposition principle
In an analogous manner to time–temperature super-
position, a time–stress superposition principle (TSSP)
approach is used to build creep master curves. The TSSP
assumes that an additional stress provides energy to the
tested material similar to the heat effect. Mathematically,
the TSSP can be expressed as follows:
εðs1 ; tÞ ¼ εðs2 ; t:as Þ
where s1 and s2 are creep stresses while as is the stress
shift factor.
The TSSP has been used [4–9] to predict the long term
creep of viscoelastic materials. Luo et al. [4] used the TTSP
to construct, for a PMMA commercial grade, a smooth
master creep compliance curve for a period of 290 days
from creep curves achieved over 4000 seconds and under
various stress levels. The authors verified that the time–
temperature shift factors are dependent on stresses at
which the shifts are applied, and that the time–stress shift
factors are dependent on temperatures at which the shifts
are applied. Jazouli et al. [5] evaluated the long term non-
linear creep of a polycarbonate at room temperature.
Compliance creep tests, obtained at nine different stress
levels of one hour duration, were determined and shifted
along the logarithmic time axis to get a master compliance
curve. Qaiser et al. [6] generated a creep curve for an
extended time period for an amorphous polycarbonate
from short-term creep curves obtained at different stress
levels. Additionally, they checked the effects of physical
aging on the master curve. Hadid et al. [7] investigated the
non-linear creep of fiber reinforced polyamide by the use of
an improved empirical creep model [8], on its basis they
constructed a smooth creep master curve [7]. Considering a
nanocomposite with polyamide matrix, Starkova et al. [9]
(1) were able to build a master curve for time periods more
than 60 times than the test time.
Furthermore, it should be mentioned that the stress–
time shift factor can be interpreted as the transition from
the laboratory testing time to the intrinsic time of the
material. In the construction of the master curve, Hadid
et al. [7] found an exponential relationship between the
stress shift factor and the applied creep stress:
Logðas Þ ¼ b:s (4)
However, Jazouli et al. [5] and Qaiser et al. [6] proposed
another expression for the stress shift factor using the free
volume approach. The expression assumes an exponential
(2)
relationship in terms of stress:
C1 ðs þ sr Þ
Logðas Þ ¼ (5)
ðC2 þ s  sr Þ
where C1 and C2 material constants
In summary, the works mentioned above [4–9] consti-
tute examples of the use of the classical time–stress su-
perposition principle for different kind of materials; neat
polymers, reinforced polymers and nanocomposites. This
confirms the huge interest and the need for the super-
position tool in the prediction of the viscoelastic material
behavior in the very long term.
1.3. The stepped isothermal method SIM
Stepped Isothermal Method (SIM) is a derivative of the
classical TTSP method. The SIM was first established by
Thornton et al. [10] to predict the long-term creep behavior
of geogrids used in soil reinforcement applications. Later,
Alwis [11,12] applied this technique to characterize the
creep of Kevlar 49 yarns.
(3)
In TTSP testing, a single specimen is subjected to a
constant load at a certain temperature and a plot of creep
strain vs. log (time) is produced. Similar experiments are
conducted on different specimens at different tempera-
tures. A reference temperature is selected, by applying the
principle of superposition, all individual curves are shifted
along the log time axis. Then, a creep master curve is pro-
duced at the reference temperature. In contrast to TTSP, SIM
involves loading a single specimen subjected to a constant
load. The test temperature is increased in a series of
controlled steps. At each temperature step, a creep curve is
obtained; those curves are processed and transformed in
several independent creep curves for each temperature
level. A creep master curve, at a reference temperature, is
then created in analogous manner to the TTSP method.
Several authors used the SIM technique to predict the
long term creep or the creep rupture time [13–20]. SIM
exploits a specificity of technical yarns, the high surface-to
thickness ratio, which allows very fast heating of the whole
specimen. When dealing with thick specimens concern
regarding the rapid heating and the non-uniform temper-
ature distribution in the specimen needs to be investigated.
 M. Hadid et al. / Polymer Testing 34 (2014) 113–119
1.4. The stepped isostress method SSM
Recently, a few studies [21,22] considered the stepped
isostress method (SSM), a derivative technique of the SIM
method. The approach in SSM testing is similar to that of
the SIM method but the acceleration of the creep is ob-
tained by increasing the stress level in steps rather than
stepping up the temperature. The increase in stress pro-
vides energy to the tested material similar to that provided
by heat in the SIM method. The SSM method is advanta-
geous when compared to the SIM method since there is no
need to use elevated temperatures; chemical properties of
the tested materials are not affected and the problem
concerning the slow and non-uniform heating of thick
polymeric samples is avoided. Indeed, when dealing with
thick specimens, the SIM method confronts the problem of
the poor thermal conductivity of the polymer which re-
quires a heating period at each step to reach a uniform
temperature through the specimen, resulting in a transition
phase with uncontrollable and variable creep conditions.
The few previous works dealing with the SSM technique
[21,22] have considered only technical fibers which have a
relatively lower viscoplastic behavior. In contrast, in the
present paper more pronounced viscoplastic behavior
related to the nature of the thermoplastic polymer (i.e.,
polyamide) is expected. Indeed, the stress/strain traction
curve shown in Fig. 1 shows a small elastic zone followed by
an extended nonlinear section.
The key to produce the creep master curve based on
SSM or SIM methods is the processing steps of the raw test
data [18–22]. The following sections will illustrate the steps
of data handling of the SSM method. Also, its validity for the
polyamide specimens with thick thickness will be checked.
The master curve obtained with the SSM method will be
compared to that obtained with the classical TSSP method.
2. Experimental testing
The material in the present work is a polyamide 6, trade
mark TECAMID 6, provided as extruded 7 mm thick sheets.
Specimens were cut from these sheets and stored in an
atmosphere with 20% of humidity for at least one month.
The experiments were conducted on an Instron 5969
testing machine equipped with a 10 KN cell force. The creep
Fig. 1. Traction curve.
115
strain was measured with a clip-on extensometer. The
control and data acquisition software of the machine allows
conducting complex loading sequences, which are well
suited for SSM tests.
The tests carried out were: i/ Tensile which is absolutely
necessary to determine the linear section of the tensile
curve, and to estimate the jaw effects and the true starting
point of the stress/strain curve. ii/ Classical creep tests of
three and six days duration at constant temperature. iii/
SSM creep tests with different stepwise increases of stress
(1.5 MPa, 2.5 MPa and 4 MPa) and with different creep
dwell time (2 hours and 5 hours). The testing temperature
was 300 K and the humidity 20%.
3. Test results
3.1. Traction test
Since the polyamide has some slack, the jaws effects
significantly impact the recorded specimen response,
especially in the case of creep deformation. That is why
tension tests are needed and performed to determine
accurately the true initial deformation which allows elim-
ination of the initial sliding from the recorded deformation.
Figure 1 illustrates the tensile stress/strain curve of the
polyamide. It allows determining the initial point, the
linear part of the curve and the stress levels which will be
considered for creep tests.
3.2. Classical creep tests
Two conventional creep tests of few days duration are
conducted on the same testing machine. The first test was
conducted for 6 days under a stress level of 2.5 MPa. The
second test was conducted for 3 days under a stress level of
5 MPa. These two relatively long tests are used to construct
the creep master curve according to the conventional TSSP
procedure. This curve is subsequently compared to that
obtained with the SSM method. To make up for the jump in
the creep deformation between two consecutive curves,
the last part of the creep curve obtained at 2.5 MPa was
extrapolated with a power law function in order to match it
to the beginning of the creep curve obtained at 5 MPa. This
helps to fill the gap in the creep strain between the two
stress levels during the construction of the master curve
and offers a guide for a best fitting of the master curve
based on the actual test data.
Fig. 2. SSM loading sequence.
116 M. Hadid et al. / Polymer Testing 34 (2014) 113–119

3.3. SSM creep tests

Figure 2 illustrates the specimen loading sequence for

the SSM testing procedure. This sequence is characterized
by a stepwise jump stress si and a dwell time Dt. Three
different test parameters with the corresponding configu-
ration are reported in Table 1. The first configuration has six
bearings with a stress increment of 2.5 MPa and a dwell
time of five hours at each load level. The second configu-
ration has five load levels with a smaller stress increment of
1.5 MPa but with the same dwell time. The last configura-
tion has shorter dwell time of two hours for each load level Fig. 3. Typical SSM creep curve.
and larger stress increment of 4 MPa. Figure 3 shows a
typical SSM creep curve. idea of the rescaling is based on the Boltzmann super-
position principle. It concerns all loading levels except the
reference stress, which usually corresponds to the first level.
3.4. SSM data processing and handling
Indeed, to separate the creep of each stress level, one
must take into consideration that the creep in each step is
The raw data from the SSM tests must be handled
the accumulation of the creep strain resulting from the
following a four-step process: (a) Initial vertical adjust-
applied stress in the current stress level, and also from the
ment, (b) vertical shifting, (c) rescaling and (d) horizontal
creep strain of the previous steps. In order to express the
shifting.
creep of a given bearing, solely resulting from its stress
level, one must find, the expected loading start point,
3.4.1. The initial vertical adjustment
assuming that the test had been conducted under the same
Due to the softness of the polyamide, it is inevitable for
stress but on a previously unstressed specimen.
there to be some sliding of the jaws when loading the spec-
The estimation of the rescaling value is achieved ac-
imens [19–22]. For this reason, this initial vertical adjustment
cording to the following process: Considering a given creep
step, illustrated in Fig. 4a, is absolutely necessary to allow
bearing, the primary creep is eliminated from the curve. A
matching of the true starting point for each curve. This en-
power function is used to extrapolate the creep curve. This
ables one to adjust all the creep curves with respect to a
function allows the estimation of the virtual creep starting
common reference before any other data handling.
point, under the present stress, but on a previously
unloaded sample and assuming operating the same creep
3.4.2. The vertical shifting
mechanism. Having this estimation of the rescaling time, a
Figure 3 shows that for each stress jump, an immediate
slight adjustment of the obtained rescaling value may be
strain jump is recorded due to the material elasticity. Since
necessary during the master curve construction.
the stress jump is instantaneous, it is assumed that there is
no creep strain recorded during the jump. Based on this
3.4.4. The horizontal shifting
assumption, a vertical shifting is carried in order to link the
As a result of the rescaling operation, a set of indepen-
beginning of the current bearing curve with the ending of
dent creep curves conducted under different stress levels is
the previous bearing curve. This shifting results in the
obtained. At this stage one proceeds exactly as in the
elimination of the elastic component in the recorded
classical TSSP procedure. Then, the master curve con-
deformation, as shown in Fig. 4b. Therefore, after this
struction requires certain horizontal shifts along the time
processing, the obtained curves correspond solely to the
axis of the rescaled creep curves [4–7]. The magnitude of
creep deformation.
the horizontal shift factor is a function of the stress level.
Figure 4d illustrates the end result master curve obtained
3.4.3. The rescaling
after this process.
The purpose of the rescaling operation is the transition
from a series of creep curves resulting from a step-wise
sequential loading of a unique sample, and during a 3.5. Validation of the SSM procedure for thick polyamide
known dwell time, to a set of independent creep curves for specimens
each stress level, as shown in Fig. 4c. These curves will be
used in the same manner as the conventional superposition The assessment and the validation of this procedure
principle for the construction of the creep master curve. The involve two aspects: the first is its comparison to the

Table 1
The different testing parameters for the SSM creep tests.

Stress (MPa) 2.5 5.0 7.5 10.0 12.5 15.0

Dwell time (s) 18,000 18,000 18,000 18,000 18,000 18,000
Stress (MPa) 2.5 4.0 5.5 7.0 8.5
Dwell time (s) 18,000 18,000 18,000 18,000 18,000
Stress (MPa) 2.5 6.0 10.0 14.0 18.0
Dwell time (s) 7200 7200 7200 7200 7200
 M. Hadid et al. / Polymer Testing 34 (2014) 113–119 117

Fig. 5. Master curves for different SSM testing parameters. Stress ref.
2.5 MPa.

respectively. It should be mentioned that the reference

stress for all master curves is 2.5 MPa.

4. Discussion

The objective of the present paper is to assess, for a

thermoplastic specimen having a large thickness, the
capability of the SSM method in the prediction of its long-
term creep. The creep master curves shown in Figs. 5 and 6
have been constructed stretching over dozens of years. An
excellent match is obtained between the curves con-
structed according to the classical TSSP method and those
constructed according to the SSM method. Additionally, a
unique master curve was obtained using the SSM method
with a variety of test parameters.
The use of the SSM method allows reduction of the
required number of test specimens and less handling of the
operator for the construction of the same master curve.
However, when dealing with the SSM method, one is faced
with a few challenges related to two aspects: i/ The
handling and the processing of the SSM raw test data. ii/
The effect of the stepwise jump stress and the dwell time
on the obtained creep data.

4.1. The data handling challenges of the SSM method

Some challenges should be mentioned regarding the

SSM creep data processing:
1/ The issue related to the last step in the construction of
a smooth master curve illustrated in Fig. 4d, which is the

Fig. 4. The required handling of the tests data of the SSM method.

classical TTSP method, and the second is its robustness to

different loading sequences and dwell times (Table 1).
Indeed, it is expected that the variation in the dwell time or
the change in the stress increment do not affect the creep
master curve. Figure 5 illustrates the master curve con-
structed from the three different SSM testing sequences,
reported in Table 1. Figure 6 illustrates the master curves Fig. 6. Master curve built according to SSM and TSSP methods. Stress ref. 2.5
constructed according to the SSM and to TSSP methods, MPa.
118 M. Hadid et al. / Polymer Testing 34 (2014) 113–119
result of the horizontal shifting and the connecting of the
rescaled creep curves shown in Fig. 4c. Some authors [20–
22] have utilized a third order polynomial function to link
smoothly the beginning of one curve with the end of the
previous curve. Moreover, to calculate the virtual starting
time and to connect curves of Fig. 4c (in the case of the SIM
method), Achereiner et al. [19] have tried to link two
consecutive creep curves by seeking, in an iterative way, to
have the same creep rate of the beginning of one curve and
the end of the previous curve.
2/ Furthermore, it is noted that the works dealing with
the SIM or the SSM methods do not mention the role of the
primary creep stage and its effect on the processing of the
SSM creep curves. In fact, the polyamide creep curves have
an important primary creep stage; that should not be
included in the construction of the master curve, since it
relates to the response of the material in the very short
term. For the SIM method, the problem of the primary
creep may be worse, because of the non-uniform temper-
ature inside the specimen during the stepwise increase of
temperature.
Currently, there are no available systematic treatments
of the SSM data for the construction of the master curve.
More works are required to achieve this goal. The con-
struction of master curves in Figs. 5 and 6 is based on a
graphical and empirical approach. This approach must
ensure the consistency in the evolution of the shape of each
master curve and between the different master curves
generated from different test conditions. It is clear that the
adopted empirical approach involves elements of subjec-
tivity, but it has merit prior to the emergence of a more
automated and systematic approach.
Regarding the graphical manipulations, it is found that
the most critical step is the rescaling operation shown in
Fig. 4c. It is worth mentioning that the construction of the
master curve reflects the transition of the measured creep
strain in the laboratory time scale to the intrinsic material
time scale, or to ’the material time’ scale. The two opera-
tions that allow this transition are the rescaling and the
horizontal shifting operations. The rescaling operation
corresponds, relative to the ‘material time’ scale, to an
initialization and a synchronization of the individual
curves, obtained at each stress level. However, the hori-
zontal shifting operation reflects the time/stress super-
position in the ‘material time’ scale. Then, an error in the
rescaling operation is equivalent to a loss in the ‘material
time’ scale, and will result in an aberration and probably
make it impossible to correctly achieve the next step, which
is the horizontal shifting.
4.2. The SSM testing parameters effect on the master curve
The last point to mention concerns the stress shift factor.
Figure 7 shows the curve of the shifting factor data from
different test conditions in terms of the acceleration creep
stress. It is found that the trend of its evolution follows a
power function.
The evolution of the shift factor that expresses the creep
rate with the stress can be represented by two models,
namely the modified model of Williams-Landel-Ferry given
in Eq. (5) or the Eyring model given in Eq. (6). The latter is
Fig. 7. The horizontal shift factor in terms of the stress.
more suitable for the present case, since the creep tem-
perature is below the glass transition temperature. In
addition, polyamide 6 is a semi-crystalline polymer, and it
is more appropriate to use the Eyring model [21–23]:
  
ε_ V
logðas Þ ¼ log ¼ ðs  sr Þ (6)
ε_ r 2:30kT
where V* is the activation volume, k is Boltzmann’s con-
stant, T is the absolute temperature, ε_ is the creep rate, ε_ r is
the creep rate at the reference stress.
The alignment of the entire shift factors, resulting from
different test conditions, on the curve with low scattering
constitutes proof of robustness of the SSM method. The
curve also proves that the modification of the SSM testing
parameters has no significant effect, in contrast to what has
been previously reported [10,19]. The novelty in the last
result is the trend of the shift factors data shown in Fig. 7
which follow a power function. Previous studies [21,22]
found that the variation of the logarithmic shift factor in
terms of the stress follows a linear function. Since the
activation volume varies linearly in terms of the creep
stress [21–23], a parabola trend of the shift factor shown in
Fig. 7 is more consistent with the Eyring model given by Eq.
(6). Indeed, every stepwise increase of the accelerating
creep stress affects the activation volume and creates a new
configuration of the molecular chains. The current chain
configuration is affected by all the previous creep bearings.
It is clear that the creep process of a previously strained
material occupies a larger activation volume than an un-
strained material; this is consistent with the fact that the
polymer molecules are continuously pulled during the
creep. This explains the variation of the activation volume,
and thus the nonlinear relationship of the logarithm of the
shift factor with the accelerating stress. Finally, from the
curve in Fig. 7, it can be concluded that the shift factors
from different testing parameters lies on the same curve,
this means that, for each stress level, the same creep
mechanism is occurring regardless of the loading sequence.
This result constitutes another validation of the SSM
technique.
5. Conclusions
The stepped isostress method (SSM) is a new technique
used to predict the creep behavior of technical yarns. The
 M. Hadid et al. / Polymer Testing 34 (2014) 113–119
present work demonstrates that SSM can be used to predict
the long-term creep of a thermoplastic specimen having a
large thickness.
The considered material is a commercial polyamide 6
supplied as an extruded thick sheet. Creep tests were per-
formed according to the classical TSSP method and the SSM
method. Different SSM testing parameters were considered
by varying the dwell times and the stepwise jump creep
stress. A unique master curve was obtained for different
SSM testing parameters. The obtained master curves
correlate well with that obtained with the classical TSSP
method. The handling and processing of the SSM raw data
has revealed that the rescaling step is the most decisive
compared to the other steps. The construction of the master
curve shows that the Eyring equation depicts well the
relationship between the shift factor and the creep stress.
The fact remains that the SSM method uses less sam-
ples. However, the different processing steps of the SSM
test data are currently achieved graphically and empirically.
More research is required to develop an automated and
standard approach for the handling and the processing of
the SSM raw data.
References
[1] H. Leaderman, Elastic and Creep Properties of Filamentous Materials
and Other High Polymers, The Textile Foundation Inc., Washington
D. C, 1943.
[2] A.V. Tobolsky, R.D. Andrews, Systems manifesting superposed
elastic and viscous behavior, Journal of Chemical Physics 13 (1)
(1945) 3.
[3] D.J. Plazek, Temperature dependence of the viscoelastic behavior of
polystyrene, Journal of Physical Chemistry 69 (10) (1965) 3480.
[4] W. Luo, C. Wang, R. Zhao, Application of time-temperature-stress
superposition principle to nonlinear creep of poly(methyl methac-
rylate), Key Engineering Materials 340 (2007) 1091.
[5] S. Jazouli, W. Luo, F. Bremand, T. Vu-Khanh, Application of time-
stress equivalence to nonlinear creep of polycarbonate, Polymer
Testing 24 (4) (2005) 463.
[6] A.A. Qaiser, J. Price, Estimation of long-term creep behavior of pol-
ycarbonate by stress-time superposition and effects of physical
aging, Mechanics of Time-dependent Materials 15 (2011) 41.
[7] M. Hadid, S. Rechak, A. Tati, Long-term bending creep behavior
prediction of injection molded composite using stress–time corre-
spondence principle, Materials Science and Engineering A 385 (1–2)
(2004) 54.
119
[8] M. Hadid, S. Rechak, A. Zouani, Empirical nonlinear viscoelastic
model for injection molded thermoplastic composite, Polymer
Composites 23 (5) (2002) 771.
[9] O. Starkova, J. Yang, Z. Zhang, Application of time-stress super-
position to nonlinear creep of polyamide 66 filled with nano-
particles of various sizes, Composites Science and Technology 67
(13) (2007) 2691.
[10] J.S. Thornton, S.R. Allen, R.W. Thomas, D. Sandri, The Stepped
Isothermal Method for TTS and its Application to Creep Data on
Polyester Yarn, Sixth International Conference on Geosynthetics,
Atlanta, USA, 1998.
[11] K.G.N.C. Alwis, Accelerated Testing for Long-term Stress-rupture
Behavior of Aramid Fibers, PhD thesis, University of Cambridge,
2003.
[12] K.G.N.C. Alwis, C.J. Burgoyne, Accelerated testing to predict the
stress-rupture behavior of Aramid fibers, in: T.A.N. Kiang Hwee
(Ed.), Fiber Reinforced Plastics for Reinforced Concrete Structures
(FRPRCS-6), 2003, p. 111. Singapore.
[13] J.H. Greenwood, W. Voskamp, Predicting the Long-term Strength of
a Geogrid Using the Stepped Isothermal Method. Proceedings of
the Second European Geosynthetics Conference, Bologna, Italy,
2000, p. 329.
[14] S.E. Lothspeich, J.S. Thornton, Comparison of Different Long Term
Reduction Factors for Geosynthetic Reinforcing Materials. Pro-
ceedings of the Second European Geosynthetics Conference,
Bologna, Italy, 2000, p. 341.
[15] W. Kongkitkul, F. Tatsuoka, A theoretical framework to analyse
the behavior of polymer geosynthetic reinforcement in
temperature-accelerated creep tests, Geosynthetics International 14
(1) (2007) 23.
[16] C.J.F.P. Jones, D. Clarke, The residual strength of geosynthetic rein-
forcement subjected to accelerated creep testing and simulated
seismic events, Geotextiles and Geomembranes 25 (3) (2007) 155.
[17] K.G.N.C. Alwis, C.J. Burgoyne, Accelerated creep testing for aramid
fibres using the stepped isothermal method, Journal of Materials
Science 43 (2008) 4789.
[18] S.S. Yeo, Y.G. Hsuan, Evaluation of creep behavior of high density
polyethylene and polyethylene-terephthalate geogrids, Geotextiles
and Geomembranes 28 (5) (2010) 409.
[19] F. Achereiner, K. Engelsing, M. Bastian, P. Heidemeyer, Accelerated
creep testing of polymers using the stepped isothermal method,
Polymer Testing 32 (3) (2013) 447.
[20] I.P. Giannopoulos, C.J. Burgoyne, Stepped Isostress Method for
Aramid Fibers. 9th International Conference on Fiber Reinforced
Polymers for Reinforced Concrete Structures (FRPRCS-9), Sydney,
Australia, 2009.
[21] I.P. Giannopoulos, C.J. Burgoyne, Prediction of the long-term
behaviour of high modulus fibres using the stepped isostress
method (SSM), Journal of Materials Science 46 (2011) 7660.
[22] I.P. Giannopoulos, C.J. Burgoyne, Accelerated and real-time creep
and creep-rupture results for aramid fibers, Journal of Applied
Polymer Science 125 (5) (2012) 3856.
[23] I.M. Ward, J. Sweeney, An Introduction to the Mechanical Properties
of Solid Polymers, Ed. Wiley, 2004.