Elsevier Editorial System(tm) for Engineering Failure Analysis

Manuscript Draft

Manuscript Number: EFA-D-10-00228R1

Title: Uni-axial and multi-axial creep behaviour of P91-type steel under constant load

Article Type: Original Research Paper

Keywords: short-term creep; uni-axial creep test; multi-axial creep test; constant load; Monkman-
Grant relationship

Corresponding Author: dr Aleš Nagode, Ph.D

Corresponding Author's Institution: Faculty of Natural Sciences and Engineering

First Author: Aleš Nagode, Ph.D

Order of Authors: Aleš Nagode, Ph.D; Ladislav Kosec, professor dr.; Boris Ule, PhD

Abstract: The uni-axial and multi-axial creep behaviour of P91-type steel under constant load was
studied. The conventional, uni-axial, constant-load creep tests were performed at initial stresses
ranging from 120 MPa to 240 MPa, and at temperatures from 625 to 675°C, while the multi-axial,
small-punch, constant-load creep tests were performed at loads ranging from 350 N to 550 N, and at
temperatures from 650 to 690°C. Both types of test can be considered as short-term creep tests
because the maximum time-to-rupture was less than 500 hours. Since it is well known that the creep
behaviour of P91-type steel cannot be satisfactorily described by a simple, Arrhenius-type, power-law
constitutive model, an improved stress-dependent, energy-barrier model was used for a description of
the uni-axial as well as the multi-axial creep behaviour of the P91-type steel. It was found that the
obtained values of the apparent activation energies Qc during the uni-axial and multi-axial creep tests
were very close, and in both types of test they were considerably higher than the activation energy for
the lattice diffusion in iron.
Cover Letter

Dear Sir/Madame,

I have a pleasure to send you the manuscript entitled ˝ Uni-axial and multi-axial creep
behaviour of P91-type steel under constant load˝ for publication in your estimated journal.
The manuscript has been prepared in accordance with the notified instructions. The authors
of the manuscript Aleš Nagode, Ladislav Kosec are from the Faculty of Natural Sciences and
Engineering, Aškerčeva cesta 12, SI-1000, Slovenia. Co-author Boris Ule has passed away. He
was working at the Institute of Metals and Technology, Ljubljana, Lepi pot 11, SI-1000
Ljubljana, Slovenia. The corresponding author is Ales Nagode.

Looking forward to hear from you soon,

Sincerely yours

Dr. Aleš Nagode

Faculty of Natural Sciences and Engineering
Ačkerčeva cesta 12
SI-1000 Ljubljana
Slovenia

e-mail: ales.nagode@omm.ntf.uni-lj.si
phone: 00386 1 2000 433
Detailed Response to Reviewers

Response to Reviewer’s and Editor´s comments

Ms. Ref. No.: EFA-D-10-00228

Title: Uni-axial and multi-axial creep behaviour of P91-type steel under constant load
Engineering Failure Analysis

Dear Editor/Reviewer,

I have the pleasure to send you the revised version of my paper in the light of the comments
as follows:

1. As regards the comparison between the results obtained with improved model and
generally accepted results for creep of P91 steel from the literature a new text on page 4 in
chapter 3. Result in discussion, 3.1. Conventional uni-axial creep tests was added:

The apparent activation energies Qc calculated from the exponential part of Eq. (5) (Qc =
decrease from 580 kJ mol-1 to 548 kJ mol-1 as the stress
increases from 120 MPa to 240 MPa.

Due to the same reason a new text was also added on page 6 in chapter 3. Result in
discussion, 3.2. Small-punch multi-axial creep test:

The apparent activation energies Qc for the small-punch creep test calculated from the
exponential part of Eq. (10) (Qc = decrease from 543 kJ mol-
1
to 533 kJ mol-1 as the load increases from 350 N MPa to 550 N. These values of the
apparent activation energies as well as the values obtained during the conventional uni-
axial creep tests are in good agreement with the literature data [18-20].

In chapter 4. Conclusions on page 6 the comparison between results from improved model
and results from literature has been made. Thus, a new text has been included:

This value is somewhat smaller than the values from the literature [18-20], where the stress
exponent defined as strongly depends on the temperature and may, thus, increase
from 8 to approximately 16 as the temperature decreases from 650 °C to 550 °C.

and

These values of the apparent activation energies are in good agreement with the literature
data [18-20]; however, they are significantly higher than the activation energy for self-
diffusion in -iron. They represent the activation energy of the slowest, dominant rate
controlling process of the supposed multiple creep mechanism.

The text in chapter 4. Conclusions on page 6 has been also slightly rearranged due to
additional text in this chapter.

From the same reason the following three references were added on page 8 in chapter 6.
References:
 [18] Sklenička, V. Kucharová K, Dlouhy A, Krejči J. Materials for Advanced power
engineering. Dordrecht: Kluwer;1994.
[19] Čadek J, Šustek V, Pahutová M. An analysis of a set of creep data for a
9Cr-1Mo- 0.2V (P91 type) steel. Mater. Sci. Eng. 1997;225A:22-28.
[20] Spigarelli S, Cerri E, Bianchi P, Evangelista E. Interpretation of creep behaviour
of a 9Cr-Mo-Nb-V-N (T91) steel using threshold stress concept. Mater. Sci. Tech.
1999;15:1433-1440.

2. As regards English grammar, the manuscript has been proof read by a native speaker.

3. Figure 2 was redrawn since there was no information about the testing temperatures.
Thus, a legend was included in the Figure 2 and extra lines were removed.

Because Figure 2 shows the experimental obtained relationship between minimum creep
rate and time-to-rupture at different temperatures some new comments (in italic bold
letters) are included on page 3 in chapter 3. Result in discussion, 3.1. Conventional uni-
axial creep tests:

Hence, the relationship between the experimentally obtained minimum creep rate ( and
the time-to-rupture (tr) at three different testing temperatures can be plotted for uni-axial
creep tests in a bi-logarithmic diagram, as shown in Fig. 2.

In the Figure 2 caption an information about testing temperature was added as well:

Fig. 2. Minimum creep rate dependence of the time-to-rupture for uni-axial, constant-load
creep tests performed at 625 °C, 650 °C and 675 °C.

4. As regards the References format:

The References on page 7 and 8 was rewritten in accordance with the author´s guide.

Hoping that the above mentioned changes in the manuscript satisfy all the submitted
comments, I look forward to hearing from you soon.

Yours sincerely,

Aleš Nagode
*Manuscript
Click here to view linked References

Uni-axial and multi-axial creep behaviour of P91-type steel under

constant load

Aleš Nagode1, Ladislav Kosec1, Boris Ule2†

1
Faculty of Natural Sciences and Engineering, Aškerčeva cesta 12, 1000 Ljubljana, Slovenia
2
Institute of Metals and Technology, Lepi pot 11, 1000 Ljubljana, Slovenia
ales.nagode@omm.ntf.uni-lj.si, ladislav.kosec@omm.ntf.uni-lj.si

The uni-axial and multi-axial creep behaviour of P91-type steel under constant load was
studied. The conventional, uni-axial, constant-load creep tests were performed at initial
stresses ranging from 120 MPa to 240 MPa, and at temperatures from 625 to 675°C, while the
multi-axial, small-punch, constant-load creep tests were performed at loads ranging from
350 N to 550 N, and at temperatures from 650 to 690°C. Both types of test can be considered
as short-term creep tests because the maximum time-to-rupture was less than 500 hours. Since
it is well known that the creep behaviour of P91-type steel cannot be satisfactorily described
by a simple, Arrhenius-type, power-law constitutive model, an improved stress-dependent,
energy-barrier model was used for a description of the uni-axial as well as the multi-axial
creep behaviour of the P91-type steel. It was found that the obtained values of the apparent
activation energies Qc during the uni-axial and multi-axial creep tests were very close, and in
both types of test they were considerably higher than the activation energy for the lattice
diffusion in iron.

Keywords: short-term creep, uni-axial creep test, multi-axial creep test, constant load,
Monkman-Grant relationship

1. Introduction

Creep tests can be conducted either at constant load or at constant stress. However, from the
engineering point of view, constant-load creep tests are more important than constant-stress
creep tests, because it is the load and not the stress that is normally kept constant in
engineering applications. When studying the creep properties of materials, the most
conventional tests are uni-axial creep tests using cylindrical specimens 1-3. However, in
some cases it is not possible to have available the large amount of material required for a
conventional cylindrical specimen. Therefore, some new testing techniques are being
developed: tests that are able to extract the mechanical properties from a specimen with a
small volume 4,5. One of these tests is the so-called miniaturised disc-bend test, also known
as the small-punch test. This technique can also be used to measure the creep properties 6.
In the small-punch creep test a thin, circular disc is supported over a recessed hole and forced
under constant load to deform into the hole by means of a spherically shaped punch or a
ceramic ball. This gives rise to a rotationally symmetrical, multi-axially stressed state in the
specimen, in contrast to the conventional uni-axial creep test, where a multi-axially stressed
state occurs with the formation of the neck in the specimen. However, this happens during a
very late stage of the creep test, just before the rupture. Thus, it has almost no influence on the
time-to-rupture.

P91-type steel is widely used for the high-temperature pipework components in advanced
power plants. It exhibits a considerable high-temperature creep strength, a high corrosion-

1
cracking resistance, a low oxidation rate, and good weldability 7-11. However, due to the
characteristics of P91-type steel, the creep behaviour of this steel cannot be accurately
described using a simple, Arrhenius-type power law (Eq. (1)):

(1)

where is the minimum creep rate,  is the applied stress, Qc is the apparent activation
energy, G is the shear modulus (GMPa=97400-0.039T), n is the stress exponent, T is the
absolute temperature, R is the universal gas constant (R=8.314 J mol-1K-1), and A is a
constant.

However, the improved, power-law, stress-dependent, energy-barrier model (Eq. (2))

developed by Ule and Nagode 12,13 has been quite successful at describing the creep
behaviour of P91-type steel under constant stress since a good correlation between calculated
values and experimental data has been obtained.

(2)

where Q c max is the limiting value of the apparent activation energy when →0, and C1 and
C2 are constants.

In order to find a creep equation that would satisfactorily describe the uni-axial as well as the
multi-axial creep behaviour of P91-type steel under constant load a large number of short-
term creep tests, i.e., conventional uni-axial and small-punch creep tests that give rise to a
rotationally symmetrical, multi-axially stressed state in the specimen, were carried out. The
obtained values of the apparent activation energies and stress exponents will help make it
possible to understand the dominant rate-controlling mechanism of the supposed multiple
creep mechanism during the uni-axial as well as the multi-axial creep test.

2. Experimental procedure

P91-type steel with the chemical composition 0.104 C, 0.30 Si, 0.43 Mn, 9.01 Cr, 0.01 Ni,
0.95 Mo, 0.22 V, 0.06 Nb and balance Fe (all values in wt. %) was used in our research. In
order to homogenise the microstructure, the investigated steel was subjected to an additional
two-stage heat treatment (1050 °C for 1 h, vacuum cooling + 750 °C for 1 h, vacuum
cooling). The initial, tempered martensite ferritic, microstructure of the investigated steel is
shown in Fig. 1. The specimens for the uni-axial creep tests were cut along the axial direction
of the pipe. The diameter of the specimen was 6 mm and the gauge length was 50 mm. The
discs for the small-punch multi-axial creep test were cut from the same pipe in such a way
that the normal to the disc surface was parallel to the axis of the pipe. The diameter of the disc
test specimens was 8 mm and thickness was 0.50 mm. All the discs were ground and polished
to 1200 grit.

2
 Fig. 1. Tempered martensite ferritic microstructure of the investigated P91-type steel.

The conventional uni-axial creep tests were conducted at initial stresses ranging from 120
MPa to 240 MPa and at temperatures of 625, 650 and 675°C, while the small-punch, multi-
axial tests were performed at temperatures of 650, 675 and 690 °C and at loads from 350 to
550 N. For each of the test conditions at least three measurements were made in both cases.
The temperature of the cylindrical specimen was measured with a thermocouple at three
points in the direct vicinity of the specimen, while the temperature of the disc was measured
with a thermocouple, placed as close as possible to the specimen. The changes in the
specimen’s length (uni-axial creep tests) or the displacement of the punch, i.e., the central
deflection of the disc specimens (small-punch creep tests) were measured using an inductive
transducer with a high measuring accuracy (repeatability approximately 1 μm) and were
recorded continuously by a computer. Both types of tests were performed in a protective
atmosphere of purified argon. All the tests were run to the final fracture.

3. Results and discussion

3.1.Conventional uni-axial creep tests

For the description of the creep behaviour of the P91-type steel an improved power-law,
stress-dependent, energy-barrier model (Eq. (2)) was used. Monkman and Grant 14 (Eq. (3))
found a relationship between the minimum creep rate ( and the time-to-rupture (tr) based
on the following equation:

(3)

where M and p are the Monkman-Grant constants.

Hence, the relationship between the experimentally obtained minimum creep rate ( and
the time-to-rupture (tr) at three different testing temperatures can be plotted for uni-axial creep
tests in a bi-logarithmic diagram, as shown in Fig. 2.

Fig. 2. Minimum creep rate dependence of the time-to-rupture for uni-axial, constant-load
creep tests performed at 625 °C, 650 °C and 675 °C.

The statistical analysis showed a very high correlation coefficient (R2 =0.99). Hence, the
validity of the Monkman-Grant relation (Eq. (3)) was experimentally confirmed. However,
the values of the Monkman-Grant constants M and p are 0.289 and 0.851, respectively.
Because the constant p is not very close to 1, the time-to-rupture (tr) cannot be simply
expressed with . Thus, the minimum creep rate ( in Eq. (2) can be replaced by
the time-to-rupture (tr) and the equation for the time-to-rupture (tr):

(4)

3
can only be written if the constants in Eq. (4) are obtained by a statistical multi-regression
analysis in such a way that the value of the Monkman-Grant parameter p is 1. A statistical
multi-regression analysis confirmed the following values for the constants of C1´ = 2.16∙10-40,
C2´=2.50∙107, Qc max=610 kJ mol-1, and n=4.5. Finally, the following model-based equation
for the conventional, uni-axial, constant-load creep test is obtained:

(5)

In Fig. 3, the times-to-rupture (tr) calculated from the improved, power-law, stress-dependent,
activation-energy model (Eq. (5)) are plotted against the applied stress on a bi-logarithmic
scale, together with our short-term, creep-test data. It is clear that all the measured data can be
predicted very well with this model.

Fig. 3. Applied stress dependence of the time-to-rupture for the conventional, uni-axial creep
test. The points represent the experimental values, whereas the dashed lines represent the
values obtained from Eq. (5).

The apparent activation energies Qc calculated from the exponential part of Eq. (5) (Qc =
decrease from 580 kJ mol-1 to 548 kJ mol-1 as the stress
increases from 120 MPa to 240 MPa.

3.2.Small-punch multi-axial creep tests

For a description of the creep behaviour of the P91-type steel during the multi-axial, small-
punch creep test the improved power-law, stress-dependent, energy-barrier model (Eq. (2))
was also used. The relationship between the minimum deflection rate and the time-to-
rupture (tr) was plotted in the bi-logarithmic diagram shown in Fig. 4. In this way, the perfect
validity of the modified Monkman-Grant relation [14] in which the minimum deflection rate
is used, instead of the minimum creep rate ( was experimentally confirmed. The
modified relationship thus obtained is

(6)

where M=0.90 and p=0.94. A very high correlation coefficient was obtained (R2=0.96).

Fig. 4. Dependence of the time-to-rupture tr on the minimum deflection rate

4
Since the Monkman-Grant parameter p in this case is close to 1 (p=0.94) the minimum
deflection rate is almost inversely proportional to the time-to-rupture tr. Thus, the
minimum deflection rate , i.e., the minimum creep rate in an improved power-law,
stress-dependent, energy-barrier model (Eq. (2)), can be replaced with the time-to-rupture tr,
so that Eq. (2) transforms to:

(7)

where  is the stress, Qc max is the limiting value of the apparent activation energy during the
multi-axial, small-punch, creep test when →0, and D1 and D2 are constants.

The exact meaning of the stress in Eq. (7) is not yet known, but it could represent a scalar
value of those stress components that are crucial for the time-to-rupture of the disc, and thus it
is called the critical stress. Although the distribution of the stress in the disc is rotationally
symmetrical, the multi-axial stress state, as well as the deformation state, is heterogeneous. A
Numerical simulation 15 shows that the heterogeneous shear strain during the creep is
restricted only to the circular area in the strongly deflected disc (Fig 5). Therefore, the most
convenient stress in Eq. (7) seems to be the maximum effective stress for which the von
Misses effective stress was adopted as well as the assumption that not the spherical part, but
only the deviatoric part, of the stress tensor has an influence on the creep behaviour. The
value of this stress is not known, since only the load P is measured for a particular disc
thickness and geometry of the small-punch testing device.

Fig. 5. Shear-strain distribution in the thin disc during the small-punch creep test 15.

Since Milička and Dobeš [16,17] experimentally determined the linear dependence between
the stress σ during a conventional, uni-axial, tensile creep test and the load P during a small-
punch creep test, which results in an identical time-to rupture tr:

P = kσ , (8)

Eq. (5) can be rearranged to:

(9)

where P is the applied load when measuring the creep properties of the disc, and D1, D2 and
D3 are constants, is the apparent activation energy for a small-punch creep test, i.e.,
the limiting value of the activation energy when P→0, and nSP is the load exponent.

5
The values of the constants D1, D2 and D3 as well as the apparent activation energies
and the load exponents nSP were obtained by means of a statistical, multi-regression
analysis and are D1= 2.95∙10-16, D2=2.04∙10-2, D3=2.87, =560 kJ mol-1 and nSP=4.5.

The final equation for the multi-axial, small-punch creep is then

(10)

In Fig. 6, the times-to-rupture calculated from Eq. (10) are plotted against the applied loads on
a bi-logarithmical scale together with our short-term, multi-axial, small-punch creep data. It is
clear that there is a good correlation between the model and the experimental data.

Fig. 6. Correlation of the model (Eq. (10)) with the experimental data. The points represent
the experimental values, whereas the dashed lines represent the values obtained from Eq. (10).

The apparent activation energies Qc for the small-punch creep test calculated from the
exponential part of Eq. (10) (Qc = decrease from 543 kJ mol-1 to
533 kJ mol-1 as the load increases from 350 N MPa to 550 N. These values of the apparent
activation energies as well as the values obtained during the conventional uni-axial creep tests
are in good agreement with the literature data [18-20].

4. Conclusions

Two different types of creep test for the P91-type steel under constant load were carried out,
i.e., conventional uni-axial creep tests, where cylindrical specimens are used, and
miniaturised, multi-axial, small-punch creep tests, where thin discs are used.

The relationship between the minimum creep rate during the uni-axial, conventional,
constant-load creep test and the minimum deflection rate during the multi-axial, constant-load
creep test against the time-to-rupture was studied. The validity of the Monkman-Grant
relation was experimentally confirmed in both cases. However, the Monkman-Grant
parameter p during the multi-axial creep test was very close to 1 (p=0.94), while during the
uni-axial creep test p is somewhat less than 1 (p=0.851).

Since it is well known that the creep behaviour of the P91-type steel cannot be properly
described by a simple, Arrhenius power-law, the improved power-law, stress-dependent,
energy-barrier model was used to describe both types of tests. Using this model the measured
times-to-rupture can be very accurately predicted in the case of the conventional uni-axial
creep test as well as in the case of the multi-axial, small-punch creep test. The improved
model is theoretically well founded since it assumes the stress dependence of the energy
barrier.

6
The stress exponent n obtained during the uni-axial creep test and the load exponent nSP
obtained during the multi-axial, small-punch creep test were, in both cases, 4.5, and this is not
too high for climb-controlled creep. This value is somewhat smaller than the values from the
literature [18-20], where the stress exponent defined as strongly depends on the
temperature and may, thus, increase from 8 to approximately 16 as the temperature decreases
from 650 °C to 550 °C. In the case of the uni-axial, constant-load creep test the apparent
activation energies calculated from the time-to-rupture decreased from 580 kJ mol-1 to 548 kJ
mol-1 as the stress increased from 120 to 240 MPa, whereas in the case of the multi-axial,
small-punch creep test the activation energies decreased from 543 kJ mol-1 to 533 kJ mol-1 as
the load increased from 350 to 550 N. These values of the apparent activation energies are in
good agreement with the literature data [18-20]; however, they are significantly higher than
the activation energy for self-diffusion in -iron. They represent the activation energy of the
slowest, dominant rate controlling process of the supposed multiple creep mechanism. The
reason for the slightly lower activation energies in the case of the multi-axial, small-punch
creep tests probably lies in the increasing dislocation density because of the high plastic
deformation at the onset of the creep measurement. However, such an activation energy is
considerably higher than the activation energy for lattice diffusion in iron and can be
identified as the stress-dependent activation energy of the supposed multiple-creep
mechanisms.

Acknowledgments

The experimental work was conducted at The Institute of Metals and Technology, Lepi pot
12, SI-1000, Ljubljana.

5. Literature

[1] Aktaa J, Schinke B. Creep lifetime under constant load and constant stress: theory and
experiment. Journal of Testing and Evaluation 1996;24(4):212-228.
[2] André Meyers M, Kumar Chawla K. Mechanical behaviour of Materials. New Jersey:
Prentice Hall; 1999:543-544.
[3] Nagode A, Ule B, Jenko M, Kosec L. A constitutive creep equation for 9Cr-1Mo-0.2V
(P91-type) steel under constant load and constant stress. Steel Research int.
2007;78(8):638-642.
[4] Lucas GE.The Development of Small Specimen Mechanical Test Techniques. Journal
of Nuclear Materials 1983;117;327-339.
[5] Lucas GE. Review of small specimen test techniques for irradiation testing.
Metallurgical Transaction A 1990;21A:1105-1119.
[6] Ule B, Šturm R, Leskovšek V. Effects of test specimen geometry on creep behaviour
of 12Cr steel in miniaturised disc bend tests. Materials Science and Technology
2003;19:1771-1176.
[7] Blum W, Straub S. Subgrain growth during creep of a tempered martensitic 12% Cr-
steel. Steel Research 62;1991:72-74.
[8] Jones WB, Hills CR and Polonis D H. Microstructural evolution of modified 9Cr-1Mo
steel. Metall. Trans. 1991;22A:1049-1058.
[9] Hald J. Metallurgy and creep properties of new 9-12%Cr steels. Steel Res. 1996;67:
369-374.
[10] Kosec B, Kosec G. Temperature field analysis on active working surface of the die-
casting die. Metall (Berl. West) 2003;57(3):134-136.
7
[11] Vojvodič-Tuma J, Kosec G. Effect of microstructure on the accelerated creep of
20CrMoV12-1 and P 91 steels. Steel research international 2007;78(8):643-647.
[12] Ule B, Nagode A. A model based creep equation for 9Cr-1Mo-0.2V (p91 type) steel.
Materials Science and Technology 2007;23(11):1367-1374.
[13] Ule B, Nagode A. The improved power-law, stress-dependent, energy-barrier model
of 9Cr-1Mo-0.2V steel using short-term creep data. Scripta Materialia 2007;57:405-
405.
[14] Monkman FC, Grant NJ. An empirical relationship between rupture life and
minimum creep rate in creep-rupture test. Proc. ASTM 1956;56:593-605.
[15] Ule B. Šuštar T. Lovrenčič Saražin M, Rodič T. Copernicus-Small punch test method
Assessment for the determination of residual creep life of service exposed, ˝Small
Punch˝. Twelve month progress report period 1.2. 1996 to 31.1. 1997. Project
documentation classification code: 12MPR2_(IMT Ljubljana). Ljubljana: IMT;1997.
[16] Milička K, Dobeš F. Relation between uniaxial and equi-biaxial creep and creep
fracture behaviour in P91 steel. Materials Science Forum 2005;482:407-417.
[17] Milička K, Dobeš F. Small punch testing of P91 steel. International Journal of
Pressure Vessels and Piping 2006;83:625-634.
[18] Sklenička, V. Kucharová K, Dlouhy A, Krejči J. Materials for Advanced power
engineering. Dordrecht: Kluwer;1994.
[19] Čadek J, Šustek V, Pahutová M. An analysis of a set of creep data for a
9Cr-1Mo- 0.2V (P91 type) steel. Mater. Sci. Eng. 1997;225A:22-28.
[20] Spigarelli S, Cerri E, Bianchi P, Evangelista E. Interpretation of creep behaviour of a
9Cr-Mo-Nb-V-N (T91) steel using threshold stress concept. Mater. Sci. Tech.
1999;15:1433-1440.

8
FIGURE CAPTIONS

Fig. 1. Tempered martensite ferritic microstructure of the investigated P91-type steel.

Fig. 2. Minimum creep rate dependence of the time-to-rupture for uni-axial, constant-load
creep tests performed at 625 °C, 650 °C and 675 °C.

Fig. 3. Applied stress dependence of the time-to-rupture for the conventional, uni-axial creep
test. The points represent the experimental values, whereas the dashed lines represent the
values obtained from Eq. (5).

Fig. 4. Dependence of the time-to-rupture tr on the minimum deflection rate .

Fig. 5. Shear-strain distribution in the thin disc during the small-punch creep test [15].

Fig. 6. Correlation of the model (Eq. (10)) with the experimental data. The points represent
the experimental values, whereas the dashed lines represent the values obtained from Eq. (10).

9
FIGURES

Fig. 1

10
Fig. 2

11
Fig. 3

12
Fig. 4

13
Fig. 5

14
Fig. 6

15
Figure1
Click here to download high resolution image
Figure2
Click here to download high resolution image
Figure3
Click here to download high resolution image
Figure4
Click here to download high resolution image
Figure5
Click here to download high resolution image
Figure6
Click here to download high resolution image