Materials Science and Engineering, A145 ( 1991 ) L27-L29
L27
Letter
Creep behaviour of service exposed
Cr-Mo steels
V. M. Radhakrishnan*
Metallurgy Department, Indian Institute of Technology,
Madras-600036 (India)
(Received April 10, 1991)
Abstract
Service exposed Cr-Mo steels exhibit different creep
behaviour from fresh materials, mainly because of
structural changes. This note is a discussion on the
recent paper (P. K. Liaw, G. V. Rao and M. G. Burke,
Mater. Sci. Eng., A131 (1991) 187) on creep fracture
behaviour of 2JCr-lMo welds from a fossil fuel power
plant 31 years old. It points out the necessity of testing
the exposed materials at low stress levels to obtain a
meaningful assessment of their remaining life.
1. Introduction
Liaw et al. [1], in a recent paper, made a
thorough analysis of the creep fracture behaviour
of 2¼Cr-lMo welds from a fossil fuel power plant
31 years old. Power plants designed and commissioned in the late 1950s have had a lifetime of
around 30 years. Many components may have
longer periods of service, but other components,
containing critical regions such as welds, might be
at the end of their useful life. The detailed investigations by Liaw et al. are valuable additions to the
existing data of Cr-Mo steels. One or two important factors that should be borne in mind while
comparing the data of service exposed material
and those of fresh materials, are discussed in this
note, based on similar work carried out on 1Cr½Mo steel, service exposed for a period of 29.7
years at a temperature of 530 °C [2].
*Present address: Fatigue and Fracture Group, Structural
Division, NASA-Lewis Research Center, 21000 Brookpark
Road, Cleveland, OH 44135, U.S.A.
0921-5093/91/$3.50
2. Analysis
Creep data of both fresh and service exposed
materials at specimen temperatures of 530°C,
550 °C and 600 °C were analysed using constitutive equations [3, 4] of the type
e = tilt 1/3 + fl2t+ f13 t3
(1)
where fli is a function of stress and temperature, e
and t are creep strain and time. The minimum
creep rate groin was computed and the relation
between groinand the applied stress o is shown in
Fig. 1 for both fresh and service exposed materials (the full lines are for fresh material and
those with data points are for exposed material).
It can be seen that, in general, the minimum creep
rate of service exposed material is much higher
than that of the fresh material, particularly at
higher stress levels. At lower stress levels, the difference is not very large.
Creep properties of ferritic steels are greatly
affected by the size and distribution of the carbides. The fine carbide dispersion in the matrix,
which initially strengthens the material, will
coarsen on exposure to high temperatures for a
long duration. Metastable carbides of the type
M3C, VC and M2C coarsen in the range
500-700°C, obeying the t 1/3 kinetics. These
agglomerated carbides will reduce the resistance
to creep deformation. As a result, the creep properties of fresh material will be superior to those of
service exposed material. The creep rate is
governed by the effective stress o r given in the
form
or = ( o -
o~,,)
(2)
where o~0 is the internal stress. The steady value
of the internal stress o,0 is dependent on the
structure and the applied stress. In their analysis on the impact of structural instability on
the extrapolation of short-term creep test results,
Steen and Witte [5] showed that the internal stress
is a function of the applied stress. They also
indicated two regions: a high stress region, in
which the Orowan looping around the incoherent
carbide precipitates will be the dominant mechanism, and a low stress region where the
© Elsevier Sequoia/Printed in The Netherlands
L28
dislocations bypass the carbides by climb and the
internal stress is much less dependent on the
precipitation characteristics. Thus, if the creep
experiments of the exposed material are carried
out at higher stress levels, the creep rate will be
affected by (]rio, because of the change in structure
and high stress levels, and will be much higher
than that of the fresh material. This is what is reflected in the gmin--Orelation shown in Fig. 1. An
enhanced creep rate at higher stress levels for service exposed material compared with regenerated
material can be seen in the analysis of Sklenicka
et al. [6].
Figure 2(a) shows the relation between Emin
and the rupture time tr which is given by
( £min)1/1.3tr = constant
(3)
The full line is for the fresh material on which the
data points of service exposed material are
c -8~
-- qOI
30
50
I%TRESS IMPel 500
Fig. l. Relation between minimum creep rate and applied
stress; full lines are for fresh material.
plotted. This type of relation has been found to
be valid where large creep strain is experienced in
the third stage [3, 7].
Figure 2(b) shows the stress-rupture time
relation on the log-log plot for both fresh and
exposed materials. The full lines are for fresh
material. It can be seen that at low stress levels the
rupture time of exposed material approaches that
of fresh material. At high stresses the rupture life
of exposed material is very much reduced. This is
reflected in the Emin--Orelation with a high value
of Eminat higher stress levels compared with that
of flesh material. In their comparison of the Larson-Miller parameter (LMP) for data of the preexposed materials and the ASTM scatter band,
Liaw et al. [1] found that data points pertaining to
high stress levels fell outside the scatter band and
the LMP values were less than those of fresh
materials. Such different behaviour at high and
low stress levels could be understood because of
the different dominant mechanisms that control
the deformation.
Thus it is important to take into account the
influence of the internal stress ai0 in enhancing
the creep rate at higher stress levels. At lower
stress levels the difference appears to be minimal.
Hence to compare the creep fracture behaviour
and compute the remaining lifetime of the component, it is better to test the service exposed
material at lower stress levels.
In a recent paper [8], it was shown that the rupture time tr in creep crack growth conditions can
be related to the steady state energy rate line integral Cs* if the crack nucleation takes place when
a steady state condition is established in the entire
material. The relation is given by
Cs*tr=R
_/.
~-6
L
q _t2~la)
[
I
I
I
i
I
I
f
I
i 0[6
where R is a constant. The data presented by
Liaw et al. [1] were analysed and the relation
between C* and tr is shown in Fig. 3, C* corresponding to the initial advancement of the crack.
The relation can be given by
( c*)l/l4t r = constant
I00
---.
53 C
lb.,,
~o'
%.,
,d
,o~
,o~
~s
\1.
,e
RUPTURE TIME (h)
Fig. 2. Variation of minimum creep rate and stress with rupture time; full lines are for fresh material.
(4)
(5)
The dotted line, corresponding to a slope equal
to unity, is also shown in the figure. Since a CT
type of specimen was used by Liaw et al. in their
study, crack initiation may take place in the nonsteady state condition. Further, at high stress
levels (high values of C*) the deformation rate
will be more in the service exposed material
because of structural changes; hence C* will be
L29
_ , 0 3_
~
lower stress levels to determine the remaining
lifetime.
(3) The role of internal stress and its dependence on the structure and applied stress level
should be considered in the creep analysis of service exposed materials.
\\
-
_
~o2 _
Acknowledgment
Io°
i
i
lo 1
lo2
tb
RUPTURE TIME (hi
Fig. 3. Relation between C* at crack initiation and rupture
time.
higher for the exposed material than for fresh
materials. This is reflected in the value of C*
which is higher than that given by eqn. (4) with the
slope of the line equal to unity.
3. Conclusions
From the study made on creep deformation
and fracture of service exposed material the following can be concluded.
(1) Since the functional dependence of the
creep rate of service exposed material on the
applied stress is different from that of fresh materials, values measured for fresh materials cannot be used for extrapolation and evaluation of
the remaining lifetime.
(2) Since the results for the exposed material
at lower stress levels are nearer to those for fresh
materials, it would be better to carry out tests at
The author is very thankful to Professor Dr. H.
Nickel for making available the base creep data
and to the Alexander von Humboldt foundation
for financial gupport.
References
1 E K. Liaw, G. V. Rao and M. G. Burke, Mater. Sci. Eng.,
A131 (1991) 187.
2 Creep Data Bank (Institut fiir Reaktorwerkstoffe, KFA,
Jiilich, ER.G.).
3 V. M. Radhakrishnan, unpublished work, Institut fiir
Reaktorwerkstoffe, KFA, Jfilich, 1991.
4 V. M. Radhakrishnan, Analysis of creep curve by constitutive equations, Mater. Sci. Technol., in the press.
5 M. Steen and M. de Witte, Impact of structural instability
on the extrapolation of short term creep test results. In
B. Wilshire and R. W. Evans (eds.), Proc. 3rd Int. Conf. on
Creep and Fracture of Engineering Materials and Structures, The Institute of Metals, London, 1987, p. 773.
6 V. Sklenicka, K. Kucharova, V. Foldyna and J. Cadek,
Interrelationship between creep deformation and creep
rupture in a low alloy CrMoV steel after service. In B.
Wilshire and R. W. Evans (eds.), Proc. 3rd Int. Conf. on
Creep and Fracture of Engineering Materials and Structures, The Institute of Metals, London, 1987, p. 361.
7 V. M. Radhakrishnan, Characterization of creep deformation of a Ni-base superalloy by constitutive equation, lOth
Congress on Material Testing, Budapest, October 1991.
8 V. M. Radhakrishnan and M. Kamaraj, Mater. Sci. Eng.,
A127(1990) L I S - L 1 8 .