Int. J. Pres.Ves.

&Piping8 (1980) 165-185

CREEP A N D C R E E P - R U P T U R E BEHAVIOUR OF A
BAINITIC 2 1/4 Cr-1 Mo STEEL*

R. L. KLUEH

Metals and Ceramics Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830, USA
(Received: 9 August, 1979)

ABSTRACT

The creep-rupture behaviour o f a bainitic 2 1/4 Cr-1 M o steel was determined at 454,
510 and 566 ° C. Tests were made on specimens taken f r o m a commercial heat of steel
in the normalised-and-tempered condition. In addition to the creep-rupture
behaviour, secondary and tertiary creep behaviour were examined. A relationship
between the onset o f tertiary creep and rupture liJe was obtained. The resultsJi'om
these tests were compared with available data on annealed, normalised-and-tempered
and quenched-and-tempered base metal and with tempered weM metal. The results o /
these comparisons gave insights into the types o f metallurgical change that occur in
2 1/4 Cr-1 M o steel at elevated temperatures.

INTRODUCTION

The mechanical properties of 2 1/4 Cr-I Mo steel depend on the microstructure

which, in turn, depends on the cooling rate from the austenitising temperature, in
the annealed steel, properties are determined primarily by the proeutectoid ferrite
that makes up the major part of the microstructure. ~'2 For the normalised-and-
tempered or quenched-and-tempered steels, large amounts of bainite are present in
the microstructure. Depending on the section size heat treated, the microstructure
can contain up to 100 per cent bainite (the smaller the section size heat treated, the
more bainite formed). Since the weld metal in a given weld pass will usually have
quite a small cross-sectional area and will be air cooled, 2 1/4 Cr-I Mo steel weld
metal is expected to be almost entirely bainite. 3 Thus, if the properties are
determined by microstructure, the properties of the weld metal should approach
those for the base metal in the bainitic condition.

* Research sponsored by the Division of Reactor Research and Technology, US Department of Energy,
under Contract W-7405-eng-26, with the Union Carbide Corporation.
165
Int. J. Pres. Ves.& Piping 0308-0161/80/0008-0165/$02.25 ~.~AppliedSciencePublishers Ltd, England, 1980
Printed in Great Britain
166 R.L. KLUEH

We previously reported on the creep-rupture properties of annealed 2 1/4 Cr-1

Mo steel taken from a 25.4-mm thick plate, l In this paper we report on the
creep-rupture properties of this same heat of material, but in the normalised-and-
tempered condition.

EXPERIMENTAL PROCEDURE

The heat of steel tested was a 25.4-mm thick plate obtained from Babcock and
Wilcox (B&W heat 20017) (Table 1). To normalise the plate, sections of it were
heated for 1 h at 927°C, then air cooled; the plate was tempered for 1 h at 704'~C.
After this heat treatment, the steel was entirely bainitic (Fig. 1).
Creep-rupture tests were made at 454, 510 and 566 °C on specimens with a 6.4 mm
diameter x 63.5 mm long reduced section. The tests were made in air on lever-arm

TABLE 1
CHEMICAL COMPOSITION OF 25-4-mm THICK PLATE OF 2 1/4 Cr-I Mo STEEL

Source Chemical composition (wt per cent)

oJ results C Mn St' Cr Mo Ni S P

Vendor 0,1 l 0.55 0-29 2.13 0.90 0.014 0.011

ORNL 0.135 0.57 0.37 2.2 0.92 0.16 0.016 0.012

Fig. 1. Microstructure of the normalised-and-tempered 2 1/4 Cr-t Mo steel (100 x ).

 CREEP OF BAINITIC 2 1/4 Cr-I Mo STEEL 167

creep frames with 12:1 and 20:1 ratios; the specimens were heated by a Marshall
resistance furnace. During testing the temperature was monitored and controlled by
three chromel versus alumel thermocouples attached along the specimen gauge
section. Temperatures were controlled to ___1 °C and the temperature varied less
than +__2 °C along the gauge section. Creep strains were measured with a mechanical
extensometer attached to the specimen shoulders and the elongation was read
periodically on a dial gauge with a sensitivity of 0-3 mm.

RESULTS

The creep-rupture properties were determined at 454, 510 and 566 °C (Table 2).
Logarithmic plots of stress to rupture against rupture life (Fig. 2) and stress against
minimum creep rate (Fig. 3) curved downward at all three temperatures. The data

TABLE 2
THE CREEP--RUPTURE PROPERTIES OF A BAINITIC (NORMALISED-AND-TEMPERED) 2 1/4 Cr-I Mo STEEL

Stress Rupture Elongation Reduction Minimum creep

( M Pa) liJb (per cent ) oJ area rate
(h) (per cent) (per cent/h)

454°C
379 2218.3 9.9 a 73.0 0.000133
393 2015.9 9.8 a 69.7 0.00024
407 1483.9 18.5 67-5 0.00033
441 831.1 20-1 76.5 0.00086
448 513.3 20-0 73.3 0.0011
455 542.3 18.4 78-2 0.0015
482 125.0 16.8 72.5 0.010
517 53.5 18.4 74.9 0.045
544 14.4 18.0 74.9 0-167
510°C
241 4133.5 19.4 40.7 0.001
276 1558.9 20.8 65.3 0.0029
310 416.0 26.4 79.9 0.0075
324 428.1 25.1 82.7 0.008
345 130.0 27.2 80.8 0.013
379 36.8 23.2 77.7 0.070
407 28-5 22.4 79.9 0-077
566°C
124 6945-0 7.0 24.3 0.00029
152 1747.2 10.0 23.2 0.00185
172 1020.9 15.2 39.1 0.003
207 362.2 15.7 37.8 0.010
207 436.8 14.8 42.4 0.0084
241 86.1 22.1 74.4 0.052
276 22.0 23-1 84.5 0.19
276 15.4 21.0 84-9 0.30
310 3.8 18.5 85.3 1.13

Specimen failed near the end of the gauge section.

168 R.L. KLUEH

1000 I I IIIIIll I I I111111 I I IIIIIll 1 I IIIII11

- t00

so

u~ 200 u~

20
100 I I Illllll I I lllllll I I IllLIII I I IIlllll
~00 ~0 ~ 102 t03 104
RUPTURE LIFE (hr)

Fig. 2, Creep-rupture c u r v e s for a b a i n i t i c ( n o r m a l i s e d - a n d - t e m p e r e d ) 2 1/4 Cr- 1 M o steel p l a t e at 454,

510 and 566°C.

points were visually fitted with two straight lines. Except at 454 °C, the ductility, as
determined by the reduction of area and total elongation measurements, decreased
with increasing rupture life (Fig. 4). Curves drawn through the ductility data points
showed certain trends. The reduction of area data at 454 °C showed little change for
all the tests, although the total elongation did fall off quite sharply above 1500 h.
However, both the specimens that showed the reduced elongation fractured near the
end of the gauge section. At 510 and 566°C there was a substantial decrease in
ductility with increasing rupture life.
Visual observation of the fractured specimens coincided with the observations on
ductility (Fig. 4). At 454 °C, all the specimens had a ductile cup-cone type of fracture.
Similarly, at 510°C all but one specimen showed cup-cone failures. The only
exception was the specimen with the longest rupture life (the test at 241 MPa), which
had a flat fracture with what appeared to be cracks on the external surface of the
gauge length. At the highest temperature, 566°C, all specimens tested at stresses
below 207 MPa had fiat fractures. Above this stress, fractures were of a cup-cone
type.
Metallographic examination of selected fractured specimens verified the visual
observations described above. At 454 °C the fractures appeared to be almost entirely
transgranular. Although the lowest stress test (379 M Pa) appeared to fracture
transgranularly (a cup-cone fracture), there was isolated evidence of the start of
intergranular separation back from the fracture surface; small cracks and holes were
visible on grain boundaries perpendicular to the specimen axis. As further evidence
that this was the beginning of grain boundary separation, the specimens at the
higher stresses contained rows of cavities parallel to the specimen axis, while the test
at 379 MPa contained no such cavities.
At 510 and 566 °C, several of the low-stress tests displayed intergranular fractures.
The specimens from tests at 510°C at 241 and 276 MPa had wedge-type grain
 P~
J CREEP RATE (% / h )

I I I I I I II I ! I I I i I II I I I I 1 I]11] I I I I IIlI~

~_. ~

o O

-- 0 1 > 0
I I I -- t'~
-- ~ -

1 1 I I I IIII I I I I I Illl I I I I Illll I I I t IIIII

0
170 R.L. KLUEH

30 I II[lli I I I II1[1[ A I [ I I1[111 [ [ ! [rill

i - ---'"~4L I
z (o)
0 2O o "° ~'~''c:~ o -- n~g. /
.-----., _._...~m- -m.~,. -I
:7
o
t0 t I "un~°e • \ t
0 I ILIIIII I k I111111 I I illllll I I llJtll!
90 I lrlll~t ] r t tt~ttF [ I r tlt~tt t I [r]ftl I
(b)
80

70 -
i,i
nt-
"~ 6 0 -
[a.
o
_g 5o
• ~54 °c k
ow 40 - • 5 t 0 °C ek~._..~e ~A
• 5 6 6 °C

30 -
ke~LI
20 I I IIIllll i I IIIJlll I J I Ililll
100 I0 ~ t0 2 t0 3 t0 4
RUPTURE LIFE (hr)

Fig. 4. Variation in rupture elongation and reduction of area with rupture life for a bainitic
(normalised-and-tempered) 2 1/4 Cr- 1 M o steel plate at 454, 510 and 566 °C. The curves are visual fits just
m e a n t as an aid to the visualisation of the data trends.

boundary cracks that formed perpendicular to the specimen axis. Both specimens
showed some neck formation, with the amount decreasing with decreasing stress.
The number of cracks increased with decreasing creep stress. The cracks formed in
prior austenite boundaries which have become boundaries between bainite grains
(i.e. the austenite grain boundaries served as bainite nucleation sites).
At 566 °C, the fractures at 124, 152 and 172 M Pa were quite flat with a high density
of cracks oil the grain boundaries perpendicular to the specimen axis. An abundance
of cracks was noted all along the specimen gauge length. Although a slight neck
formed at 207 MPa, there was extensive grain boundary separation. At stresses of
241 M P a and greater, fractures appeared to be entirely transgranular.
The change in fracture mode agrees roughly with the change in slope in the
creep-rupture curves at 510 and 566 °C (Fig. 2). A change in slope was also noted at
454 °C, although there was limited indication of grain boundary crack formation at
this temperature.
One of the limiting criteria used in the ASME Boiler and Pressure Vessel Code
Case N-47 to set allowable stresses for design in the creep regime is the time to the
onset of tertiary creep (at present Code Case N-47 applies only to annealed 2 1/4 Cr- 1
 CREEP OF BAINITIC 2 1/4 C r - I M o STEEL 171

M o steel). All the c r e e p c u r v e s for the b a i n i t i c 2 I/4 Cr-1 M o steel w e r e f o u n d to be

classical in the sense t h a t t h e y c o u l d be s e p a r a t e d into p r i m a r y (transient), s e c o n d a r y
(steady state) a n d t e r t i a r y stages. W e d e t e r m i n e d the o n s e t o f t e r t i a r y c r e e p by t a k i n g
a 0.2 per c e n t offset f r o m the linear s e c o n d a r y c r e e p r a t e line ( T a b l e 3).
F o r this s a m e h e a t o f steel in the a n n e a l e d ( f u r n a c e - c o o l e d ) a n d i s o t h e r m a l l y
a n n e a l e d ( f u r n a c e - c o o l e d to 704 °C, held for 2 h t h e n c o o l e d to r o o m t e m p e r a t u r e )
c o n d i t i o n s , a r e l a t i o n s h i p o f the type:

t 2 = At~ (1)

was f o u n d , 4 w h e r e t 2 is the t i m e to the o n s e t o f t e r t i a r y creep, t R the r u p t u r e life a n d

A a n d c~are c o n s t a n t s . W h e n the d a t a f r o m T a b l e 3 a r e p l o t t e d a c c o r d i n g to e q n (1),
it is f o u n d t h a t the d a t a fall on a s t r a i g h t line (Fig. 5). A l e a s t - s q u a r e s fit o f the d a t a
gives:
TABLE 3
TIME AND STRAIN TO ONSET OF TERTIARY CREEP OF BAINITIC (NORMALISED-AND-TEMPERED) 2 1/4 Cr-I Mo
STEEL

Stress Rupture Time to Strain l o e2 a

(MPa) I~e tertiary tertiary --
(h) (h) (per cent) ~R

454°C
379 2218.3 1475 0-5 0.05 h
393 2015.9 1440 0.7 0.07 b
407 1483.9 960 0.6 0.03
441 831. I 450 0.8 0.04
448 513.3 290 0.8 0.04
455 542.3 330 0-7 0.04
482 125.0 80 1.4 0"08
517 53.5 35 2-4 0.13
544 14-4 9.5 2"2 0.12
510°C
241 4133"5 2600 3-1 0" 16
276 1558"9 1050 3-8 0.18
310 416.0 260 2.6 0-10
324 428-1 215 2-4 0' 10
345 130.0 85 2.0 0.07
379 36.8 24.5 2.5 0' 11
407 28.5 18.5 1.8 0.08
566°C
124 6945.0 4750 1"8 0.26
152 1747.2 1030 2.5 0"25
172 1020.9 550 2.5 0.16
207 362.2 205 2.6 0"17
207 436.8 250 2-7 0.18
241 86" 1 45 2'9 0.13
276 22.0 12 3.3 0'14
276 15.4 9 3.5 0'17
310 3.8 2.3 3.1 0.17
ae 2 = creep strain to end of tertiary creep; e~ = rupture strain.
bThe total elongation for these two specimens may be greater than measured; these specimens failed near
the end of the gauge section.
172 R . L . KLUEH

1o4 - I I lilllll I IllrllN I I IIIIIII I I IIHIII/-~

"2
Z
_-
_ • 454 oC
//
" 510°C t'

n 10:3 _ • 566"C ~e --_

b.J ~ e
t'K _ ,~a
(J --
>_ -- ~ --
(~ -- ea --

102 -- • Z
b.I -- .,* --

0 -- a/ --
l-- -- a~
"' /.
~7 101 ____ . '
0 •
OJ -- /

10o I11111 J I I[11111 L I IIIllll I I lilllll

100 101 102 103 104
/R, RUPTURE LIFE (hr)

Fig. 5. Time to the onset of tertiary creep plotted against rupture life for a bainitic (normalised-and-
tempered) 2 1/4 Cr-1 Mo steel plate at 454, 510 and 566°C.

t 2 = 0"590t 1°°° (2)

(the correlation coefficient, R 2 = 0.998).
The strain to the onset of tertiary creep was also determined (Table 3). At both 454
and 566 °C, this strain decreases with decreasing stress (increasing rupture life), while
at 510 °C it appears to increase with decreasing stress. The strain to the onset of
tertiary creep has decreased quite dramatically at 454 °C (it falls considerably below
any of the values achieved at the other two temperatures). Also shown in Table 3 are
the ratios of the strain to the onset of tertiary creep to the rupture strain, %/eR. This
ratio increases with decreasing stress at both 510 and 566°C, while at 454°C it
appears to go through a minimum.

DISCUSSION

Although large quantities of normalised-and-tempered 2 1/4 Cr-1 Mo steel are used

by the petroleum and utility industries, probably only a small portion of that
material is entirely bainitic when placed in service. On the other hand, quenched-
and-tempered steel--even quite thick sections--is often entirely bainitic, s6
 CREEP OF BAINITIC2 1/4 Cr-I Mo STEEL 173

Similarly, because of the small cross-sectional area of a weld pass, the weld will be
entirely bainitic. 3
When Smith 7 collected and compiled all the available elevated temperature data
for 2 1/4 Cr-1 Mo steel in 1966, he combined the normalised-and-tempered (ASTM
Specification 387, Grade D) and quenched-and-tempered (ASTM Specification 542)
data. The quenched-and-tempered data were further broken down into material
with ultimate tensile strengths greater, or less, than 792 M Pa. For the temperatures
of interest in this study, Smith had limited data at 454 and 510 °C and none at 566 °C.
The creep-rupture strength of the material tested in the present study was slightly
greater than the strength of the normalised-and-tempered data given by Smith 7 at
454 and 510 °C. However, this is to be expected since most of the data given by Smith
were from plates between 100 and 180 mm thick. Assuming that the test specimens
were taken from within these plates (not on the surface), considerable proeutectoid
ferrite would probably be present in the microstructure, a thus decreasing its strength
relative to the strength of a completely bainitic microstructure. Such a conclusion is
justified by an examination of the data of Mandich et al., 9 who studied the
properties of normalised-and-tempered 152-mm thick plate and found values of 462
and 586 MPa for the room temperature yield and tensile strengths at the 1/2 T
position, respectively. This compares with values of 599 and 668 M Pa for the plate
used in the present study.l° Mandich et al. 9 also carried out a limited number of
creep-rupture tests; these properties also fell below those obtained in the present
study. The results were more in line with Smith's data for the normalised-and-
tempered steel. * Although Mandich et al. did not discuss microstructure, the above
observations on strength differences, along with the fact that they used a
normalising-and-tempering treatment similar to that used in this study, tend to
confirm a difference in microstructure.
For plates between 100- and 178-mm thick, quenched-and-tempered steels are
expected to be almost entirely bainitic. 5"6 The ASTM Specification A542 for the
quenched-and-tempered grade divides the material into three classes, according to
the room temperature strength level. Classes 1, 2 and 3 have minimum tensile
strengths of 723, 792 and 655 MPa, respectively, and minimum 0.2 per cent offset
yield stresses of 586,689 and 517 M Pa, respectively. These strengths are obtained by
tempering the quenched steel at different temperatures for various periods of time.
When the results of this study were compared with the limited quenched-and-
tempered data given by Smith, 7 who divided his data according to whether the
tensile strength was greater or less than 792 MPa, our data at 454 and 510°C agreed
well with the data for a tensile strength less than 792 MPa and fell below the data for
steel with a minimum of 792 MPa. Since it has been shown that the creep-rupture
strength is directly proportional to the tensile strength, 7' 11 this comparison is quite
reasonable.
Part of the data reported here was previously reported in a study on the effect of
carbon on the mechanical properties. 12 In that study, data were also obtained for a
174 R . L . KLUEH

normalised-and-tempered 22-mm rod, which was taken from a commercial heat

(0.12 per cent C). This steel also had a bainitic microstructure with a room
temperature tensile strength of 655 M Pa. 13 The creep-rupture properties for the rod
were quite similar to those of the plate reported here. Smith 7 has given a curve for the
relationship between the 10,000-h rupture life at 454 °C and the room temperature
tensile strength. The extrapolated results from the present study and those for the
rod 12 fall on that curve. Since that relationship was developed with data from
quenched-and-tempered materials, these results enhance our conclusion that
microstructure (for a given strength) is the determining factor (tensile strenths of the
order of 689 M Pa can only be developed with material that is essentially 100 per cent
bainite).
Obviously, the gross microstructure (i.e. bainite) is not in itself the determining
factor. As discussed later in this section, the properties of these steels are largely
determined by the precipitate structure. When the steel is tempered, changes occur
in the precipitate morphology and distribution, which results in strength changes.
Evidently, the tensile strength for these bainitic steels is a good measure of the
precipitate microstructure.
To obtain the desired strength for a quenched steel, it is tempered. For the
higher strengths--greater than 6 8 9 M P a - - t h e steel is usually tempered below
649 °C. 5- 7.9. ~ Hence, the 704 °C temper used in this study accounts for the lower
strength obtained for our bainitic structure. Before tempering, this material had a
room temperature tensile strength of 1054MPa. x3 A tempering temperature of
704 °C is in the range of temperatures used for a post-weld heat treatment for welds
made on annealed 2 1/'4 Cr-1 Mo steel. If our conclusion that the microstructure
determines the strength is correct, then, since weld metal should contain a bainitic
microstructure, the weld metal properties given such a temper should be similar to
those obtained in this study.
We previously reported on the effect of carbon on the weld metal properties of
2 1/4 Cr- 1 Mo steel. 3. ~4 One of the weld metals tested was from commercial material
(0.11 per cent C) and the room temperature tensile strength of that weld metal was
approximately 689 M P a 3 not too different from the material tested in this study.
Tests were made at 454, 510 and 566°C and the creep rupture curves are given in
Fig. 6, along with the curves shown in Fig. 2. Since only limited data were obtained
on the weld metal, curvature could be estimated only for the 566°C results.
Nevertheless, the agreement between the base metal and the weld metal is excellent,
again enhancing the conclusion that microstructure determines the strength,
regardless of the way in which the steel was fabricated or cast prior to heat
treatment. Comparison with creep rates can also be made for these same materials
(Fig. 7). Everything considered, the agreement is excellent. Smith 7 gives rupture
data for weld metal at 566~'C which fall between the curves in Fig. 6.
In our study of the effect of carbon on the creep properties of 2 1/4 Cr-1 M o steel,
we also analysed the tertiary creep behaviour. ~5 To determine the onset of tertiary
creep in that study, we tried to estimate the time at which the creep rate began to
 CREEP OF BAINITIC 2 1/4 Cr-I Mo STEEL 175

1000
- I illlifll I I Illllll I Illiiiii I I IIIIIII
- - - ~00

500 -- ~

-- 50

kd bJ
cr ~

~n
~- 200 -- ~ ° c -
WELD METAL
- - - - B A S E METAL (BAINITIC) " " t : ~ -- 20

too I I IIit111 I I llliill I I III1111 I I 1111111

t0 0 10 ~ 10 2 t0 3 10 4
RUPTURE L I F E (hr)

Fig. 6. A comparison of the creep rupture curves for the tempered bainitic 2 1/4 Cr- 1 Mo steel plate and
tempered bainitic 2 1/4 Cr-I Mo steel weld metal at 454, 510 and 566°C.

increase after the period of steady-state creep. A correct estimate of t 2, the onset of
tertiary creep, using such a technique, is difficult. The estimated t 2 will depend upon
how much the creep curve is magnified. A much more accurate (and consistent)
technique is the use of a 0.2 per cent offset technique, as was employed in the present
study (Table 3).
In the previous study,l 5 the data were analysed according to relationships defined
by Garofalo et al.16 and by Leyda and Rowe. 1~ Garofalo wrote:
t, = Bt~ (3)
where B and/3 are constants. For austenitic stainless steels, Garofalo found fl ~ 1
and B depended on temperature. On the other hand, Leyda and Rowe ~7 gave:

t 2 = Fst R (4)

where F s is a constant. With/3 = 1, eqns (3) and (4) are identical. The Leyda and
Rowe results for various alloys, including 2 1/4 Cr-1 Mo steel, showed that F s
decreased slightly with temperature.~ 7
We previously analysed our data for the 25.4-mm plate and the 22-mm rod
according to eqn (4). However, F s was not a decreasing function of temperature.
Rather, F s was found to be 0.402, 0.312 and 0.344 at 454, 510 and 566°C,
respectively. ~5 When analysed according to eqn (3),/3 approached unity and B was
essentially independent of temperature. As pointed out by Booker and Sikka, 18 the
nature of the d a t a - - u s u a l l y many short-time tests and a few long-time tests--lends
itself to analysis in logarithmic space. Since it is often of interest to know what
fraction of life occurs before the onset of tertiary creep, we chose to express our data
according to eqn (1):

t 2 = At~ (1)
176 R . L . K'LUEH

STRESS ( kll )
2O 40 60 80 400

! I I /i I f I r i
,o o _ WELD METAL m
i ~ BASE METAL
( BAINfTIC )

L
~0"t __

,///
/
t 0 "2
w

/
/
566
/ [
iO-3
4~4
~(0"C /
!
/ / //
/
10 "4 1 I /1 I 1 I 1 I
tOO 200 400 600 0OO 4000
STRESS ( MPO )

Fig. 7. A comparison of the stress-minimum creep rate relationships for the tempered bainitic 21/4 Cr.
1 Mo steel plate and tempered weld metal at 454, 510 and 566°C.
 CREEP OF BAINITIC 2 1/4 Cr-I Mo STEEL 177

where A and ~ are constants. A similar relationship was found to hold for annealed
2 1/4 Cr-l Mo steel. 4
As seen from Fig. 5, the data for the 25-mm plate follow a relationship of the type
given by eqn (1). Furthermore, A in eqn (l) is independent of temperature and ~ is
essentially unity. Thus, the Leyda and Rowe equation is still obeyed, but F s is
independent of temperature.
Since there are few data available on tertiary creep of bainitic 2 1/4 Cr- 1 Mo steel,
we redetermined t 2 for the 22-mm rod by the 0.2 per cent offset method (Table 4);
these data are shown in Fig. 8, along with the data for the 25-mm plate. The equation
for the combined data is:
t2 = 0"563t~ °°8 (5)

TABLE 4
TIME AND STRAINTO ONSETOF TERTIARY CREEP OF 22-ram ROD OF (NORMALISED-AND-TEMPERED)2 1/4 Cr- 1
Mo STEEL

Stress Rupture Time to Strain to e2°

( MPa) life tertiary tertiary ~
(h) (h) (per cent)

454°C
448 707.5 450 1'2 0"06
465 242.9 147 1.9 0.10
483 136.9 79 1"3 0"09
517 18.0 8 2.3 0'05
510°C
276 829"9 430 2"4 0" l0
310 173-5 60 1"0 0"04
345 104'3 66 2"5 0"09
379 lY4 8 2"5 0"08
414 2"9 1"6 3"0 0"ll
566°C
138 2319.5 1000 2'2 0"15
172 435"9 195 2"4 0"10
207 254"3 145 3"8 0'12
207 194'5 120 4"0 0"16
241 25"0 13 4"5 0'16
241 32"4 20 2"8 0" 1 !
310 1'3 0'7 3"0 0"12

°e 2 = creep strain to end of tertiary creep; eR = rupture strain.

( R 2 = 0.997). The A has decreased slightly (from 0.590 to 0.563) but, considering the
data scatter, such a slight deviation is not unexpected. It is interesting to note that ct
has changed very little; again, it is essentially unity.
The onset of tertiary creep for the 0.11 per cent C steel weld metal was also
determined (Table 5) and in Fig. 8 these data are shown along with the rod and plate
data. Again, the fit of the data appears excellent. Equation (l) applies for these
bainitic steels, regardless of whether they are base metal or weld metal. Of course, the
present data were determined for a bainite tempered to about 689 MPa.
178 R . L . KLUEH

-- I f I rillll I I [[1[[11 [ I I I I I T ~ ] ~ ~
-_ WELD ,/
"3 -- ROD PLATE METAL / 5
v -- 454°C o . a" /
510oc A • ""
a_ 10 3
w " a~o
bJ
rr

>-
rr
o . ° l

~. 10 2
bJ

,/
bJ
~ 101 ,_

./
~0 ° 1 JJJl / [/_Jllllt l J J_llllJl I--] [ JlXlt '
10 o 101 t0 2 10 3 fO4
IR, RUPTURE LIFE (hr)

Fig. 8. Time to the onset of tertiary creep plotted against rupture life for a tempered bainitic plate and
rod and for tempered weld metal. The least-squares line was determined from the plate and rod data.

TABLE 5
TIME A N D STRAIN TO ONSET OF TERTIARY C R E E P OF 1 E M P E R E D 2 1/4 Cr-I Mo SIEEL

Stress Rupture Time to Strain to c2"

( M Pa ) liJe tertiary tertiary
~R
(h) (h) (per cent)

454 ° C
434 801.3 557 0.8 0-06
448 497.5 349 0-9 0.08
483 72-2 54 1-7 0.15
517 8.0 5 1.7 0.14
510°C
276 680"7 395 3.5 0"21
276 862.6 490 3.3 0.22
310 247"1 130 2.3 0.14
379 31-3 18 1.9 0.11
566°C
138 2017-2 890 1.7 0.15
172 714.5 545 2.5 0.37
207 137.0 84 3.6 0.22
276 14.5 8.4 2.9 0.18

ae 2 = creep strain to end of tertiary creep; eR = rupture strain.

 CREEP OF BAINITIC 2 1/4 Cr-I Mo STEEL 179

In Tables 3, 4 and 5 the creep strain to the onset of tertiary creep is given for the
two base metals and the weld metal. At all temperatures and stresses there is
considerable scatter and no definite trends can be discerned. The primary difference
is that the strain is less at 454°C than at 510°C and 566°C. At 510°C and 566°C it
ranges from 1.7 to 4.5 per cent whereas, at 454°C, the range is from 0.5 to 2.4
per cent.
Our studies on 2 1/4 Cr-I Mo steel were motivated by its use as the structural
material for the steam generator of the demonstration liquid metal fast breeder
reactor. In that application, the steel is to be used in the annealed condition.
However, as discussed above, any welds in that structure will have properties similar
to the material discussed in this paper. Hence, it is of interest to compare properties
for the different microstructures that will be present in a weld area (i.e. weld metal,
base metal and heat-affected zone). Whereas the creep and creep-rupture properties
discussed above are determined by bainite, the proeutectoid ferrite determines the
properties in the annealed condition.1 The properties of both proeutectoid ferrite
and bainite are determined by the carbide precipitates present in the constituent
after heat treatment and the evolution of these precipitates during elevated
temperature exposure (during a creep test or in service).
The carbide precipitation reactions in 2 1/4 Cr-1 Mo steel were studied by Baker
and Nutting 19 and were previously used to explain the effect of carbon on the
creep-rupture properties. 12 The matrix of both proeutectoid ferrite and bainite is a
ferrite solid solution. The major differences in the ferrite of these two austenite
decomposition products come from the differences in the decomposition kinetics.
Proeutectoid ferrite forms as relatively clean equiaxed grains. Eventually a high
density of needle- or plate-like Mo2C precipitate particles form. This unstable
precipitate is replaced by eta-carbide (eta-carbide is referred to in most of the
literature as M6C ; recent studies 2° have shown the stoichiometry to more closely
approach M4C ). Bainite, on the other hand, when formed, contains a high density of
precipitate particles--mostly M3C; the bainitic-ferrite matrix also contains a high
density of dislocations.13 Upon tempering, MoeC is the first precipitate to form.
Then, during further elevated temperature exposure, M3C and M02C are replaced
by M7C3, M23C 6 and eta-carbide; the MTC 3 is also eventually replaced by eta-
carbide. Since the annealed steel contains some bainite and pearlite, which contains
M23C 6 platelets, 21 the annealed and bainitic steels both approach an equilibrium
structure o f a ferrite matrix that contains large globular particles of M23C 6 and eta-
carbide.19'21.22
A comparison of the creep-rupture properties of proeutectoid ferrite and bainite
(Fig. 9) shows the strength differences for the annealed (proeutectoid ferrite) and
normalised-and-tempered (bainitic) 2 1/4 Cr-1 Mo steel. The curves for the
annealed steel were obtained from the same heat of steel as was used in the present
study. ~ The curves for the normalised-and-tempered steel show a downward break,
180 R . L . KLUEH

NORMALIZED AND TEMPERED (BAINtTIC)

~ ANNEALED
--ISOTHERMAL ANNEALED
t000
~MTTT~MTFITr~-T--FTlqqTIF 7 TTTTTll _] ~oo
- - .,..// 4 5 4 °C
-8 500 - ~/

~ rY
~ 2o0 % .... -"-~ ~ ~ ~

oo _~.AAJ21ill _ .[ I A I I [ ~ ±~JJZLLLL. ~1. ~ _ J

10 0 40 t ~0 2 10 3 10 4
RUPTURE LIFE (hr)

Fig. 9. A comparison of the creep-rupture curves for 2 1/4 Cr-I Mo steel with microstructures that are
primarily bainite (normalised-and-tempered) and proeutectoid ferrite (annealed). The curves are for the
same heat of steel. At 454 °C a curve for this same steel in an isothermal annealed condition is also shown.

whereas those from the steel given an anneal (full anneal) do not. At 510 and 566 °C,
the strength of the bainitic steel is definitely superior to that given the anneal
treatment. However, at 454°C this is not the case. The extrapolated 104h rupture
curves (Table 6) indicate that the curves cross. The observed behaviour is
explainable in terms of the types of precipitation reaction that occur in bainite and
proeutectoid ferrite. The Mo2C in both constituents provides most of the particle
dispersion hardening by precipitates. 19.23 Hence, the properties should be affected
when the Mo2C transforms to eta-carbide. In bainite, the kinetics of the
precipitation reactions that transform the M%C to eta-carbide are much faster than
they are in the proeutectoid ferrite. This gives rise to the approach of properties at
566 °C where these reactions proceed quite rapidly. At the intermediate temperature
of 510 °C, the reactions do not proceed as rapidly and the approach of properties is
much more gradual. At 454°C, the crossover of the curves is a result of a solid
solution strengthening effect that occurs in the proeutectoid ferrite ( annealed steel)
prior to the completion of Mo2C precipitation.
For a ferritic steel that contains molybdenum and carbon, Baird and

TABLE 6
STRESS TO RUPTURE FOR 2 1/4 Cr-I Mo STEEL

Temperature Rupture stress j b r 104 h ( M P a )

( ° C) Normalised-and-tempered Annealed
( Bainitic ) ( Pr oeutect oid Jbrrite )

454 320 330

510 215 150
566 120 100
 CREEP OF BAINITIC 2 1/4 Cr-I Mo STEEL 181

Jamieson 24"25 showed that dislocation motion in creep is hindered by interaction of

dislocations with m o l y b d e n u m and carbon atoms or atom clusters. In annealed
2 1/4 Cr- 1 Mo steel, this interaction gives rise to non-classical creep curves in certain
temperature and stress regimes. 26 Instead of a classical creep curve with a primary
(transient), secondary (steady state) and tertiary creep stage, a curve with two
steady-state stages is observed. The transition from the first to the second steady-
state stage involves quasi-tertiary (increasing creep rate) and quasi-transient
(decreasing creep rate) creep stages.
The first of the steady-state stages in the non-classical creep curves was concluded
to be the result of interaction solid solution hardening. ~- 3 With time, the amount of
molybdenum and carbon in solution decreases (due to precipitate formation) to the
point where the interaction no longer hinders dislocation motion. At this point the
creep rate increases and eventually establishes a new steady state. In this new steady
state the creep rate is controlled by atmosphere-free dislocations moving through
the Mo2C precipitate field of the proeutectoid ferrite.
For a given temperature, the non-classical curves occurred over an intermediate
stress regime. 26 Below some stress, which depends on temperature, the creep curve
becomes classical. This occurs at the point at which the molybdenum and carbon
atom clusters can diffuse with the moving dislocations. It was also found that above
some stress--also temperature dependent--another type of 'classical' curve
developed that was concluded to be the result of the larger number of dislocations
generated at the higher stress. At these higher stresses, the rate of dislocation
generation exceeds the rate at which they can be tied up by interaction solid-solution
strengthening.
The transition stresses, where the different controlling mechanisms come into
play, increase with temperature. 26 At 454°C, only the high-stress 'classical' creep
curves were observed for the annealed steel. It is the enhanced strengthening due to
interaction solid solution hardening that leads to the approach and crossover of
rupture properties for annealed and normalised-and-tempered steels (Fig. 9 and
Table 6). At 510 °C and 566 °C, precipitation has removed sufficient molybdenum
and carbon from solution to lessen the interaction solid-solution hardening effect in
the annealed steel.
With time at 454 °C, precipitation occurs and strengthening by interaction solid-
solution hardening no longer occurs. At that time, the creep-rupture curve will show
a downward break and probably fall below the normalised-and-tempered curve.
This was shown to be true when an alternate 'anneal' treatment was used. t Instead of
a full anneal (continuously furnace-cooled from the austenitising temperature), the
steel was isothermally annealed (furnace-cooled to 704°C, held for 2h, then
furnace-cooled to room temperature). The curve at 454 °C for this heat treatment is
also shown in Fig. 9. In this case, the downward break occurs quite early and the
estimated 104 rupture life is considerably below that for the normalised-and-
tempered steel (275 M P a versus 320 MPa). The difference in these curves is caused
by the greater amount of carbide precipitation that occurs during the isothermal
182 R.L. KLUEH

anneal heat treatment. More molybdenum and carbon precipitation occurs during
the isothermal anneal heat treatment, thus leaving less of these constituents in
solution for interaction solid-solution hardening, 26
Interaction solid-solution hardening also occurs in bainite. ~°'2~' However, we
previously concluded that the effect in this case was probably due to
c h r o m i u m - c a r b o n interactions instead of m o l y b d e n u m - c a r b o n interactions. 226
This difference is caused by the different precipitation processes in bainite and
proeutectoid ferrite. In bainite, most of the M02C forms during the normalising-
and-tempering treatment. 2 ~'~When the specimen is put into test, the chromium-rich
carbides are in the process of forming and the pre-precipitate clusters of chromium
and carbon give rise to interaction solid-solution hardening. The chromium-carbon
interaction is weaker and has its maximum effect in a tensile test at a lower
temperature than the m o l y b d e n u m - c a r b o n interaction. 2"~ 2~, Evidently, these
differences, along with the more rapid precipitation processes in bainite, minimise
the interaction solid-solution hardening effect on ~reep. Thus, only classical creep
curves were observed in the bainitic steel.
The creep rate, ~i, is generally related to the stress, c,, by:
~i = Ka" (6)
where K and n (the stress exponent) are constants. Since there are breaks in the log ~!
versus log a curves (Fig. 3), two n values are required to describe the behaviour at
each temperature. At 510 °C and 566 °C, the values in both stress regimes are quite
similar, while much larger values are observed at 454 °C (Table 7). Generally, a large
stress exponent is associated with dispersion-hardened alloys. The higher n values at
454°C may be indicative of enhanced dispersion strengthening at the lower
temperature.

"FABLE 7
STRESS EXPONIeNTS FOR BAINITIC 2 1/4 Cr-I Mo STEEL

]k'mperature ~5"lre,~.s' exponent, n

((') t l i g h ,stress l.on ,~tress

454 25'0 10"5

510 12'8 7"8
566 13'6 6'5

Since a normalised-and-tempered 2 1/4 Cr-1 Mo steel can contain large amounts

of proeutectoid ferrite, it might be expected that the initial properties of a
normalised-and-tempered steel could range between those of proeutectoid ferrite
and bainite. Murphy and Branch 8 studied several normalised-and-tempered heats
and found the bainite content to vary between 34 to 95 per cent. Similarly, an
annealed steel may contain as much as 25 per cent bainite. Thus, a comparison of
the properties of bainite with those for proeutectoid ferrite gives some indication of
the kinds of strength variation possible in a normalised-and-tempered steel, tt
 CREEP OF BAINITIC 2 1/4 Cr-1 Mo STEEL 183

should also be noted that for a normalised-and-tempered steel with large amounts of
proeutectoid ferrite, non-classical creep curves would be expected, for the same
reasons as discussed for the annealed steel. 26
There is evidence that the effect of tensile strength on the long-term creep-rupture
properties of 2 1/4 Cr-1 Mo steel in the normalised-and-tempered condition applies
only to temperatures of about 538 °C. v"~1 Above 538 °C, the 10,000- and 100,000-h
rupture stresses are little affected by the original strength, since the increased
reaction kinetics at these temperatures causes the microstructures to rapidly become
similar. As discussed above, steels that are primarily bainite or proeutectoid ferrite
approach the same equilibrium. Thus, the elevated temperature strengths for the
two should eventually approach, and this was observed at 566 °C (Fig. 9). Although
an approach of properties is indicated at 566 °C, extrapolation of the creep-rupture
curves (Fig. 9) shows that they will not intersect until well beyond 105 h. Similar
observations on the approach of properties have been made by other investigators
who compared annealed and normalised-and-tempered properties;27,2s however,
for these studies the normalised-and-tempered steel contained considerable
proeutectoid ferrite, thus hastening the approach of properties.
It appears that the comparison of properties for different heat treatments
(normalised-and-tempered or annealed), rather than microstructures (bainite and
proeutectoid ferrite), may be the reason that normalised-and-tempered and
quenched-and-tempered 2 1/4 Cr-1 Mo steels are not widely used above 500 °C. The
approach of properties for normalised-and-tempered and annealed steel at
temperatures above 500 °C has generally been assumed to mean that the bainitic
steel rapidly loses its strength above that temperature. The results of this
study indicate that although the bainitic steel develops a downward break in
the stress-rupture curves (Fig. 9), the original superior strength over the steel with a
predominantly proeutectoid ferrite microstructure can be maintained for long times
at temperatures as high as 566 °C. Further work to quantify the relative strengths of
the different microstructures would appear appropriate.

SUMMARY AND CONCLUSIONS

The creep and rupture properties of a commercial normalised-and-tempered 2 1/4

Cr-1 Mo steel were determined at 454, 510 and 566°C. The following observations
and conclusions were made.
(1) Logarithmic plots of stress versus rupture life and stress versus minimum
creep rate showed a downward curvature, indicative of a change in the rate-
controlling mechanism. At 510 °C and 566 °C, this change in slope appeared
to coincide with a change in failure mode from transgranular to
intergranular. Although the 454 °C curve also showed downward curvature,
no correlation with change in fracture mode was possible.
184 R, L. KLUEH

(2) The onset of tertiary creep was found to be related to rupture life according
to:

t 2 = 0"590/~'°°8 (7)
This equation applied at all three temperatures.
(3) The creep-rupture behaviour of the normalised-and-tempered plate was
compared with literature data on bainitic 2 1/4 Cr-i Mo steel in the
normalised-and-tempered and quenched-and-tempered conditions; in
addition, the results were compared with weld metal data which also has a
bainitic microstructure. Similar creep and creep-rupture properties of base
metal and weld metal were found when bainitic steels with the same room
temperature tensile strength were compared. Even the tertiary creep
behaviour of weld metal was similar to base metal of similar tensile strength.
(4) Lacking detailed 2 1/4 Cr-1 Mo steel weld metal creep data, it should be
possible to estimate values from the base metal, if the tensile strength is the
same for the materials in the two different conditions.

ACKNOWLEDGEMENTS

Special thanks are due the following people who aided in this work: J. L. Griffith for
carrying out the experimental work; C. W. Houck for the metallography; T. L.
Hebble for statistical analysis of the data; R. K. Nanstad, T. K. Roche, C. R.
Brinkman and G. M. Slaughter for reviewing the manuscript; George Griffith for
editing the manuscript and Alice Rice for typing the manuscript.

REFERENCES

1. KLUEH, R. L. Heat treatment effects on creep and rupture behavior of annealed 2.25 Cr-I Mo steel,
Met Trans A, 9A (1978), pp. 1591 8.
2. KLUEH, R. L. Heat treatment effects on the tensile properties of annealed 2.25 Cr- I Mo steel, J. Nuc/.
Mater., 68 (1977), pp. 294-307.
3. KLUEH, R. L. and CANONICO, D. A. The microstructure and tensile properties of 2 1/4 Cr-1 Mo steel
weldments with varying carbon content, Weld J. (Miami), 55 (1976), pp. 253-s 9-s.
4. KLUEH, R. L. Tertiary creep behavior of annealed 2 1/4 Cr-I Mo steel, J. Nucl. Mater., 79 (1979),
pp. 363-71.
5. BROWN,R. M., REGE, R. A. and SPAEOER,JR., C. E. The properties of normalized and tempered and
quenched and tempered 6- I/4-inch-thick plates of 2 I/4 Cr- 1 Mo steel, In: Heat-treated steels Jot
elevated temperature service, New York, American Society of Mechanical Engineers, 1966,
pp. 27-44.
6. CARPENTER,O. R. and EMMANUEL,G. N. The use of quenched and tempered 2 1/4 Cr-I Mo steel for
the fabrication of advanced design pressure vessels, in : Heat-treated steelsJor elevated temperature
service, New York, American Society of Mechanical Engineers, 1966, pp. 27-44.
7. SMITH, G. V. Supplemental Report on the Elevated-Temperature Properties oJ
Chromium-Molybdenum Steels (An Evaluation oJ 21/4 Cr- I Mo Steel), ASTM DATA Ser. Publ. DS
6 $2, Philadelphia, American Society for Testing and Materials, March, 1971.
8. MugpHY, M. C. and BRANCS, G. D. Metallurgical changes in 2.25 chromium-molybdenum steels
during creep-rupture tests, J. Iron Steel Inst., 209(1971), pp. 546-61.
 CREEP OF BAINITIC 2 1/4 Cr-1 M o STEEL 185

9. MANDICH,L. 1., FOGELMAN,E. S. and GULYA,J. A. Elevated-temperature properties of heavy gauge

chromium-molybdenum plates. In: Heal-treated steelsJor elevated temperature sert, ice, New York,
American Society of Mechanical Engineers, 1966, pp. 45-65.
10. KLUEH,R. L. and OAKES,JR., R. E. Strain rate effects on the tensile properties of a bainitic 2 1/4 Cr- l
Mo steel, J. Eng. Mater. Tech., 99 0977), pp. 350-8.
I I. SANGDAHL,G. S. and VOORHEES,H. R. Quenched-and-tempered 2 1/4 Cr-I Mo steel at elevated
temperatures--Tests and evaluation. In: 2 I/4 Chrome-I molybdenum steel in pressure vessels' and
piping, New York, American Society of Mechanical Engineers, 1972.
12. KLUEH,R. L. The effect of carbon on 2.25 Cr-I Mo steel (ll), creep-rupture properties, d. Nucl.
Mater., 54 (1974), pp. 55-63.
13. KLUEH,R. L. The effect of carbon on 2.25 Cr- l Mo Steel (I), microstructure and tensile properties, J.
Nucl. Mater., 54 0974), pp. 41 54.
14. KLUEH,R. L. and CA~ONXCO,D. A. The creep-rupture properties of 2 1/4 Cr-l Mo steel weldments
with varying carbon content, Weld J. (Miami), 55 (1976), pp. 381-s-8-s.
15. KLUEH,R. L. The relationship between rupture life and creep properties of 2 1/4 Cr- 1 Mo steel, Nucl.
Tech., 26 (1975), pp. 287-96.
16. GAROFALO,F., WHITMORE,R. W., DOMIS,W. F. and VONGEMMINGEN,F. Creep and creep-rupture
relationships in an austenitic stainless steel, Trans. AIME, 221 (1961), pp. 310-19.
17. LEYDA,W. E. and ROWE, J. P. A Stud)' oJ Time Jor Departure Jrom SecondaJs" Creep Jor Eighteen
Steels, ASM Tech. Rep. Metals Park, Ohio, American Society for Metals, 1969, p. 6.1.
18. BOOKER,M. K. and SIKKA,V. K. Interrelationships Between Creep Lije CriteriaJor Four Nuclear
Structural Materials, ORNL/TM-4997, August, 1975.
19. BAKER, R. G. and NUT'rING, J. The tempering of 2 1/4 Cr-I Mo steel after quenching and
normalizing, J. Iron Steel Inst., 192 (1959), pp. 257 68.
20. LErrNAKER,J. M., KLUEH,R. L. and LAING,W. R. The composition of eta-carbide phase in 2 1/4 Cr- 1
Mo steel, Metall. Trans., 6A (1975), pp. 1949-55.
21. HALE, K. F. Study of creep in a chromium-molybdenum steel, Proc. 4th Intern. ConJ~ Electron
Microscopy, Vol. 1. Springer-Verlag, Berlin, 1960, pp. 650-8.
22. KLUEH,R. L. and LEITNAKER,J. M. An analysis of the decarburization and aging processes in 2 1/4
Cr-I Mo steel, Metall. Trans., 6A (1975), pp. 2089-93.
23. IRWNE, K. J., MURRAY,J. D. and PICKERING, F. B. Structural aspects of creep-resisting steel. In:
Structural processes in creep, London, Iron and Steel Institute, 1961, pp. 246-75.
24. BAIRD,J. D. andJAMIESON,A. Creepstrengthofsomesynthesizedironalloyscontainingmanganese,
molybdenum, and chromium, J. Iron Steel Inst., 210 (1972), pp. 847-56.
25. BAIRD,J. D. and JAMIESON,A. High-temperature tensile properties of some synthesized iron alloys
containing molybdenum and chromium, J. Iron Steel Inst., 210 (1972), pp. 841-6.
26. KLUEn,R. L. Interaction solid-solution hardening in 2.25 Cr- 1 Mo steel, Mater. Sci. Eng., 35 (1978),
pp. 239-53.
27. KRISCH,A. and NAUMArqN,F. K. Creep rupture strength and structure of the steel 10 CrMo 9 10,
Arch. Eisenhiittenwes., 41 (1970), p. 835.
28. YUKITOSHI,T., NISHIDA,K. and IDA,Y. Effects of heat treatment on creep rupture strength of low C 2
1/4 Cr-I Mo steel, Proc. Jap. Soc. Promotion Sci. Committee, 12 (1971), pp. 123-31.