The Effect of Carbon Content on the Microstructure

of an Experimental Heat-Resistant Steel

Frances Wang and Derek O. Northwood
Engineering Materials Group, University of Windsor, Windsor, Ontario, N9B 3P4, Canada

The effect of carbon content on the microstructure of a HP-Nb type heat-resistant steel
(35Ni-25Cr-1Nb) is examined for carbon contents of 0.35, 0.44, 0.56, and 0.62 wt.%. In the
cast state, the alloys contain both eutectic Cr7C 3 and eutectic NbC. Upon aging, Cr23C6 car-
bides precipitate out from the supersaturated austenite. The distribution of the Cr23C6
carbides was better, that is, more uniform, in the alloys containing 0.44 or 0.56 wt.%C than
in the alloys containing 0.35 or 0.62 wt.%C. Both the amount of the eutectic Cr7C 3 carbides
at the grain boundaries, and the Cr23C6 carbides within the grains, increased with increas-
ing carbon content. The relationship between the microstructures developed and the creep
and high-temperature tensile properties is briefly discussed.

INTRODUCTION which degrades the ductility. The preferred

carbon content is usually specified as being
Heat-resistant steels are used extensively for 0.2-0.6 wt.% [1].
reformer tubes in chemical fertilizer and Carbides can significantly affect the prop-
ethylene production plants. The most com- erties of the steel by precipitating at grain
monly used steel is HK40 (25Cr-20Ni-0.4C), boundaries and changing the strength of
which has a cast structure of supersaturated such regions [2]. The strengthening effect
austenite and eutectic carbides. Secondary of M23C6 carbides also depends on their
carbides, which are formed upon aging, im- precipitation morphology. When M23C6 car-
part good creep rupture properties to the bides are precipitated as dispersed particles
alloys. New developments include the ad- within grains and at the grain boundaries,
dition of carbide forming elements such as the alloy properties are excellent. If the car-
Nb, Ti, and Zr to the base alloy. This pro- bide precipitates as a continuous film, prop-
duces more secondary carbides, and the erties can be severely degraded. The pres-
eutectic carbides are more complex and more ence of a continuous film provides an easy
stable, resulting in improved ductility and fracture path. Also, grain boundary sliding
creep rupture strength. is restricted, leading to excessive stress
Carbon is a necessary element to improve buildup. On the other hand, the absence of
the strength of the cast steel. When the car- carbides at grain boundaries can lead to
bon content is below 0.2 wt.%, the cast steel premature failure. Therefore, for a given ma-
cannot be strengthened sufficiently. How- trix strength, optimum alloy properties are
ever, when the carbon content exceeds 0.6 achieved by a specihc distribution of carbides
wt.%, and the steel is used for a long time at the boundaries. The carbide distribution
at high temperatures, the carbides coarsen, required is one such that stresses can be re-

Originally presented at the 1990 IMS Convention, Cincinnati, Ohio, USA.

3
© Elsevier Science Publishing Co., Inc., 1993 MATERIALS CHARACTERIZATION 31:3-10 (1993)
655 Avenue of the Americas, New York, NY 10010 1044-5803/93/$6.00
4 F. Wang and D. O. Northwood

Table 1 Chemical Analysis for HP-Nb Alloys

Chemical composition (wt. %)
Identification
no. C Si Mn P S Cr Ni Nb Mo

4 0.35 0.84 0.79 0.013 0.008 25.05 35.70 0.89 0.082

7 0.44 0.76 0.60 0.019 0.009 24.31 34.28 0.93 0.080
3 0.56 0.88 1.02 0.013 0.003 24.83 35.43 1.05 0.080
11 0.62 1.29 0.89 0.019 0.006 24.71 34.95 0.93 0.090

lieved by a restricted degree of sliding, but the shoulders adjacent to the gauge section
where a continuous path is avoided. The of the specimen.
condition where the carbides are relatively
uniform and small seems to lead to the best
combination of properties [3]. EXPERIMENTAL RESULTS
In the present study, the structure and
high-temperature mechanical properties of MICROSTRUCTURES
four experimental HP-Nb alloys, Fe-35Ni-
To assist in the phase identification, the
25Cr-1.0Nb with 0.35-0.62 wt.%C, were de-
second-phase particles (carbides) in the as-
termined. The alloys were tested in both the
cast alloys were electrolyticallyextracted and
as-cast condition and after aging at 950°C
subjected to x-ray powder diffraction analy-
or 1100°C.
sis. Figure I shows a typical x-ray diffraction
pattern for these extracted carbides. In all
four alloys, the carbides were a mixture of
EXPERIMENTAL DETAILS
eutectic M7C3 [(Cr,Fe)7C3] and eutectic NbC.
Light micrographs of all four alloys in the
The materials were received in the form of
as-cast state are shown in Fig. 2. The "skele-
centrifugally cast tubes. The chemical com-
tal" carbide is eutectic Cr7C3, and the
position of the alloys are given in Table 1.
"blocky" carbide is NbC. Each carbide has
The main difference of the four alloys is that
its "own" eutectic area (i.e., they tend to con-
the carbon content varied from 0.35 to 0.62
centrate in areas that are either CrTC 3 o r
wt.%, while the Nb content remained al-
NbC). The variation in the amount (volume
most constant at 1%. The microstructures
fraction or wt.%) of the eutectic carbides is
were determined in the as-cast condition and
a function of the carbon content of the as-
after aging at either 950°C for 100, 200, 500,
cast alloys (Fig. 3 and Table 2).
and 1000 h or 1100°C for 10, 100, and 200 h.
The results, as illustrated in Figs. 2 and 3,
Standard sectioning and polishing tech-
indicated that as the carbon content was in-
niques were used for the preparation of the
metallurgical specimens. The electrolytic
etching solution used to reveal the micro-
structure was 20 mL of a 40%HNO3 solu-
tion and 5 mL HF.A voltage of 8-12 Vdc was
used for 15-20 s. A grid measurement
method was used to determine the quanti-
ties of Cr7C3, Cr23C6, and NbC. The number
of image helds per specimen was 20. lO0~J~ 0o
The creep and creep-rupture tests were
conducted using a uniaxial, constant load ~, 2e
tensile machine. The amount of creep was FIG. 1. Typicalx-ray diffractionspectra from electro-
measured with an extensometer clamped to lyticallyextractedcarbide particles fromas-cast alloys.
Nb-Containing Type HP Heat-Resistant Steel 5

ii!iiiiii ili ii!iii!i!ilii!i !iii!ii'i!

? !!!il~I ! i . . . . . . . . . . C i~!~

FIG. 2. Optical micrographs of the four alloys in the as-cast state.

creased, the amount of eutectic carbide also examined after aging at 1100°C and the
increased. The main part of this increase was change in microstructures with aging time
eutectic Cr7C3, while the NbC eutectic con- is shown in Figs. 4 and 5. The eutectic NbC
tent remained almost unchanged. Nb is a has good stability, and its appearance and
stronger carbide-forming element than Cr. amount are almost unchanged upon aging.
When the eutectic reaction occurs, Nb com- The eutectic Cr7C3 coarsened faster in alloy
bines with C much easier than Cr does. 3 than in alloy 7 because alloy 3 has a higher
Thus, when the carbon content is low (i.e., carbon content. The secondary carbide par-
carbon content is 0.35 wt.%), eutectic NbC tides, Cr23C6, precipitate first among the
is the main component of the eutectic car- eutectic carbides, then throughout the whole
bides. After the carbon content is high grain. As the carbon content is increased,
enough to use up the 1% Nb forming eutec- there is more Cr23C6 formed that shows a
tic NbC, then as the carbon content is further faster growth (coarsening) rate, and, hence,
increased, there is increased precipitation the particles are larger. After 200 h aging,
of eutectic Cr7C3, which becomes the main the eutectic Cr7C3 has slowly changed to a
component of the eutectic carbide. "broken chain" morphology, and the extent of
The microstructures of alloys 3 and 7 were coarsening of Cr23C6 is more readily visible.
 F. Wang and D. O. Northwood

14
I I I ' I Table 2 Eutectic Carbide Content of As-Cast
Alloys

12 ¸ Carbon Carbide content (wt. %)

O Alloy content
no. (wt.%) CrTC3 NbC (Cr7C3 + NbC)
" " 10
4 0.35 1.50 0.74 2.24
7 0.44 2.00 0.78 2.78
o ~ " ~ 3
3 0.56 2.80 0.80 3.60
11 0.62 3.20 0.80 4.00

The creep-rupture ductilities of the four

alloys at 870 and 950°C under varying ap-
plied stress are summarized in Table 3. Ex-
I ! I I
amination of Table 3 shows that, in general,
0.3 0.4 0.5 0.6 0.7 0.8 the creep-rupture ductilities are highest for
~Cortent:~% carbon contents between 0.45 and 0.56
wt.%C.
FIG. 3. Variation of the volume fraction, Vv% , of the
eutectic carbides, Cr7C3 and NbC, with the carbon con-
The change of creep rate with time for an
tent of the as-cast alloys. applied stress of 70 MPa and a temperature
of 950°C is shown in Fig. 9 for alloys 3, 4,
and 11. Alloy 11 exhibits the slowest creep
MECHANICAL PROPERTIES
rate, and alloy 4 has the highest creep rate.
The creep-rupture properties of the four al- The general trend is that as the carbon con-
loys are shown in Figs. 6 and 7 for test tem- tent is increased, the creep rate is decreased.
peratures of 870 and 950°C, respectively. To Vickers hardness was determined for the
compare the present results for alloys 3 and as-cast alloys and after aging at 1100°C for
7 (i.e., the intermediate carbon content 10-200 h; the values are given in Table 4. A
alloys) with those of the same type (i.e., trend to be noted is that in the as-cast state,
HP-Nb) produced in Japan [4], a Larson- an increase in carbon content and, hence,
Miller plot was used as shown in Fig. 8. The in the amount of eutectic carbides, gives a
points for alloys 3 and 7 are all on the upper higher hardness. Aging of the as-cast alloys,
side of the experimental data bands for the and the precipitation of secondary carbides
Japanese alloys. (Cr23C6), increases the hardness with the

FIG. 4. Optical micrographs of alloy 3 aged at 1100°C for 10, 100, and 200 h.
Nb-Containing Type HP Heat-Resistant Steel 7

FIG. 5. Optical micrographs of alloy 7 aged at 1100°C for 10, 100, and 200 h.

-_.__~__.___.__. ~ 871°C
1100

""
o
50
30 I
f!,! I I I I i IIIIII I i I I 1 I I JIIl I

10 100 1000
FIG. 6. Cree p rupture strength at 870°C for al-
loys 4, 7, 3, and 11. Time : H o w

L 950o C

30 lOI
o Illll
7 ' I t I I I I,,I I t I I I ~ i i i
FIG. 7. Creep rupture strengths at 950°C for 4 10 100 1000
alloys 4, 7, 3, and 11. Time : H o w

maximum hardness being produced after

about 100 h at 1100°C. This increase in hard-
ness caused by the precipitation of second-
ary carbides is higher for alloy 3 (0.56% C)
than for alloy 7 (0.44% C).
=""50 \ \ . N) ~
° )-x
H P - I Ib --
DISCUSSION OF RESULTS 30 ~ -~
\
FCC alloys can be hardened by the addition 27 28 29 30 31 32 33
of carbon in relatively large amounts (0.5 P = 1G'T'(23÷logt)
wt.%) to give a general carbide precipitation.
FIG. 8. Larson-Miller plots for alloys 3 and 7; the re-
Carbon may also promote the formation of sults are compared to those of Japanese researchers [4]
grain boundary carbides such a s M23C 6 and for similar HP-Nb alloys.
 F. Wang and D. (9. Northwood

Table 3 Creep-Rupture Ductilities of Four Alloys at 870°C and 950°C

Creep-rupture ductility (%)

870°C 950°C
Alloy
no. 70 MPa 80 MPa 90 MPa 60 MPa 70 MPa 80 MPa

4 6.5 11.7 12.9 14.6 10.6 14.0

7 50.8 21.8 17.3 25.7 17.9 25.5
3 32.7 19.8 19.4 16.0 19.7 25.5
11 21.0 22.2 7.3 13.0 16.2 14.1

M7C3 to provide strength in these regions. With respect to the carbon content of the
S o m e alloys also contain the strong p r i m a r y austenite, Tomoyoshi [4] has given the fol-
carbide forming element Nb. This element lowing equation:
produces, in addition to M7C3 and M6C, an
MC phase that has greater high-temperature C = _610___O0+ 4.21 - 0.0186(Cr + Ni),
T
stability. The precipitation of p r i m a r y NbC
carbide occurs o n matrix dislocations in where:
some iron-containing alloys [5].
C is the carbon content in austenite (wt. %)
W h e n the N b content is hxed at about 1
wt.%, the carbon content only influences the T is the t e m p e r a t u r e (K)
quantity of Cr7C3, but not NbC (Table 2). Cr a n d N i are the weight percent of these
This is because the Nb solubility in austen- two elements
ite is - 0 . 5 wt.%. The remaining 0.5 wt.%
From the previous equation, the carbon
Nb requires 0.065 wt. % carbon to form NbC.
content in the austenite is calculated at 950°C
In our alloys, the carbon content is m u c h
as 0.013 wt. %. T h r o u g h calculation, the ex-
greater t h a n 0.065 wt.%; thus, the quantity
tent of carbon supersaturation in the austen-
of N b C is controlled by the Nb content. Be-
ite is determined. Alloy 3 shows that from
cause the Nb content of all four alloys was
the grid m e a s u r e m e n t , the a m o u n t s of eu-
about 1 wt.%, the a m o u n t of NbC was al-
tectic Cr7C3 and eutectic NbC are 2.8 wt.%
most the same for the four alloys.
a n d 0.8 wt.%, and the carbon content of the
eutectic carbides is 0.35 wt.%. The remain-
ing carbon content is thus 0.56-0.34 = 0.22
9 6 0 * C p 7 0 MPa wt.%. Therefore, the extent of carbon super-

°l~N
a saturation in the austenite in alloy 3 is 0.22
4
s Table 4 Vickers Hardness of As-Cast Alloys
and Selected Alloys Aged at
1100°C
.¢,0

! Alloy
no.
Carbon
(wt.%) As-cast
Vickers hardness

10 h
Aged at 1100°C

100 h 200 h
0 2 4 6 II 10 tl 14 16 111 | 0
4 0.35 182 - - -
Time : How
7 0.44 183 212 245 236
FIG. 9. Creep rate-time plots for alloys 4, 3, and 11. 3 0.56 187 221 265 240
Creep properties determined at 950°C at an applied 11 0.62 192 - - -
stress of 70 MPa.
Nb-Containing Type HP Heat-Resistant Steel

wt. %. Similarly, the extent of carbon super- From Figures 6 and 7, it can be seen that
saturation in the austenite in alloys 4, 7, and alloys 3 and 7 have higher creep-rupture
11 is 0.14 wt.%, 0.17 wt.%, and 0.24 wt.%, strengths than alloys 4 and 11 (the creep-
respectively. Thus, the higher the carbon rupture strengths are similar for alloys 3 and
content, the more carbon is in supersatu- 7). In explaining these observations, it is
rated solution in the austenite, and the more noted that alloy 4 had the lowest carbon con-
secondary particles precipitate upon aging. tent. Thus, both the matrix and the grain
The carbon content of the austenite as cal- boundaries contained the least amount of
culated previously for the as-cast alloys is carbides. The strength was low, and defor-
much higher than that obtained from the mation was relatively easy in the matrix; dis-
Tomoyoshi [4] equation because the Nishino locations glide easily to the grain boundaries
equation calculates the carbon in solution where cracking can occur. Alloy 11 had the
following an infinitely slow-cooling (equilib- highest carbon content, and both the matrix
rium) process. However, alloys 4, 7, 3, and and grain boundaries had a higher strength.
11 had undergone a nonequilibrium process. However, the ductility in the matrix was rela-
Thus, there are large amounts of carbon left tively low (see Table 3). Relaxation of the
in the austenite matrix after the fast cooling, stresses built up during deformation became
and this will precipitate out in the aging pro- difficult, and premature failure occurred. Al-
cess. Another reason for the differences is loys 3 and 7 had a moderate carbon content;
that the Tomoyoshi [4] equation did not con- the grain boundaries had a high enough
sider the effect of Nb on carbon solubility strength to prevent cavity formation, and the
in austenite. matrix had su~cient ductility to impede
Examination of the hardnesses of alloys crack growth. Thus, alloys 3 and 7 had the
3 and 7 in the aged condition shows that the highest creep-rupture strengths.
higher the carbon content, the higher is the Figure 9 shows that the higher the carbon
hardness and the greater the difference content, the lower was the creep rate. As the
when compared with that value for the as- carbon content was increased, there were
cast state. This indicates that during aging, more eutectic carbides at the grain bound-
the higher carbon content alloy (No. 3) pre- aries, and these carbides were effective ob-
cipitates more Cr23C6 than the lower carbon stacles to grain boundary sliding.
content alloy (No. 7). This is in agreement
with the above calculations and the results
of Morris and Harris [6] and Miller and Pik-
CONCLUSIONS
ington [7].
Figures 4 and 5 show that carbon influ-
The main conclusions to come from this
ences the stability of the aged structure. At
study are as follows:
1100°C and 200 h aging, eutectic Cr7C3 and
eutectic NbC are generally stable. The sec- • The as-cast structure of the as-cast HP-Nb
ondary particles of Cr23C6 particles in alloy alloys consisted of "skeletal" eutectic Cr7C3
3 precipitate faster than those in alloy 7 and carbide and "blocky" eutectic NbC carbide
tend to be evenly distributed. For aging times in a supersaturated austenitic matrix.
of 10-200 h, the Cr23C6 particles in alloy 3 • Increasing the carbon content from 0.35
are larger than those in alloy 7. This indi- to 0.62 wt.% while keeping the Nb content
cates that increasing the carbon content constant at about 1 wt.% increased the
was deleterious to the structural stability amount of eutectic Cr7C3 but not NbC.
of the carbides. But there was no change • The morphology of the eutectic Cr7C 3
in the morphology of the carbides, especially and eutectic NbC did not change with
the NbC carbides. The same tendency was changing carbon content.
visible for aging at 950°C. It merely takes • The optimum creep-rupture properties
a longer time for Cr23C7 to coarsen at the were obtained for carbon contents of 0.45
lower temperature. to 0.56 wt.%C.
10 F. Wang and D. O, Northwood

The experimental work for this article was carried 3. B. Weiss and R. Stickler, Phase instabilities during
out by Ms. Frances Wang when she was a gradu- high temperature exposure of 316 austenitic stain-
less steel, Metallurgical Trans. 3B:851-866 (1972).
ate student in the Materials Engineering Dept.,
4. Nishino Tomoyoshi, Ironand Steel1:12-16 (1972) (in
Dalian Institute of Technology, Dalian, Liaoning, Japanese).
People's Republic of China. The supervision and
5. C. P. Sullivan and M. J. Domachie, Jr., Some effects
guidance of Associate Professors Wang Fugang of microstructure on the mechanical properties of
and Li Dejun is gratefully acknowledged. Fund- nickel-base superalloys, Met. Eng. Q. 7:36-45 (1967).
ing for the continuation of this work is being pro- 6. D. G. Morris and D. R. Harris, Creep and rupture
vided by the Natural Sciences and Engineering in type 316 stainless steel at temperatures between
Research Council of Canada. 525 and 900°C, Part III. Precipitation behaviour, Metal
Science 11:542-549 (1978).
7. D. A. Miller and R. Pikington, The effect of tem-
perature and carbon content on the cavitation be-
References haviour of a 1.5 pct Cr-0.5 pct V steel, Metallurgical
Trans. 9A:489-493 (1978).
1. Sasaki Yoshikazu, Researchon Metal Materials, Vol.
3, No. 6 (1980) (in Japanese).
2. MetallurgicalResearchof High TemperatureAlloys, 1st
ed., Metallurgy Publishing House, Beijing, China
(1984) (in Chinese). Received July 1992; acceptedFebruary 1993.