Professional Documents
Culture Documents
MTI Interior 1 PDF
MTI Interior 1 PDF
AF234/7226/MSch
Prepared by:
Ellen Berghof-Hasselbcher 1)
Peter Gawenda 1)
Monika Schorr 1)
Michael Schtze 1)
John J. Hoffman 2)
The results, conclusions and recommendations given in this report refer to the specimens and data submitted as well as to the system
conditions mentioned. Therefore, the results cannot be applied to other conditions. DECHEMA e.V. accepts no liability.
It is only permitted to reproduce this report unabridgedly. Publication of the report or of single experimental results are subject to
approval by DECHEMA e.V.
Atlas of Microstructures
ii
Atlas of Microstructures
iii
Table of Contents
Introduction . .......................................................................................................................... xx
Description of Test Specimens................................................................................................. xx
Creep rupture test specimens........................................................................................................ xx
-
Service-exposed specimens............................................................................................................ xx
- History of Service Exposed Samples
- Exposure Conditions of .................................................................... Table 7
- Chemistry of Service Exposed Samples............................................... Table 8
Alloy HPMA.............................................................................................................................. xx
- Results of image analysis of Alloy HPMA..............................Table 10
- As Cast............................................................................................... Fig: 2-1a to 2-1j
- Creep Rupture Samples..................................................................... Fig. 2-2a to 2-2r
..................................................................... Fig. 2-3a to 2-3r
..................................................................... Fig. 2-4a to 2-4r
Alloy HP Modified...................................................................................................................... xx
- Results of image analysis of Alloy HP Modified.................................. Table 11
- As Cast . ............................................................................................ Fig. 3-1a to 3-1j
- Creep Rupture Samples .................................................................... Fig. 3-2a to 3-2r
..................................................................... Fig. 3-2a to 3-2r
..................................................................... Fig. 3-3a to 3-3r
..................................................................... Fig. 3-4a to 3-4r
..................................................................... Fig. 3-5a to 3-5r
..................................................................... Fig. 3-6a to 3-6r
..................................................................... Fig. 3-7a to 3-7r
vi
Summary................................................................................................................................. xx
Composition of the Precipitates. ................................................................................................... xx
Average Compositions of the Precipitates in atomic-%.......................... Table 13
vii
viii
Introduction
Up to and through the 1940s, furnace tubes were typically fabricated
from wrought chromium steels and/or austenitic stainless steel alloys.
Having rather low carbon concentrations, these alloys had poor creep
strength and limited service lives. To increase tube life, greater carbon
concentrations were required to promote precipitation hardening upon
elevated temperature exposure. The increased carbon concentrations,
however, generally resulted in excessive work hardening during conventional processing of wrought materials. The use of centrifugal casting,
pouring molten metal in a horizontal spinning mold, allowed founders
to develop high carbon alloys as the molten metal solidified into near
final shape without the need for subsequent metal working operations.
Thus, refinement of centrifugal casting processes was the gateway to
further alloy development and more aggressive furnace operations.
The first widely used centrifugal cast alloy for steam-methane
reformer tube applications was HK40 in the 1950s timeframe. HK40 is
essentially the cast equivalent to wrought 310 stainless steel nominally
containing 25 wt% chromium, 20 wt% nickel, with iron as the balance.
However, HK40 nominally contains 0.40 wt% carbon while wrought
310 stainless steel contains only 0.08 wt% carbon. The increased carbon
content and precipitation of primary carbides resulted in HK40 having
greatly improved high temperature strength as compared to wrought 310
stainless steel. In the 1960s, the cast HP alloys (nominally 25 wt% chromium, 35 wt% nickel, 0.50 wt% carbon, with iron as the balance) were
developed to provide greater creep strength as compared to HK40.
The HK and HP alloys rely on precipitation of M23C6 and/or M7C3
carbides (where M is primarily chromium) for elevated temperature
creep strength. The precipitated chromium carbides in the HK and HP
alloys tended to coalesce as exposure temperatures approached 1800oF
(982oC). The carbide coalescence decreased the creep strength and,
therefore, limited the strength of these alloys at elevated temperatures.
Microstructural changes that occur in the HK and HP alloys with
Atlas of Microstructures
severe conditions. Unlike the HK40 and HP50 alloys, there is virtually no published data on the microstructural changes that occur in the
HP-modified, HPMA, or 35Cr/45Ni alloys upon long term aging. In
an effort to address industry need, the Materials Technology Institute
(MTI) sponsored the Atlas of Microstructures project. The specimens
analyzed in the Atlas of Microstructures were foundry stress rupture
specimens generously donated by MetalTek International, Duraloy
Technologies, Kubota Metal Corp., and Manoir Industries. In addition,
service exposed samples, having longer aging times than the foundry
stress rupture specimens, were donated by operating companies including Air Products & Chemicals, Syncrude Canada, Eastman Chemical,
and MetalTek International. The thorough, detailed microstructural
analyses were completed at the Karl-Winnacker Institut der DECHEMA
in Frankfurt, Germany.
The Atlas of Microstructures documents microstructural changes
with increased aging time and temperature, identification and chemical
composition of precipitated phases, as well as diagrams characterizing
the kinetics of phase transformation for the HP-modified, HPMA, and
35Cr/45Ni alloy classes. This MTI Atlas of Microstructures bridges the
gap from the Battelle project from the early 1970s to the most important alloys used in the petrochemical industry today.
Description of the
Test Specimens
Creep rupture test specimens
The specimens that had been taken from creep rupture tests can
be summarized with regards to their exposure conditions and their
chemistry in Tables 1-6. For each of the three materials investigated an
as-delivered (as-cast) specimen as reference sample was available. For
alloy 35Cr/45Ni and alloy HP Modified twelve creep specimens had
been supplied for the investigations. For alloy HPMA sixteen creep specimens existed. The test temperatures varied between 1675 and 2100F
(913 and 1149C), the maximum exposure times reached 12,289 h.
The stress range was between 1.0 and 8.5 ksi (6.9 and 58.8 MPa). The
scatter of the chemical compositions was in the allowed range for these
materials.
continued on page 8
Atlas of Microstructures
Test-Temp F
as cast
1800
1800
1922
1900
1900
1900
2012
2000
2012
2057
2050
2100
Atlas of Microstructures
Test-Temp C
as cast
982
982
1050
1038
1038
1038
1100
1093
1100
1125
1121
1149
Mn
0.60
1.32
1.46
1.02
1.40
1.04
1.31
1.02
1.37
1.46
1.02
1.43
1.46
Stress Ksi
Stress MPa
3.30
2.90
3.40
2.09
1.60
1.54
2.00
1.80
1.15
1.60
1.00
1.50
22.8
20.0
23.4
14.4
11.0
10.6
13.8
12.4
7.9
11.1
6.9
10.3
1,509
7,217
367
2,023
6,331
10,247
335
1,297
3,545
277
2,606
248
Composition (wt%)
Si
C
1.55
0.38
1.83
0.46
1.86
0.43
1.52
0.41
1.85
0.42
1.30
0.37
1.78
0.42
1.52
0.41
1.71
0.42
1.86
0.43
1.52
0.41
1.82
0.42
1.86
0.43
Nb
0.91
0.78
0.83
1.21
1.25
1.22
1.20
1.21
1.14
0.83
1.21
1.25
0.83
N
0.080
0.060
0.040
0.065
NR
NR
NR
0.065
NR
0.040
0.065
NR
0.040
Other
Ti, Zr & W
Ti, Zr & W
Ti, Zr & W
Ti
Ti
Ti
Ti
Ti, Zr & W
Ti
Ti, Zr & W
Test-Temp. F
as cast
1700
1700
1700
1750
1800
1800
1800
1832
1922
1950
1922
1922
2012
2012
2012
2012
Test-Temp. C
as cast
927
927
927
954
982
982
982
1000
1050
1066
1050
1050
1100
1100
1100
1100
Stress Ksi
Stress MPa
8.00
6.00
5.30
4.45
5.97
3.51
3.90
3.00
3.48
2.80
2.32
2.08
2.39
1.80
1.45
1.35
55.2
41.3
36.6
34.1
41.2
24.2
26.9
20.7
24.0
19.1
16.0
14.3
16.5
12.4
10.0
9.3
183
2,338
6,072
8,359
177
2,436
6,478
12,289
137
2,558
4,864
11,778
105
2,714
5,580
8,990
Atlas of Microstructures
Atlas of Microstructures
Mn
0.65
0.78
1.31
0.65
0.65
0.89
0.79
0.56
0.56
1.37
0.91
0.65
0.56
1.37
0.56
0.56
0.65
Composition (wt%)
Si
C
1.08
0.45
0.80
0.50
1.37
0.42
1.08
0.45
1.08
0.45
1.12
0.47
0.85
0.46
1.14
0.44
1.14
0.44
1.73
0.46
1.33
0.48
1.08
0.45
1.14
0.44
1.73
0.46
1.14
0.44
1.14
0.44
1.08
0.45
Nb
0.65
0.49
0.40
0.65
0.65
0.95
1.08
0.51
0.51
0.83
0.65
0.51
0.83
0.51
0.51
0.65
N
0.06
NR
NR
0.06
0.06
NR
NR
0.05
0.05
NR
NR
0.06
0.05
NR
0.05
0.05
0.06
Other
Ti, Zr & W
Ti
W
Ti, Zr & W
Ti, Zr & W
Ti & W
Ti, Zr & W
Ti, Zr & W
Ti & Zr
Ti & W
Ti, Zr & W
Ti, Zr & W
Ti & Zr
Ti, Zr & W
Ti, Zr & W
Ti, Zr & W
Test-Temp. C
as cast
913
927
955
970
970
982
982
982
982
1010
1038
1038
Stress Ksi
Stress MPa
6.50
8.50
5.10
5.15
3.77
5.40
5.42
5.52
2.80
3.60
2.30
1.90
44.6
58.8
35.2
25.5
26.0
37.4
37.4
37.4
19.3
24.8
15.8
13.1
659
59
794
286
1,185
191
4,467
7,833
10,637
707
2,555
5,373
Atlas of Microstructures
Atlas of Microstructures
Mn
1.29
0.74
0.72
0.74
1.34
1.34
0.84
0.90
0.83
0.72
0.94
0.72
0.72
Composition (wt%)
Si
C
1.40
0.44
1.67
0.43
1.23
0.42
1.67
0.43
1.79
0.45
1.79
0.45
1.29
0.41
1.37
0.41
1.35
0.42
1.23
0.42
1.71
0.48
1.23
0.42
1.23
0.42
Nb
1.06
0.75
0.74
0.75
0.83
0.83
1.18
1.22
1.20
0.74
0.82
0.74
0.74
N
NR
0.06
0.06
0.06
NR
NR
NR
NR
NR
0.06
0.06
0.06
0.06
Other
W & Zr
W & Zr
W & Zr
W
W
W
W & Zr
W
W & Zr
W & Zr
Metallographic
Preparation
From the specimens available sections were taken for metallographic
preparation with the surfaces to be investigated oriented parallel to the
longitudinal direction of the tensile creep rupture specimens or perpendicular to the longitudinal axis of the tube specimens. Cutting was
performed by using a precision sectioning machine with direct water
cooling of the specimen. Some of the specimens had been delivered
in the embedded state from which smaller sections were taken by saw
cutting which were embedded again. Hot embedding in a conductive
epoxy resin was used with a diameter of the moulds of 25 mm and after
this the specimens were ground on SiC papers of the grades 180, 220,
320, 500, 1000 down to grit 2400 (Struers Standard 43-GB-1984, DIN
69176, Part 1,2,4) with a pressure of 70-80 N and water as a lubricant.
Fine polishing was performed with a two step diamond polish of 3 m
and 1 m followed by the finalizing polish with SiO2 suspension of
size 0.02 m. The area on the specimen surface to be investigated was
marked by four Vickers hardness imprints (HV1) with a distance of
500 m. Usually these areas were marked in the center of the specimen
(in some cases, however, in addition regions close to the surface were
investigated after marking).
Investigation Techniques
Differential Interference Contrast Technique (LM-DIC)
For documentation of the microstructure in the light microscope
at high magnification (500x and 1000x) the differential interference
contrast (LM-DIC) technique2 was used by which the different phases
due to the different hardness reveal certain topographical structures
which facilitates distinction of these phases.
Atlas of Microstructures
Atlas of Microstructures
The matrix was not binarized and was used as a background in true
colors behind the false color image.
Table 9: Results of image analysis of Alloy 35Cr45Ni Creep Rupture Stress Samples
Sample_ID Sample
Temp. Temp F Stress Stress Expo.
Time
core
C
F
MPa
Ksi
hrs M(C,N) M7C3 M23C6 M6C
A234_003 35/45-1 (c) as cast as cast
0.8
3.9
A234_004
35/45-2
982
1800
22.8
3.30 1,509
0.5
7.9
A234_005
35/45-3
982
1800
20.0
2.90 7,217
0.3
7.8
A234_006
35/45-4
1050
1922
23.4
3.40 367
0.4
6.9
A234_007
35/45-5
1038
1900
14.4 2.090 2,023 < 0.05
8.3 < 0.03
A234_008
35/45-6
1038
1900
11.0
1.60 6,331 < 0.05
10.5
7.5
A234_009
35/45-7
1038
1900
10.6
1.54 10,247
0.1
15.1
10.7
A234_010
35/45-8
1100
2012
13.8
2.00 335
1.1
7.6
A234_011
35/45-9
1093
2000
12.4
1.80 1,297
0.4
10.8
0.3
A234_012 35/45-10
1100
2012
7.9
1.15 3,545
0.4
12.0
3.8
A234_013 35/45-11
1125
2057
11.1
1.60 277
1.1
7.2
A234_014 35/45-12
1121
2050
.06.9
1.00 2,606
0.2
12.7
9.2
A234_015 35/45-13
1149
2100
10.3
1.50 248
0.6
8.6
<0.3
Annotations: + few, ++ many, +++ very many
10
area fraction in %
M2(C,N) G-Phase M(C,N)
1.5
1.8
0.7
3.0
1.2
0.8
0.3
0.3
< 0.02
< 0.05
1.1
0.2
0.6
0.2
5.9
6.6
7.4
8.8
+
+
7.8
10.2
8.9
7
0.6
9.9
6.3
12.5
4.2
13.1
+++
+++
6.8
10.8
8.7
4.5
+++
7.6
M2(C,N)
+
+
+
++
++
+
++
++
+
++
++
Atlas of Microstructures
Results of the
Microstructural Analyses
colored purple and the M(C,N) were colored blue, Figure 1-1j.
Alloy 35Cr45Ni
The results from the image-analytical determination of the fractions of the precipitates in the microstructure are summarized in Table
9 (previous page) for all specimens investigated.
As-Cast
In the as-cast condition the microstructure consists of an austenitic matrix with a eutectic mixture of carbides of the types M7C3 and
M(C,N) which is situated between the dendrites and along the grain
boundaries respectively. The fraction of M7C3 amounts to about 4%,
while that of M(C,N) lies at about 0.8%. The M7C3 carbides are rich
in chromium, while the M(C,N) carbides are rich in niobium and/or
titanium, Figures 1-1a to 1-1j. In the DIC representation the M7C3
carbides appear light gray and the M(C,N) have a reddish appearance
and are slightly elevated above the matrix due to their higher hardness,
Figures 1-1a and 1-1b. In the BSE image the Nb-rich M(C,N) particles
appear significantly brighter than the Cr-rich M7C3 particles and the
Ti-rich M(C,N) which are almost black, Figures 1-1e and 1-f. After
coating the specimen surface with ZnSe the MC particles containing
high amounts of Nb appear in blue color, Figures 1-1g and 1-h, while
the Ti-rich M(C,N) which could be detected in the DIC image by their
strong intrinsic orange color were colored in light blue to white. The
austenitic matrix and the M7C3 particles both have a reddish color,
Figures 1-1g and 1-1h; however, the carbides can be distinguished quite
clearly from the matrix as they are significantly darker. After etching
the carbides, which have a dark seam, can be clearly distinguished from
the matrix, Figure 1-1i. In this figure it becomes evident that these have
been precipitated in regions of interdendritic residual melts or at grain
boundaries, respectively. In the false color image the M7C3 particles were
Atlas of Microstructures
11
structure, Figure 1-3a, which is also true for the secondary precipitates
of M23C6, Figures 1-3a to 1-3j. However, the volume fraction of the
primary and secondary M23C6 carbides remained constant with 7.8%.
The fraction of the G-phase has slightly increased with 1.8%. One striking feature is the bright, high Nb-containing dot-like inclusions in the
G-phase which are residues of dissolved M(C,N), Figures 1-3c to 13h. The fraction of M(C,N) has decreased from 0.5 to 0.3% due to
increased formation of G-phase. In the metal sub-surface zone M6(C,N)
can replace the G-phase with a volume fraction of 12.5% and is most
frequently directly in contact with the M23C6 carbide, Figures 1-3k to 13r. The composition of this phase can be taken from the table in Figure
1-3. In the DIC image this phase appears gray, while in the BSE image
it can scarcely be distinguished from the matrix, Figures 1-3m and 1-3n.
With increasing amounts of Ni this phase becomes brighter in the BSE
image. In the ZnSe-coated condition the phase appears violet, Figures
1-3o and 1-3p, and in the false color image it had been assigned a red
color, Figure 1-3r. The phase is resistant to V2A etchant, Figure 1-3q.
The amount of M23C6 is slightly lower in the sub-surface area with 6.6%
compared to the core area with 7.8%. In the outermost zone due to
uptake of nitrogen the phase M2(C,N) appears which is slightly elevated
in the metallographic section in the DIC image due to its higher hardness (not documented here).
1900F/1038C
1922F/367 h/3.4 ksi (1050C/367 h/23.4 MPa)
Pores and cracks from the creep process can be clearly seen in the
microstructure, Figure 1-4a. In the core of the specimen the phases
M23C6, G-phase and M6C could be identified by EPMA. In the light
microscopic investigations and BSE image furthermore the phase
M(C,N) could be found, Figures 1-4a to 1-4j. Since the G-phase and
M6C could not easily be distinguished from each other in the contrasted
condition both phases were taken together in the image analysis procedure. The fractions given for these two phases in the table of Figure 1-4
12
Atlas of Microstructures
from M6C, Figures 1-6u and 1-6v. After contrasting the phases by ZnSe
M2(C,N) has a violet color similar to M6C with, however, a slightly
brighter tone, Figure 1-6y. It was, however, not possible to distinguish
between the two phases in the image analytical process due to an overlap
in the color distributions in the red and blue range. In the false color
image the M2(C,N) was colored green, Figure 1-6z.
phase, while 0.4% are M(C,N). The total fraction of these two phases
was determined to be 1.6%. In the false color image the phases were
partially re-drawn manually, Figure 1-9j. In the sub-surface zone again
higher amounts of M6C (10.8%) are found beside M23C6 (10.3%),
while the G-phase is missing, Figures 1-9k to 1-9r. The phase M2(C,N)
is only found in the outermost area again, Figures 1-9s to 1-9z.
2000F/1093C
2012F/335 h/2.0 ksi (1100C/335 h/13.8 MPa)
After 335 h of exposure at 2012F (1100C) the microstructure in
the core reveals about 7.6% M23C6 (primary and secondary carbides)
and about 1.1% M(C,N), Figures 1-8h to 1-8j. Again a large number of
creep pores and cracks are observed. In the sub-surface zone furthermore
about 6.8% of M6C carbides are present which frequently contain residues of the Nb-rich M(C,N) in their center, Figures 1-8k to 1-8r.
2000F/1,297 h/1.8 ksi (1093C/1297 h/12.4 MPa)
With increasing exposure time under creep conditions the pores have
grown significantly, Figure 1- 9a. The M23C6 carbides have also grown
in their size with the volume fraction in the core area of the specimen
increasing up to 10.8%, Figures 1-9a and 1-9b. Besides a small fraction (0.3%) of M6C the microstructure also reveals the presence of the
G-phase and of M(C,N). Due to an overlap in the colors the fractions
could only be determined as a sum on the ZnSe-coated metallographic
section, Figures 1-9g and 1-9h. It is estimated that about 1.2% are G-
Atlas of Microstructures
2050F/1121C
2057F/277 h/1.60 ksi (1125C/277 h/11.1 MPa)
After 277 h of exposure at the further increased temperature of
2057F (1125C) several rows of creep pores oriented perpendicular to
the tensile direction are observed, Figure 1-11a. The microstructure in
the core of the specimen contains 7.2% of carbides of the type M23C6
(primary and secondary carbides) and 1.1% of the type M(C,N), Figures
1-11c to 1-11j. Furthermore in the sub-surface zone, Figures 1-11k to
1-11r, 4.5% of M6C are observed while the fraction of M(C,N) has
significantly decreased (0.2%). M2(C,N) is observed only locally in the
outermost sub-surface area.
2050F/2,606 h/1.00 ksi (1121C/2606 h/6.9 MPa)
Compared to the shorter exposure time coarsening and an increase
in the number of M23C6 carbide to a fraction of 12.7% has taken place
13
in the core area of the specimen with a simultaneous decrease of the fine
secondary carbides, Figures 1-12a to 1-12j. The formation of a significant fraction of carbides of the type M6C (9.2%) is associated with a
dissolution of the carbides of the type M(C,N) whose fraction amounts
only to 0.2%. In the metal sub-surface zone the carbide of the type
M2(C,N) is found in a significant amount, Figures 1-12k to 1-12z.
2100F/248 h/1.50 ksi (1149C/248 h/10.3 MPa)
After brief exposure at the highest temperature of 2100F (1149C)
a fraction of 8.6% of M23C6 primary and secondary carbides as well as
of about 0.6% of M(C,N) is observed after 248 h with a small fraction
of M6C (< 0.3%), Figures 1-13a to 1-13j. In the sub-surface zone the
amount of M6C carbides is significantly higher (7.6%) with also the
amount of M23C6 having increased (9.9%). Furthermore in the outermost sub-surface zone which had been strongly oxidized and depleted in
chromium M2(C,N) is also observed, Figures 1-13s to 1-13v.
Conclusions
The results of these investigations can be summarized as follows:
The centrifugally cast material 35Cr/45Ni in the as-cast condition
consists of an austenitic matrix with eutectic areas of chromium-rich
carbides of the type M7C3 and of Nb or Ti-rich particles of the type
M(C,N) which have precipitated between the grains or along the grain
boundaries respectively. Already after 200-300 h of exposure in the
temperature range of 1800-2100F (982 - 1149C) a complete conversion of the carbides of the type M7C3 into chromium-rich carbides of the
type M23C6 takes place with the latter type of carbides forming secondary precipitates in addition. With increasing exposure time the number
of these secondary carbides increases and also the fraction of the primary
M23C6 carbides increases, accompanied by significant growth. At the
same time additional M6C and/or Gphase are formed. Formation of the
latter phases initiates at the phase boundary between M23C6 and matrix.
M6C is rich in Cr, Ni and Nb, while the G-phase (Ni16Nb6Si7) contains
14
less Cr, but is rich in Ni, Si and Nb. Frequently the G-phase shows
a seam of small round M23C6 secondary carbides. M6(C,N) shows a
stronger tendency of formation in the metal sub-surface zone compared
to the core area which may be explained by depletion processes due
to outward diffusion of oxide-forming elements (e.g. Cr, Si) or by the
uptake of nitrogen. The appearance of M6C seems to be stimulated at
higher temperatures, i.e. it appears after shorter times. Furthermore the
fraction increases with increasing exposure time. The G-phase is only
found in the temperature range of 1800-2000F (982 - 1149C). In the
lower part of this temperature range (1800F (982C)) this phase is still
observed after 7,217 h, while at 2000F (1093 C) it has disappeared
after 6,331 h. At 1900F (1038C) it can be found after 1,300 h, while
after 3,545 h it no longer exists and seems to be replaced by M6C. At
2012F (1100C) this phase is not observed even after 335 h. In this
case the lower Si content (1.52 mass-%) may have played a role9. An
increased amount of silicon (> 1.6% mass-%) promotes the formation
of G-phase by facilitating the conversion of NbC into G-phase. In the
G-phase as well as in the M6C particles enrichment of Nb is observed
which directs towards the dissolution of former MC10-15. The dissolution
of MC with increasing exposure time is confirmed by a decrease of their
volume fraction. In the cast state the fraction amounts to about 0.8%,
while after short exposure times in which no G-phase has formed yet it
amounts to 0.6-1.1%. When G-phase formation or after longer times
G-phase and M6C formation start the fraction goes down to less than
0.02-0.4%. After several thousand hours above 1900F (1038C) the
secondary carbides (M23C6) dissolve and form blocky, elongated particles. Besides these coarse carbides of the type M23C6 where the particles
sometimes also have grown together only coarse carbides of the type
M6C and in the sub-surface zone of the type M2(C,N) are present. The
G-phase is no longer found.
Atlas of Microstructures
Table 10: Results of image analysis of Alloy HPMA Creep Rupture Stress Samples
Sample_ ID
A234_018
A234_019
A234_020
A234_021
A234_022
A234_023
A234_024
A234_025
A234_026
A234_027
A234_028
A234_029
A234_030
A234_031
A234_032
A234_033
A234_034
Sample
HPMA-1 (c)
HPMA-2
HPMA-3
HPMA-4
HPMA-5
HPMA-6
HPMA-7
HPMA-8
HPMA-9
HPMA-10
HPMA-11
HPMA-12
HPMA-13
HPMA-14
HPMA-15
HPMA-16
HPMA-17
Temp.
C
as cast
927
927
927
954
982
982
982
1000
1050
1066
1050
1050
1100
1100
1100
1100
Temp F
F
as cast
1700
1700
1700
1750
1800
1800
1800
1832
1922
1950
1922
1922
2012
2012
2012
2012
Stress
Stress
MPa
Ksi
Expo.
Time
hrs
55.2
41.3
36.6
30.7
41.2
24.2
26.9
20.7
24.0
19.1
16.0
14.3
16.5
12.4
10.0
9.3
8.00
6.00
5.30
4.45
5.97
3.51
3.90
3.00
3.48
2.80
2.32
2.08
2.39
1.80
1.45
1.35
183
2,338
6,072
8,359
177
2,436
6,478
12,289
137
2,558
4,864
11,778
105
2,714
5,580
8,990
area fraction in %
M(C,N)
0.9
0.6
<0.3
0.8
0.8
0.6
1.2
0.5
0.8
0.7
1.0
1.1
0.8
0.8
0.8
1.0
1.1
M7C3
3.4
core
M23C6
M6C
7.0
8.3
8.0
8.7
7.0
7.9
8.1
9.0
9.5
8.8
9.7
13.1
7.7
9.6
11.1
10.9
M2(C,N) G-Phase
0.8
0.7
0.5
1.3
<0.2
1.7
M(C,N)
+
+
0.8
+
0.4
0.4
0.5
+
0.4
0.9
+
0.8
+
+
+
+
9.8
5.1
15.2
+
Atlas of Microstructures
15
Alloy HPMA
The results of the image analytical determination of the volume
fractions of all specimens of this material are summarized in Table 10
(previous page).
As-Cast
Figures 2-1a to 2-1j show the as-cast condition after application of
the different imaging techniques. In the austenitic matrix two different
types of precipitates can be distinguished which are located on the grain
boundaries or in areas of the residual melt. As the EPMA analyses reveal
these precipitates are Nb or Ti-rich carbides MC or carbon nitrides of
the type M(C,N) and Cr-rich carbides of the type M7C3 which can
be distinguished even in the DIC image due to their shape, color and
size, Figures 2-1a to 2-1d. M(C,N) precipitates have different modifications with varying contents of Nb, Ti and Cr. In the BSE image 2-1b
the heavier elements Zr and Nb appear brighter than the lighter ones.
Therefore the Ti-rich core of M(C,N) appears darker. A better contrast
is achieved by ZnSe coating, Figures 2-1g and 2-1h, after which the
M7C3 appear in purple, the Ti-rich cores of the M(C,N) particles in
light blue to white and the highly Nb-containing species in blue. Truecolor imaging allows a reliable identification of the different phases and
a quantification by the binary image, Figure 2-1j. The fractions of M7C3
amount to 3.4% and of M(C,N) to 0.9%.
16
phase containing Cr, Ni, Si and Nb this creep specimen exhibits the
phase M6C which also contains some Fe. This phase appears as light gray
in the DIC image and is partially encapsulated by M23C6, Figures 2-2c
and 2-2d. Frequently in the M6C particles, which can hardly be distinguished in the BSE image, Figures 2-2e and 2-2f, from the surrounding
matrix bright Zrrich MC areas are also present. Coating of the specimen
with ZnSe facilitates the detection of the M6C phase (which can easily
be overlooked in the BSE image) by changing the color to violet. This is
illustrated by the comparison of Figure 2-2f and Figure 2-2h. In the false
color image, Figure 2-2j, the M23C6 (7.0%) are colored yellow, the M6C
(0.8%) are colored red and the M(C,N) (0.6%) are colored blue.
1700F/2,338 h/6.00 ksi (927C/2338 h/41.3 MPa)
Even after more than 2,000 h of exposure the microstructure has
not changed significantly, Figures 2-3a to 2-3r. In the core about 8.3%
M23C6, 0.7% M6C and less than 0.3% MC are observed. The decrease in
the amount of MC is probably due to the fact that according to chemical analysis this specimen contains less Nb and no Ti, as well as a lower
amount of C, so that in the end fewer M(C,N) particles were formed.
The sub-surface zone has a similar appearance of the microstructure.
1700F/6,072 h/5.30 ksi (927C/6072 h/36.6 MPa)
Even after 6,000 h of exposure no significant changes in the microstructure can be found. The fraction of M23C6 in the core area amounts
to 8.0%, while the amount of M(C,N) has increased to 0.8% due to
the chemical composition of the specimen, Figures 2-4a to 2-4f. In the
sub-surface zone of the specimen some M6C (< 0.3%) had precipitated,
Figures 2-4k to 2-4r.
1750F/8,359 h/4.45 ksi (954C/8359 h/30.7 MPa)
After more than 8,000 hours at a slightly increased temperature the
amount of M23C6 has increased slightly from 8.0% to 8.7% which is
probably due to an increased precipitation of secondary carbides, Figures
2-5a to 2-5r. The amount of M(C,N) of 0.8% has remained constant
Atlas of Microstructures
Atlas of Microstructures
17
18
Conclusions
The results of the investigations can be summarized as follows:
The centrifugally cast material HP-MA (microalloy) in its cast state
consists of an austenitic matrix with a eutectic structure at interdendritic
sites or along grain boundaries respectively, formed by Crrich carbides
of the type M7C3 and Nb or Ti-rich carbonitrides or carbides of the
type M(C,N). Already after 100-200 h of exposure in the temperature
range of 1700-2012F (927 - 1100C) a complete conversion of the
carbides of the type M7C3 into Cr-rich carbides of the type M23C6 takes
place. The M23C6 carbides form secondary precipitates. With increasing exposure time the number of secondary carbides increases and later
on also the number of primary M23C6 carbides rises accompanied by a
strong growth in size. The fraction increases from about 7% to a maximum of 11-13%. At the same time some M6C is precipitated in the core
Atlas of Microstructures
Atlas of Microstructures
Alloy HP Modified
The results of the image analytical determination of the volume fractions of the precipitates are summarized in Table 11.
As-Cast
Figures 3-1a to 3-1j show the as-cast condition after application of
the different imaging techniques. The microstructure is comparable to
that of the as-cast condition of the material HPMA. The precipitates
along the grain boundaries or in the interdendritic residual melt areas
are the Cr-rich carbides of the type M7C3 (3.9%), which appear bright
gray in the DIC image, and Nb-rich carbides of the type M(C,N) with
0.9% which appear slightly red in the DIC image, Figures 3-1c and
3-1d. In the ZnSe-coated state M7C3 appears purple and M(C,N) blue,
Figures 3-1g and 3-1h.
19
Table 11: Results of image analysis of Alloy HP Modified Creep Rupture Stress Samples
Sample_ID
A234_036
A234_038
A234_039
A234_040
A234_041
A234_042
A234_043
A234_044
A234_045
A234_046
A234_047
A234_048
A234_049
Sample
HPNb-1 (b)
HPNb-2
HPNb-3
HPNb-4
HPNb-5
HPNb-6
HPNb-7
HPNb-8
HPNb-9
HPNb-10
HPNb-11
HPNb-12
HPNb-13
Temp.
C
as cast
913
927
955
970
970
982
982
982
982
1010
1038
1038
Temp F
F
as cast
1675
1700
1750
1778
1778
1800
1800
1800
1800
1850
1900
1900
Stress
Stress
MPa
Ksi
Expo.
Time
hrs
44.6
58.8
35.2
35.5
26.0
37.4
37.4
37.4
19.3
24.8
15.8
13.1
6.5
8.5
5.1
5.2
3.8
5.42
5.42
5.42
2.8
3.6
2.3
1.9
659
59
794
286
1,185
191
4,467
7,833
10,637
707
2,555
5,373
area fraction in %
M(C,N)
0.9
0.3
1.0
0.4
0.5
0.4
3.0
2.2
1.7
0.4
0.7
0.6
0.6
M7C3
3.9
core
M23C6 M6C
8.0
8.0
7.5
6.7
7.1
7.6
6.4
7.1
11.1
7.8
8.4
9.5
0.5
0.7
2.4
1.1
0.7
3.9
0.9
2.5
1.0
M7C3
++
+
+
++
++
+
(+)
(+)
(+)
++
++
++
++
(+)
(+)
+
+
+
+
Annotations: (+) very few, + few, ++ many, +++ very many, ++++ abundant
20
Atlas of Microstructures
guished from the G-phase due to their blocky shape. After ZnSe coating
the M(C,N) particles appear in light blue to white, the G-phase in blue,
the M6C particles in violet and the M23C6 carbides in magenta and are
thus easy to distinguish. The G-phase particles and the M6C particles
had to be measured together in the image analytical investigations since
it was not easy to distinguish between the two of them and to measure
their fractions separately. The fractions of the two phases were estimated
manually. The false color image in Figure 3-2j needed some manual
adjustment and the phases were colored as follows: M23C6 yellow, Gphase magenta, M(C,N) blue and cyan, M6C red. Etching with
V2A etchant improves the visibility of the precipitates in the matrix,
Figure 3-2i. The microstructure in the sub-surface zone is not different
from that in the core, Figures 3-2o to 3-2r.
1700-1750F/913-955C
Atlas of Microstructures
1778F/970C
1778F/286 h/5.15 ksi (970C/286 h/35.5 MPa)
After this relatively short exposure time at 1778F (970C) the core
area contains 6.7% M23C6, 1.1% M6C and 0.5% M(C,N), Figures 3-5a
to 3-5j. Despite the relatively high amount of Si of 1.79 mass-% the
G-phase could not be identified reliably. The metal sub-surface zone
does not show any differences compared to the core area, Figures 3-5k
to 3-5r.
1800F/982C
1800F/191 h/5.42 ksi (982C/191 h/37.4 MPa)
The specimens HP-Nb-7 to HP-Nb-9 possess a high amount of Nb
(1.2 mass-%) which has a significant effect on the microstructure. Besides
M23C6 a large number of M(C,N) are precipitated in very fine structures
and a eutectic appearance, Figures 3-7a to 3-7r. Their volume fraction
amounts to 3% with an amount of 7.6% of M23C6. In the outermost
zone of the specimen some M6C and M2N are encountered.
1800F/4,467 h/5.42 ksi (982C/4467 h/37.4 MPa)
After an exposure time of 4,467 h the microstructure has not changed
compared to specimen HPNb- 7, Figures 3-8a to 3-8r. The volume fractions of the carbides determined in the image analytical investigations
21
22
Conclusions
The results of these investigations can be summarized as follows:
The centrifugally cast material HP-Nb (modified alloy) in its cast
version consists of an austenitic matrix with eutectic structures of the
Cr-rich carbides of the type M7C3 and of Nb and Zr carbides of the
type M(C,N). These structures are found along grain boundaries and in
interdendritic areas. Already after an exposure time of 59-700 h and a
temperature range of 1700-1900F ( 927 -1038C) a complete conversion of the carbides of the type M7C3 into Cr-rich carbides of the type
M23C6 takes place with the latter also forming secondary precipitates.
In addition in specimens with Nb contents between 0.74 and 0.83
mass-% M6C carbides are precipitated. The amount of M6C seems to
increase in the lower temperature range up to 1750F (954C) with
increasing exposure time. At the higher temperatures above 1778F
(970C) rather a dissolution of M6C seems to take place with increasing
exposure time. An exception is found for the specimen which had been
exposed for 10,637 h at 1800F (982C) which had shown the highest
amount of M6C (3.9%). Higher amounts of Nb (1.2 mass-%) seem to
promote the precipitation of finely dispersed NbC as eutectic structures
evidently stabilizing the microstructure and suppressing the formation
of M6C. After longer exposure times at temperatures of 1800F (982C)
the amount of M23C6 increases after times higher than 8,000 h and
Atlas of Microstructures
Atlas of Microstructures
23
Service-Exposed Samples
Below the history of the Service-Exposed Tubes is described in detail. Exposure conditions and chemical composition are summarized
in Tables 7-8.
Table 7: Exposure Conditions of Service-Exposed Samples
Tube Skin
Tube Skin
Hoop
Hoop
Expo.
Sample_ID
Sample ID
Temp.
Temp.
Stress
Stress
Time
Donated
Dechema
MTI
Alloy
F
C
ksi
MPa
hrs
by
A234_050
SE-A
HPNb
1630
888
2.12
14.60
108,306
MetalTek
A234_051
SE-B
HPNb
1607
875
2.12
14.60
108,306
MetalTek
A234_052
SE-C
HPNb
1650
899
0.72
4.96
75,000
Eastman
A234_053
SE-D
HPNb
~1650
899
1.16
8.00
148,920
Eastman
A234_054
SE-E
HPNb
~2350
1288
1.30
9.96
35,000
APCI
A234_055
SE-F
HPMA
~1800
982
1.79
12.30
65,664
APCI
A234_056
SE-G
HPMA
1673
912
2.04
14.00
94,900
APCI
A234_057
SE-H
HPNb
1661
905
1.98
13.60
131,000
Syncrude
A234_058
SE-I
HPNb
1661
905
1.98
13.60
143,000
Syncrude
A234_059
SE-J
HPNb
1661
905
1.98
13.60
143,000
Syncrude
Table 8: Chemistry of Service Exposed Samples
Sample_ID Sample ID
Dechema
MTI
Cr
Ni
A234_050
SE-A
24.5
34.1
A234_051
SE-B
24.5
34.0
A234_052
SE-C
24.9
31.7
A234_053
SE-D
24.5
34.5
A234_054
SE-E
25.3
32.6
A234_055
SE-F
24.9
32.8
A234_056
SE-G
24.7
34.0
A234_057
SE-H
23.9
31.8
A234_058
SE-I
24.4
33.3
A234_059
SE-J
24.3
33.4
24
Fe
bal
bal
bal
bal
bal
bal
bal
bal
bal
bal
Mn
0.40
0.34
0.68
0.83
1.20
1.00
0.70
0.50
0.60
0.57
Composition (wt%)
Si
C
0.90
0.45
0.90
0.39
1.10
0.42
1.90
0.40
2.50
0.40
1.30
0.42
0.90
0.57
1.10
0.42
1.40
0.42
1.40
0.39
Nb
0.80
0.75
0.74
1.40
1.20
0.98
0.50
1.20
1.20
1.20
N
0.047
0.060
0.074
0.062
0.074
0.038
0.104
0.128
0.090
0.102
Other
Ti,Zr & W<0.1
Ti,Zr & W<0.1
Ti,Zr & W<0.1
Ti,Zr & W<0.1
Ti,Zr & W<0.1
Ti & W>0.1, Zr<0.1
Ti,Zr & W<0.1
Ti,Zr & W<0.1
Ti,Zr & W<0.1
Ti,Zr & W<0.1
Atlas of Microstructures
UTS
Yield Strength
Elongation
SE-A
66.9 Ksi / 462 MPa
37.6 Ksi / 259 MPa
5.9%
SE-B
65.8 Ksi / 454 MPa
40.1 Ksi / 277 MPa
3.0%
Sample SE-C:
This sample was removed within 6 inches (15.2 cm) from the outlet
end of an HP-modified tube. The tube had been in service approximately 75,000 hours with no operational anomalies. Sample was
included to observe the microstructural changes at relatively mild operating conditions.
Sample SE-D was extracted from the failed tube at the same elevation as the through-wall rupture. However, the sample was removed 180
degrees around the circumference from the rupture in an effort to obtain
a microstructure representative of long-term ageing.
Sample SE-E:
The HP modified reformer tube was in service for approximately
four years prior to rupture (see Table 8). The ruptured area of the
tube had a maximum diametral growth of only 2.3%. Metallographic
analysis revealed numerous, tight interdendritic cracks running nearly
through wall with minimal creep voids. Higher magnification examination revealed a lamellar-like phase adjacent to the interdendritic cracks.
The tight intergranular cracks and lamellar phase are characteristic of
incipient melting. Incipient melting is liquation of low melting point
constituents along the dendrite boundaries. Incipient melting of cast
HP alloys generally occurs at approximately 2350F (1288C) or 100F
(56C) below the bulk melting temperature of approximately 2450F
(1343C).
Sample SE-D:
This tube failed approximately four months after a scheduled plant
outage. The failure was likely the result of debris inadvertently left in the
tube during the outage. The debris caused channeling of the feed gas
resulting in a local hot spot and subsequent rupture. Immediately adjacent to the rupture, the secondary carbides were completely dissolved
with notable coarsening of the primary carbides. The microstructural
features suggested that the metal temperature at the ruptured area of the
tube exceeded 2012F (1100C).
Atlas of Microstructures
25
26
Sample SE-G:
The HP micro-alloyed tube was removed from the furnace for
destructive remaining life assessment. The tube had been in service
approximately 95,000 hours with a design life of 100,000 hours (typical
operating conditions listed in Table 8). The tube had a maximum diametral growth of 5%. Metallographic analysis found the onset of creep
void alignment but no creep fissure formation suggesting considerable
remaining life.
Remaining life assessment was performed via the Omega creep strain
method. Despite the actual service life reaching the original design life,
the measured strain rates and calculated Omega indicated the tubes have
an estimated remaining life of nearly 17 years at the normal operating
conditions. The estimated tube life decreased approximately 44% with
each 25 F (14C) increase in tube metal temperature.
Sample SE-H:
This HP-modified sample was extracted from a tube removed from a
furnace after approximately 131,000 hours of operation (typical operating conditions in Table 8). The sample was removed from the fired side
of the tube approximately 16 inches (40.6 cm) above the floor of the
top-fired, downflow reformer.
Eddy current inspection of the tube immediately prior to removal
indicated that the tube had a 30% through-wall indication and 2.0-
Atlas of Microstructures
Atlas of Microstructures
27
Service-Exposed Samples
The results of the image analytical determination of the volume fraction of the precipitates for the service-exposed tubes of the centrifugally
cast materials are summarized in Table 12.
Table 12: Results of image analysis of Service Exposed Samples
Sample_ID
A234_056
A234_055
A234_051
A234_050
A234_052
A234_053
A234_057
A234_058
A234_059
A234_054
Sample
SE-G
SE-F
SE-B
SE-A
SE-C
SE-D
SE-H
SE-I
SE-J
SE-E
Material
HPMA
HPMA
HPNb
HPNb
HPNb
HPNb
HPNb
HPNb
HPNb
HPNb
C
912
982
875
888
899
899
905
905
905
1288
Hoop Stress
Ksi
2.04
1.79
2.12
2.12
0.72
1.16
1.98
1.98
1.98
1.30
MPa
14.0
14.0
12.30
14.60
4.96
8.00
13.60
13.60
13.60
9.96
Expo.
Time
hrs
94,900
65,664
108,306
108,306
75,000
148,920
131,000
143,000
143,000
35,000
area fraction in %
M(C,N)
0.3
<0.2
0.6
1.6
0.1
1.3
M23C6
12.9
6.2
7.0
9.2
8.0
5.5
7.4
6.5
7.9
4.0
core
M6C
1.6
1.5
0.8
2.2
M2(C,N) G-Phase
+
2.0
1.2
0.4
1.0
M(C,N)
+
+
(+)
0.2
2.0
2.2
2.5
0.5
1.5
1.6
+
28
Atlas of Microstructures
Alloy HPMA
1673F/94,900 h/2.04 ksi (912C/94900 h/140 MPa)
This specimen contains a very large number of M23C6 primary and
secondary carbides, Figures 2- SE-Ga to 2-SE-Gr. Their volume fraction
amounts to 12.9%. Furthermore in the center of the tube wall as well as
in the metal sub-surface zone carbides and nitrides of the type M6C and
M2(C,N) are present. The sum of these carbides amounts to about 1.6%.
The M(C,N) have evidently been almost completely dissolved or were
oxidized and appear white in the cross-section which had been coated
with ZnSe. Their fraction amounts to 9.3%. The G-phase could not be
identified in this specimen. Besides the precipitates a large number of
creep pores could be observed. In the BSE image the M2(C,N) appear
bright due to the high Nb content, see Table in Figures 2-SE-Ga and
2-SE-Gb. The high amount of N (0.1 mass-%) could explain the formation of M2N, which contains nearly 50 atomic-% Nb. As Nb is bound
to N it is not available for the formation of G-Phase. High N-contents
seem to prevent G-Phase formation.
1800F/65,664 h/1.79 ksi (982C/65664 h/12.3 MPa)
After operation for 7.5 years at 1800F (982C) the microstructure
of the tube contains a high level of creep pores and shows M23C6 primary
and secondary carbides all over the entire cross-section with a fraction
of about 6.2%. In addition less than 2.0% G-phase due to the high Si
content of 1.4 mass-% and the low N content of 0.04 mass-% as well
as residues of M(C,N) with about less than 0.2% are observed, Figures
2-SE-Fa to 2-SE-Fr.
Alloy HP Modified
1607F/108,306 h/2.12 ksi (875C/108306 h/14.6 MPa)
After an operation time of more than twelve years at 1607F (875C)
only a few creep pores are observed. The carbides are arranged in a
distinctive eutectic structure, Figure 3-SE-Ba to 3-SE-Br. Besides about
7.0% M23C6 primary and secondary carbides about 1.2% G-phase and
Atlas of Microstructures
29
1.5% M6C are present in the microstructure. M6C and G-phase are
located in contact with M23C6 carbides or are found in areas of secondary M23C6 carbides. Possibly a small fraction of the G-phase particles are
rather M(C,N) particles which can, however, not be distinguished due
to their small dimensions. In the outer sub-surface zone some M2(C,N)
is present.
1630F/108,306 h/2.12 ksi (888C/108306 h/14.6 MPa)
The microstructure of this part of the tube which had been exposed
to a slightly higher temperature is significantly different from that of
the section exposed to 1607F. (875C) The primary and secondary
carbides have changed their shape and have coarsened and the number
of secondary carbides has decreased due to agglomeration, Figures 3SE-Aa to 3-SE-Ar. The section contains besides 9.2% M23C6 also about
0.4% G-phase, 0.8% M6C and 0.6% M(C,N) in the core. In the metal
sub-surface zone some M2(C,N) is observed, however no G-phase could
be identified, Figures 3-SE-Ak to 3-SE-Ar. The different microstructures
of both samples, A and B, cannot be explained by differences in composition as their compositions are almost identical, Table 8.
1650F/74,000 h/0.72 ksi (899C/74000 h/4.96 MPa)
The microstructure of this specimen which was exposed for 8.5
years reveals a similar appearance to that of the specimen exposed for
108,306 h at 1630F. (888C) Both show except of a higher Si content
nearly identical compositions, Table 8.The structure is characterized by
a eutectic-type of precipitate arrangement in areas of residual melt and
along grain boundaries, Figures 3-SE-Ca to 3-SE-Cr. Besides about 8%
M23C6 primary and secondary carbides, 1% G-phase and 2.2% M6C
were measured. The M6C particles are sometimes found in combination
with M23C6carbides in the dendrites of the matrix.
1650F/148920 h/1.16 ksi (899C/148920 h/8.0 MPa)
After seventeen years of exposure the microstructure contains a large
number of creep cracks resulting from creep pores which had grown
30
Atlas of Microstructures
Conclusions
Among the two service-exposed tubes of the alloy HPMA only the
one exposed at 1800F (982C) for 7.5 years contained, besides M23C6
and residues of undissolved M(C,N), some G-phase in an amount less
than 2%. The tube which had been exposed for eleven years at the lower
temperature of 1673F (912C) showed M6C and M2(C,N) instead of
G-phase. The chemistry of this specimen shows a very high N content
which obviously leads to the formation of Nb-rich M2N. G-Phase
cannot be formed because not enough free Nb is present. Possibly the
highly Nb-containing M2(C,N) represents a precursor stage of the
formation of the G-phase. The HP Modified service-exposed tubes with
the exception of the two tubes with exposure time higher than 75,000
h contain between 0.4 and 1.6% G-phase besides 7-9% M23C6 and 0.82.5% M6C. In microstructures containing more than about 1% G-phase
Atlas of Microstructures
no M(C,N) can be observed. One tube peaks out due to its high amount
of M(C,N), while the amount of M23C6 is low. In this tube no G-phase
had been formed which may be due to its chemical composition. Indeed
a high content of Nb besides a high Si content was found which however
does not explain why the G-phase had not formed here. Another tube
contains Nb-rich M2(C,N) instead of G-phase. Again a high N content
was found by chemical analysis. It can be suspected that the formation of Nb-rich M2N is enhanced by high N contents. Again it can be
suspected that M2(C,N) is a precursor stage of the formation of G-phase.
In other words M2(C,N) is possibly dissolved after longer exposure
times and G-phase is formed. Evidently at 2350F (1288C) a critical temperature has been reached at which the material HP Modified
shows molten areas in the microstructure due to the formation of lowmelting phases which can reduce the creep strengths locally to a dramatic
extent.
Summary
Composition of the Precipitates
In Table 13 the average compositions in atomic-% of the different types of carbides and phases are listed for the different alloys. The
values of all elements except for C, (N), which were taken from the
difference of the sum of the determined metal fractions and 100%, rest
upon several averaged spot measurements for all samples of each alloy
by EPMA. Comparison of the results shows a good agreement within
the accuracy of the measurements for each type of phase. The existence
of M2(C,N) of different compositions in alloy HPMA und HPNb is
remarkable. Both alloys show probably Cr-Nitrides and Cr- and Nbrich M2(C,N).
31
Table 13: Composition of the precipitates in atomic-% (Mean values of several measurements)
Type
Alloy
Cr
Ni
Fe
Si
Nb
35/45
63.7
1.1
2.4
0.0
0.0
M7C3
HPMA
58.9
1.1
7.0
0.0
0.2
HPNb
61.9
1.0
5.9
0.0
0.2
35/45
4.0
2.1
1.0
0.0
49.0
M(C,N)
35/45
3.1
1.3
0.5
0.1
8.6
HPMA
3.6
2.1
1.6
0.6
48.4
HPNb
2.8
1.2
1.4
0.1
49.9
35/45
68.9
3.4
3.3
0.0
0.1
M23C6
HPMA
64.0
3.4
8.9
0.0
0.1
HPNb
66.5
3.4
8.7
0.0
0.1
35/45
34.0
27.1
1.9
13.0
8.3
M6C
HPMA
38.0
23.9
3,4
12.0
4.6
HPNb
31.3
25.8
3.6
13.1
10.2
35/45
5.1
49.4
2.9
22.0
18.9
G-Phase
HPMA
2.7
47.3
5.3
22.9
20.0
HPNb
4.0
47.7
5.8
21.8
19.6
35/45
64.8
0.7
0.8
0.0
0.3
HPMA
62.6
0.3
1.5
0.0
0.3
M2(C,N)
HPMA
32.0
1.8
3.0
0.0
29.8
HPNb
63.0
0.5
1.9
0.0
0.3
HPNb
32.2
1.3
2.5
0.0
31.5
Annotations: n. a. - not analysed
32
Ti
0.2
0.0
n.a.
3.5
30.5
2.1
0.1-1.1
0.0
0.0
0.0
0.2
0.0
0.0
0.5
0.4
0.0
n.a.
n.a.
1.4
n.a.
0.1
Zr
0.0
0.0
0.0
0.0
9.4
2.1
1.6
0.0
0.0
0.0
0.0
0.0
0.0
0.6
0.0
0.1
n.a.
n.a.
0.0
0.0
0.0
C, (N)
32.6
32.8
30.6
40.1
46.8
39.5
42.9
23.9
23.6
21.1
15.5
18.0
15.9
0.2
0.9
1.2
33.4
35.0
32.0
34.4
32.4
Atlas of Microstructures
Alloy 35Cr45Ni
The development of the M23C6-Carbides is shown in Fig. 1-PKD-a.
The content of M23C6 increases the faster the higher the temperature. The
content of M(C,N) strongly depends on the composition of the alloy,
especially on the alloy content of the M(C,N) forming elements Nb and
Ti. The plot of a tendency does not make much sense here, Figure 1PKD-b. For the formation of carbides of the M6C-type a similar trend is
observed, Figure 1-PKD-c. The area fraction is up to 10.7% after exposure times of 5,000-10,000 hours at higher temperatures (1900-2050F,
1038-1121C). Higher temperatures and exposure times >1,000 hours
seem to promote the formation of M6C, which is rich in Cr, Si and Nb.
The G-phase is observed in the temperature range of about 1800-2000F
(982C-1093C) after exposure times of 500 <T<10,000 hours, Figure
1-PKD-d. For the formation of the G-phase high Nb fractions are necessary. Due to diffusion processes the NbC dissolves, which is shown by a
decreasing content of the M(C,N). In addition a large fraction of Si and
Ni is needed for the formation of the G-Phase. The supply of Ni and Si
results from the matrix. It is remarkable that the M6C-content is lower
or M6C does not exist at all, in samples, in which the G-Phase is present
Atlas of Microstructures
and vice versa, cf. Figures 1-PKD-c and 1-PKD-d. The scatter bands
of the area fractions of the precipitates for the different temperature
ranges are shown in Fig. 1-PKD-e. From these scatter bands the user
can easily extract the expected area fraction of a certain precipitate (MC,
M23C6, M6C and G-phase) for the specified temperature ranges. It must
be pointed out, that the fraction of MC strongly depends on the alloy
composition and on the formation of M6C or G-phase, respectively. It
is noticeable that the MC-content is lower or MC does not exist at all,
in samples in which M6C or the G-phase are present and the that M6Ccontent is lower or M6C does not exist at all, in samples, in which the
G-Phase is present.
Alloy HPMA
The growth of the M23C6-fraction with increasing exposure time for
three temperature ranges is similar to the development of this type of
carbides in alloy 35Cr45Ni, Figure 2-PKD-a. The scatter of the amount
of M(C,N) is shown in Figure 2-PKD-b. In the actual case its fraction
predominantly depends on the composition of the sample, because here
only in one case G-Phase is present. Figure 2-PKD-c shows the change
in the fraction of M6C for the different temperature ranges. No clear
tendency can be derived. In comparison to alloy 35Cr45Ni less M6C
(max. 1.7%) is present. Only one of the samples was containing the Gphase. The scatter bands of the area fractions of the precipitates for the
different temperature ranges are shown in Fig. 2-PKD-d. As only one
of the SE-samples was containing the G-phase, no scatter bands for the
G-phase were drawn. Attention should be given to the different scale of
the y-axis for the M6C-area fractions in comparison to alloy 35Cr45Ni
(Fig.1-PKD-e).
33
Alloy HP Modified
For this alloy the content of M23C6-Carbides is more stable during
exposure time. No clear dependence on the temperature range is shown
in Fig. 3-PKD-a and 3-PKD-e compared to 35Cr45Ni (Fig. 1-PKD-a
and 1-PKD-e) and HPMA (2-PKD-a and 2-PKD-d). The maximum
fraction of M23C6 is 11%. One sample shows a very low fraction because
of dissolution processes due to a temperature near the melting point (red
value (SE-E) in Fig. 3-PKD-a and the beginning of melting, marked
with an arrow in Fig. 3-PKD-e). The distribution of the M(C,N) fractions shows two levels depending on the respective content of Nb. High
Nb-contents 1.18 mass-% seem to result in high M(C,N) fractions
(1.0-3.0 area-%). Low Nb contents (0.74-0.83 mass-%) seem to result
in lower M(C,N) fractions (1.0 area -%). In Fig. 3-PKD-e two green
hatched scatter bands representing M(C,N) are shown for the temperature range of 1787-1800F (970-982C). The formation of M6C initially
shows an increase in its fraction (max. 4%) up to an exposure time of
about 10,000 hours. After long exposure times >50,000 hours this type
of carbides partly seems to dissolve, while simultaneously in certain cases
G-Phase (max. 2%) is formed (Fig. 3-PKD-c and 3-PKD-e). G-phase
was only observed in the temperature range of 875-955C for exposure
times between 750- 150,000 hours (Fig. 3-PKD-d and blue hatched
scatter band in Fig. 3-PKD-e). In this temperature range in addition to
M6C the G-phase exists. For higher temperature ranges G-Phase does
not exist at all whereas M6C is sometimes present (green and red scatter
bands in Fig. 3-PKD-e).
The Project Team that reviewed this document was surprised at the
rapid conversion of MC to M23C6 carbides and at the absence of G
phase in some samples. Following receipt of the manuscript Dr. Schtze
offered some additional insights. This discussion is in the Addendum.
Acknowledgement
34
Literature
L. Dillinger and D. Roach, Metallurgy and Metallography of Reformer
Alloys, Battelle Columbus Laboratories, May 1, 1975
2
Metallografie, 14. Edition, Editors H. Schumann and H. Oettel,
2005 Wiley-VCH Verlag GmbH Co. KGaA, Weinheim, ISBN
3-527-30679-X
3
Atlas der Interferenzschichten-Metallographie, Editors H.- E. Buehler
and H. P. Hougardy, 1979 Deutsche Gesellschaft fr Materialkunde
e.V, Frankfurt am Main, ISBN 3-88355-015-9
4
I. Aydin, H. - E. Buehler, Contrasting the Carbides M23C6, M7C3
and M6C in High Temperature Resistant Cast and Wrought Alloys,
Praktische Metallografie 17, 1980, no. 9, 441 446
5
I. Aydin, H. - E. Buehler and A. Rahmel, Precipitation in the heat
resistant nickel-base cast alloys G-NiCr 28 W and G-NiCr 50 Nb,
Arch. Eisenhuettenwesen 54, 1983, no. 11, 461 466
6
I. Aydin, H. - E. Buehler and A. Rahmel, Determination of optical constants of carbides and carbonitrides in refractory steels,
Microstructural Science 9, 1981, 55 63
7
I. Aydin, H-E. Buehler, A. Rahmel, Observations concerning nitridation of refractory austenitic materials in air and combustion gases,
Werkstoffe und Korrosion 31, 1980, 675 682
8
ASM Handbook, Volume 9, Metallography and Microstructures, 1998,
ASM International, Materials Parc; OH 44073
9
Dr. D. Flahaut, D. T. Pool, 35Cr45Ni Micro-alloy Friend or Foe,
presented at: ERTC Petrochemical Conference, Vienna, 2004
10
John J. Hoffman, G.E. Gapinski, Properties and Microstructures of
Outlet Manifold Components, Ammonia Technical Manual, 2002
11
R. C. Ecob, R. C. Lobb and V. L. Kohler, The formation of G-phase
1
Atlas of Microstructures
Addendum
1. Etchant Cookbook
Dr. Vogels pickle:
Vogelsche Sparbeize (German name) is a commercial pickle, named
after its inventor Dr. Otto Vogel (Patents: DRP. 309264, 346530 und
350771; Brit. P. 158768,
188713;.)
Dr. Vogels Sparbeize is a chemical solution whose exact recipe is not
known for the public. It partially acts as an inhibitor.
Known contents are: 20% H2SO4, 1% H3PO4, 40-50% 1methoxy-2-propanol, 5-7% nonylphenol-ethoxylate, 3-5% thio-uric
acid. It can be purchased e.g. from the following supplier:
Buehler GmbH
In der Steele 2
D-40599 Dsseldorf
http://www.buehler-met.de
http://www.buehler.com
Purchase order number : 170014 quantity: 0,25l price: 26,00
Safety data sheet: www.buehler-met.de/pdf/msds/170014.pdf
Most probably there will also be an American Buehler subsidiary
with the address given on the above .com web page.
V2A etchant
This is a traditional etchant dating back to the middle of the last
century which had been developed for the etching of austenitic stainless
steels. We took the composition from the chapter etchants of Ni and
Ni-alloys: recipe m10 of the following German book.
Metallographisches, keramographisches und plastographisches
tzen
6. berarbeitete Auflage
[Metallographic, keramographic and plastographic etching. A
Atlas of Microstructures
Cookbook.]
Unter Mitarbeit von: Carle, Veronika.
1994. 6. Auflage. XII , 320 Seiten, 23 Abbildungen, 21x15cm
(Materialkundlich-Technische Reihe, Band 1)
Publisher: Gebr. Borntraeger, Berlin
ISBN 3-443-23014-8 brosch., EUR 26.00
Similar etchants may be described also in American textbooks for
metallography.
As a cooking guide you can use:
Composition: take 100 ml distilled H2O
add successively
100 ml HCl 1.19 M,
10 ml HNO3 1.40 M,
0.3 ml of an inhibitor e.g. Dr. Vogels Sparbeize
stir up the mixture and heat up to 50C
Etching: move the sample in the hot etchant for 25-30 sec by using
a pair of tweezers
2. Rapid conversion of M7C3 to M23C6
We know from experience, that the conversion of M7C3 to M23C6
is rapid in other comparable alloys, too (see Table 2 and Fig. 12, Fig.
38 and Fig. 66). If you consider the Time-Temperature-PreciptationDiagramm of the cast alloy 1.4857 whose composition is similar to
HP-MA and HPNb, you see that for a temperature of 900C the transformation of M7C3 to M23C6 is already completed after 20 hours of
exposure, Fig. 38. For higher exposure temperatures the conversion is
even faster.
So it is not surprising that we didnt find M7C3 and M23C6 in
coexistence, because the exposure times and temperatures are outside
the the band of conversion. The reaction of the conversion is 2M7C3 +
9M M23C6.
35
36
Atlas of Microstructures
Atlas of Microstructures
Silicon
The higher the Si content, the higher is the upper limit temperature of formation for G-phase, and the faster is the kinetics of
formation
Niobium
The higher the content of Nb the higher is the content of Gphase. Only at very high contents of Nb (more than 1.9 wt-%)
the G-phase may partially be formed directly from NbC
Titanium
Ti addition transforms NbC into NbTiC which is more stable
and less easy to destabilise to form G-phase. Ti therefore inhibits
G-phase formation. Maximum efficiency is 0.1% wt-% Ti
The following chemical composition (wt-%) is preferred in order
to limit Gphase formation:
Min
Max
C
0.40
0.45
Si
1.3
1.6
Ni
42
44
Nb
0.7
0.9
37
38
Atlas of Microstructures