MTI Interior 1 PDF

Atlas of Microstructures
AF234/7226/MSch
Prepared by:
Ellen Berghof-Hasselbcher 1)
Peter Gawenda 1)
Monika Schorr 1)
Michael Schtze 1)
John J. Hoffman 2)
1) Karl-Winnacker-Institut der DECHEMA e.V.

Theodor-Heuss-Allee 25
60486 Frankfurt am Main
Germany
2) Air Products and Chemicals, Inc.
7201 Hamilton Boulevard
Allentown, PA 18195-1501
USA
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.
ii
All rights reserved, including translations

ISBN 978-1-57698-046-0
No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the publisher. This
document was prepared under the sponsorship of the Materials Technology Institute Inc. (MTI) and is approved for release.
All data and information contained in this document are believed to be reliable; however, no warranty of any kind, express
or implied, is made with respect to the data, analyses, or author of this document; and the use of any part of the document is
at the user's sole risk. MTI, the author, or any person acting on its behalf, assume no liability and expressly disclaim liability,
including without limitation liability for negligence, resulting from the use or publication of the information contained in this
document to warrant that such use or publication will be free from privately owned rights.
Published by Materials Technology Institute
www.mti-global.org
iii
Table of Contents
Introduction . .......................................................................................................................... xx
Description of Test Specimens................................................................................................. xx
Creep rupture test specimens........................................................................................................ xx
-
Alloy 35/45 Exposure Conditions...................................................... Table 1

Chemistry of Alloy 35/45 Creep Rupture Stress Samples................... Table 2
Alloy HPMA Exposure Conditions................................................... Table 3
Chemistry of Alloy HPMA Creep Rupture Stress Samples................. Table 4
Alloy HP Modified Exposure Conditions.......................................... Table 5
Chemistry of Alloy HP Modified Creep Rupture Stress Samples........ Table 6
Service-exposed specimens............................................................................................................ xx
- History of Service Exposed Samples
- Exposure Conditions of .................................................................... Table 7
- Chemistry of Service Exposed Samples............................................... Table 8
Sample Preparation and Microstructural Characterization....................................................... xx

Metallographic Preparation...................................................................................................... xx
Investigation Techniques.............................................................................................................. xx
- Differential Interference Contrast Technique (LM-DIC)
- Scanning Electron Microscopy (SEM-BSE)
- Electron Probe Microanalysis (EPMA)
- Interference Layer Metallography (LM-ZnSe)
- Image Analysis by the False Color Technique (LM-FC)
- Development of the Microstructure by Etching (LM-etched)
iv
Results of the Microstructural Analyses.................................................................................... xx

Alloy 35/45................................................................................................................................ xx
- Results of image analysis of Alloy 35/45............................................. Table 9
- As Cast............................................................................................... Fig. 1-1a-j
- Creep Rupture Samples..................................................................... Fig. 1-2a to 1-2j

..................................................................... Fig. 1-3a to 1-3r

..................................................................... Fig. 1-4a to 1-4r

..................................................................... Fig. 1-5a to 1-5r

..................................................................... Fig. 1-6a to 1-6z

..................................................................... Fig. 1-7a to 1-7r

..................................................................... Fig. 1-8a to 1-8r

..................................................................... Fig. 1-9a to 1-9z

..................................................................... Fig. 1-10a to 1-10r

..................................................................... Fig. 1-11a to 1-11r

..................................................................... Fig. 1-12a to 1-12z

..................................................................... Fig. 1-13a to 1-13v
- Conclusions. ........................................................................................................................ xx
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

- Creep Rupture Samples (continued)....................................................... Fig. 2-5a to 2-5r

..................................................................... Fig. 2-6a to 2-6r

..................................................................... Fig. 2-7a to 2-7z

..................................................................... Fig. 2-8a to 2-8r

..................................................................... Fig. 2-9a to 2-9r

..................................................................... Fig. 2-10a to 2-10r

..................................................................... Fig. 2-11a to 2-11r

..................................................................... Fig. 2-12a to 2-12r

..................................................................... Fig. 2-13a to 2-13r

..................................................................... Fig. 2-14a to 2-14r

..................................................................... Fig. 2-15a to 2-15r

..................................................................... Fig. 2-16a to 2-16r

..................................................................... Fig. 2-17a to 2-17s
- Conclusions. ........................................................................................................................ xx
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
- Creep Rupture Samples (continued)....................................................... Fig. 3-8a to 3-8r

..................................................................... Fig. 3-9a to 3-9r

..................................................................... Fig. 3-10a to 3-10r

..................................................................... Fig. 3-11a to 3-11r

..................................................................... Fig. 3-12a to 3-12r

..................................................................... Fig. 3-13a to 3-13r
- Conclusions. ........................................................................................................................ xx
Service Exposed Samples.............................................................................................................. xx

- Results of image analysis of Service Exposed Samples ........................ Table 12
- Alloy HPMA.................................................................................... Fig. 2-SE-Fa to 2-SE-Fr

..................................................................... Fig. 2-SE-Ga to 2-SE-Gr
- Alloy HP Modified ........................................................................... Fig. 3-SE-Ba to 3-SE-Br

..................................................................... Fig. 3-SE-Aa to 3-SE-Ar

..................................................................... Fig. 3-SE-Ca to 3-SE-Cr

..................................................................... Fig. 3-SE-Da to 3-SE-Dr

..................................................................... Fig. 3-SE-Ha to 3-SE-Hr

..................................................................... Fig. 3-SE-Ia to 3-SE-Ir

..................................................................... Fig. 3-SE-Ja to 3-SE-Jr

..................................................................... Fig. 3-SE-Ea to 3-SE-Ez
- Conclusions. ........................................................................................................................ xx
Summary................................................................................................................................. xx
Composition of the Precipitates. ................................................................................................... xx
Average Compositions of the Precipitates in atomic-%.......................... Table 13
vii
Precipitation Kinetics Diagrams............................................................................................... xx

- Alloy 35/45 . .................................................................................... Fig. 1-PKD-a to 1-PKD-d

..................................................................... Fig. 1-PKD-e
- Alloy HPMA ................................................................................... Fig. 2-PKD-a to 2-PKD-c

..................................................................... Fig. 2-PKD-d
- Alloy HP Modified . .................................................................. Fig. 3-PKD-a to 3-PKD-d

..................................................................... Fig. 3-PKD-e
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
extended aging time and temperature have been well documented by

Battelle Columbus Laboratories1. The Battelle data has proven to be a
valuable resource in estimating reformer tube exposure conditions associated with reformer tube failure analyses and remaining life assessments.
User demand for higher temperature/stronger alloys fueled continued
alloy development resulting in the introduction of the HP-modified alloy
in the 1970s. The HP-modified alloy had the same nominal chemistry
of the HP alloy along with the addition of typically 1 wt% niobium. The
niobium addition results in precipitation of M23C6, M7C3, and MC type
primary carbides upon solidification. In the M23C6 and M7C3 carbides,
niobium substitutes for some of the chromium with the complex
niobium-chromium carbides being more stable at elevated temperatures
as compared to chromium carbides. In the HP-modified alloy, niobium
is the primary carbide forming element in the MC type carbides.
In the 1980s, the demand for more severe design conditions and/or
design lives in excess of 100,000 hours led to the introduction of the
HP-micro-alloyed material. HP-micro-alloyed (or commonly designated as HPMA) material was based on the HP-modified chemistry
with micro additions of alloying elements. In general, micro-alloying
refers to intentional alloying additions at a concentration of 0.10 wt% or
less. Titanium is the most common micro-alloying addition with other
micro-alloying additions including zirconium, tantalum, or tungsten.
Rare earth elements such as lanthanum, cerium, and/or yttrium may
also be added. The micro-alloying additions provide a fine dispersion
of MC type carbides that are stable at temperatures well in excess of
2,000oF (1093oC).
In the 1990s, the 35Cr/45Ni alloy family grew in popularity. The
35Cr/45Ni alloy has similar creep strength to the HPMA alloy but with
notably improved carburization resistance. The excellent carburization
resistance made the 35Cr/45Ni alloy well suited for ethylene pyrolysis
furnace tubes.
As outlined above, alloy development has continued in centrifugally
cast heat resistant alloys allowing users to design for and operate at more
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
Table 1: Alloy 35/45 Exposure Conditions

Sample_ID Dechema
Sample MTI
A234_003
35/45-1 (c)
A234_004
35/45-2
A234_005
35/45-3
A234_006
35/45-4
A234_007
35/45-5
A234_008
35/45-6
A234_009
35/45-7
A234_010
35/45-8
A234_011
35/45-9
A234_012
35/45-10
A234_013
35/45-11
A234_014
35/45-12
A234_015
35/45-13
Test-Temp F
as cast
1800
1800
1922
1900
1900
1900
2012
2000
2012
2057
2050
2100
Table 2: Chemistry of Alloy 35Cr/45Ni Stress Rupture Samples

Sample_ID
Sample
Dechema
MTI
Cr
Ni
Fe
A234_003 35/45-1 (c)
35.85
44.59
bal
A234_004
35/45-2
35.12
43.69
bal
A234_005
35/45-3
34.84
44.43
bal
A234_006
35/45-4
32.30
43.40
bal
A234_007
35/45-5
32.39
44.77
bal
A234_008
35/45-6
34.34
45.42
bal
A234_009
35/45-7
34.41
45.37
bal
A234_010
35/45-8
32.30
43.40
bal
A234_011
35/45-9
32.65
44.61
bal
A234_012
35/45-10
34.84
44.43
bal
A234_013
35/45-11
32.30
43.40
bal
A234_014
35/45-12
32.59
44.34
bal
A234_015
35/45-13
34.84
44.43
bal
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
Expo. Time hrs
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
Table 3: Alloy HPMA Exposure Conditions

Sample_ID Dechema
Sample MTI
A234_018
HPMA-1 (c)
A234_019
HPMA-2
A234_020
HPMA-3
A234_021
HPMA-4
A234_022
HPMA-5
A234_023
HPMA-6
A234_024
HPMA-7
A234_025
HPMA-8
A234_026
HPMA-9
A234_027
HPMA-10
A234_028
HPMA-11
A234_029
HPMA-12
A234_030
HPMA-13
A234_031
HPMA-14
A234_032
HPMA-15
A234_033
HPMA-16
A234_034
HPMA-17
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
Expo. Time hrs
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
Table 4: Chemistry of Alloy HPMA Stress Rupture Samples

Sample_ID
Sample
Dechema
MTI
Cr
Ni
Fe
A234_018 HPMA-1 (c)
25.01
33.79
bal
A234_019
HPMA-2
24.92
34,54
bal
A234_020
HPMA-3
24.29
32.51
bal
A234_021
HPMA-4
25.01
33.79
bal
A234_022
HPMA-5
25.01
33.79
bal
A234_023
HPMA-6
25.85
35.25
bal
A234_024
HPMA-7
24.38
35.13
bal
A234_025
HPMA-8
24.49
33.53
bal
A234_026
HPMA-9
24.49
33.53
bal
A234_027
HPMA-10
24.36
33.18
bal
A234_028
HPMA-11
25.11
36.13
bal
A234_029
HPMA-12
25.01
33.79
bal
A234_030
HPMA-13
24.49
33.53
bal
A234_031
HPMA-14
24.36
33.18
bal
A234_032
HPMA-15
24.49
33.53
bal
A234_033
HPMA-16
24.49
33.53
bal
A234_034
HPMA-17
25.01
33.79
bal
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
Table 5: Alloy HP Modified Exposure Conditions

Sample_ID Dechema
Sample MTI
Test-Temp. F
A234_036
HPNb-1 (b)
as cast
A234_038
HPNb-2
1675
A234_039
HPNb-3
1700
A234_040
HPNb-4
1750
A234_041
HPNb-5
1778
A234_042
HPNb-6
1778
A234_043
HPNb-7
1800
A234_044
HPNb-8
1800
A234_045
HPNb-9
1800
A234_046
HPNb-10
1800
A234_047
HPNb-11
1850
A234_048
HPNb-12
1900
A234_049
HPNb-13
1900
Test-Temp. C
as cast
913
927
955
970
970
982
982
982
982
1010
1038
1038
Stress Ksi
Stress MPa
Expo. Time hrs
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
Table 6: Chemistry of Alloy HP Modified Stress Rupture Samples

Sample_ID
Sample
Dechema
MTI
Cr
Ni
Fe
A234_036 HPNb-1 (b)
25.30
34.93
bal
A234_038
HPNb-2
25.34
34.88
bal
A234_039
HPNb-3
25.26
33.56
bal
A234_040
HPNb-4
25.34
34.88
bal
A234_041
HPNb-5
25.11
32.85
bal
A234_042
HPNb-6
25.11
32.85
bal
A234_043
HPNb-7
24.88
33.33
bal
A234_044
HPNb-8
24.86
33.79
bal
A234_045
HPNb-9
24.86
33.30
bal
A234_046
HPNb-10
25.26
33.56
bal
A234_047
HPNb-11
25.62
34.75
bal
A234_048
HPNb-12
25.26
33.56
bal
A234_049
HPNb-13
25.26
33.56
bal
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.
Scanning Electron Microscopy (SEM-BSE)

After the LM-DIC investigations the same spot was investigated
by the scanning electron microscope using the back-scattered electron
imaging technique (SEM-BSE). In this case the differences in contrast
of the different precipitates resulting from differences in the density can
be used in order to distinguish between the phases.
Electron Probe Microanalysis (EPMA)

As a next step the phases documented by the other two techniques
were analyzed quantitatively for selected specimens with regard to their
chemical composition. In each case several spot measurements were
taken with a beam diameter of 2 m. The results of these measurements were averaged. From the ratio between the different metals in the
precipitates conclusions were drawn on the respective carbide type by
taking carbon and nitrogen respectively from the difference of the sum
of the determined metal fractions and 100%.
Interference Layer Metallography (LM-ZnSe)

After the EPMA investigations the specimens were coated by a PVD
process in an evaporation equipment (Edwards) at a vacuum of about
10-3 mbar3-7. As a coating material zinc selenide (ZnSe) was used which
allowed the comparison of these marked areas in the colored state with
the images from the other techniques. The evaporation source consists
of a little vessel made of tantalum with ZnSe grains filled into this vessel.
By resistance heating the vessel is heated to a point where the ZnSe
starts to evaporate. The evaporation rate is controlled by controlling the
heating current together with observation of changes of the color of
the section surface. When reaching the desired color (violet as a macroscopic color) of the specimen surface the heating current is switched off
immediately in order to achieve a reproducible thickness of the interference layer. Since the reflection characteristics change periodically with
the layer thickness and manual control of the heating current is not easy
this coating technique requires significant experience in order to achieve
a suitable interference layer. In some cases the layer had to be removed

again from the section surface several times and a new coating had to be
applied in order to come to satisfactory results. Due to the small dimensions of the different phases there may be slight deviations between the
different photographs taken by the different techniques.
Image Analysis by the False Color Technique (LM-FC)

By the use of the automated image analysis system Leica QWin in
combination with an automated laboratory microscope Leica DMLA
the different fractions of the precipitations which had been characterized
before by the other techniques were measured. For these measurements
specimens were used which had been contrasted by the interference
layer technique beforehand. An automated program routine was developed by which for each specimen several focused images were taken
at representative spots in the specimen center and partially also at the
specimen edges at a magnification of 1000x. The images were stored
and in a second run the lower and upper threshold values for the RGB
(red, green, blue)-colors of the different phases were determined and
recorded. Since, during the coating process the ZnSe layer can vary
from specimen to specimen, this procedure had to be performed for
all specimens at least once. For specimens where the microstructure in
the center and in the outer region was different this procedure had to
be applied separately for the two regions. The program routine allows
manual interaction with the measurement procedure, i.e. artifacts like
creep pores or cracks can be eliminated in order not to influence the
measurement results. Based on the information in the binary memory,
each of the phases was represented by a defined (false) color. For the
different types of carbides and other phases the following false colors
were selected:

M7C3
- purple
M6C
- red

M23C6
- yellow
M2(C,N) - green

M(C,N), MN - blue, cyan
G-phase - magenta

The matrix was not binarized and was used as a background in true
colors behind the false color image.
Development of the Microstructure by Etching (LM-etched)

Finally the interference layers were polished off and sections were
etched for 30 seconds at 50C in etchant V2A (composition: 100 ml
H2O, 100 ml HCl 1.19, 10 ml HNO3 1.40, 0.3 ml Dr. Vogels pickle.
Dr. Vogels pickle is a mixture of organic solvents with Thiourea. It
consists of 1-Methoxy- 2-propanol (40-50%), Thiourea (3-5%) and
Nonylphenol-ethoxylate (5-7%).
The interesting areas of the specimen were photographed at 200x
and 500x magnification. By etching the matrix is partially removed
so that edges are formed at the transition from matrix to precipitate.
Due to the local reflection situation of the light which hits the surface
perpendicularly the edges of these phase boundaries appear dark in the
photograph so that all phases have a dark seam and the grain boundaries
become visible.
(For more discussion on the etching technique, see the
Addendum).
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
outer zone / edge

M7C3 M23C6 M6C
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)
+
+
+
++
++
+
++
++
+
++
++
Results of the
Microstructural Analyses
colored purple and the M(C,N) were colored blue, Figure 1-1j.
Alloy 35Cr45Ni
1800F/1,509 h/3.3 ksi (982C/1509 h/22.8 MPa)

During exposure a transformation of the M7C3 particles had taken
place. The microstructure reveals some single creep pores, Figure 1-2a,
and a dominance of M23C6 carbides (7.9%) which had also formed very
fine secondary precipitates, Figures 1-2a to 1-2d. Furthermore a small
amount (1.5%) of G-phase can be observed which is frequently in direct
contact with the M23C6 carbides and shows a gray color with a slightly
reddish tone in the DIC image. It cannot be excluded that among those
particles which had been identified as G-phase and which have a dot-like
appearance (i.e. very small particles) could also be M6C carbides in some
spots. Due to their small dimensions it had not been possible to identify these precipitates exactly either by their color contrast or by EPMA
analysis. The G-phase (in the present analysis Ni16Nb6Si7) (according
to the ASM Handbook of Metals also Ni18Ti8Si168) is supposed to have
a negative influence on creep lifetime9. Due to its high amount of Nb
this phase appears bright in the BSE image, Figures 1-2e and 1-2f. After
coating with ZnSe the color becomes dark blue, while the Ti and Nbrich M(C,N) particles (0.5%) which had appeared as orange precipitates
in the DIC image become light blue to white, Figures 1-2c, 1-2d, 1- 2g
and 1-2h. After etching, the contrast between these precipitates increases
slightly, Figure 1-2i. In the false color image the following colors were
chosen for the different phases: M23C6 yellow, M(C,N) and MN,
respectively blue and cyan, respectively, G-phase magenta, Figure
1-2j. In the outer metal sub-surface zone of the specimen in addition
the carbide M6C (6.3%) was found which has, however, not been documented photographically in the present case.
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
Creep rupture samples

1800F/982C
1800F/7,217 h/2.9 ksi (982C/7217 h/20.0 MPa)

After 7,217 h the number of creep pores had increased in the micro-
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
are based on an estimation. The primary and secondary M23C6 particles

amount to 6.9%, the G-phase to about 0.7%, the M6C carbides to 12% and the M(C,N) particles to 0.4%. Due to the high amount of Nb
the M6C carbides in the BSE image appear with a similar level of gray
as the G-phase, Figures 1-4e and 1-4f, while M(C,N) has the brightest appearance. In the sub-surface zone the amount of M23C6 increases
slightly (7.4%), that of the M6C carbides reveals a significant increase
(4.2%), while the G-phase no longer exists in these positions, Figures 14k to 1-4r. Again, directly under the outer part of the metal sub-surface
zone M2(C,N) can be detected.
1900F/2,023 h/2.09 ksi (1038C/2023 h/14.4 MPa)
With increasing exposure time the amount of pores and cracks from
creep has significantly increased, Figure 1-5a. The fraction of primary
and secondary M23C6 (8.3%) and of the G-phase (3.0%) increase in
the core of the specimen, while almost no M6C (< 0.03%) and M(C,N)
(< 0.05%) can be found, Figures 1-5a to 1-5j. In the sub-surface zone
the amount of M23C6 (8.8%) and of M6C (13.1%) has significantly
increased. G-phase and M(C,N) are scarcely found there with a fraction
of less than 0.02%, Figures 1-5k to 1-5r. Again, M(C,N) can only be
found in the outermost zone of the sub-surface area.
1900F/6,331 h/1.6 ksi (1038C/6331 h/11.0 MPa)
After exposure of more than 6,000 h a significant coarsening and
increase of the volume fraction of M23C6 (10.5%) and of M6C (7.5%)
had taken place in the core and in the sub-surface area. The size of the
pores has increased, Figures 1-6a and 1-6b. The secondary carbides have
been dissolved or have become larger and blocky, Figures 1-6a to 1-6r. In
the outermost zone of the sub-surface again M2(C,N) is detected which
is slightly elevated in the LM-DIC image and appears somewhat darker
than M23C6 and brighter than M6C, Figure 1-6d. In the SEM-BSE
image this phase appears very dark so that it can be clearly distinguished
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.
1900F/10,247 h/1.54 ksi (1038C/10247 h/10.6 MPa)

After more than 10,000 h the amount of M23C6 has increased to
15% and that of M6C to 10.7%. The size of the carbides has increased
even more, Figures 1-7a to 1-7j, and a larger number of M23C6 carbides is
found than after 6,331 h. A large number of pores from creep are found
which have agglomerated increasingly to cracks oriented perpendicular
to the stress direction. In the sub-surface zone additionally M2(C,N) is
found, Figures 1-7k to 1-7r.
2012F/3,545 h/1.15 ksi (1100C/3545 h/7.9 MPa)

In the core area of the specimen the fraction of the significantly
coarsened M23C6 carbides has increased up to 12% and the fraction of
M6C amounts to 3.8%, Figures 1-10a to 1-10j. Furthermore some very
fine secondary precipitates of M23C6 are also found. The G-phase is no
longer present. The fraction of M(C,N) amounts to 0.4%. In the subsurface zone of the specimens about equal fractions of M23C6 and M6C
are observed (8.8%). Furthermore the microstructure contains about
0.6% M(C,N) and M2(C,N), Figures 1-10k to 1-10r.
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-
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.
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
+
+
+
+
outer zone / edge

M7C3
M23C6
M6C M2(C,N)
+
++
+++
+
+++
+
++
<0.3
+++
5.3
4.9
6.4
4.7
6.1
1.0
5.6
7.1
+++
0.0
+++
+++
++++
++++
9.8
5.1
15.2
+
Annotations: + few, ++ many, +++ very many, ++++ abundant
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%.
Creep Rupture Samples

1700F/927C
1700F/183 h/8.00 ksi (927C/183 h/55.2 MPa)
Already after 183 h at a temperature of 1700F (927C) and a load
of 8.0 ksi the M7C3 has been converted into the more stable M23C6.
Furthermore M23C6 secondary carbides are observed in the grain interior, Figures 2-2b to 2-2d. The EPMA analyses reveal that the conversion
of M7C3 into M23C6 is mainly due to an uptake of Ni from the matrix.
The Ni content of M23C6 amounts to about 4-5 mass-%. As a further
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
which is consistent with the identical compositions and phases formed

in the specimens numbers HP-MA-4 and HP-MA-5.
1800F/982C
1800F/177 h/5.97 ksi (982C/177 h/41.2 MPa)
In the austenitic matrix of the specimen again about 7.0% M23C6
(primary and secondary carbides) as well as 0.6% M (C,N) have been
found, Figures 2-6a to 2-6r. The outermost sub-surface zone additionally
contains about 5% M6C and M2(C,N) respectively (not documented
here). The preferred formation of M6C in the sub-surface area is presumably due to the relatively high amount of Nb in the specimen and the
outward diffusion of Cr and Si which are needed for the formation of
the oxide scale. Compared to M23C6 the carbide M6C is less rich in Cr
and Fe, but contains more Ni, Si and Nb. The formation of M2(C,N)
which is formed in the outermost sub-surface zone is probably due to
outward diffusion of Cr and nitrogen uptake from the environment.
The matrix in the metal sub-surface zone was strongly depleted in Cr
around these M2(C,N) particles.
1800F/2,436 h/3.51 ksi (982C/2436 h/24.2 MPa)
After 2,436 h the amount of secondary carbides evidently increases
slightly which is reflected by an increase of M23C6 to 7.9%, Figures 2-7a
to 2-7j. The high amount of M(C,N) of 1.2% is due to the composition of the material, cf. Table 4. Again in the metal sub-surface zone the
formation of M6C and M2(C,N) is observed, Figures 2-7k to 2-7z.
1800F/6,478 h/3.90 ksi (982C/6478 h/26.9 MPa)
After more than 6,000 h besides M23C6 (8.1%) and M(C,N) (0.5%)
smaller amounts of M6C (0.5%) are also found in the core area, Figures
2-8a to 2-8j. In the metal sub-surface zone the amount of M6C is twice
as high as in the core, Figures 2-8k to 2-8r.
1832F/12,289 h/3.0 ksi (1000C/12289 h/20.7 MPa)
After more than 12,000 h a coarsening of the primary carbides and a

coagulation of secondary carbides have taken place, Figures 2-9a to 2-9r.
The fraction of M23C6 amounts to 9.0%, M6C is observed neither in the
core nor in the metal subsurface zone. The fraction of M(C,N) amounts
to 0.8% and is clearly higher than in the case of specimen HP-MA-8
where in the core M6C had also been found.
1920F/1050C
1922F/137 h/3.48 ksi (1050C/137 h/24.0 MPa)
At this further elevated temperature a significant increase in the
number of creep pores and cracks can be observed, Figure 2-10a. The
M23C6 primary carbides seem to have coarsened and the secondary
carbides show a needle-like and blocky appearance, Figures 2-10c to
2-10r. Their fraction amounts to 9.5% in the core area with that of
M(C,N) amounting to 0.7%. Furthermore the core area contains 1.3%
M6C. The metal sub-surface zone shows the following fractions: M23C6
5.6%, M23C6 0.4% and M6C and M2(C,N) (as a sum) 9.8%,
Figures 2-10k to 2-10r.
1950F/2,558 h/2.80 ksi (1066C/2558 h/19.1 MPa)
After more than 2,500 h the amount of creep damage has further
increased, Figure 2-11a. In the core area the beginning of M6C formation (< 0.2%) has been found. The fraction of M23C6 amounts to about
8.8% and that of M(C,N) to about 1.0%, Figures 1-11a to 1-11j. The
microstructure of the sub-surface zone contains the following phases:
M(C,N) 0.9%, M23C6 7%, and M6C and M2(C,N) as a sum 5%,
Figures 2-11a to 2-11r.
1922F/4,864 h/2.32 ksi (1050C/4864 h/16.0 MPa)
After almost 5,000 h of exposure the amount of M23C6 carbides has
increased up to 9.7%. The fraction of M(C,N) amounts to about 1%.
M6C and M2(C,N) are neither observed in the core nor in the metal
sub-surface zone, Figures 2-12a to 1-12r.
17
1922F/11,778 h/2.08 ksi (1050C/11778 h/14.3 MPa)

After almost 12,000 h a significant coarsening and an increase in
the amount of M23C6 carbides has taken place in the core area of the
specimen with about 13.1%. M2(C,N) has spread from the metal subsurface zone into the core area, Figures 2-13c to 2-13j. Due to its higher
hardness M2(C,N) can easily be distinguished from the other carbides in
the DIC image. Even in the BSE image it shows a good contrast, Figures
2-13e and 2-13f. EPMA investigations reveal that this carbonitride is
almost free of Ni and has the composition (Cr, Fe)2 (C,N), cf. table
in Figures 2-13a and 2-13b. Values for the ratio between carbon and
nitrogen cannot be given here since C and N had not been analyzed.
Although the particles of M2(C,N) are rather coarse, which should facilitate their identification, it is not easy to distinguish them from M6C in
the ZnSe-coated section, Figures 2-13e to 2-13r. The false color image,
Figure 2-13j, was constructed by image manipulation in which the color
green was assigned to M2(C,N). Besides about 0.8% M(C,N) about
1.7% M6C plus M2(C,N) were measured in the core area, Figures 2-13a
to 2-13j. In the metal sub-surface zone no M23C6 was found. Besides
about 15% M2(C,N) about 0.8% M(C,N) were present, Figures 2-13k
to 2-13r.
2012F/1100C
2012F/105 h/2.39 ksi (1100C/105 h/16.5 MPa)
Compared to specimen HP-MA-10 (1922F/137 h) (xxx) the
amount of M23C6 carbides is lower with the precipitates being coarser,
Figure 2-14a to 2-14r. Almost no differences between the core area and
the metal sub-surface zone are observed with 7.7% M23C6 beside 0.8%
M(C,N).
2012F/2,714 h/1.80 ksi (1100C/2714 h/12.4 MPa)
After 2,714 h the fraction of M23C6 carbides has increased due to an
increase of secondary precipitates up to 9.6%. The fraction of M(C,N)
of 0.8% remained about constant. The microstructure in the core and in
the metal sub-surface zone is identical, Figures 2-15a to 2-15r.
18
2012F/5,580 h/1.45 ksi (1100C/5580 h/10.0 MPa)

With increasing exposure time the amount of M23C6 increases further
(11.1%). The number of the secondary carbides decreases since the very
fine particles combine with the larger ones to form coarse particles and
some of them seem to agglomerate with the primary carbides, Figures
2-16a to 2-16r. The fraction of M(C,N) lies around 1%. Occasionally
high Nb-containing needles are found which most probably are nitrides,
Figure 2-16b. The thickness of these needles is less than 1 m and their
length reaches up to 100 m. In the DIC image they appear gray and
after ZnSe coating violet.
2012F/8,990 h/1.35 ksi (1100C/8990 h/9.3 MPa)
After almost 9,000 h of exposure time no visible changes in the
microstructure are observed compared to the former specimen, Figure
2-17a to 2-17r. Even the fractions of M23C6 (10.9%) and of M(C,N)
with 1.1% are comparable. Again highly Nb-containing precipitates are
found, Figures 2-17b and 2-17s.
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
area, while this is more pronounced in the metal sub-surface zone. At

1700F (927C) the M6C precipitates with an amount of less than 1%
are dissolved after times longer than 2,500 h in the core of the specimen and exist in the metal sub-surface area up to about 6,000 h where,
however, only an amount of less than 0.3% can still be detected. After
more than 8,000 h no more M6C can be found. At the higher temperature of 1800F (982C) M6C can be found only in the metal sub-surface
zone at times up to 2,500 h. After 6,500 h traces (< 0.5%) of M6C are
observed in the core area, while the fraction in the metal sub-surface
zone decreases significantly. Due to the uptake of nitrogen M2(C,N) is
also formed whose fraction was measured as part of the sum with M6C.
After 12,000 h at 1832F M6C is neither found in the core nor in the
metal subsurface zone. At temperatures above 1900F (1038C) small
amounts of M6C (< 0.2-1.3%) are observed in the core after exposure
times of 2,500 h which becomes dissolved after longer times. The formation of M6C and M2(C,N) occurs only in the metal sub-surface zone to
a significant extent with their fraction decreasing markedly after 2,500 h
due to outward diffusion processes. However, after almost 5,000 h M6C
and M2(C,N) are observed neither in the core nor in the metal subsurface zone. After almost 12,000 h the extension of M2(C,N) is marked
and reaches from the sub-surface zone into the core of the specimen. In
the core only single precipitates of M6C are observed besides M23C6 and
M(C,N), while in the outer part of the specimen only M2(C,N) and
M(C,N) are present. At the maximum temperature of 2012F (1100C)
the increase of the fraction of M23C6 with time is very marked. M6C and
M2(C,N) are not present in the core and are observed only after short
exposure time (100 h) and sporadically in the metal sub-surface zone.
After longer exposure times in the core area some highly Nb-containing needle-shaped phases are formed. In the metal sub-surface zone an
increasing depletion of oxide-forming elements takes place so that finally
M6C and M2(C,N) are dissolved completely. In none of the specimens
investigated was the G-phase detected.
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.
Creep Rupture Samples

1675F/913C
1675F/659 h/6.50 ksi (913C/659 h/44.6 MPa)
At the lowest temperature of 1675F (913C) after 659 h in the
core primary carbides of the type M23C6 are found together with elongated secondary carbides of the same type (together 8%). Furthermore
some carbides of the type M6C (0.5%) and its secondary precipitates
are observed together with about 1.8% G-phase and 0.3% M(C,N),
Figures 3-2a to 3-2j. The M(C,N) precipitates besides a high amount
of Nb contain also Zr, see table in Figures 3-2a and 3-2b. Presumably
the formation of G-phase was favored by the high amount of Si of 1.67
mass-%, see Table 6. In the DIC image it was not easy to distinguish
between the G-phase and the carbides of the type M6C since both of
them appear gray, Figures 3-2c and 3-2d. In the BSE image the G-phase
appears bright due to its high Nb content, while M6C does not differ
much from the matrix, Figures 3-2e and 3-2f. The M(C,N) particles
appear bright in the BSE image. These particles can, however, be distin-
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
M2(C,N) G-Phase M(C,N)

+
1.8
(+)
+
(+)
(+)
+
(+)
++
++
++
(+)
+
+
+
M7C3
++
outer zone / edge

M23C6 M6C M2(C,N) G-Phase
+++
+++
+++
+++
+++
+++
+++
+++
++++
+++
+++
++++
+
+
++
++
+
(+)
(+)
(+)
++
++
++
++
(+)
(+)
+
+
+
+
Annotations: (+) very few, + few, ++ many, +++ very many, ++++ abundant
20
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.
of the specimen residues of dissolving M(C,N) are frequently detectible

due to their blocky shape and are embedded into M6C, Figures 3-4k to
3-4r. Bluish colorings which would indicate the presence of G-phase
were not detected here.
1700-1750F/913-955C
1778F/1,185 h/3.77 ksi (970C/1185 h/26.0 MPa)

After this extended exposure time of almost 1,200 h no differences
with regard to the volume fractions of the phases identified by EPMA
are observed compared to specimen HP-Nb-5, Figures 3-6a to 3-6r. The
G-phase is not detected despite the high Si amount of 1.79 mass-%.
1700F/59 h/8.5 ksi (927C/59 h/58.8 MPa)

After this short exposure time at 1700F (927C) the microstructure
is full of secondary precipitates. The M23C6 carbides (8.0%) show typical eutectic appearance, Figures 3-3a to 3-3r. Besides a small amount of
M6C (0.7%), some M(C,N) are also observed with about 1% in the core
area as well as in the metal sub-surface zone.
1750F/794 h/5.10 ksi (955C/794 h/35.2 MPa)
After this increased exposure time at 1750F (954C) the amount of
M6C in the core area has slightly increased compared to specimen HPNb-3, while the amount of M(C,N) has clearly decreased, Figures 3-4a
to 3-4j. Possibly some G-phase is present in the M6C carbides which,
however, could not be evidenced by EPMA. The bluish coloring in the
violet M6C seemed to indicate their presence, Figure 3-4a. Since the
chemical composition of this specimen is identical to specimen HP-Nb2, the appearance of the G-phase would have been expected here. In the
image analysis this phase was measured together with M6C or M(C,N) so
that their fractions may appear slightly too high. In the sub-surface zone
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
are about equal to the former specimen.
image in Figure 3-12r.
1800F/7,833 h/5.42 ksi (982C/7833 h/19.3 MPa)

Even after almost 8,000 h no changes in the microstructure can be
found compared to the short exposure times, Figures 3-9a to 3-9r. Again
the volume fractions of the carbides are unchanged.
1900F/5,373 h/1.90 ksi (1038C/5373 h/13.1 MPa)

After doubling the exposure time at a test temperature of 1900F
(1038C) the M6C carbides have become partially dissolved, while
again increasingly M23C6 secondary carbides were precipitated, Figures
3-13a to 3-13r. Their volume fraction of M23C6 increases up to 9.5%,
while that of M6C decreases to 1%. The amount of M(C,N) with 0.6%
remains stable.
1800F/10,637 h/2.80 ksi (982C/10637 h/19.3 MPa)

Compared to specimens HP-Nb-7 to HP-Nb-9 this specimen
HP-Nb-10 contains only 0.74 mass-% Nb, which is reflected by the
appearance of the M(C,N) precipitates. The microstructure does not
show the typical eutectic characteristics, Figures 3-10a to 3-10r. The
long exposure time of more than 10,000 h has led to a coarsening of the
primary carbides of M23C6. The fraction of M(C,N) plus M23C6 amounts
to 11%. In addition about 4% M6C and 0.4% M(C,N) were formed.
In the metal sub-surface zone the amount of M6C is significantly higher
and due to the uptake of nitrogen some M2(C,N) is observed, Figures
3-10k to 3-10r.
1850-1900F/1010-1038C
1850F/707 h/3.60 ksi (1010C/707 h/24.8 MPa)
This specimen contains a high amount of Si with 1.7 mass-%. The
G-phase, however, could not be detected here. The following precipitates were observed: M23C6 7.8%, M2(C,N) 0.7% and M6C 0.9%.
M(C,N) or their undissolved residues are frequently observed in the core
of M6C particles, Figures 3-11a to 3-11j. In the outermost sub-surface
zone additionally some M2(C,N) is present, Figures 3-11k to 3-11r.
1900F/2,555 h/2.30 ksi (1038C/2555 h/15.8 MPa)
After 2,555 h at slightly increased temperature growth of the M23C6
carbides had taken place. The number of the secondary carbides has
decreased. The volume fraction of M23C6 has increased to 8.4% and that
of M6C to 2.5%, while the fraction of M(C,N) has remained almost
constant with 0.6%, Figures 3-12a to 3-12r. In the outermost zone
some M2(C,N) is also present, which is colored green in the false color
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
at 1900F (1038C) after times higher than 2,500 h from about 8%

to maximum 11%. The G-phase was observed only for one specimen
of this material which was subjected to the lowest test temperature of
1675F (913C) for 659 h. At higher temperatures for specimens with
comparable or even higher Si content the G-phase could not be detected
with confidence.
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
Samples SE-A & SE-B:

These samples were removed from the same furnace after approximately 108,306 hours of service. The available history indicated that the
furnace had 53 cold start-ups and 70 hot restarts over the service life.
Tube samples SE-A & SE-B were removed from different tube segments
and elevations within the furnace resulting in a slight difference in estimated tube skin temperatures (see Table 7).
Post service analysis found that samples SE-A and SE-B both had
maximum diametral growths of less than 1%. Ambient temperature
tensile testing of the aged material found the following:
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).
25
Chemical analysis found deficiencies in micro-alloying and/or

carbon concentrations in many of the tube segments that experienced
large diametral growths. Remaining life assessment via the Omega creep
strain methodology was performed finding the creep strain rates as least
60 times greater than the average for HP micro-alloyed material in the
Materials Property Council (MPC) database. The alloying deficiencies
resulted in the material being highly strain tolerant, likely capable of
15% diametral growth prior to rupture, but with poor creep strength.
The remaining life testing indicated that the material had essentially no
remaining life at the normal operating conditions.
Micrograph of fine, interdendritic cracks and lamellar-like phase.

(oxalic acid electrolytic, 500X)
The plant had a history of operational problems that tended to crush

the reformer tube catalyst and promote channeling of the feed gas. The
tube rupture resulted from a short-term temperature excursion. This
sample was included to show the microstructural features as a result of
incipient melting.
Sample SE-F:
During an outage after 7.5 years of service, diametral growths
in excess of 10% were measured at an isolated area in the furnace.
Metallographic analysis revealed only aligned creep voids with no creep
fissures despite the large diametral growth. During normal operation,
the tube metal temperature was in the 1650-1690F (899-921C)
range. The microstructural features (i.e. carbide coalescence) and review
of historical operating data indicated that the tube metal temperatures
likely exceeded 1800F (982C) at an isolated area of the furnace for
approximately 6 months leading up to the outage.
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-
2.4% diametral growth. Radiography could not confirm the indication.

Subsequent metallography revealed small, linear indications, especially
at the tube ID, where creep voids began to coalesce and align. Creep
voids were present throughout the cross section but with little to no
creep void alignment. Remaining life based on metallography was likely
considerable, perhaps another 50,000 hours at the current design conditions. Caliper and strapping measurements found up to 5.6% diametral
growth over the nominal diameter.
Upon removal from service, tube remaining life assessment was
performed using the Omega creep strain methodology. Even though the
tube had surpassed the design life by nearly 50%, the calculated Omega
value and measured creep strain rate values indicated that the tube still
had 20,000 to 60,000 hours of remaining life at the current operating
conditions. The obtainment of life well in excess of the original design
life was likely due to the operating temperature being approximately
45F (25C) lower than the 1710F (932C) design temperature.
Samples SE-I & SE-J:
Samples SE-I & SE-J were obtained from a tube removed from the
same reformer as sample SE-H (above). However, samples SE-I & SE-J
had increased service time of approximately 143,000 hours. Sample SE-I
was extracted approximately 24 inches (61 cm) above the reformer floor
while sample SE-J was extracted from the same tube but 84 inches (213.4
cm) above the reformer floor. The samples were included in an effort to
observe any microstructural differences resulting from the slight elevation changes within the furnace. Strapping and caliper measurements
indicated that sample SE-I had a diametral growth of 2% while sample
SE-H had 3.5% diametral growth. Tube metal temperature measurements taken at these two elevations indicated a difference of 9 F (5C),
consistent with the diameter change. Omega creep strain remaining life
assessment has not yet been completed in these samples.
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
Tube Skin Temp.

F
1673
1800
1607
1630
1650
1650
1661
1661
1661
2350
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
+
outer zone / edge

M23C6
M6C + M2(C,N) G-Phase
+++
++
+
+++
+
+++
+
+
+++
++
(+)
+++
++
++
4.6
6.2
+++
++
+
7.9
6.2
7.8
6.5
+
(+)
Annotations: (+) very few, + few, ++ many, +++ very many
28
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
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
together, Figures 3-SE-Da and 3-SE-Db. The M23C6 primary carbides

have coarsened and secondary carbides have not been found. In the core
area about 5.5% M23C6 carbides are present besides 1.6% M(C,N),
Figures 3-SE-Da to 3-SE-Dj. The composition of this sample shows
higher Si and Nb contents. In the sub-surface zone 6.2% M6C are found
together with 4.6% M23C6 and the M(C,N) particles had been almost
completely dissolved in M6C (0.2%), Figures 3-SE-Dk to 3-SE-Dr.
1661F/131,000 h/1.98 ksi (905C/131000 h/13.6 MPa)
After about fifteen years of exposure at 1661F (905C) this specimen contains about 7.4% M23C6 primary and secondary carbides of a
coarse structure. Furthermore in the core area as well as in the metal
sub-surface zone about 2.0% M6C, 0.5% M2(C,N) and 0.1% M(C,N)
are present, Figures 3-SE-Ha to 3-SE-Hr. The G-phase was not observed
although the composition of the sample shows high Si and Nb contents.
Possibly the high N content leads to bonding of Nb by the formation
of M2N. In the metal sub-surface zone increasingly Nb and Cr-rich
M2(C,N) are present.
1661F/143,000 h/1.98 ksi (905C/143000 h/13.6 MPa)
After slightly longer exposure of sixteen years under comparable
conditions this specimen contains 6.5% M23C6 primary and secondary
carbides, 2.2% M6C and 1.5% G-phase in the core, Figures 3- SE-Ia- to
3-SE-Ij. M(C,N) could not be detected. The composition of specimen
I shows high Si and Nb contents. Although the N content is high, no
M2N was formed. In the sub-surface zone no G-phase was present. The
fraction of M23C6 primary and secondary carbides amounts to 7.9% and
that of M6C to 6.2%, Figures 3-SE-Ik to 3-SE-Ir.
1661F/143,000 h/1.98 ksi (905C/143000 h/13.6 MPa)
This specimen originates from the same tube as the two foregoing
specimens (H and I), however from a higher position. No differences
in the microstructure were found compared to the specimen exposed
for the same duration and same temperature shown before, Figures 3-
SE-Ja to 3-SE-Jr. Moreover the volume fractions of the precipitates, as

determined by image analysis, are almost the same as for the foregoing
specimen. The chemical composition shows nearly identical contents, as
well, Table 9.
2350F/35,000 h/1.3 ksi (1288C/35000 h/8.96 MPa)
The microstructure of this specimen which had been exposed to
the highest temperature over a duration of four years contains a high
number of interdendritically oriented cracks which reach almost through
the entire wall thickness. The precipitates are about 4% M23C6, mostly
of the primary type, and 1.3% M(C,N) which was rich in Nb and also
contained some Ti, Figures 3-SE-Ea to 3-SE-Ej. Locally clear indications of beginning melting was observed. In these spots a fine lamellar
eutectic M23C6 structure was present which partially contained some
M6C, Figures 3-SE-Ek to 3-SE-Er. In the outer sub-surface zone around
the M23C6 particles and within the dendritic matrix some local depletion features were observed which resulted in local over-etching of the
microstructure, Figures 3-SE-Es to 3-SE-Ez.
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
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).
Precipitation Kinetics Diagrams

The development of the contents (area-%) of the different types of
carbides and phases as a function of exposure time on a logarithmic
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
scale for different temperature ranges are represented in the diagrams in

Figures 1-PKD-a to 1-PKD-d, 2-PKD-a to 2-PKD-c and 3-PKD-a to
3-PKD-d. In the plots, which reveal trends for the increase or decrease
of the contents of the different precipitates, error bars for the standard
deviation are drawn. As these trends are rather complex and the diagrams
are not easy to understand in a further step for each temperature range
the upper and lower limits for each type of precipitate were plotted in
the form of scatter bands. The results are given in Figures 1-PKD-e, 2PKD-d and 3-PKD-e, where the scatter bands are hatched in different
colors for different temperature ranges. These plots show the range of
fractions of MC, M23C6, M6C and Gphase expected for aged alloys of
the types 35/45, HPMA and HP Modified for different temperature
ranges based on the results of the image analysis measurements.
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
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
The financial support by MTI and the very pleasant collaboration

with Galen Hodge, MTI are gratefully acknowledged.
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
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
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
Source: I. Aydin, Ausscheidungsverhalten einiger hitzebestndiger,

chromhaltiger Gulegierungen auf Eisen und Nickelbasis, Dissertation,
RWTH Aachen, 1982
36
3. Precipitation of the G-Phase

The G-phase is an intermetallic silicide of the type Ni16Nb6Si7.
Many of the intermetallic phases do not show a broad range of compositions, some are even socalled line-phases. Presumably stoichiometry
does not allow for major variations concerning the composition also in
the case of the G-phase. Intermetallics precipitate if specific equilibrium
conditions are fulfilled.
Indeed the G- phase was found most frequently in the 35/45 alloy in
a temperature range of 982C-1150C. Possibly the higher Ni-content
in comparison to the HP-MA and HP-Nb alloys beside Nb and Si
favours the G-phase formation. In the HP-MA alloy it was only found
after exposure at 982C / 65664hours. This service exposed sample
contains 1.3 wt-% Si and 0.98 wt-% Nb measured after exposure. For
the HP-Nb alloy the G-phase is only present in the temperature range
of 875-955C. The Si content of these samples varies between 0.9-1.7
wt-% and the Nb content between 0.74-1.2 wt-%. The service exposed
HP-Nb samples show G-phase in a temperature range of 875-905C for
long time exposure of 75000-143000 hours. The wall thickness of these
samples is higher in comparison to the creep rupture
samples. Probably the diffusion of carbide and G-phase forming
elements is much slower at this lower temperature and the diffusion
path of Si to the surface of the sample is longer in samples of larger
dimensions. Si may be present for a longer time for the formation of Gphase than in thinner samples where it may be consumed by oxidation
at the surface.
For the G-phase formation the following informations are known
from the literature:
Ageing of Nb-containing castings results in the replacement of
Cr7C3 carbides by M23C6, secondary precipitates of M23C6 and
NbC, with the replacement of some of the pre-existing NbC by
the G-Phase
The initial volume fraction of NbC and its rate of transformation
to G-Phase were found to increase with the Si content
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
Source: Dr. D. Flahaut, D.T. Pool, 35Cr45Ni Micro-alloy Friend

or Foe, presented at: ERTC Petrochemical Conference, Vienna, 2004
37
38

MTI Interior 1 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MTI Interior 1 PDF

Uploaded by

Copyright:

Available Formats

Atlas of Microstructures

1) Karl-Winnacker-Institut der DECHEMA e.V.

All rights reserved, including translations

Alloy 35/45 Exposure Conditions...................................................... Table 1

Sample Preparation and Microstructural Characterization....................................................... xx

Results of the Microstructural Analyses.................................................................................... xx

- Creep Rupture Samples (continued)....................................................... Fig. 2-5a to 2-5r

- Creep Rupture Samples (continued)....................................................... Fig. 3-8a to 3-8r

Service Exposed Samples.............................................................................................................. xx

Precipitation Kinetics Diagrams............................................................................................... xx

extended aging time and temperature have been well documented by

Table 1: Alloy 35/45 Exposure Conditions

Table 2: Chemistry of Alloy 35Cr/45Ni Stress Rupture Samples

Expo. Time hrs

Table 3: Alloy HPMA Exposure Conditions

Expo. Time hrs

Table 4: Chemistry of Alloy HPMA Stress Rupture Samples

Table 5: Alloy HP Modified Exposure Conditions

Expo. Time hrs

Table 6: Chemistry of Alloy HP Modified Stress Rupture Samples

Scanning Electron Microscopy (SEM-BSE)

Electron Probe Microanalysis (EPMA)

Interference Layer Metallography (LM-ZnSe)

a suitable interference layer. In some cases the layer had to be removed

Image Analysis by the False Color Technique (LM-FC)

Development of the Microstructure by Etching (LM-etched)

outer zone / edge

1800F/1,509 h/3.3 ksi (982C/1509 h/22.8 MPa)

Creep rupture samples

1800F/7,217 h/2.9 ksi (982C/7217 h/20.0 MPa)

are based on an estimation. The primary and secondary M23C6 particles

1900F/10,247 h/1.54 ksi (1038C/10247 h/10.6 MPa)

2012F/3,545 h/1.15 ksi (1100C/3545 h/7.9 MPa)

outer zone / edge

Annotations: + few, ++ many, +++ very many, ++++ abundant

Creep Rupture Samples

which is consistent with the identical compositions and phases formed

After more than 12,000 h a coarsening of the primary carbides and a

1922F/11,778 h/2.08 ksi (1050C/11778 h/14.3 MPa)

2012F/5,580 h/1.45 ksi (1100C/5580 h/10.0 MPa)

area, while this is more pronounced in the metal sub-surface zone. At

Creep Rupture Samples

M2(C,N) G-Phase M(C,N)

outer zone / edge

of the specimen residues of dissolving M(C,N) are frequently detectible

1778F/1,185 h/3.77 ksi (970C/1185 h/26.0 MPa)

1700F/59 h/8.5 ksi (927C/59 h/58.8 MPa)

are about equal to the former specimen.

image in Figure 3-12r.

1800F/7,833 h/5.42 ksi (982C/7833 h/19.3 MPa)

1900F/5,373 h/1.90 ksi (1038C/5373 h/13.1 MPa)

1800F/10,637 h/2.80 ksi (982C/10637 h/19.3 MPa)

at 1900F (1038C) after times higher than 2,500 h from about 8%

Samples SE-A & SE-B:

Chemical analysis found deficiencies in micro-alloying and/or

Micrograph of fine, interdendritic cracks and lamellar-like phase.

The plant had a history of operational problems that tended to crush

2.4% diametral growth. Radiography could not confirm the indication.

Tube Skin Temp.

outer zone / edge

Annotations: (+) very few, + few, ++ many, +++ very many

together, Figures 3-SE-Da and 3-SE-Db. The M23C6 primary carbides

SE-Ja to 3-SE-Jr. Moreover the volume fractions of the precipitates, as