IMPROVING C1023

MANUFACTURABILITY
USING TWO-STEP HEAT TREATMENT

N
The inherent high- ickel-base C1023 super-
alloy (Ni, 15.5% Cr, 9.7%
temperature strength of Co, 8.3% Mo, 4.1% Al) is
superalloys that makes widely used in the man-
ufacture of equiaxed nozzle
these materials ideal for guide vanes and seal seg-
applications at elevated ments for aero gas turbines,
especially in low-pressure
temperatures, such as gas turbines (Fig. 1). However, in
turbine components in spite of its extensive applications
in many different types of engines, Representative
aero gas turbine engines, this alloy is considered difficult to aero gas-turbine
engine.
works against the weld, repair, and machine (al-
though it is considered weldable).
fabricator in The high strength of this alloy is
manufacturing the parts. mainly due to the presence of a fine investigation addressed heat treat-
distribution of J’ (gamma prime) pre- ments that would achieve opposite
Thermal processing is cipitates that are formed directly after results. That is, first, it is necessary to
required that puts the casting. The quick formation of these achieve a heat treatment capable of
precipitates does not allow pro- softening the material to improve all
material in a workable cessing of C1023 components in a soft the aspects related to its manufac-
state and then reverts it condition, such as that of other su- turability. After this, the development
peralloys in a solution annealed state. of a second heat treatment that
back to its high strength However, it is desirable to achieve would be introduced at the end of the
condition for service. some kind of temporary soft condi- manufacturing process is necessary
tion that could improve the manu- to recover the original high strength
facture of C1023 components. of the material.
This article discusses this alterna-
Iñigo Hernández*,
tive with a detailed study on the re- C1023 Microstructural Features
Amaia Subinas, sponse of C1023 material to different C1023 is a face centered cubic
Dr. Iñaki Madariaga, heat treatments. The focus of the (FCC) superalloy strengthened by a
and Dr. Koldo Ostolaza
Industria de Turbo Propulsores (ITP)
Materials & Processes Dept.
Zamudio, Spain

*Member of ASM International Fig. 1 — Nozzle guide vanes.

HEAT TREATING PROGRESS • MAY/JUNE 2007 25

 fine distribution of J’ precipi- • As cast + TT1190ºC/1h furnace
tates Ni3(Al, Ti). Most of the J’ cool
precipitates have sizes around • As cast + TT1190ºC/1h water
1 Pm, but this material can also quench
have submicron sized J’ precip- • As cast + TT1190ºC/1h slow cool
itates. MC, M23C6, and MC6 car- (5ºC/min) to 1000ºC and then gas
bides are formed during the so- fan quench
lidification along the grain • As cast + TT1190ºC/1h gas fan
boundaries with M being typi- quench + TT1100ºC/1h slow cool
Fig. 2 — Typical C1023 microstructure
containing Mo (white) and Ti and Mo (black) cally Mo and Ti. Figure 2 shows a rep- to 700ºC and then gas fan quench
carbides; fine Jc is distributed in the J matrix. resentative C1023 microstructure. In • As cast + 1100ºC/8h furnace
2000 × some cases, P phase, which is consid- cooling
ered detrimental, can appear if the • As cast + 1100ºC/8h water
solidification of the casting is too quench
slow in the temperature range of 800 A key factor in this study is that
and 900ºC (1470 and 1650ºF). there are no visible J’ precipitates in
The typical material features such the microstructures of samples
as grain size, carbides, J’ and hard- heated at 1190ºC/1 hour and water
ness can be influenced by heat treat- quenched (Fig. 4). This means that
ment. However, in the range of tem- complete solution of the J’ occurs at
peratures where this material has this temperature. Therefore, it is pos-
been studied in the present work, the sible to return to the standard mi-
main microstructural changes have crostructure if at the end of the man-
Fig. 3 — Sample microstructure in the as
been observed in the J’ size and dis- ufacturing process a stress relief/
cast condition consists of a dendritic mi-
crostructure with fine Jc precipitates (Jcis tribution and, to a lesser extent, in the solution heat treatment is carried
about 0.5 Pm in the interdendritic regions) volume fraction of carbides. out. However none of the heat treat-
The typical service condition of ments carried out holding the mate-
this material is as cast plus a stress re- rial at 1190ºC and modifying the
lief treatment carried out at a temper- cooling rate was capable of pro-
ature between 1150 and 1190ºC (2100 ducing the desired coarsening of the
and 2175ºF). This heat treatment is J’ and then the intended softening
usually conducted to relieve the of the material.
stresses produced during the casting The analysis of the samples heat
process. treated at 1100ºC for 8 hours followed
by slow cooling under high magni-
Heat Treatment Trials fication (20,000×) shows the forma-
The high strength of this alloy is tion of two different types of J’ pre-
Fig. 4 - Sample microstructure in 1190ºC
heat treated and water quenched condition. mainly due to the presence of a fine cipitates (Fig. 5). However, as shown
Nothing is resolvable with the SEM because distribution of J’ precipitates that are in the sample quenched after 8 hours
Jc has gone into solution. formed directly after casting (Fig. 3). at 1100ºC (Fig. 6), it is clear that the
Therefore, to achieve some kind of fine precipitates were not formed
Primary J ’ temporary soft condition that could during the 8 hour hold at 1100ºC, and
improve the manufacture of C1023 instead, must be the result of the pre-
components, it is necessary to
achieve coarser J’ precipitates by con-
trolling the cooling rate or overaging
Secondary J ’ in the J’ formation temperature
range. A second treatment that
would be able to dissolve the J’ and
precipitate it again with a fine distri-
bution to restore the original proper-
Fig. 5 — Sample microstructure in the as
cast + 1100ºC/8 hours and slow cooled con- ties would also be necessary.
dition. A very fine distribution of secondary With this purpose in mind, mi-
gamma prime exists between the primary pre- crostructural studies were carried out
cipitates that precipitate during slower on specimens in the as cast condition Fig. 6 — Sample microstructure in the as
cooling. and after various heat treatments. cast + 1100ºC/8 hours and water quenched
condition. Quenching from 1100ºC avoids
The different conditions included: the precipitation of the secondary gamma
• As cast prime, which means the precipitates of the
• As cast + TT1190ºC/1h gas fan sample in Fig. 5 formed during slow cooling
quench from 1100ºC.

26 HEAT TREATING PROGRESS • MAY/JUNE 2007

cipitation occurring during the slow HIP Study very slow cooling rate such as that
cooling from 1100ºC. Thus, it appears During this heat treatment study, associated with HIP would be able
that a partial dissolution of the J’ oc- C1023 material was also the focus of to achieve the coarsening of the J’ that
curs at 1100ºC, but no coarsening ef- an investigation directed toward the was not achieved with the heat treat-
fect occurs. reduction of the shrinkage formed on ments investigated.
Therefore, none of these heat treat- C1023 components during the Using as a basis the standard HIP
ments is capable of producing the de- casting process. The material was conditions of 1200ºC for 4 hours and
sired softening of the material. This processed by hot isostatic pressing 100 MPa (14,500 psi) pressure, labo-
was confirmed not only by means of (HIP), a process used to help close the ratory samples were processed using
microstructural analysis but also internal porosity of castings, and a variety of thermal cycle combina-
with the microindentation hardness widely used in aircraft gas turbine tions including:
results, which are similar in all the industry. 1. As cast
samples except the material The cooling rate of the process is 2. As cast + 2HT
quenched in water (425 HV1). The very slow from the very high temper- 3. As cast + HIP without pressure
hardness value of the rest of the sam- atures of the HIP cycles, in this case (NP)
ples is about 380 HV1. 1200ºC (2190ºF), and such a cooling 4. As cast + HIP (NP) + HT
The results of the first trials led to rate would most likely lead to a slight 5. As cast + HT + HIP (NP)
the exploration of using HIP (hot iso- modification of the microstructure 6. As cast + HT + HIP (NP) +HT
static pressing, or sometimes pro- and the properties of C1023. Based 7. As cast + HIP
cessing) cycles, taking into account on the results of the heat treatment 8. As cast + HT + HIP
that the cooling rate of the HIP studies, it was believed that a holding 9. As cast + HT + HIP +HT
process is very slow from a very high period at 1200ºC would be able to The first part of the study covered
temperature. fully dissolve the J’. Afterward, a the cycles 1 to 6 described. At this
1. As cast 2. As cast + 2 HT 3. As cast + HIP

4. As cast + HIP + HT 5. As cast + HT + HIP 6. As cast + HT + HIP + HT

7. As cast + HIP 8. As cast + HT + HIP 9. As cast + HT + HIP + HT

Fig. 7 — SEM photomicrographs of samples in various conditions show little difference in microstructure (samples 3-6 HIPed without pres-
sure; samples 7-9 HIPed with pressure). There are differences in hardness among the samples possibly due to the presence of very fine Jc precip-
itates that are not resolvable at the SEM magnification of 5,000×.

HEAT TREATING PROGRESS • MAY/JUNE 2007 27

Table 1 — Hardness comparison after different HIP cycle combinations
Trial Condition of sample Microindentation hardness, HV1 Sample Ref. No.
1 As cast 374.6 ±6.4 S-067-05A
2 As cast + 2 HT 378.9 ±6.3 S-067-05B
3 As cast + HIP (no pressure) 347.4 ±6.4 S-067-05C
4 As cast + HIP (no pressure) + HT 371.8 ±6.2 S-067-05D
5 As cast + HT + HIP (no pressure) 341.8 ±7.4 S-067-05E
6 As cast + HT + HIP (no pressure) + HT 379.5 ±6.3 S-067-05F
7 As cast + HIP (with pressure) 356.5 ±6.5 S-067-05G
8 As cast + HT + HIP (with pressure) 358.4 ±6.5 S-067-05H
9 As cast + HT + HIP + HT 371.9 ±5.7 PCB ref 05/0637

stage of the study, it was considered Cycles 7 to 9 were conducted in the duction parts.
that the influence of the HIP pressure second part of the study. These cy- The heat treatments were analyzed
on the microstructural evolution of cles were carried out on samples ex- by comparing the microstructures
the material would be negligible and tracted from components that had obtained after each cycle and also by
all the cycles were conducted without been HIPed in production furnaces, measuring microindentation hard-
pressure. The solution treatment at in an attempt to determine if the mi- ness. The results of all combinations
1190ºC/1h/slow cooling to 1000ºC crostructural changes observed on of heat treatments and HIP cycles are
and then gas fan quench is defined laboratory scale furnaces and sam- presented in Fig.7, and the correspon-
as HT. ples could also be observed on pro- ding hardness values are listed in
Table 1. There is not a great deal of
variation in the microstructures of
SR 850°C no HIP
SR 850°C HIP + HT
the different treatments, but hardness
SR 850°C HT + HIP + HT decreases for samples that were
SR 850°C database typical processed using HIP cycles (with or
SR 850°C database minimum
without pressure), and that did not
receive a final heat treatment at
Arbitrary units

1190ºC. In the particular case of sam-

ples 3 and 5, the obtained hardness
value is below 350 HV compared
with the original values of 380 HV for
the as cast material and the as cast +
stress relieved material. The results
also demonstrate that although HIP
produces a decrease in hardness, the
0 200 400 600 800 1,000 1,200 original hardness of the as cast ma-
Time to rupture, h
terial could be recovered by treating
Fig. 8 — Relative stress rupture time of HIP and non-HIP material. the material at 1190ºC.
The differences between the dis-
No HIP
similar hardness values can not be
HIP + HT explained on the basis of the mi-
HT + HIP + HT crostructural features observed on
Database typ. the material at SEM resolution. For
Database minimum example, Fig. 7 does not show any
clear microstructural differences be-
Arbitrary units

tween any of the samples analyzed.

However, apart from the micron-
sized precipitates that can be seen
using SEM, this material also has
much smaller J’ precipitates that can
not be resolved using SEM, and that
can explain the differences among
the samples.
0 200 400 600 800 1,000 1,200
Temperature, °C Mechanical Testing
Fig. 9 — Relative ultimate tensile strength of HIP and non-HIP material. The results above show it is pos-
30 HEAT TREATING PROGRESS • MAY/JUNE 2007
sible to achieve a softening of C1023 cooling, which is able to restore its For more information:
by heat treatment and then recover original properties. Iñigo Hernández
the original properties by performing Industria de Turbo Propulsores, ITP, S.A.,
a second heat treatment. To verify Conclusions Parque Tecnológico
Edificio 300
these results, mechanical tests were Test results show that it is possible
48170 Zamudio, Spain
conducted, including tensile, stress to achieve a significant reduction in
tel: 34 94 466 2924
rupture, low cycle fatigue (LCF), and the hardness of C1023 after 4 hours e-mail: inigo.hernandez@itp.es
impact tests. Figs 8 and 9 show the at 1200ºC followed by controlled Internet: www.itp.es
result of the stress rupture and ten- slow cooling. A starting microinden-
sile tests conducted on samples tation hardness of 375 HV was re-
processed with and without the HIP duced to 341 HV on laboratory sam-
cycle. In both cases, the results ob- ples and to 356 HV on real
tained with and without HIP are components. This suggests it could
comparable. be possible to improve the machin-
LCF and impact tests also show the ability and weldability of this alloy
same trend observed in the stress after HIP. The application of an
rupture and tensile tests with very 1190ºC heat treatment for 1 hour is
little differences among sample able to fully dissolve the coarse pre-
processed with and without HIP, cipitates formed during HIP, and it
with all the results within the typical is possible subsequently to form a
scatter of properties of this material. fine distribution of the precipitates,
Therefore, while it seems the mate- which restores the original proper-
rial is softer immediately after the ties. Results of tensile, stress rupture,
HIP treatment (as indicated by the impact, and hardness tests are better
reduced hardness), the final heat after HIP + HT than those in the as
treatment is capable of fully dis- cast + HT condition. These results
solving the J’ and producing an im- confirm the restoring effect of
proved precipitation pattern upon the 1190ºC heat treatment.

Proven and Reliable

Combustion
Solutions for
Heat Treating

i As the North Our extensive

American source for product lines can be
Kromschroder burners used for direct and
and recuperative indirect fired
burners like the applications.
Ecomax from LBE, in
addition to our own Our high velocity
SVG line, Hauck has burners combined
an extensive array of with pulse firing
burners, control control technologies
strategies, and total will improve your
system solutions for product quality,
the heat treating reduce emissions
industry. and optimize fuel
efficiency.

Hauck Manufacturing Company, PO Box 90, Lebanon, PA 17042

Phone: 717-272-3051 Fax: 717-273-9882

To Earn our Customer’s Loyalty - Every Day www.hauckburner.com

HEAT TREATING PROGRESS • MAY/JUNE 2007 31