NASA / TM--2002-211796
Characterization of the Temperature
Capabilities of Advanced Disk Alloy
Timothy P. Gabb and Jack Telesman
Glem_ Research Center, Cleveland, Ohio
Peter T. Kantzos
Ohio Aerospace Institute, Brook Park, Ohio
.-a i -_K_,m-teth O Com_or
Gleru_ Research Center, Cleveland, Ohio
ME3
August 2002
https://ntrs.nasa.gov/search.jsp?R=20020081280 2020-03-07T13:40:53+00:00Z
The NASA STI Program Office... in Profile
Since its founding, NASA has been dedicated to
the advancement of aeronautics and spacescience. Tile NASA Scientific and Technical
Information (STI) Program Office plays a key part
in helping NASA maintain this important role.
Tile NASA STI Program Office is operated by
Langley Research Center, the Lead Center forNASA's scientific and technical information. The
NASA STI Program Office provides access to the
NASA STI Database, the largest collection of
aeronautical and space science STI in file world.
The Program Office is also NASA's institutional
medlanism for disseminating the results of its
researd3 and development acti vities. These results
are published by NASA in the NASA STI Report
Series, which includes the following report types:
TECHNICAL PUBHCATION. Reports of
completed :research or a major significant
phase of research that present the results of
NASA programs and include extensive data
or theoretical analysis. Includes compilations
of significant scientific and technical data and
information deemed to be of continuing
reference value. NASA's counterpart of peer-
reviewed formal professional papers but
has less stringent limitations on manuscript
lengfl3 and extent of graphic presentations.
TECHNICAL MEMORANDUM. Scientific
and tedmical findings that are pre, liminary or
of specialized interest, e.g., quick release
reports, working papers, and bibliographiesthat contain minimal annotation. Does not
contain extensive analysis.
CONTRACTOR REPORT. Scientific and
technical findings by NASA-sponsored
contractors and grantees.
CONFERENCE PUBLICATION. Collected
papers from scientific and technical
conferences, symposia, seminars, or other
meetings sponsored or cosponsored byNASA.
SPECIAL PUBLICATION. Scientific,
technical, or historical information from
NASA programs, projects, and missions,
often concerned with subjects having
substantial public interest.
TECHNICAL TRANSLATION. English-
language translations of foreign scientific
and technical material pertinent to NASA'smission.
Specialized services that complement the STI
Program Office's diverse offerings include
creating custom thesauri, building customized
data bases, organizing and publishing research
results.., even providing videos.
For more information about the NASA STI
Program Office, see the following:
® Access the NASASTI Program Home Page
at http:lhuww.sti.nasa.gov
® E-mail your question via the Intemet to
* Fax your question to the NASA Access
Help Desk at 301-621-0134
* Telephone the NASA Access Help Desk at301-621-0390
Write to:
NASA Access Help Desk
NASA Center for AeroSpace Information7121 Standard Drive
Hanover, MD 21076
NASA/TM--2002-211796
Characterization of the Temperature
Capabilities of Advanced Disk Alloy
Timothy R Gabb and Jack Telesman
Glem_ Research Center, Cleveland, Ohio
Peter T. Kantzos
Ohio Aerospace Institute, Brook Park, Ohio
.-a i -_K_,m-teth O Com_or
Glepa_ Research Center, Cleveland, Ohio
ME3
National Aeronautics and
Spa ce Administration
Glelm Research Center
August 2002
Acknowledgments
The authors gratefully acknowledge the support of: the NASA GRC Ultra Efficient Engine Technology Program,
managed by Robert J. Shaw, Ajay Misra, and Robert Draper. The support of the NASA GRC Materials Division andStructures Division, managed by Hugh Gray and James Kiral?; is also acknowledged. The authors also wish to
acknowledge the many helpful discussions with Kenneth Bain, ]on Groh, Robert Vanstone, and David Mourer,General Electric Aircraft Engines, and Paul Reynolds, Pratt & Whitney: Subscale disk forgings and heat treatments
were performed at Wyman-Gordon Forgings under the direction of William.
NASA Center for Aerospace Information71121Standard Drive
Hanover, MD 211076
Available frorn
National Technical Information Service
5285 Port Royal RoadSpringfield, VA 22100
Available electronically at http://gltrs.zrc.nasa.gov
Characterization of the Temperature Capabilities ofAdvanced Disk Alloy ME3
Timothy P. Gabb and Jack Telesman
National Aeronautics and Space AdministrationGlenn Research Center
Cleveland, Ohio 44135
Peter T. Kantzos
Ohio Aerospace InstituteBrook Park, Ohio 44142
Kenneth O'Connor
National Aeronautics and Space AdministrationGlenn Research Center
Cleveland, Ohio 44135
Abstract
The successful development of an advanced powder metallurgy disk alloy, ME3,
was initiated in the NASA High Speed Research/Enabling Propulsion Materials
(HSR/EPM) Compressor/Turbine Disk program in cooperation with General Electric
Engine Company and Pratt & Whitney Aircraft Engines. This alloy was designed using
statistical screening and optimization of composition and processing variables to have
extended durability at 1200 °F in large disks. Disks of this alloy were produced at the
conclusion of the program using a realistic scaled-up disk shape and processing to enable
demonstration of these properties. The objective of the Ultra-Efficient Engine
Technologies disk program was to assess the mechanical properties of these ME3 disks
as functions of temperature, in order to estimate the maximum temperature capabilities of
this advanced alloy. These disks were sectioned, machined into specimens, and
extensively tested. Additional sub-scale disks and blanks were processed and selectively
tested to explore the effects of several processing variations on mechanical properties.
Results indicate the baseline ME3 alloy and process can produce 1300-1350 °F
temperature capabilities, dependent on detailed disk and engine design property
requirements.
Introduction
The advanced powder metallurgy disk alloy ME3 was designed in the NASA
HSR/EPM disk program to have extended durability at 1200 °F in large disks. This was
achieved by designing a disk alloy with moderately high 7' precipitate content and
refractory element levels, optimized with supersolvus solution heat treatments to produce
balanced monotonic, cyclic, and time-dependent mechanical properties. The resulting
baseline alloy, processing, and supersolvus heat treatment has shown extended durability
capabilities, combined with robust processing and manufacturing characteristics (ref. 1).
NASA/TM--2002-211796 1
There is a long-term need for disks with higher rim temperature capabilities of
1300 °F or more. This would allow higher compressor exit (T3) temperatures and allow
the full utilization of advanced combustor and airfoil concepts under development. The
balance of mechanical properties necessary to achieve these temperature capabilities
could vary with engine size and engine cycle design, as well as the particulars of a
selected potential disk design and location in an engine. Such detailed preliminary and
detailed design assessments are beyond the scope of this study. However, a general
characterization of the mechanical properties of ME3 as functions of temperature would
allow initial assessments of the balance of properties produced by the current baseline
processing conditions and how these properties would impact such advanced
applications.
The objective of this study was to assess the mechanical properties of ME3 as
functions of temperature. This would enable assessments of the maximum temperature
capabilities of this disk alloy for different potential applications in the engine community.
Scaled-up disks processed in the HSR/EPM Compressor/Turbine Disk program were
sectioned, machined into specimens, and extensively tested in tensile, creep, fatigue, and
fatigue crack growth tests by NASA Glenn Research Center (GRC). Additional sub-
scale material was processed and selectively tested to explore the effects of several
processing variations on mechanical properties.
Materials and Procedure
Twelve scaled-up baseline ME3 disks were either subsolvus or supersolvus
solution heat treated (ref. 1). They were then removed for brief fan air cooling followed
by oil quenching. Subsequent stress relief heat treatment and aging heat treatment steps
were then applied. These disks each had an outer diameter of near 24 in., maximum bore
thickness of near 4 in., and rim thickness of near 2 in.
A remnant section of a scaled-up ME3 extrusion used for the scaled-up disks was
machined to mults 3.5 in. dia. and 7 in. long, then forged into 15-20 pound sub-scale
disks about 5-7 in. in diameter and 1.6 in. thick by Wyman-Gordon Forgings. Specimen
blanks were machined using electro-discharge machining from one forging before heat
treatment. The other disks were heat treated at Wyman-Gordon Forgings, Houston Div.,
Research & Development Shop. Solution heat treatment complexity and soak time
effects were studied in the ME3 subscale disks and blanks. They were either given a
simple, short "direct heatup" (DH) supersolvus heat treatment or a longer, two-step "pre-
annealed" (PA) heat treatment sequence of subsolvus pre-anneal+ supersolvus solution
heat treatment (ref. 2). Stress relief heat treatment and aging heat treatment steps were
then applied to these two subscale disks. Two additional disks were given a DH solution
heat treatment then a single step combined stress relief/aging (CSRA) heat treatment,
designed using stress relaxation test data to be presented. The effects of the stress relief
heat treatment step were further explored using subscale blanks. Selected blanks were
given a stress relief heat treatment followed by the aging heat treatment, while other
blanks were just directly aged after the solution heat treatment. Additional blanks were
machined into stress relaxation specimens after just the solution heat treatment, in order
to study stress relaxation occurring during potential stress relief and aging heattreatments.
NASA/TM--2002-211796 2
It was intended that the subscale disks and blanks be quenched from the solution
heat treatments at cooling rates typically expected at near-surface to deeply imbedded
locations of large disks of several hundred pounds weight. Due to the much lower weight
and volume of the subscale disks and blanks, this required design and screening of slower
cooling procedures than typically employed for large disks. A procedure of fan air
cooling starting 2 minutes after removal from the furnace was adopted for most of the
subscale disks. An additional DH+CSRA disk was directly oil quenched starting 1
minute after removal from the furnace, to simulate faster cooling rates near the surfaces
of large disks. The cooling procedure selected for the blanks was more complicated due
to their very low mass and rapid cooling tendencies. A small resistance heating box
furnace having a translating platform was used lower the blanks out of the hot zone at a
controlled rate. The cooling temperature-time data of thermocouples embedded in the
middle ("bore") and near the corner ("rim") of a subscale disk and in the middle of a
blank are compared in Fig. 1. The temperature-time paths of cooling measured in the
subscale disks was similar to that expected for large disks. The cooling path of the
blanks closely reproduced that of the bore location of the subscale disks. The
thermocouple temperature-time data recorded from 4 thermocouples embedded in one of
the subscale disks during fan air and oil quenching cycles was analyzed using a
commercial heat transfer computer code in order to assign approximate cooling rates,
averaged over the temperature range of solution temperature to 1600 °F, for each
extracted specimen.
An extensive mechanical testing matrix was employed for the scaled-up disks
including tensile, notched tensile, creep, low cycle fatigue, and fatigue crack growth tests.
Tensile tests were performed from 75 to 1500 °F on both supersolvus and subsolvus heat
treated disk material. Other mechanical property tests were only performed on the
supersolvus heat treated material. Stress relaxation tests were performed from 1400 to
1600 °F. Creep tests were performed from 1200 to 1500 °F. Low cycle fatigue tests
were performed from 75 to 1400 °F. Cyclic crack growth tests were performed from 75
to 1500 °F, while dwell crack growth tests were performed from 1200 to 1400 °F.
Mechanical test conditions of subscale disks and blanks were selected from among these
conditions to allow direct comparisons with specimen tests from the scale-up disks.
Tensile Tests. Machining and testing of scaled-up disk tensile specimens was
performed by Dickson Testing Company. Specimens having a gage diameter of 0.25 in.
and gage length of 1.25 in. were machined and then tested in uniaxial test machines
employing induction heating and axial extensometers. The tests were performed
according to ASTM E21, using an initial test segment with strain increased at a uniform
rate of 0.2%/min., followed by a segment with displacement increased at a uniform rate
of 0.2 in./min. Tests of subscale material were performed at Dickson Testing Company
and GRC on specimens machined by Metcut Research Associates having a gage diameter
of 0.16 in. and gage length of 1 in. in a uniaxial test machine employing a resistance
heating furnace and axial extensometer according to E21. Additional tensile specimens
from subscale material were first subjected to exposures in air at 1400 and 1500 °F.
About 0.020 in. was removed from the gage diameter of some of these specimens after
exposures. Then all were tested at their exposure temperature. Notch tensile tests of
specimens with a minimum gage diameter of 0.25 in. and notch stress concentration
factor Kt =3.5 were performed at Dickson Testing Company according to E602.
NASA/TM--2002-211796 3
Notched tensile tests of subscale material were performed at Dickson Testing Company
and GRC on specimens machined by Metcut Research Associates having a minimum
gage diameter of 0.16 in. and stress concentration factor Kt=3.5 in a uniaxial testmachine.
Stress Relaxation Tests. Specimens having a gage diameter of 0.16 in. and gage
length of 1 in. were machined from supersolvus solution heat treated subscale blanks and
then tested at GRC in a uniaxial test machine employing resistance heating and an axial
extensometer. The tests were performed in general accordance with E328, using an
initial test segment having strain increased at a uniform rate of 0.2%/min., with the strainthen held constant at 1.0% to allow stress relaxation for 8-24 hours.
Creep Tests. Machining of scaled-up disk creep specimens was performed by
Metcut Research Associates. Specimens having a gage diameter of 0.25 in. and gage
length of 1.5 in. were machined and tested in uniaxial lever arm constant load creep
frames using resistance heating furnaces and shoulder-mounted extensometers. The
creep tests were performed by GRC, Metcut, and Mar-Test, Inc. according to ASTM
E139. Creep specimens of subscale material were machined and tested at Metcut. These
specimens having a gage diameter of 0.16 in. and gage length of 0.75 in. were tested in
constant load creep frames each using a resistance heating furnace and extensometer
attached to the specimen gage section. Creep-rupture specimens of subscale disks having
both a smooth gage section 0.16 in. diameter and 0.75 in. long, and a notched section of
0.16 in. notch dia. were machined by Metcut and tested at NASA GRC.
Low Cycle Fatigue Tests. Machining from scaled-up disks of low cycle fatigue
specimens having gage diameters of 0.4 in. and gage lengths of 1.25 in. was performed
by BITEC CNC Production Machining. The low cycle fatigue (LCF) specimens were
then tested at Mar-Test, Inc. using uniaxial closed-loop servo-hydraulic testing machines
with induction heating and axial extensometers. Tests were performed according to
ASTM E606. A frequency of 0.5 hertz was employed in strain-controlled fatigue testing
for the first 24 hours of cycling. Strain ratios (R,=emao,/emin) of 0.5, 0, and -1 were used.
Surviving specimens were then cycled to the same stabilized stresses using a load-
controlled cycle at a faster frequency of 5 hertz until failure. LCF specimens having gage
diameters of 0.25 in. and gage lengths of 0.75 in. were machined from the subscale disks
by BITEC and tested at Mar-Test, Inc. using the same procedures. Additional LCF
specimens from subscale material were first subjected to exposure in air at 1400 °F for
500h. They were then all tested at 1400 °F.
Fatigue Crack Growth Tests. Machining of surface flaw fatigue crack growth
specimens (ref. 3) from scaled-up disks was performed by Low Stress Grind, Inc.
Machining of specimens of the same configuration from subscale disks was performed by
BITEC. All specimens had a rectangular gage section 0.4 in. wide and 0.18 in. thick,
with a surface flaw about 0.014 in. wide and 0.007 in. deep produced by electro-
discharge machining. The fatigue crack growth specimens were then tested at NASA
GRC. Tests were performed in a closed-loop servohydraulic test machine using
resistance heating and potential drop measurement of crack growth. Pre-cracking was
performed at room temperature. Tests were then performed at elevated temperatures
using a maximum stress of 100 ksi. Cyclic tests were performed at a frequency of
0.33 hertz. Various stress ratios (Rc,=OnJOmax) were used in the cyclic tests. Dwell tests
NASA/TM--2002-211796 4
were performed with various times of dwelling at maximum stress in each cycle, usingstress ratios of 0 or 0.05.
Fracture surfaces of selected specimens were evaluated by scanning electron
microscopy. Cracking modes and grain sizes were also examined on metallographically
prepared sections. Grain sizes were determined according to ASTM E112 linear
intercept procedures using circular grid overlays, and As-Large-As (ALA) grain sizes
were determined using ASTM E930.
Results and Discussion
Typical Microstructures
Typical grain microstructures in optical images of etched metallographic sections
of tensile specimen grip sections are shown in Fig. 2. These tensile specimens were from
the disks' rim regions, which cooled more quickly during quenching than the bore
sections. Supersolvus heat treated scaled-up material had a mean grain size of ASTM 7.1
(27.5 _tm), with a standard deviation of ASTM 0.2 (2.0 _tm) and ALA grain size rating of
ASTM 3.25. Subsolvus heat treated scaled-up material had a mean grain size of ASTM
12.0+/-0.1 (5+/-0.2 _tm) and ALA grain size of ASTM 8. Typical 7' precipitate
microstructures in transmission electron microscopy superlattice darkfield images from
thinned foils of tensile specimen grip sections are also shown in Fig. 2. Within the grains
of supersolvus specimens, three populations of y' precipitates were evident. Scattered
large precipitates (0.3-0.5 _tm diameter) appeared to have preferentially grown at the
cube corners, giving consistently oriented star shapes. Selected area diffraction pattern
analyses indicated the cube sides corresponded to {001 } planes, while the extended cube
corners grew out in <111> directions, as previously reported elsewhere (ref. 4).
Intermediate size precipitates (0.15-0.3 _tm diameter) had a simpler, rounded cube shape.
Fine precipitates (0.01-0.05 _tm diameter) were spherical.
Subsolvus specimens had less distinct differences in large versus intermediate
precipitate morphology and size ranges, but still displayed some evidence of preferential
growth at the cube corners. The fine spherical precipitates were somewhat smaller in
subsolvus specimens. Coarse, undissolved "primary" y' particles (0.6-2 _tm in diameter)
were spaced along grain boundaries and sometimes widely scattered within grains.
Tensile Stress-Strain Response
The stress-strain curves of typical tensile tests are shown in Figs. 3-4. Both
supersolvus and subsolvus specimens had serrated plastic flow at intermediate
temperatures, pea_king at 800 °F then subsiding at higher temperatures. At temperatures
of 1400 °F and higher, initial peak strengths were usually attained at the slow initial
testing strain rate, followed by plastic softening to lower stresses. These tests then
generated higher stresses and a higher ultimate strength when switched to a faster
constant displacement rate in the second test segment, as shown in Figs. 3-4. This
indicated that the strength was strain rate dependent, and decreased with decreasing strain
rate at these temperatures. This variation of strength with strain rate is not usually
encountered in current disks which mn at lower temperatures, and such strength
variations could present design challenges at these higher temperatures. The strain rate
NASA/TM--2002-211796 5
sensitivity (m) of strength for these temperatures was estimated by linear regression using
the general equation (ref. 5):
o=K'(de/dt)m ; log o = log K' + m log(de/d0
Strain rate sensitivity increased with temperature, and tended to be slightly higher for
subsolvus material than supersolvus material, as shown in Fig. 5. Yield strengths at 0.2%
offset, ultimate strengths, notched strength, % elongation after failure, and % reduction in
area after failure are compared as functions of temperature in Figs. 6-8. Polynomial
regression was performed on these responses using temperature (T), T 2, and T 3 as the
independent variables. The resulting equations and correlation coefficients are listed in
the figures, for use in estimating mean strengths and ductilities. Yield strength was
sustained to a temperature of 1300 °F, then dropped oft" with increasing temperature.
Ultimate strength began dropping oft" above 1200 °F. Specimens extracted from the disk
rims usually had higher strengths than those from disk bores, possibly due to the higher
cooling rates expected in rims (ref. 6). Elongation and reduction in area did not
significantly vary as functions of temperature for supersolvus heat treated material.
Test results of specimens from supersolvus heat treated subscale disks and blanks
are shown in Figs. 9-14. The subscale material had comparable tensile properties to the
baseline scaled-up disks, for the DH and PA solution heat treatments with baseline stress
relief plus aging heat treatments. The blanks given the standard aging heat treatment
without the stress relief step also had comparable response. The DH solution with CSRA
combined stress relief/aging heat treatments gave 5-10 ksi higher strength at the highest
temperatures than the scaled-up disk specimens, with the oil quenched subscale disk
giving highest strengths.
Yield and ultimate strength of the subscale disk specimens are shown versus
approximate cooling rate in Fig. 14. Increasing cooling rate consistently increased
strength, as previously reported (ref. 6). Yield strength was usually more strongly
increased by cooling rate than ultimate strength. The effects generally decreased with
increasing test temperature from 1100 °F to 1500 °F. Simple linear regression equations
are included for estimating cooling rate effects on mean response.
The tensile properties of this alloy could be affected by service time at the
projected advanced disk operating temperatures of 1400 °F and higher. In order to
briefly assess these effects, groups of fully machined tensile specimens were exposed at
1400 °F/500 h and 1500 °F/600 h. The gage sections of some of the specimens were re-
machined after exposure to remove the oxidized surface layer, then all specimens were
tensile tested at their exposure temperatures. The resulting yield strengths, ultimate
strengths, elongations, and reductions in area are compared for specimens of low and
high average cooling rates in Fig. 15. After 1400 °F/500 h exposure, yield strength was
reduced by less than 5 ksi while ultimate strength was increased by 3-5 ksi. There was
no consistent effect on reduction in area, and machining away the oxidized surface layer
did not consistently change these results. These results suggest that extended service at
1400 °F would not substantially degrade strength or ductility due to volume-dependentmicrostructural effects or near-surface oxidation effects.
After 1500 °F/600 h, yield strength was reduced by 15-25 ksi, while ultimate
strength was reduced by 13-20 ksi. The strength effects were greater for material having
NASA/TM--2002-211796 6
slow (near 112 °F/min) average cooling rates. Machining away the oxidized surface
layer increased exposed strengths by only 2-3 ksi. Reduction in area after this exposure
was more sharply reduced to 30-50% of unexposed values. The effects on ductility were
greater for material having fast (near 160 °F/min) average cooling rates. Machining away
the oxidized surface layer increased reduction in area of exposed specimens to 50-75%
of unexposed values. These results suggest that extended service at 1500 °F could
sharply reduce strength primarily due to volume-dependent microstructural effects, and
sharply reduce near-surface ductility due to oxidation as well as microstructural effects.
Typical tensile fracture surfaces are compared in Fig. 16. Tensile specimens had
a predominantly transgranular failure mode by microvoid coalescence in tests from room
temperature to 1300 °F. At intermediate temperatures, scattered slip "facet" grain
failures were also observed. At higher temperatures of 1400-1500 °F, oxidized
intergranular surface cracks appeared to precede the transgranular microvoid coalescence.
Stress Relaxation Response
Stress versus time in typical stress relaxation tests at 1400 to 1600 °F are shown
in Fig. 17. The rate of stress relaxation decreased with increasing time, such that stress
decreased linearly with log(time). Relaxation increased with increasing temperature as
expected. Multiple linear regression was performed on stress versus log (time) and
temperature (P-to-enter=0.05). The resulting equation and correlation coefficient are
listed in the figure, for use in estimating mean stress relaxation response. This equation
showed the strong temperature dependence of stress relaxation, and indicated the
temperature dependence was enhanced at higher values of log(time). These results
indicated a combined stress relief/aging (CSRA) heat treatment of 1500 °F/8 h could
relax residual stresses from quenching to below 50 ksi, judged sufficient in this study.
Expected variations in time at this stress relief temperature due to production batching
and disk section-size effects, estimated to be at least +/- lh, were predicted to produce
only minor variations in resulting residual stresses for this CSRA combined stress
relief/aging heat treatment.
Creep Properties
Creep strain-time curves of typical creep tests lasting over 1400 h at 1200, 1300,
1400, and 1500 °F are shown in Fig. 18. Creep data was generated for tests extending
from lh to over 10,000 h in some cases. Tests at higher temperature tended to have
smaller periods of primary creep, and larger periods of tertiary creep. Times to 0.1%,
0.2%, 0.5% and rapture were first analyzed using a Larson-Miller approach (ref. 7)
commonly employed for disk alloys. Creep results were used to generate conventional
Larson-Miller curves of stress versus Larson-Miller parameter (LMP) using the equation:
LMP=(T+460°R)(log t +C)
The resulting plots are shown in Figs. 19-22. It can be seen that the LMP constant C=20
did not fully account for test temperature in modeling the time to produce low creep
strains of 0.1, 0.2%, or 0.5%, but worked well for rupture life. Regressions indicated a
constant of 28 gave the best compromise of high correlations for 0.1%, 0.2%, and 0.5%.Polynomial regression equations using the variables LMP and LMP 2 are included with
NASA/TM--2002-211796 7
correlation coefficients in the figures, for use in estimating mean life responses as
functions of temperature and stress using this Larson-Miller approach.
Times to 0.2% creep are also compared for test temperatures of 1200-1500 °F in
Fig. 23. A simpler approach using multiple quadratic regression was performed to model
time to 0.2% creep, using stress, temperature and their resulting interactions. The
resulting equation and correlation coefficient is also given, for directly estimating mean
response.
Test results of specimens from subscale pancakes and blanks are compared to the
scale-up results in Fig. 24. The subscale material did have comparable creep properties
to the scaled-up disks. The creep properties did not significantly vary between the DH
and PA solution heat treatments, however creep resistance varied when the stabilization
heat treatment step was removed. Creep life at 1200 °F/125 ksi and 1400 °F/60 ksi
significantly increased when the stress relief step was removed from the baseline SR+A
cycle, Fig. 25. Significantly more scatter in life was apparent in the subscale data than
scaled-up data. This was apparently due to extensometer slipping for the small specimen
configuration used for the DH+SR+A and PA+SR+A material. Small extensometers
were lightly attached to the gages of these small specimens, while larger extensometers
were more firmly attached to ridges on the shoulders of larger specimens. Specimens
were machined from the subscale CSRA disks using the larger specimen configuration,
as in the scaled-up material tests. The resulting 0.2% lives exhibited much lower scatter
which was comparable to the scaled-up data, and slightly exceeded scaled-up lives atboth 1300 and 1500 °F.
Times to 0.2% creep of the subscale DH+CSRA disk specimens are shown versus
approximate cooling rate in Fig. 26. Increased cooling rate improved creep life at
lower temperatures (1300 °F/100 ksi), but reduced creep life at high temperatures
(1500 °F/50 ksi). The effects on creep life were less than 2X for both cases, over the
range of cooling rates evaluated.
Creep specimens tended to fail from intergranular, surface-initiated cracks at all
creep test temperatures, as shown in Fig. 27. Specimens tested at higher stress levels had
fewer cracks than those tested at lower stresses, for each test temperature. At increasing
temperatures of 1400-1500 °F, exposed grain surfaces on the surface cracks had a more
rough, dimpled morphology and more secondary cracking, with evident grain boundary
cavitation. The final overload failure occurred by transgranular microvoid coalescence
with scattered "facet" grain failures at 1200 °F. At increasing temperatures of 1300-
1500 °F, the final overload failures increasingly favored cavitation at grain boundaries.
Low Cycle Fatigue Properties
Total strain range versus life is compared for the test temperatures at each strain
ratio of 0.5, 0, and -1 in Fig. 28. Fatigue lives at 75, 1000 °F, and 1400 °F are shown as
functions of strain range and strain ratio (R0 in Fig. 29. A generalized polynomial
regression using temperature as a variable along with strain range and strain ratio gave
unsatisfactory results, with large error. Regressions at each temperature were therefore
performed using strain range, R_, and their interactions. The resulting equations and
correlation coefficients are included in the figure. The effects of strain ratio were found
to increase with temperature. The effect of strain ratio was quite modest at 75 °F, with
higher strain ratios reducing life by less than about 80%. However, both strain ratio and
NASA/TM--2002-211796 8
the interaction between strain ratio and strain range became more significant along with
strain range at the higher temperatures of 1000 and 1400 °F. At these temperatures,
higher strain ratios reduced life by over 90%. The resulting equations are included in
Fig. 29 for estimating mean life responses at these temperatures.
Close inspection of Fig. 28 indicates fatigue life for low strain ranges was lower
in tests at 400-800 °F than at room temperature and higher temperatures up to 1400 °F.
This is shown in Fig. 30 comparing lives at strain ranges of 0.55 and 0.70% with a strainratio of 0. Simple polynomial regression equations using T and T 2 are included in this
figure, for use in estimating mean life responses for these conditions as a function of
temperature.
Test results of specimens from subscale disks are compared to the scaled-up
results in Fig. 31. Groups of six tests were run at the temperatures of 800 and 1400 °F
using a strain range of 0.55%, R_=0. The subscale material had comparable fatigue
resistance to the scaled-up disks. The fatigue properties did not significantly varybetween the DH and PA solution heat treatments. These results did confirm that mean
life, given at a cumulative probability of 50%, was lower at 800 °F than that at 1400 °F.
Six additional specimens from subscale disks were given a prior exposure in air at
1400 °F for 500 h before LCF testing at a strain range of 0.55%, R_=0, in order to briefly
screen the effects of realistic service exposure times. These results are also compared in
Fig. 32. The mean life was similar to the unexposed mean life. However, a single
exposed specimen failed at only 5% of the mean cyclic life of the other five. A dissimilar
surface initiated failure mode was responsible for the low life of this exposed specimen,as will be discussed below.
Low cycle fatigue specimens predominantly failed from cracks initiated by planar
failure of relatively large grains from room temperature to 1400 °F, as shown in Fig. 33.
These "faceted" grain failures appeared to be due to concentrated slip on {111 } planes,
which could produce slip offsets in large grains, ref. 8. The grain facets were most flat
with least texture in tests at 400 and 800 °F. The grain facets had more texture in tests at
room temperature and 1000-1400 °F. More cracks were initiated in tests at higher strain
ranges and higher strain ratios. A smaller number of specimens failed from oxidized
surface cracks. These cracks were either transgranular or intergranular. A much smaller
minority of specimens failed from ceramic inclusions. The inclusions were more often
granulated alumina inclusions often referred to as Type 2 soft, reactive inclusions (ref. 9).
Among fatigue specimens pre-exposed at 1400 °F/500 h, the five specimens
having long mean life failed from internal cracks initiated at facets or inclusions, as
typified in Fig. 34. The single specimen failing at a much lower life had a surface
initiated failure with intergranular cracking. Evaluation of a metallographic section of
this specimen prepared transverse to the loading axis indicated general oxidation damage
along the specimen surface, producing an outer layer of NiO and underlying branches
rich in A1203 extending further in, as shown in Fig. 35. The alumina-rich branches grew
in at grain boundaries as well as along the machined grain surface. The activation of this
crack initiation mode at surface oxidation during service at 1400 °F could presentsignificant fatigue design challenges, due to the 10X lower fatigue life of the exposed
specimen with this failure mode. This failure mode has been shown to be operative after
prior exposures as well as during extended cycle periods in another powder metallurgy
NASA/TM--2002-211796 9
superalloy, Udimet 720, at temperatures as low as 1200 °F (ref. 10). Cyclic life was
reduced by up to 8X in that work.
Fatigue Crack Growth Properties
Cyclic crack growth rate versus stress intensity factor range is compared for all
test temperatures at stress ratios (R_) of-0.5 and 0.25 in Fig. 36. Crack growth rates
increased with temperature at both negative and positive stress ratios, and increased with
increasing stress ratio. The increase in crack growth rates with temperature was quite
modest, increasing roughly 8-10X in going from 75 to 1200 °F. This is shown more
clearly in Fig. 37, comparing cyclic crack growth rates at a fixed stress intensity factor
range versus temperature. Linear regression equations modeling cyclic crack growth
rates versus temperature are included in this figure, for use in estimating mean crack
growth responses as a function of temperature.
Dwell crack growth rate versus stress intensity factor range is compared for all
test temperatures at each stress ratio of 0 and 0.05 in Fig. 38. Most notable is the wide
scatter in dwell crack growth rates at each temperature. This was found to be related to
cooling rate, with specimens from higher cooling rate rim locations having higher crack
growth rates than slow cooling rate bore locations. Test results and linear regression
equations modeling dwell crack growth rates at maximum stress intensities of 25 ksi*in °5
and 30 ksi*in °5 versus temperature are included in Fig. 39, for use in estimating mean
crack growth responses as a function of temperature.
Dwell crack growth rate versus estimated average cooling rate of specimens from
DH&PA+SR+A subscale pancakes are shown in Fig. 40. Dwell crack growth rates were
shown to increase by over 10X when going from slowest (116 °F/min) to fastest cooled
(168 °F/min) specimens at 1300 °F. The crack growth rate increase with cooling rate was
reduced to 5X at 1400 °F. The subscale material did have comparable crack growth
properties to the specimens from the scaled-up disks, the latter specimens extracted from
relatively fast cooled disk rim regions. The cyclic and dwell crack growth properties did
not significantly vary between the DH and PA solution heat treatments.
The cracking mode observed in fatigue crack growth tests varied most notably
between the cyclic and dwell tests. Cyclic crack growth specimens had majority
transgranular cracking at all test temperatures, Fig. 41. While the proportion of
transgranular cracking was essentially 100% at 75 °F, an increasing percentage of
intergranular cracking became apparent at temperatures of 1200°F and higher.
Specimens tested from 75 to 1200 °F displayed planar cracking of some individual grains
by facet failure, possibly related to concentrated slip on { 111 } planes as for the low cycle
fatigue specimens. At higher temperatures a more textured fracture morphology was
observed which was more nearly Mode 2.
Dwell crack growth specimens had predominantly intergranular cracking at the
temperatures tested, Fig. 42, as previously observed in other superalloys in dwell crack
growth tests (refs. 11-12). The intergranular cracking mode was mixed with minor
trangranular cracking in tests of short dwell times and lower temperatures of 1200 °F.
These exposed grain boundaries were relatively flat. However, the intergranular failure
became highly prevalent in tests at higher temperatures, with considerable secondary
grain boundary cracks obvious. The exposed grain surfaces had large dimples due tocavitation.
NASA/TM--2002-211796 10
Summary and ConclusionsScaled-up ME3 disks processed in the HSR/EPM disk program were sectioned,
machined into specimens, and mechanically tested. Additional sub-scale disks and
blanks were processed and tested to explore the effects of several processing variations
on mechanical properties. Scaled-up disks had quite comparable mechanical response to
sub-scale disks, for common test conditions where direct comparisons were possible.
The mechanical properties of ME3 can be summarized as follows:
1 Tensile: Scaled-up ME3 had stable tensile strength and ductility to at least
1300 °F. Strength generally increased with increasing cooling rate, however this
effect decreased with increasing temperature. Strength became strain rate
dependent at 1400 °F, decreasing with decreasing strain rate. Strength and
ductility also became exposure time dependent at 1500 °F, decreasing with
increasing exposure time. Microvoid coalescence within grains produced failure
at 75-1300 °F, but surface cracking interceded at 1400-1500 °F.
2) Stress relaxation: Stress relaxation increased with increasing log(time) and
temperature, and was accentuated at high temperatures and long times. A
combined stress relaxation + aging heat treatment could be designed using stressrelaxation test results.
3) Creep: ME3 would creep less than 0.2% in 100h at 1300 °F with an applied stress
of 100 ksi. At 1400 °F and 1500 °F, this applied stress dropped drastically to
about 75 ksi and 50 ksi, respectively. Creep response could be modeled versus
temperature and stress using simple regression. Alternatively, a Larson-Miller
Parameter approach using a Larson Miller constant of 28 worked well for low
creep strains, while a constant of 20 worked well for rupture. Intergranular
surface cracking limited rupture life at all test temperatures.
4) Low cycle fatigue: At strain ranges of 0.7% or less typically encountered in
applications, ME3 had good LCF resistance up to 1400 °F. However, at higher
strain ranges, life decreased at 1400 °F due to decreasing strength. Extended
exposures at 1400 °F could also reduce life at low strain ranges by up to 20X.
Slip failures of large grains initiated failure at most temperatures. However, some
failures at 1400 °F were produced by crack initiation modes at surface oxidation.
5) Crack growth: Cyclic crack growth rates only increased by 12X between 75 °F
and 1300 °F. However, dwell crack growth rates strongly increased with
temperature from 1200 to 1500 °F, by about 10X per 100 °F. Dwell crack growth
rates also strongly increased with increasing cooling rate at 1300 °F, although this
effect appeared reduced at 1400 °F.
It can be concluded from this evaluation that ME3 has at least 1300 °F general
capabilities. Potential maximum temperatures for consideration in detailed
assessments of potential applications can also be suggested according to each
property:
1) Tensile: 1250-1300 °F based on yield and ultimate strength needs in disk boresand webs.
2) Creep: 1300-1400 °F based on 100-75 ksi creep stress requirements in webs andrims.
NASA/TM--2002-211796 11
3) Low cycle fatigue: 1300-1400 °F based on strain and service exposure
requirements throughout the disk.
4) Fatigue crack growth: 1300-1400 °F based on dwell crack propagation in
limiting rim locations.
References
1. T.P. Gabb, J. Gayda, J. Telesman, "Development of Advanced Powder
Metallurgy Disk Alloys in NASA-Industry Programs," Aeromat 2001, Long
Beach, CA, June 14, 2001.
2. C.P. Blankenship, M.F. Henry, J.M. Hyza_k, R.B. Rohling, E.L. Hall, "Hot-Die
Forging of P/M Ni-Base Superalloys," Superalloys 1996, ed. R.D. Kissinger,
D.J. Deye, D.L. Anton, A.D. Cetel, M.V. Nathal, T.M. Pollock, D.A. Woodford,
TMS, Warrendale, PA, 1996, pp. 653-662.
3. R.H. Vanstone, T.L. Richardson, "Potential-Drop Monitoring of Cracks in
Surface-Flawed Specimens," ASTM STP 877, American Society for Testing andMaterials, W. Conshohocken, PA, 1985, 148-166.
4. R.A. Ricks, A.J. Porter, R.C. Ecob, "The Growth of y' Precipitates in Nickel-
Base Superalloy," Acta. Met., V. 31, 1983, pp. 43-53.
5. W.F. Hosford, R.M. Caddell, Metal Formin_ Mechanics and Metallurgy,
Prentice-Hall, Englewood Cliffs, NJ, 1983, pp. 80-81.
6. J.E. Groh, "Effect of Cooling Rate From Solution Heat Treatment on Waspaloy
Microstructure and Properties," Superallogs 1996, ed. R.D. Kissinger, D.J. Deye,
D.L. Anton, A.D. Cetel, M.V. Nathal, T.M. Pollock, D.A. Woodford, TMS,
Warrendale, PA, 1996, pp. 621-626.
7. F.R. Larson, J. Miller, Trans. ASME, V. 74, 1952, pp. 765-766.
8. T.P. Gabb, J. Gayda, J. Sweeney, "The Effect of Boron on the Low Cycle Fatigue
Behavior of Disk Alloy KM4," NASA/TM--2000-210458, NASA, Washington,
D.C., 2000.
9. D.R. Chang, D.D. Krueger, R.A. Sprague, "Superalloy Powder Processing,
Properties, and Turbine Disk Applications," Superallogs 1984, ed. M. Gell,C.S. Kortovich, R.H. Bricknell, W.B. Kent, J.F. Radavich, TMS, Warrendale, PA,
pp. 245-252.
10. T.P. Gabb, J. Telesman, P.T. Kantzos, J.W. Sweeney, P.F. Browning, "Effects of
High Temperature Exposures on Fatigue Life of U720," Fatigue-David L.
Davidson Symposium, ed. K.S. Chan, P.K. Liaw, R.S. Bellows, T.C. Zogas,
W.O. Soboyejo, TMS, 2002, pp. 261-269.
11. K.R. Bain, M.L. Gambone, J.M. Hyza_k, M.C. Thomas, "Development of Damage
Tolerant Microstructures in Udimet 720," Superallogs 1988, ed. S. Reichman,
D.N. Duh., G. Maurer, S. Antolovich, C. Lund, TMS, 1988, pp. 13-22.
12. J. Gayda, T.P. Gabb, R.V. Miner, "Fatigue Crack Propagation of Nickel-Base
Superalloys at 650 °C," Low Cycle Fatigue, ASTM STP 942, ed. H.D. Solomon,
G.R. Halford, L.R. Kaisand, B.N. Leis, ASTM, Philadelphia, PA, 1988,
pp. 293-309.
NASA/TM--2002-211796 12
0
-100
-200O
O
-300O
E-
o -400
-5OO
-600 , ,
0
÷ Fan Air Quenched Bore
x Fan Air Quenched Rim
• Oil Quenched Bore
o Oil Quenched Rim
Air Cooled Blanks
0 1 2 3 4 5 6
Time (rain)
Fig. 1. Temperature versus time for thermocouples in the mid section (bore) and corner
(rim) of subscale disks during fan air and oil quenching, compared to thermocouple datafrom air cooled blanks.
NAS A/TM--2002-211796 13
a. b.
50 _m
1 _tm 1 _tm
c. d.
Fig. 2. Typical microstructures of scaled-up disks: a. subsolvus heat treated disk, S001
rim grain structure; b. supersolvus heat treated disk, S101 rim grain structure; c. S001 rim
7 'microstructure; d. S101 rim 7 'microstructure.
NASAFFM--2002-211796 14
SupersolvnsTensileStress-StrainCurves
250
g
c_
200 _ _ ............v:
150
lOO5oi i
0 :
--75F
...............400F
..............800F
..............I000F
--ll00F
--1200F
--1300F
1350F
..............1400F
..............1500F
0 0.05 0.1 0.15 0.2 0.25
Strain-in/in
%
a.
180
170
160
150
140
130
120
110
Supersolvus Tensile Stress-Strain Curves
____ --ll00F..............1200F
1300F
--1350F
..............1400F
----1500F
::Sf #':_:
_s/:i! !'/
I/A----....---.--JI
//
/I
0 0.01 0.02 0.03
Strain-in/in
b.
Fig. 3. Typical tensile stress-strain curves from supersolvus scaled-up disks, a) entire
test, b) initial stages at high temperature.
NASA/TM--2002-211796 15
Subsolvus Tensile Stress-Strain Curves
300
250
................::.... 75F200
__- ...............400F..............800F
150 '<i_i"......................................................... -.............1000F
:--I=_j_ -- 1lOOF100 --1200F
50 _ ...............13oov..............1400F
--1500F0 1
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
Strain-in/in
a.
Subsolvns Tensile Stress-Strain Curves
220
200
180
160
140
120
100
........ .........................................................................................................................' ....................'l_[.--1300F _,j_..r_/-,J'----/ _ s ........................J I ..............1400F J " _ //'
15ooF__/ .................._ ,z.,., ii. .................................,,..,.......................................................................................................................................................................
/
fj.//'_"--, _'_ _.,_. z_-, z_,, //_e
i i
0 0.01 0.02 0.03
Strain-in/in
b.
Fig. 4. Typical tensile stress-strain curves from subsolvus scaled-up disks, a) entire test,
b) initial stages at high temperature.
NASA/TM--2002-211796 16
._.a
180
170
160
150
140
130
120
llO
100
Q
t@
• 1400F Subsolvus
1400F Supersolvus
* 1500F Subsolvus
_, 1500F Supersolvus
TO
|
l
i i 1 I
0 0.0002 0.0004 0.0006 0.0008 0.001
Strain Rate-in/in*s
2.3
-_ 2.2._.a
2.1Q
• 1400F Subsolvus
1400F Supersolvus
-5 -3
i i i T I T i
-4Log(strain rate-in/in*s)
S_p 1400F/_-g(_} ........O.(}44S'_log(daid_} + 2.352?
R _= 0.7406
Sub. 1400F log(cy) = 0.0411 *log(ds/dt) + 2.327
R2 = 0.8835
St_p, _500F lc,g(c_} : 0.062':"]og(d_;/d_} + 2,3267
]<::_....0,9469Sub. 1500F log(e) = 0,0645*log(de,/d0 + 2.3057
R _ : 0,8885
Fig. 5. Strain rate dependence of strength at 1400 and 1500 °F shown using normal and
logarithmic axes.
NASA/TM--2002-211796 17
Fig. 6.
300 Supersolvus
%
250
200
150
100
5O
0
El0.2% Yield
Ultimate
,--"Notch (Kt=3.5)
0 500 1000Temperature-F
YS = -0,0000000956T 3 + 0,000192592gT 2 - 0,1303 g62217T +
178_8337787200
R2 ....£6749797240
U'.{'S ......,..0i)000001783'r _ + 0,0003492933]i a .. 0,20113453()N." +
...... i!98 )6884
R._ ....0,9582802949
N'}I'S ........O,O£KKKKK_551"r_ .. 0 0000320968'}i'" .. (}.(}637'384334T +
272,278"0()54116
R;_ 0,6180397 ii(i2
1500
Comparisons of yield strength, ultimate tensile strength, and notch strength from
supersolvus heat treated scaled-up disks.
NAS A/TM--2002-211796 18
%
_o
Fig. 7.
300
250
200
150
100
5O
0
Subsolvus
.............................................................................................................................................. ........................................................................
O
[] 0.2% Yield
Ultimate
Notch (Kt=3.5)
0 500 1000 1500Temperature- F
YS .......0,0000001189T :_+ 0.000237519(Y[ _ - 0,1357683301T +
188.2550521542
Rs ....0,8599851979
U.[S ..........0,0000001692T _ + 0,0003003117T; .. O. 157()7359()9T +
256,95 3 :{76 S2!){-{
R_ ....0,9122119108
NTS .........,(),()()()()()()()8"08"F> + (),()()()2274434"I. ":>.. (;',;:(-99()66449T +.
2":1(),7A-2()251533
R_:-.-.-.-0,2_21i300275
Comparisons of yield strength, ultimate tensile strength, and notch strength from
subsolvus heat treated scaled-up disks.
NAS A/TM--2002-211796 19
5O
4O
30¢,
._.a
._.a
20
10
D Supersolvus Elongation
Supersolvus Red. Area
_> Subsolvus Elongation
Subsolvus Red. Area
N
N
N
1°o
o N@
i _ i I T i ? t i _ i
0 500 1000 1500Temperature-F
SubRedArea .... 0.000()()00059T _ + 0ff)000032663T 2 -- 0.0007123647T
= 0° / 184-i.-l./8,41
S_._}-d!!;/o_g= .-.0 !_,)(_t}(_0!_,)6, i + 0.00()()092076T 2 --_,k(_t>,4, _7'_b,/I +
2 I. 161 _ 870207
R: = !)°:__96 / 5/096
Fig. 8. Comparisons of elongation and reduction in area from supersolvus and subsolvus
heat treated scaled-up disks; supersolvus mean elongation and mean reduction in area did
not significantly vary with temperature.
NASA/TM--2002-211796 20
Supersolvus 0.2% Yield Strength
._.a
._.a
170
160
150
140 'i °iJE B
i Ii
130 J_ [] Scaled-Up Disks4 [] Pancake DH+SR+A _[ /_ Pancake PA+SR+A [ ]
120 _ • Blank DH+SR+A
110 L ,ABlankPA+SR+A , , ,
1000 1100 1200 1300 1400 1500
Temperature-F
a.
Supersolvus Ultimate Tensile Strength
23°I220
210
200
190
180
170 -
160
150 -
140
130
1000
I ] ::
Scaled-Up Disks
[] Pancake DH+SR+A
A Pancake PA+SR+A
• Blank DH+SR+A
• Blank PA+SR+A7 I
f l r 1 I
|l
II
1100 1200 1300 1400 1500
Temperature-F
b.
Fig. 9. Comparison of scaled-up and subscale tensile a) yield strengths, and b) ultimate
strengths with solution heat treat variations pre-annealed (PA) and direct heatup (DH),
with comparable stress relief and aging heat treatments.
NAS A/TM--2002-211796 21
Supersolvus Percent Tensile Reduction in Area4O
35
30
25<
"" 20O
15
10
!! .... ,., 0
LScaled-Up Disks
[] Pancake DH+SR+A
A Pancake PA+SR+A
5 • Blank DH+SR+A
• Blank PA+SR+A0 ...................i......................................_...................t...................f...................t.................._......................................_...................
1000 1100 1200 1300 1400 1500Temperature-F
Fig. 10. Comparison of scaled-up and subscale reductions in area with solution heat treat
variations pre-annealed (PA) and direct heatup (DH), with comparable stress relief and
aging heat treatments.
Supersolvus 0.2% Yield Strength170
._.a
._.a
160
150
140
130
120 ?
110
1000
Scaled-Up Disks
© Pancake DH+CSRA
N Blank DH+A
_{_Blank PA+A
<
1100 1200 1300 1400 1500
Temperature-F
Fig. 11. Comparison of yield strengths, baseline scaled-up versus subscale disks with
combined stress relief +aging heat treat, and blanks not given stress relief.
NASA/TM--2002-211796 22
.,_a
G_
230Supersolvus Ultimate Tensile Strength
220
210
200
190
180
170
160
150
140
130
Scaled-Up Disks
© Pancake DH+CSRA
N Blank DH+A
_ Blank PA+A
• i i i i
1000 1100 1200 1300 1400 1500
Temperature-F
Fig. 12. Comparison of ultimate tensile strengths, baseline scaled-up versus subscale
disks with combined stress relief +aging heat treat, and blanks not given stress relief.
Supersolvus Percent Tensile Reduction in Area
G_
<3
G_
40
35
30
25
208
15 _ _ i_
I _ Scaled-Up Disks
10 # Pancake DH+CSRA
5 N Blank DH+A_{_Blank PA+A
_ i [ 1 t T
1000 1100 1200 1300
Temperature-F
i
1400 1500
Fig. 13. Comparison of reductions in area, baseline scaled-up versus subscale disks with
combined stress relief +aging heat treat, and blanks not given stress relief.
NASA/TM--2002-211796 23
f-%
• ll00F DH&PA+SR+A170 _ 1300F DH+CSRA
o 1400F DH&PA+SR+A1500F DH+CSRA
160
150
14o
130
120
100
J
Yield Stren th
jJ
o.._ g]
120 140 160 180 200
Av. Cooling Rate (F/min)
YS(1100F) = 0.185CR + 123.25
R 2 = 0.8846
YS(i 400F) = 0.1323CR + 123. l
R; = 0.9408
220 240 260
_<_f ........_'f.... 1.300F! (L{4{SCR+ {25.5)7
R ! :...{),$186
YS(_ 500F) = 0,1 I06CR + 10828
R:" = 0,9793
230 Ultimate Stren
_-, 220
"_ 210
,= 200
= 190
180
170
150140 __
130
• ll00F DH&PA+SR+A1300F DH+CSRA1 _+UUI _ LIII(N_I_/_t _lKt/_
1500F DH+CSRA
.......... _
100 120 140 160 180 200 220 240 260
Av. Cooling Rate (F/min)
UTS(1100F)= 0.2611CR + 187.01 _,ITS_,1300F) = ()#)(_4CR + 1_%I4
R 2 = 0.9631 R:: = 0,479
UTS(1400F) = 0.1112CR + 151.02 U[S(15(}(}i_.) ....0,0844CR + 132,26
R 2 = 0.8637 R:" ....0,97(;
Fig. 14. Effect of cooling rate on yield and ultimate strengths of subscale disks.
NASA/TM--2002-211796 24
_D
>_
160
150
140
130
120
ll0
100
90
1350
DH&PA+SR+A
t
• 112F/min Unexposed160F/min Unexposed
o 112F/min Exposedo 160F/min ExposedX 112F/min Exposed+Machx 160F/min Exposed+Mach
1400 1450
Temperamre-F
1500 1550
#._..a
._..a
._..a
180
170
160
150
140
130
120
1350
DH&PA+SR+A
• 112F/min Unexposed160F/min Unexposed
o 112F/min ExposedO 160F/min Exposedx 112F/min Exposed+Mach× 160F/min Exposed+Mach
I I
1400 1450 1500 1550
Temperature- F
I
<
Q
50.0DH&PA+SR+A
40.0
30.0
20.0
10.0
0.0
)
° 112F/min Unexposed160F/min Unexposed
o 112F/min Exposedo 160F/min Exposed× 112F/min Exposed+MachX 160F/min Exposed+Mach
1350 1400 1450 1500 1550
Temperature-F
Fig. 15. Effects of exposures on strengths and ductilities of subscale disks.
NASA/TM--2002-211796 25
10 _tm 10 _tm
a. b.
10 _tm 50 _tm
c. d.
Fig. 16. Tensile failure modes at: a) 75 °F: microvoid coalescence, b) 800 °F: microvoid
coalescence plus grain slip failures, c) 1200 °F: microvoid coalescence plus grain slip
failures, d) 1500 °F: intergranular surface cracking plus internal microvoid coalescence.
NASA/TM--2002-211796 26
¢)
180.0
160.0
140.0
120.0
100.0
80.0
60.0
40.0
20.0
0.0
0.001
[] 1400F DH1400F PA
[] 1450F DH1500F DH1500F PA1550F DH1550F PA
D 1600F DH1600F PA
0.01 0.1 1 10 100
Time-h
- (88.618925+8.895715*log(t)-0.082203 T-0.007791 *log(t)*T) 2
R2=0.9838
Fig. 17. Comparison of stress relaxation versus time (t) and temperature (T) in tests of
specimens after PA and DH solution heat treatments.
NASA/TM--2002-211796 27
G_G_
2.00
1.00
• 1200F/125ksi
1300F/95ksi
o 1400F/55ksi
1500F/35ksi
0.00
0 5OO
Time-h
i
1500
14
G_
12
• 1200F/125ksi10 _ 1300F/95ksi
_, 1400F/55ksi8 : 1500F/35ksi
G_
4
0
0 500 1000 1500 2000 2500 3000 3500
Tin_-h
Fig. 18. Typical creep curves, tests mn to rapture.
NASA/TM--2002-211796 28
0.1% Creep Life
140
120 _
NQO100 _'_
_, 80
o60
c/z
40 o
20
0
1200F,C=20
1300F,C=20
1400F,C=20
1500F,C=20
34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64
LMP=(460+T)(C+log 10(t))/1000
Fig. 19.
rs = 0.1195LMP 2 - 21.95LMP + 916.42
C=28, R 2 = 0.9489
Larson-Miller parameter versus stress for time to 0.1% creep, using Larson-
Miller constants (C) of 20 and 28.
NASA/TM--2002-211796 29
g
140
120
100
80
60
40
20
NO
0.2% Creep Life
°,(
• 1200F,C=20
1300F,C=20
1400F,C=20
c_ 1500F,C=20
Fig. 20.
34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64
LMP=(460+T)(C+log 10(t))/1000
= -0.1715LMP 2 + 9.2468LMP + 90.233
C=28, R 2 = 0.9745
Larson-Miller parameter versus stress for time to 0.2% creep, using Larson-
Miller constants (C) of 20 and 28.
NASA/TM--2002-211796 30
140
0.5% Creep Life
120 _ • N_
100
°_, 80
60
40
20
• 1200F,C=20
1300F,C=20
1400F,C=20
1500F,C=20
Fig. 21.
0
34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64
LMP=(460+T)(C+log 10(t))/1000
= -0.1782LMP 2 + 9.9431LMP + 79.193
C=28, R 2 = 0.9139
Larson-Miller parameter versus stress for time to 0.5% creep, using Larson-
Miller constants (C) of 20 and 28.
NASA/TM--2002-211796 31
gG_
140
120
100
80
60
4O
2O
Rupture Lit_
\
¢
¢
@
@
D
@
©½
@
1200,C=201300F,C=201400F,C=201500F,C=201200F,C=281300F,C=281400F,C=281500F,C=28
>
CD@
>
@=¢
0
34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64
LMP=(460+T)(C+logl 0(t))/1000
c_ = -0.319LMP 2 + 13.274LMP + 85.391
C=20, R 2 = 0.6437
Fig. 22. Larson-Miller parameter versus stress for time to rupture, using Larson-Miller
constants (C) of 20 and 28.
NASA/TM--2002-211796 32
0.2%CreepLife
140
120
100
80
60
4O
2O
• 1200F• N 1300F
1400F........................ _ 1500F
T T ? I ?rJi t ? _ ?_JT_ 1 i f _ TT?f r r i T ttit J J 1 tlifi
1 10 100 1000 10000 100000
0.2% Creep Life-h
log(0.2% life)=l 8.230123+0.076886cy-0.009298T-0.000079cyT-0.000155cy 2R2=0.9829
Fig. 23. Time to 0.2% creep versus stress using multiple quadratic regression.
NASA/TM--2002-211796 33
0.2% Creep Life
140
120
100
80
60
4O
2O
• []
[] Pancake DH+SR+AA Pancake PA+SR+A• Blank DH+SR+A• Blank PA+SR+A• Scaled-Up 1200F
Scaled-Up 1300F;> Scaled-Up 1400F
[q r 1 1 1 1 r r ] r i i r r r 1 i 1 i i 1 1 1 r i
10 100 1000 10000
Life-h
Fig. 24. Comparison of time to 0.2% creep for baseline scaled-up case versus subscale
disks and blanks having solution heat treat variations pre-annealed (PA) and direct heatup
(DH), using comparable stress relief and aging heat treatments.
NASA/TM--2002-211796 34
140
0.2% Creep Life
&go
c_
120
100
8O
6O
4O
2O
-. .......... -..
.... ¢,.. >
_ Pancake DH+CSRA .........._
[] Blank DH+AgeA Blank PA+Age* Scaled-Up 12-00Fe Scaled-Up 1300F* Scaled-Up 1400F_ Scaled-Up 1500F
i i i i i i i i I i i i i i ] i i
[]
..................... _ O
i ; i i i i i i
10 100 1000 10000Life-h
Fig. 25. Comparison of time to 0.2% creep for baseline scaled-up case versus subscale
disks with combined stress relief +aging heat treatment, and blanks with stress reliefremoved.
NASA/TM--2002-211796 35
3.0000DH+CSRA
5="
¢..)
t"l
O
2.5000
2.0000
1.5000
100
1300F/100ksi J1500F/50ksi
120 140
] _00I_b_:,_h) = 0/)0I 2CR + 2. I I "_';'
R" = ('_.4595
160 180 200 220 240
Cooling Rate-F/min
5()()Fling(h) = _(),001CR + 2,2452
Fig. 26. Effect of cooling rate on creep resistance in DH+CSRA subscale disks.
NASA/TM--2002-211796 36
50 _tm 50 _tm
a. b.
50 _tm 50 _tm
c. d.
Fig. 27. The typical creep failure modes of intergranular surface cracking:
1200 °F/115 ksi/7090.1h; b) 1300 °F/95 ksi/2400.1h; c) 1400 °F/45 ksi/7695.1h;
d) 1500 °F/30ksi/1829h.
NASA/TM--2002-211796 37
7
_6¢j>_?
_5.a
_4
3
0.4
7
_6¢..)
o,_5
o4
3
0.4
7
6
o,
o
U=- 1
o
o[]
[]
o
[]
o8
[]o [_o
oo
r
[] 70F
• 400F
[] 800F
o IUUUU
[] I200F
o 1400F
[]
o
oF
0.6 0.8 1
Strain Range- %
1.2
a.
R=0
O
[]
[] O
8o[][]O W
o
i
[] 70F
• 400F
[] 800F
o 1000F
[] 1200F
o 1400F
O
0.6 0.8 1
Strain Range- %
1.2
b.
R=0.5
©
[]O
oo
• o[] []
io
U []0
0i
[] 70F
• 400F
o 1000F
[] 1200F
o 1400F
1.4
1.4
0.4 0.6 0.8 1 1.2 1.4
Strain Range- %
C.
Fig. 28. Low cycle fatigue life versus strain range at a) R_=-I, b) R_=0, c) R_=0.5.
NASA/TM--2002-211796 38
a.
7
6¢o>_¢o
I
5
_4o
70F
* R=0.5• R=0o R=-I
0.4 0.6 0.8 1 1.2 1.4
Strain Range-%
70 oF: log(life)=3.477516+0.013860(1/Aa)2+0.026668R_-0.000012(1/Aa)2R_R2=0.9727
7 1000F. R=0.5• R=0
6 R. o,_ o R=-I'-_ f .
_5 -_,,,
"_4 __o"'* _ -
0.4 0.6 0.8 1 1.2 1.4
Strain Range-%
b. 1000 °F: log(life)=2.096100+0.017591 (1/Aa)+0.811451R_-0.008613 (1/Aa)R_R2=0.9475
6¢..)
¢..)I
5
_4o
O
1400F
* R=0.5• R=0o R=-I
r"'O
0.4 0.6 0.8 1 1.2 1.4
Strain Range-%
c. 1400 °F: 2log(life)=2.609246+0.0000817( 1/ Aa) +0.1696561R_-0.000032 (1/ Aa) 2 R_R2=0.9420
Fig. 29. Fatigue life regressions at a) 70, b) 1000, and c) 1400 °F.
NASA/TM--2002-211796 39
7
6
¢,.)
¢,.)
o_,._
©
4
3
Fatigue Life at R=0
• 0.5% Strain Range0.7% Strain Range
I I I I
J
I I I I I I I I
0 500 1000 1500
Terrperamre-F
bg([email protected]%) = 2E-06T z- 0.0025T + 5.7968
R2 = 0.9514
I{_.i(I.i;,:<,_'0,70_7) = -71:L07_.1': .I..0,00 I:4T .I..-:LII51_-'_
R! ::::0,557)I2
Fig. 30. Simplified fatigue life versus temperature relationships at strain ratio of 0.
NASA/TM--2002-211796 40
¢..)?
o
4
Fig. 31.
©Q
i
Q
[]O
i
R=0
[] 800F, Scaled-Up Disks[] 800F, Pancake DH+SR+AA 800F, Pancake PA+SR+AU 14UUI_ _(.;_ll_(J--Up IJlbKb
[] 1400F, Pancake DH+SR+_A 1400F, Pancake PA+SR+A
i[]
0.4 0.6 0.8 1 1.2 1.4
Strain Range- %
Comparison of strain range-life responses for scaled-up and pancakematerial
+,.a
©
29
9O
8O
70
50
30
2O
10
/® 1400F /
/O 800F /
• 1400F Pre-exposed 500h /
/
/
/• /e
//
//
/
//
//
104 105 106
Fig. 32.
Life-cycles
Probability plot comparing life of subscale disk specimens at 800 and 1400 °F,
and prior-exposure effects.
NAS A/TM--2002-211796 41
50 _m
a.
50 _m
C.
50 _m
d.
N
N® N
e. f.
50 _m
Fig. 33. Failure initiation points in LCF specimens tested at R_=0: a) 75 °F, A_=0.5%:
surface grain facet; b) 75 °F, A_=1.15%: multiple surface grain facets; c) 800 °F,
A_=0.5%: surface grain facet; d) 800 °F, A_=1.15%: multiple surface grain facets;
e) 1400 °F, A_=0.45%: internal ceramic inclusion; f) 1400 °F, A_=1.15%: multiple
surface grain facets.
NASA/TM--2002-211796 42
50 _tm
a. b.
50 _tm
50 _tm
C.
Fig. 34. Failure initiation points of specimens LCF tested at 1400 °F, Ae=0.7%, R_=0
after 1400 °F/500h exposure: a. single internal grain facet, life = 499,289 cycles; b. single
internal Type 2 alumina-rich inclusion, life = 162,977 cycles; c. single surface
intergranular crack, life = 10,994 cycles.
NASA/TM--2002-211796 43
20 _tm
Fig. 35. Typical oxidized surface of 1400 °F/500h exposed LCF specimens, with outer
NiO layer and inner branches of A1203.
NASA/TM--2002-211796 44
1.E-04R=-.25
>_
""TG)
-,.....a
o
GO
1.E-05
1.E-06
1.E-07
1.E-08
• 70F H101-F1
800F W110-t
a 1000F S100-1
:_i_:1200F S101-1
• 400F H101-F
0 10 20 30 40 50
Stress Intensity Factor Range-ksi*in °5
a. R_=-0.25
1 .E-04R=0.5
1.E-05 _G_
,= 1.E-06
o
"_ 1.E-07
1 .E-08 i i i
* 400F Hlll-F10
800F Sll0-F2
800F S101-F12
1000F Hlll-L35
1200F Sll0-F7@ 1300FWll0-F8
I _ I
0 10 20 30 40 50
• • .05Stress Intensrty Factor Range-ksffm "
b. R_=0.5
Fig. 36. Typical cyclic fatigue crack growth test results, da/dn versus AK.
NASA/TM--2002-211796 45
¢.9
_5
O
O
-4
-5
-6
Cyclic Fatigue Crack Growth at AK=30ksi*in °5, R_=-0.25
200 400 600 800 1000 1200
Temperature- F
k_g(&_/dt_ ....0,000(';'1' .. 5,4_;77
R_ = 0.,9672
i
1400
a.
¢.9
_5
O
¢.9
O
-4
-5
-6
Cyclic Fatigue Crack Growth at AK=15ksi*in °5, R=0.5
200 400 600 800 1000 1200
Temperature- F
bg;dwdn) ....0,0014J --6,639
R_ = 0,9857
1400
Fig. 37.
b.
Comparison of cyclic fatigue crack growth rates versus temperature at stress
ratios R_ of a) -0.25; b) 0.5.
NASA/TM--2002-211796 46
<q.
"'T
._.a
O
1 .E-03
1 .E-04
1 .E-05
1 .E-06
1 .E-07
1 .E-08
1 .E-09
..]- 1200F/90s Sll0-Fll1300F/90s Wll0-Fll Dwells, R=0
o 1300F/90s H11 l-F11x 1400F/60s S100-F12
i i i i i i i i i i i i i i i i
0 10 20 30 40
Maxirr_am Stress Intensity Factor-ksi*in °5
a.
Fig. 38.
• 1300F/90s S100-F111300F/90s S101-F91300F/90s W110-F101300F/2h S101-F2
x 1400F/90s S100-F8× 1400F/90s S101-F11
Dwells, R=0.051.E-03
+
++
._. 1.E-04 _ x _
.= + 1400F/90s Wll0-F12 + +
_ 1.E-05 __1.E-06 ° _
1.E-07
1.E-os d4
1.E-09
0 10 20 30 40
Maxirr_am Stress Intensity Factor-ksi*in °5
b.
Typical dwell crack growth curves, da/dt versus Kma_., at a) R_=0; b) R_=0.05.
NASA/TM--2002-211796 47
_5
Q
<D
29
Q
Dwell Fatigue Crack Growth at AK=25ksi*inA.5
-3• 25ksi*inA.5, R=0
25ksi*inA.5, R=.05-4
-5 _
-6
]
]
-7
1200 1300 1400 1500
Temperature- F
Fig. 39.
log(da/dt, R=0) = 0.009T - 17.217
R2 = 0.9837
/_g(da/d<R=,()5} ....0,0()5(<['_ 12,g5P
R_ = 0,2776
a.
Q
Q
-3
-4
-5
-6
-7
Dwell Fatigue Crack Growth at A K=30ksi*inA.5
1200 1300 1400 1500
Temperamre-F
log(da/dt, R=0) = 0.0106T- 18.978
R2 =0.9934
l{_g(da£1<R=,05) = 0,00597' _ ]2,879
t._s : 0,3472
b.
Dwell fatigue crack growth rates versus temperature at different stress ratios R_at a) AK=25 and b) 30ksi*in °5.
NASA/TM--2002-211796 48
-4tt'3
c5
= -4.5
u-_ -5t"q@
-5.5
"'7._.a -6
-6.5o
l-7 I _ _ f
DH&PA+SR+A
15UU_
0 1400F
100 120 140 160 180
Cooling Rate-F/min
1400Flog(daJdt) = 0,0114CR - 7.4457 I 3(}(}}.'log(da/d_) .....0_0255CR -. 9_4929
R 2 = 0.7984 }_:_....0,9692
Fig. 40. Dwell fatigue crack growth rates at Kmax=25 ksi*in °5 versus cooling rate at 1300
and 1400 °F for subscale disk material.
NASA/TM--2002-211796 49
50 _m
a.
50 _m 50 _m
b. C.
Fig. 41.
50 _m
d.
Typical cyclic crack growth modes: a. 400 °F, R=0.25; b. 800 °F, R=-I;
c. 800 °F, R=0.25; d. 1300 °F, R=0.25.
NASA/TM--2002-211796 50
50 _m
a.
50 _m
b. C.
50 _m
50 _m 50 _m
d. e.
Fig. 42. Typical dwell crack growth modes for: a. 1200 °F, 90 s dwell; b. 1300 °F, 90 s
dwell; c. 1300 °F, 2 h dwell, d. 1400 °F, 90 s dwell, e. 1500 °F, 90 s dwell.
NASA/TM--2002-211796 51
Form ApprovedREPORT DOCUMENTATION PAGEOMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources,
gathering and maintaining the data needed, and completing and review#lg the collection of information. Send corrlments regarding this burden estimate or any other aspect of this
collection of information, including suggestions for reducing this burden, to Washington Headquarters Services. Dhectorate for Information Operations and Reports, 1215 Jefferson
Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget Paperwork Reduction Project (0704-.0188), Washington, DC 20503.
1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED
August 2002 Technical Memorandum
5. FUNDING NUMBERS4, TITLE AND SUBTITLE
Chm'actedzation of the Temperature Capabilities of Advanced Disk Alloy ME3
& AUTHOR(S)
Timothy P. Gabb, Jack Telesman, Peter T. Kantzos, and Kenneth O'Connor
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)
National Aeronautics and Space Administration
John H. Glenn Research Center at Lewis Field
Cleveland, Ohio 44135 - 3191
WU-714-04-20-00
8. PERFORMING ORGANIZATIONREPORT NUMBER
E----13491
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORINGAGENCY REPORT NUMBER
National Aeronautics and Space Administration
Washington, DC 20546-0(101 NASA TM------.2002-211796
11. SUPPLEMENTARY NOTES
Timothy P. Gabb, Jack Telesman, and Kenneth O'Connor_ NASA Glenn Rese_'ch Center; Peter T. Kantzos, Ohio Aero-
space Institute, Brook Park, Ohio 44142. Responsible person, Timothy R Gabb, organization code 5120, 216-433-3272.
12a, DISTRiBUTION/AVAILABILITY STATEMENT
Unclassified - Unlimited
Subject Category: 07 Distribution: Nonstandard
Available electronicaJly at bttp://glt:.-s._rc.nasa.aov
"l-his publication is available from the NASA Center for AeroSpace In_brmadon, 301-621-0390.
12b. DISTRNBUTION CODE
13. ABSTRACT (Maximum 200 words)
The successful development of an advanced powder metallurgy disk alloy, ME3, was initiated in the NASA High Speed
Research_nabling Propulsion Materials (HSR_PM) Compressor/Turbine Disk program in cooperation with General
Electric Engine Company and Pratt & Whitney AircrNt Engines. This alloy was designed using statistical screening and
optimization of composition and processing variables to have extended durability at 1200 eF in large disks. Disks of this
alloy were produced at the conclusion of the program using a realistic scaled-up disk shape and processing to enable
demonstration of these properties. The objective of the Ultra-Efficient Engine Technologies disk program was to assess
the mechanical properties of these ME3 disks as functions of temperature in order to estimate the maximum temperature
capabilities of this advanced alloy. These disks were sectioned, machined into specimens, and extensively tested. Addi-
tional sub-scale disks and blanks were processed and selectively tested to explore the effects of several processing varia-
tions on mechanical properties_ Results indicate the baseline ME3 alloy and process can produce 1300 to 1350 °F
temperature capabilities, dependent on detailed disk and engine design property requirements.
14. SUBJECT TERMS
Gas turbine engines; Rotating disks; Heat resistant alloys; Fatigue (materials); Inclusions
17. SECURITY CLASSIFICATIONOF REPORT
Unclassified
NSN 7540-01-280-5500
15. NUMBER OF PAGES
5716. PRICE CODE
18, SECURITY CLASSiFiCATiON 19. SECURITY CLASSiFiCATiON 20. LiMiTATiON OF ABSTRACTOF THIS PAGE OF ABSTRACT
Unclassified Uncl assifi ed
Standard Form 298 (Rev. 2-89)
Prescribed by ANSI Std. Z39-18298-102