DEVELOPMENT AND TURBINE ENGINE PERFORMANCE

OF THREE ADVANCED RHENIUM CONTAINING SUPERALLOYS

FOR SINGLE CRYSTAL AND DIRECTIONALLY SOUDIFIED BLADES AND VANES

Robert W. Broomfield, David A. Ford, Harry K Bhangu

Rolls-Royce plc
[Derby and Bristol, U.K.] 1111111111113111111111
Malcolm C. Thomas, Donald J. Frasier, Phil S. Burkholder
Allison Engine Company (Rolls-Royce plc)
[Indianapolis, Indiana U.S.A.]
Ken Harris, Gary L. Erickson, Jacqueline B. Wahl
Cannon-Muskegon Corporation
(SPS Technologies, Inc.)
[Muskegon, Michigan U.S.A.]

ABSTRACT Re) for a variety of turbine engine applications. A range of

Turbine inlet temperatures over the next few years will critical properties of these alloys is reviewed in relation to
approach 1650°C (3000°F) at maximum power for the latest turbine component engineering performance through engine
large commercial turbofan engines, resulting in high fuel certification testing and service experience.
efficiency and thrust levels approaching 445 kN (100,000 lbs). Industrial turbines are now commencing to use this aero
High reliability and durability must be intrinsically designed into developed turbine technology in both small and large frame units
these turbine engines to meet operating economic targets and in addition to aero-derivative industrial engines. These
ETOPS certification requirement applications are demanding with high reliability required for
This level of performance has been brought about by a turbine airfoils out to 25,000 hours, with perhaps greater than
combination of advances in air cooling for turbine blades and 50% of the time spent at maximum power. Combined cycle
vanes, design technology for stresses and airflow, single crystal efficiencies of large frame industrial engines is scheduled to
and directionally solidified casting process improvements and the reach 60% in the U.S. ATS programme. Application
development and use of rhenium (Re) containing high y' volume experience to a total 1.3 million engine hours and 28,000 hours
fraction nickel-base superalloys with advanced coatings, individual blade set service for CMSX-4 first stage turbine
including full-airfoil ceramic thermal bather coatings. Re blades is reviewed for a small frame industrial engine.
additions to cast airfoil superalloys not only improve creep and
thermo-mechanical fatigue strength but also environmental NOMENCLATURE
properties, including coating performance. Re dramatically slows ASMET = accelerated simulated mission endurance test
down diffusion in these alloys at high operating temperatures. ATS = advanced turbine system
A team approach has been used to develop a family of two B = boron
nickel-base single crystal alloys (CMSX-4. containing 3% Re C = carbon
and CMSX0-10 containing 6% Re) and a directionally solidified, CGR = crack growth rate
columnar grain nickel-base alloy (CM 186 LC: containing 3% DS = directionally solidified, columnar grain

CMSX-20, CMSX-30, CMSX-40, CMSX-60, CMSX0- 10, CM 247 LC® and CM 186 LC®
are registered trademarks of the Cannon-Muskegon Corporation.
LAMILLOY® is a registered trademark and CASTCOOLTI4 a trademark of Allison Engine Company.
Presented at the International Gas Turbine & Aeroengine Congress & Exhibition
Orlando, Florida — June 2-June 5, 1997
This paper has been accepted
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ for publication
on 10/26/2017 inUse:
Terms of the http://www.asme.org/about-asme/terms-of-use
Transactions of the ASME
 OPT = durability proof testing Maximum metal temperatures approaching 1130°C
EDAX = energy dispersive x-ray micro-analysis (2066°F) have been flight qualified for CMSX-4 turbine blades
EFH = engine flight hours at maximum engine power during accelerated, simulated mission
ETOPS = extended over water, twin engine certification endurance testing (ASMET) (Fullagar et al., 1994). Full airfoil
requirements and platform advanced thermal barrier coatings have been
HCF = high cycle fatigue cerrfied for commercial turbine engine use, with the capability to
HIP = hot isostatic pressing increase gas temperatures by 100°C (180°F), or reduce metal
HP = high pressure temperatures commensurately to dramatically improve turbine
IP = intermediate pressure blade life (PW, 1994).
ISA = International Standard Atmosphere (15°C) Allison's unique dual-wall Lamilloy® quasi-transpiration
kN = kilo newton cooling technology applied to CMSX-4 single crystal airfoils
LCF = low cycle fatigue facilitates a further 222°C (400°F) to 333°C (600°F) turbine inlet
LW' = Larson-Miller parameter temperature capability increase over the next five years. The
LNG = liquid natural gas CastcoolTh Lamilloy® technology combines film leading and
MFB = machined-from-blade trailing edge airfoil cooling, with the dual-wall Lamilloy cooling
MW(e) = mega-watt in the rest of the airfoil in a one piece single crystal casting,
NGV = nozzle guide vane further improving the cost of manufacture. The fine detail and
OPR = overall pressure ratio complexity of these components bring manufacturing
ppm = part per million considerations to the forefront (Harris et al., 1990, Burkholder et
S = sulfur al., 1995). (Fig. 1).
Si = silicon
SEM = scanning electron microscope
T = temperature
TBC = thermal barrier coating
TCP = topologically close-packed phase
TEM = transmission electron microscope
TEl = turbine entry temperature
IF = thermal fatigue
TFCLL = thermal fatigue crack initiation life
TMF = thenno-mechanical fatigue
TTT = transformation-time-temperature
WDX = wavelength dispersive micro-analysis
Zr = Zirconium
[NJ = combined nitrogen
[0] = combined oxygen
y = gamma phase
y' = gamma prime phase
Astr,„, = change in mechanical strain
Figure 1 - AE 301X Castcool 1st Stage Blade - CMSX-4
Kr = stress concentration factor
Alloy
INTRODUCTION The compositions of the first generation single crystal
During the last 30 years, turbine inlet temperatures have superalloys which have attained turbine engine application status
increased by about 500°C (900°F). About 70% of this increase are shown in Table I. These alloys are characterized by similar
is due to more efficient design of air cooling for turbine blades creep-rupture strength. However, they exhibit variations in
and vanes, particularly the advent of serpentine convection and single crystal castability, residual y/y' eutectic phase content
film cooling and the use of full airfoil thermal battier ceramic following solution heat treatment, absence or presence of
coatings, while the other 30% is due to improved superalloys and carbides, impact and mechanical fatigue properties (HCF &
casting processes. The greatest advances in metal temperature LCF), environmental oxidation and hot corrosion properties,
and stress capability for turbine airfoils have been the result of the coating performance and density.
development of single crystal superalloy, casting process and Turbine engine experience with the first generation single
engine application technology pioneered by Pratt and Whitney crystal alloys has resulted in process developments being
(P&W) (Gell et al., 1980). combined with Re additions to improve and maximize overall

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 Table I ALLOY DEVELOPMENT
First Generation Single Crystal Superalloys Development in the CM family of single crystal
Nominal Composition, wt. X superalloys has been in two general directions since the inception
Ain Ce Co No W Ts V Co AI 11 Mt id dewy of CMSX-2® and CMSX-3® alloys (Harris et al., 1983, Harris
(N0) Oisrent et aL, 1986):
PWA 10 5 • 4 12 • • 5.0 14 • SAL 170
RAS N4 10 I 2
Partial replacement of tungsten (W) with increasing Re.
6 5 • .5 4.2 33 2 SAL 1.56
EAR 99 1 3 • 10 3 - • 53 12 • SAL 1136
Lowering of chromium (Cr) to accommodate the increased
RR2003 10 IS 3 • • 1 • 5.5 4.0 • 841. 737
alloying with acceptable phase stability.
Pan 5 6 2 6 5 • • 5.2 12 - SAL 1139 • Partial replacement of titanium (Ti) by tantalum (Ta).
NO • I 5 2 s • • • 64 2.0 • fiAL Re is a key element, and the magnitude of the improvement
04052 $ .e 5.5 14 641. 136 which it provides in creep and LCF strength at 950°C (1742°F)
C1421143 5 3 .5 5.6 14 .1 EL4L 633
is illustrated in Figs. 2 and 3. As an example, changing from a
04113 10 5 3 44 4.7 .1 SAL 7.116
II 11 •
non-Re containing alloy SRR99 to a 6% Re alloy CMSX-10 (RR
SX 7/2 2 3.4 42 841 1.25
3000) increases creep strength at 500 hours life by 46%, and
increases fatigue strength at 20,000 cycles life by 59%. These
properties of the turbine airfoil components (Harris et al., 1990). improvements are less when corrected for alloy density
Microstructures can be optimized to be fully solutioned and differences.
H1P'ed, to contain neither y/y' eutectic phase, nor regions of
incipient melting, carbides, nor microporosity (Fullagar et al., 250

1994). The published compositions of the Re containing single

crystal alloys are shown in Table II. Streng thIncrease. MPe - DitSx-e
CUSX .10 (RR31200)
Table II
Re Containing Single Crystal Alloys
Nominal Composition, wt. %
IX
Ann Cr Co No W Ts CO Re 11 M1 NI Density
(ND) 0022dro,
••••••.„...
CIA3X-4 6.5 '9 6 6.5 3 5.6 1.0 .1 SAL 5.70
50
PWA 1434 ------------
5 10 2 6 9 • 3 5.6 • .1 BAL ILO ea/ --
qr.
Sc 100 5 10 2 $ 0.5 - 3 5.2 1.0 .1 RR. 854
Rerd 5/5 7 5 2 $ 5 • 3 62 .2 MAL 3.33
2 5 10 20 • 50 IX 829 SOO 1300
Rend $6 • 12 1 6 7 • 5 54 • .2 SAL 6.97
Tune, hours
CMS% -10 2 3 .4 5 3 .1 6 S.? .2 .03 SAL 935
Figure 2 - The Strength Advantage of Re-Containing
Component cost considerations particularly for commercial Single Crystals over SRR 99 - 1% Creep Strain, 950°C
aero and industrial turbine engines has resulted in the (1742°F) [Not density corrected.]
development of the three Re containing DS superalloys (Table
131). These alloys have similar creep-rupture strength to the first
generation single crystal superalloys. PWA 1426 (Cetel et aL,
1992) is used 50% solutionecl, Rene 142 (Ross and O'Hara, 1992)
close to 100% solutioned and CM 186 LC as-cast (Harris et 0, - - - - - - - - -

al.,1992, Camel et al., 1996). ' The absence of a solutioning - ----

requirement with CM 186 LC not only lowers cost and improves ..• .0
manufacturability (no recrystallisation or incipient melting
CIASC-4
problems) but also provides excellent transverse intermediate OASX -10 (RSCIX10)
temperature ductility and transverse low cycle fatigue (LCF)
properties.
Table III 1 ,

Re Containing DS Alloys . 2003 X0) KW 15030 2C000 X000 00030 Ic0X0

Nominal Composition, wt. Ti .C1155

Noway
Alloy Cr Co Mo w to Re Al Ti 141 C a Zr w 0021121 Figure 3 - The Strength Advantage of Re-Containing
PV0A 1425 5.5 13 2 5 • 3 0.0 • 13 .10 .015 .03 SAL 5.6 Single Crystals Over SRR 99 - Low Cycle Fatigue,
Re4 142 6.0 12 2 5 6 3 62 - 1.5 .12 .015 .02 SAL 11.6 950°C (1742°F)
OA 104 LC 6.0 9 .5 8 3 3 5.7 .7 1.4 .07 .015 .005 BAL 11.70

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 In order to understand these improvements, the distribution Table V
of Re through the microstructure has been studied in some detail Alloy and y' phase compositions.
on three scales: dendritic, microscopic and atomic. Co Cr Ma W Re At Ti Ta

Nominal wt. %
DISTRIBUTION OF RHENIUM RR2067 8.0 3.4 0.48 6.1 5.3 5.6 1.0 7.8
On the dendritic scale, it is well known that Re segregates CMSX-4 9.7 6.5 0.6 6.4 2.9 5.6 1.0 6.5
strongly to the dendrite centers and that even - alter a Gamma PAM c5, vet. %
solution/homogenisation heat treatment, a uniform distribution is RR2067 5.2 1.9 1.0 5.6 1.9 6.1 1.25 11.4
not achieved (Table IV). In the Rolls-Royce materials CMSX-4 6.1 2.7 055 5.75 1.4 6.2 1.4 10.7
specifications for CMSX-4 and CMSX-I 0 (RR 3000), a different Gamma Prime (75, at.%
approach to measuring dendritic segregation following RR2067 5.5 2.3 0.6 1.9 0.65 14.1 1.6 3.9
CMSX-4 6.4 3.2 0.35 1.9 0.45 14.2 1.75 3.6
solution/homogenisation heat treatment has been taken. A large
number of measurements of Re, W and Ta levels are made across
levels in the y phase of about 6 wt. % and 13% (by weight)
a section perpendicular to crystal growth, and the standard
respectively in the two alloys. It is not surprising therefore, that
deviations associated with the distributions for each element are
under conditions where dislocation movement is confined to the
calculated. An upper limit is set for the standard deviation for
y phase, Re additions are very powerful strengtheners.
each element; the limits vary from one alloy to the other, but they
Atom-probe micro-analyses of Re containing
do not vary by component in a given alloy. The
.modifications of PWA 1480 and CMSX-2 alloys reveal the
solution/homgenisation heat treatments developed for both alloys
occurrence of short range order in the y matrix (Blavette et al.
are designed to minimise residual dendritic microsegregation and
hence to maximise phase stability. 1986 , Blavette et al., 1988). Small Re clusters (approximately •
1.0 Ann in size) are detected in the alloys. The Re clusters act as
Table IV efficient obstacles to dislocation movement in the y matrix
CMSX-4 ALLOY channels compared to isolated solute atoms in solid solution and
0.25" (6.4 mm) 0 Test Bar Allison Solution/Homogenisation thereby play a significant role in improving alloy strength. The
Heat Treatment + Double Aged
benefits of Re to mechanical properties are seen in situations
SEM-WDX Analysis (wt. °A) where dislocation movement within the y phase matrix channels
Center of Interdendritic is controlling. Where dislocations pass readily through both y
Primary Dendrite Region and y' phases, the strength advantage is smaller. Dislocations
Cr 5.8 5.7 travel mainly within the y matrix channels at higher temperatures
Co 9.7 9.8 >850°C (1562°F). A consistent benefit for CMSX-4 for example
is seen in tensile strength, creep and stress-rupture strength over
Ni BAL BAL the temperature range 850-1050°C (1562-1922°F) where the
6.1 4.2 temperature capability advantage is at least 30°C (54°F) over
Mo .6 .7 SRR 99.
Further work on the distribution of Re on the atomic scale
Ti .9 1.0
is being undertaken at Oxford University using atom-probe
Al 5.7 6.4 micro-analysis. In filly heat-treated CMSX-10 (RR 3000), a
Re 3.7 1.9 pronounced buildup of Re has been observed adjacent to the y'
particles - as one might expect since Re is a slow-diffusing
Ta 5.0 6.5
element which is rejected by the growing y' (Fig. 4).
I-If .06 .05
7 phase r phase
On a microscopic scale, the composition of the y' phase
has been measured in both CMSX-4 and in an early variant of
CMSX-10 (RR 3000) containing 5.3% Re (RR 2067) using Al
EDAX micro-analysis on thin TEM foils. The alloy and y' phase
compositions were as shown in Table V.
The y' compositions were much more similar to each other
than were the alloy compositions, basically (Ni, Co), (Al, Ta)
with some Cr, W and Ti dissolved. The Re contents in the y'
were low, and so on the basis that both CMSX-4 and CMSX-I 0
(RR 3000) contain about 70 vol. % y', this implies average Re Figure 4 - Re Distribution Across y and y' Phases
CMSX-10 (RR 3000)

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 INFLUENCE OF Re UPON SELECTED PROPERTIES
The influence of Re is all-pervasive in this class of
superalloy, but five areas have been selected: stability of y' at
high temperatures, effect of Re distribution upon primary creep
behavior, oxidationaoerformance, thermal fatigue, and finally
LCF properties.

Stability of v' .
y' stability is important in terms of resistance to coarsening
and re-solutioning of y' during coating or brazing operations.
CMSX-4 has good performance in this respect, as indicated by
a recent paper (Miglietti and Pennefather, 1996). These workers
measured y' size following a wide range of brazing/diffusion heat
treatments up to 1240°C (2264°F). These times and
temperatures have been combined via the Larson-Miller
parameter (LMP) (Fig. 5); there is steady y' growth up to a LMP
Figure 6 - Rafted y/y1 Structure in CMSX-10 (RR 3000)
of 30,000, but much faster growth thereafter. The 30,000 value
after 1% creep strain in 60 hours at 1175°C (2147°F)
conesponds for instance to 3 hours at 1190°C (2174°F) (atypical
brazing condition). It is interesting to note that the resistance to
y' coarsening appears to be improved by the 1140°C (2085°F) Primary Creel, Behavior
intermertivf. age used for CMSX-4. As-solution-treated CMSX-4 During the development of CMSX- 10 (RR 3000) it was
was soaked for various times and temperatures by Roan (1996), noted that the magiitude of primary creep varied with the casting
and in this condition, 3 hours at 1190°C (2174°F) caused source. Times to rupture were fairly consistent, but the times to
significant dissolution of the cubic y'. Quite probably the Re 1% creep strain on samples from the "good" source were between
"wall" at the y/y' interface, observed in the atom-probe work, is two and four times as long as those from the "worst" source.
effective in restricting growth and dissolution of the y'. When the standard homogeneity check was carried out, a good
correlation Was observed between homogeneity and creep life;
12
the standard deviation for Re. was 1.23% in the worst samples,
1.1
falling to 0.6% in the best ones. The casting source with the
highest thermal gradient and solidification rate gave the lowest
1.0
standard deviation for Re. A proposed explanation for this effect
is as follows:
In cast and heat treated single crystals, the dislocations are
u5 concentrated in interdendritic regions (Pollack and Argon, 1992).
OA As deformation occurs, these dislocations multiply and spread
throughout the structure. In more heavily segregated test pieces,
0.7 the dislocation motion through the y matrix channels at the start
0 of the 1080°C (1975°F) test (ie., primary creep) will therefore
0.0
occur in regions low in rhenium, and hence low in creep strength.
05
A rapid rate of primary creep would therefore be expected. In the
vcco 29030 MOO 30000 31000 32300 most segregated sample referred to above, it was estimated that
Larson - Miller Parameter the weakest 10% of the structure contained only 3.0 - 4.3% Re
and 4.5% W, so might be comparable with homogeneous CMSX-
Figure 5 - Gamma Prime Size in CMSX-4 vs. Soaking
4 with 2.9% Re + 6.4% W. Once dislocations have spread
Condition, hrs/°C
throughout the structure the overall creep rate will be a function
of the average composition of the alloy, hence segregation has
y' stability is further improved in the 6% Re alloy CMSX- less effect upon the stress-rupture life.
10 (RR 3000). During creep tests at 1175°C (2147°F) a stable y'
rafted structure developed in a few hours, and was still stable Oxidation.Behavioi
after 60 hours testing (Fig. 6). The oxidation performance of 3% and 6% Re containing
single crystal alloys at 1100°C (2012°F) for instance, is
remarkably good bearing in mind the low Cr content of these
materials. Normally, in cast superalloys with 5.5 - 6.2%

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 r

aluminum (Al) (Strangman, et al., 1980) one would not expect a Solutioned
stable, protective a alumina film to form with less than about 8% 1177°C (2150°F) Dynamic Oxidation Test
Cr present, yet CMSX-10 (RR 3000) has quite good performance 450 hours
Mach 0.45 Cyclic
in this respect, with only 2% Cr. Once ;pin, Re could be the key (once per hr)
element. Research (Chen and Little, 1995) showed that Re did
not enter the oxide film, but it did concentrate in the y' depleted ,-
zone beneath the oxide. Figure 7 is taken from the work of Chen .—Continucas.
Min. *lumina
and Little (1995) and demonstrates mice as much Re in this Vg2 0i wide
scale Slyer
region than in the base alloy. It is proposed that this Re Son tack
concentration slows down the diffusion of elements such as Ti
into the aluminium oxide scales, so increasing the oxide scale
stability. Residual ppms (10-20 ppm) of yttrium (Y) and
lanthanum (La) have been shown to dramatically improve the
bare oxidation resistance and coating performance on CMSX-4
alloy (Fig. 8,9 & 10) (Thomas, et al, 1994, Korinko, et al, 1996)
when the sulfur (S) content of the alloy is <2 ppm.
Depleted faYor 10 srfl 20 om
3 II

CMSX-4 Mod A. (bare) (15 ppm Y) (< 2ppm S)

is
Figure 8
2
Atomic % Re

1.5

0
ro IS 20 25

Distance from Oxide/Metal Interface, microns

Figure 7 - Re Distribution Beneath The Oxide! Metal

Interface
Figure 9 - Becon Burner Rig Dynamic Cyclic
HIP & Solutioned
Oxidation Bare Alloys 1038°C (1900°F), 0.4 Mach
1177°C (2150°F) Dynamic Oxidation Test
450 hours
Mach 0.45 Cyclic
(once per hr) IS -o- 1151 -San
-go- psi -SEP KAI
-4.- 551 - ASP Akaft

JP-5 Rai
Or*: 60 Mona Rame Dwell
Weiss Coll *kr 131151

503 1500 2600 ISO) 1010 1503

Time (MOSS)

Depleted layer 94 WO
Figure 10 - Becon Burner Rig Dynamic Cyclic
20sm
Depth of wide:ion/diffusion voids 182 pm )--f Oxidation CMSX-4 + La [16ppm] Bare & Coated [CM
Heat V8614] 1038°C (1900°F) 0.4 Mach
CMSX-4 (bare) (<2 ppm S)

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 It has also been found that La gives improved control of its The important factor from a turbine engine performance
ppm chemistry in the single crystal casting process compared to viewpoint is of course the effect which these TCP phaq•s have
Y. It is now also known that the La and Y tie-up the residual S upon mechanical properties. Some TCP phases particularly in
as very stable sulfides. The latest burner rig data indicates ppm multi-grain cast superalloys have a marked embrittling effect, but
combinations of La + Y may give the best results at 1038 ° C these do no; their effect is Z.:educe creep strength when a
(1900 ° F) and 1093 °C (2000 °F) test temperatures. S wealcens the certain volume fraction is formed by concentrating Re and W into.
strong Van der Waal's bond between the alumina scale and the an ineffective form, in effect de-alloying the material. During the
base alloy. CMSX- I0 (RR 3000) development programme, the amount of
TCP phase was estimated by a point - counting technique. The
Phase Stability point was counted if it fell either on a TCP needle or on its y'
One aspect of critical importance in these high Re envelope, so in effect the proportion of the structure that was
superalloys is the metallurgical stability, ie., the rate of formation . other than the normal y + y' was measured. To give one
of topologically-close packed phases (TCPs). These are not example, unstressed exposure of CMSX-10 (RR 3000) for 250
present significantly in the practical use of CMSX-4, but in hours at 1100 °C (2012 ° F) gave an area fraction of TCP's of 5%,
CMSX-10 (RR 3000) the operating conditions of components but this had no deleterious effect upon the impact strength, high
have to be carefully considered against the TTT curve for or low cycle fatigue strength . of the alloy. A substantial deviation
formation of these phases. CMSX-4 shows continuing linear in creep strength was only seen when the area fraction of TCFs
relationships for log stress to log stress-rupture life, (Fig. 11) with approached 20%. In that condition, the creep elongation was still
no fall-off due to excessive TCP phase formation, out to the in the range of 13 to 18%, confirming these TCP's do not have an
extent of current testing: 5,600 hrs. at 1121 ° C (2050 °E), 12,400 embrittling effect upon this single crystal alloy at this area
hrs. at 1093 ° C (2000 ° F), and 17,000 hrs. at 982 ° C (1800 °F). • fraction level.
The composition of these TCP phases has been established by
Chen and Little (1995) as shown in Table VI. Thermal Fatigue (TFI
1030
The thermal fatigue behavior of CMSX-4 and SRR99 has
been investigated on blade-shaped single edge wedge specimens
500 (Meyer - Olbersleben, et al., 1992, Meyer - Olbersleben, et al,
'NYC (INV") 1996). The strain was measured at the wedge tip of the IF
specimens. Temperature-mechanical strain cycles with different
wart otorri
canes. • sa. se • ISIS mre Corn
mean strain values and strain ratios were obtained. The strain
• 100 I Hat 170C11.10C • lb in f
nrc 110:071 AC
distribution on the edge was presented. Further, TF crack
0.
2
MAO. it

"l DP.
.070•0
mire morn initiation life and total life depend on the strain cycle, which itself
50 is temperature, gradient dependent On the basis of TF stain
.64,0731 Waal
Wall WS. MS. MM.
NHS. VaVa WZR,V011,
measurements, a new TMF cycle was introduced. An integrated
NW& Sit nal. Nail. ■1561 approach for TF and TNT investigations was proposed.
1 1 I I I 1111 1 1 1 1 11111 1 1 11 1 III Preliminary investigations show that under the test conditions
10
10 50 103 SOO 1003 5000 1COSO used, TF cycling is more damaging than IMF. IT crack
Rupture Lila 000 --00 initiation mechanisms for both superalloys were identified.
Figure 11 - Stress - Rupture CMSX-4 Alloy (Average Finally, the higher TF crack . initiation resistance of CMSX-4 is
001) explained by its higher oxidation resistance combined with a
higher mechanical strength of its y'-depleted zone and y/y'-
microstructure (Fig. 12) (Meyer - Olbersleben, 1996).
Table VI
In SR1299 and CMSX-4, residual cast microporosity on the
TCP Phases Nit %) wedge tip led to stress concentration. For both alloys under high
Ni Co Cr Mo W Re Al Ti strain loading, cracks always initiate on these porosities at the
CMSX-4 20 8 9 2 28 28 1 4 specimen surface, Fig. 12a + b.
Under low strain loading the number of thermal cycles (NO
CMSX -10 (RR3000) 30 2 6 1 10 44 2 5 to crack initiation is much higher. Nevertheless, the same
mechanism of crack initiation on microporosities was observed
Unlike the y' phase referred to earlier, the compositions of for CMSX-4, Fig. 12d. For SRR99, a complex mechanism of
the TCP phases formed are clearly different in the two alloys. oxidationJspallation/reoxidation combined with the effect of the
They are both basically .Ni-Cr-W-Re, but the Re: W ratio in residual cast microporosity was identified as the crack initiation
particular is much higher in CMSX-10 (RR 3000), reflecting the mechanism, Fig. 12.
basic chemistry differences of the two alloys.

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 HIP treatments were found to enhance IT resistance by
closing the microporosity.

Low Cycle Fatigue (LCF)

CMSX-4 shows a benefit over non-Re containing alloys
when the time-dependent LCF mode is in operation where
dislocation activity and crack growth are mainly confined to the
y matrix phase. If the LCF data are plotted as stress for a given
cyclic life vs temperature, the temperature advantage of CMSX-4
over SRR 99 is about 45°C (81°F) which is similar to the creep
strength advantage at high temperatures.
Under cycle -dependent conditions, where dislocations and
cracks slice through y and y' phases ace, the improvement for
CMSX-4 is less. However, if an improved fatigue performance
• is required in this regime, there is a solution: H1P'ing. In both
plain specimens and notched specimens (KT = 22), fatigue
failures initiate at single crystal casting microporosity and the
fatigue life at a given cyclic stress and temperature can be related
to the size and shape of the micropores [one 100 an
interdendritic micropore is more damaging than hundreds of 30
jam spherical pores at their normal uniform spacing in single
crystal castings.]
The LCF life improvement at 950°C (1742°F) of the Re -
c) d) containing alloys and HIP'ing is shown in Table VII.
200 WC

a) SRR 9311 Inn =1150t, as-mec--0.75%, N1.1= 680, Table VII

b) CMSX441., Tran =1150t, M=1080,
Strain Controlled LCF 950°C (1742°F) R = 0
c) SRR 93, Trnax =1 1 50 t Umn=0.53%, Ni= 5200, NI at 0.7% Strain Cycles
d) CMSX4S, Tnn =1100t, ae,“=0.48%, M=28000..
DS Mar M 002 10,000
Figure 12 - View of the wedge tip, SEM micrographs [Long]
DS CM 186 LC 19,000
• Wedge tipsof CMSX-4 specimens remained almost intact [Long]
during low strain loading even after very high numbers of thermal
cycles (Fig. 12d), showing its high resistance to oxidation, while CMSX-4 (001) 50,000
wedge tips of SRR99 specimens were highly damaged by oxide- [Unhipped]
scale spalling.
CMSX-4 (001) -100,000
For both afoys investigated, crack initiation was always [Ripped]
observed only after an incubation period, which is dependent
upon the strain range and the maximum temperature of the
thermal cycle. Alleracks were initiated at the wedge tip surface CASTABILITY
and propagated towards the bulk For all tests, an initial increase RR investment foundries have cast well over 150 tonnes of
in the crack gowt rate (CGR) was followed by a pronounced Re containing superalloys over the last 10 years. Principle
decrease when thetrask length was between 2-3 ram. The exact applications are turbine blading for both military and civil
beginning of this crack growth retardation depends upon Ac me, engines. During the period of introduction RR has worked
and Tran closely with CM and has evaluated, in-depth, the alloys CM 186
The reduction of thermal gradients in the specimen depth LC, CMSX-4 and CMSX-10 (RR 3000). Each alloy has
and subsequent decrease in thermal strains and stresses at the presented the foundry with interesting challenges and as a result
crack tip is the mais reason for crack growth retardation on blade- considerable understanding of each alloy's behavior has been
shaped specimens. For longer cracks, the lower temperature and gained.
the lower interactias with oxygen should also be considered as an From a castability view point, CMSX-4 performed well
additional cause. No significant difference in the CGR was with little difference observed in gain selection and quality to
observed between SRR99 and CMSX-4 for the same TF test first generation alloys such as RR2000 and SRR99 when cast
conditions. under conditions developed for these alloys. However, a revision

8
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use
 Table VIII
Critical Chemistry (wt To or ppm] CMSX-4 Alloy Blend V-3 3629 kg (8000 lb) Heats
HEAT# BLEND (0) Al 11 Zr Si Fe
RATIO PPIll ppm PPm PPm Pan ppm
V8331 SOR/SOV 18 2 <20 2 2 5.63 1.02 18 .01 .040
a. V8481 50FU50V 17 2 <20 2 1 5.67 1.02 25 .02 .026
V8562 50R150V 20 2 <20 1 1 5.61 1.04 17 .02 .037
V8563 50R150V 21 2 <20 1 1 5.66 1.02 22 .01 .036
V8634 50R/50V 21 2 <20 1 1 5.63 1.02 17 .01 .040
V8640 60R/40V 25 2 <20 1 1 5.63 1.03 17 .02 .038
V8653 50FUSOV 19 2 <20 1 1 5.65 1.03 32 .01 .037
18656 50R/50V 23 2 <20 2 1 5.64 1.02 30 <.01 .032
V8657 505/SOV 22 2 <20 1 1 5.66 1.03 30 .01 .030
V8678 505/SOV 18 2 <20 2 2 5.66 1.04 17 .01 .031
V8820 605/40V 28 4 <20 1 2 5.66 1.01 22 .01 .045
V8821 60R/40V 27 3 <20 2 1 5.68 1.01 21 .01 .044
V8848 60(9140V 32 3 <20 1 1 5.70 1.01 18 .01 .059
V8857 505ISOV 30 2 <20 1 1 5.68 1.02 39 .01 .061
V8876 60R/40V 24 3 <20 2 2 5.65 1.01 32 .01 .051
V8877 60F140V 27 3 <20 3 1 5.65 1.01 32 .01 .044
V9042 60R/40V 24 3 <20 1 1 5.65 1.03 24 .01 .053
V9081 SOFVSOV 30 3 <20 1 2 5.67 1.02 39 <.02 .054
V9119 40R/60V 30 1 <20 1 1 5.64 1.02 10 <.01 .067
V9179 soRroov 25 2 <20 1 1 5.66 1.02 60 <.01 .089
V9180 60R/40V 26 3 <20 1 1 5.64 1.03 17 <.01 .067

MEAN 24 2 <20 2 1 5.65 :1.02 25 .01 .041

STD. OEV. .6 .6 .4 .02 .01 7 .010

of casting conditions showed CMSX-4 to be less prone to freckle The testabilitv assessment consisted of studies of the
chains and freckle generated defects than SRR99. Production propensity to DS grain boundary cracking, porosity and ceramic
experience has confirmed the general absence of these defects. shell/core reaction. RB211 HP and IP turbine blades were cast
CMSX-4 has shown no particular propensity to high angle with a variety of casting conditions and mould assembly designs
boundary formation and recrystallisation at 1.6% critical strain as to understand the behavior of the various chemical modifications.
is typical of other single crystal alloys. The results of these trials showed the benefit of reducing Si and
During the development of CMSX-10 (RR 3000) over 10 Zr levels to minimise any tendency for DS grain boundary
• chemistry iterations were considered to meet the testability and cracking. The optimised chemistry showed no adverse foundry
mechanical property objectives, including microstructural problems or other unusual problems. Casting yields, based on
stability. The development programme showed certain rejections associated with the alloy, were very high and
chemistries to be sensitive to freckle formation, particularly at equivalent to the best materials evaluated by RR Grain structure
low casting temperatures. However, a satisfactory combination defects such as freckle chains were not encountered.
of mechanical properties were achieved and production Although no components are in production at RR with this
experience has shown no excessive tendency to any common alloy, castings have been produced for both military and civil
single crystal defect. Foundry yields are in line with first and demonstrator engines. These castings were produced from mixed
second generation single crystal alloys. virgin/revert ingot utilising scrap 3% Re containing CMSX-4
Table VIII lists the critical chemistries of the twenty-one - alloy and balancing virgin elements to create the CM 186 LC
3629 kg (8000 lb) blend heats of CMSX-4 manufactured to date. composition.
The ability to recycle CMSX-4 foundry revert to these high
quality standards which ensure the blend heats perform quite as ALLISON ENGINE TEST RESULTS
well as 100% virgin heats, has resulted in significant single Initial engine desiga for the T406, AE 2100 and AE 3007
crystal component cost reduction, which along with significant engines incorporated a variety of nickel-base superalloys in the
advances in single crystal casting technology result in turbine high pressure turbines, including CMSX-3, IN 738 C and MAR
*airfoil yields often > 90%. M 247. These alloys launched and certified this family of gas
CM initiated a collaboration with RR in 1987 to establish turbine aero engines; however, increased demand for higher
the testability of a series of CM 186 LC alloy variants with power and lower specific fuel consumption has necessitated
chemical iterations controlling the level of the wain boundary increased turbine entry temperatures. To accommodate the
strengthening elements C, Zr, and B and the residual element Si..

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 increase in temperature, improved airfoil alloys must be utilized. destructive dye penetrant inspection, metallurgical evaluation has
Significant development, as well as certification testing, has been also been done, but this has been limited due to its destructive
done with CMSX-4. nature and the desire to reuse the turbine hardware. In addition
The family of aero gas turbines mentioned above currently to full engine testing, component testing on a hot fatigue rig has
has amassed 7700 hours of high pressure turbine testing using been used to quantify the endurance limit for a representative
CMSX-4: the distribution is 2360, 2790 and 2550 hours for the sample of CMSX-4 blades.
HPT 1 blade, HPT 2 vane, and PITT 2 blade, respectively. The Only very limited distress has been noted in the CMSX-4
HPT 1 blade (Fig. 18) is an air-cooled component, while the HPT airfoils after testing and the engine results indicate that the
2 vane and blade have been tested and certified utilizing no relative improvement for CMSX-4 is perhaps greater than
cooling air. Turbine airfoils manufactured from CMSX-4 have originally anticipated. This has been particularly true for the
performed successfully in a number of key test vehicles: engine multi-airfoil segmented 2nd vane (Burkholder, et al., 1995);
#ps468 (150 hour development type test), engine #A300717 (Figs. 13-17 incl.) this vane has passed a 500 hour ASMET and
(official 150 hour FAA type test), engine #A300130 (500 hour currently is undergoing testing in a 1000 hour ASMET test. At
development Accelerated Simulated Mission Endurance Test the 900 hour mark the engine was disassembled and visually
(ASIvIET)), engine #A300704 (official FAA overtemperature inspected. The CMSX-4 vanes were found to be in excellent
test) and engine #A300131 (official 1000 hour ASMET). Post condition with only minor indications of any hot section damage.
test inspections generally include a visual, dimensional, and non-

Figure 14 - IN 738 C Cooled Equiaxed 2nd Vane

Following 300 Hrs. of DPT Testing - Trailing Edge Airfoil
Bowing.

rRef .060 bow

0. 5 i cy New Parts
IN Mr 150 hr.
DPT
3

0 5

0. 5

0. 6
Figure 13 - IN 738 C Cooled Equiaxed 2nd Vane
F.
Following 300 Hrs. of DPT Testing - Leading Edge Becton BB Se000 MH
Outerband and Trailing Edge Airfoil Cracking.
Figure 15 - Dimensional Inspection Results of the
CMSX-4 Uncooled 2nd Vane Airfoils As-New and
Following 150 Hrs. of DPT AE Series Engine Testing.

10
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use
 ss-4.Aia

-2sSrtn.:
, 0

-.sr "'"1•11-:'
,

Figure 16 - CMSX-4 Uncooled 2nd Vane Segment • Figure 17 - CMSX-4 Uncooled 2nd Vane
Following 150 Hrs. of AE Series Engine OPT Testing Segment Following 150 Hrs. of AE Series
Engine OPT Testing.

Serpentine Cooling Configuration

Figure 18 - AE Series HP1 Turbine Blade Following

ASMET Engine Testing - CMSX-4 Alloy

Figures 19 - 22 incl. show the y, y' microstructures of a With this successful development and certification
CNISX-4 HPT 1 cooled turbine blade following 1093 hrs. of database, flight test engines have already been shipped using
ASNIET testing in a AE series engine. It is apparent that at 50% CMSX-4 airfoils, and production AE 3007A engines
span, the convex wall has only seen modest temperatures, with incorporating CMSX-4 airfoils shipped in late 1996.
higher and similar temperatures at the concave wall and leading
edge. with the highest temperatures at the trailing edge,
particularly concave side.

•1

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 10 ban Trailing Edge Convex Side
Figure 21 - SEM Micrographs y, y' Microstructure AE
Series HPT 1 Cooled Turbine Blade CMSX-4 Alloy
Following 1093 Hrs. of ASMET Testing.
50% Airfoil Span Trailing Edge

10grn Concave Wall

Figure 19 - SEM Micrographs y, y' Microstructure AE
Series HPT 1 Cooled Turbine Blade CMSX-4 Alloy
Following 1093 Hrs. of ASMET Testing.
50% Airfoil Span le 2nd RIB 10 ALT Trailing Edge Concave Side
Figure 22 - SEM Micrographs y, y' Microstructure AE
Series HPT 1 Cooled Turbine Blade CMSX-4 Alloy
Following 1093 Hrs. of ASMET Testing.
50% Airfoil Span Trailing Edge

Figure 20 - SEM Micrographs y, y' Microstructure AE

Series HPT 1 Cooled Turbine Blade CMSX-4 Alloy
Following 1093 Hrs. of ASMET Testing
50% Airfoil Span

10mni Leading Edge

12

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 ROLLS-ROYCE MILITARY ENGINE EXPERIENCE Pegasus LP1 Vane
In order to extend the life of the current equiaxed C1023
peoasus HP1 and I4P2 Turbine Blades alloy LP I NGV, it is proposed to cast the vane in CMSX-4.
The F402-RR-408 engine in service with the USMC is Castings are due in January 1997, with endurance testing starting
now being fitted with "sand tolerant" HP! and HP2 turbine blades late 1997.
in single crystal CMSX-4. Fleet service experience to date
includes 25 engines fitted with these blades, the lead engine Pegasus HP2 Vane
being at 291 hours (Jan. 1997). The current HP2 NGV is cast in equiaxed PD21 alloy and
Development bench engine testing has now demonstrated suffers from leading edge oxidation/burning" in earlier versions
a service life of 2000 hours. (2 off ASMET tests, each of 530 hrs. of the Pegasus engine, resulting in a high reject rate at overhaul.
endurance running time) after which the blades were in good To increase the life of the component, a customer-funded
condition (Fig. 23). Turbine entry temperatures reached 1397°C programme has been initiated to validate a vane cast in CMSX-4.
(2547 ° Fj (1670°K) simulating ISA+34°C conditions. It is anticipated that this will offer a 100°C (180°F), increase in
material property capability. Castings are due in January 1997
and following endurance testing, will be offered as a
modification.

Adour LP Vane
Two engine sets of castings (Fig 24) have been produced
in CMSX-4 alloy, for development testing. The current material
is equiaxed C1023. One of these sets has been machined, and is
currently carrying out engine testing with total naming time now
at 10 hrs. (Jan. 1997).

Figure 23 - Pegasus CMSX-4 HP1 and HP2 Turbine

Blades (Leading Edge)
On Right: Aluminised L106 ( x2 ASMETS)
On Left: Pack Aluminised ( x1 ASMET)
Figure 24- Adour LP Vane CMSX-4 Alloy - As-Cast
pegasus LP1 Turbine Blade
A development programme has just been carried out, with Adour HP Vane
the LP I blade cast in CMSX-4. The production part is in SRft99. CMSX-4 casting trials have been carried out as singles and
This has been pursued in order to both increase creep life in now are being carried out as triples. These vanes are currently
service, and to enable engine uprate capability. An ASMET conventionally cast in C1023 alloy. An order for 10 sets of parts
cyclic endurance test has just been successfully carried out on a of triples has been placed to support the 2000 hr life engine
set of CMSX 4 LP I turbine blades. (530 hr. test). It is planned
- programme. Finished parts are due mid 1997, for engine testing
to offer this modification to the customers for the engine. leading to certification in late 1998.

13
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use
 Adour HP and LP Turbine Blades programme in the laboratories and casting trials to establish
For the 2000 hr engine life initiative, it is planned to viable production processes (Table IX).
introduce CMSX-4 HP and LP turbine blades to replace the
existing DS MAR M 002 and SRR 99 alloys respectively. Design Table IX
work is completed. CMSX-4 and CMSX 10 (RR 3000) Applications
Lead Detect
Bench
ROLLS-ROYCE.CIVIL ENGINE EXPERIENCE Component Material Hours
Serulce Entry Into
Cycles Service
The RB211 and Trent family of engines powers many of
R821 1-5240/H CMSX-4 3300 3050 Dec 94
the large civil aircraft being used for passenger carrying service, High-Pressure
CM5X-4 5200 2000 Mar 95
Turbine sta.- Trim 700
including L1011, 8747, B757, B767, A330 and 8777. Cooled "NMI 800 CM5X-4 4750 800 Apr 98
All currently manufactured engines in the R8211 family
IP Turbine Trent 700 C24SX-4 5200 2000 ?Aar 95
feature a high bypass ratio, a wide chord fan, and a 3-shaft Slade -
(Incanted Tram 800 CMS% -10 4750 803 Apr 96
system. The 3-shaft concept apart from providing a more (FIR3CCO)
optimised matched thermodynamic cycle, incorporates an HP and iP R9211-5243A1
intermediate pressure (TP) shaft which rotates significantly slower Shroud Tram 700 C51524 - - -
Segment
than the high pressure shaft, enabling the rp turbine blade to Liners
Trent 800

remain uncooled. Figure 25 shows the general arrangement of

Trent turbofan engine, showing the components where Re
The HP turbine blade applications were driven by the need
containing superalloys are used.
to minimise the amount of cooling air used without incurring
creep penalties either in the airfoil or the shroud. All blades were
HP turbine IP turbine
blade blade subjected to rigorous bench engine test programmes which
included temperature surveys, dynamic measurements, feed air
pressure and 150 hour type tests and realistic simulated service
cycle tests. The condition of the RB211-5240/1-1 CMSX-4 blade
-taw 'iv t;Maliwil P
after 5000 cycles is compared to the DS blade after 3750 cycles
in Fig. 26.

ases
s
HPT segment IPT segmen
liner liner

Figure 25 - Trent Turbofan

Earlier versions of the RB211 family were certificated with

directionally solidified MAR M 002 nickel based alloy HP and IP •
turbine blades. However, the continual drive for improved
specific fuel consumption has resulted in increased overall
pressure ratios (OPFts) and turbine entry temperatures (TETS).
The latter are now routinely above 1527°C (2781°F) [1800°K]
during a 150 hr. type test, approximately 200°C (360 °F) higher
than the melting point of the materials used in the high pressure
turbine. The cycle efficiency is also improved by minimum use
DS MAR M 002 CMSX-4
of cooling air in the high pressure turbine which in turn requires
improved material properties and cooling techniques. Figure 26 - Comparison between DS MAR Ni 002 blade
Additionally, the airfoil material must be capable of accepting (left) and CMSX-4 blade (right) following endurance
protective coatings, and in particular thermal barrier coatings engine testing.
(TB Cs).
It was these considerations which led Rolls-Royce to select The Trent 800 HP blade, shown in Fig. 27 represents the
CMSX-4 and CMSX- l0 (RR3000) for the following engine most advanced application of CMSX-4, operating at the highest
applications after an intensive materials characterisation OPR and TET, essentially as a cantilevered blade but with a

14

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 shroud for aerodynamic performance and performance retention
advantage. This blade is currently being evaluated with a nth
airfoil advanced TBC system for higher thrust versions of the
Trent 800.

Figure 27 - Trent 800 HP Blade Cast In CMSX-4

The 1P turbine blade in a 3-shaft engine enables the blade Figure 28 - Trent 890 IP Turbine Blades CMSX-10 (RR
to operate =cooled and hence results in improved cycle 3000) Alloy Following ETOPS Engine Testing.
efficiency relative to a 2-shaft engine where the second stage HP
turbine blade must be cooled. Because the blade is uncooled, with natural gas or liquid natural gas (LNG) as the predominant
operating at about 1000°C (1832 °F), the successive increases in fuel (Harris, et al., 1992, Kubarych and Autrecoechea, 1993,
TET have demanded materials with better creep and oxidation Brentnall, et al., 1997). Figure 29 shows a photomicrograph
properties. Figure 28 shows IP turbine blades in CMSX-10 (RR from a MARS 100 first stage CM5X-4 turbine blade after over
3000) alloy which met the Trent demands for improved creep 25,000 service hours. The CMSX-4 substrate alloy is coated with
resistance. The condition of the 13) turbine blades after all of the a platinum aluminide coating applied by a pack cementation
bench tests has been excellent. process (RT-22). The photomicrograph was obtained from an
Finally, the use of CMSX-4 is being extended to HP and axial airfoil section near the blade tip, at the convex wall near the
IP shroud segments, which form the rotor path, to overcome the leading edge. Coating condition appears to be excellent
component plastic deformation seen in certain applications. SNECMA have successfully completed initial 400
Analytical work has shown that the improved creep properties of equivalent cycle ASMET type military engine testing with DS HP
CMSX-4 will result in components which will not incur the vanes in CM 186 LC alloy with good results, which confirm the
plastic strain. The hardware is currently undergoing bench excellent transverse LCF properties and coating performance of
engine evaluation. the alloy. (Bourguignon, et al., 1996)
European Gas Turbines Ltd has now validated blading in
: . OTHER ENGINE EXPERIENCE both CM 186 LC and CMSX-4 alloys for application to their
Solara Turbines, Inc. has reported that six years of field range of industrial gas turbines. CM 186 LC has been chosen as
experience for CMSX-4 first stage blades in the MARS 100 a cost effective DS HP rotor blade alloy for the Typhoon gas
industrial turbine has been excellent Total running time for the turbine to provide enhanced life margins at its latest 4.9 MW(e)
132 engines in the field is 1.25 million hrs, with the blades and rating (Figs. 30 and 31). This cooled blade (Fig. 32) is now in
Pal coatings in good condition when examined at engine overall full production and first engine deliveries with this standard of
after 26,000 - 28,000 Firs service. The MARS 100 engines tend blade commenced in September 1996. The performance of the
to spend at least 50% of their miming time at maximum power alloy in the Eaundry has been very encouraging with no problems
encountered during casting development and with yields

15

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 25,000 Service Hours 200 x
Figure 29 - MARS 100 1st stage CMSX-4 Turbine Blade -
Airfoil Axial Section Near Blade Tip - Convex Wall Near
Leading Edge Figure 31 - Typhoon Genera Arrangement

approaching 90% early in production. The first application of An extensive in-house materials testing programme on
CMSX-4 is for an uncooled HP rotor blade in the Hurricane 1.6 both materials has been in place to provide the mechanical and
MW(e) gas turbine to replace the y'/ODS alloy blade previously physical property data necessary for design, to provide long term
specified for this application. creep data and to carry out corrosion and oxidation testing and
For both applications, the blades have been fully validated coating trials. The latter have confirmed the acceptability of
in development engine testing. This has included strain gauge these alloys for long term industrial gas turbine applications with
testing to determine in-engine vibration modes and establish HCF the selected layered silicon aluminide coating system (Sermalloy
margins, infra-red pyrometry to measure blading metal 1515).
temperatures and a cyclic endurance test to simulate up to three Further applications of these alloys in the EGT product
years of cyclic operation in service. range are currently being pursued.

Figure 30- Typhoon Industrial Engine

16
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use
 Brentnall W.D., Aurrecoechea J.M., Rimlinger • C.M.,
(Solar), Harris K., Erickson G. L., Wahl J. B., (CM), June 1997,
'Extensive Industrial Gas Turbine Experience With Second
Generation Single Crystal Alloy Turbine Blades' ASME (IGTI)
Turbo Expo '97, Orlando Fl., USA.
Burkholder P. S., Thomas M. C., Frasier D. J., Whetstone
J. R. (Allison) Harris K., Erickson G. L., Silckenga S. L. and
Eridon J. M. (CM), April 25-27, 1995, 'Allison Engine Testing
CMSX-4 Single Crystal Turbine Blades and Vanes', 10M 3rd
International Charles Parsons Turb. Cant Proc.,Newcastle upon
Tyn. e, UK
Camel F., Bourguignon S., Lallement B., Fargeas S.,
DeBussac A. (SNECMA), Harris K., Erickson G. L. and Wahl J.
B., (CM), June 10-13, 1996, 'Snecma Experience With Cost
Effective DS Airfoil Technology Applied Using CM 186 LC
Alloy', ASME Gas Turbine Cong. and Exhib., Birmingham, UK
[96-GT-493].
Cetel A.D. and Duhl D.N., (PWA), Sept 1992, 'Second
Generation Columnar Grain Nickel-Base Superalloys, 7th
International Symposium , pp 287-296.
Cetel A. D. and Duhl D. N. (PWA), Sept 1988, 'Second-
Generation Nickel-Base Single Crystal Superalloy, 6th
International Symposium*, pp 235-244.
Figure 32 - Typhoon Cooled HP Turbine Blade
DS CM 186 LC Alloy Chen J. H. and Little J. A., (Cambridge University, UK),
1995, Unpublished Work.
Doner M. and Heckler J. A. (Allison), Oct. 1985, 'Effects
of Section Thickness and Orientation on the Creep-Rupture
SUMMARY
Properties of Two Advanced Single Crystal Alloys', Aerospace
The year 1997 is seventy-two years since the discovery of
Technical Conference, Long Beach, CA.
the metallic element Re. Over the last ten years Re has been
Erickson G. L. (CM), Sept. 1996, 'The Development and
successfully used as a critical strengthening element in cast
Application of CMSX-10', 8th International Symposium', pp 35-
nickel-base superalloys for single crystal and directionally
44.
solidified Mrbine airfoils.
Ford D. A. and Arthey R. P. (RR), Oct 1984,
Turbine engine test and service experience has generally
'Development of Single Crystal Alloys for Specific Engine
exceeded expectations for the Re containing superalloys. This
Application', 5th International Symposium*, pp 115-124.
paper explains these results relating alloy properties to turbine
Fullagar K. P. L., Broomfield R. W., Hulancis M. (RR),
engine component performance for three of the alloys.
Harris K., Erickson G. L. and Sikkenga S. L. (CM), June 13-16,
1994, 'Aero Engine Test Experience with CMSX-4 Alloy Single
ACKNOWLEDGMENTS
Crystal Turbine Blades', 39th ASMEIIGTI International Gas
The authors wish to acknowledge the significant
Turbine & Aero Engine Congress & Exp., The Hague and Trans.
contributions to this work from many personel at RR, Allison,
ASME Jrn. Eng. Gas. Turbines and Power, April 1996.
Solar, EGT, SNECMA, CM, Howmet and PCC (Airfoils).
Garrett, U.S. Patent #4,935,072.
Gell M., Duhl D. N. and Giamei A. F. (PWA), Sept. 1980,
REFERENCES
'The Development of Single Crystal Superalloy Turbine Blades',
Bachelet E. and Lamanthe G. (SNECMA), Feb. 26-28,
4th International Symposium', pp 205-214.
1986, National Symposium - Single Crystal Superalloys, Viallard-
Gell M., Duhl D. N., Gupta D. K. and Sheffler K. D.
de-Lans (France).
(PWA), July, 1987, JOM, pp 11-15.
Blavette D. (Facult6 des Sciences de Rouen), Caron P. and
Giamei A. F. and Anton D. L. (UTRC), Nov. 1985,
Khan T. (ONERA), Oct. 1986, Scripta Met Vol 20 No. 10.
'Rhenium Additions to a Ni-Base Superalloy: Effects on
Blavette D. (Faculte des Sciences de Rouen), Caron P., and
Microstructure', Met Trans A. 16A, pp 1997-2005.
Khan T. (ONERA), Sept 1988, 'An Atom-Probe Study of Some
Goulette M. J., Spilling P. D. and Arthey R. P. (RR), Oct
Fine Scale Microstructural Features in Ni-Base Single Crystal
1984, 'Cost Effective Single Crystals', 5th International
Superalloy', 6th International Symposium*, pp 305-314.
Symposium*, pp 167-176.

17
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use
 Harris K., Erickson G. L. and Schwer R. E. (CM), 1983, Pollack T. M. and Argon A. S. (MIT), 1992, Aeta
Development of the Single Crystal Alloys CMSX-2 and CMSX-3 Metallurgica at Materialia, 40, pp 1.
For Advanced Technology Turbines', ASME Paper 483-6T-244. PW 2037 Engine Display, Sept 1994, Farnborough Air
Harris K., Erickson G. L. and Schwer It E. (CM), Oct. Show (PWA),
1984, 'MAR M 247 Derivations - CM 247 LC® DS Alloy, Roan F. (Westinghouse Electric Corp.), 1996 - Private
CMSX41) Single Crystal Alloys, Properties and Performance', 5th Communication.
Inteitational Symposium*, pp 221-230. Ross E. W. and O'Hara K. S. (GE), Sept. 1996, 'Rene N4:
Harris K, Erickson G. L. and Schwer R. E. (CM), Oct. 6- A First Generation Single Crystal Turbine Airfoil Alloy With
9, 1986, 'CMSX Single Crystal, CM DS and Integral Wheel Improved Oxidation Resistance, Low Angle Boundary Strength
Alloys Properties and Performance', Cost 50/501 Conf. Liege and Superior Long Time Rupture Strength', 8th International
Proc., pp 709-728. Symposium*, pp 19-25.
Harris K., Erickson G. L. and Schwer R. E. (CM), Frasier Ross E. W. and O'Hara K. S. (GE), Sept 1992 b, 'Rene
D. J. and Whetstone J. It (Allison), Sept 24-27, 1990, 'Process 142: A High Strength, Oxidation Resistant DS Turbine Airfoil
and Alloy Optimization for CMSX-4 Superalloy Single Crystal • Alloy', 7th International Symposium'', pp 257-265.
Airfoil', Cost Conference Liege, Proc. Pan II, pp 1281-1300. Strangman T. E., et al. (Garrett), Sept 1980, 'Development
Harris K, Erickson G. L., Silckenga S. L. (CM), Brentnall of Exothermically Cast Single Crystal MAR M 247 and
W. D., Aurrecoechea J. M. and Kubarych K. G. (Solar0), Sept Derivative Alloys', 4th International Symposium*, pp 215-224.
1992, 'Development of the Rhenium Containing Superalloys Thomas M. C., Helmink R. C., Frasier D. J., Whetstone J.
CMSX-4 and CM 186 LC For Single Crystal Blade and R. (Allison), Harris K., Erickson G. L., Sildcenga S. L. and Eridon
Directionally Solidified Vane Applications in Advanced Turbine J. M. (CM), (Oct. 3-6, 1994) 'Allison Manufacturing, Property
Engines', 7th International Symposium*, pp 297-306. and Turbine Engine Performance of CMSX-4 Single Crystal
Holmes J. W. et al. (MIT), O'Hara K S. [GE (AE)J, 1988, •Airfoils', Cost 501 Conf Proc, Liege.
Thermal Fatigue Testing of Coated Monocrystalline Superalloys', Walston W. S., O'Hara K. S., Ross E. W., Pollock T.M.
ASTM STP 942, Phil, pp 672-691. and Murphy W. H.. (GE), Sept. 1996, 'Rene N6: Third
Khan T. (ONERA) and Brun M. (Turbomeca), June 1989, Generation Single Crystal Superalloy', 8th International
Symposium on Single Crystal Alloys, MTU/SMCT, Munich. Symposium, pp 27-34.
Korinko P. S., Barber M. J. and Thomas M. C., (Allison), Wortrnann J., Wege R. (MTU), Harris IC, Erickson G. L.
June 10-13, 1996, 'Coating Characterization and Evaluation of and Schwer It E. (CM), June 29, 1988, 'Low Density Single
DS CM 186 LC and SX CMSX-4', ASME Gas Turbine Cons and Crystal Superalloy CMSX-60, 7th World Conference on
Erhib, Birmingham, UK. Investment Casting Proc., Munich.
Kubarych K. G. and Aurrecoechea J. M. (Solar) 'Post Field • Wukusick C. S. and Buchakjian L. Jr. [GE (AE)1], March
Test Evaluation of an Advanced Industrial Gas Turbine First 13, 1991, 'Improved Property - Balanced Nickel-Base Superalloys
Stage Turbine Blade', Oct. 17-21 1993, TMS/ASM Mat Wk '93 for Producing Single Crystal Articles' UK Patent Application
Proc., PA tlGB 2 235 697A.
Meyer - Olbersleben F., Rezai-Aria F. (Swiss Fed Institute Wulcusick C. S., [GE (AE)), Aug. 25, 1980, Fnl. Rep.
of Tech) and Goldschrnidt D (MTU), Sept 1992, 'Investigation of 'Directional Solidification Alloy Development', NAVAIR Contr.
the Thermal Fatigue Behavior of Single Crystal Nickel-Based N62269-78-C-0315.
Superalloys SRR 99 and CMSX-4', 7th International Yamazaki M., et al. (NR1M), Oct 1984, 'Alloy Design for
Symposium* pp 785-794. High Strength Nickel-Base Single Crystal Alloys', 5th
Meyer - Olbersleben, F. Engler-Pinto Jr., C. C. and Rizai- International Symposium*, pp 157-166.
Aria F., 1996, 'On Thermal Fatigue of Nickel-Based Superalloys', Yukawa N., et al. (Toyohashi Univ.), Sept 1988, 'High
Thermomechanical Fatigue Behavior of Materials: Second Performance Single Crystal Superalloys Developed By The d-
Volume. ASTM 577 3 1263 Michael J. Verrilli and Michael G. Electrons Concept', 6th International Symposium*, pp 225-234.
Castelli, (1996) &is , American Socieorfor Testing and Materials.
Miglietti W.M. and Pennefather R. C. (CSIR), 19961
ASME Gas Turbine Cong and Exhib, Bilmingham, UK. • Superalloys, Seven Springs, PA., TMS Proc.
Pessah M., Caron P. and Khan T. (ONERA), Sept 1992,
'Effect of p Phase on The Mechanical Properties of a Nickel-Base
Single Crystal Superalloy', 7th International Symposium, pp
567-576.
Petrov D. A. and Tumanov A.T., 1973, The Use of Single
Crystal Blades, Aircraft Engineering No. 9.

18

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 10/26/2017 Terms of Use: http://www.asme.org/about-asme/terms-of-use