NASA Technical Reports Server (NTRS) 20010061358: Development and Evaluation of Directionally-Solidified NiAl/(CR,MO)-Based Eutectic Alloys for Airfoil Applications

10 Pages·1 MB·English

by  NASA Technical Reports Server (NTRS)

DEVELOPMENT AND EVALUATION OF DIRECTIONALLY-SOLIDIFIED NIAL/(CR,MO)-BASED
EUTECTIC ALLOYS FOR AIRFOIL APPLICATIONS
S.V.Rajt,I.E.Locci2and.I.D.Whittenbergerl
|NASA GlennResearchCenter,Cleveland,OH 44135
2CWRU atNASA GlennResearchCenter,Cleveland,OH 44135
Abstract acombination of alloying and mierostructural controls.Thus, several
investigators have studied DS NiA[ eutectic alloys, wherethe eutec-
The results of recent efforts todevelop directionally-solidi fled alloys ticphases are ideally parallel to the growth direction, in an effort to
based onthe Ni-33AI-31Cr-3Mo eutectic composition arediscussed. optimize the mechanical properties [3-6].
These developmental efforts included studying the effects of
rnacroalloying andgrowth rates on microstructure formation as well Inparticular, the Ni-33A1-34Cr [7,8] and the Ni-33Al-(34-x)Cr-xMo
as the elevated temperature compressive and tensile properties of [7-14] eutectic alloys, where 0< x<6%, have drawn greater attention
these alloys. These observations revealed thatcontrary to conven- since the earlier investigations of Walter and Cline [4-6] demon-
tional opinion, the cellular microstructure was stronger and tougher strated that the substitution of Me for Cr leads to a corresponding
than the planar eutectic microstructure due to a microstructural change in the fiber morphology from Cr rods to (Cr, Mo) lamellar
refinement of the cell size and interlamellar spacing. The high plates when the Me content exceeds 0.7%. This change in fiber
temperature strengths of these alloys are compared with those of morphology offers the possibility of improved fracture loughness
commercial superalloys and advanced NiAI single crystals. The and elevated temperature creep properties.
implications of this research on airfoil manufacturing and applica-
tions are discussed. Conventionally, ithas been assumed that these in-situ composites
require the two phases to be grown parallel toeachother and tothe
Introduction growth direction as a planar eutectic microstructure in order to
maximize the mechanical strength in the longitudinal direction.
Although high pressure turbine airfoil vanes were successfully However, planar eutectic microstructures developin the NiAi/(Cr, Mo)
manufactured from a high strength NiAI single crystal alloy and system only when the solidification rate, Vi, is typically less than
engine tested afew years ago, further developments on this material 25 mm h"Idepending onthe Me content ofthe alloy [5,12,13]. These
were not continued owing to the brittleness of the alloy [1,2]. rates are generally too slow for economic commercial production of
Directionally-solid!fied (DS) NiAI eutectic alloys have been inves- any alloy even if it is demonstrated to possess all the requisite
tigated as alternative materials for airfoil applications due to their properties required forturbine airfoil applications. However, solidi-
higher elevated temperature strength and fracture toughness com- fication at faster rates leads to the development of cellular eutectic
pared to binary NiAI [3]. Two eutectic systems, namely the NiAI- microstructures, where the (Cr,Mo) and the NiAI plates emanate
34(at.%)Cr I and tbe NiAI-9Mo, have been studied ingreater detail, radially from the cell interior to the cell boundaries [12-14]. Cellular
where the microstructure consists of either Cr or Me fibers inaNiAI microstructures are also likely to form even with small amounts of
matrix. These alloy systems offer the possibility that both the creep alloying elements varying between 0.25 and I% even when Vi is
andthe fracture toughness properties can beeffectively improved by sufficiently slow to produce a planar eutectic microstructure in the
IUnless otherwise mentioned, allcompositions are reportedinatomic
percent in this paper.
This is a preprint or reprint of a paper intended for presentation at a
conference. Because changes may be made before formal
publication, this is made available with the understanding that itwill
not be cited orreproduced without the permission of the author.
S.V. Raj I of 10
unalloyed NiAI/(Cr, Mo) material [15]. Clearly, it would be techno- NiAI/(Cr, Mo), or a three-phase, NiAII(Cr, Mo)ICr2Ta, eutectic
logically relevant tostudy the mechanical properties ofthese cellular microstructure was the desired objective for the directional-solidi-
microstructures todetermine if attractive mechanical properties can fied alloys, these microstructures were not attained when they were
be obtained under fast growth conditions. processed by arc-melting. For example, the base composition repre-
sented by AM-0 revealed an arc-melted microstructure consisting of
Historically, cellular microstructures have generally been consid- 92.3% of the two-phase (Cr,Mo)/NiAI eutectic and 7.7% of the
ered to be undesirable for high temperature structural applications primary NiAI dendrites (Table I) although the NiAI dendrites were
due poor creep properties induced by cell boundary deformation usually absent when this alloy was directionally-solidified [I0,12,13].
processes. However, growing at faster rates is likely to lead to a Thus, the volume fractions of the phases are likely to be different if
refinement in the interlamellar spacing, _ in accordance with the these alloys had been directional-solidified. Duplicate, and in one
Jackson-Hunt relationship [16]: _.2VI= constant. This relationship case triplicate, melts of five of the alloys revealed that the variation
has been shown to be valid for the NiAI-34Cr [4] and the Ni-33AI- in Viola Iwas between 0.9 and 8.5%. However, larger variations inthe
31Cr-3Mo [13] eutectic alloys with cellular microstructures. Hence, amount of the individual phases were observed in some of the alloys,
the effect of growth rate on the room temperature fracture toughness and in the case ofAM-I 1and AM-12, some of the phases were not
and elevated temperature strength isunknown for the NiAI/(Cr, Mo) consistently observed despite similarities intheir chemical composi-
alloys. Investigations on the effect of withdrawal rates onthe micro- tions. While noting these limitations of the are-melting technique for
structures and mechanical properties of these alloys could have processing the DOE alloys, this method was chosen for its relative
important commercial and engineering consequences. simplicity in these early attempts in order to narrow down and
identify suitable alloy compositions for directional solidification and
The objectives of this paper are primarily to summarize the results of detailed study atalater date. The data shown inTable 1were used as
studies on theeffect of macroalloying and growth rate on microstruc- the input inthe DOE model, which was then allowed topredict other
ture formation and elevated temperature properties of directionally- alloy compositions with the intention of maximizing the volume
solidifed Ni-33AI-3 ICr-3Mo alloys. Details of the creep properties fractions of the eutectic microstmctures and minimizingotherphases.
of the Ni-33AI-33Cr- IMo alloy and the effect of growth rate on room Unfortunately, this approach to alloy design proved tobe unsuccess-
temperature fracture toughness will be published separately [I8,19]. ful since the statistical analysis did not give reproducible results
presumably due to the wide range of compositions considered and
Background scatter present inthe arc melted microstructures. Nevertheless, three
compositions predicted by the model were directionally-solidified
Statistical design of experiments approach to alloy design and their high temperature compressive strength and room tempera-
ture fracture toughness were evaluated [I1].Since these alloys were
Aprogram todevelop directionally-solidified NiAI/(Cr, Mo) eutectic brittle with nosignificant improvement intheir elevated temperature
alloys for airfoil applications was initiated at the Glenn Research strength, it was concluded that the DOE technique was unsuitable as
Center, Cleveland, Ohio, in 1995. Itcommenced with an attempt to an alloy design method in the present instance.
design alloys using astatistical design of experiments (DOE) approach
[11,17]. Forty new alloy compositions were formulated bysimulta- _1 _,uidelines for alloy design
neously and independently varying the compositions of several
elements (AI, Cr, HI',Si, Ta and Ti) ineight-dimensional space based However, certain insights were gained from the DOE results (Table I),
on aDOE model using Ni-33AI-31Cr-3Mo as the base composition. which formed the basis for designing the next generation alloys.
The Mo content was kept constant at3% while nickel was used as a Thus, alloys containing Hf and Si generally tended to promote
filler. Duplicate compositions of five of the alloys were also cast in intercellular segregation, especially when Hf/Ni >0.025 andSi/Ni >
order to comply with the statistical requirements of the DOE model. 0.01. Later studies on DS alloys microalloyed with about 0.25% Si
The elements, Hf, Si, Ta and Ti, were chosen because they were confirmed that this element by itself does not promote intercellular
known toimprove the high temperature creep strength of NiAI. Since segregation [15].Tantalum tended to promote theformation of three-
it was impractical todirectionally-solidify allthese compositions, the phase eutectic mierostructures when the ratio Ta/Ni >0.125 although
alloys were prepared by are-melting (AM). Their microstructures higher Ta/Ni ratios also had atendency toencourage the precipitation
were qualitatively and quantitatively examined using optical and of(Cr, Ta)primary dendrites. For values ofTa/Ni < 0.125, the volume
scanning electron microscopy (SEM) techniques, and the phases fraction of the intercellular region increased with increasing values
were identified by x-ray diffraction (XRD). Details of the alloy of the Ta/Ni ratio. Titanium by itself did not promote intercellular
compositions, their mierostructures and the volume fractions of the segregation even with a 10%addition (e.g., AM-22 andAM-29) and
individual phases are tabulated in Table ! [17]. A close examination no precipitates were observed inthese alloys. Alloys containing Tias
of Table I reveals that despite large differences in the compositions well asHf, Si andTa showed agreater propensity towards intercel-
of these alloys, the commonly observed microstructures included lular segregation. As a result, Hf and Si were not used in the first
primary dendrites consisting either of (Cr, Mo), (Cr,Ta) or NiAI generation directionally-solidified alloys although detailed micro-
phases, and the two- and three-phase eutectic morphologies. The structural andmechanical properties studies were later conducted on
term "'other phases" reported inthe table includes (Hf,Ti) dendrites, alloys containing less than I% of these elements [15]. The total
unidentified phases and interdendritic segregation. volume fractions of the eutectlc microstructures decreased sharply
when the Ni content in the alloys deviated more than 3% from the
The total eutectie volume fraction, Vlota I, given in Table Irefers to ideal base composition of 33% (Fig. l(a)) and when the
the summation of the volume fractions of the two- and three-phase (AI+Cr+Mo+Ta+Ti)/Ni ratio significantly varied from about 2.3
eutectic microstructures. Although a complete two-phase, (Fig. 1(b)). These guidelines limited alloy design to anarrow range
S.V. Raj 2 of 10
i
_o_ooo _ o_o__ ._
,.._._._,, ,_ ....... _.. _.-_ -_ ....
i ...... . .< • • • "I_IIIIIIIIIII_ _I I "_II • II
i_ IIIIIII_ IIIII+ ....
iiiiiiiiiiiililiii IIiiliiiliii1
v-
S.V.R,_j 3 ofI0
100 I I I t.L L I and the chemical compositions of the alloys have been described
elsewhere [10--15,18,19] and only abrief description is included in
/' "q" J"' _N
80-- /..."_-o":E-'_'. _ -- this paper. Induction-melted rods of selected alloy compositions
, -..'.'.-.-", .'.._. -_, were directionally-solidi fled by amodified Bridgman method under
I a temper_lture gradient of 8-10 Kmm'l. The variations in chemical
60--
" '."7" _ V..:_
composition and microstructures ofthe basecomposition were found
l'.:.'."l _1 "_.?..! to be within acceptable limits both along the length of DS zone
40-- I.:; .qo ¢_I :..t
I.:•.'.'1 0 _1 (~100 ram) as well as from one batch to another for rods grown at
t'.'..'.'i _ j
/,_', 12.7 mm hl [10]. Three alloys, NASA-0, NASA-2.1 and NASA-
20-- I".o.: 71 29.1, were processed at this growth rate• Having established the
ability to dircctionally solidify the base eutectic alloy rods in a
I j I I I i
0 (a) l reproducible manner at a single growth rate of 12.7 mm h-I, addi-
0 5 10 15 20 25 30 35 40 45 tional rods of the Ni-33AI-33Cr- IMo (NASA-0. I)and the Ni-33Ai-
Ni (at. %) 3ICr-3Mo (NASA-0) eutectic alloys were grown at rates varying
between 7.6 and 508 mm h"t. The DS rods were sectioned for
I I I a
chemical and metallographic analyses aswell as for the machining of
creep and fi'acture toughness specimens as described elsewhere
[10,12,13]. Optical and scanning electron microscopy (SEM) and
back-scattered electron (BSE) imaging techniques were employed
A
1060080t I • - for microstructural observation. In addition, x-ray diffraction (XRD)
o , _".':::_'::::.:.._. and energy dispersive spectroscopy (EDS) analysis was used for
u _,._. :.:;-."_,
phase identification.
II o v".%:..'":.":.::'.:'.:_:.'.". --
20 JIi %V@Q.':'..:"._'..".;.:,:o_.-"...'"...__ -- The long axis of the creep specimens were parallel to the growth
; _.;._'," "d'." direction. Compression specimens were 8 x4 x4 mm in size while
o II I I, i I Ivq.''..':_:._:_.:3 "go tensile specimens had arectangular cross-section of3.2x 2.2 mm and
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 a gage length of 25.4 ram. Ridges, which were machined on the
(AI+ Cr +Mo +Ta + Ti)/Ni tensile creep specimens to permit the attachment of an extensometer,
allowed the gage length tobe precisely determined. The as-cast EDM
Figure I.--Variation of thetotal volumefraction ofthe
layers were removed bygrinding the surfaces on600 grit SiC paper.
eutectic phases in the DOE alloys with (a) nickel content
Constant engineering strain rate and constant load creep tests were
and (b) the ratio (AI + Cr+ Mo + Ta +Ti)/Ni.
conducted inair between absolute temperatures, T, 1100 and 1400 K
around the base compesition, AM-0, inorder tomaximize the volume under engineering strain rates, _, varying between 2 x 10.7and 2 x
fraction of the eutectic micmstructure. 10-4 s"! and initial stresses, (_, varying between 30 and 200 MPa,
respectively [I0,14,15]. The tensile creep properties compared rea-
The following selection criteria were developed to narrow the list of sonably well with compression data obtained by constant load
alloy compositions for directional solidification and detailed micro- compression creep and constant engineering strain rate techniques
structural and mechanical property studies. First, alloys with atotal [10]. Thus, it was possible to determine the elevated temperature
eutectic content less than 70% were dropped from consideration. strength of these alloys by conducting compression tests which
Second, alloys containing significant amounts of(Cr,Ta) dendrites or allowed more specimens to be tested from asingle DS rod.
those showing intercellular segregation were deleted from the selec-
Results and Discussion
tion list since it was felt that these alloys could not be effectively
toughened. Third, alloys containing significant amounts of NiAi
Mierostructures
dendrites were not considered owing to possible detrimental effects
on high temperature creep properties. However, alloys AM-2 and
AM-29, which consisted primarily of the eutectic and (Cr, Mo) The DS eutectie microstructures observed in both NASA-0.1 and
dendritic microstmctures, showed promise for further development. NASA-0 alloys were dependent on the growth velocity. Examples of
It was felt that the small amounts of (Cr,Ta) dendrites observed in planar eutectic microstructures observed on transverse sections of
AM-2 could be suppressed with either small modifications to its NASA-0. Iand NASA-0 are shown inFigs. 2(a) and(b), respectively.
composition or by directional solidification. Thus, two new alloys The dark and light phases observed inthese microstmctures areNiAI
were designed anddirectionally-solidified: NASA-2. Iwith anomi- and (Cr, Mo), respectively. Additional details pertaining to qualita-
nal composition of Ni-33AI-29Cr-2Ta, and NASA-29.1 with a tive andquantitative descriptions ofthese microstructures arediscussed
nominal composition of Ni-32.5AI-29Cr-3Mo-3Ti. inorder tomain- elsewhere [13,14,18]. Planar eutectic microstructures, which formed
tain consistency, directionally-solidified base composltion, Ni-33AI- only when V_< 12.7 mm h"l, consisted of alternating parallel plates
31Cr-3Mo, is designated as NASA-0 in this paper• of the (Cr,Mo) and NiAI phases within each eutectic colony. Although
both alloys had similar microstructures at these slow growth rates,
Experimental Procedures faceted (Cr,Mo) rods were sometimes observed inthe Ni-33AI-33Cr-
IMo alloy. At growth rates exceeding 12.7 mm hq, the microstruc-
Details of the DS technique, the test methods used fordetermining the ture changed from planar eutectic colonies to cells, where the
constant engineering strain rate and constant load creep properties, (Cr,Mo) and the NiAI plates emanated radially from thecell interior
S.V. Raj ' ' 4 of 10
atgrowth speeds _254 mm h"l, although there were regions adjacent
to the cell boundaries in the NASA-0.1 alloy, where the (Cr, Mo)
phase had broken down into particulates. However, an important
difference was observed inthe microstructures of these two alloys at
agrowth speed of 508 mm hi, where the (Cr,Mo) phase had broken
down into short plates inthe NASA-0. ialloy. Incontrast, the NASA-
0alloy had long (Cr,Mo) plates extending the entire length of thecell.
Figures 3(a) and (b) show the longitudinal and transverse sections of
alloy NASA-2. Icontaining Ta, respectively, while Figs. 3(c) and(d)
show the corresponding microstructures for the Ti-containing
NASA-29. Ialloy. Both these alloys were grown at 12.7 mm h"i.In
comparison totheplanareutectie microstmclure observed inNASA-O
grown atthis speed (Fig. 2(b)), NASA-2. Iexhibited acoarse cellular
eutectic microstructure consisting of broken (Cr,Mo) platelets (light
gray phase in Fig. 3(a)) and primary NiAI dendrites (Fig. 3(b)). In
__,N,_N,___3_3AI-31Cr-3Mo _x_ 1 addition, the Cr2Ta laves phase was also observed in this alloy (white
phase in Fig. 3(a)). Low magnification observations of longitudinal
sections of NASA-29. [showed abanded microstmcture consisting
of alternating regions of coarse and fine (Cr, Mo) lamellae [10].The
transverse sections showed that the microstructure had a cellular
morphology (Fig. 3(d)) with an internal microstructure consisting
only of (Cr,Mo) and NiAI(Ti) phases (Fig. 3(c)).
N
Elevated Tem_rature Deformation Pro_nies
The creep properties of the NASA-0. I alloy are described in great
detail elsewhere [14,19] and will not be discussed here. Inthe early
experiments conducted to determine batch-to-batch variations in
chemical composition, microstructures and elevated temperature
strength of the NASA-0 alloy grown at aconstant growth speed of
12.7 mm h"t, itwas observed that some batches contained regions
_I_N_I • - _ _-_
with NiAI dendrites (Figs. 4(a) and (b)). The presence of these
dendrites were detrimental to creep properties asshown inFigs. 5(a)
and(b), which show plots of true tensile creep strain, ¢.,versus time,
t, and tensile _ versus e, respectively, at 1300 K under an tensile
engineering stress of 50MPa. Itis clear from Fig. 5thatbatch !1/96
was considerably weaker than batch 12/96 due to the presence of
NiAI dendrites in the microstructure (Figs. 4(a) and (b)). However,
with further refinement in the processing technique, microstructural
observations on later batches of material did not reveal any NiAI
dendrites. Figure 5 also compares the constant load tensile creep
curves for NASA-0 with those for NASA-2. !and NASA-29. I.These
three alloys exhibit ashort normal primary transient, where the creep
rate decreases sharply with increasing creep strain to a minimum
Figure Z--Optical microstructures of
value before continuing toincrease tofinal rupture (Fig. 5(b)). For the
NASA-0 andNASA-0. I alloys direction-
ally solidified at(a) and(b) 12.7mmh-I creep conditions shown in Fig. 5, the Ta-containing NASA-2. I was
and(c) and(d) 50.8 mmh-t. Micmgraphs stronger than both the base composition NASA-0 as well as the Ti-
(a) and(c) representNASA-O. I while (b) containing NASA-29.1 below _' < 7%. However, under higher
and (d) are for NASA-0. stresses and lower temperatures, NASA-29.1 was observed to be
comparable or stronger than NASA-0 and NASA-2.1 and the creep
to the cell boundaries (Fig. 2(c) and(d)). in comparison to theplanar properties of the latter alloy were comparable to NASA-O [10].
eutectic microstructures, where thewidth of the colony boundaries
were narrow and often indiscernible, the average width of the cell Figures 6(a) and (b) confirm that directional solidification results in
boundary varied between 20and 25wn irrespective of growth speed animprovement in theelevated temperature strength of the NASA-0
[I3]. Microstmctural observations of longitudinal sections of these alloy for both the planar (Fig. 6(a)) and cellular (Fig. 6(b)) eutectic
alloys revealed that the eutectic grains were several millimeters long microstructures, where the specimens were grown at 12.7 and
for theboth the cellular and planar morphologies butthe orientation 50.8 mm h'l,respectively, withthestrengtheningeffectbeinggreater
of thelamellae were notalways parallel tothegrowth direction. The at the lower temperatures. The strain rateexhibits achange from a
microstructures observed in NASA-0. I and NASA-0 were similar power-law to an exponential dependence with increasing stress for
S.V. Raj 5 of 10
"ASA-2. m
NASA-29.1
es_
Figure 4.--Comparison ofthe transverseopticalmicro-
(c) )__m structures oftwo batchesof Ni-33AI-3 ICr-3Mo direc-
tionally solidified at 12.7mm h-I; (a) DSrun 1!/96 and
(0) DS run12/96.
values represent one standard deviation about the mean,and varied
between 405.1:1:29.9 and539.4 +24.3 kJtool"Ifor thedirectionally-
solidified specimens. The corresponding stress exponents, n, were
4.6 + 0.2 for the as-cast materials and varied between 4.4 +0.2 and
5.6 +0.2 fortheDS alloys. The growth ratedid notappear toinfluence
theacti ration energy inany consistent manner thereby indicating that
the deformation mechanism wasindependent of theDS microstruc-
lure. Nevertheless, the magnitudes of Qc for the DS materials are
Figure3._a) Back-scattered electronimage generally higher thanthose reportedfor polycrystallin¢ high purity
and(0)optical micrograph of longitudinaland Cr forwhich Qc_=305kJmol'l inasimilar rangeoftemperature [20].
transversesectionsofNi-33AI-29Cr-3Mo-2Ta, Although the magnitudes ofQ¢fortheas-castDS alloys arecloserto
respectively; (c) back-scatteredelectron image the values of440 and455 kJtool"1reported for NiAI <001> single
and (0) oixical micrograph of longitudinaland crystals [21], itcannot be unambiguously concluded that deforma-
transversesectionsofNi-32.5AI-29Cr-3Mo-3Ti, tionoftheNiAI phasecontrols thedeformation oftheeuleclic alloys.
respectively.
Figures 7(a) and(b) showthe variation ofthe creep ratewith stress
theas-castspecimens atall temperatures. A similar dependence is for NASA-2.1 and HASA-29.1, respectively, in the temperature
alsoobservedatall temperatures forthe planareutectic microstruc- range 1100 to 1400 K.The datainclude compression and tension
ture(Fig. 6(a)), andat!200 and1300 Kforthecellular microstructu res results fromseveralbatches[10].Thestressexponent variesbetween
(Fig. 6(0)). However, the cellular microstructures exhibit apower- 5.7 and 6.0 for NASA-2.1 while it varies between 3.8 and 4.4 for
lawdependenceat 1400 K(Fig. 6(b)). A multiple regression analysis NASA-29.1. Multiple regression analysisofthe dataforthesealloys
of theintermediate andlow stressdataassuminga power-law stress assumingapower-law stressdependence revealed thatn,',,5.5 ±0.2
dependence suggestedthattheactivation energy fordeformation, 0.c, andQc_ 419.3 ±21.8 kJmol"1forNASA-2. I andn _"4. I+0.2 and
was 372.3 -I-29.6 kJtool"ufor the as-castmaterial, where theerror Qc _ 529.7 ± 25.8 kJtool"1for NASA-29.1. Once again, no deep
S.V. Raj 6of I0
3s I I I I I I I t , I ,I,I, I t , t ,i,t,
_ NiAl/(Cr_Mo) ,, _ Ni--33AI--31Cr-3Mo (NASA-0)
30 Alloys I VI = 12.7 mm h-1
25 -- T=1"T3-_K |,,,,- Ni-32.SAI-29Cr-3Mo-3"ri -- Condition
_" 20 --
er=50 MPa f (NASA-29.1) - --t--.)
"_" 15 _NASA-O /_t-Ni-33AI-31Cr-3Mo -- 10-2 ::::i As-Cast ']
(11/96) # (NASA-0) (12/96) _
10 /// Ni-33AI-29Cr-3Mo-2"ra ) -- 10-3
o _ I _'t-'T"]_l I I (a) lO.5
0 100 200 300 400 500 600 700 800 140
t(h) _ 10-6 "_
10-04k . ' I I I --_ 10-7
N ICr'm°> Z--NASA-(01.m) 10-8 ,o: '--1200K
10-05 _ T=1300 K . ,,_ r- (NASA-0) -_ 10-9 __. 1300 K_
I_ e=soMPa_ ,_ (12/96)
--10-06__ I 10-10::(a) I' I 'l'l'l'''"l , , , ,I,n:
I = I I = I lill=
• 10_07 __ _/ _Ni-323AI-29Cr-3Mo-3"ri.._ Ni-33AI-31Cr-3Mo (NASA-0)
(NASA-29.1) 1 VI =50"8 mm h-1
10-08 _ t_Ni-33AI-29Cr-3Mo-2"l'a _ Condition
_.. (NASA-2.1) ":-I --e.. --"_
I I I (b)-'1
10-09[-" I --e--./>As-Cast
0 5 10 15 20 25 10-2 _ - -i--j t _" "_
•(%) 0-3_ --o--}DS
Figure5.---(a)Tmetensilecreepstrainvs. timeand(b)tme 10.-4 _[- -.-o---J s,_/s i_ /
tensilecreepratesvs.truecreepstrain plotsfortwobatches :- l,,'l::I(l,s (_" .._
..-Z,-'J ]
._ 10"6 ._E- ,_,1=" _ )'; /"
insights can be gainod from these values of naid Q¢except that they 10-7 _- / "_1/_ ,f¢{
suggest that the deformation mechanisms operating in these two -- _n,"" _ ! ",, i
alloys atthese temperatures are probably similar to those acting in 10"8 _ _13= K'_ "q-1200 K
NASA-0. 10-9 _r
10-10-(b) I ,¢ I=l,l,I I , I,t,rT
Figure 8 compares the compressive and tension creep strength for
0 100 1000
different batches of NASA-0, NASA-2. Iand NASA-29. !with those
o (MPa)
for polycrystalline NiAI [22], DS Ni-33AI-28Cr-6Mo [7], cnjomilled
NiAI-30.4% AIN [23] and NASAIR 100 superalloy single crystals Figure 6.---Comparison of thecompressivetrue strain rate
[24]. The data for NiAI, Ni-33AI-28Cr-6Mo and NiAl-30.4% AIN vs.compressivetruestressplots foras-cast(solid symbols)
representcompressive propertiesat1300 Kwhile thoseforNASAIR andDS (opensymbols) NASA-0 grownat(a) 12.7and
were determined by tensile creep at 1273 K. Several points may be (b) 50.8 mmh-I between 1200 and1400 K.
noted from Fig. 8.First, the DS alloys aremuch stronger than binary
NiAI. Second, the compression and tension data are comparable
although there is significant scatter in the data. Third, there is no
advantage inalloying NASA-0 either with Taor Ti since there was Effect of Growth Rate onElevated Tem.oeralure Strength
no effective improvement in the elevated temperature strength.
Fourth, the DS alloys are much weaker than cryomilled NiAI-30.4% In the absence of a suitable alloying scheme for improving the
AIN and NASAIR 100 superalloy single crystals. Furthermore, our elevated temperature strength of NASA-0, one possibility was to
studies have also revealed that many other alloying elements even verify ifmicrostmctaral refinement would result in any strengthen-
when added in small amounts effectively destroy theinitially planar ing effects. Two ways in which this refinement can be achieved isby
microstructure without any significant improvements in the high signi ficantly increasing the thermal gradient orthe equipment and by
temperature strength [! 5]. These results suggest that macro- and withdrawing the DS ingots at higher speeds. Since the thermal
microalloying of NASA-0 isunlikely tobe aneffective method for gradient of the equipment could not be increased without a major
improving itselevated temperature strengthundertheDSconditions modification of itsdesign, microstmctural refinement was achieved
usedinthepresent study. byincreasing the growth speed (Fig. 2) [13].Figu re9shows theeffect
S.V.Raj 7 of 10
10-3 5- I t I I 1,1_1 I Jjl r lll_ 1 ' I _I'l'J I _ I I I'1
B Alloy Batc.__._h Testmode
lo-4 / / 1300K -=
_=- • NASA-0 8/96 Tensioncreep
1.0 / / 1200K -
• NASA-0 12/96 Tensioncreep
lo-5 Ill=- • NASA-O 13/96 Tensioncreep
=-
6.0_ o o NASA-0 4/98-6/99 Compression
lO-6 n NASA-2.1 14/96 Compression
_=_
• NASA-2.1 14/96 Tensioncreep
10-7 0 NASA-29.1 15196 Compression
r 1400
7- __ ..___adP 1100K -. • NASA-29.1 15/96 Tensioncreep
lO-8 F" • ! 10-3 /
10-9 7_ polycrystal/ /._" -_
_r Ni-33AI-29Cr-3Mo-2Ta
1°-4_ / _ "'":'_- :
lo-lO [ I I I IIlll I [ I JIKI_
- lO- 4=
lO-3
--- I I I IlJll I ! I) ,I I
- 2 /Present .[:_:_,_s i// _--
lO-4 F 14
•"k" ";t _
lO-5 :.".9:'.:'..: I, "a.
-.//7"//
10-9 -'-/
10-6
7"4--
=- 10-10 _
.W 10-7 "=l 10 100 1000
lO-8 --- _: Jz_,444 1100K (MPa)
"=1
Figure 8.mComparison of the elevated temperature
lO-9 Ni-32.SAI-29Cr-3Mo-3Ti "el strengths of the DS NASA-0, NASA-2.1 and NASA-29.1
lo-lO "Col I J I i 111t[ I , i I ,I, n: at 1300 K with those for NASAIR 100 single crystals
[24], binary powder-metallurgy extruded NiAI polycrystals
10 100 1000
[22], DS Ni-33AI-28Cr-6Mo [7] and cryomilled
(MPa)
NiAI-30.4% AIN [23].
Figure 7,reStrain ratevs. stress data for directionally solidified
(a)NASA-2.1 and (b) NASA-29.1 between 1100 and 1400 K.
Itislikely that the elevated temperature strength ofNASA-0decreascs
in a manner similar to the observations on Ni-33AI-34Cr [4] and
of growth rate onthe elevated temperature strength atapproximately Ni-33AI-33Cr- IMo (Fig. 9(a)). Correlations of the strength observed
1300 K for < 10-4 s"] (Fig. 9(a)) and room temperature fracture at !300 Kwith the average cell size, d, (Fig. IO(b))and interlamellar
toughness, Kp (Fig. 9(b)) [18] as the microstructure changes from a spacing, _, (Fig. 10(c)) suggest that microstructural refinement is
planar eutectic toaradial cellulareutectic microstructure forNi-33AI- effective only at strain rates greate_rthan 10-6 s"]. It is clear from
34Cr [4], Ni-33AI-33Cr-! Mo and Ni-33AI-31Cr-3Mo with increas- Figs. 10(b) and (c) thatboth d and ;Linfluence the magnitudeofth..¢
ing growth rate. All three alloys show an increase in strength elevated strength at 1300 K. Since there is acorrelation between d
(Fig. 9(a) and toughness (Fig. 9(b)) as VI increases from 7.6 to and _ [13], the similar dependencies shown in Pigs. 10(b) and (c) is
50.8 mm h"I.However, Ni-33Ai-33Cr- 1Mo loses its strength as well to be expected. Nevertheless, Figs. I0(b) and (c) clearly demonstrate
as its toughness as the growth rate increases above 50.8 mm h"l. In that microstructural refinement due toan increase inthe growth speed
contrast, Ni-33AI-3 ICr-3Mo continues to increase in strength while can only lead to limited improvements in the elevated temperature
retaining itsroom temperature toughness athigher growth rates. This strength. Fun..her microstructural refinements leading toasignificant
observation demonstrates that it is possible to improve both the decrease in _,will require the thermal gradients to be increased by
elevated temperature strength as well as the room temperature factors of two or more using either the edge-defined film (EDF)
fracture toughness for some NiA! eutectic alloys by directionally growth technique, fluidized bed or molten metal cooling technology.
solidi fying them atfastergrowth rates. Although thecurrent generation This conclusion is consistent with the limited tensile data published
of alloys are still weaker than the first generation superalloy single byYang etal. [9]on EDF grown Ni-33AI-33Cr- !Mo alloy, where its
crystals (Fig. 8), the observations inFig. 9demonstrate that itmay be strength at 1273 Kiscomparable to that of NASAIR 100 [24] if itis
commercially feasible todirectionally solidify future generations of superposed on Fig. 8.
new NiAI eutectic alloys at fast enough growth rates without a
significant deterioration in their mechanical properties. l.arson-Miller Plots
The strength of NASA-0 at 1300 Kwas observed to increase with Figures 1l(a) and Co)compares the variation of the applied tensile
increasing VIonly atstrain rates above 10-5s-z(Fig. 10(a)). Similar creep stress and the density-compen sated tensile creep stress with the
observations were made at 1200 K. However, the growth rate had no Larsen-Miller parameter (LMP) [251, T(Iog tf+ 20), where tfis the
effect on the elevated temperature strength when the strain rate was creep rupture time inh, respectively, forseveral current and potential
about 10"6s"1orat 1400 Kfor t varying between 10-6 and 10-4s"l. airlbil alloys. The commonly assumed value of 20 for the LMP
S.V. Raj 8of I0
NASA-0 NASA-0_ 400 I l'l'l' i I _ I ii'ti I I ' l l'lll'
350 --
300 -- 10_4 --
_" 250 -- _10_ 5 --
g 200 --
12.7 mm h-1 lz,'mmh- _ _-_ 150 -- n o o o ",-.g. 1.0OX10-6 --
I 100 --
I
400 I
i ' I'4'I' I 'I l I llilt[ ztl _ lll'l _ 50 i-a)_ , l]Ithl I , I,lthl I , Irich
% I 0
350 1 10 100 1000
1300K-_, , __-----"_'_ -_
VI {ram ITI)
300
400 I I I I I I I
250
_-"'_ 1300 K---" "%-,--] 350- --=__..__ _ _2.1x10-4 --
o.
200m
Planar Radial 300
b eutectic eutectic 250 --
150 colonies cells 0- ".__2.1X 10._5 _-
200 --
100 Ni-33AI-34Cr 150 -- ".._1.0x10_ 6 o --
Ni-33AI-33Cr-1Mo
50: B oNi-33AI-31Cr-3Mo -- 100 --
0 (a) l , i,l,r,l i l , lililll I i I,I,I, 50--(b) I I I [ I I t
o
25 100 120 140 160 180 200 220 240 260
t ' lil'l*J =I i l'ltl'l I =I'I'I i
I (tzm)
INi-33AI-31Cr-3Mo --1
400
I i I |
A_ 15 -- -- 350 _xl0 -4 --
3oo - ._ --
_ lO- _- 250 --
et
g 200 --
e_ O_P O O=.-
150 -- "_--1.OX10-6__ o --
iNi-33AI-33Cr-1Mo
100 --
C°)l , I,l,t,I ',1 , lll_l,I I _ I,l,h Ni-33AI-31CP3Mo
0 50--
10 100 1000
o (c) I I I I
VI(mm h-l)
0.0 0.5 1.0 1.5 2.0 2.5
Figure 9._Variation in (a)the strength at 1273 or 1300 Kand [ (_.rn)
(b) the room temperature fracture toughness [18] with growth
ratefor NiAI-34Cr [4], NASA-() andNASA-0. I.The repre- Figure 10.--Effect of (a)growth rate, (b) average cell size and
sentative microstmctures at the low and high growth rates are (c) average interlamellar spacing on =hecompressive strength
also shown. of NASA-0 at 1300 K under engineering strain rates varying
between 10-6and 2.1X 10.4s-l.
constanthasbeenusedintheconstruction of theseplots.Thedata for
theDS NiA! eutecticalloys wereobcained fromseveralbatchesofthe growth speed and thermal gradient during directional solidification
firstgeneration NASA-O, NASA-2.1 andNASA-29.1 alloys whereas would result in significantly creep-resistant alloys.
thedatafor NiAI polycrysta|, advanced superalloys andNiAI single
crystal alloys were obtained from theliterature [26]. On anabsolute Summary. and Conclusions
scale,thelimited dataforthesefirstgeneration alloys revealthatthey
are weaker than thecommercial superalloys andtheadvanced NiAI The results of asystematic research program undertaken todevelop
singlecrystalalloys(Fig. I I(a)). However, theirdensity-compensated directionally solidified NiAI eutectic alloys based on the Ni-33AI-
tensile creep strengths are comparable to those of polycrystalline 3ICr-3Mo (NASA-0) composition arediscussed. Initial attempts to
Ren_80attheupperendofthescatterplot whenLMP >29x 103(K). develop alloys using astatistical design of experiments approach on
Since macroalloying and microalloying [15] have not proven very arc-melted compositions provedtobeunsuccessful.The micmstruc-
effective inincreasing the elevated temperature strengths of these tures of two new alloys, Ni-33AI-29Cr-2Ta (NASA-2.1) and
alloys, theonly viable option for improving their creep properties Ni-32.fAI-29Cr-3Mo-3Ti (NASA-29. I),, resulted in the formation
appears to be through, microstructural refinement. Since strength of coarse cellular microstructures. The creep properties of these
generally scalesas I/_L, it remains to bedemonstrated whether a alloys were not much better thanthe basealloy. Theseresults taken
reduction inthe intcrlamellar spacingbyacombination ofincreased together withthe microstructural observations from the DOE study
S.V. Raj 9 of 10
450 I l I I I 5. H.E. Cline and J.L. Walter, Metall. Trans. I, 2907(1970).
400 _ _AFN-20 Ni,_SX Alloy __ 6. H.E. Cline, J.L. Walter, E. Lifshin and R.R. Russell. MetalL
Reno'N4_ // "I-AFN-12NIAISX ONASA*0 Trans. 2, 189 (1971).
350
_.,:8o_\..' .'\ ,en.',6-, ,,,ASA-Zl- 7. DR. Johnson, X.F. Chen, B.F. Oliver, R.D. Noebe and
\ ! .\ / • I 1:2NASA-29.1
300 J.D. Whittenberger, Intennetallics 3, 99 (1995).
- ",./ - 8. T.M. Pollock and D. Kolluru, Micromechanics of Advanced
250
Materials: A Symposium in Honor of Professor James" Li's 70lh
:s
b 200 --. :._ ...._,_,_ __" _=loy_-i -- Birthday, edited by S.N.G. Chu, P.K. Liaw, R.J. Arsenault,
K. Sadananda, K.S. Chan, W.W. Gerberich, C.C. Chau and
150
T.M. Kung, The Minerals, Metals& Materials Society, Warrendale,
100 PA, pp. 205-12 (1995).
9. J.M. Yang, S.M. Jeng, K. Bain andR.A. Amato, Acta Mater. 45,
50
I I I 295 (i997).
0 10. S.V. Raj, I.E. Lecci and J.D. Whittenberger, Creep Behavior of
60 Advanced Materials for the 21st Century, edited by R.S. Mishra,
A.K. Mukherjee and K. Linga Muny, The Minerals, Metals &
IRen-N"4"7_- Rene"N6 Alloy I
Materials Society, Warrendale, PA, pp. 295-310 (1999).
11. I.E. Locci, S.V. Raj, J.D. Whittenberger, J.A. Salem and
, =NAS^- 1.l D.J. Keller, High-Temperature Ordered lntermetallic Alloys--VIII,
VoL 552, edited by E.P. George, M. Yamaguchi and M. ]. Mills,
E ...... _ r-AFN-12NiAISX Materials Research Society, Pittsburgh, PA, pp. KKS.I.1 (1999).
12. J.D. Whittenberger, S.V. Raj, I.E. Locci and J.A. Salem, Inter-
metallics 7, 1159 (I999).
13. S.V. Raj and i.E. Locei, Intermetallics, 9,217 (2001),
14. J.D. Whittenberger, S.V. Raj and i.E. Locci, High-Temperature
lO I---___" /-4 OrderedlntermetallicAlloys--lX, Vol. 646, edited byJ.H. Schneibel,
o R.D. Noebe, S. Hanada, K.J. Hemker and G. Sauthoff, Materials
i ,, ::._(:-__ .... •
Research Society, Pittsburgh, PA, N 6.9 (2001).
25 26 27 28 29 30 31 15. J.D. Whittenberger, S.V. Raj, I.E. Locci and J.A, Salem, this
volume.
T (log tf + 20) (×10 -3) (K)
16. K.A. Jackson and J.D. Hunt, Trans AIME 236, !129 (1966).
Figure 1I.--Comparison of the (a)Larson-Miller and(b) 17. S.V. Raj, 1.E. Locci, P. Dickerson, R.D. Noebe, M.V. Nathal and
density-compensated Larsen-Miller plots for DS NASA-0, D.J. Keller, unpublished research, Glenn Research Center, 1996.
NASA-2. ! and NASA-29. I with those for currant and
18. S.V. Raj, I.E. Locei, J.A. Salem and R. Pawlik, to be published.
potential airfoil materials [26].
19. J.D. Whittenberger, S.V. Raj, I.E. Locci and J.A. Salem, to be
published.
20. J.R. Stephens and W. Klopp, J. Less Com. Metals 27, 87(I972).
suggest that macroaUoying of NASA-0 is unlikely to improve its
21. J.D. Whittenberger and R.D. Noeb¢, unpublished research,
creep properties. Studies on improving the elevated temperature
Glenn Research Center, Cleveland, Ohio (1990).
strength of NASA-0 by microstructural refinement due to high
22. S.V. Raj and S.C. Farmer, Metall. Trans. 26A, 343 (1995).
growth rates proved to be somewhat more successful without any
23. M.G. Hebsur, J.D. Whittenberger and A. Garg, Structural Inter-
deterioration in the room temperature fracture toughness properties.
metallics--1997(ISSI-2), edited by M.V. Nathal, R.Darolia, C.T. Liu,
These observations clearly demonstrate that itis possible toimprove
P.L Martin, D.B. Miracle, R. Wagner and M. Yamaguchi, The
both the elevated temperature creep properties androom temperature
Minerals, Metals &Materials Society, Warrendale, PA, pp. 62i--630
fracture toughness of this alloy simply by increasing the growth
(1997).
velocity during directional solidification. Thus, DS canbeacommer-
24. M.V. Nathal and L.J. Ebert, Metall. Trans. 16A, 427 (1985).
cially viable manufacturing technique for making airfoils from NiAI
25. F.R. Larsen and J. Miller, Trans. ASME 74, 765 (1952).
eutectic alloys.
26. R. Darolia, W.S. Walston and M.V. Nathal, Superalloys 1996,
References edited by R.D. Kissinger, D.J. Deye, D.L. Anton, A.D. Cetel,
M.V. Nathal, T.M. Pollock an D.A. Woodford, The Minerals, Metals
& Materials Society, Warrendale, PA, pp. 561-570 (1996),
I. R.D. Noebe and W.S. Walston, Structural Inrermetallics 1997
(ISSI-2), edited by M.V. Nathal, R. Darolia, C.T. Liu, P.L Martin,
D.B. Miracle, R. Wagner and M. Yamaguchi, The Minerals, Metals
& Materials Society, Warrendale, PA, pp. 573-84 (1997).
2. W.S. Walston and R. Darolia, Structural Intermetallics 1997
(1551-2), edited by M.V. Nathal, R. Darolia, C.T. Liu, P.L. Martin,
D.B. Miracle, R. Wagner and M. Yamaguchi, The Minerals, Metals
& Materials Society, Warrendale, PA, pp. 573-84 (1997).
3. J.M. Yang, JOM 1997;49: 40.
4. J.L Walter and H.E. Cline, Metull. Trans. 1, 1221 (1970).
S.V. Raj ' 10of 10