Note: Descriptions are shown in the official language in which they were submitted.
2 ~
RD-19.4~h
~IG~-NIQBI ~ TIT~ ~ AL ~ INIDE A~LOYS
CROSS~REFERENCE TO RELATED APPLICATIONS
The subject application relates to copending
applications as follows: Serial Numbers 138,407; 138,408; ~-~
138,476; 138,481; 138,485; 138,48~; filed December 28, 1987
respectively.
The texts of these related applications are
incorporated herein by reference. `~
BACKGROUND OF THE INVENTION
The present invention relates generally to alloys
of titanium and aluminum. More particularly, it relates to
alloys of titanium and aluminum which have been modified both
with respect to stoichiometric ratio and with respect to
niobium addition and which contain a higher concentration of
niobium additive.
It is known that as aluminum is added to titanium
metal in greater an greater proportions the crystal form of
20 the resultant titanium aluminum composition changes. Small --
percentages of aluminum go into solid solution in ti~anium
and the crystal fsrm remains ~hat of alpha titanium. At
higher concentrations of aluminum (including about 25 to 35
atomic %) an intermetallic compound Ti3Al is formed. The
Ti3Al has an ordered hexagonal crystal form called alpha-2.
At still higher concentrations of aluminum (including the
range of 50 to 60 atomic % aluminum) another intermetallic
compound, TiAl, is formed having an ordered tetragonal
crystal form called gamma. ~ ~
:
~ !
:
.
:,
-:
- . : ~, - -
- 2 - ~ ~3 ~
RD-19 426
The alloy of titanium and aluminum having a gamma
crystal form and a stoichiometric ratio of approximately one
is an intermetallic compound having a high modulus, a low
density, a high thermal conductivity~ good oxidation
resistance, and good creep resistance. The rela~ionship
between the modulus and temperature for gamma TiAl compounds
to other alloys of titanium and in relation to nickel base
superalloys is shown in Figure 1. As is evident from the
figure the gamma TiA1 has the best modulus of any of the
titanium alloys. Not only is the gamma TiAl modulus higher
at temperature but the rate of decrease of the modulus with
temperature increase is lower for gamma TiAl than for the
other titanium alloys. Moreover, the gamma TiAl retains a
useful modulus at temperatures above those at which the other
titanium alloys become useless. Alloys which are based on
the gamma TiAl intermetallic compound are attractive
lightweight materials for use where high modulus is required
at high temperatures and where good environmental protection
is also required.
One of the characteristics of gamma TiAl which
limits its actual application to such uses is a brittleness
which is found to occur at room temperature. Also, the
strength of the intermetallic compound at room temperature
needs improvement before the gamma TiAl intermetallic
compound can be exploited in structural component
applications. Improvements of the gamma TiAl intermetallic
compound to enhance ductility and/or strength at room
temperature are very highly desirable in order to permit~use
of the compositions at the higher temperatures for which they
are suitable.
With potential benefits of use at light weight and
at high temperatures, what is most desired in the gamma TiAl
compositions which are to be used is a combination of
strength and ductility at room temperature. A minimum
-
: ;:
:
:
.
:, : :
- 3 - ~J t~ J
RD~13 426
ductility of the order of one percent is acceptable for some
applications of the metal composition but higher ductilitles
are much more desirable. A minimum strength for a
composition to be useful is about 50 ksi or about 350 MPa.
However, materials having this level of strength are of
marginal utility and higher strengths are often preferred for
some applications.
The stoichiometric ratio of TiAl compounds can vary
over a range without altering the crystal structure. The
aluminum content can vary from about 50 to about 60 atom
percent. The properties of TiAl compositions are subject to
very significant changes as a result of relatively ~mall
changes of one percent or more in the stoichiometric ratio of
the titanium and aluminum ingredien~s. Also, the proper~les
are similarly affected by the addition of relatively similar
small amounts of ternary elements.
PRIOR ART
There is extensive li~erature cn the compositions
of titanium aluminum including the Ti3Al intermetallic
compound, the TiAl intermetallic compounds and the TiAl3
intermetallic compound. A patent, U S. 4,294,615, entitled
"TITANIUM ALLOYS OF THE TiAl TYPE'I contains an extensive
discussion of the titanium aluminide type alloys including
the TiAl intermetallic compound. As is pointed out in the
patent in column 1, starting at line 50, in discussing TiAl's
advantages and disadvantages relative to Ti3Al:
"It should be evident that the TiAl gamma
alloy system has the potential for being
lighter inasmuch as it contains more
aluminum. Laboratory work in the 1950's
in~icated that ~itanium aluminide alloys
had the potential for hi~h temperature
use to about 1000C. But subsequent
;
- . ,:
4 ~ v~ 7 ?,
RD-1~,426
engineering experience ~ith such alloy~
was that, while they had the requisite
high temperature strength, they had
little or no ductility at room and
moderate temperatures, i.e., from 20 to
550C. Materials which are too brittle
cannot be readily fabricated, nor can
they withstand infrequent but inevitable
minor service damage without cracking and
subsequent failure. They are not useful
engineering materials to replace other
base allQys."
It is known that the alloy system TiAl is
substantially different from Ti3Al tas well as from solid
solution alloys of Ti) although both TiAl and Ti3Al are
basically ordered titanium aluminum intermetallic compounds.
As the '615 patent points out at the bottom of column 1:
"Those well skilled recognize that there
is a substantial difference between the
two ordered phases. Alloying and
- transformational behavior of Ti3AL
resemble those of titanium, as the
hexagonal crystal s~ructures are very
similar. However, the compound TiAl has
a tetragonal arrangemen~ of atoms and
thus rather different alloying
characteristics. Such a distinction is
often not recognized in ~he earlier
literature."
The '615 patent does describe the alloying of TiAl
with vanadium and carbon to achieve some property5 improvements in the resulting alloy.
It should be pointed out, however, with re~ard to
the '615 patent that there are many alloys listed in the
Table 2 of this patent reference but the fact that a
composition is listed should not be taken as an indlcation
that any alloy which is listed is a good alloy. Most of the
alloys which are listed have no indica~ion of any properties~
For example, alloy lT2A-ll9 of Table II is listed as Ti-45Al-
:
: . ~
- 5 - ~ J,,~ ~?,~
RD-19 426
l.OHf in atomic %. This alloy corresponds to alloy 32 of
applicant's Table II. The composition listed by the
applicant in Table II is Ti54A145Xf1 so that it is precisely
the same composition in atomic % as that lis~ed and referred
in Table II o~ the '615 re~erence~ However, as is evident
from the applicant's Table II, the titanium base alloy
containing 45 aluminum and 1.0 hafnium is a very poor alloy
ha~ing very poor ductility and, accordingly, having no
valuable properties ~nd no use as a titanium base alloy. ~he
alloy Ti-45Al-5.0Nb is listed in Table 2 in the same fashion,
i.e., without any listing of properties or indication that
the alloy has any use or any value.
A number of technical publications dealing with the
titanium aluminum compounds as well as with the
characteristics of these compounds are as follows:
1. E.S. Bumps, H.D. Kessler, and M. Hansen, "Titanlum-
Aluminum System", _, TRA~SACTIONS AIME, Vol.
194 (June 1952).pp. 609-614.
2. H.R. Ogden, D.J. Maykuth, W.L. Finlay, and R.I.
Jaffee, "Mechanical Properties o~ Hi~h Puri~y Ti-Al Alloys",
~9~c~LJIL_Y~15, TRANSACTIONS AIME, Vol. 197.tFebruary
1953) pp. 267-272.
Three additional papers contain limited information
about the mechanical behavior of TiAl base alloys modified by
niobium. These three papers are as follows:
3. Joseph B. McAndrew, and H.D. Kessler, "Ti-36 Pct Al
as a Base for High Temperature Alloys", ~G ~,
TRANSACTIONS AIME, Vol. 206 (October 1956) pp. 1348-1353.
" ~ ' '
?J'~
-- 6
RD~ 25
4. S.M.L. Sastry, and H.A. Lipsitt, "Plas~ic
Deformation of Ti~l and Ti3Al", li~QuLl~m - ~Q (Published by
American Society for Metals, Warrendale, Pennsylvania), Vol.
2 (1980) page 1231.
5. S.M.L. Sastry, and H.A. Lipsitt, "Fatigue
Deformation of TiAl Base Alloys", ~$aLl ~ U_~L _L l- L19~3
Vol. 8A (February 1977) pages 299~308.
The first paper above contains a statement that ~'A
Ti-35 pct Al~5 pct Cb specimen had a room temperature
ultimate tensile strength of 62,360 psi, and a Ti-35 pct Al-7
pct Cb specimen failed in the threads at 75,800 psi." The
two above alloys referred to in the quoted passage are given
in weight percent and have approximate compositions in a~omlc
percentages respectively of Ti4gAlsoNb~ and Ti47AlsoNb3. It is
well-known that the failure of a test specimen in the threads
is a strong indication that the specimen was brittle. It is
further mentioned in this paper that the niobium containing
composition is good for oxidation and creep resistance.
The second paper contains a conclusion re~arding
the influence of niobium additions on TiAl but offers no
specific data in support o~ this conclusion.~ The conclusion
is that: 'IThe major influence of niobium.additions to TiAl
is a lowering o the temperature at whiCh twlnning becomes an
important mode of deformation and~thus a lowering of the ~
ductile-brittle transition temperature of~TiAl." There ls no
indication in this article as to whether the ductile-brittle :
transition temperature of TiAl was lowered to below room
30 temperature. The only niobium containing titanium aluminum :~
alloy mentioned without any reference to properties or oth:er - :
descriptive data is given in~weight percent and is~Ti-36Al-
4Nb. This corresponds in atomic percent to Tig7 sAls1Nbl.s, a
compos1tion which 1s quite dlstlnct from tho~se taught and
: ~
,:
::
:
, : : .
. ~ ' , ' . '
- .' : .. . ~ :: ~ ~
- - - . :
~ 7 ~
R~-19,~6
claimed by the Applicant herein as will become more clearly
evident below.
The composition described in the fifth reference
above, which contains 36.2 weight % of aluminum and 4.65
weight % of niobium in a titanium base composition, when
converted to atomic composi~ion is Ti-51Al-2Nb. This
composition was s~udied as is reported at the l~st sentence
o~ page 301 and the first portion of page 302. As reported
on the bottom of page 301 and on top of page 302, the authors
concluded that:
"It has been found that the addition of
Nb to the TiAl base composition improves
the low temperature ductility of the base
compQsition. .... The addition of Nb does
not significantly alter the fatigue
properties of the bàse composition as can
be seen in Figure 5."
Figure 5 is quite persuasive that there is no significant
alteration of the fatigue properties. There is no~indication
in the article that room temperature ductility is improved by
Nb additions.
BRIEF DESCRIPTION~OF THE INVENTION~
One object of the present invention is to provide a
method of forming a titanium alumlnum~intermetalllc compound
having improved ductility and related properties at room
temperature.
Another object is to improve the properties of
titanium aluminum intermetallic compounds at low and
intermediate temperature~. ~
Another object is to provide an alloy of titanium
and aluminum having improved properties and processability at
low and intermediate temperatures. ;
~.
:.
` ' `
.
~3.~ 9,q26
Other objects will be in part apparent and in part
pointed out in the description which follows.
In one of its broader aspects, the objects of the
present invention are achieved by providing a
nonstoichiometric TiAl base alloy, and adding a relatively
higher concentration of niobium to the nonstoichiometric
composition. The addition is followed by ingot processing of
the niobium-containing nonstoichiometric TiAl intermetallic
compound. Addition of niobium in the order of approximately
6 to 14 parts in 100 is contemplated and additions in the
order of 8 to 12 parts is preferred.
BRIEF DESCRIPTION OF THE DRAWINGS
FIGURE 1 is a graph illustrating the relationship
between modulus and temperature for an assortment of alloys.
FIGURE 2 is a graph illustrating the relationship
between load in pounds and crosshead displacement in mils for
TiAl compositions of different stoichiometry tested in 4-
point bending.
FIGU2E 3 is a bar graph illustrating alloyproperties on a comparative basis.
FIGURE 4 is a graph in which weight gain in mq/cm~
is plotted against dynamic exposure time in hours.
DETAILED DESCRIPTION OF THE INVENTION
It is well known, as is discussed above, that
except for its brittleness and processing difficulties the
intermetallic compound gamma TiAl would have many uses in
industry because of its light~wei~ht, high strength at high
temperatures, and relatively low cost. The composition would
have many industrial ~ses ~oday if ie we e not for this basic
:
' ; '
'
-
... . .
.. . .
- . : .: : . :
~ .
: - ~' ' : ' ~ :;~ ,
9 _ ~J 'V ~ f,
RD~ 426
property defect of the material which has kept it from such
uses for many years.
The present inventor found that the gamma TiAl
compound could be substantially ductilized by the addition of
a small amount of niobium. This finding is the s~lbject of
copending application Serial No. 332,088, filed April 3,
1989.
Further, the present inventor found that a chromium
ductilized composition could be remarkably improved in its
oxidation resistance with no loss of ductility ox strength by
the addition of niobium in addition to the chromium. This
later finding is the subject of copending application Serial
No. 201,984, filed June 3, 1988.
The inventor has now found that substantlal further
improvements in ductility can be made by additions of higher
concentrations of niobium alone in the range of 8 to 13
atomic percent where this addition is coupled with ingot
processing as discussed more fully below.
To better understand the improvements in the
properties of TiAl, a number of examples are presented and
discussed here before the examples which deal with the novel
compositions and processing practices of this invention.
EXa~2LE~ 1-3:
Three individual melts were prepared to contain
titanium and aluminum in various stoichiometric ratios
approximating that of TiAl. The compositions, annèaling
temperatures an~ test results of tests made on the
compositions are set ~crth in Table I.
For each example, the alloy was first~made in~o an
ingot by elec~ro arc melting. The ingot was processed into
ribbon by melt spinning in a partial pressure of~argon. In
both stages of the melting, a water-cooled copper hearth was
used as the container for the melt in order to avoid
'' ~ .
:
- 10 - h 'v 2
R~-19~?~
undesirable melt-container reactions. Also, care was used to
avoid exposure of the hot metal to oxygen because of the
strong affinity of titanium for oxygen.
The rapidly solidified ribbon was packed into a
steel can which was evacuated and then sealed. The can was
then hot isostatically pressed (HIPed) at 950C (1740F) for
3 hours under a pressure of 30 ksi. The HIPing can was
machined off the consolidated ribbon plug. The HIPed sarnple
was a plug about one inch in diameter and three inches long.
The plug was placed axially into a center opening
of a billet and sealed therein. The billet was heated to
975C (1787F) and was extruded through a die to give a
reduction ratio of about 7 to l. The extruded plug was
removed from the billet and was heat treated.
The extruded samples were then annealed at
temperatures as indicated in Table I for two hours. The
annealing was followed by aging at 1000C for two hours.
Specimens were machined to the dimension of 1.5 x 3 x ~5.4 mm
(0.060 x 0.120 x 1.0 in.) for four point bending tests at
room temperature. The bending tests were carried out in a 9-
point bending fixture having an inner span of 10 mm (0.4 in.)
and an outer span of 20 mm (0.8 in.). The load-crosshead
displacement curves were recorded. Based on the curves
developed, the following properties are defined:
(1) Yield s~rength is the flow stress at a cross
head displacement of one thousandth of an inch. This
amount of cross head displacement is taken as the first
evidence of plastic deformation and the transition from
elastic deformation to plastic deformation. The
measurement of yield and/or fracture strength by
conventional compression or tension methods tends to
give results which are lower than the results obtained
by four point bending as carried out in making the
. ~
:
: .
- . - , ~. . . . , :
- - . . :~. . . . . . :
-. , ~. , ~ . : '
- : :
.. . . .
- 11 ?J~32c~ fJ
~D-19 426
measurements reported herein. The higher levels of the
results from four point bending measurements should be
kept in mind when comparing these values to values
obtained by the conventional compression or tension
methods. However, the comparison of measurements'
results in many of the examples herein is between four
point bending tests, and for all samples measured by
this technique, such comparisons are quite valid in
establishing the differences in strength properties
resulting from differences in composition or in
processing of the compositions.
(2~ Fracture strength is the stress to fracture.
(3) Outer fiber strain is the quantity of 3.71hd,
where "h" is the specimen thickness in inches, and "d"
is the cross head displacement of fracture in inches.
Metallur~ically, the value calculated represents the
amount of plastic deformation experienced at the outer
surface of the bending specimen at the time of fracture.
The results are listed in the following Table I.
Table I contains data~on the propertieS of samples annealed
at 1300C and ~urther data on these samples in particular is
given in Figure 2.
~; .
. '
` ` ' ' '
.
.
,
- 12 - ~ ~ 2 ~ 9 ~ ~
9,426
Outer
Gamma Yield Fracture Fiber
Ex. Alloy Composit. Anneal Strength Strength Strain
No. No. Sat.%) Temp(C)(ksi) (ksi) (%)
. -- _
1 83 Tis4Al46 1250 131 132 0.1
1300 111 120 0.1
1350 * 58 0
2 12 Tis2A148 1250 130 180 1.1
1300 98 128 0.9
1350 88 122 0.9
1400 70 85 0.2
3 85 Ti50Also 1250 83 92 0.3
1300 93 97 0.3
1350 78 88 0.4
* - No measurable value was found because the sample
lacked sufficient ductility to obtain a measure-
ment
- It is evident from the data of this table that
alloy 12 for Example 2 exhibited the best combination of
properties. This confirms that the properties of Ti-Al
compositions are very sensitive to the Tl/Al atomic ratios
and to the heat treatment applied. Alloy 12 was selected as
the base alloy for further property improvements ba~ed on
further experiments which~were per~ormed as described below.
It is also evident that the anneal at~temperatures
between 1250C and 1350C results in~the test specimens
having desirable levels of yield strength, ~fracture strength
and outer fiber strain. However, the anneal at 1400~C
results~ in a te~t specimen having a significantly lower yield
strength ~about 20% lowex); lower fracture strength ~about
30% lower) and lower ductility ~about 78% lower) than a test
40 speclmen annealed at 1350C. The sharp~decline ln properties ~`
is due to a dramatlc change ln~microstructure due, iD~ turn,
~: :
::
`:
- '
,~ ,:
- :. .: . . , . . ~ . '
:.:: `.,
:
- 13
R~=19,~26
to an extensive beta transformation at temperatures
appreciably above 1350C.
EX~E~ L=1~:
Ten additional individual melts were prepared to
contain titanium and aluminum in designated atomic ratios as
well as additlves in relatively small atomic percents.
Each of the samples was prepared as described above
with reference to Examples 1-3.
The compositions, annealing temperatures, and test
results of tests made on the compositions are set forth in
Table II in comparison to alloy 12 as the base alloy for this
comparison.
. ~
.,:
,
' ~
., .
- . , ~ . . ., . ~ , -
,~,J ij ~;J1
-- 1'1 --
~ 9,426
X
Outer
S Gamma YieldFracture Fiber
Ex. Alloy Composition Rnneal Strength Strength Strain
No. No. (at.%) Temp( C)(ksi~ (ksi) (%)
10 2 12 Tis2Al4s1250 130 180 1.1
1300 9~ 128 0.9
1350 88 122 0.9
4 22 TisoAl47Ni3 1200 * 131 0
S 24 Tis2A146Ag2 1200 * 114 0
1300 g2 117 0.5
6 25 TisoAl48Cu2 1250 * 83 0
13~0 80 107 0.8
. 1350 70 10~ 0.9
7 32 Tis4Al4sHfl 1250 130 136 0.1
1300 72 77 0.2
8 41 Tis2A144Pt4 1250 132 150 0.3
9 45 TislAl47C2 1300 136 149 0.1
30 10 57 TisoAl4gFe2 ~ 1250 * 8:9 0
1300 * 8I 0
1350 86 111 0.5
11 82 TisoAlq8Mo2 1250 128 140 0.2
1300 : 110 136 0.5
13S0 80 g5 0.1
12 39 TisoAl46Mo4 1200 * 143 0
~ l~S0 135 154 0.3
1300 131 14g 0.2 :
13 20 T$4g~sAl4s.sErl +
* - See~asterisk no~e to Table I ::
+ - Material fractured during machining to prepare
test specimens
: ~:
,
- 15 ~ J ~ ~ ~?j ~ ~,
, RD-19,426
For Examples 4 and 5, heat treated at 1200C, the
yield strength was unmeasurable as the ductility was fou~d to
be essentially nil. For the specimen of Example 5 which was
annealed at 1300C, the ductility increased, but it was still
undesirably low.
For Example 6, the same was true for the test
specimen annealed at 1250C. For the specimens of Example 6
which were annealed at 1300 and 1350C the ductility was
significant but the yield strength was low.
None of the test specimens of the other Examples
were found to have any significan~ level of ductility.
It is evident from the results listed in Table II
that the sets of parameters involved in preparing
compositions for testing are quite complex and interrelated.
One parameter is the atomic ratio of the ~itanium relative to
that of aluminum. From the data plotted in Figure 4, it is
evident that the stoichiometric ratio or nonstoichiometric
ratio has a strong influence on the test properties which
formed for different compositions.
Another set of parameters is the additive chosen to
be included into the basic TiAl composition. A first
parameter of this set concerns whether a particular additive
acts as a substituent for titanium or or aluminum. A
specific metal may act in éither fashion~and there is no
simple rule by which it can be determined which role an
additive will play. The significance of this parameter is
evident if we consider addition of some~atomic percentage of
additive X. ~ -
If X acts as a titanium substituent, then a
composition ~i4gA14gX4 will give an effective aluminum
concentration of 48 atomic percent and an~effective titanium
concentration of 52 atomi~c percent.
`:
~:
; - ~ .. :
~ lJ ~ ~ / J ~
- 16 -
If, by contrast, the X additive acts as an aluminum
substituent, then the resultant composition will have an
effective aluminum concentration of 52 percent and an
effective titanium concentration o~ 48 atomic percent.
Accordingly, the nature of ~he substitution which
takes place is very important but is also highly
unpredictable.
Another parameter of this set is the concentration
of the additive.
Still another parameter evident from Table II is
the annealing temperature. The annealing temperature which
produces the best strength properties for one additive can be
seen to be different for a different additive. Thls can be
seen by comparing the results set forth in Example 6 with
those set forth in Example 7.
In addition, there may be a combined concentration
and annealing effect for the additive so that optirnum
property enhancement, if any enhancement is found, can occur
at a certain combination of additive concentration and
annealing temperature so that higher and lower concentratlons
and/or annealing temperatures are less effective in providing -
a desired property improvement.
The content of Table II makes clear that the
results obtainable from addition of a ternary element to a
nonstoichiometric TiAl composition are highly unpredictable
and that most test results are unsuccessful with respect to
ductility or strength or to both.
EXAMPLE~ 14-?4:
Eleven additional samples were prepared as
described above with reference to Examples 1-3 to contain
titanium aluminide having compositions respectively as listed
in Table III.
- . . ', ' , ,"'. ' '
. . .
17
RD-19,426
In addition to listing the test compositions, the
Table III summarizes the bend test results on all of the
alloys both standard and modified under the various heat
treatment conditions deemed relevant.
: : :
" ... .. .
. . : i : : .
2 ~ 2 ~ ~
RD-19,426
TAB~
Four-Point Bend Propertie~ of Nb-Modified TiAl Alloys
Outer
GammaYield Fracture Fiber
Ex. Alloy Composit. Anneal Strength Strength Strain
No. No. (at.%) Temp(C) tksi) (ksi) (~)
~ _ , .
212 Ti52A148 1250 130 180 1.1
130~ 98 128 0.9
1350 88 122 0.9
1400 70 85 0.2
15 1478 Ti5oAl48Nb2 1250 139 143 0.1
1300 111 134 0.4
1350 57 67 0.1
15119 Ti51A145Nb4 1250 150 178 0.4
1300 *-- 69 0
1640 Ti5oAl46Nb4 1250 136 167 0.5
1300 124 176 1.0
1350 86 100 0.1
1766 Ti4gA147Nb4 1250 138 160 0.4
1300 126 167 0.8
1350 *-- 64 0
30 1855 Ti4gA148Nb4 1300 126 147 0.4
1350 10~ 135 0.6
1992 Ti46Al48Nb6 1350 *-- 8B 0
35 2052 Ti4gA144Nb8 1250 125 172 0.4
1300 *-- 131 0
1350 : *-- 125 0
2167 Ti44A148Nb8 1250 151 161 0.2
1300 140. 161 0.2
1350 119 153 0.7
2253 Ti46Al42Nbl2 1250 *-- 152
1300 *--~ 138: 0
1350 *-- :181 0
23123 Ti40Al48Nbl2 1300 *-- 67 0
~350 107 : 138 0.8
50 24137 Ti36A148Nbl6 **~~
:
, ~
2~
- 19
RD-19 42
* - N~ measurable value was found because the sample
lacked sufficient duc~ility to obtain a measurement
S **- The material was too brittle to be machined into
samples for test
From Table III, it is evident tha~ alloys 12, 78,
55, 92, 67, 123, and 137 contained 0, 2, 4, 6, 8 j 12, and 16
atomic percent of niobium respectively as an additive to the
base composition Ti52Al48. From the data listed in Table III,
it can be concluded that the rapid solidification of the
listed compositions does not improve room temperature
lS ductility.
If the results are compared based on the same heat
treatment (1300C) being applied to each sample, then it may
be concluded from the data of Table III, for the yield
strength which could be measured, tha~ the progressive
addition of greater concentrations of niobium xesults in a
progressive increase in the yield s~rength but also resulted
in a progressive decrease in the ductility. This finding is
consistent with the teaching o~ McAndrew in his article 3
above, but contradicts the Sastry teaching in his above
articles 4 and 5.
From Table III it is also~evident that at~the 8 and
12 atomic percent additive level ~see alloys 67 and 123) a
better combination of strength and dycti~lity can be obtained ;~
if the specimens are heat treated at the 1350C Ievel but
ductility is still below 1~.
For samples having lower concentrations~of niobium, : :
such as samples 78:and 55, it was:found that imparting
improvements to the samples by:such heat treatment~is not
feasibIe as the improvement achieved are not as:s~ignificant.
A finding results ~rom comparing the test resul~s
for alloys 55, 66, 40, and 119 in Table III. This comparison
, ~ ~
,, .
r
-- 20 ~
RD-19,426
is made with respect to samples having a 4 atomic percent
level of niobium additive but different stoichiometric ratios
of titanium and aluminum. It has been discovered based on
the study of these compositions ~hat the aluminum
concentration can be reduced slightly to obtain significant
increases in ductility without sacrificing the attractive
strength. However, aluminum concentration canno~ be reduced
below 46% without substantial elimination of ductility. Even
where the aluminum is at 46% or above the ductility is at or
below 1%.
Considering the data of Table III it is apparent
that there is an optimum concentration of the niobiu~
additive of between 4 and 12 atomic percent if appropriate
adjustments are made in the aluminum concentration and the
annealing temperature according to ~he teaching contained in
Table III.
All of the foregoing test samples were prepared by
rapid solidification. Also, the testing of all of the test
samples listed in the foregoing ~ables was done by four-point
bending tests.
TENSILE TESTING vs. FOUR-POINT BEND TESTING:
As noted above, all of the foregoing examples were
prepared by rapid solidification processing and the testing
was done by four-point bending tests. All of the data lis~ed
in the above tables is from this source.
The results of such preparation and testing as set
forth in Examples 20 through 22 is that the material having 8
to I2 atomic percent of niobium in the titanium aluminide had
very limited ductility for the most paxt with the one
exception that the Ti44Al4gNbg which was processed at 1350
annealing temperature.
I have now discovered that compositions having
niobium additive in the relatively larger quantities of 8-12
~ ''
.
.,. . . ~ . . , . : .
:, , : ~ , : ~ :
: - ' .
~ - ,
- 21 - ~2~ 2
R~-19
or more atomic percent can be given very significant
ductility if ~he processing is carried out by conventional
ingot metallurgy techniques and by conven~ional tensile
testing techniques rather than the rapid solidification and
four-point bending tests as set forth in the Examples 20
through 24.
The principal distinguishing processing step here
is that the ingo~ metallurgy technique involved a melting of
the ingredients and solidification of the ingredients into an
ingot. The rapid solidification method by contrast involves
the formation of a ribbon by ~he melt spinning method
followed by the consolidation of ~he ribbon into a fully
dense coherent metal sample.
However, before getting to the ingot processing, a
note of caution is warranted. The caution concerns the
different measurements which are usually used in testing
ingot processed samples.
The ingot processed samples are usually tested by
conventional tensile tests employing tensile bars which are
prepared expressly for this purpose.
In order to make a fair comparison between the
properties of alloys prepared by rapld solidification and
alloys prepared by conventional ingot processing a series of
tests were conducted of the proper~ies of rapidly solidified
alloys using conventional tensile bar testing.
TENSILE BAR TESTING OF RAPIDLY SOLIDIFIED SAMPLES
For this purpose, a series of conventional pins
were prepared from the alloy samples which had been prepared
by rapid solidification, most of which are listed in Table
III above. In addition, however, a gamma TiAl alloy with
niobium doping was prepared by the rapid solidification
method described above. This alloy is identified as alloy
-
.
:
'3
- 22 -
RD-19,426
132 and it contained 6 atom percent o~ the niobium dopant. A
set of pints were prepared ~rom each o~ the test alloys
listed in Tahle IV below including a set of pins prepared
from alloy 132.
S The different pins were separately annealed at the
different temperatures listed in Table IV below. Following
the individual anneals, the pins were aged at 1000C for two
hours. After the anneal and aging, each pin was machined
into a conventional tensile bar and conventional tensile
tests were performed on the resulting bars. The results of
the tensile tests are listed in Table IV immediately below.
- :
.
f ~ Y J .~
- 23 -
RD-1~,926
TA3~ ~V
Conventional Tensile Bar Testing of
Room Temperature Tensile Properties of Gamma RSG Alloys
Weight Loss
After 48hrs
Plastic Q982C in
Ex. CFG Compo- Heat Treat Strength Strength tionStatic Air
10 No. No. ~ition Temp. C (ksi~ (k~ ) (mg/cm2)
,, ., . _ . _ . .. .
2 12Ti-48A1 1250 --* 88 0
1300 77 92 2.1
1350 68 81 1.1 31
14 78 Ti-48Al-2Nb 1300 90 103 1.7
1325 82 82 0.2 7
~0 15119 Ti-45Al-4Nb 12~5 124 124 0.2
1250 120 120 0.2
1275 --* 87 0
16 40 Ti-46Al-4Nb 1275 --* 105 0
1300 101 110 0.7 4
1325 96 96 0.2
1766 Ti-47Al-4Nb 1275 109 110 0.4
1300 100 lOI 0.3
1325 95 105 0.8
1855 Ti-48Al-4Nb 1275 102 105 ~ 0.5
1325 84 ~3 1.2
1350 81 87 0.7
35 25132 Ti-46~1-6Nb 1275 --~ 120 0
1300 125 12~ 0.4
}325 --* 71 ~ 0
i992 Ti-~8Al-6Nb 1325 96 103 ~ 0.5 5
23123 Ti-48Al-12Nb 13?5 --* 106 0
1350 92 ~ 99 1.3
1375 ~4 90 0 5
I400 --* 82 0.1
- No measurable value was found because the sample
lacked sufficient ductility:to obtain a measure- : ::
ment
In addition~, as is evident from the~data presented
in Table IV, oxidation resi~tance tests were carried out. ::
:~ ..
:
::
~ r~
~ 2~ ~
RD-1~,426
If a comparison is made between the alloys listed
in Table IV which contained different percentages of niobium
dopant and the base gamma TiAl alloy which was free of the
niobium ~alloy 12) it is evident that there is essentially no
overall improvement in ductility. There are some alloys for
which significant strength improvement is formed but in
general where the strength is significantly hiqher the
ductility is quite low. For example, for alloy 119, alloy
strength is quite high (124 ksi and 120 ksi) but the
corresponding ductility is quite low ti.e. 0.1).
There is an overall improvement in oxidation
resistance from the data shown in Table IV.
E~:
INGOT METALLURGY AND TENSILE BAR TESTING
A second lot of a number of th~ alloy compositions
which are listed in the tables above were prepared by
conventional ingot metallurgy processing rather than by the
rapid solidification processing used in the first lots ~-
prepared as described in the first lots prepared as described
in the earlier examples. Where the alloy composition of the
ingot processed alloy is the same as an alloy of an earlier
example, the same example number is repeated but the ingot
processing is evidenced by adding an "A" to the example
number. One additional alloy designatedi as alloy 26A was
also prepared by ingot processing..
The properties of the alloys so prepared were
tested and the test results are listed in Table V immediately
below.
~.~r~
-- 25 --
RD-19, a26
7!A.BL~ Y
Room Temperature Tensile Properties of Cast and Forged
Gamma TiP,l Alloys
Weight Loss
Homo- Plastic After 48hrs
Gamma Atomic geni- Yield Frac~ure Elon~a- Q 982C in
Ex. CFG Compo- zation Heat Treat Strength Streng~h ~ion Static Air
No. No. sition TempCTemp.C tk~i)(ksi) ~) (mg/cm2)
. ~
2A 12A Ti-48Al 1250 130054 73 2.6 32
1250 1325 50 71 2.3
1250 1350 57 77 2.1
16A 40A Ti-46Al-4Nb 1250 l~S0 93 96 0.8
1250 1275 89 99 1.4
1250 1300 87 100 1.6 3
18A 55A Ti-48Al-4Nb 1250 1275 70 77 1.3
1250 1300 57 73 2 2
1250 1325 59 71 2
1250 1350 57 78 2.3
1400 1300 65 79 2.2
: 1400 1325 62 77 2
1400 ~ 1350 63 82 2.2
26A 151A Ti-49Al-4Nb 1400 1300 53 60 1.4
1400 1325 50 63 2.1
1400 1350 52 65 2.1
1400 1375 52 66 1.6
3~5 21A 67A Ti-4~Al-8Nh 1400 1300 74 82 1.7 2
1400 1325 70 82 2
1400 1350 67 83 2.2
1400 1375 70 87 ~ 2.6
23A 123A Ti-48Al-12Nb 1400 132S 7~ 82 1.6
1400 13S0 72 8~ 2
1~00 1375 69 87 2.3
* - Example 2A corre3pond~ to Example 2 above in the
composition of the alloy u~ed in the example. However,
Alloy 12A of ~xamp}e 2A wa~ prepared by ingot metallurgy
rather than by the rapid solidifioation method o Alloy 12
of Example 2. The tensile and elongation properties were
tested by the ten3ile bar method rather than the four point
S0 bending tèsting u~ed for Alloy 12 o~ Example 2. The other
alloys listed in Table V were also prepared by conventional
ingot metallurgy. AIl ten~ile data in Table V wa~ obtained
by conventlonal to-Yil- b-F te~ing.
::
:
:: :: : - :: ~. ', :
: ,
.3
- 26 ~
~ 426
The ingot processing procedure, which is also
designated cast and forge processing herein, was essentially
the same for each of the alloy samples prepared and was as
S follows:
In the ingot melting procedure, the ingot is
prepared to a dimension of about 2" in diameter and about
1/2" thic~ in the approximate shape of a hockey puck.
Following the melting and solidification of ~he hockey puck
shaped ingot, the ingot was enclosed within a steel annulus
having a wall thickness of about 1/2" and having a vertical
thickness which matched identically that of the hockey puck
ingot Before being enclosed within the retaining ring, the
hockey pucked ingot was homogenized by being heated to
15 1250C-1400C for two hours. The assembly of the hockey puck
and retaining ring were heated to a temperature of about
975C. The heated sample and contàining ring were forged to
a thickness of approximately half that of the original
thickness.
After the foryed ingot was cooled, a number of pins
were machined out of the ingot for a number of dif~erent heat
treatments. The different pins were separate~ly annealed at
the different temperatures lis~ed in Table V above.
Following the individual anneals, the pins were aged at
1000C for ~wo hours. After the anneal and aging, each pin
was machined into a conventional tensile bar and conventional
tensile tests were performed on the resulting bars. The
results of the tensile tests are listed in Table V above.
As is e~ident from the table, the four samples of
alloy 67A were individually annealed at the four different
temperatures and specifically 1300, 1325, 1350, and 1375C.
The yield strength of these samples is significantly improved~
over the base alloy 12A. For example, the sample annealed at
1300C had a gain of about 37% in yield strength over the
c~
- 27 -
~2~g~
alloy 12A which was annealed a~ a same temperature. Other
gains are of the same order of magnitude. This gain in
strength was realized with a reduction in ductility but the
ductility of the sample of alloy 67A annealed at 1300C is
remarkably improved over a similar sample for Example 21 of
Table III. The other heat-treated samples show comparable
gains in strength wi~h modest reduction in ductility over the
base alloy 12A and in some cases with a modest gain in
ductility. The combination of improved strength with
moderately reduced ductility or even moderately increased
ductility when considered together make these gamma titanium
aluminide compositions unique.
Returning again to consideration of the test
results that are listed in Table V and by comparing it with
the data, for example, listed in Table IV, it is evident thht
the yield strengths determined for the rapidly solidified
alloys-as reported in Table IV are somewhat higher than those
which are determined for the ingot processed metal specimens
as reported in Table V. Also, it is evident that the plastic
elongation of the samples prepared through the ingot
metallurgy route have higher ductility than those which are
prepared by the rapid solidification route. The results
listed, however, provide a good comparative basis in having
alloy 12A which was prepared by ingot mstallurgy listed in
Table V and alloy 12 which was prepared by rapid
solidification listed in Table IV. However, from a general
comparison of the data of Table V, with the data of Table IV,
it is evident that for the higher concentration of niobium
additive, the preparation of the alloy samples by the ingot
metallurgy processing technique and the testing of the
samples by conventional tensile bar t sting techniques
demonstrates that the higher niobium alloys prepared by ingot
metallurgy techniques are very desirable for those
applications which require a hlgher ductillty. Generally
. . : ', '
. .
. .
., , , , ' ' . .
- - 28 - ~;J~
RD-19.926
speaking, it is well known that processing by ingot
metallurgy is far less expensive than processing through melt
spinning or rapid solidification inasmuch as there is no need
for the expensive melt spinning step itself nor for the
consolidation step which must follow the melt spinning when
the rapid solldification processing is employed.
OXIDATION ~ESISTANCE
The alloys of this invention also display superior
oxidation resistance. The oxidation tests reported in Table
IV are static tests. The static tests are performed by
heating the alloy sample to 98~C for 48 hours and then
cooling and weighing the heated sample. The weight gain is
divided by the surface area of the sample in square
centimeters. The result is stated in rnilligrams of weight
gain per square centimeter of surface area for each sample.
The data given in Table V is determined on the same
static basis.
A number of dynamic oxidation resistance tests were
performed on a number of the alloys as lis~ed in Table V.
The data from these tests are plotted in Figure 4. In Figure
4, the weight gain in mg/cm2 from oxidation of alloy samples
as marked is plotted against dynamic exposure to oxidation at
850C. By dynamic or cycled exposure to an oxidizing
a~mosphere at elevated temperature is meant that the test
sample is cycled through a series of heatings and coolings
and that the sample is weighed each time it has cooled to
room temperature. The heating is to 850~C in each case and
the sample is maintained at the 850C temperature during each
cycle for 50 minutes. Cooling is ~ot a forced cooling but
rather is a cooling in an ambient room temperature
atmosphere. The cooling, weighiny, and return to the furnace
for testing to the 850C temperature takes in the order of
ten minutes for an average size sarnple. The heating to
.
- ., , . , ,: -
.
.
. - . .
- 29 -
~D-19,~2
temperature and cooling from temperature is not part of the
S0-minute period during which the sample is maintained at
temperature.
The data plotted in Figure 4 is a plot of the
weight and of the changing weight of the four samples tested.
From the plot of Figure 4, it i~ evident that the alloys
having 8 and 12 atom percent niobium dopant were by far the
best compositions from the point of view o~ cyclic oxidation
resistance.
Figure 3 presents similar data but on a different
basis. In Figure 3, the oxidatlon resistance is displayed on
the basis of the time needed for the sample to reach a weight
gain level of 0.8 mg/cm2. For the Ti~4A14gNbg alloy, the time
is 500 hours.
Figure 3 also presents the relevan~ strength and
ductility data for the respective alloys.
Clearly, from the data plotted in Figures 3 and 4,
it may be seen that the ingot processed alloy Ti48-37Al46-
4gNb6-l4 is a novel and unique alloy having unusual and novel
sets of properties.
.
