Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
~ he subject invention is addressed primarily to powder
metallurgy, particularly to the production of superalloy metal
powders through the disinkegration of molten metal streams by
atomization.
As is known, over the years research efforts have been
intensified in respect of the development of superalloys capable
of withstanding the increasingly severe operating conditions,
notably higher temperatures and stress, imposed by reason of
advanced designs, intended applications, etc. This has been
particularly evident, for example, in turbine engine development.
And in response to such demands, a number of alloys, usually of a
nickel, cobalt or iron base, possessing the necessary metallurgical
properties have been developed. -
But, the achieving of such properties has given rise to
a serious attendant problem, to wit, poor hot workability and
fabrication characteristics. Some of the most promising alloys
developed, usually those of highest strengths at the more elevated
temperatures, upon melting and casting cannot, as a practical
matter, be hot worked, let alone otherwise fabricated. As a
consequence, such alloys have been normally produced and used in
the cast form. Apart from such drawbacks as "segregation" abnor
malities associated with cast structures, the cast form is
inherently self-limiting with regard to properties and product
shapes that can be produced.
In an effort to circumvent the hot working and
fabrication difficulties, the art has turned to powder
metallurgy. And incident to this, the production of the
powder per se has heen accorded particular attention. ~n
this connection and as has been suggested elsewhere, there
are nearly as many ways to produce metal powders as there are
~07~3S67
metals from which they are produced, including atomization,
chemical reduction, mechanical comminution, etc. It is the
former which is of concern here, particularly gas atamization.
While both gas and water atomization are commonly used,
the latter has probably seen greater general use ostensibly because
of its seemingly innate ability to deliver powders of high density,
its ability to provide quenched powders more readily and, not
unimportantly, it is normally more economical. sut these powders,
irregular in shape, contain appreciable quantities of oxygen, an
undesirab~e contaminant in a number of alloys.
While it would be impossible to consider the innumerable
proposals heretofore advanced in respect of gas atomization, a
cursory review of the literature would seem to suggest there are
two general approaches to directing gas, usually argon or nitrogen,
at a metal stream. One technique involves the use of a plenum
chamber having an annular opening in the bottom thereof to exhaust
gas in a downward manner concentric with the flowing metal stream.
The lower bore of the metal teeming nozzle diverges outwardly such
that the metal stream is caused to spread outwardly in a controlled
manner under the influence of a vacuum generated within the cone
of surrounding gas. In a variation of this, high velocity gas is
passed through two plenum chambers to swirl the gas and enhance
the atomization process.
In the second approach a series of separate openings
are used in the plenum chamber rather than the annular concept.
The benefit here is said to follow from directing the gas at
the circumference of the metal stream as opposed to its axis,
thus causing a swirling effect to aid atomization. Modifications
have included the use of nozzles in lieu of bored holes in the
bottom of the plenum chamber. This is said to be expensive such
that water cooled sur~aces have been used as the cooling medium.
~ 7
In contrast to the emphasis accorde~ the direction ~
gas should be caused to 1Ow to produce powder, our review o the
literature, albeit not exhaustive, rather indicates that compara-
tively scant attention has been given to the energy level that
should, as we believe, be delivered to the molten metal stream.
This aspect forms a part of the instant invention since it has
been found that finer powder particle sizes obtain, a result
probably due to the larger surface areas generated by reason of
the high energy plateaus developed. But imparting high energy to
a molten metal stream would normally suggest even higher costs
than that heretofore associated with gas atomization.
In accordance herewith, however, a most substantial cost
reduction is effected. This is largely brought about by the fact
that we have devised a process by which low pressure gas can be
used, e.g., as low as 120 psig, and a minimum mass of gas as well,
the gas being controlled through a multiplicity of venturi-type gas
jets in combination with a correlated teeming nozzle, etc., as
will be herein described. Argon costs (perhaps the most costly
item) alone are reduced by at least one-half in comparison with
conventional processes (which virtually all rely on much higher
gas pressures, e.g., 240-270 psig), since, given the present
invention, low pressure gas can be supplied with standard equip-
ment directly from a liquid phase evaporator. Yet powder yield
is exceptionally high. Normally lower gas consumption is the anti-
thesis of higher yields.
Moreover, the invention in its most advantageous
embodiment (double impact mode) provides for a narrow profile,
i.e., a narrow cone of falling powder such that there is a
more intimate gas-to-metal contact. This has several benefits.
7~S~
First, there is a marked reduction in both flake and filigree
formation. By way of explanation, in conventional processes
as the metal droplets are formed they descend in a most distinct
"diverging" pattern. A goodly portion of the droplets hits the
sides of the enclosing chamber. The droplets, being very hot, if
they have not sufficiently solidified and cooled, either de~lect
off the inner chamber wall and form flakes or stick and form
filigrees. In either case, powder recovery obviously suffers,
and both clean-up and tap-to-tap times are increased. This is
most substantially minimized by the instant invention, yields as
high as 96-98~ being obtainable.
Second, the narrow cone of falling powder results in
such gas-to-metal interface and efficient heat transfer there-
between that the temperature range of atomized powder reaching
the bottom of the enclosing chamber is greatly reduced. In the
case of superalloys, we have found that temperatures as low as
550~F. to 750F. are not uncommon. This compares with tempera-
tures well over 1500F. characteristic of other processes. Too,
it is thought that the intimate gas-to-metal contact in the narrow
cone likely helps decrease the volume of gas otherwise required
to cool the falling powder. In any event, the need for such
currently used equipment and techniques as elaborate gas
recirculation-refrigeration cycles, inert cryogenic liquids,
externally cooled hearths, etc., have been obviated.
Third, the narrow cone profile enables smaller holding
tank diameters to be used and a "multiple" of atomized streams
can be formed in a single tank without the spray of one stream
detrimentally interfering with another. Therefore, production
rates can be readily increased without the necessity of having to
extrapolate beyond available gas and metal flow rates.
-~ "- : '' ' ' ' . ' ,
,
. . . . . .. .
7~35t;~7
Fourth, apar~ ~rom excellent pow~er yields, practically
all the powder can be produced within a desired particle range
and of desired particle distribution. Of the many superalloys
experimentally produced to date, particle size has been within
the desired range of -40 +325 mesh (U.S. Series). And since argon
was used, subversive oxygen contamination was avoided.
Other advantages of the invention will become apparent
from the following description and accompanying drawings in which:
Figure 1 is simply a schematic arrangement in depicting
the principal components of the subject atomization assembly;
Figures 2(a) and 2(b) illustrate two different tundish
teeming nozzles, a smooth bore venturi and sharp-edge orifice,
respectively;
Figure 3 schematically represents in section a plenum
chamber;
Figure ~ reflects the profile of a preferred gas jet
embodiment;
Figure 5 schematically depicts one of the most advanta-
geous embodiments the plenum plus the gas jets arranged to strike
a molten metal stream.
Figure 6 depicts a relationship among argon driving
pressure, jet exit diameter and energy generated at jet exit.
Generally speaking, the present invention contemplates
the atomization of molten metal to powder whereby high total
yields of powder are obtained with low gas consumption, the
invention generally involving tapping a molten metal stream
into a teeming vessel, particularly vessels of the tundish
type, teeming the metal through a nozzle to form a molten
stream and subjecting the molten stream to the action of an
7~35~7
atomizing gas, the gas being dispensed through a plurality of
jets arranged such that the mol~en stream is impacted by gas
from at least two sets of jets (can be ~hree or more), i.e.,
a primary set of jets and at least one secondary set as described
above, with the included angle of the primary set in relation
to the molten stream advantageously being greater than the
included angle of the secondary set. This double impact mode
system (or greater) provides for high total yields of powder
with minimum argon consumption as will become more eviden-t
from the following.
The following information is given on the basis of
treating 100-lb~ melts from a furnace. Larger tanks, plenum
chamber, etc., could be used upon scale~up to larger size melts.
TUNDISH
The tundish (holding vessel) should be capable of
holding a portion of a melt at depths up to 10 inches or more, -
a preferred depth being from 6 to 10 inches depending upon
teeming rate. A 6-inch diameter vessel has been found quite
satisfactory for 100-lb. melts, though larger sizes would be
desirable for larger size melts. The tundish should preferably
be heated separate from the furnace and be capable of maintaining
the melt up to desired temperature, advantageously about 100F.
above the liquidus (approximately up to 2900F. in the case of
nickel and/or cobalt base superalloys).
It might be mentioned that the temperature at which
the melt is tapped from the melt furnace to the tundish is
important. While it must, of course, be sufficiently high to
prevent freeze-up in the tundish nozzle, it should be low
enough such that the atomized particles solidify rapidly with
the fine grains and low oxygen pick-up.
.. ..
: ,
~(~7~S~;7
TEEMING NOZ ZLE
The teeming nozzle is supported in the tundish, its
function being to meter the molten metal into the atomization
zone. While a teeming nozzle of the smooth bore venturi, Figure
2(a), is generally used, it is sometimes deemed, however, more
advantageous to use a sharp-edged orifice nozzle of the type
illustrated in Figure 2(b). And khis obtains even though this
type of nozzle might offer less resistance to turbulence in
the tundish than would the venturi profile.
The orifice type nozzle above mentioned was arrived -
at as a result of extended investigation and experimentation.
We have found that it is beneficial by reason of a low discharge
coefficient, approximately 0.65-0.75 in comparison with unity
as is the case generally with standard nozzles. This offers
a larger opening for a given flow rate. Yet, it maintains
sufficient stream stability. Therefore, alloys prone to "nozzle
blockage", e.g., those having a large solidification range, can
be teemed more successfully because of the larger opening required
for a given flow rate. It has the additional advantage as a
result of the smaller mass of nozzle to conduct heat away from
the metering restriction. This results in reduced heat losses
from the metering zone. Moreover, our investigations reflect
that the atomiæing medium tends to accelerate about the sharp
orifice edge and this lends to removing initial precipitation
of a nozzle blockage.
Beneficially the teeming or tundish nozzle is of a
ceramic such as zirconia. In minimizing nozzle blockage, a
throat diameter of about 3/16 to 11/32 inch is generally satis-
factory. For venturi or smooth bore nozzles, a throat diameter
of 1/8 to 5/16 inch is generally suitable.
--7--
:
6~
As an aside and with other atomization parameters
held constant, the smaller nozzle diameters give smaller po~7der
particles, this at the expense of slower -teeming rates and
higher gas consumption. Conversely, the larger nozzles result
in coarser particles, faster teeming rates and lower gas con-
sumption.
TEEMING RATE
The metal teeming rate from the tundish is influenced
principally by the throat diameter of the nozzle (they are
approximately proportional) and by the head of metal in the
tundish (teeming rate being virtually proportional to the square
root of the melt height in the tundish). For a given gas flow
rate, the lower teeming rates produce smaller powder particles.
To avoid excessive fines and unnecessary gas consumptionr it is
of benefit to control the rate of teeming between about 10 and
65 kg/min. and more preferably from about 18 to 40 kg/min., the
teeming nozzle throat diameter being above about 0.2 inch and
up to about 0.34 inch, particularly from about 0.23 to 0.30 inch.
PLENUM C~AMBBR
An illustrative plenum chamber is set forth in
Figure 3, again in rather schematic fashion.
While the plenum can take virtually any shape, it is
preferred that it be substantially that of a hollow torrid
(akin to a hollow annulus), to thereby permit the metal being
teemed to pass through the central hole thereof and to feed
argon to the gas jets at the bottom. The outside surface can,
of course, be modified for ease of fabrication. The diameter
of the central hole should be at least about 1-1/2 or 1-3/~
inches to permit sufficient clearance for the metal stream. On
the bottom surface of the plenum, spaces are provided to
receive into place the desired number of gas jets.
7~S~ 7
The diameter of the circle through the center o
the holes ("jet circle diameter") used to secure the jets can
be from about 2 to 6 inches or more, it being preferred that it
be from about 2-1/2 to 4 inches. A jet circle diameter of 3
to 3-1/8 inches is a good compromise, given the need to keep
the metal stream away from the gas jets and the need to extend
the gas jets close to the atomization zone to minimize energy
losses in the gas.
The chamber should withstand pressures up to at least
600 psi, and be adapted to receive gas on both sides as shown
in Figure 3. A gauge can be used outside the atomizer to record
the driving gas pressure for the gas jets via a third tube
into the plenum.
GAS JET PROFILE
The gas jets, which can be formed of any suitable
material, e.g., brass, should be of the venturi converging-
diverging type. Such jets accelerate the gas smoothly up to the
throat where it reaches, say, Mach I, and then accelerate it
along the gradually diverging bore to from, say, Mach I up to
Mach 5 at the exit. Past the exit, gas velocity decreases, but
maintains a supersonic tongue up to 3 inches or more.
The two most important dimensions of the jets are
throat diameter and length of tapered section. The finish of
the base should be as smooth as possible without abrupt changes
in cross section. The design and dimensions of preferred jet
embodiments is depicted in Figure 4. Jet No. AA differs from A
in the l/2-inch additional lengths, i.e., length of exit, exten-
sion from plenum and length of jet. The same applies to Jets B
and ss and Jets C and CC.
s~
GAS JET ASSEMBLAGE
While the invention is no-t res-tricted -to any speci~ic
number of jets, in accordance herewith it is most preferred
that eight, approximately equally spaced, jets be u-tilized.
Four of the jets, the "second set", strike the falling molten
stream below the other fourl i.e., the "first set" as shown in
Figure 5. This provides for the above-mentioned "double impact
mode" with the second set creating the narrow powder cone profile
as depicted in Figure 5.
To secure the jets, plugs can be welded inko the
plenum and allowed to protrude slightly beyond thé bottom surface.
The faces of the plugs can be machined to provide seats for the
gas jets and to ensure they are aimed correctly.
The direction in which the jets exhaust the gas is of
considerable importance. The included angle of the jets
(Figure 5) of the "first set" should not exceed 30 and most
beneficially is not more than about 25 to 27, the preferred
angle being about 24 to 26, given preferred teeming rates.
With regard to the "second set" of jets, while the included
angle could be that of the primary set, it is much preferred
that it be less than that of the "first set" and preferably
be at least 2 or 3 less, the preferred included angle being
from about 21 to 23. The two angles for alternate opposed
jets contain the powder spray in a tight or narrow cone. It
might be added that lower energy jets perform better if the - -
included angles are increased slightly to decrease the distance
over which the energy decays. Higher energy jets require the
smaller included angles~
--10--
.
,
1,~7f~51~i7
In terms of the mass flow rate of gas delivered
from the jet exits, it is most preferred that an exit velocity
of at least Mach No. 1.5 be reached, particularly a veloci-ty
greater than Mach No. 2Ø In this connection, the energy
(kinetic) available at the jet exit largely depends upon the
gas driving pressure and throat diameter. This is depicted
in Figure 6, the information being based upon theoretical con-
siderations. Thus, the same energy generated with a relatively
large throat diameter can be generated with a jet of smaller
diameter provided the driving pressure is increased. The
reduction in gas con~umed by reason using a high driving gas
pressure and smaller throat diameter is balanced by the higher
gas velocity, hence higher kinetic energy, at the jet exit.
However, there is a limit to how far the jet exit
diameter can be decreased since to maintain the mass flow rate
of gas requires that the gas be disproportionately increased.
For a given Mach number the length of supersonic cone of gas
delivered to the atomization zone decreases much in proportion
to the decrease in exit diameter. Put another way, the smaller
~0 the exit diameter, the less effective is the energy transfer
from jet to atomization site.
ATOMI_ATION TANK
The above components, tundish, plenum, nozzles, etc.,
operate within a chamber (not shown in Figure 1) which for
many alloys, including the superalloys, is maintained under
vacuum during melting. This chamber should be capable of
holding a vacuum of 10 um of Hg or less, and can be varied in
size depending upon the number of metal streams to be atomized,
--11--
5~7
given a narrow powder cone proE:Lle. For l00 lb. melts and
atomizing a single molten stream a tank 4 feet in diameter and
20 feet below the tundish, has proved satisfactory.
The bulk of the powder can be collected in a water
cooled skip car at the bottom of the tank. By holding under
argon in a quasi-fluidized state for a predetermined time, e.g.,
2 hours, oxygen pick-up is minimized. It is also to advantage
that above about 3 psig in the tank, the argon be exhausted
through a cyclone or the like to remove entrained particles.
ATOMIZATION PREPARATI0~
T avoid contamlnation from one alloy to the next, it
may be necessary to "blow down" any prior accumulated powder
in the atomization vessel or exhaust gas scrubber. Compressed
air can be used.
Raw materials should be free from refractory phases
to minimize tundish nozzle blockage.
The tundish nozzle should be arranged such that the
molten stream is teemed vertically down onto the focal points
of the jets. An offset of even l¦4-inch reduces efficiency.
Backfilling with the gaseous medium, e.g., argon, is
recommended prior to atomization. This reduces extreme pressure
difference between the plenum and ~et e~its.
The following information and data are given as
illustrative of the invention:
.
1 - 12 -
~785~
Terminology:
Recovery (%) = Total Powder Weight x 100
Melt Weight
Powder Yield (%) = Powder Weight at Size x 100
Total Powder Weight
Total Yield (%) = Powder Weight at Size x 100
Melt Weight
Argon Used (ft.3/KG) = Argon ~olume
Powder Welg t at Slze
0 Note: Powder "recovery" seldom exceeds 92-94% since the
furnace and tundish can each contain up to, for example,
0.9 lb. of a 100 lb. melt as skulls. Thus, total yield
values are influenced by the size of the skulls. There-
fore, powder yield is considered more accurate than
either powder recovery or total yield. With units larger
than the laboratory scale used herein, this problem will
be reduced. Longer runs minimize the beginning and end
effects which are present eor most of a 100 lb. melt run.
A 200 lb. scale-up confirmed this.
A number of tests were conducted in respect of various
well-known superalloys, the nominal aim compositions being
given in Table I with the processing conditions, gas pressures,
teeming rate, teeming and gas jet nozzle parameters, etc., being
varied as detailed hereinafter.
TABLE I
.
Alloy: Cr %: Co %: Ti %: Al %: Mo %. W % : Cb %: B % : Zr %: C %
1 : 10 15 : 4.7 : 5.5 : 3 : -- : -- : .01 : .06 : .03
2 : 15.3: 16.9: 3.5 : 4.0 : 5 ~ - : .03 : -- : .06
- 3 : 13 : 8.0: 2.5 : 3.5 : 3.5 : 3.5 : 3.5 : .01 : .05 : .06
4* : 12.4: 9.0: 3.9 : 3.2 : 2.0 : 3.9 : -- : .01 : .01 : .05
5**: 19 : -- : 0.3 : 0.5 : 3.1 : -- : 5 : .004 : -- : .04
6 : 48 : -- : 0.35: -- : -- -- __ __ __ __
*alloy 4 contained 3% Ta
**alloy 5 contained 52.5~ Ni, balance Fe
balance Oe alloys otherwise nickel
Generally, the tests involved tapping a 45 kg melt
of superalloy into a tundish, the tappiny temperature reflec-ting
composition and being generally from about 2700 to 2~50F.
-13-
:1~78S~o~
For purposes of comparison, Examples I and II are
included to give some idea as to what might be e~pected with
processing procedures which might be representative of some
prior art procedures.
EXAMPLE I
45 kg of Alloy 2 were vacuum melted in an atomizer,
tapped into a tundish and then teemed through a 1/4 inch venturi
t~vpe teeming nozzle at an average rate of about 23 kg/min.
Argon was exhausted from a plenum chamber at 260 psig through
a single set of four equi-spaced subsonic eliptical orifice jets
(gas velocity Mach I or less~ at an included angle of 30, (no
double impact mode). Each of the jets was rotated through 45
to impart a downward swirling motion to the gas (swirl mode).
The resulting powder was collected, yield, particle size, etc.,
being determined.
The results are reported in Table II.
EXAMPLE II
In an attempt to reduce the average particle size
of Alloy 2, the metal flow rate was restricted to give an
increase in the ratio of mass rate of gas flow to alloy teeming
rate. In this instance, a 7/32 inch (I.D.) orifice teeming
nozzIe was used, the argon pressure and orifice jet arrangement,
including swirl mode, being much ~he same as in Example I.
The results are given in Table II.
... . . .
S~;7
TABLE I I
_ __ _
Example I (Recovery 86.8%)
-40 +325 -60 +325 -80 +325
Powder Yield 88.8% 68.8~ 49.8%
Total Yield 77.0% 59.7% 43.2%
Argon consumed 47.8 ft3/kg 61.7 ft3/kg 85.3 ft3kg
Example II (Recover~ 84.4%)
Powder Yield 93.2% 81.6% 63.2%
Total Yield 78.7% 68.9% 53.3%
Argon consumed 53.08 ft3/kg 94.8 ft3/kg 122.0 ft3/kg
. . . _ _
While particle size was reduced under the conditions
of Example II, the results are not deemed very outstanding,
given the amount of argon consumed.
EXAMPLE III
To demonstrate that simply directing more energy to
the atomization zone by using supersonic venturi jets is not a
panacea, 42 kg of Alloy 1 were vacuum melted in an atomizer,
tapped at 2650F. into a preheated tundish and teemed through
a 5/16 inch diameter orifice nozzle at an average rate of about
18 kg/min. Argon was exhaused through 1.5 inch long venturi
jets having throat diameters of 7/32 inch, a Mach No. of 1.7
being reached at the exit. Using a 2-1/2 inch jet diameter
circle four jets spaced 90 apart were aimed at the metal stream
at an included angle of 25~ (vs. 30 in Examples I and II).
In this instance the total yield of powder was 61.7%
for -40 +325 mesh and 22.6% for -80 +325 mesh.
A satisfactory volume of argon was used, a higher
Mach No. was achieved and more energy was delivered to the
atomization zone. However, this single impact mode system
resulted in a wide cone of coarse droplets and much flake at
the tank walls, the result being a low total yield of powder.
.. . .
~` 3L0~78S~i'7 1
EX~MPLE IV
This test serves to illustrate the marked improvement
obtainable using 2 sets of venturi type gas jets (double -impact
mode) in combination with an orifice type teeming nozzle.
46 kg of Alloy 3 were vacuum melted, tapped at 2700F.
into a tundish preheated to 2200F. and then teemed through a
9/32 inch diameter orifice nozzle at an average rate of about
16 kg/min. Argon was exhausted at 150 psig (vs. 260 in Examples I
and II) through 1-1/2 inch long venturi jets, the jets having a
throat diameter of 5/32 inch. A Mach No. of 2.8 was de~ermined
at the exit. Using a 2-1/2 inch jet diameter circle, four jets -
spaced 90 apart were directed at the stream at an included
angle of 30, with a second set of jets being alternately spaced
at an included angle of 25.
The results are given in Table III.
EXAMPLE V
The test of Example IV was repeated with an average
teeming rate of 18 kg¦min., the argon being exhausted at 180
psig through 2 inch long venturi jets, having throat diameters
of 5/32 inch, a 3.4 Mach No. being achieved at the exit. In
this instance, the jet diameter circle was 3-1/2 inches with
the first set of jets being at an included angle of 25 (vs 30
in Example I~) and the second set at 22 (vs 25 in Example IV).
In this case, powder temperature was measured at the bottom of
the tank. A maximum temperature of approximately 600F. was
determined.
- 16 -
` ~0~5~
TABLE III
.. . . , _ _ . . . . .
Example IV (Recovery 94.8%)
-40 ~325 -60 + 325 -80 ~325
Powder Yield 91.6~ 71.8% 54.1%
Total Yield 87.0% 68.2~ 51.3%
Argon consumed 29.3 t3/kg 37.8 ft3/kg 49.8 ft3/kg
Example V (Recovery 96.0~)
Powder Yield 91.5% 74.0% 55.1
Total Yield 87.8% 71.0~ 55.7%
Argon consumed 41.1 ft3/kg 50.8 ft3/kg 64.8 ft3/kg
. . . _
A comparison of the data in Tables II and III reflect
the dramatic reduction in argon consumption with marked improve-
ment in recovery and yield.
As will become evident from additional data presented
infra, the results given in Table III are by no means the best
that can be achieved. Bu~ in simply comparing Tables II and III
it will be noted that the total yield of the finer particle
size, i.e., -30 -~325 powder, was greater in respect of Examples IV
and V and yet the argon consumed was decidedly less.
EXAMPLE VI
As indicated above herein, a correlation of ori~ice
nozzle, teeming rate, venturi jets, etc., is required to generate
high efficiency. This is reflected by virtue of a test run in
which a high metal ~eeming rate was used in conjunction with a
large nozzle, though a double mode impact system was employed
at favorable included angles to the falling molten streams.
Thus, approximately 45 kg of Alloy 4 were vacuum
melted, tapped at 2700F. into a tundish, preheated to 2200F.
and teemed through an 11/32 inch diameter orifice nozzle at an
average rate of 34 kg/min. Argon at 120 psig was exhausted
through 2 inch long venturi jets, having throat diameters of
5/32 inch. A relatively high Mach No. 3.4 was reached at the
-17-
, . . . . . .
. .: - . ~ :
-. - . ~ . . :,
,... . . :
~L~78~67
exits. With a 3-l/8 inch jet diameter circle, 4 jets were
spaced 90 apart at an inclwded angle of 25 with a second set
of jets being alternately spaced at an included angle of 22.
Total powder ~ield was but 72.~ over the -40 ~325
mesh range and ~2.7% for the -80 +325 mesh. Argon consumptions
were 28.4 ft.3/kg and 48.1 ft3/kg, respectively. It is believed
the high teeming through a large nozzle exceeded the capability
of the gas flow to cool the powder as would otherwise be the
case. As this obtained notwithstanding that the energy level
of the argon stream was high. It might be added that much of
the powder, though useable, had caked together.
In Table IV, are set forth data derived by reason of
varied operating parameters, e.g., teeming nozzle diameter,
argon pressure, jet design type, etc. These later developed
jets were made not only to accelerate the gas through the throat
and diverging sections, but also to maintain that velocity in
the confines of a tubular section (D of Figure 4) to the exit.
The teeming rate for Runs l-A, l-B, l-C 2 and 3 was about
23 kg/min., the jet-to-impact distance being about 5.2 inches.
Castellated nozzles (Runs 4-A and 4-~) were also used,
the purpose being to further minimize melt turbulence (which
was low) around the tundish nozzle and to produce smoother
teeming streams~ Such nozzles do not significantly affect any
melt turbulence, at least on a small scale basis.
A constant argon driving pressure was attempted for
the complete cycle run, the results being shown under tests 5
and 6. As will be understood by those skilled in the art,
start-up and end conditions in terms of driving pressure dif~er
from that experienced over the major part o~ the teeming cycle.
-18-
5~7
The jet-to-impact dis-tance was altered rom the 5.2
inch distance in the case of tests 7 and 8 (also tests 9-13).
In these instances the distance was 4.7 inches ~jet "AA" being
1/2 inch longer than "A"). It will be observed that generally
there was an increase in powder yield in the small mesh sizes,
i.e., -80 and -100 mesh.
In tests 9 through 13 the teeming nozzle diameter was
varied over the range of 0.29 to 0.36 inch. In all instances
a baffle was used (terminated about 1.5 inch above the nozzle),
the objective being to minimize any melt turbulence which might
arise while filling or any tendency for the melt to form a vortex
while emptying. The 0.36 inch nozzle diam~ter did result in
lower yields. However, given a relatively short steady teeming
rate for a 100 lb. melt test (approx. 15 seconds) it is thought
that this masks the true effects of changes in nozzle diameters.
As referred to above herein, a distinct commercial
advantage of the subject invention is that the narrow cone
profile permits of multiple atomization of molten streams in
apparatus which could not be so used with other processes. In
this connection the two liquid streams were spaced about 5-3/4
inches apart. One tundish was used, the tundish being fitted
with two nozzles, two plenum chambers and two sets of jets. The
venturi jets used here, E, had a converging section (120), a
parallel throat section and a diverging section (6) extending
to the exit. These differed from the others largely in that
the latter had tubular extensions to the exit. While the tests,
Nos. 14-16, Table IV, were far from being refined, nonetheless
they confirmed that multiple stream atomization could be
conducted.
--19--
.
: .. - .: ' . ' ' ~ ; '
107~S~;~
T~b]e V offers a comparison of powder yields and
argon consumption as a function oE particle size for each of
the three different ~et embodiments ~et forth in ~ig~re 4, to wit,
AA, sB~ and C. The teeming noz~le diameter was 0.27 inch, jet
circle diameter 3.125 inch~ with the same nozzle arrangement
of 4 jets at 25 and 4 alternate at 22~. It should be poin~ed
out that 20~ lb. melts of superalloy were used. As can be seen
from the data, performance was quite satisfactory.
- In the above discussion of the invention, reference
has been made to argon as the gaseous medium. However, other
gases can be used, includlng inert gases generally nitrogen,
carbon monoxide, helium can be used. Depending upon the nature
of the alloy processed, oxidizing gases, including air and oxygen
can be used~ Even water might be used as the atomizing medium.
Too, while the invention as above described contemplates the
atomization of alloys, it is equally applicable ~o the atomization
of metals per se. Also, conceptually the invention might be
- applicable to the disintegration of molten streams other than
alloys or metals.
While a number of specific superalloys have been above
referred to, and while the invention is particularly directed
to otherwise difficulty workable alloys, notably those con-
taining more than about 4 or 5% of the precipitatlon hardening
elements aluminum and titanium, or a goodly percentage of ~atrix
stiffening elements, molybdenum, niobium, tantalum, tungsten,
vanadium, etc., the invention is, o course9 applicable to
alloys in general. Among the superalloys are those containing
up to 60%, e.g., 1% to 25%, chromium; up to 30%, e.g., 5% to 25%,
cobalt; up to 10%, e.g., 1% to 9%, alumlnum; up to 8%, e.g., 1%
to 7%, titanium, and partlcularly those alloye ~ontalning ~ or 5%
or more of aluminum plus titanium; up to 30%, e.g., 1% to 8%,
-~1
- 20 -
.. . .
7BS~;7
molybdenum; up to 25~, e.g., 2~ to 20%, -tungsten; up to 10
columbium; up to 10~ tantalum, up to 7% zirconium; up to 0.5
boron; up to 5% hafnium; up to 2% vanadium; up to 6~ copper;
up to 5% manganese; up to 70% iron; up to 4% silicon, and the
balance essentially nickel. Cobalt-base alloys of similar
composition can be treated. Among the speci~ic superalloys
might be listed IN-738 and 7g2, Rene alloys 41 and 95, Alloy 718,
Waspaloy, Astroloy, Mar-M alloys 200 and 246, Alloy 713, Alloys
500 and 700, A-286, etc. Vaxious of these alloys are more
workable than others. Other base alloys such as titanium can be
processed as well as refractory alloys such as SU-16, TZM,
Zircaloy, etc. Prealloys contemplated herein can contain up
to 10% or more by volume of a dispersoid such as Y2O3, ThO2,
2 3~ tc.
Finally, it will be understood that modifications
and variations o~ the invention may be resorted to without
departing from the spirit and scope thereof as those skilled
in the art will readily understand. Such are considered to
be within the purview and scope of the invention and appended
claims.
-21-
,67
g ~ ~ ~ ~ ~! o co ~ ., o ~ ~ o ~ ~ o l~
~ ~ ~ o ~ ~r ,i ~ ~ o ~7 ~ o ~i o ~ Ln o o ~D
S ~ g r~ u~ D ~ ~D ~D ~D ~D ~D u~ ~ ~ D CO O ~D
~ ~,~ ,,
o ...... ............ .... .... .... .. .. ............
o ~ ~ ~ D ~ ~ ~ ~ ~ ~_ CO ~ O C~
3 ~D oP o ~ ~D o t`~ r` ~r ~ ,i ~D ,i ~ ~D co ~ r~1 o ~ ~
0'~ ~ D In ~ ~D ~D ~r ~ ~ ~ u~ L~
.. ...... ............ .... .... .... .. .. ............
O .Y O O CO 00 r-l O ~ ~I ~1 ~ ~I t') Ci) N N 11 ~ 1`
~U~ ............. .. .. .~ . . ...... .........
S ~ OD O ~9 ~ CO ~ ~` ~ ~r ~ ~ ~D ~ $ ~r
.... ............ .... .... .... .. .. ............
CO ~ ~ ~ ~ ~9
l 3 0 o\ ~ ` O ~r o _i t`l ~r co ~I 1~ Lo r~
pO, ~ ~` t~ r` a) I` ~ CO C~ C~ ~D g CO ~D I_ I` ~
.. .... ............ ~ ..... ~ .... .... .. .. ............
O ~ ~ OD ~ .~ ~ ~ X h ~ ~` 'a u~ ~9 ~ ~D ~r ,~ o ~ ,,
.C; ~ O ~ O ~i r~ r~ U~ ~:i N ~ ) ~ O CO (~ O co h r` o~ co a~ O (5) ~ D
O ~J N ~ 1 ~ ::~
O ...... ,......... t~ .. gP~ .. ~3 .. ~1 .. .. ...... ......
~:) ~ Ci~ ~ OD O ~1 ~ ~r C,~ O r~ o o ~ N I~ ~ O ~` ')
3 ~ ,\o co 1~ ~ co o~ 4~ h~ c~ ~ E~ ~ O ~ ~ ~r o ~
~ O ,1 ~ ~ ~ ~ O ~ ~ ~ ~ ~ o~ æ a~ ~ ~
H Pl ~ C ) ~ ~1 ~
~ .... .. ............ .... 07 .. H .. .~ .. .. ...... ., ,
O ~ ~ 0 ~ ~D g ~ ~ aJ O O ~ O O O
~ u~ ,1 ~ ~ ~ ~ o o ~ ~ ~ E~ ~ ~1 ~1
h u~ u~ 1: ~ I I I ~ E~ o r) ~: l l l ~ = ~
~: ~ P. ~ ~ ~ ~ 8 ~, ~ ~o o o æ ~
.... .... ............ .... .... .... .. .. ....... ,
'0~ ~ =S ~ ~C
1~ '~ ~
'0~ U~ :
H : ~ ~: ::_ :: : : ::: : ~ ~
...... ............ .... .... .... .. .. ............
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ I I I
........ ............ .... .... .... .. .. ............
~ h --~ ~ : ~~: :: _ : ~ :
-~1 ~ ~
0~ . .
........ ............ .... .... .... .. .. ............
o
~ N --~: ~ :: :: ~ r-l ~ r~) ~O 11') Ll~ OD
O N 111 . . . . . . . . .
~zO'~ O O O 000 000
........ ............ .... .... .... .. .. ............
~ ~ ~ m ~ ~ ~ *
I:o o I I I ~I ~l l~ ~o * * a) o ~
E~ Z -~ ~ ~ 1- a~ ~1 ~ 1
~7~35~
O ~ U~ N ~ N
tJ' o7 ~C (~; Ni ~Yi O O ~`i
S-l ~ ~ N N N N N N
~0~
~E: ... ............
O
o a~ ~ a~
~1 ~ ~1 .. .. ..
3 a) d~ ~ ~I CO
~ I I`~ r
........ .... .... ....
~a
~ ' ~ IS ~ 1 O
O ~ ~ ~ 0~ CO LS) 9 ~D O
~ ~ ~ ,~ ~ ~ ,1,~ ~
U~ '¢ ~) ~1 N
:E: ... .... .... ....
O ~ .,~.,1 ~
OD ~J ~S ~9 ~) ~) ~1 a) N ~J U~ r-l
.. .. .. u
3o a~ o\o ~D O ,1 ~ 1_ S~
::~ P~ ~ 1
~1 .. ...... .... .... .... "i
N N a) ~I N IS)
::~ ~ ~r N r) ~i ,
r~ 1 ~ N ..
a) .... .... .... .... ~ a
o ~
~ O ~
N Ir)~r 1~ C~ 1` ~ rl
Ni r~ ~) 10 ~ ~ ~ Cl
~4 ~ ~5) 1~ ~ O~ (sl t~ ~ h ,~ ~:
41 S~
........ .... .. " ,, ,,
P~ '
~ U~ ~
g ~ ~ O O O O O O ~ ~
tJ` u~ Ul ~ ) O O 10 ,1
Il) ~r) NN N N ~1 U
~1 S l
~ ~ ~0 rl
3 ~
O Q~ N
Pl N
.IJ ~ ~ 41 'U~ o
1~ 1~ ~ o a) f3
~ ~ r~
V
~ ~ m~ a: ~ ~ ~
U~ O l l l l l l ~I P.l N
O ~z; ~ ~:; N
E~ ~ C~ o æ
--23--
.
' : . :
.. . . . . .
