U.S. patent application number 10/820248 was filed with the patent office on 2004-10-28 for infiltration of a powder metal skeleton by a similar alloy to achieve uniform composition.
This patent application is currently assigned to Massachusetts Institute of Technology. Invention is credited to Allen, Samuel M., Lorenz, Adam M., Sachs, Emanuel M..
Application Number | 20040211538 10/820248 |
Document ID | / |
Family ID | 25340164 |
Filed Date | 2004-10-28 |
United States Patent
Application |
20040211538 |
Kind Code |
A1 |
Lorenz, Adam M. ; et
al. |
October 28, 2004 |
Infiltration of a powder metal skeleton by a similar alloy to
achieve uniform composition
Abstract
In infiltrating a porous metal skeleton an infiltrant
composition is used similar to that of the powder skeleton, but
with the addition of a melting point depressant. The infiltrant
quickly fills the skeleton. As the melting point depressant
diffuses into the base powder, the liquid may undergo diffusional
solidification and the material eventually homogenizes. Maintaining
the infiltrant at a liquidus composition for the infiltration
temperature typically ensures that the bulk composition or
properties will remain uniform throughout the part, particularly in
the direction of infiltration. Success of such an infiltration is
enhanced by effective means of maintaining the molten infiltrant at
a liquidus composition. It is also beneficial, in some cases, for
the time scale of the infiltration to be much faster than the time
scale of the diffusion of the melting point depressant and the
subsequent solidification and homogenization. The relative rates of
infiltration and diffusion/solidification rate are significantly
impacted by the choice of materials system. Other factors also
influence these rates. They include: selection of powder size
(diameter), shape, surface roughness, and size distribution,
feeding liquid from different locations, liquid feeder channels,
smoothing of the part surface with fine powder and affecting the
infiltrant fluid properties. Various material systems are also
disclosed, as are methods of designing a process of infiltrating a
part, including binary and ternary and higher component systems.
Homogeneous composition may be achieved using these techniques,
particularly along the direction of infiltration.
Inventors: |
Lorenz, Adam M.;
(Somerville, MA) ; Sachs, Emanuel M.; (Newton,
MA) ; Allen, Samuel M.; (Jamaica Plain, MA) |
Correspondence
Address: |
STEVEN J WEISSBURG
238 MAIN STREET
SUITE 303
CAMBRIDGE
MA
02142
|
Assignee: |
Massachusetts Institute of
Technology
Cambridge
MA
|
Family ID: |
25340164 |
Appl. No.: |
10/820248 |
Filed: |
April 6, 2004 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
10820248 |
Apr 6, 2004 |
|
|
|
09863073 |
May 21, 2001 |
|
|
|
6719948 |
|
|
|
|
60206066 |
May 22, 2000 |
|
|
|
Current U.S.
Class: |
164/98 |
Current CPC
Class: |
B22F 3/26 20130101; B33Y
80/00 20141201 |
Class at
Publication: |
164/098 |
International
Class: |
B22D 019/02 |
Goverment Interests
[0002] The United States Government has certain rights in this
invention pursuant to the Office of Naval Research Award #
N0014-99-1-1090, Research in Manufacturing and Affordability,
awarded on Sep. 30, 1999.
Claims
Having described the inventions, what is claimed is:
1. A method for fabricating a substantially metal part, comprising
the steps of: a. providing a skeleton of interconnected adhered
metal powder having a network of interconnected porosities
throughout, said metal powder composed of at least two elements; b.
providing an infiltrant comprising: i. the same at least two
elements as are in the skeleton; and ii. melting point depressant
(MPD); the infiltrant having a composition that is a liquidus
composition for an infiltration temperature; and c. infiltrating
said skeleton with said infiltrant, at approximately said
infiltration temperature, whereby essentially no erosion of said
skeleton transpires.
2-13. (Canceled)
14. A method for fabricating a substantially metal part, comprising
the steps of: a. providing a skeleton of interconnected adhered
metal powder having a network of interconnected porosities
throughout, said metal powder composed of two or more elements,
chosen as in step e below; b. providing an infiltrant comprising:
i. the same elements as are in the skeleton; and ii. melting point
depressant; the infiltrant having a composition that is a liquidus
composition for an infiltration temperature, the liquidus
composition and infiltration temperature chosen as in step e below;
c. infiltrating said skeleton at said infiltration temperature with
said infiltrant in liquid form; d. subjecting said infiltrated
skeleton to conditions such that a portion of said melting point
depressant diffuses from said infiltrated porosities into said
metal powder, and at least partial diffusional solidification
occurs; and e. choosing said metal powder composition, melting
point depressant, infiltrant composition and infiltration
temperature such that during diffusional solidification of said
infiltrant, relative ratios, of components other than melting point
depressant, in said liquid infiltrant not yet solidified, remain
substantially constant.
15. The method of claim 14, said melting point depressant
consisting essentially of a single element.
16. The method of claim 14, said melting point depressant
consisting essentially of two or more elements, all of which have
similar mass transport characteristics relative to said elements of
said skeleton.
17. The method of claim 14, said step of subjecting said
infiltrated skeleton to conditions such that at least partial
diffusional solidification occurs comprising subjecting said
infiltrated skeleton to constant temperature conditions such that
at least partial isothermal solidification occurs.
18. The method of claim 14, said step of subjecting said
infiltrated skeleton to conditions such that at least partial
diffusional solidification occurs comprising subjecting said
infiltrated skeleton to reducing temperature conditions.
19. The method of claim 14, said skeleton further comprising
melting point depressant.
20. The method of claim 14, said skeleton being substantially free
of melting point depressant.
21. The method of claim 14, said step of choosing comprising
choosing said metal powder composition, melting point depressant,
infiltrant composition and infiltration temperature such that a
liquidus composition and a solidus composition of said infiltrant,
that are joined by a tie line on an equilibrium phase diagram, both
lie on a line of substantially constant relative proportions of
non-MPD components of said infiltrant.
22. The method of claim 21, said step of choosing comprising
choosing said metal powder composition, melting point depressant,
infiltrant composition and infiltration temperature such that the
composition of said skeleton, lies on said line of substantially
constant relative proportions of non-MPD components of said
infiltrant.
23-80. (Canceled).
Description
PRIORITY CLAIM
[0001] This claims priority to U.S. Provisional application No.
60/206,066, filed on May 22, 2000, the full disclosure of which is
fully incorporated by reference herein.
[0003] The inventions disclosed herein will be understood with
regard to the following description, appended claims and
accompanying drawings, where:
BRIEF DESCRIPTION OF FIGURES
[0004] FIG. 1 is a generic equilibrium phase diagram for a mixture
of skeleton material and a single element as a melting point
depressant;
[0005] FIG. 2 shows schematically infiltration of an idealized
capillary channel with an infiltrant at liquidus composition and
subsequent diffusion and diffusional solidification to achieve
uniform final bulk composition;
[0006] FIG. 3 shows schematically, the percentage of melting point
depressant within the capillary channel and the surrounding
skeleton walls of FIG. 2, at the locations along the capillary
channel designated I and II, at three different times (unprimed, ',
");
[0007] FIG. 4 shows schematically dissolution of a pure nickel
skeleton after dipping into an undersaturated (off liquidus) pool
of Ni-11 wt % Si infiltrant for 5 minutes at 1200.degree. C.;
[0008] FIG. 5 shows schematically erosion at the base of a
cylindrical skeleton, which also progresses several centimeters
into the part;
[0009] FIG. 6 shows schematically a skeleton composed of .about.300
micron powder infiltrated to a height of about 22 centimeters
before freezing choked off the flow of infiltrant;
[0010] FIG. 7A is a nickel-silicon equilibrium phase diagram;
[0011] FIG. 7B is an enlargement of a portion of the Ni--Si
equilibrium phase diagram of FIG. 7A, which shows schematically an
infiltrant at liquidus composition within a processing window,
using the nickel-silicon binary system as an example;
[0012] FIG. 8 is a nickel-phosphorous equilibrium phase
diagram;
[0013] FIG. 9 is an aluminum-silicon equilibrium phase diagram;
[0014] FIG. 10 shows schematically a cross-section of a base powder
particle coated with surface powder {fraction (1/50)} the size of
the base powder, to increase capillary pressure;
[0015] FIG. 11 shows schematically smoothing effect of a moving
solidification front, with the initial surface matching that of
FIG. 10, with the interface shown moving in steps of 1/4 of the
diameter of the surface powder;
[0016] FIG. 12 is a digital image that shows schematically a
cross-section showing enhanced surface texture of nickel powder
made by hydrometallurgical processing;
[0017] FIG. 13 shows schematically external supply tabs used to
feed infiltrant to several entry points of a part skeleton from an
infiltrant reservoir above the part;
[0018] FIG. 14 shows schematically a network of internal feeder
channels built into a part skeleton to facilitate liquid flow to
remote areas without freezing, the channels having variable
diameter (left) or change in direction (right) including both
horizontal and vertical runs;
[0019] FIG. 15 shows schematically, in cross-section, surface
texture refinement achieved by adding a paste of fine powder to an
external surface of a skeleton of larger powder, the paste forming
a thin shell outside the skeleton (left) or penetrating and filling
the space between powder near the surface (right);
[0020] FIG. 16A shows schematically distortion of a nickel skeleton
(first letter is deformed) which occurred while hanging the
skeleton at 1200.degree. C.;
[0021] FIG. 16B shows schematically a similar part without
distortion that was resting on a flat crucible bottom;
[0022] FIG. 17 is a portion of a titanium-silicon equilibrium phase
diagram;
[0023] FIG. 18 is a ternary nickel-silicon-chromium equilibrium
phase diagram at 1250.degree. C.;
[0024] FIG. 19 is a ternary nickel-silicon-iron equilibrium phase
diagram at 1200.degree. C., and shows schematically how the
relative proportions of nickel and iron change during diffusional
solidification; and
[0025] FIG. 20 is a generic ternary equilibrium phase diagram, and
shows schematically desirable characteristics to achieve uniform
final bulk composition.
DETAILED DISCUSSION
[0026] Traditional manufacturing processes using powder metallurgy
("PM") produce a near net shape part which is only initially 50-70%
dense. These `green` parts then undergo further processing to
achieve full density and the desired mechanical properties either
through lightly sintering and infiltrating with a lower melting
temperature alloy or through a high temperature sintering alone. In
the first method, the part's dimensional change is typically only
.about.1% making it suitable for fairly large (.about.0.5 m on a
side) parts, but the resulting material composition will be a
heterogeneous mixture of the powder material and the lower melting
temperature infiltrant. In the second method, sintering the powder
to full density will result in a homogeneous final material, but a
part starting at 60% density will undergo .about.15% linear
shrinkage. For this reason, full-density sintering is typically
only used for smaller (<5 cm on a side) parts.
[0027] In some cases, infiltration can be done extremely rapidly by
the application of external pressure. However, this requires a mold
and typically expensive processing equipment. The inventions
disclosed herein are directed to pressureless infiltration, where
the primary driving force is capillarity and in some cases,
gravity.
[0028] In many critical applications (structural, aerospace,
military), a material of homogeneous composition (or with
homogeneous properties) is preferable because of certification
issues, corrosion issues, machinability, or temperature limitations
that might be imposed by the lower melting point infiltrant.
Further, because designers of metal components are not accustomed
to working with composites of heterogeneous composition, they
experience a psychological barrier to adoption.
[0029] Creation of very large parts with homogeneous composition or
properties via powder metallurgy builds on all of the benefits of
PM processing. This can be done using an infiltration step to
densify the green part without any significant dimensional change,
but in such a way that the final material has a homogeneous
composition or properties to enable significant advantages over
tradition processing. It is also beneficial to ensure the bulk
material composition or properties are consistent throughout the
entire part. Solid freeform fabrication processes, (such as
three-dimensional printing, selective laser sintering, etc.) metal
injection molding, or other PM processes will be enabled to make
homogeneous net shape parts in a wide variety of sizes by methods
described herein. Also disclosed is the potential of matching the
final part composition or properties to existing commercial
material systems.
[0030] By three-dimensional printing, it is meant the processes
described generally in U.S. Pat. Nos. 5,204,055, 5,387,380,
5,490,882, 5,775,402, which are incorporated herein by
reference.
[0031] A general concept, explored more fully below, is to use an
infiltrant composition similar to that of the powder skeleton, but
with the addition of a melting point depressant. The infiltrant
quickly fills the powder skeleton. Then, as the melting point
depressant diffuses into the base powder, in some cases, the liquid
undergoes diffusional solidification and the material eventually
homogenizes. The diffusional solidification may be isothermal, but
need not be. Maintaining the infiltrant at a liquidus composition
for the infiltration temperature typically ensures that the bulk
composition or properties will remain uniform throughout the part,
particularly in the direction of infiltration.
[0032] Success of such an infiltration is enhanced by effective
means of maintaining the molten infiltrant at a liquidus
composition. It is also beneficial, in some cases, for the time
scale of the infiltration to be much faster than the time scale of
the diffusion of the melting point depressant and the subsequent
solidification and homogenization. Methods of establishing the
liquidus composition include all of the following or a combination
of any of these: separating the infiltrant melt supply from the
skeleton prior to infiltration, adding excess skeleton material to
the melt, overshooting the infiltration temperature, and agitating
the melt. The relative rates of infiltration and
diffusion/solidification rate are significantly impacted by the
choice of materials system. But other techniques have been
developed, and are disclosed herein, to influence these rates. They
include: selection of powder size (diameter), shape, surface
roughness, and size distribution, feeding liquid from different
locations, liquid feeder channels, smoothing of the part surface
with fine powder and affecting the infiltrant fluid properties.
[0033] In some cases, even after the part has reached its
equilibrium condition at the infiltration temperature, some of the
infiltrant in the skeleton will remain liquid after diffusional
solidification has ceased. In some such circumstances, the final
microstructure that results is not homogeneous, but rather is
similar to that typically obtained with a cast part, which is also
a useful result.
[0034] Because significant mass transport occurs in parts after
infiltration has occurred, there exists the potential for very
small voids to develop within the part due to differential movement
of species. These voids are generally referred to as Kirkendall
porosity, and may appear to varying degrees based on factors such
as the mechanism of diffusion and the relative size of mobile
species. This porosity is typically very fine scale and may not
affect mechanical properties. Heat treatment, including hot
isostatic pressing, can be used to reduce these voids in the event
they are significant.
[0035] Infiltrant with Single Element as Melting Point
Depressant
[0036] The initial discussion is limited to the important case of
an infiltrant composed of the skeleton material with the addition
of a single element from the Periodic Table to serve as a Melting
Point Depressant (MPD). In this case, it is generally possible to
design the infiltrant composition and the infiltration temperature
in such a manner as to guarantee that the infiltrated body is
uniform in bulk composition, that is, that there is no gradient in
bulk composition along the path of infiltration.
[0037] A Melting Point Depressant (MPD) is a material which, when
added to a metal, produces a new alloy which melts at a lower
temperature than the metal itself. The MPD typically is composed of
a single element, however, multiple elements are also possible. The
metal itself may be a single element, however, alloys composed of
two or more elements are most common. Alloys typically have
temperature ranges over which they melt and not just a single
melting temperature. An alloy begins to melt at the solidus
temperature and becomes fully molten at the liquidus temperature.
The addition of an MPD to a metal produces a new alloy with a lower
liquidus temperature (temperature at which the alloy is fully
molten) than the liquidus temperature of the metal itself. (In the
case of the addition of an MPD to a metal composed of a single
element, the liquidus temperature of the alloy is lower than the
melting temperature of the elemental metal.) In a preferred
embodiment of some of the inventions disclosed herein, the MPD will
result in an alloy whose liquidus temperature is below the solidus
temperature of the original metal of a skeleton. In this way, the
infiltrant alloy formed of the metal with MPD added can be fully
molten while the metal is still fully solid and thus infiltration
can take place without the skeleton beginning to melt.
[0038] In a binary system there are only two elements present in
the skeleton and infiltrant. In the simplest case, the skeleton is
composed entirely of a single element and the infiltrant is
composed of this element with the addition of a second element as a
melting point depressant. However, the same principles apply if the
skeleton started with some of the MPD in it and the infiltrant
simply has a higher concentration of this MPD.
[0039] An equilibrium phase diagram for a generic mixture of a
skeleton material and a melting point depressant (MPD) is shown in
FIG. 1. The infiltration temperature can be chosen anywhere between
the eutectic temperature (.about.1100.degree. C.) and melting
temperature of the skeleton (.about.1440.degree. C.). If the
skeleton is a pure metal, such as nickel, it will have a discrete
melting temperature. If, however, it is an alloy containing two or
more elements, discussed more below, there is a range of
temperatures over which different components of the skeleton begin
to melt. The lower limit of this temperature range is the solidus
temperature of the skeleton, the maximum temperature at which no
liquid is present. Since there is always some variation in
processing temperature, the infiltration temperature should remain
safely below the solidus temperature of the skeleton.
[0040] The skeleton will also be prone to sinter and start to sag
as it loses strength near its melting point, discussed below. As
long as the skeleton maintains dimensional stability, the
infiltration temperature can also be selected to influence the
diffusivity and solubility of the MPD in the skeleton material.
Generally, the diffusivity increases and solubility decreases with
increasing processing temperature. These tendencies are relevant to
the challenge of ensuring that the entire skeleton fills with
liquid before solidification chokes off the liquid flow, discussed
in more detail below.
[0041] It is helpful to now consider the case where the composition
of the infiltrant liquid at the infiltration temperature lies along
the liquidus of FIG. 1. More specifically, consider the case of
infiltration at temperature T and liquidus composition C. At the
infiltration temperature shown (.about.1300.degree. C.), the
infiltrant would be liquid at any composition between 10% and 50%
MPD. The minimum composition that allows the material to remain
completely liquid is 10% MPD--the liquidus composition at the
designated temperature. Liquid infiltrant at this composition is
considered saturated with the skeleton material, for a binary
system. Any removal of MPD from the infiltrant at this temperature
will result in solidification of some of the infiltrant at the
corresponding solidus composition at point S. Such removal of MPD
will typically take place during the infiltration of the skeleton
by diffusion of MPD into the skeleton. Thus, as the MPD diffuses
into the skeleton, the infiltrant will solidify on the skeleton at
the solidus composition and the remaining liquid will still be at
the liquidus composition--unchanged by the process of infiltration.
Thus the infiltrant flowing through the part will always be at the
liquidus composition and the bulk composition throughout the part
will be ensured to be uniform.
[0042] Reference to FIGS. 2 and 3 further illustrates this concept.
FIG. 2 shows schematically a saturated melt 110 at a liquidus
composition filling a capillary channel 112. The capillary channel
112 is formed between two identical sheets 114 of skeleton material
with spacing and void fraction chosen so that the volume fraction
of solid to void space is 60:40. FIG. 2 shows three different
moments in time, with the left most (unprimed) being the earliest
and the right most (double prime ") being the latest. FIG. 3 shows
the expected MPD concentration profile for two locations at the
three moments in time. The vertical axis shows the local
composition (percent of MPD) as a function of the position on the
horizontal axis, with the coordinates c, d, e and f, representing
corresponding locations of the sheets of skeleton material and
capillary channel shown in FIG. 2. Profile I in FIG. 3 represents
the initial condition, when and where the liquid first comes into
contact with the skeleton. This profile is found just below the
meniscus M as the liquid is flowing up the capillary 112 indicated
at I in FIG. 2. At a slightly later time shown in the middle, an
identical condition would be found further along the capillary at
position II'. The composition profile is 10% MPD in the liquid
region and near zero in the solid 114, since the MPD has just begun
to diffuse into the solid 114 at the interface. At the position
marked as I', the liquid has been in contact with the skeleton for
a given time period and diffusion has caused some degree of
solidification of the infiltrant. The composition profile labeled
I' in FIG. 3 corresponds to the cross-section at the position I' in
FIG. 2, with an increased MPD composition in the solid and
resulting motion of the solid/liquid interface inward (from d away
from c and from e away from f). The right most image of FIG. 2
portrays the capillary 112 after the liquid 110 has reached the top
of the capillary channel 114, and the system has had additional
time to equilibrate. Since the composition of all of the liquid
remains constant regardless of whether the liquid is flowing or
not, the solidification behavior and the profile corresponding to
the cross-section at position II" will be identical to I'.
Similarly, point II" at a future moment in time will be identical
to the current profile of I". After the entire system reaches
equilibrium, the composition profile at I (lower position) and II
(upper position) will be identical and the final bulk composition
throughout the capillary will be uniform.
[0043] In contrast, in the case of a liquid infiltrant which has a
concentration of MPD above that of the liquidus composition, (such
as corresponding to the composition for the point indicated at U,
FIG. 1), mass transport of both the skeleton material and the MPD
has the potential to change the infiltrant composition. If this
happens while the liquid is still flowing into the skeleton, the
last areas of the skeleton to be reached by the liquid will have a
different final composition of MPD than the first areas reached.
Such a variation in bulk composition throughout the part could not
be rectified by a homogenizing heat treatment in a reasonable
time.
[0044] To visualize this concept, consider another similar
idealized capillary channel made by two identical sheets of
skeleton material with spacing chosen so that the volume fraction
of solid to void space is 60:40. Filling the void space within the
control volume with an infiltrant composed of 15% MPD would result
in an average bulk composition of 6% MPD. For the phase diagram
shown in FIG. 1, the infiltrant would be unsaturated (off the
liquidus) at the infiltration temperature. The infiltrant would
exchange mass with the skeleton material upon contact at its entry
point, until reaching its equilibrium liquidus composition of 10%
MPD. For the case of a capillary being filled from the bottom, this
could occur while the liquid is flowing through the capillary such
that the bulk composition at the top would be only 4% MPD, due to
MPD depletion. At the bottom (the entry point), the bulk
composition would be greater than 6% MPD because of the loss of
skeleton material and subsequent replacement with infiltrant having
a composition of 15% MPD. Such a variation of bulk composition
would result in undesirable variation of properties throughout an
infiltrated part. In simple terms, this is the result of lower
sections of a part skeleton being dissolved into the liquid and
then carried by the liquid to other regions of a part. (If
infiltrant enters the part from its top rather than its bottom,
such as by simply placing a slug of infiltrant supply material on
top of a skeleton and heating it, then the top regions would have a
higher contribution based on the MPD. Basically, the region
adjacent the infiltrant supply will have increased contribution to
composition from MPD.) The case of ternary and higher compositions
is similar, with an important added consideration. It is still true
that the infiltrant should be designed and controlled to be at a
liquidus composition for the chosen temperature of infiltration
(see subsequent discussion on erosion). It is further true that the
removal (by diffusion) of any of the MPD will result in
solidification of material at a solidus composition. However, in
the case of a ternary system, for example, at a single temperature
there are a range of possible liquidus compositions and a range of
possible solidus compositions (as opposed to just a single liquidus
and a single solidus composition in the binary case).
[0045] FIG. 18 shows a ternary phase diagram for the system
consisting of nickel, chromium and silicon at a temperature of
1250.degree. C. (This and all subsequent ternary phase diagrams
were generated using Thermo-calc, a Computational Thermodynamics
program used to perform calculations of thermodynamic properties of
multi-component systems based on the Kaufman binary thermodynamic
database.) Line 780 is the liquidus (at this temperature) and any
composition falling along this line is a liquidus composition. Line
782 is the solidus and any composition falling along this line is a
solidus composition. Further, tie lines 784 connect specific
liquidus and solidus compositions. The liquidus and solidus
compositions at the end of each tie line can co-exist in two-phase
equilibrium.
[0046] To ensure that the infiltrated part does not develop a
composition gradient along the path of infiltration due to
diffusional solidification during infiltration, the liquid
infiltrant must solidify with no change in the relative
contributions to the infiltrant composition of the non-MPD
elements. To illustrate, consider the general case where an
infiltrated part would develop a gradient in bulk composition due
to diffusional solidification--shown schematically in FIG. 19 using
the ternary phase diagram for the system consisting of nickel,
iron, and silicon at a temperature of 1200.degree. C. In this case,
line 890, which passes through the corner of the diagram
corresponding to pure Si, corresponds to compositions that have a
constant ratio of Ni and Fe (approximately 72:28). There are other
lines of constant ratio, which represent different ratios. A
liquidus composition marked 820 is on the line. The solidus
composition marked 822 that is connected to the liquidus by a tie
line, is not on the line. Thus, the solidus composition has a
different ratio of Fe to Ni than does the liquidus composition. In
such a case, if the infiltrant begins at a liquidus composition
such as 820, diffusional solidification of the infiltrant will
result in the remaining liquid becoming relatively richer in Ni
(and poorer in Fe). This Ni enriched infiltrant will travel up the
skeleton, resulting in the further reaches of the infiltrated part
having a higher composition of Ni than the first areas to
infiltrate. This is somewhat undesirable.
[0047] The desirable circumstance is best illustrated by reference
to FIG. 20, a ternary phase diagram identical to FIG. 18 with the
elements labeled A, B and MPD to represent a generic system. In
this case, line 990, which passes through the corner of the diagram
corresponding to pure MPD, corresponds to compositions that have a
constant ratio of A and B, the non MPD components of the
infiltrant. The desirable case then is an infiltrant liquidus
composition whose tie line lies along this line, as shown. In this
case, the liquidus and solidus that are in equilibrium with each
other have the same relative concentration of A and B and
diffusional solidification will not result in a change in the
relative composition of A and B. This will guarantee that there is
no variation of composition along the path of infiltration. Not all
material systems and infiltration conditions will allow for this
condition. Rather, the material system, liquidus composition
(infiltrant composition) and infiltration temperature must be
chosen by these criteria.
[0048] In the most desirable circumstance, the tie line lies along
a line of constant relative proportions of A and B as above, and
the composition of the skeleton also lies on this same line of
constant relative proportions of A and B. In the case of FIG. 20,
the skeleton material composition could be chosen at point 992.
Thus, as the infiltrant undergoes diffusional solidification, the
solidified infiltrant has the same relative proportions of A and B
as the skeleton and this will be true along the entire path of the
infiltrant. Thus, there will be no need to wait for diffusion of
either or both species A and/or B between the skeleton and
solidified infiltrant in order to attain uniform composition
between them.
[0049] In the case of ternary and higher alloys, not all materials
systems will allow for the selection of infiltrant alloy such that
the tie line is along a line of constant relative proportions of A
and B. Further, the added desirable feature of having the skeleton
composition lie on this same line of constant relative proportions
is more restrictive.
[0050] Thus, an aspect of one of the current inventions is to
select and design materials systems according to the criteria
described. This includes the selection of the elements in the
skeleton, the selection of the MPD, the selection of the relative
amounts of the non-MPD elements in both the skeleton and in the
infiltrant (if different), and the selection of the infiltration
temperature. The most preferred case is that the tie line lies
along a line of constant relative composition of non-MPD elements
and that the skeleton composition lies on this same line. However,
having the tie line lie on this line with the skeleton composition
not lying on this line is sufficient to guarantee uniform
composition along the infiltration path. Uniform composition
between infiltrant and skeleton might then be attained by
diffusional homogenization. Note that a change in the infiltration
temperature will change the orientation of the tie lines and so,
the selection of infiltration temperature must also be based on
this consideration.
[0051] The principles explained herein in the context of the
ternary phase diagram also apply to systems with four or more
alloying elements. In particular, a system is to be chosen such
that, during diffusional solidification, the relative ratio of the
non-MPD elements remains substantially constant and, preferably
that this ratio is substantially the same in the skeleton as in the
infiltrant.
[0052] Methods of Ensuring a Liquidus Composition (Saturation)
[0053] If the infiltrant composition is known exactly, the process
temperature can be selected to exactly match the liquidus
temperature for that composition, but this requires very accurate
process control. A more robust method for ensuring that the liquid
lies at an appropriate liquidus composition, is to put the liquid
in contact with sacrificial excess solid skeleton material and
allow it to reach an equilibrium liquidus composition corresponding
to the actual processing temperature. Having a high interfacial
surface area between the liquid and solid can help promote mass
transport and speed the process of equilibration. For this reason,
it is beneficial if the excess solid material is supplied in powder
form, which has large surface area, but this alone may not suffice
to guarantee reaching the liquidus composition in reasonable times.
In the case of a binary system, as explained above, the liquidus
composition is also referred to as a composition saturated with
skeleton material.
[0054] If raising the MPD concentration lowers the liquid density
(which is typically the case), the liquid will stratify with the
higher density liquid of low MPD concentration in contact with the
solid at the crucible bottom. The liquid of higher MPD
concentration will remain at the surface with no mixing by natural
convection. Stirring the melt or using some other means of
agitation to force convection promotes mixing. A ceramic propeller
has been used to stir the infiltrant supply by running a shaft
through the furnace roof with a Teflon seal a small motor to power
the propeller. Other possible mechanical stirring methods that may
be used include tipping the crucible back and forth, flowing the
liquid through a sacrificial porous network of skeleton material,
shaking, vibrating or sonicating the melt, or bubbling gas through
the melt. Placing the liquid in an inductive AC electromagnetic
field can also generate substantial mixing by inducing currents in
the molten metal. Heating the infiltrant supply by induction is one
means for preparing a well-mixed liquid at the equilibrium liquidus
temperature. (As used herein, "infiltrant" typically means liquid
material that actually infiltrates a skeleton. "Infiltrant supply"
means the source material that will melt to become the infiltrant.
The infiltrant source material becomes liquid infiltrant plus
residual solid at the infiltration temperature; liquid infiltrant
is what is available to enter the skeleton.) Another method to help
ensure the liquid is at its liquidus composition is overshooting
the infiltration temperature to dissolve excess solid skeleton
material that has deliberately been added to the melt. Once the
excess material is dissolved, the temperature is slowly ramped back
down to the infiltration temperature while agitating the melt. This
promotes re-solidification of material, with the remaining liquid
at the desired liquidus composition. As long as the liquid is in
contact with some solid, it is unlikely that any under-cooling
would occur.
[0055] The amount of excess skeleton material added to the melt
must be sufficient to saturate the melt, but not so much that the
melt solidifies. The proper amount is a function of the skeleton
material's solubility (maximum capacity to absorb MPD). For
example, with a nickel-silicon infiltrant, sacrificial excess
nickel powder is added to the crucible of infiltrant. The
appropriate amount is determined by considering the extreme cases
for a range of processing temperatures. FIG. 7A shows an
equilibrium phase diagram for nickel and silicon, and FIG. 7B shows
an enlargement of a portion of FIG. 7A. FIG. 7B illustrates how
this would be done for a desired infiltration temperature of
1180.degree. C. and maximum temperature variation (due to
uncertainties) of plus or minus 20.degree. C. Above and to the
right of the liquidus, all compositions are liquid. Below and to
the left of the solidus, all compositions are solid. Between, there
are compositions that have two coexisting phases, liquid and solid.
The bulk composition of infiltrant supply is chosen from the
intersection of the maximum anticipated temperature (in this case
1200.degree. C.) with the liquidus line, marked as A on FIG. 7B (in
this case, 10% Si and 90% Ni). This ensures that some solid will be
present at any temperature below this expected maximum temperature
and all of the liquid present will be saturated with nickel and be
at the liquidus composition for that temperature. If the
temperature is at the lower limit, the total amount of the material
provided as the infiltrant supply will be partially liquid and
partially solid, in a two-phase field between the liquidus and the
solidus. The ratio of liquid to solid will be given by the lever
rule. For this example, at 10% Si and 1160.degree. C., it would be
approximately 30% solid. This will determine the total quantity of
infiltrant supply needed, since only 70% of the infiltrant supply
is guaranteed to be liquid infiltrant available for filling the
part in this example.
[0056] For the case of ternary or higher alloys, there can be a
range of liquidus compositions at a given temperature. The
infiltrant supply composition can be selected such that it contains
the elements of the skeleton material in a relative ratio similar
to their elemental composition in the skeleton material alone. Once
again, the bulk infiltrant supply composition can be selected to
lie in the two-phase region between the solid and liquid at the
infiltration temperature, such that any liquid present will be at a
liquidus composition. As described previously in FIG. 19, the
solidus and liquidus compositions corresponding to a given tie line
may have different relative proportions of the elements of the
skeleton material. Solidification due to diffusion of MPD may
result in depletion or enrichment of other species in the liquid,
but this would only result in the liquid moving to a different
liquidus composition. The new liquidus composition would still not
allow dissolution of any skeleton material. Any enrichment or
depletion of the solidifying composition would come from the
changing liquid composition rather than the existing skeleton
material.
[0057] Erosion
[0058] As mentioned above, if the liquid infiltrant has a
composition that is not saturated in the skeleton material, in
other words, where the concentration of the MPD is greater than in
the equilibrium liquidus composition for a given temperature, the
liquid infiltrant will have the capacity to absorb additional
material from the skeleton and partially dissolve the skeleton.
This can happen very quickly, because of high diffusivity in
liquids and can be a significant problem, especially when a large
melt pool is used. FIG. 4 shows a pure nickel skeleton, originally
a cylinder, with its bottom section dissolved. It was dipped into a
pool of molten Ni-11 wt % Si for 5 minutes at 1200.degree. C. Since
the equilibrium liquidus composition has significantly less than 11
wt % Si at that temperature, the pool of liquid absorbs the solid
nickel as it contacts the skeleton.
[0059] To a lesser degree, a liquid infiltrant not already at its
equilibrium liquidus composition has a tendency to leave an erosion
path as it enters the skeleton. This occurs to some extent in most
powder metal infiltrations, but usually is limited to the initial 1
cm at the base of a part. In such cases, the part to be infiltrated
can be placed on a sacrificial stilt. In the nickel-silicon system
with an unsaturated infiltrant containing more than the liquidus
silicon concentration, the erosion tends to propagate for several
centimeters into the part and resembles a riverbed (one example is
shown in FIG. 5). This part is approximately 10 cm. long. It was
infiltrated from the end near to the zero of the scale, the bottom
of FIG. 5, as shown. Studying the erosion pattern on several
different shaped parts suggests that erosion occurs in the areas of
highest liquid flow. Once erosion begins, a larger channel is
created, which has less viscous drag and allows even more liquid to
flow through the newly formed channel. An instability such as this
explains why the erosion progresses so far into the part (almost 4
cm.). Through metallographic study of cross sections, the eroded
areas are found to be high in silicon content. This is not
surprising, since those compositions would be liquid at the
infiltration temperature. The areas of erosion are not limited to
the surface. Voids have been found within the interior of a part in
a region of high silicon content.
[0060] Using a liquid infiltrant that is at its equilibrium
liquidus composition before it contacts the skeleton removes the
driving force for diffusion and has proved to be a good method of
preventing erosion. Since temperature affects both the diffusion
rate and the infiltrant liquidus composition, temperature variation
with time or a temperature gradient set up within the part, could
be used as a further method of erosion prevention.
[0061] Infiltrant with Multiple Elements Used to Depress Melting
Point
[0062] Two or more elements can be used as a melting point
depressant. In that case, the mass transport can become more
complicated and having a liquidus composition no longer provides a
guarantee of uniform bulk composition throughout the part. The
diffusivity and solubility in the skeleton material of the various
MPD elements will determine their mass transport during the
infiltration. If the elements behave similarly to each other, the
bulk composition would likely become uniform for the same reasons
discussed in the previous section. However, significant diffusion
of one element without the other, while the liquid is still
flowing, would likely result in variation of final bulk
composition. This is the more typical situation. For example, if a
second element has significantly lower solubility in the skeleton
material, less of that element will be absorbed in the solidifying
material and the remaining liquid will be enriched in that second
element. This enriched liquid would be carried to other regions of
the part, while a fresh supply of infiltrant at the original
composition replaces it.
[0063] In these cases of multiple elements having different mass
transport properties, uniform bulk composition can be achieved by
filling the part with liquid in a much shorter time scale than the
time scale of diffusion and solidification. Factors that influence
the rate of filling and of diffusion are discussed below.
[0064] Relative Rates of Infiltration and Diffusion
[0065] Due to the diffusion of MPD into the skeleton and
corresponding infiltrant solidification, in many cases the liquid
has only a limited time to fill the part skeleton before the flow
is choked off by solidification. In addition, fast infiltration
relative to diffusion may be necessary to ensure uniform bulk
composition as mentioned above for cases with multiple elements
diffusing. The rate of infiltration is determined by various
factors, including: the surface tension, wetting angle, viscosity,
and density of the liquid infiltrant as well as the geometry of the
skeleton, determined by powder size and shape, size distribution,
packing density and part geometry. Specific relations are provided
below in the discussion of each topic. FIG. 6 shows a Ni--Si
skeleton which has been infiltrated to a height of 22 cm by
controlling the relative rates of infiltration and diffusion such
that diffusion had not proceeded to the point of choke off before
the infiltrant reached 22 cm.
[0066] Some typical infiltration rates have been measured of the
Ni-10 wt % Si infiltrant, filling a skeleton of 50-150 micron
nickel powder. This was done by hanging the skeleton from a wire
through the roof of the furnace and measuring the force on the
wire. By compensating for the surface tension and buoyancy forces,
it was possible to relate the force to the increasing mass of the
part due to the addition of liquid. The liquid filled an 8 cm tall
skeleton in approximately one minute. Other liquid metals have
similar viscosity and surface tension so this rate should not
change drastically with material system.
[0067] The diffusion rate controls diffusional solidification, a
special case of which is isothermal solidification of the
infiltrant and the eventual homogenization of the skeleton. Since,
in a typical case, the liquid fills a small skeleton in
approximately one minute, diffusional solidification would ideally
take place over a much longer time period, e.g., an hour or two.
The diffusion rate will be controlled primarily by the material
system chosen. Selection of a material system is critical to
controlling the time scale of the isothermal solidification. In
particular, the diffusivity of the melting point depressant in the
solid skeleton will have the greatest effect on the freezing. The
basic components of the skeleton are usually dictated by other
requirements. A part is typically specified as steel, or aluminum,
etc. and thus, that metal or an alloy thereof will be the basic
component of the skeleton. Part geometry is also typically not a
variable the process designer can change. Using a slower diffusing
melting point depressant can drastically increase the amount of
time the skeleton has to fill with infiltrant before freezing
begins to occur. Si, B and P can all be used as a melting point
depressant in Ni. Of these, Si diffuses much more slowly than do B
and P. Diffusivity also has a strong dependence on temperature,
since it is an activated process that follows Arhennius dependence.
Controlling infiltration temperature allows for some control of the
diffusivity for a given material system. Reduced temperature
decreases diffusivity and should allow more time for the liquid to
fill the skeleton before freezing.
[0068] Coating the powder skeleton (or just the raw powder) with a
finite time diffusion barrier slows the freezing by keeping the
melting point depressant from leaving the infiltrant until the
liquid has filled the part. Such a diffusion barrier can be another
metal that has a lower diffusivity of MPD. The thickness of the
barrier can be selected so that it only lasts for the duration of
the infiltration. As the coating material begins to break down, it
allows the MPD to diffuse through, allowing isothermal
solidification and eventual homogenization. Ideally, the coating
material itself is also relatively homogeneous throughout the
part.
[0069] Distinction between Low and High Solubility Systems
[0070] Returning to the case of a single element as melting point
depressant, the solubility of the melting point depressant in the
skeleton material also influences the diffusional solidification
behavior. The solidus line on the phase diagram describes the
solubility of the MPD in the solid, and the location of the final
part bulk composition relative to this line will determine whether
the part will completely solidify at that temperature or if there
will always be liquid present at the infiltration temperature. The
two most important factors in determining the bulk composition of
the final part are: the packing density of the powder skeleton
(which determines the liquid void fraction); and the infiltration
temperature (which determines the infiltrant composition for a
given material system assuming the infiltrant is at the liquidus
composition).
[0071] In the first case of relatively high solubility, such as
Ni--Si, characterized by the equilibrium phase diagram shown in
FIG. 7A, solidification will proceed to completion and choke off
the flow of liquid. It is therefore necessary to ensure that the
liquid fills the entire part before the infiltrant solidifies.
Several mechanisms for increasing the rate of infiltration relative
to the rate of solidification are presented below.
[0072] As an example, a pure nickel skeleton infiltrated with an
alloy containing silicon as a melting point depressant would likely
fall into this category. The equilibrium phase diagram for Ni--Si
is shown in FIG. 7A. Assuming a skeleton with packing fraction of
60% and an infiltration temperature of 1180.degree. C., a bulk
composition of .about.4% would result from the liquidus composition
of .about.10% times the liquid volume fraction 0.4. Since this is
significantly less than the solidus composition of .about.7%, such
a part would completely solidify diffusionally.
[0073] In a second case, of low solubility, when the bulk
composition of the part lies in a two-phase equilibrium field,
liquid will remain present in the infiltrated skeleton until the
part is cooled below the infiltration temperature. It is possible
that the liquid volume fraction may remain high enough to allow
continuous flow through interconnected pores, such that any
diffusional solidification that occurs would not prevent the part
from being completely infiltrated. For example, consider a pure
nickel skeleton infiltrated with a nickel alloy containing
phosphorous as a melting point depressant. The Ni-P equilibrium
phase diagram is shown in FIG. 8, with the liquidus line indicated
at 270. The solidus line 272 is not discernible on this diagram,
because it is so close to 0% P line, between 1455.degree. C. and
870.degree. C. A lower infiltration temperature of 1000.degree. C.
could be used and the liquidus composition of 7% P would result in
a bulk composition of 2.8% P. Since the solubility of P in Ni is
only 0.17%, only a small amount of P would diffuse into the
skeleton at the infiltration temperatures and there would be very
little solidification and restriction to the liquid flow.
[0074] Another material system with low solubility is the binary
alloy of aluminum and silicon, with the equilibrium phase diagram
shown in FIG. 9. The liquidus 280 and the solidus 282, along with
the line at 577.+-.1.degree. C. bound a two-phase region. At an
infiltration temperature of 600.degree. C., there is approximately
2% solubility of silicon in aluminum. If liquid infiltrant at a
composition of 10% Si filled the void space of a 60% dense pure
aluminum skeleton, the bulk part composition would be 4% Si and
would lie in a two-phase field of .about.25% liquid as long as the
skeleton remained at the infiltration temperature. In this case,
the skeleton would absorb some Si, but diffusional solidification
would occur only until the part was 75% solid. The liquid flow
would not be completely choked off by the solidification, so the
rate of diffusion and solidification need not be slow compared to
the infiltration rate.
[0075] The final microstructure resulting in these cases of low
solubility of the MPD in the skeleton material will not be a
single-phase solid solution of the MPD in the skeleton material.
The original interconnecting porosity space that was filled with
liquid will have a two-phase microstructure. The aluminum-silicon
system's result would resemble that of a cast microstructure. The
powder particles that have absorbed Si would substitute for the
primary dendrites that solidify first in a casting. The remaining
infiltrant would have a eutectic microstructure similar to the
regions of a casting between the dendrites that are the last to
freeze. Since this type of microstructure is sometimes desirable
and widely accepted in industry, the fact that the material
composition is not uniform throughout is not a drawback. Two-phase
strengthening is common for commercial net-shape casting alloys and
can also be achieved in cases of infiltrated systems with low
solubility.
[0076] By contrast, high solubility cases are more typical of
commercial wrought alloys, relying on solid solution strengthening
or precipitation hardening. Either the high or low solubility case
will result in more uniform properties than traditional
heterogeneous infiltration and will eliminate the disadvantages of
poor machinability, poor corrosion resistance, temperature
limitations, and difficulty in material certification.
[0077] It should be noted that even when diffusional solidification
takes place, it is not necessary to wait for it to complete before
lowering the processing temperature. Once the infiltrant has
completely filled the skeleton, the skeleton can be cooled to
another temperature, subsequent solidification and homogenization
can continue to take place by diffusion. This could be useful
because the solubility of the MPD in the skeleton typically changes
with temperature.
[0078] Gating
[0079] If the infiltrant begins to wick in to the part as soon as a
small portion of the infiltrant supply is molten, two problems
result. First, as soon as infiltration begins, diffusion and
homogenization also begin and the pores of the skeleton may become
occluded by material that has undergone diffusional solidification.
Thus, a small amount of molten infiltrant may cause the pores to
clog before the general mass of infiltrant supply is available to
enter the body. Second, if the infiltrant supply, during its
preparation, has solidified into multiple phases (which will
generally be the case), these phases will melt sequentially as the
infiltrant supply heats up. Thus the first liquid available to
enter the skeleton will not have the average composition of the
infiltrant supply. These problems can be avoided by first creating
a melt of infiltrant and allowing it to equilibrate thermally and
chemically before putting it in contact with the skeleton to be
infiltrated.
[0080] Further, it is advantageous to preheat the skeleton. If the
skeleton is not preheated, the infiltrant will heat up the exterior
of the skeleton where they contact each other and the infiltrant
will begin to penetrate. The rate of penetration will be limited by
the heating of the skeleton rather than just fluid mechanics, since
the liquid would be unable to flow into colder areas of the part.
In these initially penetrated regions, diffusion will begin upon
contact and the pores may become choked off and prevent subsequent
flow. Since the infiltration of the skeleton as a whole is limited
by the need to heat up the interior of the skeleton, this problem
can be avoided by preheating the entire skeleton.
[0081] Several gating methods have been used to initially separate
the melt from the skeleton, then control the introduction of the
liquid. By "gating," it is meant mechanically separating the
skeleton and the liquid infiltrant supply, and then bringing them
together. The motion of a linear or rotary feedthrough from outside
the furnace can be translated to open a `gate` and introduce the
liquid to the skeleton.
[0082] One gating method is to suspend the skeleton before
infiltration and dip it into a pool of the molten infiltrant.
Either the skeleton can be lowered, or the pool can be raised, or
both, to bring the skeleton and the pool together.
[0083] If the skeleton is too delicate to hang under its own
weight, then a mechanism should be used to allow a gated
infiltration with the part resting in a crucible. It can be
difficult to create a hermetic fluid seal that will hold at the
infiltration temperature, but using a crucible material that is not
wet by the infiltrant makes a seal possible. Two such mechanisms
have been used successfully. The first is a vertical alumina plate
used to separate a rectangular crucible into two halves. The shape
of the plate must match the cross-sectional profile of the
crucible, so a bisque-fired alumina plate was cut and filed to
maintain less than 1 mm gap when fitted to the crucible. This gap
was sufficient to hold a 2 cm deep pool; a deeper pool would
require closer tolerances or filling of any gaps with a coarse
alumina powder. A more elegant solution is to use an alumina tube
with a cleanly cut end to sit vertically, with the end flush with
the bottom of the crucible. The infiltrant supply is placed inside
the tube and the melt is contained until the tube is lifted from
above.
[0084] Several other methods can be used for gating the
infiltration. One method involves a custom crucible that has a hole
at the bottom. This hole is plugged with a ceramic rod to prevent
infiltrant flow until the rod is removed. The infiltrant flows
through the hole into another vessel, below, that holds the
skeleton. Another method is to tip a container of infiltrant
supply, allowing the liquid to flow out of the container. Further,
the vessel used to contain the infiltrant supply can be flexible. A
woven cloth of alumina fibers has been used to contain liquid
metal. Such a cloth bag can be used to contain the melt and then
opened up to allow the liquid to flow out.
[0085] The actuation of any type of gate requires a linear or
rotary motion actuator passing through the gas-tight shell of the
furnace. In the case of nickel parts fired in a forming-gas
atmosphere, the feedthrough can be a rod sliding through a slightly
oversized hole in the shell. If the internal pressure in the
furnace is maintained to several inches of water, the leak will not
allow air into the furnace to contaminate the atmosphere. In
applications where atmosphere purity is more critical, several
linear and rotary motion feedthroughs designed for high vacuum
applications are available commercially.
[0086] Powder Size and Size Distribution
[0087] The choice of powder size, defined for spherical powder by
the diameter, has a substantial effect on the depth of penetration
of infiltrant into the skeleton. For simplicity of discussion, the
case of particles which are spherical and which are substantially
mono-modal will be considered. This means that a given skeleton is
made of spherical particles that are all approximately the same
size. Initially, it will be assumed that the particles have smooth
surface texture. (Particles with non-smooth surface texture are
discussed below). Four physical phenomena are influenced by the
particle size:
[0088] 1. The capillary pressure developed by the infiltrant
increases as particle size decreases. The capillary pressure is the
pressure developed across the interface between the ambient gas and
the liquid infiltrant due to the curvature of the surface of liquid
infiltrant between the skeleton particles. It is the capillary
pressure that causes the infiltrant to be wicked into the skeleton.
Thus, other things being equal, the higher the capillary pressure,
the faster the infiltrant will wick into the infiltrant. If the
infiltrant wicks in quickly, it can penetrate farther before any
choking off of the pores due to diffusional solidification. An
expression for the capillary pressure may be developed by applying
the Laplace equation to the meniscus between the particles.
Alternatively, the capillary pressure .DELTA.p may be expressed as
a function of the surface area per unit volume of a powder bed as
follows [G. Scherer, "Theory of Drying," J. Am. Ceram. Soc., 73,
pp. 3-14 (1990).]: 1 p = L V cos ( ) S p V p ( 1 )
[0089] where .gamma..sub.LV is the liquid/vapor interfacial energy,
.theta. is the contact angle of the liquid with the solid, S.sub.p
is the surface area of the pore space and V.sub.p is the volume of
the pore space. For mono-modal spheres it can be shown that 2 S p V
p = 6 ( 1 - ) D ,
[0090] where .epsilon. is the void fraction and D is the powder
diameter.
[0091] 2. The maximum height to which infiltrant can rise in the
skeleton, (in the absence of diffusional solidification and choking
off of the pores) increases as particle size decreases. This effect
is due to the increase in capillary pressure with decreasing
particle size. As the infiltrant rises up the skeleton, the
capillary pressure must be sufficient to overcome the pressure due
to the static head of the liquid metal in the skeleton. This static
head is related to the density of the liquid .rho., acceleration of
gravity g, and height h above the free liquid surface as follows
[James A. Fay. Introduction to Fluid Mechanics. MIT Press:
Cambridge, Mass. 1994]:
P.sub.gravitational head=.rho.gh (2)
[0092] The maximum possible height of this liquid is attained when
the gravitational head (Eq. 2) is equal to the capillary pressure
(Eq. 1).
[0093] 3. Increasing the size of the particles leads to larger pore
spaces between them and a reduction of the effect of viscous drag
on the flowing infiltrant. Darcy's Law describes how the pressure
gradient in a porous medium is directly proportional to the
volume-averaged velocity of the fluid: 3 p = - K V ( 3 )
[0094] where .mu. is the fluid viscosity, and K is the permeability
of the medium [Fay].
[0095] For the case of mono-modal smooth spherical powder, the
permeability of a powder bed can be predicted by the Carman-Kozeny
relation [Phillip C. Carman. Flow of gases through porous media.
Butterworths:London. 1956.]: 4 K = 3 5 S 2 ( 1 - ) 2 ( 4 )
[0096] where .epsilon. is the void pore fraction of the powder bed
and S is the specific surface area, which is equal to 6/D for
mono-modal spheres.
[0097] Thus, other things being equal, it is believed that a
powderbed with a higher permeability will allow the liquid
infiltrant to penetrate faster and therefore penetrate farther
before the pores are choked off by diffusional solidification.
[0098] 4. Increasing the size of the powder reduces the surface
area of the powder per unit volume of the skeleton. The diffusion
of the melting point depressant occurs through the surface and
therefore reducing the surface area in turn slows down the
diffusional solidification and allows for a greater infiltrant
penetration distance before the pores are choked off. Similarly,
larger powder requires the MPD to diffuse over a longer distance to
reach the interior volume of each particle. Thus, other things
being equal, it is believed that increasing the size of the powder
results in a longer time available for infiltration before
diffusional solidification chokes off the pores and therefore,
greater penetration.
[0099] It is important in many cases to attain greater penetration
of the infiltrant before choke-off occurs. Thus, the powder size is
a very important variable. If the skeleton is made from very fine
powder (for example metal powder of 20 microns and smaller, down
to, for instance, even 1-3 microns), the capillary pressure will be
high and the maximum height to which infiltrant can rise will be
high. However, the viscous drag of the penetrating infiltrant and
the surface area available for diffusion leading to isothermal
solidification will also be high. Because decreasing the powder
contributes toward the greater penetration distance in two ways and
also detracts from greater penetration distance in two ways, the
details of the relationships must be examined to gain guidance
about the choice of particle size that will maximize penetration
distance.
[0100] To understand the relationships, a relatively simple system
is considered first. In this simple system no diffusion, and
therefore, no diffusional solidification, takes place. Further, the
melt is penetrating horizontally relative to a vertical
gravitational field. In such a case, only two of the four factors
above are operative--the change in capillary pressure with particle
size and the change in viscous drag with particle size. While these
factors act in opposite directions, the viscous drag is sensitive
to particle size squared, while the capillary pressure is only
directly proportional to particle size, as can be seen from
equations 1, 3 and 4. In other words, as particle size increases,
the viscous drag drops off faster than the capillary pressure. The
result is that the penetrating liquid moves faster through the
skeleton as the particle size increases.
[0101] The next case to consider is one where diffusion and
diffusional solidification take place also. The only one of the
four effects listed above which is not at play is the need of the
infiltrant to rise against gravity. An increase in powder size will
even more strongly favor penetration, because increased powder size
reduces the rate of diffusion of the melting point depressant into
the powder.
[0102] Only when one considers an infiltration that is proceeding
vertically (against gravity), does one see an effect that limits
the effectiveness of increased particle size in attainment of
greater penetration distance. As the particle size is increased,
the maximum height to which infiltrant can rise, decreases. Thus,
as particle size is increased, the height attained by the
infiltrant will increase only up to this limit imposed by
capillarity and gravity. In fact, as this limiting height is
approached, the infiltration will proceed ever more slowly, as
there will be little pressure remaining to drive the flow.
Diffusion and diffusional solidification will have more time to act
and thus, it will be difficult to ever attain the full value of
this limiting height.
[0103] A discussion of a method of designing a process to
manufacture a part follows. The designer is typically faced with
infiltrating a body of a height specified by the design project. In
such a case, the designer would first choose a relatively small
particle size (to attain the best surface finish possible) and
increase the size of the particles as needed to gain infiltration
throughout the body and up to the top (if proceeding vertically
against gravity) of the designed part. However, if the body is too
tall for infiltration it may not be possible to pick a particle
size large enough, because the limitation imposed by the
gravitational head may be reached before reaching the top of the
part. The combination of equation 1 and 2 predicts the maximum
capillary rise height. For example, for liquid metal, with a
surface tension of .about.1 N/m, density of 8 g/cc, and a 60% dense
skeleton of 250 micron diameter powder, the rise height would be: 5
h = L V cos ( ) 6 ( 1 - ) D g = 0.45 meters . ( 5 )
[0104] The discussion above has been in the context of
substantially mono-modal powders. In a bimodal powder, where fine
powder is added and used to fill the interstices between the larger
powder, the fine powder increases the capillary pressure, but it
also very substantially increases the viscous drag and results in a
decrease in the infiltration speed of the molten infiltrant.
[0105] Surface Area
[0106] A further method to attain greater penetration distance of
the molten infiltrant before diffusional solidification is to
increase the surface area of the powder, but without changing its
basic size. FIG. 10 shows schematically a powder particle 330 which
has a texture on the surface resulting in increased surface area.
By such means, it is possible to increase the surface area of a
powder particle by a factor of two or more. The capillary pressure
is related to the surface area per unit volume. Thus such texturing
will increase the capillary pressure proportional to the increase
in surface area because the volume remains approximately the same.
Further, such texturing has only minimal effect on the size and
shape of the pore spaces between the particles and thus has minimal
effect on the viscous drag of the infiltrant through the skeleton
(although the roughness does very slightly increase the drag).
Following the reasoning above, the penetration of a non-diffusing
infiltrant is faster in a skeleton made with powder with non-smooth
surface texture, because the capillary pressure increases much
faster than the small increase in viscous drag. The increase in
surface area will, however, lead to an increase in the rate of
diffusion in the case of a melt with a diffusing species. However,
this increase in diffusion will apply only at the initial contact
between the melt and the powder, because the initial solidification
will tend to smooth out the powder particle, as shown in FIG. 11.
The solid/liquid interface is moving in the direction of the arrow
marked A. The initial surface 332 has relatively sharp indentations
and greater surface area, as compared to subsequently formed
surfaces 334, 336. Thus, the net effect of surface texture on the
penetration of a melt with a diffusing species is beneficial--that
is, greater penetration distance before diffusional
solidification.
[0107] FIG. 12 is a digital image that shows a cross section
through nickel powder particles made by hydrometallurgical
processing. This process results in some degree of surface
texturing of the type desired. For the particle illustrated, the
increase in surface area over a spherical particle is only about 25
percent. Changes in the deposition parameters may result in a more
accentuated surface area. A method to achieve a surface texture
similar to that shown in FIG. 10 is to coat large powder particles
with a single layer of much finer powder (50:1 powder diameter
ratio shown in the figure) and to sinter that finer powder
particles into place. In general, the coated powder is between 10
and 1000 times the size of the coating powder, and preferably
between 20 and 200 times the size. For example, 200 micron nickel
powder is coated with 2 micron nickel powder. In principle, an
increase of a factor of five in surface area is possible using such
a technique. Alternatively, etching techniques can be used to
create surface textures. One such technique is vapor-phase etching.
This would tend to create grooving along the grain boundaries and
other crystallographic defects in the powder.
[0108] Fluid Supply Tabs
[0109] To fill skeletons with dimensions larger than the
penetration distance limit due to freezing, other techniques are
required. Variation of the entry point of the liquid infiltrant can
be used to alleviate some of these problems. FIG. 13 shows a
skeleton 370 and an infiltrant reservoir 372. Infiltrant can be
supplied to multiple areas 374, 376, 378, 380, 382 of the part 370
rather than just the bottom surface. External fluid supply tabs
384, 386, 388, 390, 392 can bring liquid infiltrant to any area of
the skeleton's surface. This reduces the limitation on a maximum
dimension to a less stringent limitation of maximum part thickness.
FIG. 13 shows the tabs supplying infiltrant with the aid of
gravity, in which case they could be hollow tubes allowing the
infiltrant to easily flow through them. However, they can also be
arranged to provide fluid from a melt pool underneath the part. In
this case, they would need to have a porosity suitable to draw
infiltrant up to the required height by capillarity. The tabs could
be inert relative to the infiltrant to prevent any change in
composition or undesirable closing off of pores. The porosity of
the tabs would be relatively coarse as compared to the skeleton to
permit liquid infiltrant to travel through quickly.
[0110] Feeder Channels
[0111] Also, as shown in FIG. 14, solid freeform fabrication
technologies used to create the powder skeleton can create internal
feeder channels 360, 362, 364, 366 to carry the liquid to remote
areas of the skeleton 470. Such channels are considerably larger
diameter (by a factor of 5 or more, preferably between 5 and 10)
than the pore sizes and allow the liquid infiltrant to flow through
the feeder channels quickly without freezing. Indeed, for some SFF
processes, the size of such channels would need to be at least
three times larger than the powder diameter to facilitate powder
removal during fabrication of the skeleton. A network of such
channels can be designed into a part of complex geometry and
function as major arteries to supply liquid infiltrant to the
extremities. A relatively simple example is shown in FIG. 14, but
the channel geometry could be much more sophisticated if necessary.
The feeder channels can have a uniform cross section 362, or
varying 360 (for instance being larger nearer to the infiltrant
supply contact surface than farther from it). The feeder channels
can be vertical, horizontal, inclined, interconnected, or
independent.
[0112] In the case of relying on capillary forces to fill the
feeder channels, their size must be small enough to reach
sufficient rise height, given by the following equation [Fay]: 6
Pressure = gh = 2 cos ( ) r ( 6 )
[0113] with the definition of variables from equations 1 and 2, and
r is the radius of the channel. For a typical liquid metal surface
tension of 1 N/m, a 1 mm diameter channel would provide 4 kPa
capillary pressure. For liquid Ni of density 8 g/cc this would
correspond to a rise height of 5 cm. Channels 360 can be made with
variable diameter, starting larger at the bottom and decreasing in
size at the top to facilitate greater capillary rise. Note that
this would be a limitation on the height, but not on distance;
horizontal sections 366 would result in no loss of head.
[0114] Feeder channels can prove useful for overcoming a short
penetration distance limit when small powder, such as 20 micron, is
used. Small powder is more likely to have a short penetration
distance limited by freeze-off of the infiltrant. For instance, if
the penetration limit for a skeleton of 20 micron powder were only
2 cm, a network of internal feeder channels can be designed into a
5 or 10 cm part such that no section is more than 2 cm from a
feeder channel supplying liquid infiltrant. In general, feeder
channels can be arranged so that no region of the skeleton is
spaced from a feeder channel more than the penetration limit.
Infiltrant will pass from the feeder channels to the body of the
skeleton through the walls of the feeder channels along essentially
their entire length. The composition in the solidified feeder
channels would match that of the infiltrant rather than the bulk
composition of the homogenized part.
[0115] Skeletons with Fine Surface Texture Relative to Interior
[0116] A large penetration distance may be attained together with
good surface finish in another manner shown schematically in FIG.
15. First, create a skeleton 570 out of a powder 530 that is large
enough to allow infiltration up to the desired height without
choking off the pore space. Next, apply a paste 520 of fine
metallic powder 522 with a particle size significantly smaller than
the size of the particles 530 constituting the skeleton 570. The
paste is then applied to the surface of the skeleton to create an
outer layer that has a surface finish superior to the original
skeleton, this outer layer is referred to as the covering layer.
The paste may be made with polymeric vehicles as thickeners and
binders and may be formulated to have a solids loading of typically
20-50% by volume metal powder. The skeleton 570 with the paste
applied is then fired to burn out any polymer in the paste 520 and
to sinter the fine powders in place. The skeleton with fine outer
covering layer is now infiltrated according to the manner of this
invention described above. The liquid infiltrant penetrates through
the main body of the skeleton traveling rapidly through the large
particle core. The infiltration slows appreciably as the infiltrant
reaches the covering layer 520 of the fine powder paste. However,
the infiltrant will only have to penetrate a small distance through
this layer of fine material and thus it will not choke off due to
diffusional solidification. The thickness of the covering layer
must be less than the penetration distance limit due to diffusional
solidification, but this constraint is easily satisfied because the
typical layer thickness is less than one diameter of the larger
particles. The fine powder could also be applied during the
fabrication of the part by selective deposition of slurry during
the SFF process. Such slurry deposition processes are described in
PCT/US98/12280, JETTING LAYERS OF POWDER AND THE FORMATION OF FINE
POWDER BEDS THEREBY, filed Jun. 12, 1998, published Dec. 17, 1998,
which is incorporated fully herein by reference.
[0117] The size of the powder 522 in the paste 520 should be
between approximately {fraction (1/100)} to {fraction (1/10)} the
size of the powder 530 in the main body of the skeleton. Thus, if
200 micron powder is used for the skeleton, the paste should
contain particles in the size range of 2-20 microns. The particles
in the paste may be all of approximately one size, or might span a
range of sizes.
[0118] FIG. 15 shows two approaches to application of the paste 520
to the skeleton 570. 1) The paste may be applied to the surface of
the skeleton to create a skin 524 of finer powder over the surface
(as shown on the left side of figure). 2) The paste can be designed
to penetrate into the pore spaces 526 and to smooth the surface by
filling in the space between the larger powder particles 530, but
not result in a layer on top of the larger particles (as shown on
the right side of figure). The second approach 526 has the
advantage of accurately maintaining the geometry of the original
component. However, if the composition of the fine powder in the
paste is the same as that of the large powder, the composition of
this region after infiltration and homogenization will be different
than that of the interior of the skeleton. This is because the
packing density of the larger powder with the additional fine
powder in the pore will exceed that of the original skeleton. This
is an advantage of the first approach, in that the packing density
of the applied layer will be approximately the same as that of the
bulk of the skeleton. However, the final composition using the
second approach can be made the same by using a fine powder in the
paste that has a composition different from that of the large
powder. The composition of the finer powder would actually have to
match that of the infiltrant, but some combination of the two
approaches (skin over the surface and penetration of the paste)
along with a carefully selected fine powder composition could
provide a desirable result. It may also be desirable to alter the
properties of the part near the surface, which could be done
through appropriate material selection for the fine powder. For
instance, high surface hardness can minimize wear due to friction
and a material with higher hardness could be selected for the fine
powder.
[0119] Maintaining Part Shape
[0120] Since infiltration is accomplished at temperatures close to
the melting point or solidus temperature of the skeleton, the
mechanical strength of the skeleton at the infiltration temperature
might be very low. Part distortion has been encountered when
suspending odd shaped parts above the melt. Distortion can happen
during the high temperature sintering, prior to infiltration. A
first step in minimizing part distortion can be achieved either
through changing the shape of the part or by supporting the part
from beneath rather than suspending it. FIGS. 16A and 16B show how
a large part that underwent distortion while hanging (16A) (note
holes for suspension support) experienced little or no distortion
while resting on the floor of a crucible (16B). For intricate part
shapes, simple floor support may not suffice. A loose ceramic
powder can be filled around the metal part to support parts with
intricate geometry. The infiltration can occur even while the part
is embedded in ceramic, because the ceramic powder is typically not
wet by the infiltrant, and thus the infiltrant will not enter those
regions.
[0121] Material Systems
[0122] Selection of appropriate material systems involves the
choice of skeleton material and MPD, with consideration for the
degree of infiltrant melting temperature depression, diffusivity
and solubility of the MPD in the skeleton material, and the desired
final material composition.
[0123] The inventors have conducted extensive experimental work
involving the binary Ni--Si material system, using a skeleton
material of pure nickel and an infiltrant of .about.90% Ni with the
addition of .about.10% Si. The specific amount of silicon used
depends on the infiltration temperature. Additions of other
alloying elements to this binary system can provide different, and
for some applications, more desirable mechanical properties. Such
possible alloying elements include, but are not limited to Chromium
(Cr), Iron (Fe), Cobalt (Co), and Molybdenum (Mo). Several
commercial alloys such as Inconel 617, HX, and G3 contain a
combination of those alloying elements along with 1% Si. For
example, chromium added to the Ni--Si system acts as a solid
solution strengthening element. (Commercial nickel brazing alloys
containing silicon typically contain 20% chromium for this
reason.)
[0124] Other possible melting point depressants for nickel-based
alloys include boron (B), phosphorous (P), and tin (Sn). Boron and
phosphorous are used extensively in commercial brazing alloys. They
both have very low solubility, and would result in a two-phase
final part composition. Tin has a fairly high solubility that would
enable homogenization. Antimony (Sb) and sulfur (S) also provide
deep eutectics with nickel. Addition of large quantities of copper
(Cu) can significantly depress the melting point of nickel. As an
extreme case, a nickel skeleton infiltrated with pure copper would
also undergo diffusional solidification, due to the complete
solubility of the two elements with each other.
[0125] Aluminum (Al) offers many potential melting point
depressants. Table 1 summarizes the effect of several alloying
elements commonly used in aluminum. Pure aluminum has a melting
point temperature of .about.660.degree. C. Copper and magnesium
(Mg) are typically used to provide strengthening at small
concentrations. Silicon is used extensively in die casting alloys
to improve fluidity of the melt. Matching existing commercial
die-casting alloy concentrations would be useful, and can be done
as an aspect of an invention disclosed herein. Ternary and
quaternary alloys can provide additional melting point depression.
For example, an aluminum alloy commonly used in die casting of
automotive pistons (336.0) contains 12Si-2.5Ni-1Mg-1Cu, has a
solidus of 540.degree. C. and a liquidus of 565.degree. C.
1TABLE 1 Effect of various alloying elements on the melting point
of aluminum. Alloying Element Eutectic Melting Point Melting point
in Aluminum Comp (wt %) (.degree. C.) depression (.degree. C.)
Silicon (Si) 12 577 83 Magnesium (Mg) 35 450 210 Copper (Cu) 30 548
112 Germanium (Ge) 50 420 240 Lithium (Li) 8 596 64
[0126] One challenge for aluminum is to conduct the transient
liquid-phase infiltration within a small temperature window, since
the melting point depression of the infiltrant may be less than
100.degree. C. Fortunately, the lower operating temperature of
aluminum allows for easier manipulation of the part and the
melt.
[0127] The diffusivity of silicon in aluminum is .about.10.sup.-12
m.sup.2/s at 600.degree. C., which is about one order of magnitude
higher than that of silicon in nickel at 1200.degree. C. The
diffusion distance is only affected by the square root of the
diffusivity, but this still presents more of a challenge in
achieving large infiltration depth as compared to a Ni--Si system.
Factors such as grain size and the presence of other species can
influence the diffusivity. Addition of iron to the infiltrant may
be used to slow mass transport, because of a high affinity of iron
for ordering with silicon and a lack of solubility of iron in
aluminum. Copper acts as an excellent barrier to Si diffusion and
can be electroplated on aluminum powder. The diffusivity of Cu in
Al is similar to that of Si, so the coating will not last long at
the infiltration temperature, but a higher concentration of Cu at
the surface of the powder could still slow the mass transport
appreciably. The lower solubility of silicon in aluminum (as
compared to in Ni) will typically result in the liquid flow never
being choked off due to solidification. This is because the part
will only undergo partial diffusional solidification at the
infiltration temperature if the MPD final bulk composition is
greater than the MPD solidus composition.
[0128] Another challenge of processing aluminum alloys derives from
the natural formation of a thin surface layer of aluminum oxide,
potentially preventing wetting of the infiltrant, and having other
detrimental effects. The oxide grows faster at higher temperature.
Thus, minimizing the time the skeleton is exposed to elevated
temperature is beneficial and can be done through fast temperature
ramp rates and short dwell times. The furnace atmosphere can also
be controlled to slow the oxidation process. Using flux can also
help break down the oxide layer. Specific flux materials include,
but are not limited to boric acid or others commonly used in the
aluminum welding and soldering industries. Adding small amounts of
magnesium also has a beneficial effect at breaking up the surface
layers of aluminum oxide. Using detergents or wetting aids to allow
the molten infiltrant to wet the oxide layer would also facilitate
infiltration.
[0129] It should be understood that in the claims appended hereto,
if the transitional phrase "consisting essentially" is used, the
inventors intend the claim to read on a composition that has the
materials identified in the claim and also small amounts of flux,
detergent, wetting agent, or magnesium, or other similar materials,
which small amounts do not adversely affect the depression of the
melting point.
[0130] An approach for infiltrating steel skeletons involves using
multiple alloying elements to achieve the melting point depression
of the infiltrant. The ternary, quaternary, and greater complexity
alloys can provide significantly more depression of the melting
point than is achieved through any of the individual binary alloys.
Further, the concentration of alloying elements in the infiltrant
can be more than double that of the desired final composition,
because the infiltrant fills less than half of the total part
volume.
[0131] Table 2 below shows the melting range (characterized by the
liquidus and solidus temperatures) and composition of two common
stainless steels, 316 and 17-4 PH, along with the melting range of
an infiltrant that may be used to reach that standard
composition.
2TABLE 2 Melting ranges of 316 and 17-4 PH stainless steels and
potential infiltrants. Liquidus Solidus C Mn Si Cr Ni Mo Cu Nb
Material (.degree. C.) (.degree. C.) (%) (%) (%) (%) (%) (%) (%)
(%) 316 1400 1339 0.08 2 1 17 12 2.5 Infiltrant 1292 1135 0.2 5 2.5
17 30 6.25 17-4 PH 1406 1237 0.07 1 1 16.5 4 4 0.3 Infiltrant 1298
1205 0.175 2.5 2.5 16.5 10 10 0.75
[0132] For these cases, the skeleton is composed of pure iron with
a melting point of 1538.degree. C. or iron and chromium with a
similar melting point. The infiltrant contains all of the necessary
alloying elements for the final composition to match that of the
standard stainless steel. For a 60% dense skeleton, this requires
the contribution of each alloying element to the infiltrant
composition to be 2.5 times greater than the desired final content.
Chromium does not have a significant impact on the melting point.
Thus, its concentration can be kept the same in the skeleton and in
the infiltrant. The processing window for the infiltration is over
200.degree. C., and the diffusivities of Ni, Mn, and Cu in iron at
1300.degree. C. are all approximately 10.sup.-14 m.sup.2/s, which
is slow enough to allow infiltration before freezing. (The liquidus
and solidus information presented in Table 2 was calculated using
Thermo-Calc, a Computational Thermodynamics program used to perform
calculations of thermodynamic properties of multi-component systems
based on the Kaufman binary thermodynamic database.)
[0133] The infiltrant liquidus temperature dictates the minimum
infiltration temperature. In the case of 316 stainless steel, this
infiltration temperature (1292.degree. C.) lies below that of the
bulk material solidus (1339.degree. C.). This means that the
material will undergo complete diffusional solidification at the
infiltration temperature and the liquid flow will be choked off in
a time period determined by the solidification rate. If the
infiltrant liquidus is above that of the bulk material solidus, as
is the case with 17-4PH steel, then the final part composition will
lie in a two-phase field and liquid will always be present at the
infiltration temperature. These two conditions are analogous to the
previous distinction made between skeleton material systems of low
and high solubility of MPD.
[0134] Titanium alloys have important uses in high temperature
applications where high specific stiffness and strength are
required. The binary Ti--Si phase diagram shown in FIG. 17 shows
very similar characteristics to the Ni--Si system discussed in
detail above. Although the processing of Ti parts is more
challenging, the materials behave similarly. Other alloying
elements that are common in commercial Ti alloys include but are
not limited to Al, Sn, Zr, Mo, V, Cu and Cr. Copper has a fairly
significant impact on the melting point, reaching a eutectic
temperature of 1005.degree. C. at a composition of 45 wt %.
Chromium and zirconium also work as melting point depressants in
titanium, although to a lesser degree.
[0135] Copper-based material systems are also good candidates for
the infiltration of a higher melting temperature skeleton with a
similar material used as an infiltrant. Potential melting point
depressants that can be found in cast copper alloys are Ag, Mg, Mn,
Si, Sn, and Ti.
[0136] Many techniques and aspects of the inventions have been
described herein. The person skilled in the art will understand
that many of these techniques can be used with other disclosed
techniques, even if they have not been described as being used
together. Thus, the fact that a subcombination of features that are
described separately, may not be described in subcombination, does
not mean that the inventors do not regard any such subcombination
as an invention that is disclosed herein.
[0137] For instance, any of the following techniques and features
can be used with any of the others: skeleton with feeder channels;
skeleton with a relatively coarse inner powder, with surfaces
covered with a paste formed from relatively finer powder; liquid
infiltrant supply tabs to introduce liquid to the skeleton at
multiple locations; providing the infiltrant supply at a desired
liquidus composition to facilitate uniform bulk composition along
the path of infiltration; agitating the infiltrant supply to insure
that it remains at such a desired composition; using an infiltrant
with a melting point depressant that diffuses into the skeleton
material, thereby tending to homogenize the composition of the
finished part, in some cases (high solubility) completely, and in
other cases (lower solubility) to a degree similar to cast
products; choosing powder size to insure infiltration to the full
extents of the skeleton in consideration of a penetration distance
limit imposed due to diffusional solidification, with relatively
larger particle sizes allowing greater penetration distance before
freezing, and relatively smaller particle sizes having a greater
capillary rise limit in the absence of freezing; choosing powder
surface area (roughness) to achieve penetration to the desired
extent, with relatively rougher surface area particles providing
greater capillary driving force, faster and thus, deeper
penetration than relatively smoother surface area particles, other
factors being equal; choosing material systems with MPD that will
diffuse within the skeleton material to a degree necessary to
achieve homogenization of bulk properties, and, if possible,
composition, but at a rate that is slow enough to permit full
infiltration of the skeleton before diffusional solidification (if
any) occurs to a degree sufficient to choke off flow of liquid
infiltrant into the skeleton. Any of these general principles and
techniques can be applied to any of the material systems disclosed,
or hereinafter developed.
[0138] Some of the inventions disclosed herein are methods of
fabricating metal parts. However, other inventions disclosed herein
are methods of designing processes for fabricating such metal
parts. In other words, the process design inventions are methods
for designing manufacturing processes. For instance, it is
disclosed how a designer, challenged with the task of fabricating a
metal part of a specified shape, and specified basic metal (e.g., a
predominantly nickel part, or a predominantly aluminum part) will
proceed to design the process to make the part. The disclosure
herein teaches how the designer shall select a powder composition
including the base metal and alloying elements, and also how to
select an infiltrant, composed of the metal of the powder, and
melting point depressant agents. (The skeleton metal can also
include some smaller amount of these MPD agents.) The disclosure
also teaches that a relatively small powder size should be first
considered if smooth surface finish is desired, and then if such
size is too small to permit full infiltration due to penetration
distance limits, instead a larger particle size must be selected.
The designer is also informed by this disclosure of the effects of
particle size, surface roughness, density, viscosity, and myriad
other factors that can be considered in the selection of the
materials of the skeleton and MPD. Additional mechanical techniques
are disclosed to overcome, or minimize the effect of material based
infiltration penetration limits. These mechanical techniques
include but are not limited to: feeder channels, fluid supply tabs,
covering a relatively coarse skeleton with a paste of finer
particle, and using particles with a rough surface. Thus, the
designer is taught how to achieve infiltration penetration distance
greater than would be achieved in a comparison system, without the
enhancement, for instance, feeder channels, rougher particle
surface, or relatively fine powder surrounding a skeleton of
relatively coarser powder.
[0139] Further, the disclosure teaches how to achieve various
degrees of uniformity in composition, including substantially fully
homogeneous, homogeneous along the direction of infiltration, and
non-homogeneous, but similar in microstructure to cast products,
resulting in essentially homogeneous properties. These teachings
are based on maintaining the infiltrant at a liquidus composition,
and more subtle selection criteria related to the ratios of
components in the infiltrant as diffusional solidification takes
place in ternary and higher infiltrant systems, facilitated by
resort to equilibrium phase diagrams. All of these tools relate to
the inventions of designing a process of fabricating a metal
part.
[0140] This disclosure describes and discloses more than one
invention. The inventions are set forth in the claims of this and
related documents, not only as filed, but also as developed during
prosecution of any patent application based on this disclosure. The
inventors intend to claim the various inventions to the limits
permitted by the prior art, as it is subsequently determined to be.
No feature described herein is essential to each invention
disclosed herein. Thus, the inventors intend that no features
described herein, but not claimed in any particular claim of any
patent based on this disclosure, should be incorporated into any
such claim.
[0141] An abstract is submitted herewith. It is emphasized that
this abstract is being provided to comply with the rule requiring
an abstract that will allow examiners and other searchers to
quickly ascertain the subject matter of the technical disclosure.
It is submitted with the understanding that it will not be used to
interpret or limit the scope or meaning of the claims, as promised
by the Patent Office's rule.
[0142] This is being filed of even date with a patent application
under the Patent Cooperation Treaty, designating The United States
of America, in the names of the same inventors (Sachs, Lorenz and
Allen), entitled INFILTRATION OF A NET SHAPE POWDER METAL SKELETON
BY A SIMILAR ALLOY WITH MELTING POINT DEPRESSED TO CREATE A
HOMOGENEOUS FINAL PART, Attorney Docket No. MIT 8873 PCT, being
filed under Ex. Mail Label No. EL662947541US the full disclosure,
of which is incorporated fully herein by reference, including the
specification, claims and figures.
[0143] The foregoing discussion should be understood as
illustrative and should not be considered to be limiting in any
sense. While the inventions have been particularly shown and
described with references to preferred embodiments thereof, it will
be understood by those skilled in the art that various changes in
form and details may be made therein without departing from the
spirit and scope of the inventions as defined by the claims.
[0144] The corresponding structures, materials, acts and
equivalents of all means or step plus function elements in the
claims below are intended to include any structure, material, or
acts for performing the functions in combination with other claimed
elements as specifically claimed.
* * * * *