Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Chapter 1
INTRODUCTION
OVERVIEWWHAT IS MIM? the same as with other metalworking processes. Metal powder
Metal powder injection molding (MIM) is a shaping injection molding succeeds as a net-shaping approach, mean-
process focused on forming and sintering small metal pow- ing there is little material loss, and in several situations 98%
ders. It is widely used in the fabrication of complex-shaped, of the purchased raw metal powder ends up in final products.
high-performance components in large production quantities. Further, MIM enables shape complexity in materials that are
The Venn diagram shown in Figure 1.1 illustrates how the often difficult to fabricate by other processes, such as com-
convergence of four main considerations makes for the MIM posites (W-Cu), electronic alloys (Fe-Ni, Fe-Co), hard mate-
sweet spot. An example component is pictured in Figure 1.2, rials (tool steel), titanium, stainless steels, nickel-base
showing a stainless steel cellphone hinge formed by MIM superalloys, and tungsten.
cost, performance, complexity, and production quantity are all This guide is for engineers seeking to design components
evident in this application. that intersect with the MIM technology. The following chap-
The concept of using plastic-forming equipment to shape ters provide information on the process, design rules, materi-
a metal, such as surgical stainless steel, excites most design als, properties, applications, costs, and markets. The MIM
engineers, especially when they learn the final properties are design process combines decisions on the material, compo-
nent geometry, component function, and fabrication process.
It is important to balance economic and technical considera-
tions; hence, beyond engineering specifications this book
1
Table of Contents
Chapter 1: Introduction
provides information on the cost aspects. Further, winning thermoplastic binders. By the mid-70s metallic products were
technicaleconomic successes are used throughout the book offered for commercial use by a few firms. However, up to
to illustrate the various concepts. This information is useful the 90s, MIM lacked the required infrastructure for broad
in rationalizing the process with the applications. success. As the field grew large, several users stepped in to
So, to the question of this sectionWhat is MIM? Metal ensure proper standards, providing the quality gains important
powder injection molding is a net-shaping process that starts to widespread acceptance. Today, MIM is essentially a three-
with a small metal powder that is dispersed in a thermoplastic step process involving:
polymeric binder. Paraffin wax is a common thermoplastic formulation of feedstock from appropriate metal powders
polymer distinguished by the fact that it can be repeatedly and polymers,
melted. Most plastics are thermoplastics, and several are used molding of that feedstock into tooling that is designed
in binder formulationspolyethylene, polypropylene, poly- for the final part and includes dilation of the size in
acetal, and ethylene vinyl acetate are examples. anticipation of sintering shrinkage, and
Most common engineering alloys are possible by MIM, but thermal processing of the shaped part to remove the poly-
about 30 alloys dominate the applications. The most popular mer and sinter the powder.
alloys are surgical stainless steel (commonly called 17-4 PH, The last step might be accomplished in a single cycle or in
or American Iron and Steel Institute 630 or AISI 630) and two separate steps of first debinding to remove the polymer,
austenitic stainless steels (AISI 304L and AISI 316L). After followed by sintering.
injection molding, the polymeric binder is extracted and the To begin the process, MIM mixes a small powder typically
particles are subjected to a high-temperature heat treatment in the size range between 1 and 40 m with a plastic binder.
termed sintering. During sintering the capillary forces between The typical ratio is about 60 v/o powder, requiring then about
the small particles densify the structure to give a shrunken 40 v/o binder. A common binder would be two-thirds paraffin
component that is essentially full density with tensile and wax and one-third polyethylene or polypropylene. Depending
hardness properties equivalent to handbook values. on the powder, the mixture constitutes about 94% powder by
Alloys can be formed by mixing elemental powders to weight, but for low-density powders such as aluminum this is
match the desired alloy, for example, iron, nickel, and 86% powder, and for high-density powders such as tungsten
chromium to form a stainless steel. The powders homogenize this is 97% powder. The binders melt at relatively low tem-
during sintering to deliver the alloy composition and proper- peratures, up to 130C (266F), to allow easy mixing and
ties. Another approach is to use prealloyed powder where each molding. The combination of powder and binder is termed
particle contains all of the elements. Ferrous alloys are popular feedstock, and feedstock production is a significant business
by MIM, and nearly half of the commercial activity is in stain- supporting MIM.
less steels. Complex-shaped, small components (10 g median A MIM process flow chart is given in Figure 1.3. The heart
mass and 25 mm median size) formed by MIM are widely of production is molding, where the heated feedstock is
used in dental orthodontics, portable computers, automotive shaped in a custom-designed tool. Prior to molding, the pow-
engines, surgical instruments, electronic packages, hand tools, derbinder feedstock is heated inside the injection molding
firearms, sporting devices, and cellular telephones. machine to melt the polymer phase. The molding machine
Since small powders are costly, one of the keys to economic rapidly presses the molten feedstock mixture into the mold
success is to apply MIM to more complicated geometries, cavity to avoid any premature freezing. During this molding
those that would require extensive machining. A further gain
comes by adding cores, holes, or other mass-reduction fea-
tures, the opposite to the situation in a machined component.
Thus, MIM excels vs. other fabrication approaches when the
component complexity is high and performance demands re-
quire excellent properties. It wins against casting since it has
better dimensional control and a smoother surface finish. Un-
like machining, MIM is best applied to large production quan-
tities. Accordingly, several situations have reached production
quantities in excess of 10 million parts per year.
PROCESS ESSENTIALS
The clear advantage in MIM is the use of plastic molding
concepts to form structures with properties not possible Figure 1.3. Outline of the key steps in MIMstarting by mixing a
from plastics. This idea has its origins in the 1930s when small metal powder with a polymeric binder, followed by molding,
automotive ceramic spark-plug bodies were fabricated using binder removal, and sintering
2
Table of Contents
Chapter 1: Introduction
step the powder remains unaffected by the molten binder, and machining. It is most competitive when used to form com-
since the peak temperature is far below the powders melting plicated and durable components, especially when the com-
temperature. After injection into the mold cavity, the pressur- ponents require high properties above those associated with
ized feedstock cools to solidify the polymer, thereby freezing die-cast zinc. Further, MIM is best suited for mass production
the particles into the desired shape. Cooling is usually the slow and is commonly applied to production quantities over
step, so molding cycles often are 10 to 20 seconds. After ex- 200,000 parts per year. Recognition of the wide range of
traction from the mold, the polymer portion is removed by applications, materials, component features, and properties
heat, solvents, or catalysts without disturbing the shaped pow- is now helping to identify new uses that are contributing to a
dera process known as debinding. The final portion of the period of sustained application expansion.
polymer decomposes while heating the shaped body to the
sintering temperature. WHERE MIM DIFFERS FROM PM
Sintering is a particle-bonding heat treatment that naturally Sometimes confusion exists about the difference between
densifies packed powders when they are heated to a tempera- metal powder injection molding and traditional presssinter
ture where atomic motion is active. For most powders, sinter- powder metallurgy (PM). The standard powder metallurgy
ing starts well below the melting temperature. Besides
considerable strengthening, one common result of sintering
small powders is component densification; the sintered com-
ponent is smaller than the molded component. Indeed, in MIM
the final component is about 15% smaller in each dimension
due to annihilation of the space initially filled with binder.
Figure 1.4 provides before and after versions of a molded
component to illustrate sintering shrinkage. After sintering
there is no evidence of the initial powders or polymers. Thus,
performance attributes of MIM products are similar to the
properties cited in common engineering handbooks.
Secondary steps after sintering are very much the same as
applied to all metals. The components can be heat treated,
electroplated, drilled and tapped with threads, and welded.
After sintering there is nothing different in the MIM material
compared with that of standard metallurgical treatments; there Figure 1.5. The key forming step in traditional presssinter powder
are no artifacts from the powders. metallurgy is uniaxial die compaction. The powderpolymer mixture
Metal powder injection molding is a widely accepted net- is fed into the tool cavity by gravity and the upper and lower punches
shape production route that competes with investment casting press the powder with a high pressure, which is followed by ejection
3
Table of Contents
Chapter 1: Introduction
forming step, sketched in Figure 1.5, relies on uniaxial (ver- if sintered to full density, may have poor dimensional
tical axis) compaction to press the powder into a green body. uniformity, and
Pressing is a long-established technology for shaping a broad if not sintered to full density, then their properties are
range of materials into low-height parts. It is the mainstay in degraded by pores.
compacting pharmaceuticals pills. The compacted shape is In short, with die compaction the choice is either full
formed using upper and lower punches working on the pow- density giving high properties with poor dimensional control,
der placed inside a laterally constraining die. Die compaction or good dimensional control with lower density and lower
is best at forming squat shapes that are easily ejected from the properties.
tooling. The ejection step requires the sides be parallel, a re- As an illustration, Figure 1.8 compares the scatter in final
striction in geometry not seen in MIM. Thus, undercuts and dimensions for a die-compacted-and-sintered copper compo-
holes perpendicular to the main axis are possible in MIM, as nent and the same component fabricated by molding and sin-
illustrated by the molded component shown in Figure 1.6. tering. The scatter is measured by the standard deviation in
A typical compaction pressure for PM die compaction is dimensions, both as-formed and after sintering. Note that,
near 700 to 840 MPa (100,000 to 120,000 psi). Unfortunately, after sintering, the MIM component has a significantly lower
the proportionately high ejection stresses can crack the weak standard deviation since MIM forming is hydrostatic, thereby
pressed powder during removal from the tooling. A minimized minimizing green-density gradients. Thus, after sintering to
height helps avoid ejection cracks, so as is evident in Figure full density, the MIM process is capable of tighter dimensional
1.7, the shape complexity is all in the pressing direction. Little control, because it avoids the density gradients encountered
shape detail is permitted perpendicular to the pressing direc- in compaction. For die compaction, the standard protocol is
tion. Further, because of friction between the powder and to use larger powders vs. MIM and to sinter at lower temper-
moving tooling, die compaction results in density gradients in atures for shorter times to avoid distortion during sintering
the pressed powder. This contrasts with MIM, which is uni- densification.
form in forming pressure with a lower pressure, meaning In powder metallurgy systems, residual pores dispropor-
smaller pressure and density gradients. Since density gradients tionately reduce the particleparticle bonding. Under stress,
result in distortion during sintering densification, the homo- fracture occurs by cracks jumping between residual pores.
geneous density in MIM gives uniform dimensions after high- MIM products are sintered to near-full density while 85% den-
temperature sintering densification. In die compaction, the sity is typical in PM. Consequently, the properties fall rapidly
typical sintering temperature is lower to avoid distortion that due to a short-circuit failure path associated from pores. As a
accompanies densification. Consequently, components fabri- consequence, the MIM tensile properties are almost twice that
cated from pressed powder: of presssinter powder metallurgy. Differences in component
are sintered at lower temperatures where sintering den- shape complexity, density, and performance means that tradi-
sification is avoided,
are often machined after sintering to correct critical
dimensions,
4
Table of Contents
Chapter 1: Introduction
tional presssinter powder metallurgy and MIM do not com- 1999. Clearly, disproportionate growth took place in the
pete with one another. The performance penalty from pores is higher density range, which matches with MIM.
evident in Figure 1.9, which plots the tensile-strength change A key point in the comparison of MIM with PM is the link-
with sintered density for a steel (Fe-2Ni-0.5Mo-0.3C) fabri- age of performance to microstructure. Figures 1.11 and 1.12
cated by the PM presssinter route (sintered at 1,120C or provide a pore-structure contrast between presssinter 316L
2,050F). At 15% porosity the properties are all less than half stainless steel and MIM 316L stainless steel, which is sintered
the full-density values, and impact energy is only 16% of to near-full density. In light microscopy the pores appear black
full density. For comparison, when fabricated by MIM and since there is no reflection. Because the sintered MIM product
subjected to heat treatment, the tensile strength is 2,000 MPa is nearly free of pores, the performance characteristics
(290 ksi). compete with other metal forming processes. Admittedly, die
Improved properties from higher densities are driving the compaction results in a lower-cost product, so when it can de-
shift to MIM vs. presssinter powder metallurgy. Full density liver the shape and required properties, it is the economical
is increasingly of value, as evident by the trend shown in Fig- choice. As outlined in Table 1.1, PM and MIM only have the
ure 1.10. This plot gives the relative change in product density use of metal powders as a common base, but otherwise are
for presssinter ferrous powder metallurgy between 1989 and very different. Thus, the two technologies are not competitors,
since they have significant differences in powders, processing
steps, properties, shapes, costs, and applications.
Figure 1.10. The relative trend in final sintered density for ferrous
presssinter powder metallurgy components. These data show the Figure 1.12. A typical sintered microstructure for a MIM 316L
progressive move to higher performance increasingly forced the stainless steel with about 2% porosity, a level comparable to
production at higher density levels, an area compatible with MIM castings (pores are black), and a strength comparable to the
(data courtesy of Diran Apelian). handbook value
5
Table of Contents
Chapter 1: Introduction
6
Table of Contents
Chapter 1: Introduction
Figure 1.16. Hexagonal screw holders fabricated for a hand tool Figure 1.17. An example of the shape complexity and
using MIM stainless steel (component courtesy of Advanced perpendicular holes and undercut features possible using MIM
Powder Products)
7
Table of Contents
Chapter 1: Introduction
8
Table of Contents
Chapter 1: Introduction
9
Table of Contents
Chapter 1: Introduction
10
Table of Contents
Chapter 2
PROCESS OPTIONS
PROCESS OVERVIEW terial such as epoxy, which can only be heated once and
Background remains rigid since heat causes the polymer chains to link
Most engineers think of injection molding as a forming together.
technology devoted to plastics, useful for material such as After molding with a thermoplastic binder, the shaped
polyethylene, polystyrene, and glass-filled nylon. On the other powder is sintered to attain the target properties. A schematic
hand, they know that metals such as steel, ceramics such as of the process is given in Figure 2.1. The polymeric binder
alumina, cemented carbides, and other inorganic materials lubricates and bonds the particles during the shaping process.
have several desirable properties when compared with plastics. Subsequently, the binder is extracted in a process termed de-
The metals and ceramics are stronger, tougher, and stiffer, and binding where the polymer is dissolved or decomposed by
offer significant latitude in electrical and thermal conductivity, burning, technically known as pyrolysis. This leaves a porous
magnetic response, and wear behavior, and can operate to very powder structure that looks like the molded shape, but it is
high temperatures. Powder injection molding is a synergistic weak. However, when the powder is heated to a high temper-
combination of the performance properties associated with ature, the particles bond and densify due to capillary attrac-
ceramics and metals and a shaping option that is associated tion, in what is called sintering. There is a tremendous strength
with plasticsinjection molding. High performance, shape gain in sintering, to the point that the sintered solid is essen-
complexity, and low cost are attributes that indicate a good tially indistinguishable in strength from wrought metal. So un-
application of powder injection molding. Indeed, most appli- like in plastic injection molding, where the plastic is the
cations for powder injection molding are in metals, especially object, in MIM the plastic or polymer is only a transient phase,
stainless steelswhat is termed MIM (metal powder injection added to help the metal particles flow and fill out the mold.
molding). The MIM process has many variants, reflecting different
In both MIM and PIM, small powders are molded in the combinations of powders, binders, molding techniques, de-
desired shape using a thermoplastic binder during the shaping binding routes, and sintering furnaces. However, in all variants
stage. A common thermoplastic is paraffin wax (also termed there are three key technology areas:
candle wax). It softens and melts on heating and then freezes feedstock: powders, binders, mixing
on cooling, and repeats this behavior with each heating cycle molding: rheology, tool design, machine operation
without degradation. Compare this with a thermosetting ma- thermal processing: debinding, sintering, heat treatment
11
Table of Contents
Chapter 2: Process Options
FeedstockPowder Plus Binder of a typical stainless steel powder fabricated by gas atomiza-
Feedstock is the mixture of powder and binder used for tion for MIM. It is nearly spherical and exhibits a range of
molding. Figure 2.2 is a scanning electron microscope picture particle sizes that allows good fitting together of the particles.
Actually, there are many ways to form the small particles used
in MIM, ranging from vapor condensation to oxide reduction.
The best choice changes with material and some of the inher-
ent material attributes, such as melting temperature and reac-
tivity, as well as the desired particle size, particle shape, alloy,
purity, and cost. For example, it is very difficult to melt tung-
sten due to its high melting point (3,410C or 6,170F) so hy-
drogen reduction of a milled tungsten oxide is appropriate,
but for copper it is common to form a melt and to spray that
melt into droplets that solidify as particles (atomization).
A few example MIM powders are compared in Table 2.1.
More than 800 different powder variants have been used in
MIM, so the few examples in this table are simply for illus-
tration purposes. The common name is given in the first col-
umn, and the nominal composition is listed next. The third
column gives the fabrication process. (For those seeking
details on the powder fabrication routes, typical powder char-
acteristics, and standard tests to measure the characteristics
Figure 2.1. Process outline for MIM, where the powder and binder
phases are combined by premixing and intense hot mixing, then Figure 2.2. A scanning electron micrograph of a typical gas-
granulated into feedstock. This feedstock is molded, debound, and atomized stainless steel powder used in MIM. These particles are
sintered to give the final shape about 15 micrometers across, but intentionally have a range of sizes
12
Table of Contents
Chapter 2: Process Options
should see the textbook by the author; R. M. German, Powder powderbinder cold-mixed feedstock is fed into a twin-screw
Metallurgy and Particulate Materials Processing, Metal mixer where two threaded shafts rotate inside a barrel to shear,
Powder Industries Federation, Princeton, NJ, 2005.) compress, split, and recombine the feedstock as it moves into
In the powder data, the median particle size corresponds to a final extrusion step. When the mixture is improperly mixed,
the 50% point on the cumulative particle-size distribution and the feedstock inhomogeneities carry over to become molding
it is also termed the D50 or D50 size. At the median size half defects; thus, high shear is required in mixing to force the
of the powder mass corresponds to particles smaller than the binder molecules between all particles without contamination.
stated size and half of the powder mass corresponds to particles In formulating the feedstock, a typical binder content is
larger than the stated size. This is different from the mean or near 40 v/o of the mixture; for steel, that corresponds to about
average particle size, and is different from what would be seen 6 w/o binder, but for a low-density ceramic such as alumina it
in a microscopic picture of the particles. If the number of par- will be 14 w/o, while for a high-density metal such as tungsten
ticles is counted from a micrograph, then it would appear the it will be less than 3 w/o. Feedstock formulation balances sev-
size is smaller than from the particle-mass distribution. It takes eral considerations. Sufficient binder is needed to fill all voids
many small particles to equal the mass of one large particle. between the particles and to further lubricate particle sliding
Hence, invariably mass-based powder statistics skew toward during molding. Too much powder makes the feedstock stiff
the larger particles. In the last column, the apparent density and difficult to mold, but too much binder leads to distortion
tells how the powder packs when poured into a container with- during thermal processing. To find the right ratio of powder
out vibration. For example, copper had a theoretical density of to binder requires a simple mixing experiment. For example,
8.9 g/cm3, but the water-atomized small cooper powder used in the kitchen we can mix baking flour with water or milk, as
in MIM packs to 3.3 g/cm3 or about 37% of theoretical. Often illustrated in Figure 2.3. Flour is a powder and milk acts like
suppliers of metal powders for MIM include more exact details a binder. The mixture is a liquid at low concentrations of flour,
in the data sheet accompanying the powder. and will flow, but as more flour (powder) is added, the mixture
Common binders in MIM are waxpolymer mixtures. becomes viscous like a paste. Too much flour makes the mix-
When heated, these binders provide a low-viscosity glue-like ture a solid that will not flow. As the Goldilocks Principle says,
liquid for transporting the powders into the mold cavity. At there is a balance that is just right.
the end of the molding cycle, the cooled binder freezes to pro- Best molding success is attained when the hot feedstock
vide sufficient strength to resist damage during ejection. A fa- mixture viscosity is near 100 Pas (1,000 poise). As a bench-
vorite binder system is 65% paraffin wax, 30% polypropylene, mark, this is thicker than most paints (which are small parti-
and 5% stearic acid. It is fully molten near 130C (266F), but cles such as titania dispersed in binders) and more like
this temperature changes with the molecular weight of the toothpaste (which also consists of small oxide particles such
polypropylene. Stearic acid is a polar molecule that attaches as silica, calcium phosphate, alumina, or calcium carbonate
to the metal powder, yet is mostly the same chemical structure in a binder formed from cellulose or gum dissolved in water),
as paraffin wax. It helps connect the powder and binder and but not like cold ice cream.
further eases ejection from the mold. Other polymers such as The final step in feedstock preparation is to form granu-
polyethylene, polyacetal, and ethylene vinyl acetate are used, lated pieces such as shown in Figure 2.4. The pellets are
and even cellulose, water, beeswax, carnauba wax, or peanut loaded into a molding machine via the hopper. Premixed feed-
oil are used in some binders. The binder system always con- stock is available from several companies. Accordingly, in
sists of two or more components; the stronger polymer is MIM, two variants exist with respect to feedstock:
called the backbone. The idea is to remove the wax phase of older and larger MIM operations that formulate and
the binder, while leaving the polypropylene to hold the parti- mix their own feedstock
cles in place as the compact is heated to the sintering temper- newer or smaller MIM operations that purchase feed-
ature. All of the backbone is removed during heating before stock from a supplier
the temperature passes over about 550C (1,022F). The situation is comparable to buying premixed cookie
Mixing involves first melting and then dispersing the pow- dough as opposed to making cookies from scratch. There are
derbinder feedstock. It often starts at room temperature to merits to both approaches, but the time, expertise, and capital
mechanically disperse the phases and, as heat is applied, the cost of the mixing equipment often leads to a decision to pur-
molten binder is forced between the particles. Mixers come in chase feedstock. More than a dozen feedstock suppliers exist,
two basic classes, batch and continuous. The batch mixers are each with a range of materials.
reserved for smaller lots, but up to 700 kg (about 1,500 lbs.) A homogeneous feedstock mixture is needed to maintain
at a time is mixed in the largest batch mixers. Batch mixers process control. Because feedstock viscosity is sensitive to
involve a rotary paddle or blade that splits the molten feed- composition, inhomogeneities impede uniform flow in the
stock repeatedly over a mixing time from 30 min to a few mold cavity. At higher temperatures the polymer thermally
hours. More typical to production is a continuous mixer. The expands more than the powder, effectively increasing the rel-
13
Table of Contents
Chapter 2: Process Options
(b)
Figure 2.3. Demonstration of how powderbinder rheology varies
with the ratio of powder to fluid phase, in this case using flour
and milk; (a) low solid-particle content giving a fluid that flows, Figure 2.5. A sintered tungsten medallion resting on the tooling
(b) a high solid-particle content mixture that is stiff and not able to insert used for molding, illustrating typical sintering shrinkage, in
flow. MIM relies on a critical viscosity that allows flow in a high this case for a tungsten heavy alloy (components courtesy of Anita
solids loading Hancox)
14
Table of Contents
Chapter 2: Process Options
15
Table of Contents
Chapter 2: Process Options
During molding the feedstock granules are loaded at room When ready to fill the mold, the hot feedstock accumulated
temperature into the hopper of the molding machine. Figure at the tip of the screw is rammed into the mold cavity. At this
2.7 is a picture of a production molding machine. It is essen- point the feedstock consistency is similar to toothpaste. In
tially the same as used for plastic injection molding. Attached molding, the feedstock flows through the nozzle, at the end
to the molder is robot system for removing the component and of which it enters the tooling through the sprue, runner, and
staging it for debinding and sintering. Such an integrated han- gate prior to filling the cavity. Once the mold is filled, the
dling and quality system is shown in Figure 2.8 in the form of feedstock is cooled to freeze the powderbinder mixture into
an unloading robot and conveyor system. the desired shape. When sufficiently cool, the rigid component
A cross-section schematic helps to understand the feed- is ejected from the mold and the molding cycle is repeated,
stock flow through the molder. As sketched in Figure 2.9, the typically once every ten to twenty seconds.
hopper for loading the pellets feeds into a heated barrel with Figure 2.10 is a picture of an open mold located in a mold-
an internal screw for stirring, compressing, and melting the ing machine. This is for a turbocharger rotor, so the tool has
feedstock. The end of the barrel has a nozzle that butts against eleven slides that are activated by the pins when the mold
the mold during filling. The mold is clamped closed up to the closes; the pins are evident on the right side of the picture.
point of component ejection. It is important for the screw to The turbine part is still in the mold to show the possible blade
heat, melt, and stir the feedstock to ensure homogeneity and shape. A screw, barrel, and mold are shown in Figure 2.11a;
no trapped air bubbles. The screw has a taper that compresses
the hot feedstock to squeeze out any trapped air.
Figure 2.10. The open mold inside the molding machine. This is a
Figure 2.8. An integrated robot and conveyor system to move the turbocharger tool cavity that has several sliding motions to form
green shapes to a staging area prior to debinding (photograph the vanes on the turbine which is still in the mold (photograph
courtesy of Arburg) courtesy of T&N Technology)
16
Table of Contents
Chapter 2: Process Options
normally the screw is inside the barrel and not seen. An form heating while maintaining mixture homogeneity. As
alternative tool design based on a circular mold is given in shown in the cross section of Figure 2.12, after heating and
Figure 2.11b. stirring the feedstock the screw acts as a plunger to generate
A motor turns the screw to stir the feedstock to ensure uni- the pressure needed to fill the die. The screw tip has a check
valve, the action of which transforms the rotating screw into
a forward moving plunger to generate pressure to fill the mold.
Figure 2.13a shows a sketch of the internal flow path, indi-
cating the component (two cavities in this case), gate, runner,
and sprue, and Figure 2.13b is a picture of molded parts from
(a)
(a)
(b)
Figure 2.11. Mold tooling; (a) is a picture of a mold, screw, and
barrel used in MIM, and (b) shows a circular design for the mold
(b)
Figure 2.12. The upper sketch illustrates the rotating screw inside Figure 2.13. The flow path in molding is shown in terms of
the barrel prior to mold filling. To fill the mold, as shown in the (a) a sketch of the internal flow path in the tooling for the molten
lower sketch, the screw becomes a plunger and the check ring feedstock exiting the nozzle, and (b) the sprue, runner, and molded
stops the back flow of molten feedstock during the forward motion automotive components aligned to show the cavity in the mold
17
Table of Contents
Chapter 2: Process Options
18
Table of Contents
Chapter 2: Process Options
between the molding problems, tooling cost, and machine pro- tion the majority of the binder is selected to evaporate at lower
ductivity often results in a compromise of about two- or four- temperatures while a higher-temperature backbone polymer
cavity molds. holds the particles in place. In this regard, paraffin wax evap-
orates first, leaving polypropylene to hold the particles in
Debinding place. Finally, the residual binder evaporates as the porous
After ejection from the mold, the binder is removed from powder structure approaches the sintering temperature. A
the component, hopefully without damaging the powder struc- common burnout cycle then is at least 12 hours for the com-
ture. Thermal debinding is the easiest to envision and the most bined heating, hold, and cooling. A few batch-debinding fur-
widely practiced. The molded component is slowly heated to naces are shown in Figure 2.17. Trays of parts are inserted
evaporate and decompose the polymeric binder. For this op- through the front door and a burnoff stack is evident where
the decomposed polymer is combusted to form water and car-
bon dioxide. Note that many newer operations integrate de-
binding and sintering into a single furnace.
An alternative to thermal debinding is to immerse the com-
ponent in a solvent that dissolves some binder, leaving an
insoluble backbone polymer behind to hold the particles in
place for handling. That remaining backbone binder is ther-
mally extracted as part of the sintering cycle.
Early in the development of MIM the selected solvents
were hazardous chlorinated hydrocarbons such as carbon
tetrachloride. Heptane is popular for debinding paraffin wax
and is widely used in spite of flammability concerns. Since
the 1990s, solvent debinding turned to water-soluble polymers
mixed with insoluble backbone phases. Debinding is achieved
by immersion in hot water as illustrated in Figure 2.18, where
Figure 2.17. Batch thermal-debinding furnaces for heating of the polyethylene glycol is substituted for the paraffin wax.
molded component to burn out the binder phase and presinter Another option involves catalytic-phase erosion of a poly-
the particles (photograph courtesy of Advanced Materials acetal binder. The polyacetal binder is attacked by nitric acid
Technologies) vapor during heating, generating formaldehyde that must be
properly handled. A polyethylene-type backbone polymer pro-
vides handling strength during heating to the sintering tem-
perature. The laboratory-scale catalytic debinding unit shown
in Figure 2.19 has a front door that is used to insert the com-
ponents and, once a nitrogen atmosphere is established, nitric
19
Table of Contents
Chapter 2: Process Options
acid is added and the byproducts combusted in the stack at the cemented carbides, and some polymers. Sintering dramati-
top of the unit. cally changes a powder structure into a solid structure. This
In all of the debinding approaches, the erosion of the binder is illustrated in Figures 2.20a through 2.20f, a sequence of
is from the outside surface. If the system is heated rapidly, then high-magnification cross-section pictures of 17-4 PH stainless
internal vapor pockets form and create blow holes, blisters, or steel powder taken after heating to various temperatures up to
even miniature volcanic eruptions. Generally, slow heating to 1,365C (2,460F), showing progressive particle bonding and
600C (1,112F), or higher, over several hours in nitrogen re- pore elimination during sintering. In these images the pores
moves the binder and presinters the powder. The heating rate are black.
and time for debinding increase as the section thickness in- The goal of sintering is to densify the powders by removing
creases. For most metal powders and binders, a rule of thumb the void space previously occupied by the binder. Obviously
is that debinding can penetrate about 2 mm/h (0.08 in./h). full binder burnout is required prior to densification, or some
Since the erosion is from both the front and back (or top and polymer residue will be trapped in the pores. When properly
bottom) surfaces, the time for debinding is approximated by performed, the sintered microstructure shows no evidence of
considering the thickest half-section. For example, a wall that the original powders and little to no porosity. For example,
is 10 mm across requires heating at 1C/min (1.8F/min) to at Figure 2.21 is a micrograph taken of tool steel after MIM pro-
least 450C (842F) with a hold of 2.5 h (5 mm divided by 2 cessing to near-full density. This microstructure shows desir-
mm/h) as long as both surfaces can be accessed during debind- able small carbides dispersed in the steel structure with a small
ing. This says that heating requires about 6 hours, with a peak amount of residual porosity.
hold of 2.5 hours, and about 6 to 8 hours of cooling time, so a The typical sintering shrinkage is from 12% to 18% on
12-to-24-hour cycle is typical. each dimension, nominally 15%, so the molded component is
oversized to deliver the desired final dimensions. Figure 2.22
Sintering illustrates the typical uniform, isotropic shrinkage attained in
Independent of the initial debinding route, final polymer going from the molded to sintered body, in this case for a med-
extraction usually occurs during heating to the presintering ical-pump housing. However, since gravity acts in one direc-
temperature. Full sintering occurs as the powder compact ap- tion, the component mass gives frictional drag on the sintering
proaches its melting temperature. Sintering was used thou- support. As a consequence, the reality is slightly less than
sands of years ago to form bricks and pottery by firing the isotropic shrinkage. Thus, the common approach is to leave a
shaped green ceramic body to a high temperature. Today, little extra on the mold for final dimensional corrections after
the same idea is applied in powder metals, technical ceramics, trial sintering-shrinkage measurements.
20
Table of Contents
Chapter 2: Process Options
(a)
(b)
Figure 2.24. Two examples of a batch vacuum-atmosphere
sintering furnace used for production MIM sintering: (a)
Figure 2.23. A typical small box furnace for sintering MIM refractory metal hot zone, and (b) graphite hot zone (photographs
components up to 1,600C in a protective atmosphere of hydrogen, courtesy of Powder Injection Moulding International, Advanced
nitrogen, or argon Materials Technologies, Elnik Systems, and Key Safety Systems)
21
Table of Contents
Chapter 2: Process Options
Figure 2.25. Continuous debinding and sintering is possible in a furnace such as shown here (photograph courtesy of CM Furnace)
of 2 m (over 6 ft.) diameter and 3 m (about 10 ft.) long, reach- This is an ideal way to evaluate design variations without
ing load capacities up a ton per day. For larger production committing to tooling changes. Commonly green machin-
quantities it is appropriate to move to continuous furnaces, ing is used to remove gate blemishes or parting line flash.
where sintering and debinding are combined into a single de- controlled porosityImportant powder metallurgy prod-
vice with a sequence of zones that require from 6 to 24 hours ucts are created with controlled porosity and controlled
to traverse. Figure 2.25 is a photograph of a continuous fur- pore size for use as flow-control devices, filters, and
nace capable of 1,600C (2,912F) operation. The central bearings. In addition, controlled porosity is a means to
zones are profiled to hold precise temperatures while the com- form air bearings and tools for aeration. If an insoluble
ponents move via a pusher-plate conveyor mechanism from phase is mixed into the feedstock prior to molding, then
the cold entry to where they are cooled prior to exit. the size and amount of that insoluble phase determines
Since shrinkage is associated with sintering, precise final the pore size and porosity in the final sintered object.
dimensions rely on uniform molding and uniform heating, Several sacrificial additions serve this function, including
such that the component shrinks to the specified final size. sodium chloride (salt). Other sacrificial phases include
When MIM is properly executed, the sintered compact has the polymer particles that are insoluble in the binder.
shape and precision of an injection molded plastic, but the sin- green joiningTwo green MIM parts are joined to make
tered metal delivers performance levels unattainable with a larger, more complex shape by effectively gluing the
plastics. This is because the sintered microstructure is essen- two bodies together prior to debinding and sintering.
tially fully dense, so there is no residue from the original pow- Variants on this idea include first molding a core material
der. Further, proper cooling in the furnace can impart a heat and then over molding a second material or combining
treatment to the structure to induce excellent mechanical prop- two pieces with binder along the bond interface. The lat-
erties. The sintered product may be further densified, coined, ter is a simple means to form a hollow body using two
plated, or machined to complete the fabrication process. halves. Demonstrations have even included ball and
socket creations assembled green.
NOVEL OPTIONS two-material moldingIn some molding operations it is
One fact evident in MIM is how inventive everyone can be possible to use molding machines with two barrels, each
once acquainted with the basic concepts. A large range of op- barrel containing a different feedstock material. The two
tions exist to form novel sizes or shapes, novel material com- barrels work in sequence to fill the mold, allowing place-
binations, or even new compositions. Here are some of the ment of a different feedstock in different regions. Special
options that fall under auspices of MIM: two-barrel molding machines, similar to that shown in
green machiningA green molded compact is machined Figure 2.26, allow the tooling to move between the two
to add extra detail or to remove blemishes after molding. nozzles for sequential filling. The first step molds the
22
Table of Contents
Chapter 2: Process Options
inner material, and the second step molds the outer ma-
terial. This approach is used to combine magnetic and
nonmagnetic materials in automotive timing sensors.
gas- or water-assisted moldingA fluid phase can push
on the inside of the feedstock during molding to pressur-
ize a hot feedstock bubble against the mold walls, like
blowing bubble gum into a cavity. Warm water or gas is
pressurized to expand the feedstock bubble. An early
demonstration component was a ceramic spoon fabricated
with a hollow gas-filled handle, shown in Figure 2.27.
large structuresCosts usually increase rapidly with
component size in MIM, so novel means are required to
enable large structures in the 200 g and larger range. One
Figure 2.26. A two-barrel MIM molding machine intended to make approach for limited production is to form soft silicone
multiple material components by first filling part of the mold with one
feedstock, dilating the mold to make room for the second feedstock, rubber molds around a model and this silicone rubber
and then rotating to enable final filling with the second feedstock mold is filled with hot feedstock at low pressures. Once
(photograph courtesy of Advanced Materials Technologies) the feedstock has hardened, the body is removed from the
mold, debound, and sintered. Examples of the shape com-
plexity, size scale, and feature preservation are given in
Figure 2.28 which is a picture of a bronze component 120
mm (4.8 in.) high, 2 kg (4.4 lb.) in weight that illustrates
the ability to create undercuts, holes, and thickthin
sections.
microminiatureAt the other end of the size scale are
tiny medical and electronic components produced using
small powders and automated molding. Molds are fabri-
cated using ultraviolet lithography techniques developed
for semiconductor fabrication. Since the parts are very
small, part handling must be automated. An early exam-
ple was a stainless steel biopsy tool with 100 m wall
Figure 2.27. A hollow handled ceramic spoon formed using gas thickness. The next wave of applications are seen in
pressure filling to blow the feedstock against the mold walls
(component courtesy Arburg). Figure 2.29, in this case for microarrays of features, such
as used for blood analysis, chemical studies, and even
ultra-capacitors.
composite materialsDifferent powder chemistries are
mixed into the feedstock to form a particle composite
after sintering. If the added particles are insoluble and
hard, then the sintered structure has enhanced wear re-
sistance. Examples include borides in a stainless steel
matrix, titanium carbide in a titanium matrix, and silicon
carbide in an aluminum alloy matrix. An example par-
ticulate composite microstructure is illustrated in Figure
2.30. Wear trials show a small concentration of hard par-
ticles, in the range of 3 v/o, increases wear resistance
from ten-fold to fifty-fold for many alloys.
high elastic modulusThe addition of a ceramic phase
to a metal is useful for increasing the elastic modulus.
For example, steels have been loaded with titanium
boride (TiB2) to increase elastic modulus. With 40 v/o
boride the steel stiffness increased by 60%. An example
Figure 2.28. An example large component molded using a low-cost
tool material. This bronze statue weighs 2 kg and was molded in
micrograph of a titanium carbidereinforced titanium
silicone rubber using low-pressure molding (component courtesy alloy is given in Figure 2.31.
of Aesthetic Materials). heterogeneous microstructuresWhen mixed powders
23
Table of Contents
Chapter 2: Process Options
are sintered, regions of different composition are retained greatly improved fatigue properties. For example, an
even after full density is reached. Often the sintering time Fe-2Ni-0.5Mo-0.4C alloy formed from mixed powders
is prolonged to homogenize these gradients, but in some gives 20% higher tensile strength with a heterogeneous
cases the heterogeneous (hardsoft) microstructure has microstructure vs. that attained with a homogeneous mi-
Figure 2.29. Microminiature molding is possible in several array geometries as used for bioassay, electronic, capacitor, and small-scale
testing applications (photograph courtesy of Taisei Kogyo)
Figure 2.30. Improved wear properties are possible by the Figure 2.31. A high-elastic-modulus composite formed by MIM;
addition of hard particles in a MIM feedstock. Shown here is the in this case the microstructure corresponds to titanium carbide
microstructure of stainless steel reinforced with tungsten carbide reinforcement in a titanium matrix (micrograph courtesy of
(micrograph courtesy of Anthony Griffo) Timothy Weaver)
24
Table of Contents
Chapter 2: Process Options
crostructure. An example of the microstructure hetero- tissue scaffolds A relatively important new development
geneity is shown in Figure 2.32, where the difference in is the creation of MIM tissue scaffolds for implants in hu-
properties point-to-point is tailored to improve fracture mans. Usually the metal is titanium and the structure is
properties. molded to about 40% solid density to match human bone,
nitrogen alloyingIt is well known that nitrogen has with an elastic modulus near 10 GPa and strength near
similar effects on the strength of steel as carbon. In sin- 100 MPa (about 15,000 psi). The ideal pore size is about
tering it is possible to infuse nitrogen into the iron in a 100 to 200 m, much larger than the particle size. These
way not possible with casting. Thus, new alloys, such as scaffolds are treated with hydroxyapatite (calcium phos-
extra-high-strength tool steels with carbon-nitrides or al- phate) to induce bone growth to serve as anchors for
loys devoid of nickel (both nitrogen and nickel are implants, artificial teeth, or trauma repairs. Early demon-
austenite stabilizers) become available from MIM and strations relied on sacrificial pore-forming agents, such
nitrogen sintering. One very successful product is based as plastic particles or salt, mixed with MIM feedstock.
on forming austenite (face-centered cubic iron) as a non- When that pore former is removed the resulting void cre-
magnetic stainless steel for use in watches. The nickel, ates a distended structure with desired pore size and
which is a common austenite former in stainless steels, porosity, such as shown by the computer tomography
is eliminated and that avoids concern over skin irritation scan in Figure 2.33. Another variant relies on a preform
from the common nickel allergy found in humans. polymer foam that is filled with molten feedstock. When
burned out and sintered the titanium takes on the negative
image of the foam (where there previously were pores it
is now solid and where is was previously solid there are
now pores).
25
Table of Contents
Chapter 2: Process Options
26
Table of Contents
Chapter 2: Process Options
SECONDARY OPERATIONS
After the component is sintered, many additional changes
in size, shape, surface finish, or heat treatment are possible. Figure 2.38. An example of value added via secondary operations
Since the sintered structure is the same as wrought material, in a solenoid valve body, formed from a nickeliron alloy. The
sintered component has machining on the base, added holes,
the response to various secondary fabrication steps is the same threads, and a staked core (components courtesy of Advanced
as seen with wrought materials. Materials Technologies)
Some of the common steps include the following:
coining or cold deformationA sintered compact is
forced to conform to a rigid mandrel or substrate and in
doing so the body is slightly deformed to match the sub-
strate. This allows for proper sizing of slots or holes, or
reduced spread in dimensions. Coining is also used to
flatten thin sections that tend to warp in sintering.
hot deformationThe sintered compact is heated and de-
formed by a rapidly applied loading stroke to ensure
proper size and density, and proper heat treatment (often
the compact is quenched from the hot deformation step).
For a steel, the sintered strength jumps from 500 MPa to
Figure 2.39. An example microstructure from the laser-welded
720 MPa after hot deformation. In a few materials hot
joint between two MIM components showing excellent bonding
deformation is a means to remove residual porosity. (photograph courtesy of Rene Cooper)
machiningAll common machining operations are ap-
plied to MIM components, often to add threads, under- properties.
cuts, grooves, or other special features that would be surface carburizationCarbon is important to attaining
difficult or expensive to place in the tooling. It is a good high strengths in steels. A high surface hardness is at-
means to add drilled holes after sintering, as evident in tained with carbon surface additions using a heating
the solenoid housing shown in Figure 2.38. Careful ex- cycle with an atmosphere containing methane. Surface
amination shows this body has holes and a staked post carburization cycles result in some loss of dimensions,
added after sintering. so trade-off is required between surface hardness due to
heat treatmentSintering leaves the material in an an- the addition of carbon (carburization) and dimensional
nealed condition due to the high temperature and slow accuracy.
cooling rate. For low-carbon ferrous alloys, this is of lit- joiningLike other metallic components, MIM compo-
tle consequence, but for high carbon levels there is much nents are joined by welding, brazing, or even adhesive
change possible via post-sintering heat treatment. For techniques. For the most part, MIM materials behave the
MIM materials, heat treatment is possible within the same as standard metals. Figure 2.39 is a cross section
cooling cycle or by use of a properly designed post-sin- through a weld between two MIM stainless steel parts,
tering cycle. In some cases a cycle is designed to maxi- showing a refined microstructure with excellent strength.
mize strength and hardness, while in other cases the surface treatmentsSurface treatments such as polish-
material is cooled to induce a soft structure. Precipita- ing, coating, painting, cleaning, anodizing, plating,
tion-hardened stainless steels especially require cycles sealing, and laser glazing are all applied to MIM com-
of heat and hold to properly optimize the mechanical ponents. Surface-hardening treatments sustain a tough
27
Table of Contents
Chapter 2: Process Options
core with a hard surface. Electroplating is used for either ponent sticks in the mold and will not eject when cooled, while
improved aesthetics or corrosion resistance. Most of the inadequate pressure results in sink marks where the component
electroplated coatings are chromium, nickel, copper, shrinks away from the mold during cooling before ejection.
gold, zinc, silver, or cadmium. Nickel electroplate is a After packing, the calculations shift to the cooling stage and
favorite on instrumentation, firearm, or magnetic com- the extraction of heat from the molded feedstock. Most of the
ponents, and cadmium was a favorite on many electronic solutions are based on finite difference and finite analysis so-
components, but its use is declining because of environ- lutions, starting with the input material and mold parameters.
mental problems. An important supplement takes the subtle density gradients
Clearly, after sintering, MIM components behave similarly calculated in the molded green body and imports the gradient
to other metallic components. No special handling procedures information into subsequent sintering simulations to predict
or steps are required in the secondary operations for MIM parts. density, microstructure, and final component size and shape.
Generally a skilled engineer is required to fully benefit from
PROCESS SIMULATIONS the process simulations, and many of the MIM shops are too
Since the early 2000s, the metal powder injection molding small to make this commitment. Consequently, hands-on ex-
process has been supported with a variety of process simula- perience is favored in the smaller shops over process simula-
tions. About eight simulations currently exist, and of those, tions during tool and process design. However, considerable
six are available as commercial codes. Much difference exists sophistication exists and a few firms have made excellent
in the sophistication of the codes, the type of computer re- progress in integrating process simulation into the product of-
quired for the solutions, and the time for the solution. Some fering. Reports on savings show that on average the extra cost
of the codes are written exclusively for MIM, while others are for process simulations is regained by faster component deliv-
adaptations of casting codes or plastic injection molding ery with fewer defects. In an overall cost analysis, the consen-
codes. Only one integrates molding, debinding, and sintering sus is a 15% cost savings when simulation is used. It slows the
into a stepwise set of simulations. initial phase of process development, but helps avoid surprises
The simulations are focused first on the filling stage, and and delays when the tool is placed into production.
determining how the hot feedstock enters, flows, and cools in
the mold. These require conservation of mass, energy, and mo- VENDOR DIFFERENCES
mentum. Images from the filling stage are useful in determin- There is no standard MIM process. Thus, variations in
ing the sequence of feedstock flow and filling in the cavity so MIM components occur between vendors. Several factors
that gate and vent are properly located and sized to avoid jet- contribute to the variation, including differences in starting
ting, incomplete filling, or trapped air. Figure 2.40 is an image powders and binders, processing cycles, equipment design and
from one such filling simulation, showing via coloration the operation, impurities, and post-sintering heat treatment.
time-based advance of the flow front. Within one MIM operation, the property scatter is usually very
Next is the packing stage where the proper pressure is de- small. For ductile systems, the strength will typically have a
termined for the hydraulic system based on calculation of the standard deviation of less than 20 MPa (about 3 ksi) and the
required pressure in the mold. Too much pressure and the com- elongation will have a standard deviation of approximately
1%. Yet when compared between operations, properties that
tend to be associated with crack propagation in impact, frac-
ture, and fatigue vary considerably.
The large sensitivity seen in MIM fracture properties traces
to the interaction of impurities (small contents of oxygen, ni-
trogen, carbon, and semi-metals such as phosphorous, sulfur,
silicon, and boron) and the microstructure during cooling from
the sintering temperature or the final heat-treatment tempera-
ture. Usually the cooling is slow, so impurities migrate and
accumulate on grain boundaries or phase boundaries. The
change in atomic bonding due to the concentrated impurities
substantially alters the bulk properties.
For example, stainless steels can absorb nitrogen in sinter-
ing and this tends to increase strength, but also reduces duc-
tility to half the value without nitrogen. Further, if slowly
Figure 2.40. Computer simulation of mold filling for a copper cooled, residual nitrogen reacts to from a compound with
fitting, showing the gate location and progressive fill vs. time chromium with the effect that chromium is removed from its
(images courtesy of Seong Jin Park) role in corrosion, leading to significant corrosion problems.
28
Table of Contents
Chapter 2: Process Options
Hence, depending on the processing atmosphere and final Society of Powder and Powder Metallurgy, 2002, vol. 49, pp.
density there will be property differences. 841845.
One of the challenges facing MIM is to settle on standard-
J.A. Sago and J.W. Newkirk, The Effects of MIM Processing
ized materials and processes to deliver consistent properties.
Parameters on Variations in Part Weights and Dimensions,
For this reason, when a MIM part is to be used in a dynamic
Advances in Powder Metallurgy and Particulate Materials
loading situation, where fracture toughness, impact toughness,
2005, compiled by C. Ruas and T.A. Tomlin, Metal Powder
or fatigue resistance is a concern, it is imperative to perform
Industries Federation, Princeton, NJ, part 4, pp. 7283.
in-service testing to ensure adequate properties. Performance
testing needs to be under the intended loading conditions and G. Schlieper, Back to Basics: Putting the Finishing Touches
in the intended service environment as a basis for establishing to Metal Injection Moulded Components, Powder Injection
the vendor credibility. Moulding International, 2009, vol. 3, no. 3, pp. 3540.
Thus, to summarize vendor effects, between MIM vendors
T. Shimizu, K. Matsuzaki and Y. Ohara, Process of Porous
there is
Titanium Using a Space Holder, Journal of the Japan Society
little difference in hardness and bulk chemistry
of Powder and Powder Metallurgy, 2006, vol. 53, pp. 3641.
small difference in yield strength, tensile strength, and
ductility L.K. Tan, R. Baumgartner and R.M. German, Powder Injec-
potentially larger differences in impact toughness, tion Molding of Bi Metal Components, Advances in Powder
fracture toughness, corrosion resistance, and fatigue Metallurgy and Particulate Materials2001, compiled by
endurance strength. W.B. Eisen and S. Kassam, Metal Powder Industries Federa-
tion, Princeton, NJ, part 4, pp. 191198.
RESOURCES
E.L. Wolf and S.R. Collins, GTA Welding of AISI 316L
S. Ahn, S.T. Chung, S.J. Park and R.M. German, Modeling
Metal Injection Moulded Components, Powder Metallurgy,
and Simulation of Metal Powder Injection Molding, ASM
2002, vol. 45, pp. 201203.
Handbook, Vol. 22, Modeling and Simulation, 2010, ASM In-
ternational, Materials Park, OH.
S.V. Atre, S.J. Park, R. Zauner and R.M. German, Process
Simulation of Powder Injection Moulding: Identification of
Significant Parameters during Mould Filling Phase, Powder
Metallurgy, 2007, vol. 50, pp. 7685.
M.J. Edirisinghe and J.R.G. Evans, Review: Fabrication of
Engineering Ceramics by Injection Moulding. I. Materials Se-
lection, International Journal of High Technology Ceramics,
1986, vol. 2, pp. 131.
M.J. Edirisinghe and J.R.G. Evans, Systematic Development
of the Ceramic Injection Molding Process, Materials Science
and Engineering, 1989, vol. A109, pp. 1726.
J. Fleischer and A.M. Dieckmann, Automation of the Powder
Injection Molding Process, Microsystem Technologies, 2006,
vol. 12, pp. 702706.
R.M. German, Powder Injection Molding, 1990, Metal
Powder Industries Federation, Princeton, NJ.
R.M. German and A. Bose, Injection Molding of Metals
and Ceramics, 1997, Metal Powder Industries Federation,
Princeton, NJ.
A.D. Hanson and S.C. Perruzza, Optimizing Component De-
signs for Metal Injection Molding, International Journal of
Powder Metallurgy, 2000, vol. 36, no. 3, pp. 3742.
T. Harikou, Y. Itoh, K. Satoh and H. Miura, Joining of
SUS316L by Insert Injection Molding, Journal of the Japan
29
Table of Contents
Table of Contents
Chapter 3
DESIGN PRINCIPLES
DESIGN CYCLES alized design cycle is illustrated in Figure 3.1, wherein peri-
Presented here is a condensed approach to the identifica- odic input from MIM vendors ensures manufacturability as
tion of good-design matches with the process capabilities of the design is refined. On each contact the details increase,
metal powder injection molding. Each cycle involving a new touching on component features, materials selection, tooling
design or product redesign provides an opportunity for em- issues, and process identification. In general, costs incurred
bracing MIM. To ensure success, it is important to translate during design have a relatively small impact on final product
performance attributes into quantitative-design requirements cost, but decisions made during the design process have a
that spell out the form (size, shape), function (properties, ma- major impact on production costs.
terials), tolerances, and cost factors (primary fabrication route
and secondary operations). RECOGNIZING GOOD CANDIDATES
Much experience exists on how to deliver the required per- Qualitative Criteria
formance while ensuring success within the MIM process. The strengths of MIM are outlined in a qualitative manner
During the design cycle, it is best is to have frequent contact in Figure 3.2. This Venn diagram shows major process
with the fabrication community to rationalize design goals and strengths to help in selecting good candidates. The initial qual-
process constraints. If this contact is delayed to the quotation itative criteria look at production quantities, shape complexity,
stage, then experiences and knowledge that might improve material performance, and component cost. Metal powder in-
productivity and lower cost will be lost. Conceptually, the ide- jection molding is not low-cost, except in those instances
31
Table of Contents
Chapter 3: Design Principles
Quantitative Criteria
The attributes of production quantity, shape complexity,
cost, and performance are reflected by several quantitative
factors. First, we will quantify the production-quantity aspect.
Small quantities, at a few thousand per year, give less op-
portunity to amortize tooling and engineering costs and are
generally discouraged by the production community. Often the
best match is with production quantities from 5,000 per year
to more than 100 million per year, such as for cellphone vi-
brator weights, computer logos, eyeglass hinges, and ortho-
dontic brackets. One example disk-drive latch is shown in
Figure 3.3. It is a small stainless steel component produced at
a rate of about 30 million per year in multiple-cavity tooling.
Soft tooling combined with low molding pressures is one
Figure 3.1. The classic design cycle involves identification of the prob- option for the production of trial quantities. One application
lem followed by a product concept development, on to engineering for soft tooling is the sculpted female image photographed in
conceptualization, design, and final definition. Often the MIM technol-
ogy is only brought in after the design is advanced, but this limits the Figure 3.4. This bronze was produced in an edition of three
cost savings and quality improvements. For best MIM designs, it is replicas using low-pressure molding in rubber tooling. Tool
important to have reviews with process experts to incorporate changes
in the design to better advance quality while reducing cost
Figure 3.2. This Venn diagram suggests that most justifications for
MIM result from various intersections of considerations of cost, Figure 3.3. This small stainless steel latch was one of the first
shape complexity, production quantity, and performance. In each high-production-volume disk-drive MIM components, reaching
region, various example components are named as illustrations; production rates exceeding 30 million per year (components
many of these are pictured in this text courtesy of Advanced Materials Technologies)
32
Table of Contents
Chapter 3: Design Principles
wear limits such production to a few hundred parts. At the ing. In the context of steel tooling, soft implies the steel is not
level of 5,000 per year there are several reports on success for heat treated after machining. Figure 3.5 is an example of a
jet-engine components, specialty firearm sights, military com- gunsight fabricated in this manner. When the production quan-
ponents, surgical tools, and luxury items using soft steel tool- tities are large, the tooling and processing technology are well
positioned to deliver outstanding performance levels.
Shape complexity is the next consideration, and at first pass
it depends on the number of engineering specifications re-
quired to define the component. Metal powder injection mold-
ing is best applied when there are more than 20 specifications
(dimensions, locations, surface finish, and such) on the engi-
neering definition. However, the technology struggles when
there are more than 250 specifications. Components with hun-
dreds of features often require several post-sintering machin-
ing operations to hold tolerances. Figure 3.6 plots the
distribution in dimensional specifications taken from a survey
of several production parts. In this survey the average com-
plexity was 74 features. Clearly a vast majority of production
components have less than 100 features. For example, the link
shown in Figure 3.7 has 30 dimensional specifications and is
33
Table of Contents
Chapter 3: Design Principles
at the lower end of complexity, while the optical housing a good choice from a cost standpoint. The powder-cost issue
shown in Figure 3.8 has 165 dimensional specifications and is a circular argumentstainless steel powders for MIM are
is challenging. Components, such as wristwatch cases, as low cost because of the high use, so the low cost drives many
shown in Figure 3.9, are good examples of the complexity designs to use stainless steel. For low-melting-temperature
level where the MIM technology excels. materials the large range of forming options generally make
Material and material properties constitute the third factor MIM less economically attractive for materials such as alu-
to consider. Many materials are available, but a few dominate minum, magnesium, and zinc.
the field. When the material is difficult to machine, such as As a simplification, the properties attained after sintering
tool steels, titanium, ceramics, stainless steel, or nickel-base largely converge to the range found in engineering handbooks.
alloys, then net-shaping is most beneficial. As production Yet, new materials are emerging within MIM, custom de-
quantities increase the material cost decreases. Because of signed for the process, that are not in the handbooks. These
widespread use, stainless steels or oxide ceramics are usually include laminated (hardsoft, magneticnonmagnetic, and
conductiveinsulator) combinations in a single component,
nickel-free nonmagnetic stainless steels, ultrahigh-strength
steels, foamed metals, and mixed-phase wear-resistant com-
posites. These options provide distinct advantages emerging
from basic process attributes. Properties for some special
materials are given in Chapter 5.
The rest of this chapter provides details on the allowed de-
sign features and restricted options. Subsequent chapters pro-
vide details on tolerance capabilities, process yields, materials,
properties, and costs. To summarize, the key criteria needed
to identify good MIM candidates are as follows, recognizing
that a wide variation exists:
production quantity more than 5,000 per year usually,
often exceeding 200,000 per year
number of engineering specifications or features between
20 and 250 and most often in the 50 to 75 range
availability of many materials, although stainless steels,
oxide ceramics (alumina, zirconia), steels, tool steels,
nickel alloys, iron-nickel soft magnetic alloys, and spe-
Figure 3.8. Optical housing with 165 specified features, reflecting
one of the more demanding MIM products (components courtesy cialty alloys of tungsten, titanium, cobaltchromium, and
AT&T) electronic alloys are most popular
property levels generally rivaling handbook tabulations
for similar compositions and heat treatments
GEOMETRIC DETAILS
This section gives more focus on the component size and
mass attributes. Besides the attributes listed above (production
quantities, shape complexity, and performance), several other
geometric design factors are important in MIM.
Effective-Density Concept
A good MIM design is one that would require much mass
removal if it were formed by machining. Here our concept is
similar to that encountered in the aerospace community, what
they term the buy-to-fly ratio. Often the purchased raw mate-
rial is seven times that of the final component, meaning that
85% of the starting material ends up as machining scrap.
In MIM this is viewed in terms of a low effective density.
Figure 3.9. Three different watchcase designs all fabricated by
Density is mass over volume, so an effective density is calcu-
MIM, reflecting good examples of the size, shape, and complexity lated as the component mass divided by its outer envelope
where MIM has historically been most successful volume (outer embracing volume) from which grinding or
34
Table of Contents
Chapter 3: Design Principles
machining would start. The concept is illustrated in Figure verse correlation shows that long items are especially low in
3.10, where the sketched outer volume is slightly more than effective density. For example, a stainless steel soup spoon
15 cm3 for a final mass of 27 g. From these two parameters has a length of about 200 mm (8 in.). Based on this plot, an
the effective density is calculated at 1.8 g/cm3 using the mass effective density near 12% (1 g/cm3 out of 8 g/cm3) would
divided by the outer embracing volume. This value is about merit consideration of MIM. Conceptually there would be
23% of the wrought steel density. To phrase it differently, if much mass and time savings using MIM vs. forging followed
the device is machined from a wrought block, assuming that by machining and polishing (the typical process). No surprise,
even the proper sized 15 cm3 block is available, about 77% of the idea of a MIM spoon such as the stainless steel soup spoon
the initial material would be machining scrap. Such waste of shown in Figure 3.13 has been demonstrated by several firms
machining time and initial material adds to the cost of manu- in recent years. Although other tableware has been demon-
facturing by machining. strated, the MIM spoon proved most attractive.
Many cost-effective MIM components have especially low
effective densities, indicating that machining would be time Surface Finish
consuming and wasteful. To reinforce this point, Figure 3.11 Related to the material loss in machining is a similar con-
plots the effective density distribution from a large collection cern over the time spent in reaching a desired surface finish.
of production MIM steel and stainless steel components (gen-
erally these alloys are similar with a range from 7.7 to 8.0
g/cm3 theoretical density). The MIM components are gener-
ally shapes with a low effective density, typically in the range
of 25% of the material density (2 g/cm3 out of nearly 8 g/cm3).
Thus, historically the injection molded components have gen-
erally avoided wasting of 75% of the initial material.
To further illustrate the design window, Figure 3.12 is a
scatter plot of effective density vs. largest dimension. The in-
35
Table of Contents
Chapter 3: Design Principles
36
Table of Contents
Chapter 3: Design Principles
one is for the size of the thickest wall and one is for the
thinnest wall. The overall median wall thickness is in the 1 to
3 mm range (0.04 to 0.12 in.).
Further, MIM works best when the wall thickness is uni-
form to ensure rapid debinding. The scissor blades from zir-
Figure 3.15. A long, thin MIM medical hand tool that represents a
Figure 3.18. These scissors represent an example of metal powder
larger maximum size produced in the technology; in this case the
injection molding used to form a thin-walled shape that has a
components are 250 mm or about 10 in. long (components courtesy
nearly constant wall thickness. The constant wall thickness avoids
of Injectamax)
distortion stresses and a thin wall shortens heat transfer to make
for shorter cycle times in molding and debinding
37
Table of Contents
Chapter 3: Design Principles
Figure 3.19. Mass distribution for MIM ferrous (steel, stainless Figure 3.22. Scatter diagram of component weight (mass) vs.
steel) components, showing a broad range from fractions of a maximum size for ferrous MIM components, giving a general
gram to nearly 1 kg, with a median mass near 10 g correlation over a considerable size range
Figure 3.20. Orthodontic brackets are one of the more popular Geometric Summary
MIM products that have small mass, in the 0.1 g range Summary statistics on the design window are given in
(components courtesy of Ortho Organizers) Table 3.1. These statistics are based on the number of designs,
showing the median MIM design is just over 25 mm (1 in.) in
maximum dimension and 10 g in mass.
In this sample set some individual components are worthy
of mention. Lager components are in production, such as the
38
Table of Contents
Chapter 3: Design Principles
200 mm diameter (8 in.) compressor disk shown in Figure to that for investment casting and machining. As a result,
3.23, but these are specialized, lower-production-quantity MIM of smaller turbocharger units from nickel alloys is well
structures. One medical device was 193 mm (7.5 in.) in length advanced, but the larger units are restricted to high-tempera-
with 43 g mass. On the other hand, a large component at 1,097 ture ceramics such as silicon nitride, where there is no casting
g (2.4 lb.) has a maximum dimension of 137 mm (5.4 in.). option.
Even larger bodies are used in very limited quantities. One Thus, to summarize the geometric design window, the
pilot effort was devoted to bronze filters used in forming egg quantitative attributes of a successful MIM concept tend to be
cartons out of recycled paper pulp. That was targeted at 400 as follows:
per year at 40 kg each. Handgun bodies with mass ranges up low effective density where over 75% of the starting
to 500 g are in production. But, to be clear, most MIM designs mass would be lost during machining
are smaller devices that are lower in mass. surface finish of 0.2 m or rougher; matte finishes or in-
Related to the metals, ceramic injection molding tends to tentional embossing or texturing are best, smoother sur-
favor larger components. For example, ceramic pouring faces require post-sintering polishing
spouts used in steel processing reach up to 17 kg (37 lb.). Also the typical component in production ranges from a size
porous ceramic filters for aluminum casting tend to be large. of 1 mm (0.04 in.) to 125 mm (5 in.) and generally the
However, many factors work against injection molding large maximum dimension averages near 25 mm (1 in.)
structures in general, and specifically from metals. The tooling wall thickness tends to be small, with 3 mm (0.125 in.)
is more expensive as the component size increases. Further, being typical, but production capabilities exist from 0.1
the processing equipment is larger and more expensive, re- mm to 25 mm wall thickness
quires more floor space, runs slower, and incurs a higher de- wall thickness variation tends to be small, with a maxi-
preciation charge. Most important, for large structures the ebb mum-to-minimum wall-thickness ratio in a single com-
and flow of heat requires more time during molding, debind- ponent averaging 2.3
ing, and sintering; thus, process cycles slow as size increases. the typical mass in production is 10 g, but ranges from
The net effect from these several factors is a significant in- 0.02 g to almost 1,000 g (2.2 lb.)
crease in component cost as size increases. Consequently, cost
factors make MIM economically attractive for higher produc- FEATURES
tion quantities and smaller components. For example, an op- Certain features and feature combinations are well aligned
portunity given considerable attention is the production of with MIM. Early identification of good matches ensures tech-
turbochargers. For small automotive units the MIM approach nical and economic success. Table 3.2 lists examples of al-
is 40% lower in cost vs. investment casting and machining. lowed and desirable features, as well as features that require
But for larger, diesel truck units the cost by MIM is close special efforts. The balance of this section gives details to
these points.
Flat Face
A flat surface on the MIM design is useful to provide sup-
port during sintering. For many parts, the component weight
exceeds the low strength that exists during debinding and sin-
tering, resulting in distortion or even cracking. In measure-
ments during sintering the component is similar to artist clay
in strengthso small features tend to move and uneven sur-
faces tend to flatten.
If a flat surface is not possible on the design, then two op-
tions exist to avoid distortion. The first is to form conformal
setters to hold the component during thermal processing. Each
setter is fabricated from ceramic specifically to the component
geometry, so this is an additional manufacturing expense. An-
other option is to add intentional features for stability during
debinding and sintering, possibly with an added machining
step to remove the added feature after sintering. In some cases
splines or stiffeners not critical to the design are incorporated
into the body simply to provide strength or support. If possi-
Figure 3.23. Injection molded compressor disk with outside diame- ble, they are located where the shape is stabilized, but where
ter of 200 mm (8 in.) (picture courtesy of Ford Motor Company) the added feature need not be removed. For example,
39
Table of Contents
Chapter 3: Design Principles
Figure 3.24 shows the underside of a cobaltchromium indus- then the need for a custom-designed conformal setter is
trial housing with added stiffeners. Such an added feature avoided.
helps avoid warpage while being located in a noncritical spot.
If a flat face is added to provide support during processing, Closed Features
In recent years, what were previously considered to be de-
sign restrictions for MIM have led to design or process innova-
TABLE 3.2. NOMINAL MIM COMPONENT DESIGN
CONCEPTS tions. Closed internal cavities are now possible. One route is to
mold over a sacrificial plastic shape. During solvent debinding
restrictions the plastic is dissolved out of the green body to leave a hollow
inside closed cavities or undercuts on internal bores cavity. Figure 3.25 captures such an example, where stainless
require two pieces steel feedstock is molded over a sacrificial plastic insert. The
must have round corners and slight draft angles plastic then forms the circular internal cavity inside the stainless
(0.5 to 2) for ejection
corner radius larger than 0.075 mm (0.003 in.) steel. The plastic is dissolved out during debinding to produce
features (wall thickness, holes, etc.) generally must be the feature evident in the cutaway on the right of this figure. An
larger than 0.1 mm (0.004 in.) alternative is to mold the component in two pieces and then join
always below 20 kg (44 lb.) mass and usually below the parts either as green bodies or after sintering.
0.5 kg (1 lb.) Progress has been made on constructing ball-socket-type
desirable features moveable joints using two different feedstock solid loadings.
largest dimension typically below 125 mm (5 in.) The conceptual steps are outlined in Figure 3.26. At first a
gradual section-thickness changes feedstock with more powder and less binder, termed a higher
mass below 100 g (0.2 lb.) solids loading, is used to form the outside geometry, in this
wall thickness typically less than 10 mm (0.4 in.), case in two pieces. The high-solids-loading feedstock gives
most desirable is 3 mm (0.125 in.)
flat surfaces to provide support less shrinkage in sintering. A second feedstock with a lower
solids loading is used to fill the internal cavity. Because of the
allowed design features low solids loading, this portion shrinks more in sintering. The
shape complexity such as blind and flat bottom holes combination then produces a ball inside a socket with room
noncircular holes for the ball to move but not pull out of the socket.
holes at angles to each other
undercuts Related ideas consider how to form undercut features on
tapers an internal bore. The simplest concept is to use a sacrificial
ribs and protrusions plastic mounted on the core pin. After molding, the sacrificial
knurled and waffle patterns, identification number or part is dissolved out as part of the solvent debinding cycle.
logo, and surface textures Previously this was one of the excluded features in MIM,
helical threads, internal threads, and external threads if
needed but with diligence it now can be attained, at a cost, but no
longer is out of reach.
40
Table of Contents
Chapter 3: Design Principles
Small Details
Sharp corners are usually added by machining or grinding
after sintering. So scissors, scalpels, and knife blades require
post-sintering grinding. However, very tiny features are pos-
sible; for example, Figure 3.27 is a scanning electron micro-
graph of microminiature pillars formed by MIM. Note the
corners are rounded, a natural consequence of sintering. The
equipment and tooling required to form these small features
come from semiconductor fabrication, namely, lithography
and laser ablation to form the tool cavities.
Surface embossing is a related small-feature idea, where the
use of finely detailed tools and small powders leads to mi-
crochannels. A few examples are shown in the scanning elec-
tron micrographs of Figure 3.28. Microarrays are potentially a
Figure 3.26. Schematic outline on how a ball-in-a-socket joint is very large and significant area for MIM, as small genetic-se-
fabricated using MIM: (a) the half cavity is molded from a high- quence tests and blood-chemistry tests become widely applied.
solids-loading feedstock, (b) two half cavities are bonded to form
the hollow chamber, (c) the lower-solids-loading feedstock is
molded into the hollow chamber, (d) after sintering the inner Holes, Slots, Undercuts
material has more sintering shrinkage and can swivel in the Metal powder injection molding enables a wide variety of
chamber hole, slot, and undercut geometries. Undercuts on the outside
41
Table of Contents
Chapter 3: Design Principles
42
Table of Contents
Chapter 3: Design Principles
Figure 3.33. Rifle sight fabricated from steel powder. One version
on its side is prior to black oxidation treatment. The complexity of
this MIM part would necessitate multiple machining operations Figure 3.36. Example of a high-tolerance, high-reliability Kovar
and the ability to avoid those steps makes MIM cost competitive electronic package with glass-metal sealed lead wires (component
(component courtesy of Remington Arms) courtesy of Hunan Injection High Technology)
43
Table of Contents
Chapter 3: Design Principles
especially aided by using D shapes to allow for more seating desirable, so design changes can be made to better match the
area and stronger tooling at the intersection. MIM process. As an example, component redesign options
A surprise in design comes from added features that nor- focused on a constant wall thickness are illustrated for some
mally are expensive if made by machining. In MIM, the cost typical situations in Figure 3.40. Sketches are given for meet-
increases with component mass due to longer cycle times and ing the design objective while improving the wall uniformity.
more powder consumption. Thus, coring a design and adding The best option in each case is a design with constant wall
extra holes is desirable in MIM but not in machining. Already thickness and simplified tool design. The key point is that
the point has been made that nearly constant wall thickness is holes, cores, and various efforts to reduce wall thickness are
not detrimental or expensive in MIM.
44
Table of Contents
Chapter 3: Design Principles
Figure 3.43. Textured waffle pattern on a MIM component Figure 3.45. Protrusion options possible in MIM
45
Table of Contents
Chapter 3: Design Principles
Protrusions are useful in forming assembly attachment de- ing tends to be most cost effective for applications where high
tailsfor staking, cold heading, threading, or alignment. To performance is required, especially in medical or dental de-
sustain functionality, while easing molding, Figure 3.45 illus- vices, defense and aerospace systems, sporting goods, indus-
trates three different protrusion designsa stabilized design trial components, telecommunication devices, hand tools,
with support ribs, a tapered design with constant wall thick- instruments and sensors, cutting tools, automotive engines,
ness, and a dead-end design. The component shown in various electronic packaging, or marine equipment. This implies the
views in Figure 3.46 shows several protrusions that take ad- material is valued for delivering the required performance.
vantage of MIM to create functional details. Reinforcing ribs
are effective in providing alignment and support during de-
binding and sintering, so some are added to help preserve
geometry if there is no interference with function.
Threads
Both internal and external threads are possible in MIM. It
is most fruitful to add partial threads that do not go all the way
around the component. A flat surface on the parting line is
characteristic of a simple tool motion on a thread. The concept
is illustrated in Figure 3.47. Actual threaded features are
shown on MIM components in Figure 3.48 (external threads
with flats on the parting line) and Figure 3.49 (internal threads
added via motorized unscrewing motions).
The expense for including threads comes from the required
motorized unscrewing motion that must be added to the tool-
ing. It is much easier to form external threads where a flat sur-
face is allowed on the parting line.
From this basis it is possible to sort out good candidates
Figure 3.48. Stainless steel surgical device with a flat on the
for metal powder injection molding based on a series of tests. threads corresponding to the parting line (component courtesy of
Based on these several tests it is possible to determine if a MIM Ecrisa)
component design is suitable for powder injection molding.
Cost is the next consideration. Metal powder injection mold-
46
Table of Contents
Chapter 3: Design Principles
MATERIALS nickel particles are mixed and during sintering they form
The component material must be available in the small par- an alloy
ticle sizes needed for sintering densification. Commercial fully precompounded or prealloyed powdersfor
powder production tailored to MIM is a relatively new activ- example, each particle is the same composition with the
ity. Even so, a wide variety of materials are available in the desired ratio of ingredients
small particle sizes needed by the process. Globally, hundreds hybrid systems where a partially compounded master
of firms fabricate metal powders, but usually only a few com- alloy is mixed with an elemental powderfor example,
pete for any given composition, particle size, and chemistry. a chromiumnickel alloy powder is mixed with an iron
For example, nearly 50 firms make titanium powder, but powder to form stainless steel during sintering
spherical alloy powders in the size and purity desired for MIM Often the precompounded powder is easier to process, but
is only available from three firms. it is more costly.
In MIM, the bulk of the sintered products fall into a few There are many available compositions; however, there are
categories: also some limitations. For metals, there is difficulty with
ferrous alloyssteels, stainless steels, tool steels, iron strong oxide formers, so metals such as magnesium are
nickel magnetic alloys, and specialty ferrous alloys such avoided. Likewise, toxic metals such as lead are avoided. An-
as Invar and Kovar other limitation comes with low-melting-temperature materi-
common oxide ceramicsalumina, silica, and zirconia, als, since they are easily fabricated by routes such as blow
and mixtures of these molding, thixomolding, or die casting. Generally, a metal that
tungsten alloysboth tungsten heavy alloys and tung- melts at temperatures over 1,000C (1,832F) is most suc-
stencopper cessful, since options like die casting are not available at these
cermets (such as Fe-TiC), cemented carbides (WC-Co), temperatures.
and related hard materials
Besides these popular materials, there is a rapidly emerging PROPERTIES
activity in special materials, which include aluminum, pre- More than any other attribute, MIM product testing em-
cious metals, titanium alloys, nickel, nickel-base superalloys, phasizes mechanical properties. For metals, the tensile prop-
molybdenum, molybdenumcopper, and nonoxide ceramics erties of yield strength, ultimate tensile strength, and
such as silicon carbide and silicon nitride. elongation are primary concerns. The transverse rupture
Modified chemistries have emerged to take advantage of strength (sometimes called the rupture modulus) is usually the
the technology. For example, in stainless steels, the 316L com- focus for ceramics and cemented carbides. Some work has
position (Fe-19Cr-9Ni-2Mo, in weight percent) is used fre- been performed on creep, fatigue, fracture toughness, and in
quently, because of its combined strength and corrosion a few cases high-temperature tensile properties.
resistance. Compositions with more chromium allow for more Most MIM products are sintered to near 100% density, so
corrosion resistance and easier sintering, especially composi- the load-bearing cross section in the material gives strength
tions that have high levels of molybdenum (up to 6%), about the same as seen with other fabrication routes. Thus, for
chromium (up to 22%), or silicon (up to 3%). Although more MIM the densities are close to that obtained in cast products,
expensive, they provide properties superior to that attainable in the 96% to 100% range.
via casting, forging, or machining. The tensile properties attained in MIM converge to the
Powder availability is a consideration. However, with same range customarily found in engineering handbooks. For
many vendors of small powders almost any material can be example, a direct comparison of MIM nickel-base superalloy
custom fabricated for MIM. For complex chemistries, there (IN718) tensile properties is given vs. cast and forged material
are three possible starting forms: in Table 3.4. All of the test samples were given the same heat
mixed powder compositionsfor example, iron and treatment to develop a high strength. The two entries for MIM
47
Table of Contents
Chapter 3: Design Principles
material are without and then with a post-sintering hot isosta- and machined). Likewise, the elongation to fracture is com-
tic pressing (HIP) cycle. This extra step closes any residual parable between MIM, cast, and wrought.
pores by the simultaneous application of heat and pressure, To give more details, Table 3.5 compiles the alloy property
but in this case gives only minor property gains. Note how the ranges for the three different processing conditions. Because
MIM product exceeds the cast material and rivals the wrought sintering is an annealing process, MIM 17-4 PH stainless steel
material. Since each property has some testing variation, it is in the as-sintered condition has an annealed strength. Avoiding
necessary to design at the lower property levels. For MIM the heat-treatment step lowers costs, and for many applica-
products the variation in properties is small so the design level tions this still provides satisfactory properties. However, op-
is relatively high because of the reduced property scatter. timal properties require post-sintering heat treatment. As with
Another example comes from a common surgical stainless many alloys, post-sintering heat treatments allow manipula-
steel, 17-4 PH (AISI 630). It is widely used in medical and tion of the strength and ductility.
aerospace applications. It nominally consists of 17% Cr, 4% In a few situations the MIM products fall short of the hand-
Ni, and 4% Cu. When sintered and heat treated to the H1025 book properties, such as for steels that normally obtain strength
condition (aged at 552C or 1,025F), the MIM product typi- via deformation, as by forging. Consider that the sintering step
cally has a yield strength of 980 MPa (142 ksi). Detailed tests produces an annealed microstructure; without post-sintering
between different vendors give some variation from a low of heat treatment the sintered materials tend to be more ductile
965 MPa (140 ksi) to a high of 1,040 MPa (151 ksi). As cap- but with a lower strength. From an atomistic view, the forged
tured visually in Figure 3.50, when compared to a cast prod- product has a very high dislocation density that is missing in
uct, this range generally is positioned higher than the the sintered product, so normally the MIM product is high in
comparable range seen in wrought products (hot deformed ductility while the forged product is high in strength. It is pos-
sible to compensate for the strengthening effect via changes in
the alloy composition. One good illustration would be MIM
steels with nitrogen strengthening as obtained from the sinter-
ing atmosphere. The novel addition of nitrogen delivers
strength in the 2 to 3 GPa (280 to 420 ksi) range.
To emphasize the key point, Figure 3.51 plots yield
strength of MIM products including aluminum, copper, iron,
low-alloy steels, stainless steels, magnetic alloys, titanium al-
loys, tungsten alloys, nickel-base superalloys, electronic al-
loys, and tool steels. The MIM yield strength is on the y-axis
and the handbook yield strength for the same composition and
heat treatment is on the x-axis. Since MIM is a net-shape tech-
nology, the yield strength is a most important design parame-
ter. The plot shows a strong correlation between the MIM and
handbook strengths, with a statistically significant correlation
of 0.953. Note that 40% of the alloys are stronger as MIM
Figure 3.50. The yield strength ranges for heat treated 17-4 PH products. The standard error of the estimated MIM strength
(AISI 630) stainless steel fabricated by three routespowder
injection molding, casting, and machined wrought material. These is only 104 MPa (15 ksi) if the handbook strength is known,
plots show the high and low of the range of reported yield strength a reasonable scatter considering differences in impurities, ven-
for the three process options dors, or microstructures.
TABLE 3.5. COMPARATIVE ROOM-TEMPERATURE TENSILE PROPERTIES FOR 17-4 PH STAINLESS STEELS
(AISI 630) IN THE H1025 CONDITION
(range of several reports for each)
property MIM product cast product wrought product
yield strength, MPa 965 to 1,040 830 to 1,000 890 to 1,105
(ksi) (140 to 151) (120 to 145) (129 to 160)
ultimate tensile strength, MPa 1,030 to 1,190 890 to 1,070 1,030 to 1,180
(ksi) (149 to 173) (129 to 155) (149 to 171)
hardness, HRC 36 to 38 37 to 38 38 to 39
elongation to fracture, % 8 to 12 6 to 11 4 to 12
48
Table of Contents
Chapter 3: Design Principles
Similar correlations exist between handbook properties and In addition to mechanical properties, other properties such
sintered properties for brittle materials. There is no significant as machinability, thermal conductivity, wear resistance, and
difference from the handbook rupture strength when the com- corrosion resistance are competitive for MIM products. Where
positions are injection molded and sintered to near-full density special properties are desired, some of the most successful
using MIM. Figure 3.52 correlates the fracture strength (also MIM materials are those that prove difficult to fabricate via
termed the transverse rupture strength, measured in three- alternative routes.
point bending) for several ceramics and cemented carbides as Chapter 5 provides much detail on the thermal, corrosion,
a comparison of sintered properties with handbook values. In- mechanical, magnetic, electrical, and other properties of MIM
cluded in this plot are data from alumina, zirconia, silicon car- materials. In general terms, the MIM products nominally
bide, silicon nitride, and various cemented carbides. A perfect match handbook properties.
correlation would correspond to the straight solid line. About
50% of the molded compositions are stronger than the hand- DESIGN SUPPORT
book entries. The correlation between the fracture strength The MIM process excels in the production of complex
and the handbook is 0.955, which is statistically very signifi- geometries. Early identification of matches between a design
cant. In other words, if the handbook properties are known for and the MIM process best ensures success. To understand the
a brittle material, then injection molding can match those decision points a few key considerations were listed earlier
properties. and several photographs have already emphasized these
points.
The metal and ceramic powder injection molding opera-
tions live and breathe the manufacturing details. These pro-
ducers are aware of technology variants that can benefit a
customer, so they advocate for design changes to improve
yields and lower costs. For example, a long molding cycle
adds to the component cost; roughly two-thirds of the molding
expense depends on the cycle time. A design change that
shortens the molding cycle time, such as reducing wall thick-
ness, is beneficial since it makes the component lighter and
significantly lower in cost. The injection molding vendor will
provide guidance on assembly, part combinations, and mate-
rials. Likewise, with respect to tool cost, the mold can be ma-
chined with a single cavity, two cavities, and so on. Each
cavity adds to the tool expense that must be amortized over
Figure 3.51. The handbook yield strength is compared with the the production run. However, multiple cavities reduce the
value attained via MIM for several common alloys, showing the variable cost associated with molding. Guidelines exist to help
handbook property is a good predictor of that attained via metal optimize tool cost vs. component cost decisions.
powder injection molding
As treated later in Chapter 6, costs in powder injection
molding reflect the component size, tooling complexity, mold-
ing time, debinding and sintering speed, and other relatively
straightforward factors. While many shapes are possible, cer-
tain features greatly impact fabrication ease. Here are a few
examples:
High-vapor-pressure species are difficult to control so it
is appropriate to avoid materials that evaporate during
sintering; for this reason brass is avoided (since the zinc
evaporates during sintering).
Certain chemistries are used frequently, resulting in
lower powder prices, so consultation on materials selec-
tion can result in better properties at a lower price.
Relying on standard equipment enables a variety of ven-
dors to compete on a project, reducing costs.
Figure 3.52. Like the prior correlation for metals, this plot
Gradual changes in section thickness along the flow path
compares handbook fracture strength for ceramics with the ensure uniform final dimensions and are more desirable
measured value attained by CIM vs. abrupt changes in section thickness.
49
Table of Contents
Chapter 3: Design Principles
Holes and cores can be used to reduce mass and cost, and powder might dramatically reduce tool wear, improve
at the same time help create uniform section thicknesses, product quality, and generally prove a better choice.
such as illustrated in Figure 3.53. Molding blemishes can be located in noncritical loca-
Feedstock must be able to access all of the component tions to eliminate any expensive post-sintering finishing
details; for example, intersecting circular holes are very steps.
troublesome, so making one of the holes D shaped For improved dimensional control, the addition of a flat
provides a flat surface for proper sealing as sketched in surface to provide support during thermal processing (re-
Figure 3.54. moved after sintering if necessary) helps reduce distortion.
Mold-filling simulations are used to ensure all features Tight tolerances greatly add to the cost and reduce
are properly filled. process yield, so geometric tolerances need to realisti-
Feedstock must properly cool at the end of the molding cally balance function against cost.
cycle; thick sections have poor heat transfer and prolong Some features can be included in the part at almost no
the molding cycle, thereby adding cost; coring is one extra cost (e.g., part identification numbers); consultation
means to help improve the cooling rate. with an injection molding engineer helps attain the in-
At the end of the molding cycle the component must tended function with the least cost.
eject from the tooling, so undercuts, threads, and textures Secondary operations can be a lower-cost means to meet
can be added, but in certain locations they increase cost specifications vs. suffering a low process yield with a
considerably. high number of tight tolerances.
Mold life depends on many factors, including the powder Quality and ability to inspect critical features can be de-
chemistry, particle shape, binder, solids loading, and signed into both the part and process.
molding conditions; in some cases a more expensive Typically a few features dominate cost, so early discus-
sions with a MIM vendor at the concept stage helps avoid
difficult constructs.
RESOURCES
M. F. Ashby, Materials Selection in Mechanical Design, sec-
ond edition, 1999, Butterworth-Heinemann, Oxford, UK.
T. Baba, T. Honda and H. Miura, Effects of Homogeneous
and Heterogeneous Microstructures on the Fatigue Properties
of 4600 Steels Produced by MIM Process, Journal of the
Japan Society of Powder and Powder Metallurgy, 1997, vol.
44, pp. 443447.
S.R. Collins, Corrosion Resistance of MIM 316L, Advances
Figure 3.53. This comparative sketch suggests how holes and in Powder Metallurgy and Particulate Materials2002,
cores can be used to reduce mass and component cost, and at the compiled by V. Arnhold, C.-L. Chu, W.F. Jandeska, Jr., and
same time help create uniform section thickness H.I. Sanderow, Metal Powder Industries Federation, Prince-
ton, NJ, 2002, part 10, pp. 240254.
G.E. Dieter, Engineering Design, 2000, McGraw-Hill,
Boston, MA.
R.M. German and A. Bose, Injection Molding of Metals and Ce-
ramics, 1997, Metal Powder Industries Federation, Princeton, NJ.
H.O. Gulsoy, S. Ozbek and T. Baykara, Microstructural and
Mechanical Properties of Injection Moulded Gas and Water
Atomized 17-4 PH Stainless Steel Powder, Powder Metal-
lurgy, 2007, vol. 50, pp. 120126.
D. Shetty, Design for Product Success, 2002, Society of Man-
ufacturing Engineers, Dearborn, MI.
Figure 3.54. Intersecting holes are more successful in MIM when
the feedstock does not penetrate into the intersection. Shown on the P. Suri, B. P. Smarslok and R.M. German, Impact Properties
right is a comparison of circular and D shaped holes, showing of Sintered and Wrought 17-4 PH Stainless Steel, Powder
how the flat surface gives proper sealing Metallurgy, 2006, vol. 49, pp. 4047.
50
Table of Contents
Chapter 4
MANUFACTURING CONSIDERATIONS
MOLD ATTRIBUTES AND LIMITATIONS tween MIM vendors. This variation is due to differences in
The forming mold for MIM has a large influence on the equipment, binder, and powder. Mold design and fabrication
product dimensional capability during production. Prior to are treated in detail in plastics molding handbooks. In metal
molding, the feedstock lacks size and shape specificity. Once powder injection molding there are additional concerns be-
the component is ejected from the tool there remains little ca- yond plastic moldingnamely, heat transfer and feedstock
pability to adjust dimensions except with extra cost. So the density. The metal particles make the binder much more ther-
mold cavity inherently sets a limit on the dimensional control mally conductive when compared with plastics. Likewise, the
of the sintered product. An open mold is shown in Figure 4.1; metal particles increase the feedstock density significantly
the tooling for MIM looks similar to that used in plastic in- over typical plastics. Otherwise, the metal powder injection
jection molding, although molds are oversized to account for molding route uses similar principles as found in plastic mold-
the large sintering shrinkage. In most instances, the tool set ing. Whole books are available on the latter subject. A few
consists of one to four cavities serviced by tool pieces that references are given at the end of this chapter.
move in coordination with the opening, molding, and closing
actions, such as ejector pins. Cavity Design
Tool design and process design go together and vary be- The mold cavity is the heart of the tooling. It is dilated to
51
Table of Contents
Chapter 4: Manufacturing Considerations
1 .
Z = (4.2)
1Y
LO = LFZ. (4.3)
[ ]
Y = 1
/T
1/3
(4.1)
52
Table of Contents
Chapter 4: Manufacturing Considerations
as standard pre-machined pieces, so only the mold cavity These factors result in dimensional variation that decreases
needs to be custom machined. process yield and drive MIM toward single cavities for very
Within a mold it is possible to have multiple cavities. The precise products.
number of cavities in use across the MIM industry ranges up
to 40, and in one rare case is even higher. A single-cavity tool Parting Line
is satisfactory for low production quantities, say, in the range The location of the parting line is an important decision in
of 100,000 shots per year. Optimization calculations show that MIM tool design. The parting line is the trace left on the
for components ranging from 1 to 100 g, the lowest project molded component surface where the mold sections meet, as
cost (tooling and piece cost) results in the following general- illustrated in Figure 4.3. Cosmetic and functional require-
ization: ments may force parting line locations in less than optimal po-
for production below 100,000 parts per year1 cavity sitions. Part of the problem is that the tool set slightly deflects
for production quantities near 300,000 parts per year during mold filling due to elastic strain, in spite of close align-
2 cavities ment and high clamping forces. The deflection varies directly
at 1 million parts per year4 cavities with the injection pressure and part cross-sectional size, al-
at 3 million parts per year8 cavities lowing slight feedstock intrusion between the moveable tool-
at 10 million parts per year12 cavities ing plates.
For larger parts, the number of cavities is limited by the Although the parting-line mark is a common blemish, an
molding machines clamping force or shot size, which are not extreme situation occurs with feedstock intrusion along the
critical issues for a small component. parting line to give flashing. Figure 4.4 is an extreme example
The savings from multiple cavities comes about due to in- of flashing. Note the corner above the flash also has an un-
creased molding productivity achieved by reducing the re- healed weld line. Removal of flash after molding is costly and
quired number of molding machines. As the number of is avoided when possible. On the other hand, in regions where
cavities increases, the cost of manufacturing the mold in- labor is inexpensive it is common to have a technician at the
creases, but the subsequent molding cost per component de- molding machine perform flash removal by hand. An alterna-
creases. However, the optimal quantity of cavities may not be tive is to place the parting line in a region where the blemish
determined solely by costs. Other limitations include clamp- is not of concern. A few MIM firms have adopted novel
ing force, shot volume, maintenance costs, and added compo- grooved tool designs that remove the parting line.
nent dimensional variability. For example, with more than a
few cavities, a 100 g component might exceed the clamping Sprue, Runner, and Gate
force. Another factor is nonuniform filling. For very tightly The sprue connects the molding-machine nozzle to the run-
toleranced components, statistical tests show that the cavity- ner network in the mold that delivers molten feedstock into
to-cavity size effect is measurable. Further, the cavity location the cavity. In many MIM operations sprues are tapered with a
with respect to gravity and cooling channels contributes to typical diameter near 6 mm and 5 degree taper. When the
sintered dimensional variation in multiple cavity tooling. mold opens for ejection, the sprue is pulled from the sprue
bushing and usually is recycled. For faster production and the
minimization of regrind, the sprue and runner are kept hot to
avoid reheating and recycling this portion of the feedstock.
This option is called a hot runner system.
From the molding nozzle the sprue leads into the runner,
which then directs the melt into the mold cavity through the
gate. Small runners cause slow filling, so a large runner is de-
sirable, with a typical diameter from 3 to 6 mm. Of course, a
large runner means relatively less of the shot goes into the part
cavity. A circular runner design is most desirable since this re-
duces friction and heat loss during filling, but other designs
are less expensive to construct. Hot runners are available to
keep the flow path heated in the tooling. They use a valve in
the gate to seal the runner between shots and are usually re-
stricted to high-volume production where the additional cost
is justified by productivity gains.
Figure 4.4. This plate-shaped microelectronic package exhibits When a mold contains multiple cavities, balanced runner
flashing along the parting line. Note the crack in the corner above designs with an equal flow length to each cavity are required
the flash to ensure simultaneous filling of each cavity. Two factors are
53
Table of Contents
Chapter 4: Manufacturing Considerations
requiredfirst, the flow length from the sprue to gate must Additionally, gate placement should consider minimizing
be the same for each cavity and, second, the cooling channels weld lines. Weld lines form when the feedstock splits and then
must be balanced to ensure similar cooling temperatures along rejoins after flow around posts during cavity filling. If the
each flow path. A valve can be incorporated into the runner molding conditions are such that the feedstock is warm, then
system to isolate any cavities that might become defective the weld line heals. Alternatively, if the feedstock is too cool,
during production. then the weld line does not heal and a defect forms. Figure
At the end of the runner is the gate leading into the mold 4.7 offers two pictures of weld-line defects; one is a scanning
cavity. The gate leaves a surface blemish on the part, such as electron micrograph showing such a subtle molding defect in
shown in Figure 4.5. It is a small opening designed to freeze a green body and the second is an optical micrograph showing
before the part, runner, or sprue freeze. Ideally, the solidified how the defect is enlarged after sintering. A defective weld
gate sustains pressure in the cavity to counteract the shrinkage line is difficult to detect after molding, but invariably it is ev-
associated with feedstock cooling and binder contraction. Too ident after sintering, since the material pulls away from the
low a pressure when the gate freezes results in voids or sink crack. Computer simulations help avoid these difficulties by
marks, while excessive pressure causes sticking in the mold. suggesting either different molding parameters or different
It is desirable to have the gate located on the thick portion of gate locations.
the component. This reduces heat loss and premature freezing
or short shots, and lowers the pressure required for mold Vent
filling. When a molding cycle starts, the mold cavity is filled with
Gate size is determined by the filling speed and the com- air. During molding this air is pressurized by the intruding hot
ponent thickness. It is desirable to have the gate freeze before feedstock and forced to escape through a vent. The vent is lo-
the compact freezes. This requires a gate size smaller than the cated at the end of the mold cavity opposite the gate. It should
component wall thickness. Yet, if the gate is too small, then be the last portion of the cavity to be filled. Vents are very thin
the resulting higher feedstock, flow velocity during cavity fill- reliefs typically 0.015 mm (0.0006 in.) deep, and up to 12 mm
ing results in jetting, high tool wear, or powderbinder sepa- (0.5 in.) wide for large MIM parts. Ideally, the vent leaves no
ration. Figure 4.6 is picture of the powderbinder separation observable trace on the part surface.
around the gate on the bottom of a MIM electronic package, Venting is also necessary to allow the part to be ejected
most likely due to the gate being too small. On the other hand, from the tooling. For example, a shape with a closed end can-
if the gate is too large, then molding is slowed by the need to not be ejected without creating a vacuum; otherwise, an addi-
use the screw to control pressure on the component during tional ejector pin must be built into the forming tool which is
cooling instead of preparing the next shot. In such a situation pulled first to vent the mold.
a fan gate or even multiple small gates might be required. If An alternative is to rely on loose-fitting ejector pins to pro-
the gate is on a thin section, then filling is hindered because vide an opening for venting in the annular gap around the pin.
of flow resistance and feedstock cooling. Likewise, the blem- Another option is to use a stepped mold closure procedure,
ish from the gate impacts the decision on gate location be- where only a light clamp force is applied through the filling
cause of cosmetics. stage, allowing air to vent along the parting line. At the end
of the mold stroke the tool is fully clamped as the feedstock
is packed into the mold.
54
Table of Contents
Chapter 4: Manufacturing Considerations
Ejector
Once the part has cooled in the mold it must then be
ejected. The ejection force depends on the contact area be-
tween the tooling and component, tool surface finish, coeffi-
cient of friction, and thermal contraction in the cavity. Also,
adhesive polymer phases in the binder that wet and adhere to
the tool surface can cause sticking. A slight taper in the tooling
greatly aids in reducing ejection force and often just 0.5 is
sufficient. Corners in the cavity are usually rounded for easier
ejection; a radius of 0.2 mm (0.008 in.) is satisfactory, but they
(a)
can be as tight as 0.05 mm (0.002 in.).
To accomplish ejection, pins in the mold body move for-
ward with the ejector plate and push the component from the
cavity. Inserts, internal cores, or threads must be retracted to
allow free ejection. Ejector pins blemish the component, since
they concentrate the ejection force on a soft and often still
warm molded shape. Thus, larger pins are desirable to reduce
the stress. Figure 4.8 is a component with several visible ejec-
tor pin marks. The location and number of ejection pins de-
pend on the component size, binder strength, and tooling
complexity. Normally, the pins are placed to impress on non-
critical locations and constitute more than 10% of the pro-
jected compact area.
55
Table of Contents
Chapter 4: Manufacturing Considerations
defect rate and labor content in molding, so it is an option gen- inserted into the sprue bushing. After contact, a premeasured
erally restricted to low production quantities. quantity of molten feedstock is forced into the mold cavity.
Figures 4.10 and 4.11 are two examples of features gener- Pressure is sustained on the cooling feedstock until the gate
ated by slide motions. Protruding pins in the tool halves force freezes, and after complete cooling the component is ejected.
motion in the parting plane as the mold closes and opens. This Between the time when the gate freezes and the part is ejected
ensures proper placement of the slide in the cavity on closing while it is still cooling, the screw is turning to plasticate and
and retraction to avoid interference with ejection. The fabrica- meter the next shot of hot feedstock.
tion of slides, cores, or unscrewing tool components (for ex- The pressure in the molding cavity undergoes a rapid in-
ample, those required to form internal threads) are expensive crease as the hot, molten feedstock enters the cavity and that
aspects of tool fabrication. Thus, efforts to redesign the com- pressure is held until the gate freezes. Once the gate has frozen
ponent to avoid these features are rewarded by lower tool costs. the molding machine no longer controls the cavity pressure.
A pressure trace is plotted vs. time in Figure 4.13, indicating
TOOL MOTIONS the pressure loss from the hydraulic-system pressure as ap-
Molding starts with the closure of the mold cavity. As plied to the back of the screw, through the nozzle, and into the
sketched in Figure 4.12, the nozzle of the molding machine is mold cavity. It is the cavity pressure that needs to be con-
56
Table of Contents
Chapter 4: Manufacturing Considerations
trolled, but the critical pressure is controlled back at the tool walls and push the component from the cavity.
hydraulic system. Hence, effort is required to ensure proper Robot coordination is an important aspect of quality con-
hydraulic pressure, and corrections to the hydraulic pressure, trol and automation. Most modern molding machines integrate
to achieve repeatable mold-cavity pressures. The shot-to-shot the robot control and coordination with the molding cycle.
mass variation is the easiest means to monitor the success of Further, simple tests such as mass or size determination are
proper control. possible after molding to ensure proper quality control.
After mold filling, as the feedstock cools, a natural pressure A die-cavity-pressure transducer is important to molding
reduction occurs due to binder thermal contraction. At the control. This pressure-measurement device is buried in the
point of ejection the residual pressure in the cavity should be tooling at two-thirds of the flow length in the cavity. Once the
very low. Otherwise, the component will stick in the mold transducer detects pressure in the cavity, then molding ma-
cavity. For ejection, pins move from flush positions on the chine control reverts to repeat the same cavity pressure profile,
via concomitant adjustments to the hydraulic pressure. This
significantly improves uniformity in packing pressure, result-
ing in a tighter part-mass distribution, which results in more
uniform sintered dimensions. Figure 4.14 is a plot of die-cav-
ity-pressure traces taken over 50 shots. The upper plot corre-
sponds to molding without die-pressure control, showing a
wide variation in pressure (resulting in a wide density and sin-
tered dimensional variation). On the other hand, the lower plot
shows pressure traces for 50 shots with die-pressure control,
obviously giving a more homogeneous product. Reduced
dimensional scatter for the sintered component is the key gain
from closed-loop cavity-pressure control.
TOOL-CAVITY DESIGN
The molds used in MIM are custom machined from
wrought stock. Ease of fabrication and subsequent durability
of the tooling are primary concerns dictating tool material se-
lection. Because the MIM feedstock is more abrasive than
pure polymer, wear resistance is a concern. Consequently,
after machining, the tooling is heat treated or subjected to a
surface-hardening treatment. To avoid component adherence
to the tooling, surfaces should be smooth and scratch free.
57
Table of Contents
Chapter 4: Manufacturing Considerations
58
Table of Contents
Chapter 4: Manufacturing Considerations
59
Table of Contents
Chapter 4: Manufacturing Considerations
The greater the area used for ejection, the shallower the ejector pability has likewise improved over time. Gains in MIM are
indents. occurring as new production technologies are adopted, but as
long as older equipment and older manufacturing practices re-
GEOMETRIC-FEATURE COMBINATIONS main in use there will be a wide distribution in capabilities
Metal powder injection molding excels when the shape is across the industry. Efforts to hold close tolerances add to the
possible via plastic molding, but plastics lack the engineering production expense, but are offset by increased process yields.
properties needed for the application. In the MIM field, toler- Post-sintering machining is a means to generate MIM
ances and engineering specifications are a subject of much dis- products that match conventional machining tolerances. How-
cussion. As MIM has evolved, new process variants have ever, MIMs greatest cost advantage occurs with net-shape
contributed to substantial precision and accuracy gains. Every production, where the part is used directly out of the sintering
year the basic process knowledge grows, while the equipment furnace, without secondary dimensional corrections.
improves and the feedstock becomes more consistent. Conse- Design engineers tend to be puzzled by the wide variability
quently, production capabilities improve over time. Thus, there in cost estimated for MIM components. However, part of the
is no single process capability, rather a spectrum of capabilities variation comes from the fact that some facilities require coin-
exist that are improved by about ten years of experience com- ing or machining to hold tolerances while other facilities can
bined with modern computer-controlled processing equipment. hold tolerances without secondary operations. The latter ap-
Powder-based technologies have an excellent history of proach is always more cost-effective.
process improvements. A good example of the progressive
learning is seen in Figure 4.18, which plots the dimensional DIMENSIONAL TOLERANCES
variability for cemented carbide (WC-Co) cutting-insert pro- Since metal powder injection molding is a hybrid between
duction over a half-century. The variability is measured by the plastic molding and sintered powder metallurgy technologies,
coefficient of variation, a dimensionless means for expressing the tolerance capabilities are equally a hybrid. Plastic molding
scatter; it is the standard deviation divided by the mean. Such generally has a tolerance range from 0.05 to 0.5 mm
scaling allows comparison of capabilities between large and (0.002 to 0.020 in.). Sintered powder metallurgy technolo-
small components. In Figure 4.18, the data are for the cutting- gies generally have a tolerance range from 0.1 to 0.3 mm
edge distance from the central locating hole, showing progres- (0.004 to 0.012 in.), and sintered ceramics are much larger.
sive improvement over 50 years. Most of the improvements Accordingly, MIM is limited by its underpinningsplastics
occurred in the first 20 years, but even today there are gains, and sintered materialsto a general tolerance range centered
with production tolerances (without secondary operations) on 0.1 mm (0.004 in.) or slightly better with special efforts.
converging toward 10 m on a 14 mm distance (0.0004 in. In discussing tolerances there are two issues. The first is
out of 0.55 in.). the location of the mean size for the specified featureaccu-
Similar plots exist for other powder metallurgy technolo- racy. The second issue is the variation about the meanpre-
gies, although none have the history of cemented carbides. In cision. The further the mean is from the target size, the smaller
presssinter ferrous powder metallurgy, the dimensional ca- the allowed variation or tolerance zone. Adjustments to the
mean size are easier to accommodate in MIM than adjust-
ments to the variability. For example, changes in molding con-
ditions, solids loading, sintering temperature, or sintering time
can be used to relocate the mean size of the product.
Current MIM component production ranges from feature
sizes below 100 m to nearly 1 m (a size ratio of 104), making
it difficult to address tolerances directly over such a wide
range. In some instances, the choice has been to create a table
of tolerance vs. size, showing capabilities as follows:
0.05 mm (0.002 in.) for features below 3 mm (0.12 in.)
0.08 mm (0.003 in.) for features between 6 and 15 mm
(0.25 to 0.6 in.)
0.25 mm (0.01 in.) for features between 30 and 60 mm
(1.2 to 2.4 in.)
An alternative is to track the tolerance as a percentage of
Figure 4.18. Dimensional precision improves as cumulative the feature size. The coefficient of variation is used to normal-
knowledge grows. This plot gathers coefficient of variation data
from cemented carbide insert production over 50 years. By
ize dimensional variability. Statistically the coefficient of vari-
implication, MIM should show similar gains as technology, ation CV is defined as the standard deviation divided by the
materials, and equipment improve mean dimension, often given as a percentage:
60
Table of Contents
Chapter 4: Manufacturing Considerations
61
Table of Contents
Chapter 4: Manufacturing Considerations
was 0.7%, while the width variation was 0.4%, and the casting cores used for jet-engine components, the mass ranges
length variation was 0.2%. This was confirmed in a different from 10 to 1,200 g (0.02 to 2.6 lb.), and part lengths range up
study, giving a similar ratio between perpendicular dimen- to 500 mm (nearly 20 in.). These cores are produced with a
sions. Most important, mass variations between shots ac- few critical dimensions held to a coefficient of variation of
counted for 87% of the dimensional variation after sintering. 0.05% at production quantities over 50,000 per week. Un-
The thickness variation reflects pressure repeatability and the fortunately, there is only 80% process yield to these toler-
slight tooling deflection. When molding is performed with ances, so it is proper to say that the coefficient of variation of
closed loop cavity pressure control, the mass variation is much the shipped components is 0.05%, but the actual process has
reduced, resulting in smaller dimensional scatter. an inherent scatter of 0.12%. Other MIM companies report
Another factor in determining dimensional tolerances is shipped tolerances of 5 m and even 3 m on critical
surface roughness. Nominally the tolerance specification can- dimensions, but little information is provided on process
not be any tighter than tenfold the surface roughness. Most yields. Obviously, production cost increases rapidly with the
MIM operations deliver an average surface roughness near 1 tighter tolerances if yield falls.
m. In a survey of the industry, the mean is 0.6 m and the Inherently, tolerance capabilities are limited by surface fin-
standard deviation is 0.3 m, but the median (half better and ish. Accordingly, surface roughness is a concern, especially
half worse) is 0.8 m, so the distribution is asymmetrical. On as MIM has entered consumer and jewelry applications. The
this basis the surface roughness limits the tolerances to prob- sintered surface roughness ranges from 0.2 to 1.6 m or 8 to
ably 10 m at best for MIM. 64 in. This range matches with a ground finish. The limit
An option for tighter tolerances is to machine critical sur- has its origin in both the tooling surface finish and the particle
faces after sintering. Post-sintering machining is often used in size. The impact of particle size is trivial unless the tool is
MIM, ensuring precise final dimensions, but in some cases it smooth; without polished tooling, the part surface simply
adds considerably to the fabrication expense. replicates the tool surface topography, an aspect that is desir-
With newer binder systems and integrated process controls, able for textured component surfaces.
dimensional scatter has improved to a coefficient of variation When the tooling is polished, the particles still create a cob-
equal to 0.05%. For example, in the production of ceramic blestone surface as diagramed in Figure 4.20. To help envision
the as-sintered surface, a scanning electron micrograph of a
MIM stainless steel component is given in Figure 4.21. Note
TABLE 4.2. PROCESS VARIATIONS OBSERVED IN MIM the residual appearance of the particles, now fused together
COMPONENT PRODUCTION
(actual values depend on many site-specific and
by sintering. Thus, smaller (and more expensive) particles are
component-specific factors) needed to form smoother surfaces and this is only justified
when using polished tooling. For very smooth (polished)
feature best nominal or
possible typical
products, it is usually most economical to add a polishing op-
eration after sintering.
absolute dimension 0.025 mm 0.05 mm Likewise, the fabrication of sharp corners and small fea-
(0.001 in.) (0.002 in.)
tures is limited by the particle size. An illustration of this is
angle 0.05 2 given in Figure 4.22. Indeed, if the particles are large com-
density 0.2% 1% pared to the feature, then the particles simply will not fill out
the feature. As a rule of thumb, no feature should be specified
dimension 0.05% 0.3%
to a size that is not at least tenfold larger than the particle size.
flatness 0.1% 0.2% For a sharp corner fabricated from 10 m powder this would
hole diameter 0.04% 0.1% say that the corner radius should be 100 m or larger (0.1 mm
or 0.004 in.). Attempts to form sharp edges on cutting tools,
hole location 0.1% 0.3%
knives, or scissors using MIM have been unsuccessful since
parallelism 0.2% 0.3% the particle size limits the sintered edge retention. This is ev-
perpendicularity 0.1% or 0.1 0.2% or 0.3 ident by the rounded edge on a MIM drill tip as seen in Figure
4.23. Smaller particles are needed to produce small features.
roundness 0.3% 0.3%
surface roughness 0.2 m 0.8 m TOLERANCE APPROACHES
(8 in.) (32 in.) There is a high cost associated with tight dimensional con-
weight 0.1% 0.4% trol. Sorting good and bad components after sintering is one
corner radius 0.5 mm 0.3 mm
approach, where those out of tolerance are discarded or re-
(0.02 in.) (0.012 in.) worked, but the cost is high. More attractive is to take steps
to minimize dimensional variation during production, but that
62
Table of Contents
Chapter 4: Manufacturing Considerations
too comes with an increased cost. For example, custom setters 4.24 is an example of one such conformal substrate. Since the
can be fabricated for use during sintering, as is the case in pro- component shrinks during sintering, the substrate design must
duction of high-precision casting cores used in airfoil appli- provide support but not restraint. In some cases the substrate
cations. These setters are fabricated from high-temperature is a ceramic peg or grooved tray, yet it can become very com-
zirconia machined to the final contours with diamond grinding plicated and expensive.
techniques. Each injection molded body is then nestled into
the setter to ensure conformal support during sintering. In
other words, the MIM component deformation in sintering is
guided to a precise final shape cut into the substrate. Figure
63
Table of Contents
Chapter 4: Manufacturing Considerations
64
Table of Contents
Chapter 4: Manufacturing Considerations
65
Table of Contents
Chapter 4: Manufacturing Considerations
66
Table of Contents
Chapter 4: Manufacturing Considerations
(g) (h)
Figure 4.29. U-shaped MIM component showing (a) the component layout and the location of the sprue, runner, and gate, (b) filling
stage and velocity profiles simulation at half filling, (c) volumetric shrinkage of the packed part, (d) freezing distribution after the gate has
frozen, (e) outline of the cooling channel layout in the mold, (f) surface-temperature gradients on ejection, (g) green-density distribution in
the component, (h) predicted sintered size and picture of actual part after sintering (images courtesy of CetaTech)
engineering materials, where reference materials are available system-wide efforts to manage the total process via an
for calibration, MIM is without such baseline materials. Con- organizational quality policy
sequently, calibration and acceptance criteria are largely set by 2. Communication, Information, and Documentation
the customer and supplier on an ad hoc basiscases are Systems: communication systems to keep the organi-
handled one at a time. Because of this situation, acceptance test zation updated on progress
samples are exchanged between the producer and users. 3. Indicators: provide quantitative objectives and report
For success in this environment, the MIM fabricator must on progress
be proactive and plan the testing procedures, document and 4. Personnel: provide training and orientation to all
obtain acceptance of those procedures, and then conform to employees
the plan. A vendorcustomer plan documents what will be 5. Product and Process Development and Industrializa-
measured, how frequently, and with what techniques, and cap- tion: propagate quality standards at the product- and
tures samples that are retained as de facto standards. A full process-conceptualization stage
quality system for MIM requires an integrated view that 6. Management of Supplier Quality: suppliers provide
ranges from first customer interactions to post-delivery conforming materials, equipment, and products with
service. As an example, the following quality plan outline was identified quality levels, including records of incoming
provided by an automotive component supplier: materials
1. Organizational, Sales, and Administrative Systems: 7. Handling of Nonconformities: incoming material prob-
67
Table of Contents
Chapter 4: Manufacturing Considerations
lems and complaints from customers are documented For certain medical and aerospace components, assurances
8. Maintenance Plans: equipment and processes remain of proper sintering require witness tensile samples in each sin-
in conformity with set standards in a preventative- tering run. Although there are well established international
maintenance system metallurgical standards for tensile geometries, some MIM
9. Quality of Measurement: inspection devices are em- firms rely on rounded-dogbone shapes. Across the MIM in-
ployed to assure proper conformance to set standards dustry test-sample sizes range from 5 mm to 140 mm length.
10. Control Systems for Products, Processes, and Proce- Generally, for a 30 to 38 mm gauge length (1.2 to 1.5 in.), the
dures: the system looks to improve procedures to prop- sample thickness is near 5 mm (0.2 in.). The gauge length as
agate the quality system throughout the organization a ratio to the gauge diameter varies from 5 to 8, which is larger
11. Audits: both internal and external audits are performed than used in traditional metallurgy where the ratio is held to
to ensure the process is conforming to standards 4. For this reason ductility should not be directly compared
with traditional property listings, especially for materials that
INSPECTION neck prior to failure.
Metal powder injection molding has a variety of standards The final phase of MIM production is to perform quality
organized by the regional governing bodies in the U.S., checks to ensure delivery to specification. Many of the MIM
Europe, and Japan: Metal Powder Industries Federation, operations have registered their quality systemsISO 9001
European Powder Metallurgy Association, and Japan Powder for captive production and ISO 9002 for custom production.
Metallurgy Association. Firms have joined one of these Testing might be performed at either the fabrication or use
regional trade associations to help promote the field and to sites, and with appropriate quality systems the user will accept
write standards. the shipment without inspection. This was the situation in the
Data sharing is progressing between regions. Currently, disk-drive and cellphone industries, where millions of MIM
Powder Injection Molding International best serves as the components were delivered without inspection (nor problem)
conduit for market news and data. Another important compi- based on the MIM vendors quality system. Benchmarks in
lation is from Inovar Communications, which produces the the MIM industry are upwards of 50 million parts shipped
International Powder Metallurgy Directory in both print and without a defect. To achieve such quality levels requires col-
online forms. laboration to ensure the producer understands the functional
attributes and delivers only acceptable components. In pro-
Quality Tests duction, the goal is to capture errors as they occur rather than
In MIM, the input materials include the polymers and pow- after fabrication.
ders. The powder composition is treated by a range of stan- Detailed quality checks are associated with first-article
dards and nominally these are the same as encountered in submission, which includes data showing the measured fea-
traditional metallurgy. For example, the specification for 316L tures and the specified featuresdimensions, hardness, mi-
stainless steel comparison between the MIM powder and crostructure, chemistry, or other attributes. Once a component
wrought material are nearly the same. However, the minimum moves into production, the inspection criteria need to establish
mechanical properties differ slightly due to the allowed poros- several factorswhat should be measured, how it is meas-
ity in the MIM variant. Current specifications for composition ured, where it is measured during production, how frequently
and sintered properties include the following: it is measured, how many samples are used each time, how
low-alloy steels (three grades and two heat treatments for those samples are captured, the lot size associated with the
4605) measurements, and the allowed ranges. Each user industry
stainless steels (three grades and two heat treatments for typically has standards that guide implementation of these
17-4 PH) tests.
soft magnetic alloys (five grades and two impurity levels An important shop-floor technique is statistical process con-
on Fe-50Ni and Fe-3Si) trol (SPC). It is used at intermediate manufacturing steps to as-
controlled-expansion alloy (one grade) sess drifts from standard conditions. The goal is to keep the
In addition, various standards cover testing procedures for product in an accepted range throughout the MIM process.
mechanical properties, corrosion resistance, magnetic proper- Some variability occurs in any operation, but SPC continually
ties, wear testing, particle size, powder density, powder pack- senses drifts that may take a process out of the acceptable range.
ing, powder flow, and microstructure evaluation. Within the Many sources of defects are possible in MIM. A first cate-
MIM community unofficial tests have emerged, such as use gorization is given by the fishbone diagram in Figure 4.30.
of the plastic-melt-flow index for viscosity evaluation, use of Most of the problems come from molding or thermal process-
witness-slug-shrinkage measurements for sintering-tempera- ing. In a well-controlled process, the unanticipated defects can
ture-uniformity tests, and helium-gas pycnometry for deter- be reduced, since there are some obvious sources and cures.
mination of granulated-feedstock density. Others problems, such as parting lines and ejection pin blem-
68
Table of Contents
Chapter 4: Manufacturing Considerations
UM U A
CP = (4.8)
3
69
Table of Contents
Chapter 4: Manufacturing Considerations
D.M. Bryce, Plastic Injection Molding Mold Design and Y. Shengjie, Y.C. Lam, S.C.M. Yu, and K.C. Tam, Thermo-
Construction Fundamentals, 1998, Society of Manufacturing mechanical Simulation of MIM Thermal Debinding,
Engineers, Dearborn, MI. International Journal of Powder Metallurgy, 2002, vol. 38,
no. 8, pp. 4155.
P.S. Cracknell and R.W. Dyson, Handbook of Thermoplastic
Injection Mould Design, 1993, Blackie Academic and Profes- B. Smarslok and R.M. German, Identification of Design
sional, London, UK. Parameters in Metal Powder Injection Molding, Journal of
Advanced Materials, 2005, vol. 37, no. 4, pp. 311.
R.M. German and A. Bose, Injection Molding of Metals and
Ceramics, 1996, Metal Powder Industries Federation, Prince- M. Trunec and J. Cihlar, Thermal Removal of Multicompo-
ton, NJ. nent Binder from Ceramic Injection Mouldings, Journal of
the European Ceramic Society, 2002, vol. 22, pp. 2,231
T. Hartwig and T.V. Duong, Miniaturisation of MIM,
2,241.
Second European Symposium on Powder Injection Moulding,
2000, European Powder Metallurgy Association, Shrewsbury, B. Williams, Setting Standards and Getting it Right, Metal
UK, pp. 229234. Powder Report, 2001, vol. 56, no. 12, pp. 1418.
U. Haupt and H. Walcher, Powder Injection Moulding Con-
cepts for Presses and Tooling, Proceedings of the First
European Symposium on Powder Injection Moulding, 1997,
European Powder Metallurgy Association, Shrewsbury, UK,
pp. 7079.
J.L. Johnson and R.M. German, Design Capabilities of
Powder Injection Molded Materials, Journal of Advanced
Materials, 2005, vol. 37, no. 3, pp. 5159.
B. Julien, M. Ste-Marie, R. Pelletier, F. Lapointe and S.
Turenne, Parametric Modeling of MIM As-Sintered Proper-
ties of Stainless Steel 17-4 PH and 316L Using a Statistical
Model, Advances in Powder Metallurgy and Particulate
Materials2010, compiled by M. Bulger and B. Stebick,
Metal Powder Industries Federation, Princeton, NJ, part 4,
pp. 110.
V. Piotter, T. Benzler, R. Ruprecht and J. Hausselt, Manufac-
ture of Micro Sized Structures by MIM and CIM, Advances
in Powder Metallurgy and Particulate Materials1998,
compiled by J.J. Oakes and J.H. Reinshagen, Metal Powder
Industries Federation, Princeton, NJ, part 5, pp. 161168.
H. Rees, Mold Engineering, 1995, Carl Hanser Verlag,
Munich, Germany, 1995.
B.O. Rhee and J.H. Lee, Investigation of the Gate Effect on
the Powder Injection Molding Process, Proceedings of the
2000 Powder Metallurgy World Congress, Part 1, K. Kosuge
and H. Nagai (eds.), Japan Society of Powder and Powder
Metallurgy, Kyoto, Japan, pp. 278281.
D.V. Rosato and D.V. Rosato, Injection Molding Handbook,
1986, Van Nostrand Reinhold, New York, NY.
N. Salk and B. Troger, Precision Manufacturing of Micro
Ceramic Injection Molded Implants, Advances in Powder
Metallurgy and Particulate Materials2010, compiled by
M. Bulger and B. Stebick, Metal Powder Industries Federa-
tion, Princeton, NJ, part 4, pp. 1117.
70
Table of Contents
Chapter 5
MATERIAL PROPERTIES
AVAILABLE MATERIALS reports over 900 formulations in use of different alloys, pow-
Component fabrication by MIM is contingent on powder der types, and particle sizes.
availability. The good news is that a very broad range of com- The powders used in MIM come in a variety of chemis-
positions are available. For example, one feedstock vendor tries, particle sizes, and particle shapes. Most typically, the
71
Table of Contents
Chapter 5: Material Properties
powders are equiaxed, rounded, and smaller than 20 m in materials include maraging steel, tool steels, cobaltchromium
size, such as shown in Figure 5.1 (nickel alloy powder) and alloys, reactive metals (such as titanium, tantalum, niobium,
Figure 5.2 (titanium powder). These powders pack to densities and rhenium), and special electronic alloys. Many of these are
over 45% of theoretical, often close to 60%. Powder avail- not used in high production volumes, so they are prepared as
ability determines which engineering materials can be pro- special ordersa factor that negatively impacts price.
duced by MIM; fortunately, most of the widely used For the more complex chemistries, there are three possible
engineering compositions are formed into small particles. starting forms:
Table 5.1 is a listing of the major chemistries in use. mixed elemental powders
This tabulation includes a variety of metals and alloys. Be- fully compounded or prealloyed powders
sides the metallic powders, a range of oxide ceramics (alu- hybrid combinations of compounded master alloys and
mina, zirconia), nonoxide ceramics (silicon nitride, silicon elemental powders
carbide), cemented carbides (tungsten carbide with cobalt), For stainless steels, all variants are in use: mixed iron,
and refractory metals (tungsten, molybdenum) are in use. Be- nickel, and chromium powders; prealloyed powders where
sides those listed, it is possible to obtain special materials on each particle is the same composition; and iron powders
a custom basis. Other custom alloys include aluminide inter- mixed with master prealloyed ironnickelchromium powder.
metallics and NiTi shape-memory alloys. Some of the special The less homogeneous the starting powder, the more effort
required in sintering to obtain high final property levels. Thus,
the mixed elemental powder route has the lowest cost while
generally the fully compounded prealloyed powders give the
greatest assurance of meeting property and performance goals.
In the case of a stainless steel, the prealloyed powder sinters
with relative ease. If iron and a prealloyed master alloy pow-
der are mixed to give the same bulk composition, then the
time at the peak sintering temperature must be extended. Cre-
ation of a homogeneous sintered microstructure is even more
difficult when starting with elemental powder mixtures. The
bulk chemistry conforms to the alloy specification and the
powder cost is lower, but the microstructure will not be ac-
ceptable unless sintering is extended for a long time.
72
Table of Contents
Chapter 5: Material Properties
Powders sity. One major trend is to rely on smaller particle sizes that
The variety of powder chemistries used in MIM requires a sinter more easily and improve surface detail on the sintered
broad array of powder-fabrication techniques. Widely used article.
powder-production technologies include gas atomization, One problem in powder purchasing is that the materials
water atomization, attritioning, carbonyl decomposition, oxide might come with different designations for the same compo-
reduction, and plasma atomization. Each fabrication approach sition; the standards come from unrelated sourcesMilitary
influences the particle size as well as the shape, microstruc- Specifications (MIL SPEC), Aerospace Materials Specifica-
ture, packing density, chemistry, and cost. Sometimes, the par- tions (AMS), American Society for Testing Materials
ticle shape is not ideal for MIM, mainly because of asperities (ASTM), International Standards Organization (ISO), Society
on the particle surface that interfere with particle flow in of Automotive Engineers (SAE), as well as German, U.S., and
molding. In such cases, higher-cost but more-spherical pow- Japanese trade associations. This results in different names for
ders might be a better choice. As consumption increases, there the same material. For example, surgical stainless steel is also
will be more fabrication of powders specifically for MIM, called 17-4 PH stainless steel. It is often referred to as AISI
with a progressive price reduction. 630 in the medical field (American Iron and Steel Institute)
Table 5.2 gives some examples of the metal powders in use, and AMS 5355 in the aerospace field and is called SUS630 in
listing the composition, median particle size, and packing den- Japan. To avoid trademark infringements, several of the pow-
der suppliers give the materials their own names.
TABLE 5.2. CHARACTERISTICS OF SOME TYPICAL
Feedstock
MIM POWDERS
Several companies supply the small powders required for
nominal median apparent MIM. In some cases the company is focused on supplying
material chemistry, particle density,
small powders specifically to this field, while in other cases
% size, m g/cm3
the company has a broad offering and injection molding
aluminum 99.7 Al 30 1.1 grades are a relatively small offering. Annual updated listings
cemented carbide WC-10Co 1.1 3.4
of firms providing different powders are given in technical
copper 99.5 Cu 14 3.3
iron 98.3 Fe 4.1 2.6 magazines, such as Powder Metallurgy, Ceramic Industry, Ce-
ironcarbon Fe-0.9C 4.7 2.7 ramic Bulletin, Metal Powder Report, or International Jour-
ironsilicon Fe-3Si 7.6 3.9 nal of Powder Metallurgy. Additional information is available
molybdenum Mo 2.5 2.1 via Web searches. Trade associations help point to vendors
nickel 99.8 Ni 7.6 3.6
through their Web sitesfor example, visit the sites of the
niobium 97.8 Nb 7.4 3.0
molybdenum 99.7 Mo 4.1 2.7 American Ceramic Society (www.acers.org), European Pow-
stainless steel Fe-Cr-Ni-Mo 11 3.9 der Metallurgy Association (www.epma.com), Japan Powder
stainless steel Fe-Cr-Ni-Cu 21 3.3 Metallurgy Association (www.icon.perf.nagano.jp), and Metal
stainless steel Fe-Cr-Ni-Cu 10 4.2 Powder Industries Federation (www.mpif.org). The Inovar
titanium Ti-6Al-4V 32 2.1
Communications group produces both a printed and electronic
tool steel Fe-12Cr-1.5C 11 4.0
tungsten 99.5 W 3.0 2.9 report on the field that keeps current with changes.
Compounded feedstock consisting of precisely mixed
binders and powders are commercially available for molding.
TABLE 5.3. ALPHABETICAL LISTING OF Feedstock vendors include the firms listed in Table 5.3. Firms
COMPANIES SUPPLYING PREMIXED FEEDSTOCK enter and exit this business frequently, but at any time there
are typically about 10 to 12 sources.
Advanced Metalworking Practices (U.S.)
Atect (Japan)
BASF (Germany) Standardized Chemistries
Di-Ichi Ceramco (Japan) The standards for MIM materials focus on composition,
eMBe (Germany) typical properties, and testing procedures. More details are
Imeta (Germany) given at the end of this chapter. Most compositional specifica-
Inmatec (Germany)
PolyMIM (Germany) tions are duplicates of traditional handbook specifications with
Ryer (U.S.) tolerances similar to those listed by the American Society for
Starfire Systems (U.S.) Testing Materials (ASTM) or similar bodies. However, there
Syroko (Netherlands) are a few chemistries available that are specific to MIM.
TCK (Dominican Republic) Mostly these represent adjustments to standard chemistries to
Tosoh (Japan)
Witech (Japan) assist in processing. For example, extra chromium in a stainless
steel improves corrosion resistance, but also increases cost.
73
Table of Contents
Chapter 5: Material Properties
The convention is to show compositions in weight percent UK), now carry MIM materials and tabulate their properties.
(w/o) unless otherwise noted, and in most cases the percent Compositions used in MIM are detailed in Table 5.4 based
designation is skipped where there is no confusion. Hence, on the common designation. Note that in some cases individ-
Fe-2Ni indicates 98 w/o Fe and 2 w/o Ni. Unfortunately, the ual vendors customize the nomenclature or compositions
same alloy might have different names depending on which slightly, but this listing is characteristic of the popular MIM
standard is adoptedfor example, Fe-2Ni alloys are desig- materials.
nated as 2200, 4605, or 4650, often with similar compositions
(formally 2200 has no carbon, but both 4605 and 4650 have COST ATTRIBUTES
about 0.5% C). Here are a few examples of the different Material selection involves a balance between cost and per-
names given to essentially the same composition. In each case formance. Lower-cost materials are used most frequently. Fig-
the name used here is given first: ure 5.3 shows how the powder injection molding sales for
ironnickel (Fe-2Ni) = FN02, MIM-2200, FN0200, Fe- 2010 partition by material class (note that this figure is based
2Ni, steel 2200, AISI 4600, FEN-2, MIM-Fe-2%Ni, 2% on material value, not on tonnage), showing that over 77% of
nickeliron, FeNi2, Fe-2%Ni all sales are metals. Ceramic injection molding is about 15%
stainless 316L = X2CrNiMo 17 13 2, DIN 1.4404, UNS and carbides constitute 3% of sales. A small amount of PIM
S31603, SUS316L, MIM 316L, MIM-316L, MP-S- is applied to composites, and about 4% of sales involve mixed
0004, SAE316LG, AISI 316L materials or are indeterminate as to material category.
stainless 17-4 PH = MIL-S-81506, X5 CrNiCuNb 17 4, If plotted on a mass basis the carbide sector would shrink.
DIN 1.4542, SAE J467 (17-4 PH), AMS 5355, UNS This is because carbide powders are generally more expen-
S17400, AISI 630, SUS630, MIM-17-4 PH, 17.4 PH, sive. Beyond the differences in ingredient costs (nickel vs.
and MP-S-0021 iron, for example), powder price depends on consumption: the
A comprehensive cross-index of names is available on var- more popular compositions are more economical. This means
ious Web sites. Additionally, the important databases, such as that ferrous alloys and oxide ceramics are the lowest-cost and
Cambridge Engineering Selector (Granta Design, Cambridge, most frequently used materials. The classic chicken-and-egg
74
Table of Contents
Chapter 5: Material Properties
question arises: are they low cost because they are used fre- 5.4 plots a generalized clustering of strength vs. relative cost
quently, or are they used frequently because of a lower cost? per unit volume. Most MIM parts are volume constrained, so
In most cases the materials are used frequently and, as a result, obtaining a high strength at a low material cost makes ferrous
are lower in cost. alloys most attractive.
Besides raw material cost, the conversion of the material Besides material cost, there are differences in processing
into a powder increases cost, usually by about $1 to $2 per steps and expenses such as sintering that complicate cost con-
kg. The preparation of feedstock further adds to the price. siderations. This chapter provides data on the materials and
High-volume mixing cuts cost, while frequent composition their properties. This will help guide one toward the selection
changes require much equipment cleaning and adds consid- of the right material for an application. Coupled with material
erably to the cost. As consumption has increased, some of the cost guides, the most effective material becomes evident by
powders have undergone dramatic price decreases. To help balancing performance attributes and cost. As a leading engi-
understand the advantage of steels and stainless steels, Figure neer phrased the situation: ... first you must meet the per-
formance criteria, then materialprocess selection is dictated
by issues of availability, capacity, reliability, quality, and
finally cost. The data presented here help sort through the
options, but in the end, economic considerations usually
dominate. Thus, the MIM industry relies heavily on a few
compositions;
two stainless steels (316L and 17-4 PH)
ironnickelcarbon steels with typically 2% Ni and 0.4%
C, maybe with Mo or Cr
two electronic alloys (Invar and Kovar)
two tungsten alloys (tungstennickelcopper and tung-
stencopper)
cemented carbides (WC-Co compositions).
MATERIALS SELECTION
Figure 5.3. Relative sales of various PIM materials, showing a A first consideration in materials selection comes by eval-
dominance by metal powder injection molding at over 77%, uating previous designs and failures. Historically, the major
ceramics at 15%, carbides at 3%, and composites at 1%. About reason for component field failures is improper materials se-
4% of the global PIM industry sales are based on mixed materials lection, including improper specification of the surface treat-
produced and combined in an assembly
ment and heat treatment. Besides historical data, other
considerations include the function (property requirements)
and economic criteria.
Performance classifications are given in Table 5.5 based
on dominant concerns. Various combinations are possible.
Prior history with a material (both successes and failures) car-
ries the greatest weight in a new design. In situations where
there is no previous design, analysis of the required properties
is combined with handbook searches for candidate materials.
This leads to a preliminary selection to meet the design
constraints.
If the component size is fixed (constrained design), then
the selection is usually based on the lowest cost per compo-
nent. This means the cost per unit mass (most typically what
is quoted) needs to be multiplied by the density to give the
cost per unit volume. For example, a cemented carbide pow-
der is priced at $50/kg while a zirconia powder is $80/kg. The
density of the former is 14.6 g/cm3 and the latter is 6.0 g/cm3;
thus, the volume-based cost is $0.73/cm3 and $0.48/cm3, even
though the zirconia is higher in cost per unit mass. Since many
Figure 5.4. Scatter plot of how material strength and cost per unit designs are volume constrained, the cost per component vol-
volume compare for several of the materials employed in MIM ume is a key economic factor in materials selection.
75
Table of Contents
Chapter 5: Material Properties
76
Table of Contents
Chapter 5: Material Properties
soft magnetic components. Most are ferrous alloys, especially magnetization field from the origin to the first tangent with
pure iron, and alloys such as Fe 2Ni, Fe 3Si, Fe 0.45P, Fe 0.6P, the magnetization curve, giving the maximum permeability.
and Fe 50Ni. Of these, the Fe 50Ni alloy has a very attractive Table 5.7 lists the soft magnetic properties. Magnetic per-
combination of high magnetization and low coercive force. formance depends on details of the fabrication cycle resulting
Soft magnetic properties are determined by applying a field in vendor differences. Generally MIM products meet or ex-
to the material and measuring the response in terms of induced ceed the handbook magnetic properties, since the process
magnetic behavior. A schematic plot of induced magnetization gives an annealed (strain-free) microstructure and the sinter-
vs. applied magnetic field is given in Figure 5.5. During this ing cycle evaporates volatile impurities that otherwise degrade
test, the applied field is increased from zero and the induced the magnetic properties in wrought products.
behavior is quantified at saturation (at, say, 500 Oe or 39,800
A/m), then with no applied field (remnant magnetization), and Corrosion and Biocompatibility
finally by measuring the applied reverse field needed to re- The resistance to environmental attack depends on several
move magnetization (coercive force). The rate of magnetiza- factors, including the test fluid, and details such as aeration,
tion is quantified by the slope of the induced-vs.-applied acidity, halide ion concentration, oxygen content, and contact
with other materials. The corrosion, oxidation, or biocompat-
ibility of a sintered material is caught up in these details and
further depends on the sintered microstructure, heat treatment,
test conditions, and even the measure used to quantify corro-
sion resistance. As a consequence, for any application it is best
to subject representative test samples to the target environment
to assure proper behavior.
MIM components are used as human implants and have
been tested, but not used, for replacement knees and hips.
Antibacterial response has been quantified and one important
finding is that the addition of 3 w/o copper to MIM 304L
stainless steel gives much improved resistance to bacterial
growth. This is one situation where MIM allows modification
Figure 5.5. Magnetization response to an applied magnetic field of the chemistry to customize the material for the application.
for a MIM soft magnetic material. The test starts with a demagnet- Several biocompatible materials have passed implant testing,
ized sample and progressively increases the applied field, then including zirconia, tantalum, titanium and two titanium alloys
decreases the field through the point of reversal. Characteristics of (Ti-6Al-4V, Ti-6Al-7Nb), hydroxyapatite, and composites of
the material are extracted from this curve during initial magnetiza-
tion (permeability), at saturation, with no applied field (remnant), titanium and hydroxyapatite. Titanium fabricated by MIM ex-
and with sufficient field to fully remove the initial magnetization hibits resistance to stress-corrosion cracking that would nor-
(coercive force) mally attack a wrought stainless steel. Further, MIM 316L
77
Table of Contents
Chapter 5: Material Properties
stainless material has been qualified for implant use. Gener- resistant than wrought stainless steel in both solutions.
ally tests in immersion situations show much benefit; for ex- There are recognized MIM vendor effects on corrosion.
ample, artificial saliva testing of a nickel-free MIM stainless This was demonstrated in exhaustive testing of MIM 316L
steel for dental use shows a 100-fold improvement over stainless steel from various vendors, showing from half to
wrought stainless steels. twice the corrosion rate vs. wrought. For example, in salt-
Evaluation tests are widely varied and range from short im- spray testing (ASTM G85) the wrought sample corroded at
mersion in chlorine bleach, to salt-spray tests lasting weeks, 0.0019 mm per year (0.075 mil per year), while the MIM sam-
and up to two months immersion in acids. From the metals, ples ranged from 0.0009 to 0.0036 mm per year (0.035 to 0.14
stainless steels and titanium are usually selected for corrosion mil per year). Hydrochloric acid tests (as per ASTM G31) re-
resistance, although a few applications exist for pure nickel. sulted in the MIM materials being attacked at 0.35 mm per
Impurities degrade corrosion resistance, but sintering tends to year (14 mil per year), which was nearly half the 0.67 mm per
evaporate impurities; hence, MIM corrosion properties often year (26 mil per year) for wrought 316L. Ferric chloride im-
are better than observed with alternative processing routes. mersion (ASTM G48) proved the most aggressive test, and in
The major effort on corrosion testing of MIM products has that test some of the MIM 316L samples gave up to 10 times
been directed at stainless steels. Stainless steels via MIM the corrosion rate of wrought. When all of the test results were
come as austenitic, ferriticmartensitic, or precipitation-hard-
enable grades. They contain at least 12% chromium. The 300-
grade austenitic stainless steels are used for consumer items
(the classic is 18-8 stainless used in cutlery with 18%
chromium and 8% nickel) due to good corrosion resistance,
excellent toughness, and nonmagnetic characteristics. Of all
stainless steels, the precipitation-hardened stainless grade 17-
4 PH (AISI 630) is the most widely used and tested in MIM.
Its mechanical properties can be adjusted via post-sintering
heat treatments. This allows its use in products ranging from
magnetic sensors to surgical tools. However, the low
chromium content does not fully prevent corrosion, as evident
in Figure 5.6, which shows dissolution on the surface of a
MIM sample after a week of immersion in nitric acid. The
400-grade stainless steels have lower alloying levels, but can
be fabricated to high-strength levels, making them useful for
components that include automotive turbochargers and knife Figure 5.6. Corrosion on the surface of a MIM 17-4 PH stainless
steel after immersion for a week in 20% nitric acid, showing some
blades. The 400-grade stainless steels generally have the low- localized surface attack
est corrosion resistance of the MIM stainless steels.
Corrosion in stainless steels is measured using salt-spray
and immersion tests. The immersion route examines discol-
oration and weight loss during prolonged exposure to solu-
tions that include artificial sweat, artificial saliva, boiling
water, copper sulfate, nitric acid, hydrochloric acid, sulfuric
acid, chlorine bleach, and ferric chloride solution. For some
immersion tests, electrochemical potentials are applied to de-
termine the passive film breakdown and onset of pitting.
When corrosion resistance is a major concern, 316L is the
most common MIM stainless steel, but if higher strength is
needed, then 17-4 PH is the typical choice. Tests on 17-4 PH
stainless fabricated by MIM illustrate the corrosion resistance.
Figure 5.7 plots the 200 h weight loss for two MIM materials
compared with wrought material tested in solutions of chlo-
rine bleach and ferric chloride. The corrosion test results are
given in units mass loss divided by sample surface area, in Figure 5.7. Corrosive weight loss after 200 h immersion in
chlorine bleach or ferric chloride for two different MIM processing
g/m2. Approximately 1 g/m2 converts to 0.0055 mm per year routes applied to 17-4 PH stainless steel. Compared with the
or 0.2 mil per year of penetration or dissolution attack. For wrought material, the MIM materials are very successful in
this testing, the MIM stainless steel products were more withstanding both corrosion environments
78
Table of Contents
Chapter 5: Material Properties
equally weighted, some of the products ranked significantly W-Cu-Co, and W-Cu) by MIM have been applied to several
better than wrought. Differences correlated to final density applications requiring high inertiamilitary projectiles,
and impurity levelsa higher density and lower impurity shaped-charge liners, golf-club weights, medical radioisotope
level benefitted corrosion resistance. shipping-and-handling containers, birdshot, bowling ball
One important issue in MIM stainless steel corrosion traces weights, fishing weights, cellphone vibrator weights, disk-
to vacuum sintering, as used by some MIM firms. Prolonged drive counterbalance weights, and military fragmentation de-
heating in vacuum results in chromium evaporation from the vices. For higher tungsten contents the density ranges up to
component surface. It is chromium that provides corrosion re- 19 g/cm3. On the other hand, aluminum and titanium are used
sistance, so its loss (it has the highest vapor pressure of the in applications requiring low densities.
iron, nickel, molybdenum, and chromium making up 316L) A plot of elastic modulus vs. density is given in Figure 5.8
is detrimental to corrosion. Also, any oxygen contamination
(which is not specified in the 316L composition) leads to a
significant loss of corrosion resistance. Chromium forms a TABLE 5.8. DENSITY AND ELASTIC MODULUS FOR
MIM MATERIALS
compound with oxygen on the component surface, removing
it from its passivating role. Tests for stress corrosion found no elastic
significant difference from wrought materials. Thus, MIM material density modulus
g/cm3 GPa
materials have the potential to perform superiorly to wrought
material in corrosive environments. cemented carbide (WC-10Co) 14.2 593
Oxidation of MIM materials is no different from the cobaltchromium or F75 7.9 240
(Co-28Cr-4W-3Ni-1C)
wrought counterparts, so long as the sintered density is over copper (Cu) 8.5 130
95% of theoretical (to avoid oxidation through pores). In the copper steel (Fe-2Cu-0.8C) 7.7 190
section dealing with mechanical properties, recommended Hastelloy X (Ni-Cr-Fe-Mo-Co) 8.3 210
maximum-use temperatures are given for several materials. Invar (Fe-36Ni) 8.0 205
With respect to oxidation resistance at 500C (932F), there Kovar (F15) (Fe-29Ni-17Co) 7.8 117
Mo-15Cu 10.0 280
is a wide offering, but also large price difference. For many stainless 17-4 PH 7.8 195
of the metallic species, oxidation behavior is unaffected by (Fe-16Cr-4Ni-4Cu-0.3Nb)
MIM. Materials for oxidation environments include silver al- stainless 316L (Fe-17Cr-12Ni-2Mo) 7.6 190
loys, titanium alloys, tungsten alloys, and nickel alloys. stainless 430L 7.5 190
Nickel-base superalloys are widely recognized for their high- (Fe-17Cr-1Mn-1Si-1Ni)
steel 2200 (Fe-2Ni-0.5Mo-<0.1C) 7.6 190
temperature-oxidation resistance. Ferritic stainless steels are steel 2700 (Fe-7Ni-0.5Mo-<0.1C) 7.6 190
lower in cost with somewhat reduced oxidation resistance, but steel 4650 (Fe-2Ni-0.4Mo-0.5C) 7.7 195
are commonly used in automotive applications. So, from a titanium 6-4 (Ti-6Al-4V) 4.4 110
pure performance consideration, stainless steels provide the tungstencopper (W-20Cu) 16.0 290
best performance per unit cost. tungsten heavy alloy (W-5Ni-2Fe) 17.7 380
tungsten heavy alloy (W-6Ni-4Cu) 17.0 280
tungsten heavy alloy (W-7Ni-3Fe) 17.0 360
Inertial Properties
Materials fabricated by MIM exhibit physical properties
such as density, elastic modulus, heat capacity, and other at-
tributes that generally are close to values listed engineering
handbooks. Since the systems are sintered to nearly full den-
sity there is no effect from the processing route. For inertial
devices, such as counterbalance weights, sporting devices, and
vibration weights, the density and elastic modulus of several
compositions are given in Table 5.8. For comparison the hand-
book values for the same compositions are also listed, where
differences between the two are generally within the test error.
In several cases the Poissons ratio was measured along
with the elastic modulus, with the finding that the value is usu-
ally the same as given in the handbooks. This makes sense,
since the elasticity and Poissons ratio both measure the
atomic spacing dilation with stress, and that is unaffected by
the injection molding route. Figure 5.8. Scatter plot of elastic modulus vs. density for a variety
Tungsten heavy alloys (W-Ni-Cu, W-Ni-Fe, W-Ni-Co, of powder injection molded materials
79
Table of Contents
Chapter 5: Material Properties
80
Table of Contents
Chapter 5: Material Properties
to demonstrate the range of common MIM materials, section 46 mm (0.20.25 in.) across, and the standard gauge
illustrating how the ceramics and cemented carbides generally length is 2538 mm (11.5 in.). The samples are pulled slowly
are much stiffer than the steel, stainless steel, titanium, copper, to avoid adiabatic heating. Ductile metals are characterized
and other MIM metals. A concentration occurs near 8 g/cm3 by yield strength, ultimate tensile strength, fracture elonga-
and 200 GPa (29 Mpsi) due to the large number of ferrous al- tion, hardness, and reduction in cross-sectional area. Included
loys produced by MIM. Note the very high elastic modulus in the table are the densities as a percent of the theoretical for
possible with cemented carbides at a density near 14 g/cm3. the composition. Heat treatments provide a means to adjust
the properties, and hot isostatic pressing is occasionally used
Structural Properties to reach full density as noted by HT or HIP in the table. Con-
Structural properties are associated with withstanding siderable variation is possible in these cycles, so only one ex-
deformation when loaded. Besides hardness, the most typical ample is listed, typically the cycle giving the highest hardness.
attribute is derived from a strength test, with tension being When post-sintering heat treatments are used, the tensile
typical for a metal and compression being typical for a strength increases, typically resulting in a lower ductility. In
ceramic. In rapid loading, the resistant to crack propagation most cases the MIM products are fairly close to the strength
is a major concern, and that is measured by toughness or fac- and hardness of the wrought counterparts, and in some cases
ture toughness. For cyclic loading, attention turns to the fa- are stronger. For example, the tensile properties of wrought
tigue strength, which is always much lower than the tensile and MIM Inconel 718 compare as follows:
strength. The mechanical properties attained in MIM are gen- wrought typical yield strength = 760 MPa (110 ksi)
erally close to those found in handbooks. MIM typical yield strength = 1,055 MPa (153 ksi)
Likewise the MIM ultimate tensile strength and elongation
Tensile Properties of Metallic Materials to fracture were attractive when compared to wrought mate-
A tension test pulls the material to fracture and, besides the rials. Indeed, the MIM elongation was nearly 6-fold the spec-
elastic behavior, it gives strength and ductility. Some typical ified minimum for wrought material.
room-temperature tensile properties are compiled in Table 5.9. As another comparison, note that the average handbook
The typical specimen has a rectangular or circular cross properties for Invar are a yield strength 240315 MPa
81
Table of Contents
Chapter 5: Material Properties
82
Table of Contents
Chapter 5: Material Properties
83
Table of Contents
Chapter 5: Material Properties
and unnotched samples along with the test device. In tradi- toughness data. In light of that caution, Table 5.12 summarizes
tional presssinter powder metallurgy the notched impact bars the impact properties reported to date, with notations on the
are low in toughness. The stress concentration at the notch in- test conditions.
duces crack propagation though the pores. Because presssin- Based on impact toughness, low-carbon steels are effective
ter powder metallurgy relies on unnotched Charpy impact bars in rapid-strain-loading applications for strength levels in the
to obtain higher values, MIM followed along. As a further 700 MPa (100 ksi) range. If corrosion and impact toughness
complication, some of the measurements are performed on are concerns, then stainless steels are the most common ma-
subsize (one-half or one-fourth cross-sectional area) un- terial choice, but at a higher cost.
notched samples and sometimes the impact energy is then mul- Fatigue data for MIM products depend on the heat treat-
tiplied by 2 or 4, but this is invalid. Unfortunately, notches and ment and surface finish. For example, 17-4 PH stainless steel
section thickness influence the fracture energy as illustrated in the as-sintered condition has a fatigue strength of 280 MPa
for 17-4 PH stainless steel in Table 5.11. Clearly, the reported (41 ksi), but after heat treatment almost doubles to reach 517
value needs to identify the sample size and notch condition. MPa (75 ksi). The tests on MIM products are summarized in
Caution is required in comparing impact toughness values, Table 5.13. Improved fatigue strength is possible by case hard-
since most of the data are not comparable to standard impact ening or shot peening the component surface. In one compar-
ison, 4640 steel was fabricated from mixed elemental (iron,
nickel, molybdenum, and graphite) powders and also from
TABLE 5.11. SAMPLE GEOMETRY EFFECT ON prealloyed powder. The mixed elemental powders gave a het-
CHARPY IMPACT FRACTURE ENERGY FOR MIM erogeneous microstructure which was beneficial to fatigue.
17-4 PH STAINLESS STEEL, HEAT TREATED TO THE
H900 CONDITION
The hardsoft structure interrupts fatigue-crack growth, re-
sulting in a higher endurance strength.
sample size notched, J unnotched, J Generally, the properties in fatigue improve significantly as
full size 2.3 1.5 pores are eliminated, as is evident by MIM cobaltchromium
(10 mm x 10 mm) (1.6 ft.lb.) (1.1 ft.lb.) subjected to a post-sintering HIP treatment. The notched fa-
half size 17.9 14.2 tigue endurance limit at 107 cycles is 210 MPa (30 ksi) as com-
(5 mm x 10 mm) (13 ft.lb.) (10 ft.lb.)
pared with 240260 MPa (3538 ksi) for cast products.
TABLE 5.12. CHARPY IMPACT TOUGHNESS PROPERTIES FOR MIM ALLOYS (see caution in text)
impact energy
alloy notes * nominal composition J (ft.lb.)
maraging steel U-notch Fe-18Ni-9Co-5Mo-0.5Ti-0.1Si 35 (25)
maraging steel U-notch Fe-18Ni-9Co-5Mo-0.5Ti-0.1Si 5 (4)
stainless 17-4 PH (HT) standard test Fe-16Cr-4Ni-4Cu-0.3Nb-0.8Si 1.5 (0.1)
stainless 17-4 PH (HT) subsize, notched Fe-16Cr-4Ni-4Cu-0.3Nb-0.8Si 14 (10)
stainless 17-4 PH (HT) subsize, no notch Fe-16Cr-4Ni-4Cu-0.3Nb-0.8Si 18 (13)
stainless 316L standard test Fe-17Cr-12Ni-2Mo-2Mn 30 (22)
stainless 316L subsize, notched Fe-17Cr-12Ni-2Mo-2Mn 10 (7)
stainless 316L subsize, notched Fe-17Cr-12Ni-2Mo-2Mn 27 (20)
stainless 316L subsize, no notch Fe-17Cr-12Ni-2Mo-2Mn 190 (140)
stainless 410 standard test Fe-11Cr-0.5C 44 (32)
stainless 420 subsize, no notch Fe-13Cr- Mn-1Si->0.15 C 40 (30)
stainless 430L subsize, no notch Fe-17Cr-12Ni-2Mo-2Mn 150 (111)
steel 2200 subsize, no notch Fe-2Ni-0.5Mo-<0.1C 135 (100)
steel 2700 standard test Fe-7Ni-0.5Mo-<0.1C 48 (36)
steel 2700 subsize, no notch Fe-7Ni-0.5Mo-<0.1C 175 (130)
steel 4140 U-notch Fe-1Cr-0.7Mn-0.3Mo-0.4C 17 (13)
steel 4140 (HT) subsize, no notch Fe-1Cr-0.7Mn-0.3Mo-0.4C 75 (56)
steel 4605 standard test Fe-2Ni-0.4Mo-0.5C 14 (10)
steel 4605 subsize, no notch Fe-2Ni-0.4Mo-0.5C 70 (50)
steel 4605 (HT) subsize, no notch Fe-2Ni-0.4Mo-0.5C 55 (40)
*standard Charpy test is a notched sample with cross section 10 mm by 10 mm with 2 mm deep 45 notch (0.4 in. by
0.4 in. with 0.078 in. notch); U-notch sample has 2 mm wide and 2 mm deep rounded notch in 10 mm by 10 mm
sample: subsize samples are reduced in cross section to 5 mm by 10 mm (0.2 in. by 0.4 in.) and may lack the stress
concentration notch
84
Table of Contents
Chapter 5: Material Properties
Optical Properties
Optical properties of concern include surface finish and
color. Surface finish is dictated by the powder size: smaller
powders give smoother sintered surfaces. Most MIM products
have a dull matte finish after sintering, but can be polished to
a high luster using standard techniques. For applications sen-
sitive to external appearance (golf club heads, watchcases,
jewelry, eyeglass components, cellphone hinges, cosmetic
cases, computer logos, and camera pieces) this final polishing
step is a routine secondary operation. The surface roughness
after sintering averages near 0.8 m (32 in.). With small
Figure 5.14. Thermal softening data for MIM 316L stainless steel
showing a decrease in yield strength on heating. For comparison,
powders this can be improved to 0.4 m (16 in.). Smoother
wrought 316L stainless steel has a yield strength of 250 MPa at surfaces are obtained by post-sintering finishing steps, with
200C and that strength declines to about 100 MPa at 800C 0.02 m (0.8 in.) being obtained on watchcases, fountain
85
Table of Contents
Chapter 5: Material Properties
pens, and jewelry. Figure 5.16 is an example of a MIM stain- fabrication process. Aluminum and titanium are colored using
less steel watchcase taken to a high polish after sintering. either heating or electrochemical treatments. Steels are ther-
For oxide ceramics, the normal opaque color is a white, mally or chemically treated to form a black magnetite surface.
ivory, light yellow, or tan (depending on composition and im- Tungsten carbide delivers a lustrous dark gray used for watch-
purities). Zirconia can be doped to create an array of brilliant cases. Alumina (Al2O3) doped with titanium carbide (TiC)
colors ranging from black to white, and including red, yellow, gives a deep black color, while alumina doped with chromia
green, and blue shades. (Cr2O3) delivers a pink color and, when sintered to translu-
Porcelain fabricated by CIM is sintered to a white condi- cency, gives a ruby red color. However, zirconia is the most
tion with a transparent glass glaze. For optical transmission popular of the colored ceramics. Applications include luxury
the ceramic must be sintered to optical transparency or semi- pens, watchcases, buttons, and decorative hardware. Figure
transparency. This implies sintering to a pore-free condition 5.18 is a picture of a luxury watch with a face and band formed
with low impurity levels. Alumina (Al2O3) containing small using over one thousand CIM black zirconia segments.
concentrations of magnesia (MgO) has controlled grain Translucent oxide ceramics are possible with excellent sin-
growth during sintering and is injection molded to form high- tering technologies. Typically a low-impurity, zero-porosity
intensity lighting components for sodiummercury vapor condition is required. Other than lighting-lamp envelopes, in-
lamps. Such a ceramic is shown in Figure 5.17, where letter- jection molding has not penetrated much of this market.
ing is visible through the sintered tube. Other ceramics are
sintered to similar high-translucency levels, including spinel
(MgAl2O4). Hot isostatic pressing is often required to obtain
the pore-free density necessary to obtain translucency.
Other than intentionally colored ceramics, most materials
are unchanged from that color established by nature. Gold, sil-
ver, and platinum alloys give colors that are unaffected by the
Figure 5.16. Example of the final polish and surface luster possi- Figure 5.18. Luxury wristwatch formed using thousands of small
ble on MIM stainless steel, in this case illustrated by a watchcase black CIM zirconia pieces
86
Table of Contents
Chapter 5: Material Properties
Thermal Properties but contracts due to cooling when turned off. Onoff cycles
Four thermal properties are of concernmelting range, create differential strains that induce failure over time if the
heat capacity, thermal-expansion coefficient, and thermal con- thermal expansion coefficients are different. Thus, thermal-
ductivity. The latter two are important to heat-dissipation ap- management materials are constrained with respect to the
plications in microelectronic packaging, where a high thermal allowed thermal-expansion coefficient, but maximized with
conductivity is desired with controlled thermal expansion. The respect to thermal conductivity.
thermal expansion needs to match with silicon or dielectric Figure 5.19 is a scatter plot of the metal, ceramic, and com-
alumina used as a circuit substrate. posite PIM materials associated with thermal management ap-
Melting temperatures for MIM alloys are essentially the plications. Generally, the thermal conductivities of MIM
same as for other fabrication routes. Hence, if the handbook materials are comparable to other forming techniques. A high-
melting range for 316L stainless steel is 1,365C1,420C thermal-conductivity material with a desirable low thermal-
(2,490F2,590F), then the MIM product will exhibit the expansion coefficient is W-10Cu. The high density of a
same range. Exceptions occur when sintering additives are tungsten system makes it best for stationary applications, such
used to lower the onset of melting. For example, boron is as network server and base-station cellphone systems. The low
added to stainless steel to lower the sintering temperature, but density of the aluminum nitride comes with a higher cost that
it also reduces the onset of melting to 1,230C (2,246F). restricts its use to portable electronics. Tungstencopper is
Heat capacity is the amount of energy stored in a material
as its temperature is increased. It relates input energy to tem-
perature change on a unit mass basis. The British thermal unit
(BTU) measured the relation between heat and 1F tempera-
ture change for one pound of water. Today, heat capacity is
measured using J/(kg K). There is no substantial difference
between MIM materials and handbook properties. For exam-
ple, MIM Kovar has a heat capacity near 500 J/(kg K) which
is near the handbook value of 460 J/(kg K).
Table 5.14 collects data on the thermal properties of heat-
dissipation MIM materials. The desire is to reduce the ther-
mal-expansion coefficient while maximizing the thermal
conductivity. Semiconductors dissipate heat and must be
mated with high-thermal-conductivity materials to operate
properly. However, the semiconductor has a low thermal-
expansion coefficient, which mandates the MIM material like- Figure 5.19. Plot of thermal-management materials associated with
wise have a thermal-expansion coefficient in the 5710-6/K heat dissipation in microelectronics, showing the relative placement
range. Note, thermal expansion is measured in 10-6/K (kelvin) of several options with respect to the thermal-expansion coefficient
or parts per million per K, sometimes abbreviated as ppm/K, and thermal conductivity. Because of silicon and the ceramic sub-
where ppm stands for parts per million. If the thermal expan- strates thermal expansion is constrained, while the highest possible
thermal conductivity is sought, leading to strong interest in tung-
sion of the heat-dissipation material increases beyond this stencopper composites. However, for low-density portable appli-
range, then the assembly will fail due to thermal fatigue. cations (such as in laptop computers or control circuits for hybrid
When power is applied, the device expands due to heating, automobiles) aluminum nitride is selected in spite of a high cost
87
Table of Contents
Chapter 5: Material Properties
88
Table of Contents
Chapter 5: Material Properties
COMPOSITE MATERIALS
One advantage of MIM is the ability to mix different pow-
ders to create composites. Insoluble phases are combined to
customize propertiesfor example, a soft jewelry alloy gives
improved wear resistance when a small quantity of oxide ce-
ramic is added in the powder stage. Other composites are de-
sired because of wear resistance and the general ability to
customize properties. A few specific examples are given
below of composites fabricated using injection molding, usu-
ally relying on mixed powders.
High-Elastic-Modulus Cermets
Following on the idea of adding a ceramic phase to im-
prove hardness, higher contents of hard phases increase the
Figure 5.22. Microstructure of injection molded titanium after elastic modulus. A well-known example comes from the com-
sintering for two hours at 1,400C (2,552F) in vacuum, a
bination of tungsten carbide with cobalt, which is very stiff
temperature that produces two crystal phases (photograph
courtesy of Anthony Griffo) when compared to cobalt640 GPa vs. 211 GPa (93 Mpsi vs.
31 Mpsi). Likewise, steels have been loaded with stiff phases
such as titanium boride (TiB2) to increase elastic modulus.
This boride is 2.5 times stiffer than steel, so a composite with
40 v/o boride is stiffer than steel by a factor of 1.6.
Heterogeneous Microstructures
Early in MIM it was common to mix elemental powders
to form alloys during sintering. Since diffusional homogeniza-
tion might not be complete, the sintered microstructure con-
sisted of pockets rich in each individual chemistry, giving
heterogeneous mechanical properties at the microstructure
level. Surprisingly, in some cases the bulk properties were im-
proved by the heterogeneity. Phase heterogeneity has been
used to develop harder, tougher, and more wear-resistant ma-
terials. Table 5.15 directly compares the properties of forged
Figure 5.23. Microstructure of 17-4 PH stainless steel after material with the MIM homogeneous and heterogeneous ma-
sintering in hydrogen at 1,365C using water-atomized powder.
The dark spherical pores are trapped steam from a high-
terials, and further compares to traditional powder metallurgy
temperature reaction between oxygen (from water atomization) (press-and-sinter) steel. The heterogeneous MIM product is
and the hydrogen sintering atmosphere nearly as strong as forged steel.
Another demonstration comes from Fe-2Ni-0.5Mo-0.4C
steel sintered to different levels of microstructure homogene-
ity. Data from fatigue testing show an impressive gain from
the heterogeneous microstructure:
fatigue endurance limit for homogeneous MIM product
= 475 MPa (69 ksi)
89
Table of Contents
Chapter 5: Material Properties
fatigue-endurance limit for heterogeneous MIM product reaction is reversible. Normally sintering of these metals must
= 575 MPa (83 ksi) avoid hydrogen, relying on inert gas or vacuum to avoid re-
Accordingly, selective heterogeneities are used to improve actions. However, most hydrides are not stable at high tem-
properties. peratures. Thus, high-temperature sintering in hydrogen is
possible without degradation of the metal. This facilitates the
Atmosphere-Stabilized Compositions use of hydrogen to remove oxygen from the component dur-
During sintering the process atmosphere penetrates be- ing heating. Cooling in hydrogen leads to hydride formation
tween the particles to alter the chemistry. Since adsorption or and a brittle material, but processing atmosphere changes are
de-adsorption are temperature dependent, composition control used prior to cooling to avoid forming the hydride. In a novel
during heating allows for selective changes in the material concept used for titanium, cooling is performed in nitrogen to
properties. produce a gold-colored titanium nitride for watches, jewelry,
There are strengthening benefits from nitrogen absorption and decorative items. Similar uses of the process atmosphere
for steels. However, ferrous alloys normally do not take ad- have been applied to cobaltchromium, aluminum nitride, and
vantage of this process, since nitrogen is unstable in molten silicon nitride compositions.
metals. Nitrogen absorption peaks at intermediate tempera-
tures. For bulk components, this means that nitrided surfaces Biocompatible Materials
are possible, but overall bulk strengthening is not realistic Hydroxyapatite is a biocompatible ceramic (calcium phos-
the nitrogen diffuses only into the surface regions. In MIM phatebased compound) with widely recognized compatibility
the opportunity is to add nitrogen during sintering. This is with bone. Unfortunately, hydroxyapatite is very weak. To pro-
used in MIM tool steels to form carbo-nitrides, giving a final vide MIM implants with tailored strength and biocompatibil-
strength of 3,000 MPa (425 ksi) with 3% elongation. ity, several ideas are based on mixed powders to form
Similar reactions are seen in nickel-free, nitrogen-strength- composites. One composition is based on a mixture of
ened stainless steel. Some humans are nickel-sensitive as ev- hydroxyapatite and titanium and this is accepted for implants,
ident by skin rashes around jewelry. Austenitic stainless steels such as posts for affixing replacement teeth. Proper combina-
with 8%10% nickel (such as 304L and 316L) can induce tion of porosity and the two phases allows matching the
these reactions. Nitrogen is an austenite stabilizer that allows strength and elastic modulus of bone, nominally with 100 MPa
removal of nickel in a corrosion-resistant composition. The (14.5 ksi) strength and 20 GPa (2.9 Mpsi) elastic modulus.
alloy has the desirable nonmagnetic character with the benefit Much current attention is on MIM tantalum or titanium to
of no nickel, while exhibiting twice the strength and 20-fold form a porous structure for tissue ingrowth. Variants are used
more corrosion resistance when compared to 316L stainless for replacement knees, hips, and shoulders where tissue at-
steel. A direct comparison of properties is given in Table 5.16. tachment is desired. The MIM porous surface structure is at-
Obvious uses for this alloy are in watch, dental, and surgical tached to a dense replacement component to provide pores for
components where prolonged wet contact with skin would tissue attachment.
lead to nickel release and an allergic response.
Some metals are hydride formers, including titanium, zir- Infiltrated Preforms
conium, tantalum, and niobium. This means they form com- Ceramics are low-thermal-expansion materials, while met-
pounds with hydrogen, in the same way that oxygen forms a als are higher in thermal and electrical conductivities. Com-
compound with hydrogen (water). The compounds are very posites are formed by first molding a porous ceramic skeleton
different in properties from the metals and often are consid- which is subsequently infiltrated with molten metal. The ce-
ered for hydrogen-storage media, since the hydridedehydride ramic skeleton provides the desired low thermal-expansion co-
efficient while the metal provides a high thermal conductivity.
Early commercial products relied on silicon carbide (SiC) with
TABLE 5.16. COMPARATIVE PROPERTIES OF copper or aluminum as the infiltrant. Two applications have
NITROGEN-STABILIZED MIM STAINLESS STEEL gained notice: electronic heat sinks and disk-brake calipers.
AND WROUGHT 316L STAINLESS STEEL
An advantage is that sintering does not have to densify the ce-
nitrogen- wrought ramic powder, so larger powders are used in molding. A few
property stabilized MIM 316L other demonstration systems are listed in Table 5.17.
nitrogen content, w/o 12 <0.05 Figure 5.25 is the cross section microstructure of a molded
density, g/cm3 7.6 8.0 Invar preform after infiltration with copper; this composite
hardness 30 HRC 68 HRB targets applications requiring a low thermal-expansion coef-
yield strength, MPa (ksi) 670 (97) 170 (25)
ultimate strength, MPa (ksi) 960 (139) 480 (70) ficient and high thermal conductivity.
elongation to fracture, % 35 45 Current efforts are looking to promote a reaction between
the infiltrant and the porous preform. One concept uses a dia-
90
Table of Contents
Chapter 5: Material Properties
91
Table of Contents
Chapter 5: Material Properties
TESTING STANDARDS
Industry standards for MIM have been formed by the re-
gional trade associations such as the European Powder Met-
allurgy Association (Shrewsbury, UK) and Metal Powder
Industries Federation (Princeton, NJ). The standards cover
testing procedures for the powders and sintered materials,
composition ranges for popular materials, and property levels Figure 5.29. Example of the molding operation for forming a
for sintered materials. totally round tensile sample
92
Table of Contents
Chapter 5: Material Properties
line defects that might cause premature fracture out of the calls for 10 mm by 10 mm cross section with a notch. But
gauge length, the pin holes are often drilled after sintering. some data are for unnotched bars, and furthermore are from
Strength is based on yield and tensile values, recognizing subsize samples. This leads to serious confusion as illustrated
that designers focus on the yield strength since component de- earlier in Table 5.11 for MIM 17-4 PH stainless steel. The
formation is usually not acceptable in service. Tensile testing smaller sample size delivers the higher fracture energy both
is performed at a slow strain rate to avoid adiabatic heating in with and without a notch.
the manner diagramed in Figure 5.31. The elastic modulus is Detailed instructions exist for magnetic property testing,
estimated from the initial stressstrain response. Fracture where the convention is to use a 25 oersted maximum applied
elongation, and the sample reduction in area at the point of field (as per ASTM A 773). Other data might include surface
fracture, gives a guide to ductility. finish, microstructure (including phases and their distribu-
Impact bars range from square cross sections 10 mm by 10 tions), grain size, porosity or density, hardness, wear resist-
mm by 55 mm length (0.4 in. by 0.4 in. by 2.2 in.) to subsize ance, corrosion resistance, chemical composition, and various
samples of one-half or one-fourth the cross-sectional area. In other parameters.
some cases the samples lack the standard 2 mm deep 45 Special concern is given to corrosion testing. Corrosion
notch. Several variants are in use. Some of the bars are testing can be simple using immersion tests. One corrosion
notched while others are unnotched, as illustrated in Figure test applied to MIM stainless steels is to immerse a sample
5.13. This unnotched variant also can have 10 mm by 10 mm 5 mm by 10 mm by 55 mm (0.4 in. by 0.4 in. by 2.2 in.) in a
or 5 mm by 10 mm cross sections. The ASTM impact standard 2% sulfuric acid solution at room temperature for 1,000 h.
However, it is faster to rely on a 6 min immersion in a copper
sulfate solution and to inspect for copper plating or other at-
tack (ASTM F 1089). Other tests for corrosion include boiling
water immersion for 30 to 180 min (ASTM F 1089). One very
effective test, commonly used but not standardized, is to im-
merse the sintered component in chlorine bleach and observe
for discoloration over one minute.
MATERIAL STANDARDS
MIM compositions are treated at several levels. The first
level deals with the nominal composition, generally expressed
by a designation such as 316L. At the next level are the de-
tailed chemical specifications for the major ingredients. It has
a nominal composition specification of 10%14% nickel,
16%18% chromium, 2%3% molybdenum, up to 2% man-
Figure 5.30. Popular MIM tensile bar that is a hybrid of the round ganese, and up to 1% silicon. These allow for reasonable com-
and flat geometries. It is for strength and ductility tests. Shown is a position variations, often in the range of 1%2% on the
sintered sample and one twisted to demonstrate ductility
major alloying ingredients. The next level of specification
treats the impurities, and for 316L the key limitation is in the
carbon level, which is limited to less than 300 ppm (parts per
million) carbon.
Impurities are important to MIM, for example, as the car-
bon level decreases in 316L, the corrosion resistance im-
proves. Most of these specifications are borrowed from
wrought alloys. Unfortunately, wrought alloys do not antici-
pate some of the MIM options. If strength is a concern, then
just a specification of less than 300 ppm C for MIM 316L is
inadequate, since new impurities arise in MIM. Two important
cases are oxygen and nitrogen, both possible impurities from
sintering atmosphere contamination. Since the wrought ma-
terials are always low in N and O, these two impurities are ig-
Figure 5.31. Schematic illustrating how a tensile sample is nored in the specifications. For MIM this is an unsafe
clamped and pulled to fracture. Instrumentation for load and
displacement during the test allow calculation of elastic modulus, assumption and a few horror stories result. In one case the in
yield strength, tensile strength, facture strength (not normally specification stainless steel components rusted prior to ship-
reported), reduction in area, and fracture elongation ment. Thus, even the MIM industry compositional specifica-
93
Table of Contents
Chapter 5: Material Properties
tions make assumptions that are not always valid. For exam- 17-4 PH, Fe-based with 3%5% Ni, 15.5%17.5% Cr,
ple, 17-4 PH stainless steel sintered in pure hydrogen atmos- 3%5% Cu, 0.15%0.45% (Ta+Nb), 0.07% max C
phere delivers a sintered tensile strength of 980 MPa, but that
strength is for no added nitrogen during sintering. If a small Most of the metallic systems allow for up to 0.6% or 1%
level of nitrogen is in the sintering atmosphere, the sintered silicon (Si) and 1% of other total contaminants and some in-
strength is 10% higher, with a concomitant loss of ductility. tentionally allow for up to 2% manganese (Mn). No oxygen
For many compositions there are no industry-accepted or nitrogen levels are included in the specifications.
standards. MIM component producers have written material
standards through regional trade associations. More rigorous RESOURCES
standards often come from the users based on application data. Materials Standards for Metal Injection Molded Parts, 2010,
This is especially true in automotive, medical, and microelec- Metal Powder Industries Federation, Princeton, NJ.
tronic industries. Otherwise, the material standards are derived
Standard Specification for Metal Injection Molding (MIM)
for other industries. Some relevant specifications for MIM
Ferrous Alloys, ASTM International, West Conshohocken,
products are found in ASTM B 883 and MPIF 35.
PA.
In the ASTM and MPIF property specifications, both typ-
ical and minimum values are given. The typical values are Standard Test Methods for Metal Powders and Powder
based on samples performed by several companies and ex- Metallurgy Products, 2010, Metal Powder Industries Federa-
press the average behavior. The minimum values are based on tion, Princeton, NJ.
the lowest values from the samples. Typical implies an indus-
T. Beck, J. Schneider and V. Schulze, Characterisation and
try self-monitored average. Minimum standards anticipate
Testing of Micro Specimen, Microsystem Technologies,
production values will be higher.
2004, vol. 10, pp. 227232.
For MIM alloys, the composition specifications include the Metals Handbook Desk Edition, second edition, 1998, J.R.
following steels: Davis (ed.), ASM International, Materials Park, OH.
2200, Fe-based with 1.5%2.5% Ni, 0.5% max Mo,
F.H. Froes and R.M. German, Cost Reductions Prime Ti
0.1% max C
MIM for Growth, Metal Powder Report, vol. 55, 2000, June,
2700, Fe-based with 6.5%8.5% Ni, 0.5% max Mo,
pp. 1221.
0.1% max C
4140, Fe-based with 0.8%1.2% Cr, 0.2%0.3% Mo, R.M. German and A. Bose, Injection Molding of Metals and
0.3%0.5% C, 1% max Mn Ceramics, 1997, Metal Powder Industries Federation, Prince-
4605, Fe-based with 1.5%2.5% Ni, 0.2%0.5% Mo, ton, NJ.
0.4%0.6% C
V.A. Hackley and C.F. Ferraris, Guide to Rheological Nomen-
clature: Measurements in Ceramic Particulate Systems, NIST
Soft magnetic alloys are as follows:
Special Publication 946, January 2001, National Institute of
2200, Fe-based with 1.5%2.5 % Ni, 0.1% max C
Standards and Technology, Gaithersburg, MD.
Fe-3Si, Fe-based with 2.5%3.5% Si, 0.05% max C
Fe-50Ni, Fe-based with 49%51% Ni, 0.05% max C K.S. Hwang and M.Y. Hsiao, Mechanical Properties of Fe-
Fe-50Co, Fe-based with 48%50% Co, 2.5% max V, Ni Injection Molded Parts Made from Composite Powders,
0.05% max C Powder Injection Molding Symposium 1992, P.H. Booker,
430L, Fe-based with 16%18% Cr, 1% max Mn, 0.05% J. Gaspervich, and R.M. German (eds.), Metal Powder Indus-
max C tries Federation, Princeton, NJ, 1992, pp. 5366.
V. Josef and L.K. Tan, Thermal Performance of MIM
The low-thermal-expansion alloys used for instrumentation
Thermal Management Device, Powder Injection Moulding
and electronic packing include Kovar and Invar, but the only
International, 2007, vol. 1, pp. 5962.
one reaching specification status is the F-15 alloy based on
iron with 29% Ni, 17% Co. W.S. Lee and C.H. Hsu, Dependence of Impact Deformation
Behaviour of Fe-2Ni Sintered Alloy on Strain Rate and Sin-
Likewise specifications for stainless steels include the tered Density, Materials Science and Technology, 2006, vol.
following: 22, pp. 1,4311,443.
316L, Fe-based with 10%12% Ni, 16%18% Cr, 2%
Y. Masai, M. Ishida, S. Tanaka and K. Nishiyabu, Evaluation
3% Mo, 0.03% max C
on Mechanical Properties and Functionalities of Anti-Bacteria
420, Fe-based with 12%14% Cr, 0.15%0.4% C
and Resist-Corrosion for Cu-Added Stainless Steel
430L, Fe-based with 16%18% Cr, 0.05% max C
94
Table of Contents
Chapter 5: Material Properties
95
Table of Contents
Table of Contents
Chapter 6
COSTS
TYPICAL PROJECT COSTS reflects differences in internal costs and capacity factors, as
In dealing with MIM production costs, two related con- well as profit and risk factors. The same is true in MIM, but
cepts are introduced for this discussion: in some cases the ratio of highest to lowest bid seems to be
manufacturing cost is the cumulative expense incurred larger, up to a factor of 8.
in the production of a product; it includes the purchase Chapter 9 contains a detailed analysis of cost factors as a
of raw materials, as well as the expenditures for basis for spreadsheet calculations. It starts with tool cost, and
energy and labor, as well prorated items such as mainte- moves sequentially through feedstock cost, batch-size effects,
nance, depreciation, insurance, and taxes unit-manufacturing costs, and analysis of sensitivity factors
sale price reflects the value of the component to the for several aspects. The mathematics presented there are com-
user; it is the amount of money exchanged for the patible with straightforward spreadsheet calculations and a
fabricated component and normally includes the manu- few software packages are available that enable rapid cost es-
facturing cost and allocations for a return on investment timation.
and risk In MIM, the most straightforward approach to estimate
Ideally, cost and price should be related, but that is not component cost is to sum the costs accumulated at each man-
necessarily the case. During the 1980s, MIM products were ufacturing step, starting with tool fabrication and extending
selling at a price that sometimes was just 25% of the manu- to secondary operations and final inspection. The increments
facturing cost; bankruptcy was a consequence of this pricing are as follows:
philosophy. Today, MIM is profitable and the manufacturing tool cost
cost is generally below the selling price. feedstock cost (powder, binder, mixing, or purchased
Cost analysis is an important aspect of engineering design. feedstock)
One common approach is simply to send out a component de- unit-fabrication costs (mixing, molding, debinding, sin-
sign and solicit several responses from vendors. That usually tering)
results in several responses, with a large variation in cost and secondary, finishing, inspection, and packaging costs
a list of vendor-desired exceptions or changes to improve facility overhead burden
process yield. This chapter focuses on understanding what de- profit, tax, interest, and risk aversion
termines the production costs, realizing that individual com- Here the emphasis is on understanding how MIM costs
pany policies on profits, taxes, interest payments, dividends, depend on design factors such as part complexity, material
and risk combine to generate a range of price quotes. We rec- selection, and component size. As background to this
ognize price variations as a natural part of life. For example, approach, note that estimation procedures for tool cost are
on a local highway repair, ten construction firms bid on the well established in plastic injection molding, providing a
project. The ratio of the highest to lowest bid was 1.48, and sound basis for MIM. Indeed some of the tool vendors have
the lowest, and winning bid, was $3.26 million. The variation online tool-cost estimators. However, the source of the tooling
97
Table of Contents
Chapter 6: Costs
is a significant factor, since tool costs for MIM shops in India, prealloyed powder in China is half of that in Germany. Poly-
Thailand, and China are 10% of those in Germany, North mers used as binders tend to be a low cost factor and final
America, and Japan. price is nearly insensitive to the polymer formulation. Unit
Raw-material cost is another factor and it is dominated by costs in molding, debinding, and sintering depend on the fa-
the alloy and source. For example, the price of MIM-grade cility and installed equipment; generally these costs vary con-
siderably between operations.
Probably the weakest cost aspect for MIM is in secondary
operations. Some firms have embraced secondary steps such
as coining, but others rely on external services for machining,
heat treating, and coining. Accordingly, secondary costs vary
considerably, and largely this is where the larger MIM opera-
tions have a competitive advantage.
Once the fabrication cost is understood for a component,
the sale price is estimated based on profit, tax, interest, and
risk factors that vary between sites. These factors are well es-
tablished in the related fields of powder metallurgy, ceramics,
cemented carbides, and plastics molding.
For the core manufacturing steps, unit operation calcula-
tions are developed for each step to accumulate a total fabri-
cation cost. An outline of the costing model is given in Figure
6.1. It shows that several decisions and factors, beyond mate-
rial, enter into the calculations. Some industry-wide cost par-
titions are given in Table 6.1 to provide a rough sense on
factors impacting cost during MIM production. Although
there are recognized differences in all cost factors, including
labor, it is interesting that labor is not a dominant factor; MIM
succeeds in both expensive and low-labor-cost regions.
TOOLING COST
Tool cost for MIM is based on principles encountered in
plastic injection molding. Although many models exist, the
most accurate approach assumes purchase of a standard mold
base. It then calculates the time needed to machine the com-
ponent features into a single cavity. For multiple cavities, the
Figure 6.1. Detailed flow chart for part costing via MIM, showing machining time for the first cavity is used to estimate the time
how several factors enter into the overall calculation. The detailed and cost for each additional cavity, recognizing each cavity
model based on this flow chart is presented in Chapter 9 encounters a lower cost due to increased experience. As a first
approximation, the second cavity is constructed for about
70%75% of the first cavity cost, and four cavities are less
TABLE 6.1. RANGES OF TYPICAL MIM PRODUCTION
COSTS AS PERCENTAGE OF TOTAL COST than three times the cost of a single cavity.
Final tool cost is calculated from the total construction time
cost item range multiplied by a shop hourly rate, with additions for profit and
direct labor 10%30% administrative costs. This approach assumes the tool is correct
powder costs 8%20% on the first cycle, so it does not include a provision for rework.
feedstock preparation 5%10% In reality, tool shops are averse to risk when fabricating a tool
tooling amortization 2%6%
molding 6%15% that is outside their experience base; thus, a risk aversion fac-
debinding 3%10% tor is often introduced to allow for possible errors. Likewise,
sintering 5%10% current business levels impact tool quotations as evident by
interest and depreciation 5%9% the mapping of cost vs. deliverybusy shops are both costly
secondary operations 4%45% and have a long delivery time.
inspection 1% 4%
cost of yield loss 1%20% Detailed quantitative cost analysis of tool cost builds from
general, administrative, sales expenses 8%20% the assumption that a standard mold base is purchased and
converted into the customized design required for the compo-
98
Table of Contents
Chapter 6: Costs
nent molding. The mold base might cost $2,000. It needs to chining times and multiplying by the machining rate. A critical
be modified to include the specific cavity and related sprue, factor is the toolmakers hourly rate for mold fabrication,
runner, and cooling channel. Major cost additions are associ- which might be in the $45 per hour range, but can vary con-
ated with hot runner systems. siderably. For example, a recent comparison of union wages
To estimate the cavity-machining cost, the tool geometry around the world is shown in Table 6.2. With a 20-fold labor
and features determine the time and then the tool shop hourly rate difference it is difficult to quote a single value for any of
rate is multiplied to estimate the cost. In addition, purchased the manufacturing parameters, including the toolmakers
tool materials and special features are added to the cost. For hourly rate. Besides the listed countries, some recent MIM ef-
example, unscrewing devices, such as those required for in- forts have set up in India, Singapore, Hungary, and Thailand,
ternally threaded features, add significantly to the machining and there are other successful operations in Taiwan, Israel,
time, upwards of 250 hours each. Moreover, these motions are Italy, Brazil, and South Africa.
relatively slow and add to the molding cycle time. For exam- The tooling cost depends on several factors. Recent MIM
ple, a small 2 g stainless steel part with an internal thread re- tool costs ranged from $2,500 to $75,000. Further, besides
quires 25 seconds molding-cycle time, although cooling in the complexity and features, mold costs vary with geographic re-
cavity is quite fast. Likewise, besides unscrewing devices, side gion. To control mold cost, it is best to specify multiple cavi-
pull motions add at least 65 hours to the tool manufacturing ties when there is a compelling economic benefit, since
time. So the number of cores, ejector pins, and overall com- tooling is a significant initial cost. Likewise, for lower pro-
plexity and tolerances lead to an estimated machining time duction volumes it is less expensive to machine threads after
that feeds into the estimated cost. Complexity is based largely sintering than to make an expensive tool.
on the number of features, including surface protrusions,
bumps, patches, holes, and depressions. Costs escalate rapidly TOOL LIFE
with component complexity. Thus, each feature in the com- Because tooling is a large initial cost, extension of the mold
ponent is assessed for a typical machining time. life is important to the economics of MIM. Initial tooling cost
Beyond the time to machine the features, there is additional must be amortized over the production quantity, so a long tool
time required to machine the parting plane. In some cases the life is advantageous.
parting line is curved, and this significantly increases the fab- As a generalization, MIM is not cost effective in hard tool-
rication cost. In other words, a flat parting line is most typical ing for production quantities below 5,000 parts per year. Using
and least expensive, while complex parting planes described soft tooling on the other hand, some ceramic and cemented
by curvatures and steps add to the tool expense. carbide components are produced in lots as small as 200.
Tooling tolerances are more stringent than part tolerances, Most MIM shops can operate for 100,000 shots without
and in many estimates the tool tolerance is held to less than tool maintenance. Several coating technologies extend tool
20% of the component allowance. Accordingly, considerable life, especially in resisting wear from harder and more abra-
time is used to fabricate a precise tool for a tightly toleranced sive particles. The coatings include intermetallic compounds,
component, especially when the number of features is large. boride layers, hard nickel-phosphorus electroplates with em-
The specified tool surface finish incurs a cost, especially bedded polytetrafluoroethylene particles, hard sputter coat-
if a polished tool is required. In a similar manner, texturing, ings, and various spray, vapor, or reaction interfaces. Although
lettering, or insignias on the tooling add to the tool fabrication they all add expense to tool creation, there are possible gains
time and cost. over time for larger production quantities.
Machining cost is determined by adding the individual ma-
EQUIPMENT COSTS
Capital equipment creates a cost burden for MIM in the
TABLE 6.2. COMPARISON OF LABOR RATES IN form of depreciation. Across the industry arises a unit manu-
SOME MIM COUNTRIES facturing cell concept roughly consisting of 1 mixer, 4 mold-
country base rate, $/h relative increase, %/y ers, and 2 furnaces. Of course, in about 28% of the operations
there is no mixer, since that portion of the MIM industry pur-
Germany 58.50 2.5
France 47.81 5.4 chases feedstock. Most operations depreciate these devices in
Sweden 41.73 2.6 5 to 7 years. This means the purchase price is subtracted as
Japan 38.63 0.4 an expense in yearly increments, independent of the business
U.S. 33.00 2.1 for the year. Some example production equipment costs are
Spain 31.62 3.8 given in Table 6.3 for common purchases.
Korea 21.10 12.4
Mexico 3.30 3.6 The sintering furnace tends to be the most expensive item.
China 3.00 7.6 For low-volume production, an output of nearly 60 kg per day
(132 lb. per day) is possible from a $300,000 furnace, and
99
Table of Contents
Chapter 6: Costs
about three times this (300 kg per day or 660 lb. per day) from and depreciation schedules.
a $900,000 furnace. Continuous furnaces are more expensive, Chapter 9 details an approach to component costing com-
but the capital depreciation on a per part basis is lower because posed of several stages. The first stage is associated with the
of higher productivity. material from which the component is fabricated. The second
Besides the devices directly associated with production, stage is associated with the operating environment. The third
additional items are needed for quality control, materials test- stage is associated with unit operations and the determination
ing, research and development, and general manufacturing of cost allocation per piece at each step. Finally, risk and pro-
support. Table 6.4 provides an assessment of the equipment duction balancing are added cost factors. This information will
options and associated capital equipment needed to start a not be repeated here.
pilot operation. In addition to the equipment, there is a larger As an illustration of the cost partition between various con-
need for operating cash during the time when the operation is tributors, two example components are introduced here.
stabilized. Golf clubs by MIM are a favorite topic. Many demonstra-
tion club styles have been fabricated (driver, wedge, and put-
COMPONENT COSTING ter) and a few reached production status, including putters out
The cost for producing a MIM component is scattered, re- of titanium, stainless steel, and zirconia. Drivers were fabri-
flecting differences between production sites in terms of ca- cated out of titanium, but the costperformance combination
pacity, equipment, technology, labor rates, and raw materials was not satisfactory to reach production. First consider the
consumption. Many MIM operations do not have the broad cost breakdown for a MIM stainless steel sand wedge, a case
experience needed for accurate costing calculations. Signifi- where MIM is not competitive vs. investment casting.
cant errors occur by improperly estimating process yield, re- Figure 6.2 is a photograph of the target wedge design. The
work rates, cycle times, furnace-loading factors, daily cost factors involved in MIM production of a sand wedge are
capacities, employee-benefit costs, mold-maintenance costs, detailed in Chapter 9. For this analysis, assume the wedge is
formed from 17-4 PH stainless steel to a final weight of 200 g.
100
Table of Contents
Chapter 6: Costs
Based on a production volume of 200,000 per year and 98% nated by the powder price. Because of the component size and
process yield, the cost breakdown is shown in Figure 6.3, as- shape, debinding is slow (20 hours) and batch sintering-fur-
suming industry typical values. The dominant feature is the nace loading is only 20% efficient due to conformal setters,
powder cost at $1.87 per part due to the component mass. The so the 18-hour sintering cycle contributes 25% of the manu-
mixing and binder cost per part is $0.23 and the material loss facturing cost. Thus, total cost for MIM production of the sand
in processing makes about half of the total manufacturing cost wedge is estimated at $4.53. The molding tool is optimized at
assignable to the feedstock cost. The feedstock cost is domi- a single cavity with a one-time cost of nearly $25,000. If
amortized over 200,000 parts, then the mold combined with
the processing cost give a per part total of $4.77.
Based on this manufacturing cost, Figure 6.4 shows the
sale price breakdown, where the above costs are supplemented
by overhead and profit, making the final price per part $6.12
with a 10% profit margin. This reflects an industry low of just
over $30 per kg while the industry average is $133 per kg ($60
per pound). For reference, a MIM putter out of stainless steel
went into production at 66,000 per year at a slightly higher
price, giving credibility to these estimates.
Figure 6.3. Pie chart showing how cost is partitioned for the Figure 6.5. Cellphone stainless steel knuckle component with 2 g
fabrication of a 200 g stainless steel sand wedge, where the high mass considered in the cost calculations
material cost is a dominant factor; the estimated manufacturing
cost is $4.53 each
101
Table of Contents
Chapter 6: Costs
As another example, consider a 2 g stainless steel cellphone quate, then MIM has a distinct cost advantage. However, for
component shown in Figure 6.5. Details of the calculations large, simple shapes, MIM routinely proves unattractive. As
are given in Chapter 9 for this component. It is fabricated early as 1988 demonstration stainless steel parts of 12 kg were
using MIM at production rates near 1 million per month. The produced, but proved unsuccessful economically. The author
calculated component price was $0.45 each after reaching full has personally been involved in the production of components
production (equivalent to $225 per kg) and this is close to the reaching up to 40 kg, establishing the technical feasibility, but
value quoted for the project. The breakdown of this price is alas financial feasibility was lacking.
illustrated in Figure 6.6. Note that the price per unit mass at Each shaping technology is best suited to a particular com-
$225 per kg makes this a more expensive component com- bination of materials, tolerances, sizes, shapes, and properties.
pared to the sand wedge. For this shape, the complexity made For example, casting techniques excel in shape complexity,
MIM the best-choice production route. Curiously, the original but lack in surface finish and dimensional precision. On the
attempt was to form this part out of titanium, but the added other hand, at high production quantities, machining tech-
cost from even inexpensive titanium powder doubled the part niques suffer with material wastage and production costs. All
cost; thus, stainless steel was used in production. technologies are evolving, making for dynamic competition.
Across the MIM industry there is considerable variation in For every component that MIM captures, there is a competing
costs, markets, and production approaches. However, surveys technology working to win that component on the next design
show that most of the successes are priced near $1 per part. iteration, resulting in an intense competition between tech-
But prices vary considerably. Yet surveys show that the typical nologies. Cost is usually the critical parameter in selecting a
MIM component is selling at from $133 per kg. However, net-shaping technology.
some of the smaller 0.1 g components equate to pricing near
$600 per kg. Considerable variation exists in the price based DESIGN EFFECTS ON PROCESSING COSTS
on different production factors. Early consultation between the designer and component
producer leads to recommendations that will lower the pro-
COMPETITIVE FORCES duction costs. These range from use of thinner cross sections
For metals there are a host of alternative routes for the fab- to combinations of parts into a single piece. The best analysis
rication of a given design. Thus, competing technologies protocol relies on asking many questions concerning the can-
greatly influence the viability of MIM on any project. Table didate design. There are subtle but important changes that im-
6.5 compares the advantages and limitations of various net- prove process yield and reduce tooling costs, and these are
shape technologies against which MIM competes. Users say generally beneficial. In any design, a few features add sub-
that the decision to use MIM over competing technologies stantially to the cost. Custom MIM producers help avoid
usually starts with cost, but includes issues of shape capabil- costly mistakes by suggesting design changes to reduce man-
ities, productivity, surface finish, and precision. The cost ad- ufacturing costs.
vantage for MIM vs. other forming techniques increases as The approach to cost reduction builds from the 8020 rule.
the number of machining, grinding, or finishing steps grows, Which 20% of the design features control 80% of the cost?
especially for small components. Further, the cost of machin- The answer provides focus to cost-reduction efforts. Surface
ing increases dramatically as the surface finish improves. finish is a good case. If the surface specification is too smooth,
Rough ground or turned surfaces can be produced at just 25% then it adds considerable finishing cost after sintering. A so-
of the cost to form the smooth surfaces associated with as-sin- lution might be to intentionally design for a rippled or textured
tered MIM structures. If an as-sintered surface finish is ade- surface that avoids polishing, yet has a suitable aesthetic
102
Table of Contents
Chapter 6: Costs
appeal. A few places to find 20% of the features that control RESOURCES
80% of the cost include: S.V. Atre, C. Wu, S.J. Park, R.M. German, C.J. Hwang and
nonstandard materials: use common materials if possible R. Zauner, Technical and Economic Comparison of Micro
and graduate to a stronger or more corrosion-resistant Powder Injection Molding, Proceedings Powder Metallurgy
material with a lower cost if it is a widely used alloy; for World Congress, Korean Powder Metallurgy Institute, Busan,
example, in one diesel-engine application the change Korea, 2006, pp. 4546.
from 15-5 PH stainless to 17-4 PH stainless resulted in
G. Boothroyd, P. Dewhurst, W. Knight, Product Design for
improved corrosion resistance and a lower cost
Manufacture and Assembly, second edition, 2002, Marcel
combine parts where possible to eliminate assembly, in-
Dekker, New York, NY, pp. 359379.
ventory, and time spent in molding: one automotive part
was molded from two stainless steels, one magnetic and G. Dieter, Engineering Design: A Materials and Processing
one nonmagnetic, saving in joining costs Approach, third edition, 2000, McGraw Hill, Boston, MA.
look for purchasing quantities that enable better cost ef-
R.M. German, Medical and Dental Applications for Micro-
ficiencies at the MIM producer: for example, if the an-
miniature Powder Injection MouldingA Roadmap for
nual requirement is small (under 20,000 pieces), it is
Growth, Powder Injection Moulding International, 2009,
better to purchase once per year
vol. 3, no. 2, pp. 2129.
components with uniform wall thicknesses are easier to
fabricate and can be fabricated with higher tolerances A.C. Goodrich, Evaluating Technology Strategies for the
and more uniform properties: remember, thick sections, Manufacture of Near-Net-Shape Metal Automotive Compo-
on a component slow the molding cycle and add cost nents, P/M Science and Technology Briefs, 2002, vol. 4, no.
plan for quick ejection from the tooling: it is beneficial 3, pp. 510.
to place undercuts, threads, and textures in locations
R.D. Stewart, Detailed Cost Estimating, Mechanical Engi-
where they do not delay ejection
neers Handbook, second edition, M. Kutz (ed.), 1998, Wiley
place molding blemishes in locations where they are not
Interscience, New York, NY, pp. 2,1172,142.
critical, thereby eliminating any expensive finishing
steps
try to incorporate a flat surface for component support
during debinding and sintering: this avoids the cost of
special fixtures
tight tolerances add greatly to the cost and need to be
critically analyzed to ensure they are justified
This final point is one of great discussion in the MIM field.
Tight tolerances are possible. However, the concomitant in-
crease in equipment sophistication (cost), lower yields, or the
reliance on expensive secondary operations all dramatically
increase component cost. A normal, or Gaussian, distribution
in dimensions typically is associated with MIM. If the toler-
ances are tight, then less of the product will be within speci-
fication. Accordingly, sorting is required to separate those in
specification from those not acceptableand this adds cost
because of the extra step. More of an impact is from the ex-
pense associated with production of the rejected components,
an expense that must be captured in the final shipped material.
One case of a fiber optic component reports only 7 out of
1,000 were in specification. Fortunately, the price for each
was sufficiently high (it was a military sale, sort of like the
$600 toilet seat), so the product was profitable.
All MIM products are eventually limited in terms of di-
mensional precision. As-sintered dimensional scatter is re-
duced by post-sintering machining, but with added production
steps and costs. Early discussions with the vendor provide one
means to ensure the lowest-cost approach to tight tolerances.
103
Table of Contents
Table of Contents
Chapter 7
APPLICATIONS
Outline Overview
Aerospace
Automotive
Business Machines
Cellphones
Computer Devices
Cutting Tools
Defense
Dental
Electrical and Electronic Components
Firearms
Hand Tools
Hardware and Industrial
Household and Consumer
Instrumentation and Sensors
Jewelry and Watches
Medical
Microelectronics and Optoelectronics
Sporting
Telecommunication
Summary
Resources
105
Table of Contents
Chapter 7: Applications
AEROSPACE
Two early MIM components attracted much attention
because they were for aerospace applications, an area where
presssinter powder metallurgy fails to qualify. One was a nio- Figure 7.2. Flow body with mass over 1 kg, formed from
bium rocket nozzle and the second was a seal ring for a Boe- superalloy IN 718
ing commercial airliner.
The niobium alloy (Nb-10W-10Ta) rocket nozzle shown
in Figure 7.1 was fabricated by Parmatech (Petaluma, Cali-
fornia, U.S.) for Rocketdyne (Canoga Park, California, U.S.).
This single-piece nozzle was coupled to a two-piece injector
that was assembled by electron beam welding. The design was
for satellite and missile control using liquid propellants, where
during operation the nozzle-wall temperature exceeds 1,375C
(2,500F).
In the aerospace applications the material and process need
to be qualified for the use temperature, stress, and environ- Figure 7.3. Sintered (left) and molded (right) swirler component
ment. Accordingly, the extreme environments often lead to (photograph courtesy of Latitude Manufacturing)
106
Table of Contents
Chapter 7: Applications
cycles are long. For flight-critical uses, the components are near the hot sections of jet engines. Figure 7.2 is a picture of
hot isostatically pressed to full density, since residual porosity a flow body after sintering and machining; it is molded at
is unacceptable. 1,630 g (3.6 lb.) from gas-atomized IN 718 nickel-base super-
In the past decade, several other MIM components have alloy powder, but after machining the mass is 1,130 g (2.5 lb.).
moved into use for aerospace applications. Prior to accept- Similarly, Figure 7.3 is a picture of a swirler aerospace com-
ance, extensive test data are required, such as fracture-tough- ponent in both the molded and finished conditions. It was sub-
ness, fatigue-life, and stress-rupture properties. Determination jected to extensive testing and found to have excellent
of these properties under a representative range of tempera- properties.
tures, stresses, and environments requires expensive testing, Another MIM success is the butterfly valve plate pictured
so companies do not embark on this application area without in Figure 7.4 in both the top and bottom orientations. This
substantial resources. Products realized to date are valves, component is fabricated by Precision Castparts Advanced
pumps, actuators, hydraulic bodies, nozzles, and components Forming Technology (Longmont, Colorado, U.S.) for Honey-
well Engine Systems and Accessories (Tempe, Arizona, U.S.)
out of the IN 718 alloy. The component is 27 g in mass and
37 mm in outer diameter (1.5 in.). Properties depend slightly
on the powder size, oxygen content, carbon content, hot iso-
static pressing cycle, and heat treatment cycle. Nominally, the
properties by MIM are as follows:
density = 8.22 g/cm3
yield strength = 1,000 MPa (143 ksi)
ultimate strength = 1,330 MPa (193 ksi)
elongation = 29%
hardness = 42 HRC
fatigue-endurance limit = 380 MPa (55 ksi)
Surprisingly, even without hot isostatic pressing the IN 718
tensile properties are nearly the same, although the fatigue-
endurance limit improves by 20% with hot isostatic pressing.
On this basis, HIP will remain a part of the fabrication process.
The MIM fabrication of high-reliability components for
demanding aerospace uses is the subject of much current
research. One example component is depicted in Figure 7.5.
It is a superalloy turbine designed for a small jet engine for
taxi aircraft. As with many MIM aerospace components, it
has been tested extensively but not yet taken to production.
107
Table of Contents
Chapter 7: Applications
AUTOMOTIVE
The use of MIM in automotive applications started in the
late 1980s with Chrysler lock components and Buick turn-sig-
nal levers. Since then a large number of automotive MIM
components have reached production, and many early suc-
cesses have passed out of production. Other successes are
found in transmission synchronizers, airbag sensors, electric-
seat gears, interior knobs, spark-plug bodies, turbochargers,
fuel injectors, valve lifters, heat dissipaters, and engine sen-
sors. Some early examples are collected in Figures 7.6 through
7.10; in order, these are an ignition key component, timing
ratchet, airbag actuator, airbag-tilting sensor, and magnetic
position sensor. Figure 7.11 is a convertible-roof top clip
(shown green in Figure 2.13b) fabricated by GKN Sinter Figure 7.8. 17-4 PH stainless steel firing initiator for deploying
Metals (Bad Langensalza, Germany) for Volkswagen. the airbag in an automotive crash (component courtesy of Key
Use of metal powder injection molding for automotive Automotive Group)
Figure 7.6. Stainless steel MIM pivot component for automotive Figure 7.9. 17-4 PH stainless steel pivot, used for sensing an
key, showing the component and the final key assembly automobile crash (component courtesy of Hunan Injection High
Technology)
Figure 7.7. Integrated ratchet and shaft formed from 316L Figure 7.10. Rotational sensor formed from a magnetic outer stain-
stainless steel (component courtesy of Advanced Materials less steel and nonmagnetic inner stainless steel for positional sens-
Technologies) ing (component courtesy of Advanced Materials Technologies)
108
Table of Contents
Chapter 7: Applications
109
Table of Contents
Chapter 7: Applications
Alternatively, several automotive applications rely on high- component has a wall thickness of just 0.7 mm along the stem
strength, low-alloy (HSLA) steels. The steels are heat treated in some regions. Besides the dimensional challenges, it is
after sintering to maximize mechanical properties. Automo- required to resist corrosion and withstand hydraulic-fluid leak-
tive structural steels are heat treated to tensile strengths near age. The small-diameter hole proved easier to fabricate with
1,800 MPa (260 ksi), with at least 3% elongation, hardness MIM than with other technologies.
up to 50 HRC, impact energy over 100 J/cm2, and fatigue A related high-strength automotive development is the
endurance strength up to 500 MPa (85 ksi). For transmission steering clevis pictured in Figure 7.15, fabricated by Precision
components, the rolling-contact-fatigue strength must exceed Castparts Advanced Forming Technology (Longmont, Col-
2,000 MPa (290 ksi). The heat-treated steels include 17-4 PH orado, U.S.) for Delphi Saginaw Steering Systems (Saginaw,
and Fe-1Cr-0.2Mo-0.8Mn, with carbon and silicon for hard- Michigan, U.S.). The ironnickel steel component weighs 65
ening. Higher alloying levels are required for applications g and is fabricated to the following properties:
involving extreme stress and wear, up to 8% Cr, 2% Mo, and density = 7.5 g/cm3 (less than 4% porosity)
1% Ni. One example is an all-terrain brake component with yield strength = 1,450 MPa (210 ksi)
overall length of 40 mm and mass of 11.6 g. Two variants fab- tensile strength = 1,680 MPa (245 ksi).
ricated from 17-4 PH are shown in Figure 7.14 as fabricated Variable-valve rocker arms are another example of high-
by NetShape TechnologiesMIM (Solon, Ohio, U.S.). The performance MIM automotive components. These were first
used by Honda (Tochigi, Japan), fabricated by Nippon Piston
Ring (Shimotsuga-gun, Tochigi, Japan), from a heat-treatable
low-alloy steel (Fe-1Cr-0.7Mn-0.2Mo-0.2Si-0.2C). The com-
ponent was processed to a tensile strength of 1,100 MPa (160
ksi) and fatigue-endurance strength of 300 MPa (44 ksi). The
version shown in Figure 7.16 is a variant fabricated by Schunk
Sintermetallwerk for BMW.
An example of MIM automotive growth is found in fuel
injectors. The molded versions were first introduced by Cum-
mins Engine for heavy-duty diesel engines, but are now used
in gasoline engines. The fuel injectors rely on high strength
in a hard, corrosion-resistant, and wear-resistant material.
Shapes vary, but generally are in the 10 to 12 mm (under half
Figure 7.14. Stainless steel brake fitting for an all-terrain vehicle
in.) length range, with diameters near 6 to 7 mm (quarter in.),
(photograph courtesy of NetShape TechnologiesMIM)
with flanges and tapered internal bores that are 100 to 150 m
in diameter (0.004 to 0.006 in.). Inner dimensions are added
by post-sintering grinding. Variants are fabricated from pre-
cipitation-hardened stainless steels and ceramics.
Metal powder injection molding is being qualified for
under-the-hood thermal-management applications: for exam-
Figure 7.15. Steering-system clevis formed from a low-alloy Figure 7.16. Variable-valve rocker arms fabricated using a heat-
nickel-steel with a final mass of 65 g (photograph courtesy of treatable steel (photograph courtesy of Powder Injection Moulding
Metal Powder Industries Federation) International)
110
Table of Contents
Chapter 7: Applications
ple, in the control systems for hybrid-vehicle power-switching a tape cutter formed from tool steel for a Pitney-Bowes
units. Copper is the leading candidate, where an internal vapor postage meter. Most of the applications used Fe-Ni steels
chamber enables heat-pipe cooling by evaporation and con- formed from mixed carbonyl iron and nickel powders.
densation inside the MIM device. Recent interest has come from computer equipment fabri-
cators for inkjet-printer nozzles and laser-printer components.
BUSINESS MACHINES Even the metallic logo for Hewlett-Packard and Apple com-
Business machines are a mature application for metal pow- puters relied on MIM. Opportunities continue to exist in var-
der injection molding. This area was early to adopt the tech- ious handheld office machines, magnetic components, and
nology, largely for printers, postage machines, typewriters, optical components, but as mechanical drives are eliminated
and dot-matrix printers. Figure 7.17 shows an early MIM from office devices the opportunities for MIM will be small.
Epson dot-matrix printer head. It was formed using 17-4 PH
stainless steel and was one of the first components produced CELLPHONES
using continuous sintering. Mobile telephones are now produced at quantities of 1
Another early success was in the tilt ring used in an electric billion per year. For many years MIM has played a key role.
typewriter printhead based on Fe-2Ni; production by MIM at Applications include hinges, swivels, knuckles, covers,
IBM reached 1.5 million units. Similarly, postage machines lanyard loops, side buttons, supports, and vibrators. Some
relied on hard materials for wear components. Figure 7.18 is designs rely on at least four MIM parts. A few early product
examples are shown in Figure 7.19: a thin-walled metallic
cover, knuckles or corner pieces that enable the flip open
action to answer the phone, and a camera phone swivel.
Recently pressures to increase features and reduce the size
have pushed MIM into microminiature components. Two
examples fabricated by Taiwan Powder Technologies
(Taoyuan, Taiwan) are shown in Figures 7.20 and 7.21. The
first is the mount for the audio jack on a smart cellphone and
the second is termed a hanger, also fabricated from 17-4 PH
stainless steel.
One of the more taxing campaigns involved the MIM fab-
rication of a flip slider and hinge barrel, pictured in Figure
7.22 along with the assembled handheld unit. These stainless
steel MIM components are used because of their low cost that
still offers three-dimensional shape complexity in a thin struc-
Figure 7.17. Dot-matrix printer head formed from 17-4 PH ture. The design positions the clamshell phone cover so it
stainless steel. This was one of the early business machine
slides down and flips open in a single motion. Since compo-
successes for MIM (component courtesy of Raymond Weich)
nent alignment was critical, coining was performed after sin-
tering. These sets were produced at rates of 6 million per
month by several firms, including Advanced Materials Tech-
nologies (Singapore) and Parmatech (Petaluma, California,
U.S.) for Motorola (Libertyville, Illinois, U.S.). Both compo-
nents are from 17-4 PH stainless steel with the following
attributes:
final density = 7.6 g/cm3 (less than 2% residual porosity)
yield strength = 1,090 MPa (158 ksi)
ultimate strength = 1,186 MPa (172 ksi)
elongation to fracture = 6%
hardness = 30 HRC
A high strength is required in this design since the corners
were very thin. A special difficulty came from impact tests
based on simulating a dropped phone. In production, the com-
ponents were measured for several critical tolerances and
Figure 7.18. Pitney-Bowes postage meter MIM tool-steel
especially were monitored for perpendicularity of 0.2 mm
component used to handle the wear and corrosion associated with (0.008 in.) at two points. Corrosion resistance was mandatory
postage tape as well, so steel was not acceptable. Indeed, the parts were
111
Table of Contents
Chapter 7: Applications
112
Table of Contents
Chapter 7: Applications
chrome finished prior to assembly. Data from Motorola hard drives are present in music players, automobiles, global
reported 95% MIM process yield. Most of the loss occurred positioning systems, and industrial tools. As the size shrank,
during debinding and sintering. For comparison, the cost to a wealth of opportunities for MIM arose to include balancing
fabricate this component by alternatives was five times more weights, latches, actuators, connectors, and other mechanical
than via MIM. or magnetic components. In the early 2000s production accel-
erated with small disk drives that critically relied on MIM
COMPUTER DEVICES components. Following along, other computer devices
Since the early 1980s the computer hard drive has under- adopted the MIM technology for connectors, latches, pivots,
gone a dramatic size decrease and capacity increase. Today sockets, and heat sinks.
An early MIM computer disk drive component is shown
in Figure 7.23 fabricated from 316L stainless steel. It is known
as a pivot hub, fabricated by Parmatech (Petaluma, California,
U.S.) for Hewlett-Packard. It was used in storage and posi-
tioning mechanisms for high-density disk drives. The sintered
density was 7.5 g/cm3 (about 5% porosity) and the tensile
strength was 500 MPa (73 ksi). However, aluminum postage-
stamp-sized disk drives for handheld Apple iPod devices
emerged and displaced the need for stainless steel.
Due to semiconductor heating during use, heat-dissipation
application arose for MIM. Early designs included high-ther-
mal-conductivity pin arrays to help cool computer chips. The
pin array was formed out of copper or other high-thermal-con-
ductivity materials. Designs targeted a thermal conductivity
of at least 200 W/(m K) and a thermal-expansion coefficient
of 6 ppm/K. More recent designs moved to pin arrays from
copper, such as shown in Figure 7.24.
One difficulty in serving the computer field is the rapid
progression in designs that make it difficult to keep a compo-
nent in production for an extended period. A six-month prod-
uct life cycle is common. However, the savings realized by
MIM vs. casting are significant; machining a casting to com-
puter industry tolerances is 22 times more expensive than via
MIM. For cost and property reasons, MIM has become the
Figure 7.23. Hard disk drive pivot hub formed from 316L stainless process of choice for small, complicated metallic components.
steel (photograph courtesy of Parmatech) Thus, it is routinely used in laptop, notebook, and other small
devices.
CUTTING TOOLS
Hard materials for wood and metal cutting are usually
molded from cemented carbides, but a few rely on cermets or
tool steels. Cermets are mixtures of metals and ceramics (TiC-
Ni, for example) and cemented carbides are also known as
hard metals (WC-Co). About 1% of the tungsten carbide cut-
ting inserts are produced by injection molding. The use of
MIM for tungsten carbide cutting tools is relatively small,
since it only amounts to about 4%5% of the commercial
MIM production. MIM products are used in woodcutting
routers, metal drills, end mills, metal-cutting inserts, and drill
tips.
An example of MIM from WC-Co is shown in Figure 7.25.
Figure 7.24. Cold plate fabricated by MIM from copper to These are concrete abrasive tips. Figure 7.26 shows two
maximize heat exchange using an array of 279 pins (photograph metal-cutting tools; the first is an end mill with titanium
courtesy of Acelent Technologies) nitride coating and the second is a rotary cutter. The drill
113
Table of Contents
Chapter 7: Applications
DEFENSE
Defense applications were one of the early success areas
for MIM. An early product was the 316L stainless steel stabi-
lizer fin shown in Figure 7.29. The fin was produced at three
MIM facilities in California for Ford Aerospace and Commu-
nications. A final coining treatment was used to straighten the
fins after sintering.
Defense applications include handguns, weapon sights,
missile fins, fragmentation projectiles, bullets, practice ammu-
Figure 7.25. MIM tungsten carbide wear tips used in the impact Figure 7.26. Cutting tools formed using tungsten carbide
abrasion of cement (photograph courtesy of Powder Injection compositions include the (a) end mill and (b) woodcutting tool
Moulding International) (components courtesy of SECO Tools and CetaTech)
114
Table of Contents
Chapter 7: Applications
Figure 7.28. Saw-wire bead (a) and wire-cutting tool (b) formed
using diamond super-abrasive compositions developed specifically
for MIM (photographs courtesy of Nick Williams of Powder
Injection Moulding International)
115
Table of Contents
Chapter 7: Applications
DENTAL ing for the process, such as the devices shown in Figure 7.34.
A mainstay application for MIM has been orthodontic These are 17-4 PH stainless steel parts for a self-aligning tooth
brackets. These have been in continuous production since the positioning system. The pieces are produced at quantities up
mid-1980s. Example brackets are shown in Figure 7.33, illus- to 600,000 per week by FloMet (DeLand, Florida, U.S.) for
trating the waffle pattern on the bottom and the wire attach- Ormco Sybron Dental Specialties (Orange, California, U.S.).
ment posts and grooves on the front. Previously, these were One bracket and one slide go on each tooth with the hook used
made from welded bands. The small MIM brackets allow in some cases. Another new design is shown in Figure 7.35.
direct attachment to each tooth. The predominant materials
are stainless steels (316L and 17-4 PH), but the offering has
expanded to include nickel-free compositions such as cobalt
chromium alloys, nitrogen-stabilized austenitic stainless steel,
and most recently nitinol (NiTi), titanium, and transparent alu-
mina.
Several larger companies currently produce dental brackets
at production volumes exceeding 250,000 brackets per day.
Recent designs have taken full advantage of MIM by design-
Figure 7.31. Steel defense component formed by MIM used as a Figure 7.34. New design orthodontic brackets with self-aligning
retrofit device on military vehicles (photograph courtesy of designs (photograph courtesy of Metal Powder Industries
Polymer Technologies) Federation)
116
Table of Contents
Chapter 7: Applications
Figure 7.36. Endodontic tips such as shown here are for root canal
treatments. They are formed using MIM with the generation of Figure 7.37. Scanning electron micrograph of a titanium implant
different burr sizes and shapes and patterns; (a) tool with a length with porous titanium band on a solid post for enhanced tissue
of approximately 35 mm, and (b) scanning electron micrograph of attachment (image courtesy of Eric Baril, National Research
the working tip (photographs courtesy of CetaTech) Council of Canada)
117
Table of Contents
Chapter 7: Applications
ment region is evident in the threaded portion, while the Electronic packaging relies on low-thermal-expansion-
remainder of the sintered structure is dense. Actually, the core coefficient materials, such as Kovar, Invar, or tungsten
is solid and the porous region is only on the outside as a band. copper, mounting ceramics such as alumina, or pure metals
such as aluminum, copper, molybdenum, or tungsten. The
ELECTRICAL AND ELECTRONIC COMPONENTS choice depends on functional aspects, and typically
Electrical connectors and various insulator or conductor if thermal expansion is a dominant concern, then the
components for electrical service have been produced by choice is Invar;
MIM since the mid-1980s. One of the early applications was if glassmetal sealing is a dominant concern, then the
in multiple-layer circuits, where molybdenum powder was choice is Kovar;
injected into alumina laminates to form co-fired circuit pat- if heat dissipation is a dominant concern, then the choice
terns. Today, most of the MIM applications are for packaging is tungstencopper.
devices, such as the hermetic (glassmetal sealed) device A MIM electronic package is shown in Figure 7.39. It is a
shown in Figure 7.38. In addition, electronic applications hearing-aid receiver capsule fabricated by FloMet (Deland,
include electronic hardware (latches, locks, hinges) and Florida, U.S.) for Starkey Laboratories (Eden Prairie, Min-
mounts fabricated from stainless steel, with 316L being a nesota, U.S.). This thin-walled component is formed using a
favorite. Insulators are commonly fabricated using powder nickelironmolybdenum alloy to provide a magnetic shunt
injection molding from ceramics such as alumina. for the hearing aid to separate the receiver signal from the tele-
coil signal. The internal undercut especially made MIM an
optimal approach so the cost reduction in shifting to MIM was
50% when compared to the previous deep-drawing-and-
annealing approach.
FIREARMS
Handguns, shotguns, rifles, and military weapons use sev-
eral metallic components fabricated by MIM. Typical firearm
components are sights and sight mounts, handgun bodies,
safety levers, sling attachments, catches, firing pins, maga-
zines, triggers and trigger guards, and slide blocks. Steels are
typical and most of the production quantities are under
100,000 per year for several years. After sintering, the steels
are heat treated and coated with protective black oxides to
inhibit rusting. Figure 7.40 depicts a steel trigger guard
Figure 7.38. Three stages in the production of an electronic
package, showing the Kovar box formed by MIM, the glassmetal
sealed leads, and the final gold electroplated device (components
courtesy of Hunan Injection High Technology)
118
Table of Contents
Chapter 7: Applications
formed by Megamet Solid Metals (Earth City, Missouri, U.S.) fabricate whole handgun bodies. Pictured in Figure 7.41 are
for Modern MuzzleloadingKnight Rifles (Decatur, Alabama, the handgun body and cylinder and other parts for a police
U.S.) to tolerances of 0.12 mm (0.005 in.). The 88 g compo- revolver fabricated for the Chinese police forces by Beijing
nent is processed to achieve 650 MPa (94 ksi) tensile strength. Jarain Powder Injection Technology Limited Company
A most impressive development has been use of MIM to (Huairou District, Beijing, China). The material is a mixture
of carbonyl iron and nickel powders, molded using a thermo-
plastic feedstock, thermally debound, and sintered in vacuum
to full density.
Another firearm component is the 420 stainless steel hand-
gun housing block shown in Figure 7.42, fabricated by Par-
matech (Petaluma, California, U.S.). It is used in a 45-caliber
handgun to hold the spring mechanism to give the desired slid-
ing action. It is a good challenge to MIM, since it has under-
cuts, wings, through holes, blind holes, and both thick and thin
sections. Some specifics on the component are as follows:
final density = 7.7 g/cm3 (about 2% residual porosity)
yield strength = 1,500 MPa (218 ksi)
ultimate strength = 1,800 MPa (261 ksi)
hardness = 50 HRC
A firearm device for military use is pictured in Figure 7.43,
fabricated by Megamet Solid Metals (Earth City, Missouri,
U.S.) for Colt Canada (Kitchener, Ontario, Canada). It is a
4140 low-alloy steel folding front grip used on military rifles.
Sintering is in a nitrogen atmosphere to improve strength.
Post-sintering machining is used to ream and tap the holes.
The component is heat treated to a hardness of 4550 HRC
and a manganese phosphate finish is added after heat treating.
A related development is taking place in the fabrication of
ammunition. Metal powders such as tungsten are molded to
form practice bullets and green bullets for sites where lead
contamination is not acceptable. Environmental concerns on
both practice ranges and waterfowl habitats have led to a ban
Figure 7.41. Police handgun components fabricated from a iron
nickel steel using MIM: (a) two views of handgun and (b) spindle
and other parts (photographs courtesy of Beijing Jarain Powder
Injection Technology Limited)
Figure 7.42. Housing block for a Colt 45-caliber handgun Figure 7.43. Front grip on a military rifle generated using MIM
fabricated using MIM from 420 stainless steel (photograph from a 4140 steel (photograph courtesy of Metal Powder
courtesy of Metal Powder Industries Federation) Industries Federation)
119
Table of Contents
Chapter 7: Applications
HAND TOOLS Figure 7.46. 316L stainless steel compressed-air nozzle to control
Early demonstrations of MIM hand tools occurred in con- air flow and noise over a range of operating environments
sumer devices, including some high-visibility brands, such as (photograph courtesy of Metal Powder Industries Federation)
the popular Sears Craftsman socket wrenches and locking
pliers. Several shapes were produced by MIM for the various nozzle controls the flow of compressed air while reducing
multi-purpose hand tools, sort of the Swiss Army knife noise in a range of temperatures and corrosive environments.
approach to hand tools. These include screwdrivers and The component is formed from 316L stainless steel to deliver
wrenches, and feature combinations as illustrated in Figure the following properties:
7.45. This 220 g component has a length of 122 mm (4.8 in.) final density = 7.6 g/cm3 (about 3% porosity)
with tolerances of 0.2 mm (0.008 in.). It is fabricated from yield strength = 172 MPa (25 ksi)
tool steel at production rates of 20,000 per year. The compo- ultimate strength = 517 MPa (75 ksi)
nent is subjected to sandblasting and polishing as final oper- elongation to fracture = 50%
ations. It is fabricated by Hunan Injection High Technology hardness = 67 HRB
(Changsha, Hunan, China). The hollow plenum chamber is laser sealed after sintering.
The most demanding hand tool applications are associated In other instances, firms have been successful in avoiding the
with professional service, such as tools used on construction laser-sealing treatment, resulting in further cost savings.
sites. Several professional-tool firms use MIM components
for high-durability wrenches, pliers, wire cutters, and pipe HARDWARE AND INDUSTRIAL
threaders. To match the target mechanical properties, the met- Considerable opportunity exists for metallic injection
als are either tool steel, high-strength stainless steels, or heat- molded products for hardware applications, such as machine
treatable steels. After sintering, these alloys are heat treated parts, components for pumps, and spray nozzles. Good suc-
to tensile strengths from 1,400 MPa to 2,000 MPa (roughly cess is seen in stainless steel, bronze, and copper for plumbing
200 to 300 ksi). valves, stems, fittings, locking mechanisms, hinges, latches,
Figure 7.46 is a picture of hand tool fabricated by FloMet decorative fixtures, and fittings such as pictured in Figure
(DeLand, Florida, U.S.) for Silvent (Boraes, Sweden). The 7.47. One example is the 17-4 PH stainless steel lock-cylinder
120
Table of Contents
Chapter 7: Applications
components shown in Figure 7.48. These are fabricated by have superior-performance locking devices and allows 30%
Kinetics Climax (Wilsonville, Oregon, U.S.) for Black & cost saving over other processes. Testing includes long-dura-
Decker Hardware (Lake Forest, California, U.S.). The locking tion (300,000 rotation) tests and 48-hour salt-spray corrosion
bar, pin, and rack operate in a lock cylinder that contains a test. The high mechanical strength from MIM 17-4 PH is a
total of 11 MIM parts. They are fabricated to the following desirable attribute for this application.
specifications: Applications in plumbing are a similar story. A 17-4 PH
sintered density = 7.7 g/cm3 (essentially full density) stainless steel shower-faucet stem product became famous in
yield strength = 730 MPa (106 ksi) the MIM community because of its production challenges.
ultimate strength = 900 MPa (131 ksi) Figure 7.50 shows the component. This shower safety-valve
Related hardware components are found in the industrial system worked to prevent scalding due to changes in water
control lock pieces (barrel pin, square, barrel boss, and cover) pressure. It was fabricated by Parmatech (Petaluma, Califor-
that form safety switches in machine tools, shown in Figure nia, U.S.) for Moen (North Olmstead, Ohio, U.S.). MIM was
7.49. The components were initially formed from 316L stain- used to form the oversized blank which was then machined to
less steel with a high ductility, but are also formed from 17-4 final 5 m tolerances on automated machining centers.
PH stainless steel for higher strength. They are molded and Typical mass was 40 g for the stem part and the sintered den-
sintered to a 7.5 g/cm3 minimum density by Advanced Mate- sity was greater than 7.63 g/cm3.
rials Technologies (Singapore). Molding incorporates multiple
side core inserts in the tooling. This enables the customer to
Figure 7.47. Example of a hardware stainless steel fitting Figure 7.49. Components for industrial safety switches in machine
fabricated using MIM tools, including the control lock: barrel pin, square, barrel boss,
and cover (photograph courtesy of Advanced Materials
Technologies)
Figure 7.48. Lock-cylinder components fabricated using MIM Figure 7.50. Moen shower safety-valve components from 17-4 PH
from 17-4 PH stainless steel (photograph courtesy of Metal stainless steel (photograph courtesy of Metal Powder Industries
Powder Industries Federation) Federation)
121
Table of Contents
Chapter 7: Applications
122
Table of Contents
Chapter 7: Applications
HOUSEHOLD AND CONSUMER Similar early consumer products were the toothed grinding
Powder injection molding has found broad applications in portion of a pepper mill, parts for sewing machines, cutlery,
household, personal care, and consumer markets. Examples and hand-held nutcrackers.
include fingernail clippers, hair-trimming shears, cosmetic Figure 7.57 is a picture of a small, portable paper-hole
cases, toy trains, gears for electric toothbrushes, luggage punch consisting of several pieces exclusively fabricated
clasps, luxury fountain pens, eyeglass-frame hinges, can open-
ers, nutcrackers, pepper grinders, camera buttons and
switches, electric knives, pot handles, kitchen knives, wine-
bottle cork extractors, and food blenders. This is one of the
largest application areas for MIM and most of the components
are formed from stainless steel. Two examples are pictured in
Figures 7.54 and 7.55. The first shows a tripod pivot base
formed using MIM (Mimest S.p.A., Pergine Valsugana, Italy),
with the foreground version in titanium and the background
version in stainless steel. The second example is a toy train
where the stainless steel wheels are formed using MIM.
An early consumer application is shown in Figure 7.56.
These are 0.34 g (maximum dimension 6 mm or 0.25 in.)
stainless steel gears for an electric toothbrush. They were pro-
duced by Schunk Sintermetalltechnik (Giessen, Germany) for
Figure 7.55. Toy train that uses wheels fabricated via MIM
Braun Division of Gillette (Kronberg im Taunus, Germany).
Production exceeded 30,000 per day using 16-cavity tooling.
123
Table of Contents
Chapter 7: Applications
124
Table of Contents
Chapter 7: Applications
Figure 7.61. Luxury watch with links, spacers, bezel, and clasp all
formed by powder molding, and even the crystal is a spinel Figure 7.62. Stainless steel connector for use in arthroscopic
ceramic sintered to optical transparency imaging surgical treatments (component courtesy of MIMECRISA)
125
Table of Contents
Chapter 7: Applications
Microminiature medical devices are quite successful using moving health care toward short stays and single-use tools.
MIM. Concepts call for surgical manipulators with maximum The resulting cost pressures justify MIM as a solution. An
dimensions less than 5 mm. For example, Figure 7.63 is a pic- example surgical device is shown in Figure 7.65; it is a 17-4
ture of MIM stainless steel components used in heart surgery PH stainless steel articulator gear for surgical stapling fabri-
fabricated by Advanced Powder Products (Philipsburg, Penn- cated by Parmatech (Petaluma, California, U.S.). The device
sylvania, U.S.). The device is about 2.5 mm long. functions as the drive and locking mechanism for the device
Small MIM parts support efforts to reduce the incision size, head. This component costs just 30% of its machined prede-
benefitting the patient because of faster surgical recovery and cessor since it is produced net shape with no secondary oper-
shorter time in the hospital. In a similar manner, Figure 7.64 ations. The device is sintered to the following properties:
shows two components, one a manipulator and the other a final density = 7.65 g/cm3 (1% porosity)
biopsy cup. These are formed out of 17-4 PH stainless steel yield strength = 730 MPa (106 ksi)
by Hunan Injection High Technology (Changsha, Hunan, ultimate strength = 900 MPa (130 ksi)
China) at a rate of 5 million per year. Fabrication is to 20 m hardness = 25 HRC
tolerances (0.0008 in.) out of 3 mm (0.12 in.) length and com- Beyond surgical tools, MIM applications include implants,
ponent mass reaches to as low as 0.02 g. Some of the new ranging from replacement bones, replacement joints (knee and
robotic and small incision surgical tools are heavily dependent hip), and various tissue affixation devices. There is even an
on MIM to form the many links that make up the manipulators. effort to qualify new magnesium MIM alloys for absorbable
One robotic surgery device requires 16 different MIM parts. stents. Most of the medical instrument applications are satis-
A factor enabling MIM growth are insurance regulations fied by stainless steels, usually 304L, 316L, 420, 440C, or 630
(17-4 PH). However, about 10% of the parts and 25% of the
value will be in titanium alloys. Comparisons of the mechan-
ical, corrosion, and biocompatibility attributes show MIM is
competitive in titanium.
Other options are open to MIM, such as tantalum and
cobalt-base, but currently stainless steels are dominant.
126
Table of Contents
Chapter 7: Applications
127
Table of Contents
Chapter 7: Applications
or glassmetal sealing attributes. Early devices used ferrous Figure 7.70. One concept consists of a glassmetal sealing
alloys to absorb electromagnetic interference while extracting Kovar package with a W-Cu insert to better dissipate heat.
waste heat. An example is shown in Figure 7.69, molded by Components for computer-networking equipment and
Precision Castparts Advanced Forming Technology (Long- Internet servers are MIM targets, and include both electronic
mont, Colorado, U.S.) for AT&T (Murray Hill, New Jersey, and optical devices. Hermetic housings are formed from low-
U.S.). Now the packaging is tailored to a low-thermal-expan- thermal-expansion alloys and the transparent optical compo-
sion coefficient and high thermal conductivity. Some of the nents are fabricated from oxide spinels. Precise dimensions
packages are large, and difficulty in machining the large pack- are necessary because of the required alignment in assembly.
ages leads to MIM. A typical design consists of a small box with a mounting
The use of MIM low-thermal-expansion Fe-Ni and Fe-Ni- flange penetrated by several feed holes. Aluminum by MIM
Co alloys is well established. A recognized alloy with low has been considered in a few situations, but the material
coefficient of thermal expansion (CTE) is Invar, with a nom- proves too soft and is too expensive for most applications.
inal composition of Fe-36Ni. This alloy exhibits anomalous To overcome the thermal-conductivity barrier, heat-pipe
thermal-expansion behavior near room temperature (CTE in designs by MIM have arisen. A material such as copper is
near 1.710-6 per C). Alloying with 5%17% cobalt further molded with intentionally open-pore interior regions, formed
lowers the expansion coefficient. It is used to match the using larger particle sizes and low solids loadings. During sin-
thermal expansion of various semiconductor or glassmetal tering pores are intentionally created on the interior while a
sealing materials, such as the electronic package shown in dense outer shell forms based on selective placement of the
two feedstocks. A fluid in the pores provides enhanced heat
transport, reaching to 2,000 W/(m K), or about tenfold over
dense MIM materials.
With respect to optical components, MIM has several suc-
cesses, including fiber optic connectors out of stainless steel.
An example optical MIM part is shown in Figure 7.71. This
flagstaff nose is used as an external connector for high-
128
Table of Contents
Chapter 7: Applications
SPORTING
First use of MIM in sporting was in the early 1990s with
the production of running spikes for track shoes. MIM track
spikes from titanium alloyed with iron with strength of 887
MPa and 4.8 g/cm3 density were used by sprinter Leroy Burrell
in 1991, leading to a 100 m time of 9.88 seconds. Subse-
quently, sporting-shoe companies have investigated MIM for
cleats and spikes, including for baseball and golf shoes. In par- Figure 7.73. Titanium putter used to demonstrate titanium MIM,
allel, there were efforts to fabricate darts and other weighted and the ability to form golf clubs: (a) bottom view and (b) top view
sport devices by MIM, including the sole plate for golf clubs. (component courtesy of Epson Atmix)
Callaway, Taylor Made, Cobra, Never Compromise, Cleveland
Golf, and others relied on tungsten weights to adjust the center
of gravity.
In the 1990s there were several efforts to form golf clubs
via MIM. This included putters, sand wedges, and drivers. The
materials included stainless steel, titanium, and tungsten. Fig-
ure 7.73 shows a putter fabricated from titanium by Epson
Atmix (Aomori-ken, Japan). Although the potential for using
MIM in golf is exciting, widespread MIM production has been
elusive. One early success was a putter designed by Ping that
reached production levels of 66,000 per year. However, at $6
each MIM was unable to match investment casting costs.
Another goal has been to use MIM to form hollow driver
golf clubs at a target cost of $15 each. Wedges and other flatter
structures are targeted at costs in the $8 range. Several com-
panies have participated in demonstration trials. Technical fea-
sibility has been demonstrated, but economic benchmarks
have been missed. Meanwhile, one operation is working on
golf clubs injection molded from ceramic and infiltrated with
Figure 7.74. Tungsten fishing weights fabricated by MIM with
molten aluminum to form a high-stiffness, low-weight struc- internal channel and identification information (components
ture. This will provide a lower density than titanium, but at a courtesy of Tungsten Heavy Powders and Zhuzhou Luke's Metal
higher expense. Powder Products)
129
Table of Contents
Chapter 7: Applications
Fishing weights shifted to MIM when lead was banned. from target shooting areas, leading to various steelpolymer,
Most are based on tungsten, some are simply bonded tungsten copperpolymer, and tungstenpolymer compositions. Tests
(molded but not debound or sintered), and others are sintered. have been conducted with three variantspowderpolymer
Figure 7.74 is a picture of four tungsten fishing weights made bullets, debound and presintered bullets (not very strong), and
by MIM. Note the hole, indent, and internal lettering included fully sintered bullets. The first choice is the lowest cost. Initial
in each piece. This is a 22 g mass. orders in the area are hundreds of millions per year, and if
Hunting knives, arrow tips, and other sharp objects used accepted by consumers could reach billions of parts per year.
outdoors are formed using MIM. Figure 7.75 shows a novel
hunting knife. A 17-4 PH stainless steel arrow tip is shown in TELECOMMUNICATIONS
Figure 7.76. The latter is fabricated by Smith Metal Products MIM components are widely used in telecommunication
(Lindstrom, Minnesota, U.S.) for Trophy Taker (Plains, Mon- devices for everything from cellphone-vibration weights to
tana, U.S.) to a final 7.6 g/cm3 density. Most of these rely on microelectronic mounts. The vibrator weights are a lower-
stainless steels and the designs initially were copies of tradi- technology product that reached enormous production quan-
tional shapes; more recently the blades have included holes tities at very low prices. An example weight is shown in
to lower mass and reduce cost. In traditional machined blades, Figure 7.77, fabricated from tungstennickelcopper by
the addition of holes increases manufacturing time and cost, Advanced Technology and Materials (Beijing, China). These
but in MIM the holes mean less powder is used and this weights are selling for less than $0.05 each. Other telecom-
reduces cost.
A related development is the production of practice bullets
using MIM. The efforts are driven by a desire to eliminate lead
130
Table of Contents
Chapter 7: Applications
munication applications for MIM include mounting hardware, are formed using MIM, such as the clip shown in Figure 7.79.
hinges and latches, antenna mounts, housings, and compo- It is fabricated out of 17-4 PH stainless steel, with gold plating
nents associated with cellphone transmission stations. The after sintering, by Advanced Materials Technologies (Singa-
latter are tungstencopper compositions for high thermal pore) for Agilent Technologies (Singapore). The stainless steel
conductivity. provides rigidity and heat dissipation.
Electronics hardware for telecommunication systems is Most of the opportunities depend on expansion of cell-
another application for MIM, where the combination of heat phone networks. The base stations are heavy users of high-
dissipation and electromagnetic shielding favor metals. Figure thermal-conductivity materials such as tungstencopper. The
7.78 is a picture of a 32 g (32 mm by 24 mm by 10 mm or future worldwide build of telecommunication systems will
about 1.3 in. by 1 in. by 0.4 in.) housing formed by MIM. The continue to drive this application area into higher production
thinnest section is below 1 mm wall and the thread is added quantities with lower unit costs. Heat plates, such as shown
as a secondary operation. This part was fabricated by Net- in Figure 7.80 after gold plating, are an example application.
Shape TechnologiesMIM (Solon, Ohio, U.S.) in 4-cavity Variants included high-thermal-conductivity copper, porous
tooling at production rates exceeding 100,000 per year. copper with internal heat pipes, tungstencopper with lower
Fiber optic connectors rely on MIM stainless steel, but thermal expansion, and Invar with electroplated gold coatings.
ceramic versions are more popular. Various alignment devices
SUMMARY
For those people long focused on powder metallurgy, one
of the common remarks is on how MIM differs from the stan-
dard presssinter powder metallurgy process. The differences
are very evident with respect to almost all aspects:
MIM powders are small and sphericaltraditional powder
metallurgy relies on much larger, irregular powders
MIM materials are dominated by stainless steelstradi-
tional powder metallurgy largely relies on iron and steel
compositions
MIM relies on a high polymer content (5 to 10 w/o)
traditional powder metallurgy relies on just 1/10 the
polymer content
sintered MIM densities are comparable to cast metals
traditional powder metallurgy usually is only 85% of
Figure 7.79. An optical transceiver housing molded from 17-4 PH cast
stainless steel. It is used in ultrahigh speed transceiver for
properties in MIM rival handbook valuestraditional
networking and telecommunication systems (photograph courtesy
of Advanced Materials Technologies) powder metallurgy properties often are 50% of hand-
book values
MIM applications are almost everywheretraditional
powder metallurgy is focused on automotive applica-
tions
The dispersion of application fields is both good and bad
news. The good news is that MIM tends to ride the offsetting
economic ups and downs of various markets. So when one
area stalls, MIM has footholds in other areas that grow. During
the terrible economic times of 200809, fear drove firearm
sales and, unlike other fields, MIM experienced a significant
growth. The bad news is simply the same situation. It requires
much effort to stay abreast of several markets. When all of the
eggs are not in one basket, then considerable work is required
to keep up with such a range of customers, markets, and
design swings.
As a final aspect in MIM applications it is noted that sev-
Figure 7.80. Large telecommunication mounting plates formed by eral artistic items are being fabricated using the technology.
MIM with gold coating (photography courtesy of Advanced Figures 7.81 to 7.83 show that not everything is precisely
Materials Technologies) engineered for function, but some of the MIM products are
131
Table of Contents
Chapter 7: Applications
RESOURCES
N. De Cristofaro and A. De Nicolo, A Case Study of MIM
Components for Automotive Applications, Second European
Symposium on Powder Injection Molding, 2000, European
Powder Metallurgy Association, Shrewsbury, UK, pp. 215219.
R. Dwivedi and J. Fauerbach, A New Method for Production
of Micro Porous Metals, P/M Science and Technology Briefs,
2001, vol. 3, no. 2, pp. 2225.
R.M. German, Medical and Dental Applications for Micro-
miniature Powder Injection MouldingA Roadmap for
Growth, Powder Injection Moulding International, 2009,
vol. 3, no. 2, pp. 2129.
J.L. Johnson, Mass Production of Medical Devices by Metal
Injection Molding, Medical Device and Diagnostic Industry,
2002, November, pp. 4853.
Figure 7.81. Ceramic pots with decorative MIM top handles
(photograph courtesy of Taiwan Powder Technologies) M. Kimura, Powder Metallurgical Manufacture of a Rocker
Arm for Variable Valve Timing Mechanism by MIM Process,
Ninth Case Studies on New Product Development, Japan Pow-
der Metallurgy Association, Kyoto, Japan, November 2000,
pp. 3544.
M. Kneuwer, K.H. Wichmann and H. Minhardt, Processing
of Tungsten and Molybdenum Copper Housing for Microelec-
tronic Packaging via Metal Injection Moulding, Second
European Symposium on Powder Injection Molding, 2000,
European Powder Metallurgy Association, Shrewsbury, UK,
pp. 221228.
Figure 7.82. Limited-production chess set created out of bronze T. Moritz and R. Lenk, Ceramic Injection Moulding: A
using a MIM approach (photograph courtesy of Mark Pilato) Review of Developments in Production Technology, Materials
and Applications, Powder Injection Moulding International,
2009, vol. 32, no. 3, pp. 2334.
B.N. Nobrega, W. Ristow and R. Machado, MIM Processing
and Plasma Sintering of Nickel Base Superalloys for Aero-
space and Automotive Applications, Powder Metallurgy,
2008, vol. 51, pp. 107110.
J.A. Sago, H. Chen, M.W. Boradley and J.K. Eckert, Devel-
opment of Cobalt-Chromium MIM Alloys Suitable for Med-
ical Device Applications, Advances in Powder Metallurgy
and Particulate Materials2009, Metal Powder Industries
Federation, Princeton, NJ, part 4, pp. 111117.
M. Sonderegger, B. Unternahrer and A. Oberli, Application
of the MIM Technology for SWATCH Irony Watch Cases and
Watch Components, Second European Symposium on Pow-
der Injection Molding, 2000, European Powder Metallurgy
Figure 7.83. Polished stainless steel chopstick rests made by MIM Association, Shrewsbury, UK, pp. 235242.
(photograph courtesy of Taiwan Powder Technologies)
132
Table of Contents
Chapter 7: Applications
133
Table of Contents
Table of Contents
Chapter 8
MARKET CONSIDERATIONS
INDUSTRY STRUCTURE
Overview
An outline of the MIM industry structure and interactions
is diagramed in Figure 8.1. The central actors in the industry
generally focus on one of the following sectors:
ingredient supplierspolymers, powders, and ingredi-
ents for either self-mixing or intermediate feedstock pro-
duction; globally there are approximately 40 firms that
provide most of the MIM powders, although about 400
firms supply metal powders of various chemistries, par-
ticle sizes, particle shapes, and purities; for example, in
titanium about 4 companies out of 40 suppliers make the
powders needed for MIM
feedstock productionpurchase raw ingredients and for-
mulate mixtures for sales to molding firms; globally
there are 12 feedstock suppliers
molding firmsboth custom and captive molders that
total to more than 260 MIM operations; about one-third
are captive and make parts for themselves, but many of
the captive firms also perform custom fabrication, so
83% of all parts production is categorized as custom Figure 8.1. Diagram of the industry relations in the MIM field.
manufacturing The central activity is at the parts fabricator, which could be either
thermal processingfirms that own sintering furnaces a custom or captive operation
135
Table of Contents
Chapter 8: Market Considerations
and debinding equipment that provide toll services; cur- stainless steels
rently 5 firms are active in this area and most are associ- special ammunition that included birdshot, armor pierc-
ated with furnace fabricators ing and frangible bullets
designerslargely systems designers at large multina- turbocharger rotors out of nickel alloys for small sports
tional firms that intersect with the MIM industry, but a cars
few independent designers are available to handle ad hoc Since each of these MIM operations had a single field of
projects focus, little was done to grow that portion of the industry.
equipment suppliersfirms that design and fabricate However, in more recent years growth in MIM has come with
custom furnaces, molders, mixers, debinding systems, the shift to custom molding which services a wide variety of
robotic systems, and other capital devices such as testing applications. The custom molding firms have joined together
devices; the majority of molding machine sales are from in efforts to advance the industry, via collaborative marketing
6 firms, furnace sales are from 8 firms, mixer sales are efforts, promotion of material standards, publicity via annual
from 4 firms, so about 20 firms constitute the key equip- awards, and sharing of business data. Although declining,
ment suppliers captive molding still remains an important part of the MIM
consumablessuppliers of process atmospheres, chem- industry.
icals, molds, heating elements, and sintering substrates
adjunctsincluding researchers, consultants, design Historical Shifts
advisors, conference organizers, trade association per- As an introduction to market progress in metal powder
sonnel, magazine editors, and patent attorneys injection molding, Table 8.1 compares the global characteris-
Component production is the central activity. It is split tics as captured in 2000 and 2010. This table includes data
between internal and external products, referred to as captive from start-up ventures as well as long-established manufac-
and custom molders. Likewise it is supported by two parallel turing firms. For many years the MIM field sustained com-
supply routes, depending on the decision to self-mix or to pur- pound annual sales growth of 22% per year with a 34% per
chase premixed feedstock. An example captive molder would year increase in the number of operations. In recent years the
be a firearms company that uses MIM to fabricate some of the growth rates have become more modest and have stabilized
safety, trigger, or sight components. On the other hand, custom near 8% per year in North America, but continue at a 30% per
molders also can make these same components, but just as year pace in Asia.
well can be involved in several application areas as deter- Although a large number of companies are active in MIM,
mined by their customer base. at any one time some are simply in an evaluation mode. These
As outsourcing increases for multinational firms, custom are designated as level 0 firms. Characteristically they consist
fabrication grows. Accordingly, MIM fabrication facilities of a small team (two or three people) and one or two molding
owned by large firms such as IBM, AMP, and GTE, which machines, usually purchase feedstock, and might even rely on
were early adopters, shifted to purchasing components from toll sintering. After initial exploration, many such efforts are
captive molders focused a variety of application areas. Some terminated or production is transferred to an outside vendor.
of the early captive applications included the following
examples: TABLE 8.1. OVERALL PERSPECTIVE ON THE
orthodontic brackets out of stainless steel or cobalt WORLDWIDE MIM INDUSTRY
chromium alloys
feature 2000 2010
business-machine components for postage meters and
typewriters number of companies 400 265
watch components including weights, bezels, cases, total annual sales $500 million $1,000 million
total number of employees 3,500 6,000
bands, and clasps mean MIM employment 9 22
camera components that included switches and buttons per operation
firearm steel parts such as trigger guards, sights, gun bod- annual sales divided by $143,000 $167,000
ies, and safeties total employment
carbide and tool-steel cutting tools such as wood router number of installed molding 560 1,000
machines
bits, end mills, and metal-cutting inserts number of installed 430 800
electronic packages for electronic systems from glass sintering furnaces
metal sealing alloys number of installed mixers 236 322
personal care items such as hair trimmers from tool steel typical annual sales per $892,000 $1,000,000
medical hand tools for special surgical operations molding machine
MIM production for internal 31% 17%
rocket engines from specialty materials such as niobium use
automotive airbag-actuator components from hardenable
136
Table of Contents
Chapter 8: Market Considerations
This is evident by the decrease in number of MIM firms from molds, typically based on a few customers or even
2000 to 2010 while the overall field grew. captive products; often production quantities are low,
so equipment utilization is in the 10% to 30% range
Production Levels (8 hours per day for 5 days per week, or 40 out of
Most of the firms advance through stages. Initially the 168 hours); feedstock use in typically less than 5 tons
focus is on evaluation of the MIM technology to assess for per year and might be purchased to avoid mixing
opportunities. Some of the evaluation efforts advance to pilot equipment purchase
production, leading to a small facility designed to satisfy a level 3full-scale production mode, represents a company
few special applications or customers. Thus, in metal powder with a history of constant parts production, charac-
injection molding there is a natural progression starting with terized by production reaching 24 hours per day up
the evaluation efforts, moving to pilot production, the next to 7 days per week; facility utilization often in the
step being small-scale production (characterized by single- 80% or higher range; other characteristics show from
shift operation), and eventually graduating to full-scale pro- 80 to several hundred employees and sales exceeding
duction characterized by round-the-clock production. In short, ten million dollars per year; this group dominates
these are segmented as follows: sales and profits
level 0evaluation mode, involving a small technical and Figure 8.2 illustrates the typical maturation cyclic behavior
business effort designed to learn more about MIM; among these four levels by plotting expected profits vs. sales
usually includes some in-house trials, not necessarily (in thousands of dollars) from start-up to full production. Most
designed for production, but for learning the process, firms start in level 0 with a fixed annual expense of $500,000
testing new concepts and designs, and often to pro- or more, and rise to break-even with about $2 million in cap-
vide marketing and engineering with a sound basis ital investment and $500,000 in annual sales. At that point
for dialog with custom fabricators; there are no they transition to a level 1 pilot effort. Unfortunately,
measurable sales and no profits (effectively a net loss increased costs hinder profitability even during sales growth.
or R&D charge at this stage) Every time staff or equipment is added to increase production,
level 1pilot mode, represents a company with some capa- the costs incurred offset the sales gains.
bilities, where they are examining possible products As business expands, eventually profitability emerges.
and developing new customers; usually equipment However, there can be profit dips even during growth as extra
has been purchased and molds are typically single equipment and extra production capacity are added (cash flow
cavity; often characterized by one or two molding can be an issue due to the time lag between equipment pur-
machines and one or two furnaces; profitability is chase and productive operation of the equipment due to long-
marginal as the operation learns key steps in stabi- duration installation and training cycles). Finally, sustained
lizing its process and qualifying early products profitability occurs in level 3.
level 2small-scale production mode, represents a company As should be evident, MIM operations range from large,
that has gone into production, usually with a single automated factories with up to 60 molders, to small shops with
shift, maybe graduating to some multiple-cavity little more than a Web page and a single molder. In some cases
the start-ups are plastic molders that purchase feedstock and
outsource debinding and sintering. They shift between plastic
molding and metal molding as required by sales. As a conse-
quence, across the MIM industry a large disparity exists in
productivity, capacity, and capabilities.
Figure 8.3 plots the relative population of operations at
each level while Table 8.2 gives a statistical sense of the oper-
ating characteristics at each level. Here Prados Principle is
evident, where 20% of the MIM operations dominate the
industry, producing 80% of the parts. Further, the top segment
employs about two-thirds of the workforce, while owning a
vast majority of the production capacity.
137
Table of Contents
Chapter 8: Market Considerations
TABLE 8.2. COMPARISON OF MIM OPERATIONS AT VARIOUS MATURITY LEVELS (2010 statistics)
attribute level 0 level 1 level 2 level 3
characteristic evaluation pilot small-scale full-scale
% of operations 14 27 29 30
% of sales 0 4 7 89
average sales per employee per year nil $60,000 $125,000 $250,000
% of total powder use nil 10 7 83
% of total employment 7 12 12 69
138
Table of Contents
Chapter 8: Market Considerations
where large orders exceed in-house capacity, or as part of sales, jumping by a factor of 3 along the lower axis. As of
equipment evaluation prior to purchase. Less than a dozen 2010, there is no MIM firm at $100 million in annual sales.
firms support toll processing, but total sales in this area are The largest ceramic injection molding firm reached $91 mil-
relatively small, probably well under 1% of MIM sales. lion in sales one year, but subsequently retrenched. On the
Today more than 360 firms practice PIM, and about 80% other hand, only a few full-time operations do less than
of those practice MIM. Most of the firms have multiple appli- $1 million in annual sales. The majority of the industry
cation areas. Of these firms, many are beyond pilot produc- generates sales per year in the $1$30 million range. These
tion, but only about 100 use their facility 24 hours per day. operations tend to focus on a few materials and typically have
Figure 8.5 estimates of the sales distribution for MIM opera- 40 customers.
tions. This bar chart has a logarithmic scaling to the annual With respect to market segments, Figure 8.6 provides a par-
titioning of 2010 sales based on major application areas. In the
1990s, major growth occurred in orthodontic brackets, com-
puter disk drive components, consumer products, firearms,
telecommunication devices, and watch components. By the
late 1990s, consumer items reached very impressive produc-
tion rates, but often the products were short lived, remaining
in production for one to two years and then going out of style.
Today, automotive, industrial, and other longer-life products
have taken up MIM and provide more production stability.
In some application areas MIM is in decline. A good exam-
ple is in computer hard disk drives. On the other hand, a recent
surge has come from automotive and medical applications,
while the firearm and general mechanical devices have
remained strong. Some fields, such as firearms and orthodon-
tic brackets, are nearing maturation while aerospace applica-
tions are still at the early adoption stage. New applications are
emerging in hand tools, plumbing hardware, biomedical sys-
tems, computer heat sinks, medical diagnostic components,
and microelectronics packaging. As a result there is a con-
Figure 8.5. Distribution in MIM company size based on annual stantly shifting distribution of applications. This is a barrier
sales volume; note, the x-axis is not linear but jumps by approxi- since the small firms have difficulty serving so many markets.
mately a factor of 3x for each category. The largest firms are over
$30 million but under $100 million in annual sales. The mode or COMPETITIVE TECHNOLOGIES
most common company size is $1 to $3 million in annual sales
Competitive technologies are the best benchmark for meas-
uring the value of MIM. Process yield and material utilization
are immediate advantages when compared with subtractive or
machining processes. In replication processes, such as press
sinter powder metallurgy, there tends to be less waste, result-
ing in a lower cost. In summary, the successes for MIM come
when key attributes are realized vs. the competitive technolo-
gies, such as:
overcoming the property limitations inherent to plastics
expanding shape capabilities beyond those possible using
stamping, forging, and fine blanking
exceeding the property and shape-range limitations
inherent in presssinter powder metallurgy
providing a lower cost when compared with machining
providing productivity levels not attainable with other
full-density approaches such as isostatic pressing and slip
Figure 8.6. Major application areas for MIM based on annual casting
sales for 2009. Historically, fields such as watch production
dominated MIM in the early years, while medical and automotive avoiding the defects, surface-finish limitations, and
are now large application areas. Industrial and consumer tolerance limits associated with casting
applications tend to be low profile, but constant, areas Metallic powders are more expensive when compared with
139
Table of Contents
Chapter 8: Market Considerations
ingot or bar stock, as might be used for machined components. material is expensive or difficult to process by alternative
Consequently, the raw material cost difference must be offset techniques. Titanium fits this niche because it is very difficult
by savings through the elimination of machining. Figure 8.7 to machine. As an illustration of the underlying cost curves,
is a schematic outlining the interplay between component Figure 8.8 plots the unit cost vs. shape complexity for a batch
complexity and production volume for several forming size of 1 million units. It is the lower slope of the MIM curve
approaches. This roughly identifies where each forming that makes it attractive. These partitions between fabrication
approach is most successful. For a given complexity, this plot approaches are for illustrative purposes, since many factors
suggests the greatest economic merit for MIM comes with impact actual demarcations.
high production quantities. Metal powder injection molding suffers when the dimen-
Advantages for MIM occur at lower quantities when the sional precision is higher than a coefficient of variation of
0.2%. However, if only a few dimensions require final
machining, then MIM can be cost effective in forming a near-
net-shape component. The mold costs about the same as in
plastic tooling, but both are expensive, so the production
quantity needs to be sufficiently large to justify mold cost.
Figure 8.9 illustrates the relative tool cost for MIM and several
alternative net-shape technologies. At the lower production
levels, metal powder injection molding has a disadvantage.
For mass production situations the low piece price is attrac-
tive. Consequently, some firms have innovated means to cir-
cumvent cost barriers at low production quantities, using
lower molding pressures (lower viscosity binders) and softer
tooling. These solutions give production quantities under
10,000 per year. As a benchmark, common high-pressure
molding tools easily produce 100,000 or more components
per year.
FINANCIAL PERFORMANCE
About every year a market report is issued on the MIM
industry, documenting the key actors, applications, markets,
Figure 8.7. Conceptual map of production quantity on the y-axis and financial performance. This short introduction is not
and shape complexity on the x-axis, illustrating the relative intended to displace those reports, but should provide a sum-
positioning of MIM with respect to competitive technologies mary on the key indices.
In reviewing the financial performance, one complicating
factor is in determining where to assign value. An example of
140
Table of Contents
Chapter 8: Market Considerations
141
Table of Contents
Chapter 8: Market Considerations
alloys and sterling silver, as well as tungsten alloys and tool of copper was demonstrated early, the cost of a high-purity
steels. Technical feasibility is well demonstrated for all of powder delayed applications. Tool steels are similar, in that
these materials. Thus, what is probably the largest barrier early demonstrations were made for applications in woodcut-
would be in the economics. For example, good quality tita- ting and metal forming, but powder cost delayed the
nium powder for MIM is between $120 and $150 per kg for applications.
spherical powder, while angular hydridedehydride powder A few electronic alloys are in use via MIM, mainly Kovar
is approximately $80 per kg. Since wrought titanium products for glassmetal sealing packages, and Invar for low-thermal-
cost $100 per kg, MIM is mostly viable for small and com- expansion applications.
plicated components, but these are sensitive to contamination Composites offer unique growth opportunities, especially
and properties. Implants are in production after several years for applications that do not require ductility. The addition of
of seeking approval. ceramic particles to a MIM stainless steel feedstock is one
Copper is a relatively new area for MIM. Although MIM means to form a hard and wear-resistant corrosion-resistant
material, such as desired for shears and hand tools.
Heat dissipation and microwave absorption in microelec-
tronic systems are growing concerns. New microcomputers
contain tens of millions of transistors on a single chip, with
connectors of submicrometer widths. The high computing
power and small chip size generate heat that must be removed
for maximum computing performance. Thermal fatigue aris-
ing from onoff heating (expansion) and cooling (contraction)
is a problem with high-thermal-expansion materials used in
early designs. Accordingly, MIM has emerged as a fabrication
route for the creation of unique designs and materials. For
example, Figure 8.12 illustrates one concept where the outer
package is formed from Kovar for glassmetal sealing at the
feed-through holes, but the embedded heat sink in the middle
of the device is molded from W-Cu. In the optimal applica-
tion, computer-chip mounts have been fabricated with a ther-
mal conductivity over 220 W/(m C) and a thermal-expansion
coefficient of 6 ppm/C. Only a few useful combinations of
Figure 8.11. Material use based on component sales, showing how
stainless steels dominate the value of MIM products. Stainless thermal expansion and thermal conductivity exist. From a cost
steels also dominate the tonnage data, followed by steels, tungsten perspective, W-20Cu is probably the most suitable material,
alloys, nickel alloys, and several other materials but has a high density. Surprisingly, copper solutions have
been accepted by many firms for the control-system heat dis-
sipaters in hybrid vehicles, including novel heat pipes and
vapor chambers.
Applications
The metal powder injection molding field enjoys applica-
tions in a broad array of fields. But this is also a difficulty,
since MIM as a technology is relatively small and new, so it
is very taxing for the industry to support and market to a wide
variety of uses.
Generally, markets for MIM can be divided into the fol-
lowing main categories:
automotive: engine timing components, interior switches,
sensors, fuel injectors, turbocharger vanes, brake calipers,
heating-system controls, steering components
consumer: watchcases, watch bands, camera switches,
Figure 8.12. Example of a novel MIM product, shown here as a eyeglass frames, eyeglass hinges, luggage buckles, cell-
microelectronic package that would be formed from Kovar to phone hinges, vibration weights, toothbrush gears
enable glass sealing at the wire holes, while a high-thermal-
conductivity W-Cu is inserted under the semiconductor to dissipate dental: endodontic tips, orthodontic brackets, hand tools,
heat cleaning tools, implants
142
Table of Contents
Chapter 8: Market Considerations
electronics: disk drive parts, computer latches, electronic tions, mostly from stainless steels. Fortunately, the tolerances
packaging, wire-bonding tools, sensors, controls for medical devices are not as tight as in automotive applica-
firearms: trigger guards, sights, sight mounts, triggers, tions, and, similarly to firearms, the product life is several
handgun bodies years.
hardware: cutting tools, hand tools, fasteners, hinges, In recent years there has been an intersection between con-
locks sumer electronics and telecommunications, where MIM is used
industrial: valves, fittings, magnetic sensors, connectors, for hinges, swivels, weights, disk drive parts, logos, and similar
solenoids, cutters components. They must be corrosion resistant, so they are
medical: surgical tools, implants, biopsy tools, retractors, chromium or nickel-plated steel, or formed from stainless steel,
scalpels the geometries are very complex, the tolerances are very tight,
military: gun mounts, bullets, projectiles, automatic and production volumes are highly variable. Industrial appli-
firearms, fuses, timing devices cations, such as in valves, fittings, and connectors, were first
other: aerospace, business machines, jewelry, chemical demonstrated in Japan and now are a routine aspect of MIM.
devices, sporting goods Most of these are ferrous materials. Hand tools emerged over
Figure 8.13 is the partition by firm specialization, based on the past five years, including many plumbing tools, universal
these areas, showing about half the firms focus on industrial, pliers, and electrical-cable-installation devices. These rely on
medical, or electronic products. Note that in parts of the world low-alloy steels, heat treated and coated with a black oxide.
the firms were set up to produce military products, typically Table 8.4 provides yet another view of the production
firearms and ammunition. activity in 2009, in this case based on material consumption
Based on sales, the three largest segments are products for by application area and geographic region. In recent years the
industrial, automotive, and consumer applications. These are major growth has been in automotive, medical, mechanical,
followed by medical, dental, electronic, firearms, computer, telecommunication devices, and consumer products. For
hand tools, sporting, watches, and so on. These partitions are example, between 2003 and 2006, the production volume
not uniform by region, since Europe is more active in watch changes in these fields were as follows:
and automotive components, North America leads in medical, automotive 33%
dental, and firearm components, while Asia is larger in com- medical 13%
puter, cellphone, and communication devices. mechanical and industrial components 23%
Firearms remain one of the stable applications for MIM. information systems and cellphones 25%
These are small, complex shapes (trigger guards, sights, safety
switches, and triggers) formed from steels, heat treated and
TABLE 8.4. PERCENT POWDER USE BY MIM
surface oxidized to provide black oxide coatings. Typical pro- APPLICATION AREA
duction rates are up to 100,000 units per design and products
North
often run for 10 years. Orthodontic brackets are one of the America Europe Asia
longest-running applications for MIM, with stainless steels
automotive 26 60 13
and nitinol (NiTi) and a few nickel-free (cobaltchromium)
mechanical 39 3 37
alloys dominating. Medical applications have grown over the (industrial & hardware)
past 15 years, with significant gains in quantities and applica- medical 23 23 2
consumer 4 4 44
other 8 10 4
143
Table of Contents
Chapter 8: Market Considerations
144
Table of Contents
Chapter 8: Market Considerations
for acceptance. Likewise, biomedical implants such as knees, Considerable research is going into the production of distended
hips, shoulders, and other body parts are formed using powder metals with surface functional ceramic coatings for implants.
injection molding, but most in use are for short-duration The pores are generated by mixing an insoluble plastic particle
implantation. However, significant effort is taking place to into the feedstock. The amount of added plastic determines the
expand the implants to heart valves and other longer-term final porosity and the size of the plastic sets the pore size.
devices. At the same time, new standards are emerging from
bodies such as ASTM for these MIM materials from titanium. New Concepts in MIM
Microminiature components are a new target for metal Future opportunities are emerging from ongoing R&D
powder injection molding. It follows from the base established efforts and although several might not progress to commer-
by the production of orthodontic brackets. Some of this is cialization, it is still curious to see how many new inventions
showing up in components for cellphones, cameras, eye- are arising based on MIM, including the following:
glasses, jewelry, and watches. Other uses are in digital stents that are absorbed into the human body over time,
systems, biomedical sensors, genomic testing, implanted removing the tendency of current stents (treatments to
transducers, telecommunication devices, optoelectronic stabilize arteries, for example) to form blood clots; read-
devices, dental implants, notebook computer latches, and sorbable stents are formed using alloys that dissolve into
replacement bones for the human ear. blood over time and already a magnesium MIM alloy has
One novel use for powder injection molding is in the for- been patented for this application
mation of small high-feature units, such as arrayed posts of ultra-high-thermal-conductivity composites (for exam-
slots in ceramics. The metalized and assembled units form ple, copperdiamond) for heat sinksdemonstrations
ultracapacitors used to store electrical energy, useful in power with 580 W/(m C) thermal conductivity have been made
tools, automobiles, buses, military vehicles, computers, and by MIM for use in supercomputers, high-end servers,
cellphones. As one firm operating in this area claims, future phased-array radar, military electronics, hybrid-vehicle
automotive power will derive from a suitcase-size unit that control systems, gaming computers, and other applica-
will power the vehicle for days at a time. Each suitcase is com- tions involving high-performance computing
posed of thousands of stacked injection molded arrays with vapor-chamber designs, typically from copper, where a
microminiature features. porous internal chamber act as a heat pipe to evaporate a
Another large application for MIM is intentionally porous working fluid for rapid heat transport from electronics,
materials. Beyond filters, several applications arise in biomed- with early applications in hybrid-vehicle control systems,
ical systems. For example, porous titanium scaffolds for tissue gaming computers, and servers
ingrowth are under testing. The elastic modulus of titanium is LED heat sinks where copper arrays are used to mount
much higher than that of human bone, but if the titanium is the semiconductor, with reports of 100 g arrays with
fabricated with large pores and a final density near 40%, then costs as low as $0.75 per mount
it is nearly a perfect match in strength and elasticity of bone. microminiature MIM for medical minimally invasive
surgical tools, involving very small components for end
manipulators, such as cutters, grasps, and drug delivery;
most are from stainless steel and example components
are being sold in the range of $2 to $15 each
electronics and related devices also require microminia-
ture MIM components in cellphones, computers, hand-
held devices
implants for a broad array of biomedical applications
tooth posts, ligament alignment, hearing-canal recon-
struction, drug delivery, and heart valves
microarray devices with hundreds to thousands of pins,
posts, or holes for disposable lab-on-a-chip devices used
in blood testing, disease assessment, analysis of DNA to
predict disease, and protein tests; the biochip market is
targeted to reach $3.8 billion in sales by 2013 and con-
siderable research is taking place on how MIM can sup-
port this effort
Figure 8.15. Relative growth rates for PIM and MIM production titanium biocompatible structures, such as for tissue
distributed over several of the commonly discussed application affixation, implants, surgical tools, tool implants, and
areas even sporting devices; several firms have focused on this
145
Table of Contents
Chapter 8: Market Considerations
possibility, and many are relying on hydroxyapatite planned expansion, but 10% are planning to exit the field. The
titanium composites industry will probably converge on about 200 actors, 10 of
hardware from tool steel, such as threading devices for which will be very large. Projections indicate the sales growth
cast iron plumbing or water pipes, hand tools, valves and rate will sustain 8% or more annual increase from year 2010
fittings, handles, forming tools, drills, dies up to 2017. At that point the estimated annual sales will be
hermetic packages for microelectronics using Kovar to near $2 billion.
enable glass-to-metal sealing; some small example MIM Growth will continue in several application areas. Many
parts when sealed with leads are selling for $30 each pilot projects are showing an expanded range of uses. Much
aerospace applications in smaller superalloy bodies such opportunity still lies in converting investment cast stainless
as IN 625, 718, 723, or Hastelloy X, where high detail, steels to MIM or die cast zinc to higher-strength MIM steel.
good surface finish, and shape complexity are financially The addition of titanium and tool steels to the list of MIM
attractive materials will accelerate future growth, since both are difficult
to process by other net-shape routes.
Industry Changes A growth barrier is the lack of industry visibility and poor
The MIM industry has established credibility through the knowledge of the process by the design community. This need
mass production of several small, complex components. The for improved customer education is not being addressed
period of intense new-company formation has ended and now directly as yet.
growth comes from the existing firms making significant To succeed against competitive technologies, MIM must
inroads with new customers and markets. Sales growth is meet the engineering requirements at the lowest cost (engi-
coming with broad acceptance of metal injection molding for neering, tooling, capital equipment, and piece cost). As indi-
automotive, medical, consumer, and electronic applications. cated in Figure 8.16, the metal powder injection molding
These are smaller devices with demanding properties. technology has high marks with respect to the variety of avail-
It is instructive to review the market forecast made in 1990 able materials, component shape complexity, surface finish,
that identified the industrys growth barriers as follows: and final product strength, with a range of sizes (from micro-
low powder consumption (with little standardization) miniature to a few hundred mm), and excellent productivity.
long debinding times On the other hand, it suffers because of the high costs of tool-
long lead times to production (especially for tooling) ing and raw materials.
poor vendor knowledge of the industry requirements The successes for MIM lie in the combined attributes that
lack of customized equipment include:
high powder cost overcoming the property limitations inherent to plastics
lack of industry standards expanding shape capabilities beyond stamping, forging,
lack of personnel who understand the process (especially and fine blanking
at the customers) exceeding the property and shape-range limitations
poor access to capital inherent in presssinter powder metallurgy
Today, those barriers are gone, but the lack of industry
coordination (marketing, materials, processes, properties, and
standards) remains as a difficulty. The formation of an inter-
national MIM alliance is one positive step toward correcting
these problems. Most important, in the past few years Powder
Injection Moulding International has become the forum for
sharing news, statistics, research results, and product
announcements in the community.
Processing-cost reductions via automation continue to be
a major need in order to expand into automotive applications.
Powder injection molding is a route to lower-cost, higher-
quality components, especially when compared with machin-
ing. However, MIM is only now becoming sufficiently
sophisticated to take on the large projects. Further, the toler-
ances need to be improved to the 25 m range and even Figure 8.16. A graphic comparison of relative technology
tighter for the high-payout areas. strengths to give a contrast between powder injection molding and
several other net-shape metal-forming technologies. The
comparison is based on technical attributes of material strength,
FUTURE TRENDS component-size range, productivity, breadth of materials,
In recent industry surveys, about 70% of the firms report component complexity, and surface finish
146
Table of Contents
Chapter 8: Market Considerations
providing a lower cost when compared with machining M. Sonderegger, B. Unternahrer and A. Oberli, Application
providing productivity levels not attainable with isostatic of the MIM Technology for SWATCH Irony Watch Cases and
pressing and slip casting Watch Components, Second European Symposium on Pow-
avoiding the defects, surface-finish limits, and tolerance der Injection Molding, 2000, European Powder Metallurgy
limits associated with casting Association, Shrewsbury, UK, pp. 235242.
In short, the best places to apply MIM are where plastic
D. Whittaker, Powder Injection Moulding Looks to Automo-
molding would be capable of forming the shape, but plastics
tive Applications for Growth and Stability, Powder Injection
lack the mechanical, thermal, or other properties.
Moulding International, 2007, vol. 1, no 2, pp. 1422.
RESOURCES International Powder Metallurgy Directory 20102011, 2010,
R.G. Cornwall and R.M. German, An Analysis of the Inovar Communications, Shrewsbury, UK.
Powder Injection Molding Industry Global Market, Advances
R. Zauner, A. Altenburger and R. Nagel, New Application
in Powder Metallurgy and Particulate Materials2001,
for MIM of Tungsten: Divertor Modules for ITER, Powder
Metal Powder Industries Federation, Princeton, NJ, part 4, pp.
Injection Moulding International, 2007, vol. 1, pp. 5052.
1116.
N. De Cristofaro and A. De Nicolo, A Case Study of MIM
Components for Automotive Applications, Second European
Symposium on Powder Injection Molding, 2000, European
Powder Metallurgy Association, Shrewsbury, UK, pp. 215
219.
R.M. German, Medical and Dental Applications for Micro-
miniature Powder Injection MouldingA Roadmap for
Growth, Powder Injection Moulding International, 2009,
vol. 3, no. 2, pp. 2129.
R.M. German, Titanium Powder Injection Moulding:
A Review of the Current Status of Materials, Processing,
Properties, and Applications, Powder Injection Moulding
International, 2009, vol. 3, no. 4, pp. 2137.
D. Gonia, MIM Parts for Automotive Applications Produc-
tion and Economic Aspects, Journal of the Japan Society of
Powder and Powder Metallurgy, 1999, vol. 46, pp. 849852.
J.L. Johnson, Mass Production of Medical Devices by Metal
Injection Molding, Medical Device and Diagnostic Industry,
2002, November, pp. 4853.
M. Kimura, Powder Metallurgical Manufacture of a Rocker
Arm for Variable Valve Timing Mechanism by MIM Process,
Ninth Case Studies on New Product Development, Japan Pow-
der Metallurgy Association, Kyoto, Japan, November 2000,
pp. 3544.
T. Moritz, and R. Lenk, Ceramic Injection Moulding: A
Review of Developments in Production Technology, Materials
and Applications, Powder Injection Moulding International,
2009, vol. 32, no. 3, pp. 2334.
J.A. Sago, H. Chen, M.W. Boradley and J.K. Eckert, Devel-
opment of Cobalt-Chromium MIM Alloys Suitable for Med-
ical Device Applications, Advances in Powder Metallurgy
and Particulate Materials2009, Metal Powder Industries
Federation, Princeton, NJ, part 4, pp. 111117.
147
Table of Contents
Table of Contents
Chapter 9
SUPPLEMENTAL INFORMATION
Outline Overview
Cost-Calculation Approach
Tool Costs
Calculation Approach
Optimization of Multiple-Cavity Tooling
Other Tool-Cost Factors
Example Tool-Cost Estimation
Summary of Important Tool-Cost Factors
Feedstock Costs
Cost of Raw Materials
Operation Parameters
Mixing Costs
Feedstock for Golf Clubs
Feedstock-Cost Summary
Batch-Size Effects
Component Costing
Operational Characteristics
Molding Costs
Debinding Costs
Sintering Costs
Finishing Costs
Tight-Tolerance Effects
Per-Piece Costing
Additional Costs
Golf Club Example
How Much Will It Cost?
Cost Sensitivity
Typical Component Analysis
Component-Mass Effect
Golf Club Example
Bottom Line on Costing
Definitions
Nomenclature
Units and Conversion Factors
Helpful Relations
149
Table of Contents
Chapter 9: Supplemental Information
Outline Prefixes
Fundamental Units
continued Derived Units
Important Constants
Conversion Factors
Misconceptions
Standards
Patent Sources
Conferences and Publications
Trade and Professional Organizations
Resources
OVERVIEW uncertain about how best to use the technology. This slows
A cost model for MIM is generated where a larger number growth, even when MIM might have a significant advantage.
of parameters are required, but via this model the important Accordingly, this chapter reviews the engineering economics
sensitivities become evident. One of the common puzzles is of MIM to help best utilize this newer process. When coupled
why pricing varies so much between vendors. Via this cost with component-design guidelines, the combination provides
model, it is evident that busy facilities have the lowest costs, a potent means to design for improved performance with
so the incremental addition of more business often comes at a lower production costs. Throughout this chapter, a sand wedge
low price. On the other hand, if the facility is too busy, then golf club head is used to illustrate the procedures.
inefficiencies become evident. One of the important points Two words are defined for this discussion:
captured in cost modeling is how incremental changes in pro- manufacturing cost is the cumulative expense incurred
duction cause large cost changes. For example, a second shift in the production of a product; it includes the direct pur-
must be added or a new furnace must be purchased. The chases of raw materials, the expenditures for energy and
model presented here is the first comprehensive compilation labor, and prorated items such as maintenance, depreci-
of MIM costing. ation, insurance, and taxes
Besides costing, other underpinnings to MIM are treated sale price reflects the value of the component to the user;
in this chapter. For example, consistent nomenclature ensures it is the amount of money exchanged for the fabricated
accurate communication between the parties involved in any component and normally includes the manufacturing cost
project. This chapter provides definitions, conversion factors, and allocations for a return on investment and risk
and several facts on the technology, its documentation, the Ideally, cost and price should be related, but early in a tech-
trade exhibits and conferences, patent situation, and other de- nology that is not the case. During the 1980s, MIM products
tails. As a newer technology MIM has reached out to generate were selling at a price that sometimes was just 25% of the
standards for the materials, testing, properties, and specifica- manufacturing cost and bankruptcy was common. Today,
tions. Through standardization the user community gains con- metal powder injection molding is profitable, meaning that on
fidence with MIM. average the cost is below the price.
The information presented in this chapter follows engineer-
COST-CALCULATION APPROACH ing economic analysis as developed in powder metallurgy, ce-
Economic viability is critical to sustained growth for a new ramics, cemented carbides, and plastics molding. Powder
technology. An understanding of engineering economics helps injection molding depends on a combination of steps and costs
designers to choose between processes or to modify a design previously encountered in those fields. For the core manufac-
to reduce its manufacturing cost. Statistics show that the de- turing steps, unit operation calculations are developed for each
sign stage typically accounts for less than 10% of a projects step to accumulate a total fabrication cost. An outline of the
costs, yet decisions made in the design stage impact the ma- costing model is repeated from earlier and is given in Figure
jority of the production cost. Metal powder injection molding 9.1, while Table 9.1 provides a rough industry-wide cost sum-
lacks a well-established cost history, so designers are often mary breakdown for MIM component production. Although
150
Table of Contents
Chapter 9: Supplemental Information
Figure 9.1. Flowchart showing the basic MIM component-costing Figure 9.2. Detailed outline of the MIM tool-costing model in
model where each manufacturing step is analyzed for the cost flowchart form
accumulation
the time to machine the features for a single cavity, and then
TABLE 9.1. RANGES OF TYPICAL MIM PRODUCTION scales that cost to multiple cavities. Tool cost is calculated
COSTS from the construction time multiplied by a shop hourly rate,
cost item range with additions for profit (before tax, dividend, and interest)
direct labor 10 to 30% and risk aversion. Such an approach does not allow for rework.
feedstock cost 15 to 50% In reality, tool shops are averse to risk when fabricating a tool
tooling and molding 8 to 20% that is different from their prior experience base; thus, a risk-
debinding and sintering 3 to 15% aversion factor is introduced to allow for possible errors. Like-
interest and depreciation 5 to 9%
inspection, secondary operations 10 to 50% wise, business levels impact tool quotations, as evident by the
general, administrative, sales expenses 8 to 20% mapping of cost vs. delivery; contrary to MIM production,
busy tool shops are often more costly and have a longer deliv-
ery time.
there are recognized differences in all of the factors, including
labor, interestingly labor is not a dominant factor; MIM suc- Calculation Approach
ceeds in both expensive and low-labor-cost regions. Figure 9.2 shows the tool-costing flowchart, adapted here
based on values taken from the MIM industry. The first step
TOOL COSTS is to calculate the mold-base cost,
Tool-cost estimates for MIM are obtained using principles
encountered in plastic injection molding. Although many mod- CB = 2 (BT + 0.45 Ad Dd0.4) (9.1)
els exist, the following approach, adapted from Boothroyd,
Dewhurst, and Knight, has proven most accurate for MIM CB is the cost of manufactured mold base including cooling
tools. This model assumes a standard mold base and calculates and ejection parts (units of $)
151
Table of Contents
Chapter 9: Supplemental Information
BT is the cost of a manufactured standard tool-steel base (units tG is the manufacturing time to cut the geometric features of
of $) the part (units of h)
Ad is die-plate area (units of cm2), which depends on the part CI is the inner complexity of the part (dimensionless)
and molding machine CO is the outer complexity of the part (dimensionless)
Dd is the combined thickness of die plates (units of cm), which NR is the number or repeating features, such as holes of the
depends on part and machine same diameter (integer)
The factor of 2 In Equation 9.1 accounts for drilling cooling The two complexity parameters are determined as follows:
channels as well as machining inserts, runners, sprues, and
other features that scale with the tool size. For hot-runner sys- CI or CO = 0.1 (NS + NH) (9.5)
tems, major cost additions are required which are not included
here. NS is the number of surface protrusions, bumps, or patches
Calculation of the time to construct the tool cavity requires (integer)
an estimate for each of the major tool actions: NH is the number of holes or depressions (integer)
tS is the estimated hours to form each side pull (units of h)
tL is the estimated hours to form each internal lifter Beyond the machined features in the tooling, additional
(units of h) time is required to machine the parting plane depending on its
tU is the estimated hours to form each unscrewing device curvature or planarity,
(units of h)
nS is the number of side pulls (integer) tP = FP AP0.5 (9.6)
nL is the number of internal lifters (integer)
nU is the number of unscrewing devices (integer) tP is the time to manufacture the parting plane (units of h)
AP is the part projected area perpendicular to the direction of
where the cumulative time to form the major tool actions is mold opening (units of cm2)
given as FP is a factor depending on the type of curvature (dimension-
less), estimated as follows:
tC = nS tS + nL tL + nU tU (9.2)
FP = [3NL + (0.45 + NN2)] (9.7)
tC is the cumulative time for forming the major tool actions
(units of h) NL is the number of radial curvatures on the parting line
Usually, the time to machine an unscrewing device is (integer)
roughly twice that for machining a lifter (say 250 h vs. 150 NN is the number of steps on the parting line (integer)
h), and in turn the lifter takes roughly twice the time to
machine a side pull (65 h). A flat parting line is most typical and least expensive, while
A separate calculation is used for the time to form the cav- complex parting planes, described by curvatures and steps,
ity and any cores in the base. The first of these is the time to add to the expense.
machine the projected area and the associated ejector pins, Tooling tolerances are more stringent than part tolerances,
and in many estimates the tool tolerance is less than 20% of
tA = 5 + 2.5 AP0.5 + 0.085 AP1.2 (9.3)
the component allowance. Accordingly, considerable time is
tA is the machining time associated with cutting the projected used to fabricate a highly toleranced tool. The estimate of ad-
area (units of h) ditional machining time attributed to tight tolerances is given
AP is the projected area of part perpendicular to the direction as follows:
of mold opening (units of cm2)
tT = m FT tG (9.8)
The square-root dependence on part area relates to the ejector
pin machining and the 1.2 power dependence relates to the tT is the manufacturing time associated with the required tol-
empirical scaling law observed in machining the projected erances (units of h)
cavity. tG is the manufacturing time associated with the geometric
Next is estimation of the machining time associated with complexity (units of h)
the geometric complexity, which increases nonlinearly with FT is a dimensionless tolerance factor given as follows:
the number of features as follows:
FT = 0.000167/TR (9.9)
tG = 5.83 (CI + CO - 0.1 NR0.7)1.27 (9.4)
152
Table of Contents
Chapter 9: Supplemental Information
where
TR = ( Ti)1/m
the single-cavity cost by a procedure described as follows:
(9.10)
CT = CS NCK (9.14)
TR is the root-mean of the tolerances (units of mm)
Ti is the tolerance on the ith dimension (units of mm)
Ti indicates the product of the tolerances (multiply all m
CT is the cumulative tool-set cost for a multiple-cavity mold
(units of $)
tolerances times each other)
CS is the cost of a single-cavity mold (units of $)
m is the number of specified tolerances (integer)
NC is the number of cavities (integer)
K is a scaling factor associated with second and subsequent
Complexity concepts are a means to approach the cost of the
cavities (dimensionless)
tolerance factor as captured in Equation 9.8. Namely, cost is
directly proportional to the number of tolerances m. Generally
Recent studies on the scaling factor K suggest it is in the range
tighter tolerances increase the tool cost in a logarithmic man-
from 0.70 to 0.74. For most of the MIM situations, a value of
ner, which is approximated by the inverse factor in Equation
0.72 is typical.
9.9. Finally, as the overall geometric complexity increases,
the cost of tighter tolerances adds to the geometric factor cap-
Optimization of Multiple-Cavity Tooling
tured by Equation 9.4, which carries over into Equation 9.8.
It is possible to optimize the economic benefit from the use
Next is an appreciation that the specified tool-surface finish
of multiple-cavity tooling based on these equations. Letting
has an added expense. The time to achieve the specified tool-
NC represent the number of cavities, it is possible to consider
surface finish is estimated as
the tradeoff between more cavities (with an increase in ma-
chining cost) and the reduced molding time. The calculation
tF = FF (tA + tG)(1 + 0.05 YT) (9.11)
of the molding rate per hour MH is covered later, but assume
such a value is known (say in the $30/h range). Then the ben-
tF is the time associated with surface finishing (units of h)
efit from shorter molding time with multiple cavities is cal-
tA is the time associated with the projected area as calculated
culated with respect to the cost of machining the extra cavities,
above (units of h)
giving
tG is the time associated with the geometric complexity as cal-
culated above (units of h)
CM = CS NCK + (NP /NC)(MH / MR) (9.15)
YT is a factor (0 for no and 1 for yes) to indicate texturing, let-
tering, or insignias
CM is the total cost of molding (units of $)
FF is the surface-finish factor (dimensionless), given as
CS is the cost of a single-cavity tool (units of $)
K is the tool cost scaling factor (dimensionless)
FF = 0.125 + 0.06/RS (9.12)
NC is the number of tool cavities (integer)
NP is the production number or number of parts in a
RS is the surface roughness specified on the tooling (units of
production batch per year (integer)
m)
MH is the molding machine cost per hour (units of $/h)
MR is the average molding rate given as the cycles per hour
The factor YT notes whether lettering, insignias, numbering,
(units of 1/h)
or special textures are added, as these add to the overall cavity
and geometric construction times.
A minimum in the combined molding and tooling cost occurs
So far we have focused on the cost of the mold base and
when the first derivative of Equation 9.15, total molding cost
then estimated the time needed to create the required cavity
CM with respect to the number of cavities NC, equals zero. Set-
and detail in that base. To convert to a cumulative machining
ting the derivative to zero and rearranging the terms gives the
cost requires summation of the individual machining times.
optimal number of mold cavities NCo as follows:
The cost of a single-cavity tool is calculated from the con-
stituent times as follows:
NCo = [NP MH / (MR K CS )]1/(1+K) (9.16)
CS = CB + RH ( tC + tA + tT + tG + tF + tP) (9.13)
Normally, Equation 9.16 is rounded down and held to an even
number, so beyond 1 cavity, the normal progression is 2, 4,
RH = is the toolmakers hourly rate for mold fabrication (units
and on up to 32 cavities.
of $/h)
Two fundamental barriers show up in this optimization.
The first is from the molding machine: either the clamp force
The cost of a multiple-cavity tool depends on the dilation of
or the shot size might not be able to handle the optimized
153
Table of Contents
Chapter 9: Supplemental Information
number of cavities. For example, the peak molding pressure says that a four-cavity tool would be the lowest project price
multiplied by the projected area of each part multiplied by the for the intended production quantity.
number of cavities must be well below the machine clamping From the above model, the calculated four-cavity tool price
force. The second barrier is associated with the space required gives a tool cost starting near $23,000 for an hourly rate of
for mold actions, with fewer cavities being possible as the $25 per hour and 20% combined overhead and profit. It in-
number of unscrewing devices, lifters, or side pulls increases creases to a high of $38,000 for $60 per hour and 35% com-
per cavity. Two cavities might be possible with two motions, bined overhead and profit. The median quote was $34,870
but only one is possible with three or more motions. which corresponds to a 35% overhead plus profit and $48 per
hour rate. The low bid was $27,800, which corresponds to
Other Tool-Cost Factors 20% overhead plus profit and $41 per hour. Other combina-
Finally, there must be consideration of profit, tax, and in- tions would be possible as well. There is evidence that the
terest. For shops that have not previously fabricated a similar high bidder was too busy and would have had to pay overtime
tool, there is a risk aversion which necessitates a large safety to meet the production schedule, since they had the longest
margin in the tool cost. For the MIM customer, the final tool delivery time. Also, one of the higher bidders had previously
cost is the only element of concern, but for the tool fabrication taken on a MIM job and had a bad experience, so they in-
shop, underpricing errors can be fatal. Alternatively, conser- cluded a high risk-aversion factor in their quotation. In other
vative costing will lead to high prices and an idle business. words, all of the quotations could be explained using reason-
Risk aversion then describes the pricing policy (beyond profit, able model assumptions.
taxes, and interest) beyond the cost calculations. For example, Although this tool-costing model cannot predict actual
if the tool shop has a goal of 15% profit (before tax and inter- quotation prices, due to shop-specific cost factors, it still pro-
est), and estimates a 10% possible error in a quote for a new vides a guide to the cost implications from various decisions
tool, aversion to risk might lead them to include a 25% margin on part features, number of cavities, tolerances, surface finish,
in the quotation. Thus, the final mold cost for the customer is and other factors that might be adjustable during the design
given as follows: stage of a project.
As a final illustration of tool-cost calculations, assume a
CP = CT (1 + CA) / [(1 - P) (1 - RA)] (9.17) new golf club head is to be fabricated using MIM. The club
154
Table of Contents
Chapter 9: Supplemental Information
TABLE 9.3. SAND WEDGE TOOL QUOTATION INPUT AND CALCULATED PARAMETERS
head is a 56 sand wedge, with a MIM mass of 200 g. The lower production volumes it might be less expensive to per-
design is shown in Figure 9.3. Sand wedges generally range form secondary operations rather than create complicated,
near 300 g, but this club head uses a 100 g tungsten heavy multiple-action toolssuch as adding threads after sintering
alloy insert to reach 300 g with a lower center of gravity (not instead of using an unscrewing tool component. Depending
included in the calculations outlined here). Production for on the cost of cutting the threads, as a secondary operation, it
such a design would typically be in the 50,000 to 100,000 per is possible the unscrewing device is not cost justified. Other
year range, so only a single-cavity tool is justified. The cal- ways that may help lower the tool-machining cost is to reduce
culation input and output parameters are given in Table 9.3. the part size and complexity by a reduction in the number of
Using a standard mold base, typical tolerances (0.1 mm), 60 dimensions, toleranced features, holes, and depressions. Fur-
specified dimensions, and a 0.2 m tool-surface finish, the ther, relaxed tolerances help reduce tool cost. In the case of
predicted tool price would be $22,525 with no risk factor, 15% the golf club, only two of the features require tight control, so
overhead, and 20% pre-tax profit. An engineering charge less stringent tolerancing helps reduce cost. Finally, a low
might be added along with a risk factor, so the final price tool-shop hourly rate greatly lowers the final price.
might be quoted closer to $25,000.
If we wanted to significantly reduce this cost, then attention FEEDSTOCK COSTS
to the side pulls, depressions, holes, and number of toleranced A few materials with the powder characteristics desired for
features would be the way to generate the most savings. With MIM are widely available, so there is a potential cost advan-
judicious reengineering, realizing that most of the dimensions tage if the designer can use the more common materials in
are not critical and do not jeopardize the function, it is possible new designs. An understanding of raw-material costs helps
to cut this tool cost by 30%. the design community appreciate the economic implications
associated with various materials. A titanium golf club head
Summary of Important Tool-Cost Factors is used to illustrate the options in powder types and feedstock
This introduction to MIM tooling shows that a few features formulations.
quickly increase the tool cost. A first point is to avoid speci- Although the MIM technology was developed in the years
fying multiple cavities unless there is an economic benefit as- between 1940 and 1980, and has been in continuous use since
sociated with the total part-production quantity. Next, for the 1970s, still there is little quantitative information available
155
Table of Contents
Chapter 9: Supplemental Information
or formal benchmarking of costs. In this analysis, the costs of spherical prealloyed titanium powders sell for $150/kg. In
powders and powderbinder feedstocks are discussed. Usually some cases the demand cycles have significant impact on both
these are purchased constituents for all variants of the MIM price and availability, shifting raw-material cost up and down
process. A comparison is given for the cost of powders, poly- over time.
meric binders, and mixing for comparison with the price of Further, there has been a progressive price reduction for
premixed feedstock. The comparison shows self-mixing is a the raw materials as overall demand increases. For example,
natural transition to maintain lower costs as production vol- the average sale price for stainless steel powder has declined
umes grow. 30% in the past 10 years. This is the largest-use segment in
Costs in MIM often trace to differing availabilities of small MIM, accounting for 50% of the metallic component sales,
powders. In some compositions, namely, cemented carbides so it provides the greatest price competition.
and ceramics, the powders were available long before injec- Table 9.4 provides powder and feedstock prices for some
tion molding started. On the other hand, some powder com- popular chemistries. This table is a sampling and is not in-
positions are custom fabricated for MIM. This leads to a wide tended to be comprehensive and will change over time. For
material-cost range when MIM is compared to competitive each composition there might be six or more offerings, but
technologies. Thus, certain formulations have almost no ma- only one was selected to provide an overview of relative
terial-cost penalty associated with the MIM route, while other prices. For example, ten different firms offer the more popular
materials have a high cost when prepared in the size and shape stainless steel powders and at least ten firms sell stainless steel
characteristics needed for MIM. Appreciation of this differ- feedstock. For a 316L stainless steel (Fe-19Cr-9Ni-2Mo), the
ence helps guide the designer and producer to fruitful devel- raw-ingredients cost varies with the metal market, but in late
opment projects. 2010 was about $5.00/kg, melting energy is about $0.10/kg,
The current analysis relies on supplier cost data and reports and atomization costs about $1.50/kg, mostly for the inert gas.
on MIM costing for the raw powders and feedstock. The treat- The difference between the direct powder-production cost
ment is extended to include the cost of mixing. The latter ($6.60/kg) and purchase price ($15.00/kg) reflects a combi-
comes from research on MIM feedstock formulation in con-
tinuous mixers. Since mixers as well as powders and binders
all affect the cost, a generalized model is provided and applied TABLE 9.4. APPROXIMATE POWDER AND FEEDSTOCK
COSTS (various global suppliers, actual price depends
to several formulations. on supplier, quantity, particle characteristics, and raw
material costing)
Cost of Raw Materials composition powder powder feedstock
The first step is to select the engineering material, which $/kg $/kg
has a cost determined by the supply chain. In many applica-
aluminum gas atomized 75 103
tions the powder or feedstock cost is small compared to the bronze water atomized 11 18
component fabrication cost, since the components tend to be cobalt reduced oxide 45 58
small. Even so, powder or feedstock cost can determine the copper water atomized 12 19
viability of MIM when compared to competitive technologies. Invar (Fe-42Ni) gas atomized 21 31
Prior publications have detailed the size, shape, purity, pack- iron carbonyl 8 16
iron-nickel (Fe-8Ni) reduced oxides 20 28
ing, and other characteristics favorable for MIM. A small pow- Kovar water atomized 25 38
der size is required to allow sufficient interparticle strength nickel carbonyl 94 112
during debinding to retain the molded shape, and a rounded stainless 17-4 PH water atomized 15 28
powder with a high packing density ensures easy filling out stainless 304L water atomized 14 23
of the mold cavity. stainless 316L gas atomized 20 29
stainless 316L water atomized 15 24
Understanding the cost penalty or advantage associated stainless 430L water atomized 15 25
with different compositions is an important aspect of the de- stainless 440C water atomized 13 22
sign process. Compositions that are popular tend to have steel 4130 water atomized 15 25
lower powder prices. As with most manufacturing, price de- steel 8630 water atomized 16 27
creases as consumption increases. For example, stainless steel superalloy Hastelloy X gas atomized 26 41
superalloy IN 625 gas atomized 30 46
powder with a median particle size of 15 m produced by superalloy IN 718 gas atomized 25 40
water atomization decreases in price from $27/kg at a volume titanium hydridedehydride 88 150
of 100 kg to $11/kg at 10,000 kg, and falls to under $8/kg in titanium 6-4 gas atomized 125 180
long-term contracts. tool steel M2 gas atomized 22 57
Likewise, for titanium, at 30 kg the price is $85/kg for an- tungsten reduced oxide 33 68
tungsten heavy alloy mixed powders 45 67
gular powder and $120/kg for coarse spherical powder, but at tungstencopper milled oxides 75 104
120 kg quantities the price falls to $100/kg, while smaller
156
Table of Contents
Chapter 9: Supplemental Information
nation of factors that includes atomization yield in the desired HD is the hourly rate for capital equipment depreciation ($/h)
particle-size range, operational overhead, equipment depreci- CE is the equipment capital cost ($)
ation, sales expenses, and profit. tU is the hours of time the equipment is used per year (h)
Binders for MIM are also highly variable in cost depending YD is the capital equipment depreciation time (y)
on the formulation. The most common formulations are based
on mixtures of waxes and polymers. Market prices for large The use time depends on the number of hours the device is
quantities of commodity polypropylene or polyethylene are operating each day, number of days per week, and weeks per
as low as $2.40/kg. Special waxes with prescribed levels of year; it typically ranges from 800 hours per year (16 hours per
molecular weight and crystallinity can range up to $5.00/kg. week) to 7,000 hours per year for full-scale operations. These
Water-soluble thermoplastic polymers are priced as high as values roughly correspond to outputs ranging from 3 to 30
$10.00/kg. Polymers with surfactants cleaved onto the back- tons/month. Normally, the use factor needs to include every-
bone molecule are also in this price range. Surfactants, plas- thing from lunch breaks to holidays, vacations, shutdowns,
ticizers, and lubricants are usually added at low concentrations and idle weekends. Most facilities have excess capacity in
(about 5% to 15%) to the waxpolymer mixture to tailor prop- mixing, so the mixer-utilization rate is important in determin-
erties. For many of the binder formulations these agents are ing the cost of mixing. It is well recognized that MIM must
the same as those encountered in plastic injection molding, so strive toward continuous, full-time operations in order to be
they are commonly available stearates or similar species with cost effective. Unfortunately, much of the MIM industry uses
modest costs in the $2.00/kg range. So, the common thermo- their mixers at a rate of less than 60% of full capacity.
plastic binder system has a direct cost in the $2.00/kg to The cost for rental of the facility RT ($/m2) is likewise pro-
$10.00/kg window, depending on the formulation. For this rated over the equipment by first determining the calculated
analysis, the typical waxpolymer system gives a binder cost cost per unit time per unit area for the space,
of $4.50/kg. Such a binder is used in concentrations from 3%
to 15% to form the feedstock: lower concentrations with high- CR = RT A F/ tU (9.19)
density metals such as tungsten and lower concentrations with
low-density ceramics. Since the polymer is generally lower in CR is the rental cost per unit time ($/h)
cost than the powder, the net effect is to lower the feedstock RT is the facility-rental rate per unit area per year ($/(m2/y))
cost per unit mass when compared to the powder cost. In other A is the approximate area required for each device (m2)
words, on a raw-material basis, feedstock is lower in intrinsic F is the ratio of total space to manufacturing space (dimen-
cost than the powder from which it was fabricated. sionless)
Finally, raw materials must be levied for purchasing, trans- tU is the use time per year (h/y)
portation, inventory, and insurance costs. These tend to range
near 10% to 15% of the material cost; these calculations use The factor F reflects the need to allocate the rental cost of idle
12% of the material cost. For smaller operations, the protocol space against the productive activities. For many MIM oper-
is to purchase feedstock, but for larger operations the cost of ations this factor is between 1.2 and 1.5. Thus, idle space in
feedstock is better controlled by internal mixing. This calcu- the facility creates a burden on each unit operation, since the
lation will analyze the latter to calculate the annual material rental must be allocated to each productive device. Alterna-
use corresponding to a shift from purchased feedstock to in- tives include a burden based on percent time use of a facility
ternal mixing. or just a multiplier on all costs.
Next are the cost allocations for utilities, maintenance, and
Operation Parameters labor. These are calculated based on electricity, gas, com-
One major impact on feedstock production comes from the pressed air, water, and other resource-use rates gathered on an
operational context, which goes beyond the raw-material cost. hourly basis,
Included in this are interest rates, building-rental rates, em-
ployee benefits, capital-equipment depreciation schedules, ad- CU = Ci (9.20)
ministrative expenses, and financial policies. These factors
affect the mixing costs. For example, assume the equipment CU is the total hourly burden for each operating device (units
depreciation time is YD (years). Then a mixer can be assessed $/h)
an hourly cost based on its use rate and occupied space. For Ci is the individual contribution to the burden, such as elec-
the feedstock-mixing operation, this means that a straight tricity use per hour ($/h)
hourly-use rate can be computed as follows: i is a counter for the several unit contributions
HD = CE / (tU YD) (9.18) For example, if the material is abrasive, then mixer mainte-
nance will be high and that cost needs to be captured on a
157
Table of Contents
Chapter 9: Supplemental Information
prorated-use basis. Annual maintenance in industries involved $9,350/month for 1 Mg (1,000 kg), plus $292/month of
in sintered materials tends to be 5% of the equipment cost. binder. There is an additional 12% cost for purchasing, trans-
Labor costs differ with the level of automation. As automa- portation, insurance, and inventory or $1,157/month. The cap-
tion is added the capital-equipment-depreciation expense in- ital investment of $195,000 provides a depreciation expense
creases about as fast as the labor expense decreases. However, of $3,250/month (five-year depreciation). Allocated space cost
in most parts of the world the workforce can be easily adjusted would be approximately $125/month. For mixing one 1 Mg/
with the workload, while capital-equipment depreciation must month, the utility and labor costs amount to $345/month, and
be recaptured independently of the annual use. The ability to an additional charge of $100/month is included for cleaning
keep costs variable with manual labor often leads to the lower- the mixer. This gives a total cost for 1,000 kg of AISI 630 of
risk solution to use manual labor in low-labor-rate regions. As $14,807/month. Purchased feedstock in low quantities is
a result there is a progressive shift in MIM manufacturing to- quoted at $26.62/kg, and 1,078 kg would be required to de-
ward low-labor-rate geographic regions with less automation. liver 1,000 kg of product. With a 12% purchasing, transporta-
tion, insurance, and inventory cost, the pre-compounded
Mixing Costs feedstock would cost $32,139/month. The saving by self-mix-
Wear-resistant twin-cam or twin-screw mixers with the ing, when compared to purchased or premixed feedstock, is
high shear required for mixing MIM feedstock can produce $17,332/month. If the operation sets a financial goal of 30%
from 15 to 50 kg/h, depending on the material, particle size, return on assets, then this annual ratio of 3.6 savings-to-asset
and operating characteristics. The capital cost for new mixers cost per month makes a clear case for self-mixing. Hence, a
is from $140,000 to $250,000, and in a few cases reaches financially astute MIM operation would probably shift from
$450,000, with additional expenses for feeders, loaders, dry- purchased feedstock to self-mixing when consumption rates
ers, and other auxiliary devices for granulation, pre-mixing, approached approximately 1 Mg/month. In fact, firms with
and forming pellets. consumption of just 150 kg/month have elected to mix their
Mixing with a high-shear MIM compounder has a direct own special formulations.
energy content estimated at 1 MJ/kg (and can be higher for Accordingly, in most full-time MIM operations, mixing
smaller powders). This is a direct charge of about $0.02/kg costs probably range near $30/h or $0.60/kg, and are higher
for energy or about $1.00/h in most locations. Purging gas, for lower-density ceramics, especially with a submicrometer
cooling water, and compressed air combine with the electricity particle size. Eventually, in large tonnage situations, mixing
to create a total variable cost in the $2.60/h range, which in- can be scaled up to 4 to 5 tons/h with an anticipated asymp-
cludes $1.35/h for maintenance. The labor cost based on U.S. totic cost near $0.20/kg. Unfortunately, MIM has not achieved
averages in plastics is $14.66/h including benefits. Thus, the this scale, resulting in a premium price for feedstock. Most
variable cost for mixing is $17.26/h. However, space rental mixers can generate 50 kg/h of homogeneous feedstock, yet
and equipment depreciation expenses are accumulated even most of the industry is consuming feedstock at a lower rate.
if the mixer is idle. For a production mixer capable of 50 kg/h, Indeed the industry average is 9,000 kg of feedstock con-
this amounts to approximately $51,000/y in fixed expenses sumed per molding machine per year (which can be estimated
independent of the output. at 180 h of mixing per molder per year) some of which is in
Assume a production compounder mixing about 50 kg/h the form of recycled sprues and runners. Since the larger op-
of feedstock. If the mixer is busy most of the time (30 tons/ erations have 30 to 60 molding machines, this works out to a
month), then as noted above the mixing charge is small and requirement for one or two mixers. Consequently, there is ex-
amounts to just $0.53/kg ($17.26/h variable expense and cess mixing capacity in the MIM industry. The largest opera-
$9.27/h prorated fixed expense based on 22 h per day, 5 days tions consume 2,000 kg of feedstock per day. As the
per week, 50 weeks per year, and $51,000 annual fixed ex- particle-size decreases, more work of mixing is required, so
pense). For lower-use rates, say 5 tons/month, the processing the cost is higher and output is lower, but these estimates still
cost is more like $1.20/kg or nearly $60/h. For comparison, show an imbalance.
toll mixing services in the plastic field range from $0.12/kg
(40 tons/day) to $5.50/kg (300 kg minimum order). On the Feedstock for Golf Clubs
other hand, a premium of from $4 to $20/kg over powder cost To finish this phase of the costing discussion, attention is
is associated with the purchase of pre-compounded feedstock. given to feedstock costs for production of the sand wedge in-
One of the burdens is the capital-equipment-depreciation troduced the tool-costing section. Once the tool cost is esti-
charge. In operations with low mixer-use rates it is not fruitful mated, the next phase in the economic analysis is associated
to purchase a mixer and it is more prudent to pay the premium with the feedstock cost. This exercise is needed during the de-
for pre-compounded feedstock to avoid the capital expense. sign stage to ensure overall viability of the concept. Produc-
As an example, consider a facility that uses 1,000 kg/month tion is targeted for 100,000 clubs over a year, with normal
of 17-4 PH stainless (AISI 630). The raw-powder cost is operating schedules (90% effective use of a facility that is run-
158
Table of Contents
Chapter 9: Supplemental Information
ning 83% of the year). Overhead is set to 18% and minimum rier to large part production via MIM is economic, not tech-
return on assets is set to 30% per year with an alternative profit nical, since large MIM parts (several kg) are in production via
goal of 15% before taxes and interest. ceramics.
If stainless steel AISI 630 grade (17-4 PH) is selected as Powder, binder, and mixing costs are illustrated here for a
the MIM material, then the powder price is $9.35/kg (discount busy MIM operation. In such a case the typical cost of feed-
price for orders over 5 tons). A total of 20 tons is needed, since stock is dominated by the powder cost. The incremental cost
each head weighs 200 g. If the feedstock is self-mixed, it will for mixing is about $0.60/kg and the polymer cost is usually
take about three weeks using a single mixer (50 kg/h). Assume less than the powder and lower in density, so it dilutes the cost
the operation is fully engaged, so there is no need to prorate per kg for the feedstock. Currently the MIM industry is in a
idle time, then the feedstock cost will be $10.80/kg with a mode of excess mixing capacity. Typical statistics are cited
95% process yield, and the piece cost for feedstock will be for the industry and forward projections suggest that mixing
$2.45 each. However, if a pre-compounded feedstock is used, could approach $0.20/kg as an asymptotic limit based on cur-
then the piece cost increases. For example, one of the com- rent costs. As such, powder costs are the main component in
mercial offerings for AISI 630 is priced at $26.62/kg; then the the feedstock cost for internally mixed feedstock. For smaller
feedstock cost increases to $6.77 for each club. These calcu- operations, the premium price paid for pre-compounded feed-
lations assume the runner and sprue are reused. An alternative stock is probably justified. These calculations show that con-
way to lower the cost would be to use mixtures of elemental sumption levels of just 1,000 kg/month provide justifications
and master alloy powders that interdiffuse during sintering to for internal mixing.
produce the desired stainless steel alloy. This could be self-
mixed or purchased as premixed feedstock. BATCH-SIZE EFFECTS
Over the years, there has been much discussion and eval- This analysis of MIM cost is based on direct allocation of
uation of the options for combining MIM with titanium to fab- charges against a project, assuming an abundance of projects,
ricate golf clubs. Taking the same operating parameters as so only that proportion of the facility used for any given pro-
used for the AISI 630 version above, it is possible to substitute duction run is assigned to the cost calculation. In reality, idle
titanium. The lower titanium density reduces the mass to 115 shops are more expensive while shops running at capacity en-
g. Two versions are possible: counter high overtime or outsourcing expenses. Ideally, MIM
1) self-mixing which gives a feedstock cost of $7.40 each should be like a hotel: the charges in a hotel depend on the
(assuming the volume of the head remains constant) number of nights you stay, not on the hotels occupancy rate.
2) purchased pre-compounded feedstock which gives a However, like airlines, even hotels reprice to maximize profits
feedstock cost of $27.58 each as demand increases. So far the opposite philosophy seems to
As will be detailed in the subsequent part of this series, this be at work in MIM. If the MIM shop is not very busy, then it
latter value is over twice the sale price of the MIM stainless allocates the cost for idle time (for example, equipment depre-
steel golf club while the former is near the cost of the invest- ciation and rent) to fewer active jobs, leading to increasing
ment cast head. It would be difficult to convince the customer prices as business declines. One attribute of this approach is
that MIM titanium is justified with this differential just in that shops that are busier than planned produce higher profits.
feedstock cost, especially since the titanium club will have a Batch-size effects are significant. For example, consider
lower mass and may not be as successful in biting into the the cost of powder. Popular compositions have lower powder
sand. prices. As with most manufacturing, price decreases as con-
sumption increases. For example, 15 m stainless steel pow-
Feedstock-Cost Summary der produced by water atomization is priced as follows:
There are many different designs of mixers, many suppliers $27/kg at 100 kg
of powders, and some significant company operating differ- $11/kg at 10,000 kg
ences that affect mixing and feedstock costs. These example $8/kg in long-term contracts
calculations are used to illustrate the costs of feedstock and As another example of batch-size effects, consider the in-
the barrier such costs represents for MIM applications. Such jection cost depends on the cycle time and number of cavities.
cost barriers are larger for metals, which must compete against There is a set-up cost to install tooling, but once it is installed
casting, forging, machining, stamping, fine blanking, and the internal operating cost per hour for a molder is a standard
other conventional fabrication routes with lower raw-material ratesay, $25/h, while toll molding is more like $45/h. This
costs. In ceramics and cemented carbides, the MIM route has cost represents a combination of factorsdepreciation, rental,
less of a raw-material-pricing difference when compared to electricity, labor, overhead, and profit. It is desirable to use
extrusion, slip casting, die pressing, and cold isostatic press- multiple cavities to improve productivity at the expense of
ing; thus, large and more massive MIM structures are more machining the tooling.
common in these materials. Contrary to popular myth, the bar- Multiple-cavity tooling implies a tradeoff between up-front
159
Table of Contents
Chapter 9: Supplemental Information
160
Table of Contents
Chapter 9: Supplemental Information
are given in Figure 9.6 using a loglog scale for three stainless nificant cost reductions as the economic batch size reaches
steel components at 100 g, 20 g, and 5 g. At a volume of 1 one million units, but the cost of producing larger components
million each, the actual cost is $3.25, $0.55, and $0.30 each, (in the 100 g range) is insensitive to the batch size once
although final values depend on the design details. However, 100,000 units are fabricated.
assuming nominally the same complexity, this plot shows the
cost reduction with larger orders. Note that the smaller com- COMPONENT COSTING
ponent enjoys a much steeper and longer price decline, since MIM component-fabrication costs show scattered behav-
the material cost is small, while for the larger component cost ior, reflecting differences between sites in productivity, capac-
is dominated by raw materials and there is relatively less price ity, equipment, technology, labor rates, taxes, depreciation,
decrease with increasing volume. As another example of the and materials consumption. Many smaller MIM operations do
volume impact on cost, 85 g diesel-engine turbochargers cost not have the experience needed for accurate costing. This is
$20 each to produce at a quantity of 250,000/year, but the cost evident by companies where sales are only 50% of operating
declines to $4 each at 3 million/year. costs. In the long term only profitable strategies are realistic.
Metal powder injection molding excels at components For some short-term customers of MIM, working with an un-
below 10 g. Only single-cavity tooling is justified for most profitable operation is not a concern and they take advantage
components at batch sizes below 100,000, and at batch sizes of the pricing errors. However, for longer-term production, re-
of one million the move would be to two- or four-cavity tool- liance on inherently unstable MIM operations is not viable,
ing. The higher the component mass, the lower the number of so firms with large orders insist on accurate costing to ensure
cavities. Smaller components (in the 520 g range) see sig- supplier stability.
Pricing variability is common in business transactions and
especially true in MIM. As an example, a medical-device quo-
TABLE 9.5. COMPARISON OF BIDS FOR MIM tation for 20,000 parts per year gave the results summarized
MEDICAL DEVICE
in Table 9.5. The ratio between the highest and lowest price
metric part cost mold cost total project is more than a factor of 4. This raises the question as to pos-
mean $2.95 $16,392 $75,368 sible quotation errors of differences in business activity at the
median $2.04 $17,027 $56,743 various sites.
deviation $2.02 $7,408 $45,821 For accurate costing calculations, considerable information
lowest bid $1.55 $7,600 $42,600
highest bid $8.00 $31,000 $191,000
is required, including details that are elusive without experi-
high/low ratio 5.2 4.1 4.5 ence. Here we rely on a unit-cost calculation, as diagramed in
Figure 9.7. It assumes a yield for each step (molding, debind-
ing, and sintering), and allocates the labor, energy, consum-
ables, rent, depreciation, and other charges against the
product, largely based on the active time spent in each manu-
facturing operation. Considerable sophistication can go into
such calculations to account for differences in handling, cycle
time, automation, and factors such as vacation policies, benefit
packages, and depreciation schedules.
The current engineering economic analysis is based on a
direct allocation of charges against a project, assuming an
abundance of projects so only that proportion of the facility
used carries into the cost calculation. In MIM, however, it
seems that operations losing money more typically bid low,
even below cost, leading to a further spiral toward bankruptcy.
By the current model, if a shop is not very busy, then it must
allocate the cost for idle time to fewer jobs, leading to higher
quotations. As a benchmark, in the summer of 2003, globally
there were 258 active MIM companies running at slightly over
70% capacity.
The overall costing calculation is composed of stages that
Figure 9.7. Involved outline of the unit-costing concept as applied consider the tool cost, material cost, batch size, and individ-
to each of the MIM manufacturing steps. This approach accounts
for each component and its mass, energy, labor, and other costs as
ual-component cost. This installment deals with the latter por-
accumulated at each manufacturing step. Obviously it is beneficial tion, the unit operations and determination of cost allocation
to reduce the number of manufacturing steps per piece at each processing step. A quotation would assemble
161
Table of Contents
Chapter 9: Supplemental Information
all of these costs into a final bid. Additionally, risk and pro- pressed air, water, and other resource rates gathered on an
duction-balancing factors might be considered in forming a hourly basis as follows:
final quotation.
CU = Ci (9.23)
Operational Characteristics
A major impact on component-production cost comes from CU is the total hourly burden for each operating device (units
the operational context. Included in this are interest rates, $/h)
building rental rates, employee benefits, depreciation sched- Ci is the individual contributions to the hourly burden, such
ules, and such. For example, the equipment-depreciation time as electricity use per hour (units $/h)
YD (years) can be different between companies or different
locations. Each production device is assessed an annual de- For example, if the material is abrasive, then maintenance will
preciation charge. When the fabrication of a component uses be high and that cost needs to be captured on a prorated use
that device for a portion of the year, then it is assessed for its basis. Depending on automation, the labor allocations to each
relative use. Such calculations would take place for each step unit process will differ. As automation occurs there is an added
giving a rate HD computed as follows: depreciation expense. In North American MIM the reductions
in labor expenses with automation are about the same as the
HD = CE / (tU YD) (9.21) increases in capital equipment depreciation expenses. How-
ever, since labor can be varied with the workload, while cap-
HD is the hourly depreciation rate for capital equipment (units ital equipment depreciation must be captured independently
of $/h) of the annual use, this leads to the natural decision for MIM
CE is the equipment capital cost (in $) to use manual labor in low-labor-rate regions with low levels
tU is the hours of time the equipment is used per year of automation.
The use time depends on the number of hours the device is Molding Costs
operating each day, number of days per week, and weeks per Because of the large plastic injection molding industry,
year; it typically ranges from 4,000 to 7,000 hours per year molding costs are well established. Molding averages $18/h
for full-scale operations. Such a factor includes everything in Asia and ranges up to $40/h in the U.S. and even higher in
from lunch breaks to holidays, vacations, and weekends. Note Germany. If toll molding is used, then the cost includes over-
the sensitivity to the utilization rate, implying low-use device head, profit, and other factors, giving rates in the $37$45/h
time has a higher rate, to the point where if it is only used once range in the U.S. for the size molders used for MIM. Most
a year there is a singular yet high cost. components are generated in tool sets with one to eight cavi-
The cost for rental of the facility is likewise prorated over ties. As the number of cavities increases, the molding cost per
the equipment by first determining the calculated cost per unit part decreases, but the tool cost increases, as illustrated earlier.
time per unit area of space as follows: The cost-justified transition from single-cavity to two-cavity
tooling is usually between 100,000 to 300,000 pieces per year.
CR = RT A / (tU F) (9.22) Four cavities are not justified until annual volumes exceed 1
million parts per year. Currently 7 million parts per month are
CR is the rental cost per unit area per unit time (units of produced on a single machine with a 32-cavity mold operating
$/(h/m2)) 24 hours per day. Recent demonstrations have gone to 336-
RT is the facility-rental rate per year (units of $/m2) cavity tooling and 12 s cycle time for a simple MIM shape.
A is the area allocated to each device (units of m2) As a benchmark, the MIM industry averages slightly more
tU is the use time (units of h) than $1 million in annual sales per molding machine, about
F is the ratio of manufacturing space to total space twice that of plastic molding.
(dimensionless) For cost estimation purposes, section-thickness effect on
cycle time must be estimated, although actual times cannot be
Idle space creates a burden since the facility-rental cost is determined until the mold is constructed. Accordingly, cycle
spread over just the manufacturing space and each devices time is estimated using a parametric equation that is sensitive
relative use time. This calculation is repeated for each unit op- to the feedstock composition:
eration, and if the equipment occupies a large area then costs
grow. Likewise, plant expansion without direct contributions tC = CM + CE + CT WE2 (9.24)
to increased output results in an increased unit cost.
Next are the required allocations for utilities, maintenance, tC is the molding cycle time (units of seconds)
and labor. These are calculated based on electricity, gas, com- CM is a parameter reflecting the empty-molding-machine
162
Table of Contents
Chapter 9: Supplemental Information
cycle time (units of seconds) debinding relies on a water-soluble binder, then the equipment
CE is the extra time associated with various tool options (units cost is dramatically lower, near $20,000. However, penetra-
of seconds) tion rates are slower, so the cost difference is not as large since
CT is a feedstock parameter associated with cycle times to more units are needed and the parts dwell longer in each unit.
mold temperature, feedstock temperature, and feedstock Retort furnaces for thermal debinding capable of 100 kg/day
thermal conductivity (units of s/cm2) can be purchased for as little as $7,000. For small MIM com-
WE is the effective or root mean square wall thickness (units ponents, production debinding costs usually range near
of cm) $2.5/kg for a cycle that takes about 8 hours. As a first guess,
most debinding operations average a cost from $25 to $100
For a typical steel or stainless steel feedstock with a wax per hour, depending on the cycles, equipment type and size,
polymer binder, the parameter CM is 7 s and CT is 30 s/cm2. labor content, and utility rates. The load and cycle time deter-
The extra-time parameter CE comes from added steps that mine how this cost is partitioned into a per-piece debinding
slow the molding operation. For example, an unscrewing cost. With some components the cycle time is long and the
operation associated with an internal thread might add up to packing in the retort is poor, leading to a high per-piece cost.
10 s, a degating operation might add 3 s, and use of a robot For illustration, assume thermal debinding in a batch retort
can add 5 s. with a capital cost of $80,000 (retort and handling equipment).
Cooling time usually dominates the total cycle time. Cal- Based on a 5,889 hour use per year and five-year depreciation,
culation of the effective wall thickness relies on the outer em- the depreciation expense is $2.38/h of use. Based on rental
bracing dimensional envelope of the component using a rates and space, the unit might accrue a rental change of
modified empirical formulation found in plastics and invest- $0.09/h, utility charge of $9.08/h (electricity, gas, water),
ment casting to calculate an effective-thickness parameter for maintenance charge of $1.36/h, and average labor charge of
heat transfer: $9.44/h (monitoring, loading, and unloading). This totals to
$27.48/h in direct costs, but because of loading and unloading,
WE = [3 (L1f1 L2f2 L3f3)2 / [(L1f1 L2f2)2 + (9.25) a higher cost of $33.43/h is used. The component thickness
(L1f1 L3f3)2 + (L3f3 L2f2)2]1/2 now enters into the determination of the debinding cycle time:
163
Table of Contents
Chapter 9: Supplemental Information
$6/kg range, but the actual value varies from a low near $1/kg supports are necessary. Depending on the numbers of setter
to a high of $30/kg. For example, a small continuous sintering trays, part size, and furnace size, the setter costs can be sig-
furnace costs approximately $30/h to operate, but one with a nificant. Most of these setters need repair and replacement
six-fold higher production rate costs $120/h (4-fold cost in- after, say, 100 cycles, so a replacement cost is also a part of
crease gives 6-fold output gain). However, if there is insuffi- the sintering calculation.
cient work to keep the larger furnace busy, then it is a burden. As an example of sintering costs, a 6 g tool-steel piece that
Cost reductions drive MIM to larger equipment, but if there must be individually staged and vacuum sintered has a sinter-
is not adequate business to fill the equipment, then per-part ing cost of $0.10 each, but an 8 g stainless steel component
costs skyrocket. Larger and busy operations inherently have that packs tightly in the furnace is vacuum sintered for $0.05
lower costs. each. As an example of a low-cost sintering, a simple 25 g
The furnace design depends on the intended production steel component is sintered in a continuous-belt furnace in a
quantity, materials to be sintered, type of atmosphere, and nitrogen-based atmosphere for just $0.02 each.
post-sintering cooling rate. Production MIM batch sintering
furnaces exist for loads ranging from less than 1 kg to over Finishing Costs
300 kg (2 lb. to 660 lb.). Continuous sintering can reach levels Finally, there are many finishing steps required prior to
up to 100 kg/h. Larger batch furnaces typically operate on cy- shipping. These include secondary treatments such as machin-
cles lasting from 12 to 36 h. On the other hand, large pusher ing, heat treatment, or electroplating. Inspection and packag-
furnaces operate on entry-to-exit times ranging up to 24 h. ing are other final steps that add expense.
Continuous furnaces tend to have lower operating costs, but Each of the finishing steps typically has a charge based on
are less flexible. Thus, they are used only for sintering similar mass, time, or area. Because of the many factors involved,
components. Large components require slower heating and these are not single-valued parameters. However, a listing of
larger furnaces, thereby incurring even higher sintering ex- some of the common steps and approximate cost factors is
penses. Contrary to popular myths, large component produc- provided in Table 9.6. The bottom line is that each unit oper-
tion by MIM is not paced by technology; rather, the rapid ation has a calculated part or hourly rate. Rarely is a value
escalation of costs with size is the major barrier. analysis performed, but in situations where constraints exist
As with molding and debinding, there is a sintering-cycle there are some interesting pricing options.
time related to section thickness; however, this is mostly re-
lated to heating rates and soak times needed to obtain uniform Tight-Tolerance Effects
shrinkage. Smaller components can be heated quickly, with One of the significant cost issues in MIM comes from tight
demonstrations in continuous sintering of times near 2 h for 3 tolerances. Tight tolerances either reduce yield or increase the
g components. Thicker and more-massive components are secondary operations such as machining or grinding. Process
heated slowly to avoid distortion due to thermal gradients. yields fall quickly as tolerances increase. The industry average
Hence, the sintering-cycle time scales with the component coefficient of variation is 0.22% (dimensional standard devi-
thickness as follows:
164
Table of Contents
Chapter 9: Supplemental Information
ation divided by the mean size); thus, specifications tighter quickly, but give a low part density in the debinding and sin-
than about 0.5% (2.25 standard deviations) impact yield and tering devices. Consequently, molding will be out of balance
increase cost. Likewise, the more dimensions that are held to from the other operations. If the molder sits idle, then there is
tight tolerance the larger the probability the part will be re- a cost for idle time that needs to be accounted for in the pric-
jected. Accordingly, sorting is required to separate those in ing, otherwise overall financial performance sill suffer. This
specification from those not acceptablethis adds a costly capacity-balance cost needs to be calculated for the idle time
extra step. More of an impact is from the expense associated based on the characteristics of the part and production facility.
with production of the rejected components, an expense that Some facilities are better suited for certain sizes or shapes;
must be captured in the final shipped material. Thus, error thus, the capacity balance cost will be highly variable with
budget concepts need to be employed to balance cost and tol- equipment and its operation. Typically this cost is distributed
erancing. Use of secondary operations to hold tolerances over the quoted production quantity, and might amount to a
(coining or machining) can have a dramatic impact on costs. few percent of the cost in some cases.
Thus, tight tolerances are possible, but usually lower
process yield and add expensive secondary operations. As the Additional Costs
number of tight tolerances increases the economics often be- Other factors impact the final price. These are largely as-
come less attractive; hence, MIM defaults to machining di- sociated with overheadpurchasing, finance, sales, market-
rectly from wrought materials in cases were several features ing, research, management, and other business expenses that
are held in close tolerances. must be covered, but are only indirectly associated with a par-
ticular component. Further, there are often one-time (nonre-
Per-Piece Costing curring) expenses associated with fixtures, inspection tools,
This cost model is an accounting schedule that has unit costs handling devices, and other custom-designed items. These can
per time, and cumulative process costs are combined with feed- easily exceed $20,000 or more and must be amortized over
stock and subcontracted secondary costs to form a production the initial production order. Finally, profit is required to invest
cost estimate. Now the attention turns to the specific compo- in the future; across the MIM industry this is highly variable
nent and its production quantity and production rate. High- and is as low as 50% loss to 20% profit.
intensity projects require multiple molders and multiple tool Insurance is a means to reduce risk. Payments are made
sets. Basically, in the unit operation approach, multiple oper- periodically and only occasionally are claims made, but these
ating units do not change the cost, since attention is on how tend to be much larger than the payments. Risk analysis is as-
many parts can pass through each step or machine per hour. sociated with determining the exposure to loss. A modified
The unit operating cost is then allocated to each part as form of risk analysis is seen in MIM pricing. When a project
follows: forces a molder into a new realm of production experience,
often the prudent manager includes a contingency or fudge
PCi = CTi / U (9.28) factor in anticipation of surprises. Some of the problems might
be slow cycle times, more tool wear, or problems with defor-
U is the average number of parts produced per hour (units of mation in sintering. Aversion to risk leads to an extra cushion
1/h) in the price quotation to accumulate reserves in case a problem
PCi is the cost per piece (units of $) arises after production starts.
CTi is the unit operating cost per time (units of $/h) Generally, the risk aversion or contingency in pricing in-
creases as the quote strays from familiar ground. This is an-
These calculations are repeated for each unit operation and other reason for variability in price quotes, since different
the values are summed to estimate the piece price, along with vendors will have different experience levels. For example,
feedstock cost and yield factors. on a high-volume project, over eight firms participated in the
A penalty should be assessed when a component creates early quotation. The highest quote was from a smaller firm
an imbalance in the production operation. Molders, debinding with less experience in automation and high-volume produc-
devices, and sintering furnaces come in integers, so facility tion, while the lowest quote was from a firm with extensive
sizing might rely on a formula such as one mixer, six molders, prior production experience in high-volume components of
eight debinding retorts, and two furnaces or something of that similar size and shape. In other words, larger firms with broad
sort to make up a manufacturing complex. Depending on the production experiences will have more accurate quotes, but
cycle times for each step, it is possible for production to be not necessarily the lower quotes. Smaller firms with less ex-
out of balance, so some of the equipment sits idle. If there is perience often are the low bidders, since they do not recognize
sufficient business, then the spare devices are shifted to other some of the problems. Even a small loss per part times a large
products (possibly overloading downstream equipment). For number of parts can quickly bankrupt a firm.
example, a spindly component with thin sections will mold
165
Table of Contents
Chapter 9: Supplemental Information
166
Table of Contents
Chapter 9: Supplemental Information
167
Table of Contents
Chapter 9: Supplemental Information
factors. A redesign to reduce the wall thickness from 2.1 mm we will use a 316L stainless steel component for the analysis.
to 1.8 mm with the same mass lowers the price per part by Figure 9.10 is a pie chart of price contributions for this com-
3%. However, if the wall thickness and mass are decreased ponent using 4-cavity tooling at one million parts per year.
by 10%, then the price drops by 8%. Continuous sintering The calculations assume 15% general overhead, 10% profit,
gives a larger savings, and if continuous thermal debinding and 95% process yield for a factory that is typical as detailed
and sintering are combined, a total 15% price reduction oc- in Table 9.8.
curs. Hence, nearly 23% price reduction is possible via section The sensitivity analysis considered several factors to de-
thickness and process changes. Finally, if the profit is cut in termine the price changes with each. The sensitivities are
half, then the price drops to less than $0.18 each, yet is still ranked in Figure 9.11. Other factors considered, but not listed,
profitable. had even smaller impact in this case. The decision on pur-
This case illustrates that some costs are not important. To chased feedstock vs. self-mixing shows up as the most signif-
attend to the important factors shows that a MIM calculation icant parameter. Efforts to lower the mass, improve process
is needed to first identify the controlling parameters. The first yield, use lower-cost raw materials, and improve sintering all
step is to apply the model, determine the sensitivities over the promise to improve efficiency.
range of practical options, then implement strategies to meet As demonstrated in Table 9.9, a few factors have signifi-
the design objectives, yet lower the price.
An important approach comes from sensitivity analysis.
Each of the cost contributors has a different amplification
factor. Small changes to dominant factors provide large gains.
For example, sintering might have ten times the impact on
final cost when compared to molding. Thus, changes in
section size to increase molding productivity might be less
important when compared to efforts to reduce sintering time.
The sensitivity factor F provides a rationalization of various
influences:
168
Table of Contents
Chapter 9: Supplemental Information
cant impact on the final price; shown here are the benefits
from self-mixing, reducing mass 10%, improving process
yield to 98%, shifting to water-atomized powder, improving
furnace loading, and shortening sintering time. The price
drops to almost half that initially calculated. These changes
are realistic, since coring or section thickness changes can be
used to reduce the mass, process yields of 98% are realized
on larger orders, water-atomized stainless steel powders are
widely used in MIM, changes in the setters and packing of
parts can improve furnace loading, and thin substrates, heat
exchangers, and convective cooling can reduce the sintering-
cycle time. Curiously, labor rates, capital-equipment cost, and
facility-rental rates were less important factors.
Figure 9.10. Further illustration of cost contributions, showing
relative contributions to the price for a typical 8 g stainless steel Component-Mass Effect
MIM component
The above example was for the median 8 g stainless steel
component. At the other extremes, 10% of the components
TABLE 9.8. TYPICAL MIM PARAMETERS USED FOR are 0.3 g or less and 90% are 60 g or less. Table 9.10 compares
COST CALCULATIONS the upper and lower component attributes with the typical
part = 8 g, 316L stainless steel, 7.9 g/cm3, 1 million per
year, 25 mm long, 60 features
tolerance = 0.1 mm
tool base cost = $1,000
number of side pulls = 1
projected area of part = 500 mm2
number of surface bumps and depressions = 10
surface finish on tooling = 0.2 m
lettering on tooling = yes
toolmaker hourly rate = $50/h
number of cavities = 4
material = 316L stainless steel
powder = -20 m water atomized
powder cost = $13.06/kg
binder cost = $4/kg
facility-rental rate = $60/m2 per year
production floor area = 75% of factory
factory labor rate = $8.50/h
benefit rate = 15%
setup-engineering cost = $16,000
depreciation schedule = 10 years
maintenance provision = 5%
mixing-facility cost new = $195,000 Figure 9.11. Price-sensitivity factors for typical MIM component,
mixing rate = 50 kg/h showing the relative impact of various process or design decisions
molding-facility cost new = $90,000 with respect to final component price. Note that self-mixing of the
molding-cycle time = 30 s feedstock and reduction of the component mass are the two largest
debinding-facility cost new = $90,000 factors
debinding-furnace volume = 0.25 m3
debinding-furnace loading factor = 10%
debinding-cycle time = 18 h TABLE 9.9. PRICE REDUCTIONS POSSIBLE FOR A
sintering-facility cost new =$590,000
TYPICAL MIM COMPONENT
sintering-furnace volume = 0.42 m3
sintering-furnace loading factor = 10% parameter initial value improved value
sintering-cycle time = 24 h
electric rate = $0.10/kW-h feedstock purchased self-mixed
cooling-water rate = $2.00/h part mass, g 8 7.2
gas rate = $0.60/m3 process yield, % 95 98
process yield = 95% powder type gas-atomized water-atomized
operational hours per year = 6,000 h furnace loading, % 10 25
overhead = 15% sintering-cycle time, h 24 18
profit = 10% unit cost, $ 0.66 0.36
169
Table of Contents
Chapter 9: Supplemental Information
(8 g) component. The larger part would be characteristic of a the mass decreased, but the price increased to $17.91 each
turbine impeller, pump housing, or solenoid housing, while using hydridedehydride titanium powder, and increased to
the smaller part would be characteristic of a fiberoptic con- $53.23 each using spherical atomized titanium powder.
nector, orthodontic bracket, or surgical biopsy tool. If we examine the price factors for the stainless steel ver-
These two cases benchmark the range of MIM products. sion, then several small changes are identified to lower the
Figure 9.12 plots the price contributions from several factors price. For example, custom fixtures would increase furnace
for the three component sizes. The 8 g component was most loading in debinding and sintering to 20%, while convective-
sensitive to feedstock (mass), yield, and sintering cost reduc- gas cooling would reduce sintering time to 18 h. Other realis-
tions. For the larger (60 g) component, the comparative plot tic changes, such as improving process yield to 98%, would
shows that feedstock cost is even more important, followed bring the price to an asymptotic level near $6.12 each. This
by sintering. On the other hand, for the small (0.3 g) compo- approaches the price currently paid for investment cast golf
nent, the price is dominated by finishing and tooling costs. club heads, so a viable project emerges, but with great effort.
From the sensitivity analysis, cost-reduction efforts are most For titanium there is little hope for MIM to compete against
fruitful if applied to the major cost contributors. investment casting, even with low-cost powder.
170
Table of Contents
Chapter 9: Supplemental Information
a high-velocity air stream or cyclone. feedstock. It is common to use trimodal mixtures in ce-
Alloy powdera powder in which each particle is composed ramic MIM.
of the same mixture of two or more constituents, especially Binderthe polymer mixture that provides lubrication and
common in gas- or water-atomized powders such as stain- strength to the powder. It is critical to the feedstock fluidity
less steels. during molding and to the strength of the molded compact.
Aluminathe compound of aluminum and oxygen Al2O3 The binder is evaporated during debinding and sintering.
that is the most common material used in ceramics, often Binder formulations generally fall into a few classes, but
used as a furnace lining, hardware, or setter support for exact formulations are usually proprietary. The binder
sintering. chemistry impacts the debinding process.
Angle of reposethe angle from the horizontal plane that a Bingham flowviscous flow of a feedstock with initial yield
pile of loose powder will assume when freely poured strength, meaning the stress must exceed the yield strength
through an orifice, providing a measure of the interparticle prior to initiation of viscous flow.
friction and ease of shape retention during debinding. Blendingthe thorough intermingling of powders of the
Apparent densitythe mass of a unit volume of powder in same nominal composition, for example, the combination
the loose condition, usually expressed in g/cm3. For MIM of water-atomized and gas-atomized stainless steels.
powders it is often 50% of the theoretical density for the Brown statea term originally used to describe a ferrous
material. MIM component following debinding by slow heating in
Aspect ratiothe ratio of the maximum particle dimension air to remove the polymer, where oxygen caused the iron
to the minimum dimension. to oxidize into a rust-colored component.
Atomizationthe dispersion of molten material into droplets Burnoffthe removal of the polymer binder via preheating
by a rapidly moving stream of gas or liquid (usually water), prior to sintering.
or by centrifugal force. The droplets solidify into particles. Capillary rheometera device for measuring viscosity by
Atomized powderalloy or metal powder produced by the applying pressure on molten feedstock and determining the
disintegration and subsequent solidification of a molten flow-rate dependence on applied stress as the feedstock is
metal stream. pushed through a small capillary tube. It is useful in deter-
Attritiona mechanical milling or grinding process that typ- mining feedstock flow and homogeneity.
ically employs a stirred or tumbled container filled with Carbo-nitridea heat treatment in a nitrogen and methane
small balls that act as a crushing medium. The tumbling atmosphere that causes both carbon and nitrogen to diffuse
action causes repeated collisions against particles mixed into the steel surface for strengthening and hardening.
with the balls. The process is widely used for deagglomer- Carbon controla measure of the ability to remove the MIM
ation, partial alloying (W-Cu, WC-Co), and particle-size polymer without contamination of the powder. Carbon con-
reduction in ceramics. trol is quantified by the final carbon level and the unifor-
Austenitethe face-centered cubic crystal structure of pure mity of that level between parts, or over subsequent
iron that is not stable below 910C. However, with alloying operations, fabrication days, or feedstock batches. For
this nonmagnetic form of iron can be stabilized at room some materials the desire is no final carbon (austenitic
temperature and even cryogenic temperatures; austenitic stainless steel, alumina, titanium), while others require a
stainless steels are a common example where sufficient precise final carbon level, and in the extreme cemented car-
nickel is employed to stabilize the nonmagnetic phase at bides and silicon carbides require very high carbon levels.
room temperature. Carbonyl powderpredominantly an iron powder, but may
Barrelthe heated portion of the feedstock-flow path before also be nickel. These powders are prepared by the thermal
the nozzle. The barrel holds the feedstock under pressure decomposition of a metal carbonyl molecule. The resulting
while providing heat to melt the binder system. particle size is typically in the size range from 1 to 10 m.
BET surface areathe specific surface area as measured by The particle shape can be spherical, agglomerated, or
gas adsorption according to the Brunauer, Emmett, and highly angular.
Teller theory. Widely used in characterization of ceramic Carburizationdecarburizationtwo events critically re-
MIM powders and expressed as square meters per gram, lated to carbon control, since the addition of carbon to a
m2/g. material occurs by carburization from binder or atmosphere
Bimodala particle-size distribution exhibiting two mode sources, while decarburization occurs by reacting carbon
(mode is the most common or most frequent) sizes. in the material with hydrogen, oxygen, or carbon monox-
Although such a powder can result from mixing two dif- ide.
ferent particle-size distributions, some powder-production Case carburizea post-sintering heat treatment aimed at dif-
techniques naturally produce this distribution. Its main ad- fusion of carbon into the surface of a ferrous MIM com-
vantage is in producing a higher solids loading in MIM pact. The carbon increases strength and hardness.
171
Table of Contents
Chapter 9: Supplemental Information
Catalytic debindingthe partial removal of the polymer via the mean value, giving a nondimensional measure of uni-
heating in an atmosphere containing an agent that induces formity such as in dimensions, weight, or properties.
depolymerization. The classic process involves extraction Coiningthe final pressing of a sintered compact to obtain a
of polyacetal using a nitrogen atmosphere doped with nitric definite surface configuration, flatness, or surface finish.
acid. Co-moldingalso known as two-color molding because of
CCIMcemented carbide powder injection molding, where the use in plastics. Two materials are shot into the same
the inorganic phase in the polymer binders is predomi- cavity from different injector units to form layers, inter-
nantly a mixture of cementing phase (cobalt is typical) and connections, or other differences in materials with position
a carbide (tungsten carbide is typical); a subset of powder in the cavity.
injection molding (MIM). Compositea mixture of two or more powders that form a
Cemented carbidea solid composite consisting of a metal multiple phase structure, typically designed to deliver prop-
carbide and a binder phase, usually cobalt or nickel alu- erties that are a hybrid of the constituent properties.
minide. The composite is formed by liquid-phase sintering Compressionthe removal of air from melting feedstock by
a mixture of the carbide and binder metal powders. The applying heat and pressure. In a reciprocating screw mold-
WC composites are known as hardmetals in Europe. They ing operation this is achieved by tapering the screw to re-
are produced by MIM and used for wear, tooling, cutting, duce the space for the feedstock during metering toward
or drilling applications. the screw tip.
Centrifugal atomizationthe formation of spherical parti- Computer-aided designthe use of a computer program
cles by combining a melt with a centrifugal force such that geared to perform stress, fit, dimensional, and other calcu-
the melt is disintegrated into high-velocity droplets that lations and to allow visualization prior to fabrication.
spheroidize prior to solidification. Computer-aided manufacturingcomputer controlled ma-
Cermeta composite body consisting of ceramic particles chines used to ensure proper fabrication and resource uti-
bonded with a metal. Cemented carbides are the most com- lization; included in the broad category of computer
mon cermets in production. integrated manufacturing and computer-aided manufactur-
CIMceramic powder injection molding, where the inor- ing activities might be inventory control, maintenance
ganic phase in the polymer binders is predominantly a ce- schedules, production scheduling, tool-path analysis, and
ramic or mixture of ceramics; a subset of powder injection cost analysis.
molding. Critical loadingthe maximum volume fraction of solid par-
Clamping forceone of several measures of the capabilities ticles that can be incorporated in a polymer binder without
of a molding machine. In this case the available force for forming pores while still allowing flow in normal injection
holding the mold together while pressurized feedstock is molding situations.
filling the cavity. If the applied pressure times the projected Cross linkingthe formation of bonds between polymer
part area exceeds the clamping force, then the cavity will chains to give rigidity and strength to the polymer. Ther-
open or flash during molding. mosetting polymers that harden on first heating are exam-
Classificationseparation of a powder into fractions accord- ples of cross-linked polymers.
ing to particle size. Cycle timea critical measure of molding-equipment pro-
Closed-loop feedback controla molding concept for pre- ductivity, it is the time for completion of one molding
cise dimensional control where the pressure inside the cycle. It can be measured from the time to start filling to
molding cavity is monitored during filling and used to ad- the start of the next fill. Cycle times from a few seconds to
just the molding machine operation to ensure repeatable 5 minutes are encountered in MIM production.
filling, weight, and final dimensions. The weight and di- Debindinga step between molding and sintering where the
mensional scatter in final parts is greatly reduced by using majority of the binder used in molding is extracted by heat,
this control logic as compared to open-loop or adaptive solvent, catalysis, or other techniques. The debinding tech-
controls. niques are highly variable between production sites. Ther-
Closed porean isolated pore not linked to the external sur- mal debinding is most common and is the oldest version,
face, usually formed in sintering after a component reaches but several operations rely on alternatives or combinations
a density of approximately 92% of theoretical, resulting in of methods.
capture of the process atmosphere in the pores. Delaminationthe cracking of a molded compact, often lead-
Coarseningthe progressive enlargement of the grain size ing to a hairline crack oriented perpendicular to the ejection
or pore size during sintering due to diffusion, coalescence, direction that is hidden until debinding or sintering.
or solutionreprecipitation processes. With respect to prop- Densificationthe change in porosity divided by the initial
erty control, microstructure coarsening is very important. porosity due to pressing or sintering. A term loosely asso-
Coefficient of variationthe standard deviation divided by ciated with property gains in sintering.
172
Table of Contents
Chapter 9: Supplemental Information
Densitythe mass divided by the volume, usually expressed MIM feedstock it also is important because it determines
in g/cm3 (equivalent to Mg/m3) or sometimes given as a residual stress and the stress relaxation time in cooling.
ratio to pycnometer or theoretical density. Elasticitythe spring-back of feedstock after ejection from
Dew pointa measure of atmosphere purity based on water a tool set. Formally, the elastic modulus is the material pa-
content, it is the temperature where moisture condenses out rameter that links stress to strain, but in MIM it is largely
of a process atmosphere. Generally it is assumed a low dew related to die sticking and dimensional-control problems
point (low condensation temperature) corresponds to a since ejection and debinding induce stresses that might
cleaner atmosphere. cause warpage or loss of component precision.
Differential scanning calorimetrya means to determine Elemental powderpowder of a single chemical species like
the heat flow into or out of a MIM sample. Usually it is ap- iron, nickel, titanium, copper, or cobalt, with no alloying
plied at lower temperatures where the polymer melts, crys- ingredients.
tallizes, or evaporates. Elongation to fracturea measure of ductility, since this is
Differential thermal analysisthe careful measure of tem- the amount of permanent plastic stretch a material under-
perature and temperature difference during heating for a goes prior to failure in a tension test. The most common
MIM material and a reference. Whenever the polymer measure used for ductility.
melts or a phase transformation occurs the test sample will Equiaxed powderparticles with approximately the same
lag behind the reference, while if there is a reaction the size in all three (perpendicular) dimensions. A sphere
sample will heat faster than the reference. formed by gas atomization is the classic example of this
Dihedral anglea microstructure feature associated with desired particle shape.
phase boundaries in a material. It is evident when a grain Feedstockthe mixture of powder and binder used in injec-
boundary intersects with a pore or liquid phase, resulting tion molding. Its formulation involves decisions on the
in a groove in the grain boundary at the point of intersec- powder composition, particle characteristics, binder for-
tion. Low dihedral angles are associated with high grain- mulation, mixing practice, and ratio of powder to binder.
boundary energies and unstable microstructures. Ferritevarious compositions based on iron oxide used to
Dilatant flowviscosity that changes with flow conditions fabricate common magnets, usually containing other metal
where the mixture actually dilates under stress. oxides or ceramics, including zinc oxide. It is also the name
Dilatometrymeasurement of dimensional change during given to the body-centered cubic form of pure iron.
thermal processing to determine the sources of sintering Ferritic steelan alloy based on iron that consists of a body-
densification, phase transformation, or other causes of centered cubic-crystal structure. It is the most common
dimensional-control problems. Usually dilatometry is con- form of low-alloy steel when slow cooled, but is usually
ducted with a laser or a pushrod that makes contact with heat treated to form martensite.
the specimen in a furnace. Sintering experiments intended Fillingthe first phase of the molding cycle where feedstock
to find the densification temperature for a powder are con- is flowing into the mold cavity under pressure from the
ducted using constant heating-rate dilatometry. screw.
Dimensional controlthe repeatability of final dimensions Finishing operationsthe steps applied to a MIM compo-
in a MIM operation as measured by part-to-part, day-to- nent after sintering to tailor the dimensions, properties, or
day, and batch-to-batch scatter. Usually quantified by the attributes to the application; examples include machining,
standard deviation (or a multiple of standard deviations, polishing, heat treatment, straightening, and electroplating
perhaps as large as six) observed in a dimension as nor- steps.
malized by that dimension, expressed as a percent, for Flashinga lip of extruded feedstock that penetrates along
example, a coefficient of variation of 0.1%. the parting line of a die cavity due to excessive pressure,
Ductilitya measure of the permanent stretching or defor- poor tool tolerances, or binder separation from the powder.
mation a material can take prior to failure. Ceramics have Flow analysiscomputer-simulated molding to assess the lo-
no ductility, while materials such as stainless steel and alu- cation of the gate, runner, vent, and cooling passages and
minum exhibit large ductilities, often measured at 30% to other important aspects of tool design and molding to min-
60% stretch or elongation prior to failure. imize errors in production.
Ejectionthe final stage of molding where the powder Flow timethe time required for a powder sample to flow
binder compact is forced out of the die. through an orifice in a standardized test. The flow time
Ejector pinsmechanical pins that insert into a die cavity to gives a measure of the interparticle friction. Most powders
act as a means to push the molded component out of the used in MIM are not free flowing due to the typical small
tooling. particle size.
Elastic modulusthe material property linking stress and Fracture toughnessa measure of the resistance to crack
strain. Also known as Youngs modulus or stiffness. In propagation in a material, related to the applied stress when
173
Table of Contents
Chapter 9: Supplemental Information
the crack starts moving and the crack size. High-fracture- properties such as strength.
toughness materials, such as stainless steels, provide more Heavy alloya class of high-density alloys based on tungsten
safety when compared with low-fracture-toughness mate- with small concentrations of alloying additions such as
rials such as glass or ceramics. nickel, iron, or copper. These alloys are liquid-phase sin-
Freeze firinga concept developed in the 1960s using water tered from mixed elemental powders to create a composite
as the binder and freezing in the tool cavity. The frozen material. Most of the applications are for weights, radiation
compact is ejected and subjected to sublimation to extract shields, thermal management heat sinks, or projectiles.
the water without melting the ice. Hot isostatic pressinga process combining temperature and
Gas-atomized powdera rounded or spherical powder high-pressure gas to fully densify a sintered MIM structure.
formed by the disintegration of a melt stream by a high- Used only for very-high-performance structures.
pressure-gas expansion nozzle. The particles solidify dur- Hot runnera tool cavity where the flow path is kept hot
ing free flight after atomization. between shots to eliminate recycle of runners and sprues.
Gatethe constricted opening into the flow path at the entry The built-in heaters and valves are coordinated with the
to the die cavity in the injection molding tool set. It should molding machine to ensure no freezing in the flow path.
be the first portion of the flow path to solidify after filling Impregnationliquid polymer filling of open pores from an
the mold. external surface, used to seal open pores after sintering for
Gelationa binder-setting process where a macromolecule improved lubrication, corrosion resistance, or machining.
grows in a binder solution to form a highly interlinked Inconela variety of nickel-base alloys invented by Interna-
structure with most of the binder water trapped in cells tional Nickel Company designed for high-temperature ap-
formed by the long-range molecule. plications, ranging from furnace components to jet engines.
Gelcastinga variant of low-pressure injection molding Infiltrationthe process of filling the pores of a compact
where the binder consists of a monomer that polymerizes with a lower-melting-temperature metal or alloy. It is one
in the die cavity, forming a rigid polymer to hold the par- means of forming low-cost tooling or making dense struc-
ticles in place, often supported by a catalyst addition just tures via molding a porous preform and filling the pores
as the feedstock is molded. with liquid metal, such as Al or Cu.
Granulationa term describing the agglomeration of pow- Injection moldinga hydrostatic forming technique for
der or the breaking apart of lumps, runners, sprues, or parts shaping powders using plastic binders and relatively low
that are reformed into pellets for reloading into the molding temperatures and pressures.
machine. Interparticle frictionthe friction between powders which
Green statethe condition of the molded component prior limits sliding, packing, and densification.
to debinding or firing. The term comes from the ceramic Invarlow-thermal-expansion alloys of ironnickelcobalt
concept of greenware in reference to formed bodies that where a martensitic phase transformation is balanced
are not sintered. against the thermal expansion coefficient to give a near-
Green strengththe strength of the as-molded component zero thermal expansion coefficient over a range of temper-
at room temperature. atures.
Guided wavean in situ inspection technique involving Irregular powdera powder which lacks shape symmetry
wave propagation through a MIM compact while still in in the individual particles.
the mold. Mathematical transform of the scattering behav- Jettinga condition that arises with the rapid filling of an in-
ior detects defects, allowing for sorting on ejection. jection mold where the feedstock shoots across the mold
Hardmateriala group of compounds that are typically and fills back toward the gate. Generally this is unaccept-
metal carbides, borides, oxides, or nitrides that exhibit a able for quality components.
high hardness. To fabricate components, various combina- Knit linethe same as a weld line. A linear defect occurring
tions of the hardmaterials are combined with cementing where feedstock streams merge in the cavity because of
metallic phases during liquid-phase sinteringfor exam- two gates or flow around a core or other solid portion of
ple, TiC is sintered with Ni or Fe to form a cermet. the die.
Hardmetallike hardmaterials, but usually reserved for the Kovara glass-to-metal sealing alloy used for microelec-
WC-Co family of hardmaterials or cermets. This name for tronic packaging and other situations where a matched
the cemented carbides is more popular in Europe but in thermal expansion or graded structure is required.
general it reflects the high hardness after sintering. Laminationa layered structure or cracking in the pressed
Hardnessa formal test of resistance to penetration by a compact resulting from ejection stresses exceeding the
pointed or rounded indenter under a given load. There are green strength.
many useful hardness tests where the depth or width of the Liquid-phase sinteringsintering at a temperature where a
penetration is measured. Often this relates to other material liquid and solid coexist due to chemical reactions, partial
174
Table of Contents
Chapter 9: Supplemental Information
melting, or eutectic liquid formation. It is most useful for are in microelectronics, medicine, sensors, and optical
stainless steels, tungsten, cemented carbides, cermets, and communication systems.
ferrous alloys containing phosphorus, boron, copper, or sil- Microstructurethe detailed information on the microscopic
icon. phases, pores, grains, defects, heterogeneities, and other
Low-pressure moldingthe use of lower pressures and low- property-controlling features.
viscosity binders (largely water-based or wax-based) to fill MIMmetal powder injection molding, where the inorganic
out a complex tool cavity without packing the shape. This phase suspended in the binder is predominantly a metal or
route is successful for components where internal flaws are alloy powder. Common engineering alloys are possible by
not a concern, such as nozzles, spray tips, or other geome- mixing elemental powders and forming the alloy during
tries where external geometry is the key concern. sintering (homogenization) or by use of a prealloyed pow-
Lubricantan organic additive which is mixed into the feed- der where each particle contains all of the elements. MIM
stock to minimize die wear and aid in ejection after com- is a subset of powder injection molding (PIM).
paction. Mixingthe thorough intermingling of powders of two or
Maraging steela high-strength, high-toughness class of more different compositions. It is also used to describe the
ironnickelmolybdenum alloys that lack carbon, yet form compounding of feedstock by the thorough distribution of
their strength by an age-hardening process in the marten- binders between particles.
sitic phase. Strength levels of 4 GPa or 600 ksi are possible Mold-flow simulationthe use of computer simulations for
with these alloys. analysis of the flow, packing, venting, sizing, cooling, and
Martensitethe distorted ferrite crystal structure (the com- other events and tool or machine parameters associated
mon variant is a distorted body-centered cubic phase that with MIM forming operations.
forms a body-centered tetragonal structure) due to carbon Mold releasea spray or coating that reduces component
supersaturation in the rapidly cooled ferrite. Martensite is sticking to the die cavity, aiding ejection without defects.
hard and brittle, hence it is usually tempered to partially Moldabilitya relative measure of the ease of filling out a
relax the hard phase for improved toughness. tool cavity during injection molding. It can be determined
Mean sizethe average value from the particle-size distribu- by the length of filling for a long, narrow passage. The cur-
tion. rent test of plastic moldability is to measure the filling of
Mechanical alloyingthe formation of an alloy powder by a spiral, which has been altered in MIM to a zig-zag fill
milling elemental powders for a prolonged time; frequently test to induce powderbinder separation along the flow
used to create amorphous or dispersion-strengthened alloy path.
powders via attritor milling. Multimodala powder-size distribution which exhibits
Median sizethe centroid of the particle-size distribution, several modes, possibly generated by blending several
where half of the particles are larger and half are smaller; monosized powders.
not necessarily the mean, but easily identified as the 50% Nanoscalepowders or microstructures with sizes that can
value. Also known as the D50 particle size. be measured in nanometers. Typically the powders are less
Melt indexa measure of flow at low-shear-strain rates, than 100 nm in size, or less than 0.1 m.
where a capillary tube is used to extrude molten feedstock Near-net shapea discrete component without a need for
under a dead load. The melt index depends on the capillary final machining which many production technologies, in-
tube diameter and applied load, but is always reported as cluding MIM, attempt to form. Together these technologies
the grams of feedstock collected from the tube in 10 min- are considered net-shape approaches, but when critical di-
utes. mensions cannot be held without a final machining step
Meteringcontrolled forward extrusion of molten feedstock they are termed near-net-shape technologies.
past the screw tip and check ring to ensure the proper shot Net shapea compact manufactured to final density and
volume is ready for the next filling event. dimensions without the need for machining. MIM is a net-
Microminiaturecomponent dimensions measured in the shape process.
micrometer size range, requiring microscopes for evalua- Newtonian flowan idealized viscosity situation applicable
tion. These are typically formed using special machines to a few fluids where there is no sensitivity to the shear-
with smaller shot size to enable better precision for small strain rate, only temperature. The stress is proportional to
bodies. the shear-strain rate.
Micromoldinga new class of MIM technologies geared to Nodular powderirregular particles with knotted, rounded
the production of components in the millimeter and mi- shapes. This is characteristic of water-atomized powders.
crometer size range, requiring new technologies in tool fab- Open porea pore completely through a compact from one
rication, nanoscale powders, and molding machine surface to another. During thermal debinding the pores must
operation and construction. Most of the early applications be open to allow the escape of evolving vapors. Open pores
175
Table of Contents
Chapter 9: Supplemental Information
close in sintering at a density near 92% of theoretical. Porosimetera device for measuring the size of the open
Oxidationreductiona combination of atmospherepow- pores using high pressures and mercury-intrusion tech-
der reactions that can extract oxygen (reduction) or deposit niques.
oxygen (oxidation) in a powder compact during heating, Porositythe amount of void space in a powder compact.
especially during debinding and sintering. Reduction con- Most MIM materials have less than 5% porosity after
ditions are usually required for sintered metallic materials. sintering.
Packing pressurethe peak pressure encountered in the Powderparticles of solid matter characterized by a small
molding operation once the die is filled, prior to freezing size, less than 1 mm in size. Most MIM powders are below
of the gate. It is precisely controlled for weight uniformity 20 micrometers in maximum size.
and optimized final dimensional control. Prealloyed powdereach particle contains an intimate mix-
Particle sizethe controlling linear dimension of an individ- ture of two or more elements in a prescribed ratio to form
ual particle, as determined by analysis with screens, lasers, an alloy; examples include brass, bronze, steel, and stain-
or other sensing techniques. less steel.
Particle-size analyzeran automated device for determina- Presinteringthe heating of a compact to a temperature
tion of the particle-size distribution. These are widely used lower than the normal sintering temperature to gain
in the research, product-development, and quality-control strength for subsequent handling, including machining.
functions as part of MIM to ensure repeatable powders for This is often performed by heating above the highest tem-
the process. perature required to thermally decompose any residual
Parting linethe linear mark on a compact where two binder polymer.
separate tool or die pieces mated during shaping. In injec- Pressure-assisted sinteringsintering with the application
tion molding it is where the two halves of the die joined of an external pressure. It is often performed by initially
together. sintering in vacuum and subsequently pressurizing the fur-
Pelleting or pelletizingthe formation of discrete chunks of nace to densify any remaining closed pores. Best applied
feedstock with repeated sizes and shapes that allow easy to high-performance alloys or difficult-to-sinter alloys and
flow and filling of the molding machine. is commonly employed for nitrides that decompose during
PIMpowder injection molding is the comprehensive term sintering (AlN and Si3N4) and brittle materials that require
for forming inorganic engineered components in a plastic full density (WC-Co for example).
molding machine using thermoplastic binders and inert Pressure controlthe final phase of molding where the gate
powders, with subsets including metal powder injection is not frozen and the quantity of feedstock in the die cavity
molding (MIM), ceramic powder injection molding (CIM), is controlled by the pressure, thereby ensuring uniform
and cemented carbide powder injection molding (CCIM). weight and dimensions in the final MIM component.
Planetary mixerusually these have two offset mixing Proper control is achieved by monitoring pressure sensors
blades that rotate around individual shafts and the two in the die-cavity wall.
blades further rotate around a center axis. The net effect is Pseudoplastic flowa form of viscous flow where there is a
intermixing and stirring and shear to produce MIM feed- strain-rate sensitivity to the viscosity; generally it is the op-
stock, usually in a heated vessel. posite from dilatant flow, since the viscosity decreases with
Plasma debindinga new technology similar to catalytic higher strain rates.
debinding it involves a low-pressure plasma for depoly- Pycnometera device for measuring the theoretical density
merization of the binder from the sample surface inward of a loose powder or preform. Helium and granular fluids
to avoid heating and softening for distortion control. are commonly employed to encase the powder or compo-
Plunger moldinga hydraulic plunger is used to push nent for determination of true volume, and when coupled
molten MIM feedstock into a die cavity, where motion and with an independent mass determination leads to a density.
pressure are controlled by the applied hydraulic pressure. PZTpiezoelectric ceramic based on oxides of lead (Pb), zir-
Pneumatic moldinguse of an air-pressure head over conium (Zr), and titanium (Ti) used for ultrasonics, sen-
molten feedstock to push it into a die cavity. This is the sors, sonar, and various signal-propagation or collection
lowest-cost and least-precise form of MIM, yet is widely applications, based on shapes fabricated by MIM.
employed in the fabrication of large ceramic structures. Rapid prototypeseveral techniques emerging that bypass
Polydisperseimplies a broad powder-size distribution, cov- a need for tooling by using computer-controlled freeform
ering a wide range of particle sizes with no clear mode size. fabrication to generate a first MIM green body, typically
Pore sizethe size of the holes or voids between powder par- via building the feedstock into a three-dimensional object
ticles, often measured by microstructure quantification, from a stack of closely spaced laminates.
mercury porosimetry (open pores only), or other tests that Reciprocating screwinjection molding with a screw lo-
include gas flow, gas condensation, and polymer intrusion. cated in a heated barrel. During compression, metering,
176
Table of Contents
Chapter 9: Supplemental Information
melting, and forward advance of the feedstock, the screw nitride, with liquid-phase formers such as alumina and yt-
is turning, while during mold filling it becomes a plunger tria. These are very strong and lightweight materials with
to quickly fill the cavity. exceptional strength and reasonably high-temperature
Reduced powdermetal powder produced by the chemical properties (up to 1,200C) useful in high-temperature aero-
reduction of a compound, most typically a metal oxide. space, diesel engine, and military systems.
Refractorya metal or ceramic having a high melting tem- Sigma mixera closed mixer consisting of two blades that
perature, usually over 1,700C. Example refractory metals look like the Greek capital letter sigma. These blades lift
are tungsten, molybdenum, rhenium, and zirconium, while and separate feedstock to form a mixture of moderate ho-
example refractory ceramics are alumina, zirconia, yttria, mogeneity. A major advantage for MIM is the small batch
and chromia. size, but with low homogeneity due to the poor shearing.
Rheologythe study of strength, and elastic, plastic, and vis- Silicon carbidethe man-made compound SiC that is used
cous flow of polymers and feedstocks used in MIM. for applications ranging from sandpaper to high-tempera-
Runnera portion of the feed path for filling an injection ture heating elements. It has excellent high-temperature
molding die; the runner is between the sprue and gate, the strength and oxidation resistance and is formed by MIM
latter being the inlet to the actual die cavity. when complex shapes are required without post-sintering
Screwthe key portion of an injection molding machine for machining. Most systems rely on boron and carbon addi-
metering and filling the die cavity. It has a taper along the tions to improve sintering.
length to compress the feedstock as it is metered for re- Silicon nitridecompositions based on the stoichiometric
moval of trapped air and a check ring at the tip to allow compound Si3N4; in MIM this compound is used with var-
forward plunging motion during mold filling. ious liquid-phase-sintering additives to form SiAlON com-
Screw moldinga reciprocating screw is used to mix, pres- positions.
surize, de-air, and convey molten feedstock to finally be Sink marka shallow surface cavity that forms during cool-
injected into a tool cavity. This is the most widely em- ing. It is usually indicative of underpacking prior to gate
ployed means of injection molding precise components. freezing and is corrected by a higher molding pressure.
Secondary operationsthose activities performed to Sinter hardencontrolled cooling from the sintering tem-
adjust dimensions or properties of a component after sin- perature designed to induce desirable transformations and
tering. Examples include heat treatment, coining, machin- microstructures, to avoid post-sintering heat treatments.
ing, electroplating, impregnation, and shot peening. Usually it is applied to ferrous alloys to control martensite
Segregationnonuniform distribution of ingredients, such as formation on cooling.
powder separation by size, shape, or density, or chemical Sinteringthe thermal process which bonds and densifies
separation in the microstructure of a solidified material. the molded powders. It increases the compact strength via
Sendustalloys of ironaluminumsilicon that have impor- diffusion or related atomic-level events. Most of the me-
tant combinations of hardness and magnetic properties suit- chanical, magnetic, or other properties of a MIM part are
able for magnetic card readers and other high-use magnetic developed in the sintering cycle.
applications. The alloy hardness is such that MIM is the Sintering diagrama process map showing the interaction
most credible manufacturing route. of the key variables with respect to the densification of a
Setterthe tray or shaped substrate for support of MIM com- powder. The variables include particle size, grain size, tem-
pacts during debinding and sintering. perature, time, and applied pressure.
Shear ratea measure of the rate feedstock is deformed and Solids loadingthe relative powder volume in feedstock de-
forced to flow into a die cavity. It has units of inverse sec- signed for binder-assisted shaping. Common injection
onds. Actually it is the shear-strain rate, representing the molding feedstocks have solids loadings near 0.6 or 60 v/o
sheared change in length divided by the original length in powder.
unit time. For MIM, shear rates of several thousand per Solvent debindingthe extraction of the binder or some
second are normal, meaning the slug of feedstock is portion of the binder by leaching into a solvent. This is per-
quickly deformed to fill out the cavity. formed by immersion or exposure to solvent vapors. The
Shot sizea historical basis for sizing molding machines. solvent might be flammable, toxic, or simply water or
This is usually given in terms of the amount of polystyrene ethanol.
that can be moved through the nozzle in a single forward Speed controlthat initial portion of the molding stroke
motion of the screw or plunger. where feedstock is flowing to fill the mold cavity and no
Shrinkagea decrease in dimensions of a compact which feedback signals are available; thus the molding machine
occurs during sintering, usually 15% or up to 25% in a few is simply controlled by the position of the screw during its
cases and as low as 2% with multimodal powders. forward motion.
SiAlONa group of commercial alloys based on silicon Spherical powderpowder with a uniform spherical shape
177
Table of Contents
Chapter 9: Supplemental Information
178
Table of Contents
Chapter 9: Supplemental Information
waves for detection of cracks or other defects. The most Youngs modulusthe elastic modulus expresses the propor-
common procedure uses through or back transmission tionality between stress and strain during elastic deforma-
waves out of the die cavity, but new technology allows tion prior to permanent deformation; it is calculated
guided wave inspection in the die cavity during molding. assuming linear behavior by dividing the applied stress by
Vacuum debindinga variant of thermal debinding that the strain and for metals such as steel can range near 200
gives extraction of binder by progressive heating in a vac- GPa (30,000,000 psi).
uum to distill the molten binder out of the powder compact. YTZyttriatoughened zirconia, where a few percent of
Usually vacuum debinding is performed with a gas inlet to Y2O3 is added to ZrO2 to help in phase stabilization such
drag escaping binder to the pumps. that a phase transformation occurs in the zirconia under se-
Ventthe thin channel in a tool set that allows air to escape vere stress, such as at a propagating crack tip, to block frac-
while molten feedstock enters through the gate. It is located ture.
opposite from the gate and should not be sealed until late Zeta potentialan electrochemical measure of the surface
in the molding cycle to avoid trapped air pockets in the charge or surface chemistry of a fine powder in a suspen-
component. sion.
Viscositythe resistance to flow of a feedstock or polymer. Zirconiaa stoichiometric compound of zirconium and oxy-
Formally it is the proportionality between the stress and gen in the ratio ZrO2. It is often phase stabilized by adding
applied shear-strain rate. A high viscosity makes molding other ceramics. It is widely employed as a MIM material
difficult. A low viscosity is easier to mold, but there are because of its exceptional wear resistance and difficulty in
often difficulties with flashing. grinding, and is colored to form a wide variety of semi-
Water-atomized powderirregular or ligamental particles precious jewelry items.
formed by impacting a molten metal stream with high-
pressure water. NOMENCLATURE
Water-based bindermostly gelation binders that become As with words, and consistent definitions, it is also impor-
rigid after molding by the formation of a long-range mol- tant to establish consistent symbols, especially since engineers
ecule with water in the structure. Other forms rely on water often speak in mathematical terms. It is frustrating to have
saturated with silicates that form glasses on freezing. symbolic information provided without a guide as to how the
Water-soluble bindera class of binders that rely on poly- symbols are defined. It is unclear why the physics community
mers that are water soluble so solvent debinding takes uses the theta symbol for time when the rest of us use t. Any-
place in water. Most common are polyethylene oxide, how, here is my pass at a consistent nomenclature for MIM
polyvinyl alcohol, and polyethylene glycol. giving commonly used mathematical symbols:
Waxpolymera general term for polymers used as binders = numerical constant in calculations, dimensionless
that consist of low-molecular-weight polymers and wax = solid cone angle, degree
fillers. The polymers are usually simple molecules such as = strain or fractional porosity (sometimes given as %)
polyethylene, polypropylene, or co-polymers such as eth- f = elongation at fracture, a strain, usually given as %
ylene vinyl acetate. These are the classic binders used in = viscosity, Pa-s
MIM processing. = surface energy, J/m2
Weld linea linear defect or mark on a compact surface = thermal conductivity, W/(m K)
where two portions were bonded, either in molding or sub- = mean free separation or mean free path, m
sequently, such as in sinter bonding. = density, g/cm3 or kg/m3
Weibull modulusa statistical parameter associated with the = stress or strength, Pa; typical units here are MPa or
spread in fracture strengths for ceramics. A high Weibull 106 Pa
modulus indicates a narrow spread in strength and few = shear stress, Pa; typical units here are MPa or 106 Pa
manufacturing defects. = friction coefficient, dimensionless
Wicking debindingthe most popular means of binder ex- max = magnetic permeability, dimensionless
traction, especially for wax-based binders, where the com- = angle, degree
pact is packed in wicking powder and heated to a = angle, degree
temperature where the wax melts. The molten wax flows = solids loading, dimensionless or often given as
by capillary action into the surrounding packing powder volume percentage
and leaves the compact supported as polymer is extracted. = 3.14159
Yield strengththe end of the linear elastic region for most = adhesion force, frequency, or other atomic-level
metals. The yield strength is the highest stress a structure parameter
can endure without permanent deformation and should be = atomic volume, m3/atom or m3/mol
the maximum stress used in engineering designs. A = area, m2
179
Table of Contents
Chapter 9: Supplemental Information
180
Table of Contents
Chapter 9: Supplemental Information
181
Table of Contents
Chapter 9: Supplemental Information
182
Table of Contents
Chapter 9: Supplemental Information
evidence suggests that there is no single manufacturer or vendors for support. A plastic molder in Canada used a net-
design that is necessarily best. work of small contracts to seed critical developments during
start-up. However, to become established at more than, say,
There is easy money to be made in MIM. $5 million in annual sales requires mastery of three technolo-
With its maturation, MIM has largely remained a healthy giesmaterials engineering, manufacturing engineering, and
technology that continues to grow by satisfying more sophis- plastic molding, but you can start with only one or two.
ticated customers. But most of the firms with favorable finan-
cial positions tend to have much practical experience, and Powder costs will continue to fall.
almost all of the leaders are well over 10 years old. These As MIM grows, some of the powders (especially preal-
leaders have established manufacturing systems which are loyed stainless steel) have declined in cost as consumption
benchmarks for all net-shape manufacturing. To assume that has increased. But in all cases, consumption has increased
a late entry will jump into this class is not realistic. Yes, there faster than the cost reduction. In other words, if powder sales
is success if you work long and hard, but MIM is not a route (dollars) have increased 20% per year then powder shipments
to quick riches. (tons) went up 40% per year. Along with improved process
yields, the increasing tonnage generated a price decrease. For
Only a few firms understand the technology. some alloy chemistries this sort of volume growth is not going
There are more than 350 firms, several university pro- to be sustained.
grams, and a dozen research institutions supporting the tech-
nology developments in MIM. None of these has a lock on Component producers that purchase the same feedstock
the technology. The basic concepts are well documented in allow for movement of an order between sites.
books, and many of the suppliers will gladly share technology Clearly, a trend has been to rely on purchased feedstock,
to help create new customers. As is often known as Prados but that does not cure the need for similar molding practices,
Principle, 20% of the firms have 80% of the sales, but it is not processing equipment, and secondary operations. In general,
wise to assume any single firm is smarter or more knowledge- tool design is highly variable and customized to the shop and
able than another. The field has many successes, failures, and molding machine. Hence, moving an order to a second site
much cross-fertilization, so there is no single pathway or tech- might encounter problems beyond the feedstock, so there is
nology that will give success. no assurance of success along these lines.
A license is required to practice MIM. Newer binders enable cost-effective production of large
Some of the first commercial developments in powder in- components.
jection molding date from the 1930s and 40s, so long ago we The binder is not a barrier to molding large components.
have trouble reconstructing the history. In the 1960s, Corning Large stainless steel components were formed from in the
used ceramic molding to form tableware, a technology that 1980s using MIM, ranging up to 12 kg (26 lb.), using a tradi-
survives in the ceramic casting core business. In MIM, the tional waxpolymer binder. Today large ceramic components
original patents were issued in the 1970s. Thus, with the tech- are in production at several sites. There are problems with
nology aging this much, many of the suppliers provide an ex- large components, but more significant are the economic bar-
cellent support base for start-ups. The generic technology riers. In metals, as the mass increases the raw material cost
works well and most of the ingredients and equipment can be difference (powder vs. casting cost) leaves plenty of margin
ordered with help from the equipment manufacturers. So, if a for machining after casting. Thus, changing the binder will
company wants to have its hand held during the start-up phase not solve the critical powder-cost issue. Indeed, several eco-
using a license, that is fine, but to think there is a proprietary nomic problems arise with large MIM components; besides
or special technology that requires a license does not reflect powder cost there are penalties because the molding machines
reality. are larger, molding cycles are slower, and debinding times are
longer. Accordingly, MIM is less cost effective as size in-
You need a materials scientist to succeed. creases, independent of the binder.
This is a myth that is close to reality in some cases. Several
plastic injection molders have successfully entered MIM, but Low-pressure molding is less costly.
at times they struggle with controlling impurities, optimizing Low-pressure MIM machines are used to reduce machine
properties, selecting proper heat treatments, and other basic cost and tool wear. They are used by about 15% of the com-
aspects of materials engineering. These can be mastered with- bined metals and ceramics industry. The manufacturing cost
out full-time staff. One start-up used a skilled metallurgist who is now lower with low-pressure molding, since these options
was staying home with young children on an as-needed basis. lack automation and usually only fill a single cavity. So with
Another plastic firm initially relied on feedstock and furnace a lower purchase price (capital cost) there is a burden of a
183
Table of Contents
Chapter 9: Supplemental Information
higher operating cost. For example, in a comparison across new alloys, processing options, or test parameters. Listed
the industry, low-pressure molding shops show lower sales below are some of the ones seen in MIM.
per employee. More important, without a high molding pres-
sure the components tend to have more internal defects. For Testing Powders
surface features (such as in sandblast nozzles and watchcases) Analysis by Microscopical Methods for Particle Size
there is no problem. Distribution of Particulate Substances of Subsieve Size
ASTM E20
Continuous sintering in hydrogen is a MIM evolution. Apparent Density of Non-Free Flowing Metal Powders Using
The use of hydrogen sintering in a pusher furnace was ap- the Hall ApparatusMPIF 04, ASTM B212, ISO 3923/1
plied to refractory metals and stainless steel from prior to the Apparent Density of Non-Free Flowing Metal Powders Using
1940s. It was widely used for that application at many sites the Carney ApparatusMPIF 28, ASTM B417, ISO
prior to the first commercial MIM use in 1985. Today, con- 3923/1
tinuous furnaces constitute about a quarter of the installed sin- Apparent Density of Metal Powders Using the Arnold
tering capacity in MIM and prove most productive for MeterMPIF 48, ASTM B703
long-running components. Thus, they are more common in Apparent Density of Refractory Metals and Compounds by
large Asian shops fabricating consumer products, cellphone Scott VolumeterASTM B329
parts, and watchcases. Average Particle Size of Metal Powders Using the Fisher
Subsieve SizerMPIF 32, ASTM B330, ASTM C72
The MIM technology is only capable of holding (see also ISO 10070)
tolerances at 0.5%. Flow Rate of Free-Flowing Metal Powders Using the Hall
High tolerances are always possible using post-sintering ApparatusMPIF 03, ASTM B213, ISO 4490
machining. The real challenge is in forming tight tolerances Oil Absorption of Pigments by Spatula Rub-outASTM
without coining or machining. The as-sintered tolerances de- D281
pend on the age of the technology, recognizing newer opera- Particle Size Distribution of Metal Powders and Related
tions employing better binders, molders, and process control Compounds by Light ScatteringASTM B822
during debinding and sintering. Essentially all of the industry Particle Size Distribution of Refractory Metals and
can hold dimensional scatter to within 0.3% (one standard Compounds by X-ray Monitoring of Gravity Sedimenta-
deviation), but a few have reached a coefficient of variations tionASTM B761
as small as 0.02% with newer controls, such as closed-loop- Particle Size Distribution of Tungsten Metal Powder by
feedback control during molding. A few products are in pro- TurbidimetryASTM B430
duction with absolute tolerances of 2 to 5 m. Sampling Finished Lots of Metal PowdersMPIF 01, ASTM
B215, ISO 3954
Metal powder injection molding is just another form of Tap Density of Metal PowdersMPIF 46, ASTM B527, ISO
powder metallurgy. 3953
Although a few companies practice both MIM and tradi-
tional die compaction and sintering (what is generally implied Testing of Sintered Products
by the term powder metallurgy), largely the two have little in Carburized Case Hardness and Case Depth of Sintered
common other than relying on metal powders. MIM powders PartsMPIF 37, ISO 4507
are much smaller and round if not spherical, and sintering tem- Density Determination of MIM Components (Gas Pycnome-
peratures are higher such that the final density and performance try)MPIF 63
are higher. Determination of Charpy Impact Energy of Unnotched Metal
Injection Molded Test SpecimensMPIF 59
STANDARDS Determination of Corrosion Resistance of MIM Grades of
The standards used in MIM usually start at the national Stainless Steel Immersed in 2% Sulfuric Acid Solution
level, say, in Japan or the U.S., and then become adopted in MPIF 62
the powder community, and spread to national bodies, even- Green Strength of Compacted Metal Powder Specimens
tually becoming part of the ISO standards. For North America, MPIF 15, ASTM B312, ISO 3995
the example would be the Metal Powder Industries Federation Impact Strength of Sintered Metal Powder SpecimensMPIF
(MPIF) as the regional trade organization, which then feeds 40, ISO 5754
its standards into the American Society for Testing Materials Material Standards for Metal Injection Molded PartsMPIF
(ASTM), which participates in the International Standards 35
Organization (ISO) global standards. The standards are Nickel-Iron Powder Metallurgy Soft Magnetic Alloys
reviewed and updated periodically, and expanded to handle ASTM A904
184
Table of Contents
Chapter 9: Supplemental Information
Nickel-Silver Sintered Metal Powder Structural Parts European Powder Metallurgy Association, Shrewsbury, UK
ASTM B458 www.epma.com
Sintered Austenitic Stainless Steel Structural PartsASTM Japan Powder Metallurgy Association, Tokyo, Japan
B525 www.jpma.gr.jp
Sintered Metal Materials and Hardmetals Determination of Metal Powder Industries Federation, Princeton, NJ, USA
Young's ModulusASTM E111, ISO 3312 www.mpif.org
Sintered Metal Materials, Excluding Hardmetals, Determina-
tion of Transverse Rupture StrengthISO 3928 In addition, a new Asian Powder Metallurgy Association
Soft Magnetic Iron Fabricated by Powder Metallurgy is forming to combine the separate organizations in Japan, Tai-
ASTM A811 wan, South Korea, and China. Until that is fully commis-
Standard Specification for Metal Injection Molding (MIM) sioned, each country has its individual trade organization.
Ferrous PartsASTM 883 Likewise, many other countries have small efforts, for exam-
Standard Test Method for Linear Thermal Expansion of Solid ple, in Turkey, South Africa, Russia, and Thailand.
Materials with a Push-Rod DilatometerASTM E228
Transverse Rupture Strength of Sintered Metal Powder Test RESOURCES
SpecimensMPIF 41, ASTM B528, ISO 3325 International Comparison of Manufacturing Productivity and
Unit Labor Cost Trends, 2001, Report USDL 01-543, Bureau
PATENT SOURCES of Labor Statistics, United States Department of Labor, Wash-
There are a large number of patents in MIM, dating from ington, DC, 26 September 2002.
about the mid-1970s. The easy means to sort through the U.S.
M. Ashby and D. Cebon, Estimating Process Cost: Cost
Patents is via the Web site www.uspto.gov by entering the
Modelling for Selection, Version 4, Cambridge Engineering
quick search for patents zone (PTO = patent and trademark
Selector, Granta Design, Cambridge, UK, 2002.
office). You can search by title, keyword, inventor, year, or
other features. Be careful in relying on MIM since this has G. Boothroyd, P. Dewhurst, W. Knight, Product Design for
many other definitions besides metal powder injection Manufacture and Assembly, second edition, 2002, Marcel
molding. Dekker, New York, NY, pp. 359-379.
Once a patent is identified, a good copy is available as a pdf
R.M. German, Engineering Economics of Powder Injection
file at www.pat2pdf.org as a free download. Happy hunting.
Molding Component Production: Part 1, Tool Costing, P/M
Science and Technology Briefs, 2003, vol. 5, no. 2, pp. 511.
CONFERENCES AND PUBLICATIONS
Every two years there is a World Congress of Powder Met- R.M. German, Engineering Economics of Powder Injection
allurgy, rotated between Asia, Europe, and North America. Molding Component Production: Part II, Feedstock Costs,
This is the typical forum for showing the latest developments P/M Science and Technology Briefs, 2003, vol. 5, no. 3, pp.
in MIM. Additionally, each regional trade organization (listed 1116.
below) has an annual conference and MIM is usually well rep-
R.M. German, Engineering Economics of Powder Injection
resented at these as well. Quite often, to foster more attention
Molding Component Production: Part III, Production Costs,
on MIM, a short course and tabletop exhibition, standards
P/M Science and Technology Briefs, 2003, vol. 5, no. 4, pp.
committee meeting, technical presentations, and company ca-
1422.
pability briefing is organized, usually in a desirable location,
such as Orlando or Los Angeles. While the World Congress R.M. German, Engineering Economics of Powder Injection
will often draw 1,500 participants, the conferences are about Molding Component Production: Part IV, Price Sensitivity,
half that, and the MIM topical meetings attract typically 120 P/M Science and Technology Briefs, 2004, vol. 6, no. 1, pp.
participants. 510.
R.M. German and D. Blaine, Production Cost Sensitivity
TRADE AND PROFESSIONAL ORGANIZATIONS
Analysis for Metal Powder Injection Molding, Advances in
There is no trade association specifically for MIM. In the
Powder Metallurgy and Particulate Materials2004, Metal
various geographic regions, various powder metallurgy mar-
Powder Industries Federation, Princeton, NJ, part 4, pp. 110.
keting and trade associations operate and they embrace MIM
as part of their activities, including coverage at annual confer- R.M. German and A. Bose, Injection Molding of Metals and
ences, articles in the magazines, and training courses. The trade Ceramics, 1997, Metal Powder Industries Federation, Prince-
associations focused on powder metallurgy are as follows: ton, NJ.
185
Table of Contents
Chapter 9: Supplemental Information
186
Table of Contents
INDEX
A bronze statue, 23, 33
acceptance criteria, 67 brown, 9
accuracy, 60 burner cover, 109
adjustable vane, 109 burnout cycle, 19
adjustment ring, 109 business-machine applications, 111
aerospace applications, 106107, 146 butterfly-valve plate, 107
aesthetics, 125
air conditioner fittings, 122 C
airbag pivot, 108 cantilever, 46
alumina, 72, 76, 83, 85, 171 capital-equipment investment, 138
aluminum nitride, 87 captive molders, 136
aluminum, 73, 74, 80 carbon control, 171
annual sales history, 138139 carbo-nitrides, 24
apparent density, 12, 73, 171 carbonyl powder, 12, 171
applications, 9, 25, 105, 139, 142, 145 carburization, 27
arrow head, 130 casting comparison, 102
arthroscopic connector, 125 catalytic debinding, 19, 172
arthroscopic implant, 127 cavity design, 51, 57
articulator gear, 126 cavity transducer, 56, 64
artistic items, 131132 cellphone applications, 1, 111113
assembly, 42 cellphone cover, 112
atmosphere sintering, 20 cellphone knuckle, 101, 168
atmosphere-stabilized compositions, 90 cellulose, 13
audio jack mount, 112 cemented carbide, 60, 47, 72, 73, 79, 83, 88, 114, 172
automation, 146 centrifuge component, 37
automotive applications, 108111, 142 ceramics, 49, 64, 8287
automotive magnetic sensor, 56 cermet, 47, 89, 172
automotive ratchet, 108 Charpy testing, 8384
check ring, 17
B chemistry specification, 74
backbone polymer, 13, 18 chess set, 132
balanced runner, 53 chopstick rest, 132
barrel, 15, 171 chromium evaporation, 79
batch mixer, 13 clamping force, 15, 172
batch sintering, 21 closed features, 40
batch size effects, 159161 closed pore, 172
bi-material, 22 closed-loop feedback control, 74, 160
binder, 9, 13, 171 coarsening, 172
binder contraction, 57 cobaltchromium, 79, 80
biocompatibility, 75, 77, 90 coefficient of variation, 60, 172
biopsy tip, 23, 126 coining, 18, 27, 172
birdshot, 120 cold deformation, 27
blemishes, 42, 50 cold plate, 113
bonded materials, 91 color, 8586
brake fitting, 110 comparison to powder metallurgy, 3
brazing, 27 competitive technologies, 104, 139140
bronze filter, 39 complexity, 32, 42, 64, 99, 140, 152
187
Table of Contents
Index
188
Table of Contents
Index
189
Table of Contents
Index
190
Table of Contents
Index
191
Table of Contents
Index
salt-spray testing, 78 stainless 17-4 PH, 20, 27, 48, 74, 76, 77, 78, 79, 80, 82, 84,
sand wedge, 101, 155 85, 88, 94
scissors, 37 stainless 304L, 80
screw, 15, 177 stainless 316L, 61, 66, 74, 77, 79, 80, 82, 84, 85, 94
screw position, 18 stainless 318, 80
screwdriver holder, 7 stainless 410, 80, 84
sculpture, 32 stainless 414, 80
seal ring, 106 stainless 420, 58, 80, 84, 94
secondary operations, 27, 177 stainless 430L, 79, 80, 84, 94
segregation, 177 stainless 440C, 58, 80, 88
self-mixing, 14 stainless 630, 48. See stainless 17-4 PH
sensor applications, 124 stainless PANACEA, 80
serrations, 6, 45 stainless steel, 68, 73, 74, 7788, 178
service failure, 58 stainless steel microstructure, 4
setter, 63, 171 stainless steel powder, 12
shape complexity, 33 standard deviation, 69
sharp corners, 41 standard process, 29
shear-strain rate, 14 standardized chemistries, 73
shear thinning, 14 statistical process control, 68
short shot, 18 stearic acid, 13
shot size, 177 steel, 73, 74, 76, 77, 80, 94
shower safety valve, 121 steel 2200, 74, 77, 79, 80, 84, 85, 94
shrinkage, 177 steel 2700, 74, 79, 80, 84, 85, 94
shrinkage factor, 15 steel 4140, 74, 80, 84, 94
silica, 76 steel 4340, 74, 80, 85
silicon carbide, 72, 85, 90, 177 steel 4605, 74, 80, 84, 94
silicon nitride, 72, 83, 85, 177 steel 4620, 74
sink marks, 17, 28, 54, 177 steel 4640, 80, 84
sintered density, 52 steel 4650. See steel 4605
sintering, 9, 20, 177 steel properties, 4
sintering costs, 163164 steering clevis, 110
sintering shrinkage, 3, 14, 20, 52 stents, 145
sintering simulation, 28, 65 stepped mold closure, 54
sintering substrate, 20 stereolithography, 59
size range, 36 sticking, 54
slenderness, 36 stiffeners, 39
slides, 15, 55 strength, 5, 47, 8081, 8384, 87, 89
slots, 41 structural properties, 8085
small details, 41 summary characteristics, 38
small-scale production mode, 137 superalloy, 85, 106, 178
soft magnetic alloys, 68, 124 superalloy Hastelloy X, 79
soft tooling, 33, 58 superalloy Inconel 718, 47, 80, 85, 106
software packages, 65 superalloy Udimet, 81
solenoid valve body, 124 surface carburization, 27
solids loading, 13, 52, 177 surface finish, 26, 35, 42, 85, 99
solvent debinding, 19, 177 surface roughness, 62, 63
soup spoon, 35 surface treatments, 27
splines, 39 surgical stainless steel. See stainless 17-4 PH
sporting applications, 129130 swirler, 106
spray nozzles, 122
sprue bushing, 56 T
sprue, 178 tableware, 35
stabilizer fin, 115 tap density, 178
192
Table of Contents
Index
193
Table of Contents
Index
194
cover_Layout 1 4/7/2011 2:05 PM Page 1
Metal
Injection
Molding
GERMAN
Randall M. German
No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including
photocopying, recording, scanning, or by any information storage and retrieval system, without the written permission of the
Metal Powder Industries Federation, 105 College Road East, Princeton, N.J. 08540. The information in this book and the
identification of design decision processes, property mapping, and material selection techniques, as well as cost and management
metrics, is proprietary intellectual property of the author and is not for distribution, reproduction, or use beyond that of a single
purchase.
Table of Contents
To Savannah
Table of Contents
TABLE OF CONTENTS
PREFACE...........................................................................ix Design Support ...................................................................49
Resources ............................................................................50
ACKNOWLEDGMENTS.................................................xi
Chapter 4. MANUFACTURING CONSIDERATIONS ...51
AUTHOR BACKGROUND ...........................................xiii Mold Attributes and Limitations.........................................51
Cavity Design.................................................................51
Chapter 1. INTRODUCTION ...........................................1 Parting Line....................................................................53
OverviewWhat is MIM? ...................................................1 Sprue, Runner, and Gate ................................................53
Process Essentials .................................................................2 Vent ...............................................................................54
Where MIM Differs from PM ..............................................3 Temperature Control Passages .......................................55
Unique Aspects of MIM .......................................................6 Ejector ............................................................................55
Historical Evolution..............................................................6 Slides and Cores.............................................................55
Market and Technology Status..............................................8 Tool Motions.......................................................................56
Contrast with Other Technologies ........................................8 Tool-Cavity Design.............................................................57
Key Terms.............................................................................9 Novel Solutions...................................................................58
Resources ..............................................................................9 Blemishes from Tooling......................................................59
Geometric Feature Combinations .......................................60
Chapter 2. PROCESS OPTIONS ....................................11 Dimensional Tolerances......................................................60
Process Overview................................................................11 Tolerance Approaches.........................................................62
Background ....................................................................11 Process Yield.......................................................................64
FeedstockPowder Plus Binder ...................................12 Computer Simulation for Defect Avoidance.......................65
Tooling and Materials ....................................................14 Quality Plans.......................................................................66
Molding..........................................................................15 Inspection............................................................................68
Debinding.......................................................................19 Quality Tests...................................................................68
Sintering .........................................................................20 Quantitative Techniques.................................................69
Novel Options .....................................................................22 Nondestructive Techniques ............................................69
Attributes and Advantages ..................................................25 Resources ............................................................................69
Secondary Operations .........................................................27
Process Simulations ............................................................28 Chapter 5. MATERIAL PROPERTIES .........................71
Vendor Differences .............................................................28 Available Materials .............................................................71
Resources ............................................................................29 Powders..........................................................................73
Feedstock .......................................................................73
Chapter 3. DESIGN PRINCIPLES.................................31 Standardized Chemistries...............................................73
Design Cycles .....................................................................31 Cost Attributes ....................................................................74
Recognizing Good Candidates............................................31 Materials Selection .............................................................75
Qualitative Criteria.........................................................31 Key Classifications ............................................................76
Quantitative Criteria.......................................................32 Electrical and Magnetic .................................................76
Geometric Details ...............................................................34 Corrosion and Biocompatibility.....................................77
Effective Density Concept .............................................34 Inertial Properties...........................................................79
Surface Finish ................................................................35 Structural Properties.......................................................81
Size Range .....................................................................36 Tensile Properties of Metallic Materials ...................81
Mass Range ....................................................................37 Rupture Strength of Ceramic Materials ....................82
Geometric Summary ......................................................38 Impact, Fracture, and Fatigue Properties ..................83
Features...............................................................................39 Elevated-Temperature Properties ..............................85
Flat Face.........................................................................39 Optical Properties...........................................................85
Closed Features ..............................................................40 Thermal Properties.........................................................87
Small Details ..................................................................41 Wear ...............................................................................88
Holes, Slots, and Undercuts ...........................................41 Sintered Microstructures.....................................................88
Imprints, Protrusions, Ribs ............................................45 Composite Materials ...........................................................89
Threads...........................................................................46 High-Elastic-Modulus Cermets .....................................89
Materials .............................................................................47 Heterogeneous Microstructures .....................................89
Properties ............................................................................47 Atmosphere-Stabilized Compositions ...........................90
vii
Table of Contents
viii
Table of Contents
PREFACE
The combination of powder metallurgy and plastic molding results in an exciting fabri-
cation technology, termed metal powder injection molding (MIM). It changes the design
rules, allowing more complex components with more features in a high-production-volume
technology.
This design guide is constructed to help the reader take advantage of MIM. The metal
powder injection molding process has a related ceramic cousin CIM. Together the variants
of MIM, CIM, and similar technologies for carbides and composites, are known as PIM or
powder injection molding. These forming routes allow designers to combine the cost-
effective shaping attributes of plastic injection molding with the performance and engineer-
ing properties associated with durable and known engineering materials. About 80% of
powder injection molding is in the form of metals, so generally MIM and PIM are used
nearly interchangeably.
This guide organizes the facts on the process options, materials, properties, design
features, manufacturing tolerances, applications, markets, and costs to enable the design
community to best utilize this net-shape production route.
Consideration of MIM during the design process leverages its unique attributes to ease
production, lower costs, and increase process yield. This is because small design changes
can greatly enhance process yield and avoid costly secondary operations.
Today, MIM involves a large cadre of highly skilled and competitive manufacturing
firms. Although MIM is not new, exceeding production rates over a billion of parts per year,
guides on how to best use the technology are not well distributed. This design guide helps
rectify the situation, by identifying good design candidates to guide the user to best practices
in materials, properties, features, sizes, shapes, tolerances, and costs.
Randall M. German
ix
Table of Contents
Table of Contents
ACKNOWLEDGMENTS
A few individuals helped tremendously with the background Tracy Potter (Advanced Powder Products, Philipsburg,
to this book. Robert Cornwall provided insight on how to Pennsylvania, USA),
collect market statistics and assess business trends. Animesh Julian Thomas (Springfield Munitions, Kersey,
Bose helped organize application and process details Pennsylvania, USA),
fundamental to this book. Kay Leong Kim, Lye King Tan, and Nick Williams (Inovar Communications, Shrewsbury,
Chee Tian Yeo provided insight to the production environ- UK),
ment. Karl Hens made sure I was fully grounded in molding Zulkefli Bin Zainal (Advanced Materials Technologies,
technologies. Seong Jin Park likewise guided the computer Singapore).
simulation gains. My son Eric German (Mitchell Silberberg & Knupp, Santa
Several former students, sponsors, and industry friends Monica, California, USA) provided legal guidance.
guided my efforts by providing pictures, samples, data, and I am especially thankful to Julian Thomas and to Rick
comments. Besides the individuals acknowledged above, Sharbaugh (Spring Mills Studio, Spring Mills, Pennsylvania,
notice is given to the special help provided by the following USA) for their help in preparing many of the images.
individuals: Besides the individuals mentioned above, thanks go to the
James Adams (Metal Powder Industries Federation, following individuals for their help in providing input data
Princeton, New Jersey, USA), relevant to the text: Antonyraj Arockiasamy (Mississippi State
Sundar Atre (Oregon State University, Corvallis, University), Abdulhakim Belhadjhamida (Deloro Stellite),
Oregon, USA), Anthony Fanelli (Honeywell), Anthony Griffo (Smith Inter-
Eric Baril (Industrial Materials Institute, National national), Anish Upadhyaya (Indian Institute of Technology
Research Council, Boucherville, Quebec, Canada), Kanpur), Barbara Lograsso (Michigan Technological Univer-
Paul Busch (Injectamax, Escondido, California, USA), sity), Barry Rabin (Sage Technology and Development),
Qiaoling Chen (Hunan Injection High Technology, Bo Li (Yuelong Supefine Metal Materials), Jim Neill
Changsha, Hunan, China), (CM Furnaces), Debby Blaine (Stellenbosch University),
Zhahan Chen (Beijing Jarain Powder Injection Thomas Gladden (SA Fabriques d'Ebauches), Haorong Zhang
Technology, Beijing, China), (NetShape Technologies), Johnny Bruhn (SECO Tool),
Santosh Das (Polymer Technologies, Clifton, Jonathan Wroe (European Powder Metallurgy Association),
New Jersey, USA), Jonathon Fischer (MoldMaster), Joseph Rose (Penn State
Kim Gon (PIM Korea, Gyoungasn-city, Gyoungbuk, University), Louis Campbell (Eaton), Mark Janney (Porvair
Korea), Advanced Materials), Maryann Wright (Remington Arms),
Uwe Haupt (Arburg, Lossburg, Germany), Matthew Bulger (NetShape Technologies), Mu-Jen Yang (Air
Kuen Shyang Hwang (National Taiwan University, Force Research Laboratory), Neal Myers (Kennametal),
Taipei, Taiwan), Pavan Suri (Heraeus), Rajiv Tandon (Magnesium Elektron),
John Johnson (Engineered Products, Huntsville, Ala- Rudolf Zauner (Verbund-Austrian Renewable Power), Shin
bama, USA), Tian Lin (National Taiwan University of Science and Tech-
Stefan Joens (Elnik Systems, Cedar Grove, New Jersey, nology), Stanley Zalkind (TekForm Management), and
USA), Yunxin Wu (Tsinghua University).
Yoshiyuki Kato (Epson-Atmix, Aomori-ken, Japan), Draft chapters were reviewed by several volunteers who
Chhattar Kucheria (Ortho Organizers, Carlsbad, suggested means to improve the presentation. My special
California, USA), thanks go to Aditya Bothate, Wei Li, Kyle Katzmiller, and
Young-Sam Kwon (CetaTech, Sacheon, Gyeongam, Ridvan Yamanoglu for taking the time to read the manuscript,
Korea), making helpful suggestions, and supporting the timely com-
Jimmy Lu (Honeywell, Tempe, Arizona, USA), pletion of the project.
Seong Jin Park (Pohang University of Science and
Technology, Pohang, Korea),
xi
Table of Contents
Table of Contents
AUTHOR BACKGROUND
Rand German is Professor and Associate Dean of Engineering
at San Diego State University. Previously he was a chaired pro-
fessor at three universities where he conducted academic re-
search on automotive, powder metallurgy, and materials
engineering problems; Rand is an emeritus professor at the
Pennsylvania State University. He has been involved in the start-
up of a dozen firms, some of which have successfully commer-
cialized core MIM technologies. He obtained his doctoral
degree in Materials Science at the University of California
Davis, and obtained his MS and BS degrees from The Ohio
State University and San Jose State University. In addition, he
has taken intensive management training at the Hartford Grad-
uate Center and Harvard University.
He is a Fellow of ASM International and Fellow of APMI International, and has awards
that include the Nanyang Professorship, Tesla Medal, Japan Institute for Materials Research
Lectureship, Penn State Engineering Society Premiere Research Award, University of Cal-
ifornia at Davis Distinguished Engineering Alumni Award, San Jose State University Award
of Distinction, Ohio State Distinguished Alumni Award, and Technical Development Award
from the Japan Institute of Metals.
Rand has more than 930 published articles, 15 books, 19 edited books, and 24 patents.
He chairs several annual short courses, workshops, and symposia; he started the annual
Powder Injection Molding Symposium in 1990. He is technical advisor or member of the
board of directors for several organizations ranging from start-up companies and venture
capital boards to research institutes and private companies. He is involved in a wide range
of research projects that include studies on spark sintering, hard materials consolidation, ti-
tanium by MIM, microminiature components, developments in superfine powders, and
tungsten processing. He and his wife Carol live in Del Mar, California, and they have two
sons and three grandchildren who all live in California.
xiii
Table of Contents