Research Summary Aluminum

Mechanism and Preventive Measures

for Die Soldering during Al Casting in
a Ferrous Mold
Sumanth Shankar and Diran Apelian

This work provides a comprehensive features, and coatings. This article

understanding of the reactions at the fer-
rous die/molten metal interface in a metal
mold casting operation. The literature
has shown that several important fac-
tors influence reactions at the ferrous
die/molten aluminum interface, including
temperature of the melt, temperature of
the die, alloy chemistry of the melt and die,
die surface engineering, topographical
discusses the effect of the more critical
factors on soldering, based on the
authors’investigations. In addition, based
on a mechanistic understanding of the
interface reactions between ferrous die
and molten aluminum, recommendations
are given for specific processing issues
to alleviate soldering during die casting
of aluminum alloys.
INTRODUCTION
To date, the mechanism of soldering
in metal mold casting has been hypoth-
esized as either an interface-controlled or
diffusion-controlled phenomenon. The
authors’ work at the Advanced Casting
Research Center (ACRC) extended to
the die surface under the soldered layer,
revealing how the intermetallic layers

a b c

Figure 1. A schematic of the mechanism of die soldering. (a) Initial attack of the grain boundaries by aluminum to loosen up the hard grains
and martensitic plates to cause pitting on the die surface. (b) Formation of the iron-aluminum intermetallic phases inside the pits and around
the broken grains close the die surface. (c) “Pyramid” growth of the ternary α-(Al,Fe,Si) phase on the pits over the η-Fe2Al5. In addition, the
pits expand laterally and in depth. Aluminum begins to stick after this layer structure is formed resulting in the beginning of soldering. (d)
Shows the growth of intermetallic layers and merging of neighboring pits. Molten aluminum encounters the die surface only through the cracks
and gaps present between adjacent pits. (e) Straightening out of the pits and closing of the gaps between the adjacent pits. The ratio of the
intermetallic layer thickness and the soldered aluminum is ~1:5. The reaction mechanism becomes very slow. Silicon is precipitated in the grain
boundaries of the η-Fe2Al5 phase and at the intersection boundaries between the two intermetallic phase layers.

2002 August • JOM 47

attach to the die at pits formed on the
die surface. These observations led to
the design of a set of diffusion-couple
experiments between molten aluminum
alloy and the ferrous die. The results
of the diffusion-couple experiments1
clearly show that soldering is a dif-
fusional process. When aluminum
comes in contact with the ferrous die
material, the iron and the aluminum
atoms diffuse into each other, resulting in
the formation of a series of intermetallic
phases over the die material. Initially,
iron and aluminum react with each
other to form binary iron-aluminum
intermetallic phases. Subsequently,
these phases react with the molten
aluminum to further form ternary iron-
aluminum-silicon intermetallic phases.1
Iron and aluminum have a great affinity
for each other, and the root cause for
soldering is the high-reaction kinetics
between iron and aluminum. Once the
initial binary and ternary intermetallic
phase layers are formed over the die
material, the aluminum sticks to the
die due to the abnormally low thermal
conductivity of the intermetallic phases,
and due to favorable interface energies
between the intermetallic layers and
aluminum.
Based on all the results, discussion of
the results, and conclusions arrived at in
the authors’ research on die soldering,2–6
this article presents a mechanistic theory
of die soldering. In addition, a specific
set of recommendations is made to
alleviate die soldering.
MECHANISM FOR DIE
SOLDERING
A thorough understanding of the prior
work on die soldering and the interface
reactions between ferrous die material
and aluminum has been presented.1,2,6
Subsequently, experiments were per-
formed and validated to propose a
detailed working mechanism 1,2 to
describe die soldering.
Figure 1 shows a schematic of the
proposed mechanism of soldering and
Figure 2 shows micrographs of soldered
samples taken during the various
stages of soldering from a variety
of sources—commercial die casting,
permanent mold casting, and laboratory
trials. The mechanism of soldering,
which has been validated through
these and many similar microstructural
observations, occurs in six stages. Each
of the six stages is further discussed
in the following.
In the first stage, the aluminum melt
repeatedly encounters the die surface at
each cast cycle. Tool steels are generally
heat treated by double tempering to a
hardness of about Rc48–Rc50. During
the casting process, the aluminum
melt attacks the softer regions of the
die surface—areas between the hard
martensitic plates and the carbide
particles. These are primarily the inter-
granular regions. When the aluminum
erodes the soft areas of the die surface,
it results in the formation of a primary
solid solution of iron by aluminum
dissolution. The solid-solution phase is
represented by α-Fe. This is illustrated
by the schematic shown in Figure 1a.
An actual grain boundary attack is
shown in Figure 2a. In this case, Figure
2a is a soldered sample obtained from
Cambridge Tools and Manufacturing,
Burlington, Massachusetts.
Once the grain boundaries and the
soft phases on the steel surface have
been attacked by aluminum, pits that are
hemispherical in shape form. The typical
high drag forces that are experienced
by the die surface eject the loosened
surface grains from the die, into the melt.
Figure 3 shows a sample micrograph to
illustrate the pitting mechanism on the
die surface after die surface erosion.
Next, the aluminum reacts with the
loosened surface grains, and, at the
surface of the pits, more iron-aluminum
binary phases are formed such as FeAl,
FeAl2, Fe2Al5, and FeAl3. The formation
of these successive layers of binary
compounds is due to the reaction of
each phase with a continuously renewed
molten aluminum and the diffusion of
the iron out of the steel surface. This
is shown in Figures 1b, 2b, and 2c.

a b c

d e f

Figure 2. Images showing the mechanism of soldering. The micrographs were obtained from soldered industrial die samples, from soldering
experiments at ACRC and samples from diffusion couple experiments. These micrographs show the corresponding features in various stages
illustrated in Figure 1. (a) Intergranular seepage of aluminum melt causes erosion of the grain boundaries on the die surface. (b) and (c) Initial
pitting and the formation of binary intermetallic phases. (d) growth of the initial pyramid shaped intermetallic layers. (e) and (f) Lateral expansion of
the pyramid shaped intermetallic phase layers. (g) and (h) Closing up of the pits and flattening of the die surface.

48 JOM • August 2002

 emanating from the pits. Figure 4 shows
soldered aluminum adhering onto the
pyramids, as well.
The last reaction product between
iron and the aluminum-alloy melt is the
ternary α-(Al,Fe,Si) phase. As soon as
the pyramid-shaped intermetallic layer
forms on the die surface, the excess
aluminum sticks to it (see Figures 1c
and 2d). The initial sticking is primarily
due to the arrest of the ongoing reaction
between the iron, the aluminum-alloy
melt, and the surface energy effect of
the protruding intermetallic layer into
the aluminum melt. Another possible
reason for aluminum to stick to these
pyramid-shaped protrusions is the low
a b thermal conductivity of the intermetallic
Figure 3. Sample microstructure of soldered samples obtained from die casting plants. The phase layers compared to the steel
die material is H-13 and the alloy is aluminum 380 series alloy. (a) A stereo photograph surface. Hence, when the rest of the
(50X) showing the erosion pits on the H-13 steel surface. (b) The counterpart of the
figure in (a) wherein the mounds of intermetallic phases are found on the surface of the casting is solid and ready to be ejected,
soldered aluminum in contact with the steel surface. (a) and (b) are also micrographs of the melt around the intermetallic layers
soldered samples obtained from industries where the intermetallic layers were mechanically is solidifying. This results in the sticking
stripped out of the steel surface.
of the melt around the protrusions
Microstructures shown in Figures 2b surface. Silicon and other minor elements of intermetallic layers even after the
and 2c are those of soldered samples (chromium, manganese, vanadium, etc.) ejection of the cast part. Figure 5 is
from Amcast Industrial Corporation, from the die and the aluminum-alloy a micrograph showing the initial pits
and a soldered sample from experiments melt precipitate on the grain boundaries formed on the die surface, the subsequent
conducted at ACRC. of the Fe2Al5 intermetallic phase. Large intermetallic phase layers that formed
In the next stage, the FeAl3 phase silicon-rich precipitates are also found at and grew out of the erosion pits, and,
reacts with aluminum and silicon in the boundary between the binary and the finally, the part of the aluminum casting
the alloy melt to form the ternary ternary phases. Zinc-rich compounds that stuck onto these intermetallic layers
α-(Al,Fe,Si) phase. The intermetallic precipitate on the grain boundaries of and remained on the die steel (after the
layers formed during this stage have a the ternary α-(Al,Fe,Si) phase layer. ejection of the cast part).
pyramid-like morphology. This is due This stage is shown in Figure 1c, where Over time, the erosion pits widen
to the radial growth of the intermetallic the pits can be seen to expand laterally and merge with each other, resulting in
phases out of the pits on the steel and in depth. In addition, Figure 2d the straightening of the pits. Once the
surface. The ternary phase has the shows the growth of pyramid-shaped initial intermetallic layer forms inside
highest thickness compared to the other structures from initial layers in a sample the pits, the pits stop growing into the
phase layers. As the volume of the obtained from soldering experiments steel surface and grow parallel to the
aluminum melt is abundant, the reaction conducted at ACRC. Figure 4, which die surface. When adjacent pits start
between the intermetallic phases and the shows the pyramids forming at the to merge with each other, the supply
melt dominates over the diffusion of the soldered interface of a die steel from of fresh aluminum melt to the steel
iron from the steel surface. However, a die-casting operation, also shows the surface is restricted to the gaps and
the overall thickness of the intermetallic pits that are initially formed under the cracks between the intermetallic layers
layer on the steel surface is controlled pyramids and the radial growth patterns growing out of the adjacent pits. In
by the diffusion of iron from the steel of the various intermetallic compounds time these cracks close, resulting in the

Figure 4. Pyramid-shaped intermetallic phase layers found at the

soldering interface. The sample shown is taken from a soldered
section of the die. Also seen in this micrograph are the pits that
initially formed under the pyramids and the radial growth patterns
of the various intermetallic compounds emanating from these pits.
In addition, the adherence of the soldered aluminum onto these
pyramids can be observed.

10 µm

2002 August • JOM 49

 Table I. Thermodynamic Quantities categories of non-ferrous die materials of the linear fit lines and the x-axis is
Accompanying the Formation of the were evaluated for soldering tenden- a significant term; it defines the time
Binary Iron-Aluminum Phases7 cies. They include tungsten-based elapsed before the material starts to form
powder/metallurgy (P/M) material erosion pits and when the intermediate
∆H0273 K ∆S0273 K ∆G973 K (Anviloy1150), two titanium-based layers begin to grow. In the case of
P/M composites (CermeTi-C-10 and H-13, this time, to, is quite small. H-13
Phase Jmol–1 K–1 mol–1 Jmol-1 CermeTi-C-5), and molybdenum- begins to form erosion pits in less
based material (TZM)—four samples than 4 s. Once the pits are formed in
α-FeAl –51,240 51.0 –11,090
altogether. The soldering mechanisms H-13, the intermetallic layers begin
ζFeAl2 –81,900 73.3 –16,999
in these materials were studied and com- to grow instantaneously. In the case
η-Fe2Al5 –194,040 166.7 –19,636
θ-FeAl3 –112,560 95.6 –22,869
pared to die soldering in conventional of CermeTi and Anviloy, the time
ferrous die materials. All the above- before pitting on the die surface was
arrest of the chemical reaction between mentioned non-ferrous die materials considerably longer than H-13, about
the aluminum alloy and iron. Figure were manufactured via P/M. 77 s and 156 s, respectively.
5 shows the closure of one such crack Metallographic analysis of the sol- The growth of the intermediate
between adjacent pyramid-shaped dered interfaces revealed that intermedi- layer in all the die materials follows a
intermetallic phase layers. In this stage, ate layers were also formed in the non- parabolic rate law, providing evidence
the growth of the η-Fe2Al5 phase is more ferrous die materials. These layers grew that the layers grow by a diffusion
pronounced than the ternary phase. out of the die surface and the aluminum mechanism. Hence, it can be concluded
This is because the iron/aluminum soldered on top of the intermediate that for both ferrous and non-ferrous die
chemical-reaction rate diminishes and layers. In addition, erosion pits were materials, die soldering is a diffusion-
the iron-diffusion rate is constant. Once formed in all the die materials before controlled process. The growth rates of
the reaction forming the intermetallic the intermediate layers began to grow the intermediate layers in the soldered
layers is arrested, soldering is terminated on the die surface. microstructure obtained from a die-
and the aluminum sticks permanently on The length and amount of intermetal- casting operation were calculated and
the intermetallic layers. Re-melting of lic phase layer in the soldering interface are shown in Figure 6 for each die
the soldered aluminum is not probable is a good measure of the extent of material tested. The equation has the
because the time of contact of the fresh soldering.1,2 Figure 6 shows that the following form:
cast melt with the soldered aluminum is kinetics of the growth of the soldered
very short (≈4 ms). Hence, most of the layer, in both ferrous and non-ferrous die y = k√t + c
soldered microstructures exhibit a fixed materials, follow a parabolic growth law.
ratio of 1:5 between the intermetallic However, there is an initial incubation where y = thickness of the intermediate
layer thickness and the soldered thick- period after which the erosion pits form layer; t = time of contact of die material
ness. This stage is illustrated in Figures and the intermetallic phase layer begins with aluminum melt during die casting;
1d, 1e, 2e, 2f, 2g, and 2h. Microstructures to grow. In Figure 6, the intercept (to) c = constant; k = rate constant.
shown in Figures 2e, 2f, 2g, and 2h are
representative of samples from GM-
Powertrain, Bedford, Indiana; from
the ACRC diffusion-couple experi-
ments; and from samples obtained from
Cambridge Tool and Manufacturing,
Burlington, Massachusetts.
Soldering on Non-Ferrous Die
Materials
Most metal mold-casting firms use
tool steels for die materials. H-13, H-11,
modified H-11(KDA-1), H-21, nickel-
based superalloy (Rene), premium
H-13 (Orvar Supreme) are examples
of some of the ferrous die materials
used in aluminum casting. From our
understanding of the mechanism of
soldering, it has come to light that the
main cause of soldering is the high
affinity between iron and aluminum. To
further understand the mechanism of die Figure 5. Cross-section of the pits, intermetallic phase layer, and soldered aluminum.
soldering and to investigate a plausible
solution to mitigate the problem, three

50 JOM • August 2002

 The rate equations for each material
evaluated are shown in Figure 6; the
respective values of k and c are also
given in the figure.
REACTIONS
The following are the plausible reac-
tions between the various iron-aluminum
phases and the aluminum melt.
Reaction 1 ⇒ Fe(s) + Al(l) → α(Fe)(s)
Reaction 2 ⇒ α(Fe) + Al(l) → β
– FeAl(s)
Reaction 3 ⇒ β – FeAl(s) + Al(l) →
ζ – FeAl2(s)
Reaction 4 ⇒ ζ – FeAl2(s) + Al(l) →
η – Fe2Al5(s)
Reaction 5 ⇒ η – Fe2Al5(s) + Al(l)
→ θ – FeAl3(s)
Reaction 6 ⇒ θ – FeAl3(s) + Al – Si(l)
→ α – (AlFeSi)(τ5) + Al(l)
Table I gives the change of free energy
and the thermodynamic quantities
accompanying the formation of various
binary iron-aluminum compounds at
700°C.
KEY CONCLUSIONS
The literature has shown that the
following important factors influence
reactions at the ferrous die/molten
aluminum interface:1,2,6 temperature
of the melt, temperature of the die,
alloy chemistry of the melt and die,
die surface engineering, topographical
features, coatings, and casting operating
conditions. Among these, the more
critical factors were selected and the
effect of these factors on soldering was
investigated. An experimental setup was
built to simulate die-casting conditions.8
Following are the salient findings from
the work conducted at ACRC.
The addition of 0.125% titanium to
the aluminum 380 alloy significantly
decreased soldering during casting;
the alloy chemistry of the melt played
a more dominant role in alleviating
soldering than the temperature of the
melt and the dip time in the melt. The
qualitative and quantitative effects of
alloying elements in aluminum 380
alloy indicate the following:
• Iron has the greatest effect on
soldering. Increasing iron to satura-
tion limits in the aluminum allevi-
ates soldering.
• In low-iron alloys, manganese can
2002 August • JOM
be used at high percentages (up to
0.85 wt.%) to control soldering.
However, the presence of nickel is
detrimental to soldering; increasing
nickel in the alloy increases the
propensity of soldering.
• Magnesium, strontium, and silicon
do not affect soldering properties.
• The only elemental interaction that
contributes to reducing soldering is
that of manganese and nickel.
RECOMMENDATIONS
Based on a mechanistic understand-
ing of the interface reactions between
ferrous die and molten aluminum,
the following recommendations are
given for specific processing issues to
alleviate soldering during die casting of
aluminum alloys.
Melt Temperature
The temperature of the melt is a
critical factor in creating “hot-spots”
on the die surface. Care must be taken
to maintain a holding temperature of
about 671°C before pouring the melt
into the shot sleeve during die casting.
This will minimize the occurrence of
hot spots on the die surface. In addition,
if premium H-13 (QRO90) is used as the
die material, the melt temperature must
be below 677°C to ensure longevity of
the die material.
Die Pre-Heat Temperature
The die pre-heat temperature must
be between 299°C and 329°C. Higher
temperatures will result in inadequate
application of the lubricant, while
lower temperatures might result in the
formation of cold solder.
Composition of 380 Aluminum
Alloy
To reduce die soldering in aluminum-
silicon alloys, aluminum alloys with a
Table II. Properties of the Chosen Non-Ferrous Die Materials and H-13
Property ANVILOY CermeTi
Composition 4%Ni, 4%Mo, MMC of Ti-6Al-
2%Fe, 90%W 4V and TiC 99.2%Mo
Hardness (HRT) 34 40
(T) = 298 K to
550 K
Thermal Conductivity 128 8 120
(W/m K)
UTS (MPa) 945 1,000
Elongation 3% 3%
high iron content should have a limiting
range between 0.9 and 1.15 percent by
weight. The iron-saturation limit in the
melt at the pouring temperature will be
reached at these values, and any excess
iron will result in a large sludge factor.
Sludge causes melt contamination and
worsens the mechanical properties of
the casting. In addition, sludge will help
nucleate the intermetallic layer on the die
surface, thus facilitating soldering.
Low-iron alloys could have iron
content around 0.4 percent by weight.
In addition, manganese must be raised
to around 0.8 percent by weight to
compensate for the low iron in the melt.
Care should be taken to minimize the
nickel and the chromium contents in the
melt. Nickel reduces the effectiveness of
manganese in the melt, and chromium
increases the sludge factor. Reducing
silicon content in a low-iron alloy to
about 7 percent by weight will increase
the chemical activity of manganese and
iron in the melt, and thus will reduce
soldering.
The addition of titanium to the melt
of about 0.125 percent by weight is
highly recommended to avoid soldering.
Greater than 0.125 percent titanium does
not contribute to reducing soldering,
but gives rise to good grain refinement
in a permanent mold-casting process.
The upper limit of titanium addition
can be about 0.24 percent by weight.
Titanium will also form aluminides with
aluminum and silicon in the melt and will
reduce the polygonal iron aluminides
sludge crystals in the melt.
Nickel additions to the melt need to
be kept to an absolute minimum (almost
negligible) to alleviate soldering.
Die Material
Ferrous die materials undergo a
vigorous reaction when they encounter
molten aluminum. Although H-13, the
TZM H-13
0.5%Ti, 0.1%Zr, 0.35%C, 0.35%Mn,
1%Si, 5%Cr,
1.2%Mo, 1%V
RB – 260 52
28
1,000 1,730
10% 13%
51

Figure 6. The thickness of the intermediate layer for various contact times between molten
aluminum and die material during die casting. The lines represent the linear fit of the
respective data points and the equations in the legend indicate the equations for the
respective linear fits. In addition, the incubation time (to) before pitting and the start
of formatting intermediate layers is shown for each material. to values are noted by
circles on the ordinate.
most common die material, cannot be
easily replaced for economic reasons,
its composition can be modified to
better resist attack by molten aluminum.
One such modification resulted in the
emergence of QRO90, which has a better
resistance to molten aluminum attack.8
However, the temperature of the melt
should be maintained below 671°C to
enhance the life of the QRO90 die.
Understanding the mechanism of
soldering has helped to narrow down
the window of solutions to mitigate the
problem. One such solution is the use
of an alternate non-ferrous die material.
For a die material to effectively resist
soldering, it should:
• Be inert to corrosion and erosion
by molten aluminum.
• Offer reduced kinetics of formation
of reaction products between the
die material and aluminum.
• Maintain compositional and micro-
structural consistency on its die
material.
• Minimize thermal fatigue cracking
of the die surface.
The die materials studied were
selected for their thermal conductivity,
strength, and room-temperature and
high-temperature hardness. Table II
shows the properties of the selected
materials compared to H-13 tool steel.
Casting a die material creates large
variations in composition, microstruc-
52
ture, and mechanical properties. Powder-
metallurgy processes tend to create a high
integrity, net-shaped part with minimum
variations in material properties. In
addition, these parts can attain 99% or
greater values of theoretical densities.
Tungsten-based ANVILOY® 1150 is
a material developed by P/M process-
ing for its high-temperature strength
and excellent thermal conductivity.
Simplicity of tooling manufacture (no
heat treatment), low erosion, and excel-
lent resistance to thermal cracking
make it a premium choice for applica-
tions where H-13 and similar grades
might be used.9 CermeTi-C-10 and
CermeTi-C-5 are metal-matrix-com-
posite (MMC) materials containing
Ti-6Al-4V matrix and 10% and 5%
of TiC particles, respectively. The
patented materials are produced through
a proprietary CHIP processing,10 where
the part is produced by a three-step
P/M process of cold-isostatic pressing
followed by a vacuum-sintering process
and a hot-isostatic-pressing operation.
The material exhibits good machin-
ability, good wear resistance, high
fracture toughness, and high corrosion
resistance. CermeTi MMC materials
have been evaluated as a liner material
for shot sleeves where they have
demonstrated considerable improve-
ment in life than H-13 tool steel.11 TZM
is a molybdenum-based material that
is also a P/M product. It offers high
ductility and provides excellent service
in applications requiring high strength,
creep resistance, and ductility.
Die Surface Engineering
One of the most effective means of
retarding the soldering reaction is to
engineer the die-surface characteristics
by coatings, heat treatment, and surface
reactions.
Good surface coatings can act as
effective diffusion barriers to the
iron-aluminum reaction and retard the
soldering phenomenon. A number of
coatings are presently in use in the die-
casting industry. For example, boron-
based ceramics have proven to be
more effective in resisting molten
aluminum attack. One such coating,
boron nitride, can be coated by plasma
vapor deposition (PVD). The biggest
drawback in coating boron nitride to
H-13 tool steel is that the substrate
and the coating are non-compatible.
Hence, it is imperative that one uses an
additional layer of coating material to
act as a good intermediate layer between
the boron nitride and H-13. One such
coating is zirconium nitride. Zirconium
nitride has excellent adhesion properties
to H-13 and to boron nitride. Hence,
a coat of zirconium nitride followed
by a coat of boron nitride is effective
in increasing tool life. Both of these
coatings can be applied using the PVD
process. Soldering tests conducted on
these coatings showed that they were
more effective in resisting soldering
than H-13 or QRO90 tool steels. When
using a die coated with boron nitride,
care should be taken to not use water-
based lubricants on the die between
shots. Some varieties of boron nitrides
are hygroscopic and can easily dissolve
in water.
Nitriding the die surface is a common
practice in the aluminum-casting
industry. Gas nitriding the die surface
results in the formation of iron nitrides
on the die surface, which act as a
diffusion barrier to aluminum-iron
interactions. Nitriding enhances tool life
to a certain extent. There are, however,
alternatives to nitriding by other surface
engineering concepts. Die life can be
enhanced by such surface treatments
as packed boronizing, ferritic nitro-
carburizing, and aluminizing ferrous
JOM • August 2002
die in pure aluminum melt (<0.03
wt.% Si).
H-13 samples can be boronized to
form a layer of iron borides on the die
surface. Borides are better resistant to
molten aluminum attack than nitrides.
Boronizing is a standard procedure in
molten-metal handling industries. In a
recent study, Tsipas et al.1,2 have shown
that boronizing the surface of an H-13 die
proved to be far better than nitriding in
resisting attack from molten aluminum.
In the experiments conducted at ACRC,
pack-boronized H-13 tool steel outper-
formed conventional H-13 and the
QRO90 die materials in alleviating die
soldering. Wheras nitriding provides
only one surface layer of iron-nitride
compound, boronizing results in duplex
layers of Fe2B and FeB+Fe2B. The latter
results in better resistance to molten
aluminum attack.
The ACRC work established that to
improve the life of a ferrous die, one has
to retard the formation of the erosion
pits on the die surface. One way to
retard the pitting process is to strengthen
the soft phases on the surface of the
die material by tempering the die.
Tempering is an essential heat-treatment
procedure to impart toughness and
resistance to hot shortness to the die.
Ferritic nitro-carburizing of the die
surface (after a double temper of
the die) will strengthen the soft non-
martensitic regions of the die. Nitriding
and boronizing the die surface will alter
the microstructure and the chemistry of
the die surface. Thus, innate hardness
and temper of the die surface can be
maintained and utilized effectively by
ferritic nitro-carburizing. During ferritic
nitro-carburizing, carbon and nitrogen
are introduced onto the surface of the
tool steel. The tool steel is heated below
the austenizing temperature (450°C
to 590°C) in an environment rich in
ammonia and methanol. This results in
the diffusion of carbon and nitrogen into
the surface to a depth of 0.1 µm to 1 µm.
Specifically, in ferritic nitro-carburizing,
the following reactions occur:
• Ferrite and cementite dissolve
nitrogen.
• Cementite transforms into ε phase
via ferrite + cementite → ferrite
+ ε.
• Epsilon phase grows at the expense
of ferrite until a covering compound
2002 August • JOM
layer is formed.
• Continued growth of the mono-
phase compound layer is controlled
by nitrogen diffusion.
Post nitro-carburizing oxidation
treatments can be carried out on the die
surface to impart excellent resistance
to erosion by molten aluminum. One
such treatment involves coating the
nitro-carburized surface with a layer of
black oxide. Quenching temperatures,
which are very important in nitro-
carburized steels, should be at least
550°C. If the quenching temperature
is below 550°C, nitride precipitation
occurs on the surface and this lowers
the strength of the steel. The higher the
temperature before quenching, the better
the properties of the nitro-carburized and
oxidized surface. Flash oxidizing the
surface of the tool after nitro-carburizing
to impart the black oxide (Fe3O4) is a
critical process that is very sensitive to the
temperature before quenching and also
to the quenching medium. A medium
of oil/water emulsion maintained at
70–80°C is ideal to form a uniform,
non-flaky, and thin black-oxide layer
of about 1 µm thickness over the nitro-
carburized surface.
Efforts to use iron aluminides as
diffusion barriers on ferrous die materi-
als have proven that, to retard ferrous
die/molten-aluminum reaction, a coating
of iron aluminides over the die surface
is effective. However, when coating tool
steel with iron aluminides, the coating
does not adhere well to the die surface. In
the type-2 hot-dip aluminizing process,7
the steel is dipped in a bath of pure
aluminum to coat the die surface with
aluminum. The difference in coating an
aluminum-silicon alloy (soldering) and
pure aluminum is that, in the former,
the intermetallic phases grow out of
the steel surface while in the latter the
intermetallic layer grows into the steel
surface. When the intermetallic layer
grows out of the steel surface, it is more
prone to flaking, resulting in soldering
of the die surface.
The following Type-2 aluminizing
process can be effectively used to impart
a coating that is resistant to soldering:
• Polish the die surface to a mirror
finish.
• Pickle the surface with a 1:1
volume ratio of HCl and water for
15–20 min.
• Dip the die material in a bath of
molten aluminum (<0.03 wt.% Si)
maintained at 760–800°C.
• Air-cool the tool steel to room
temperature.
• The die material can be kept
at 300°C for about 12–24 h to
enhance the diffusion process and
the formation of the intermetallic
layers.
• Polish the coated surface to the
required dimensions.
• Soak the die materials at 300°C
for 12–24 hours to oxidize the
aluminum coating.
• Final polish to a mirror finish.
This process will result in the forma-
tion of binary iron aluminides growing
into the steel surface and a thin layer
(2–3 µm) of pure aluminum coating on
the surface of the die.
Aluminizing the ferrous die prior
to the die-casting operation offers
numerous benefits. The coating will
be an excellent diffusion barrier to
soldering in a 300 series aluminum-
alloy environment because the pure
aluminum coating melts at around
660°C and the temperature of the cast
aluminum alloy is never above 595°C.
In addition, the pure aluminum coating
is very ductile and can easily withstand a
180° bend test of the tool-steel substrate.
Thermal conductivity of the die will
be enhanced because the coating will
cover all of the die surface, and hot
shortness and heat-checking problems
of the tool steel will be eliminated
because the surface of the tool steel will
have less exposure to thermal fatigue.
The mirror finish on the coating will
ensure uniform cooling of the melt and,
hence, the surface finish of the cast part
will be greatly enhanced, and, finally,
the aluminum-oxide layer that forms on
the coating will ensure minimal wetting
of the coating by the cast metal.
Complimentary to this work, research
is continuing at ACRC/MPI to further
develop specific die material surface
engineering concepts.
ACKNOWLEDGEMENTS
The authors are grateful for the sup-
port of the Advanced Casting Research
Center’s industrial partners and for their
guidance and assistance in carrying
out this study.
53

References
1. Sumanth Shankar and Diran Apelian, “Die Solder-
ing—Mechanism of the Interface Reaction between
Molten Aluminum Alloy and Tool Steel,” accepted for
publication in Metall. Trans. (in press).
2. Sumanth Shankar and Diran Apelian, “Die Sol-
dering—Effect of Process Parameters and Alloy
Characteristics on Die Soldering,” Int. J. Cast Metals
(accepted and in press).
3. Sumanth Shankar and Diran Apelian, “Effect of the
Die Material and Surface Treatment on the Soldering
of Aluminum Alloys to Dies,” NADCA Transactions
(Rosemont, IL: NADCA, 2000).
4. Sumanth Shankar and Diran Apelian, “The Role
of Aluminum Alloy Chemistry and Die Material in
Die Soldering,” NADCA Transactions (Rosemont, IL:
NADCA, 1999), pp. 281–290.
5. Sumanth Shankar and Diran Apelian, “Effect of
Variation of Aluminum Alloy Chemistry (Titanium and
Strontium) on Die/Molten Metal Interface Interactions,”
Proceedings of the 5th Int. Molten Aluminum Processing
Conf. (Des Plaines, IL: AFS, 1998), pp. 209–230.
6. Sumanth Shankar and Diran Apelian, “Die Soldering:
A Metallurgical Analysis of the Die/Molten Metal
Interface Reactions,” NADCA Transactions (Rosemont,
IL: NADCA, 1997), pp. 245–251.
7. R.W. Richards, Int. Metals Reviews, 39 (5) (1994)
p. 191.
8. Sumanth Shankar, “A Study of the Interface Reaction
Mechanism Between Molten Aluminum and Ferrous
Die Materials” (Ph.D. Dissertation, WPI, Worcester,
MA, April 2000).
9. James R. Woodruff, CMW, private communication
(1999).
10. Susan M. Abkowitz, Dynamet Technologies, private
communication (1999).
11. S.M. Abkowitz, A. Laitinen, and D. Apelian, “Shot
Sleeve Life Extension with New MMC Material,” Light
Metal Age, 57 (1, 2) (1999), pp. 114–116.
12. D.N. Tsipas et al., “Degradation Behaviour of
Boronized Carbon and High Alloy Steel in Molten
Aluminum and Zinc,” Materials Letters, 37 (Oct. 1998),
pp. 128–131.
Sumanth Shankar is a post-doctoral fellow and
Diran Apelian is Howmet Professor of Engineering
and Director of Metals Processing Institute at
Worcester Polytechnic Institute.
For more information, contact D. Apelian,Worcester
Polytechnic Institute, Advanced Casting Research
Center, Metals Processing Institute, 100 Institute
Road, Worcester, MA 01609; dapelian@wpi.edu.

News (Continued from Page 16.)

The 2002 Electronic Materials Con-
ference (EMC), sponsored by the
Electronic, Magnetic & Photonic
Materials Division of TMS, grew both
in the number of attendees and present-
ers over previous years. Showing
an increase of 28 percent over last year’s
conference, the 2002 meeting featured
41 sessions and 305 individual presenta-
tions, while attendance increased by 30
percent over 2001, to 518 practicing
professionals and students.
Conference abstracts were published
in the July 2002 issue of the Journal
of Electronic Materials. The Electronic
Materials Conference and Device
Research Conference will be held next
year at the University of Utah June
25–27 and June 23–25, respectively.
Washington News (Continued from Page 14.)
• Climate Change Science and Technology
• Education Research
In making funding decisions on these priorities, the Administration will apply
the following tests:
• Relevance. “R&D programs must be able to articulate why this investment is
important, relevant, and appropriate. Programs must have well-conceived plans
that identify program goals and priorities and identify linkages to national and
‘customer’ needs.”
• Quality. “R&D programs must justify how funds will be allocated to ensure
quality R&D. Programs allocating funds through means other than a competitive,
merit-based process must justify these exceptions and document how quality
is maintained.”
• Performance. “R&D programs must have the plans and management processes in
place to monitor and document how well this investment is performing. Program
managers must define appropriate outcome measurements and milestones that
can be used to track progress towards goals, and assess whether funding should
be enhanced or redirected.”
The memo also makes it plain that “We encourage agencies to fund new, high-priority
activities by reallocating resources from lower-priority or recently completed activities.
Requests for funding above guidance levels will require a compelling rationale that the
activity is important, that the agency is the best one to conduct the activity, and that funds
from lower priority or recently completed programs cannot be substituted.”

CORRECTION National Academy Calls for More Nanotechnology Investment

Due to editing gremlins, numerous roller
coaster heights were incorrectly stated as
“cm” rather than ‘m” (!) in the article
“Materials Give Roller Coaster Enthusiasts
a Reason to Scream” (May 2002, pp 16–20).
This does, to say the least, present a
deceptively mild view of these scream
machines. The correct heights are as follows:
Superman The Escape, 126 meters; Tower
of Terror, 115 meters; Steel Dragon 2000,
97 meters; Millennium Force, 94 meters;
Fujiyama, 79 meters; Titan, 75 meters;
Goliath, 72 meters; Nitro, 70 meters; Speed,
68 meters; Mr. Freeze, Six Flags St. Louis, 66
meters; Mr. Freeze, Six Flags Over Texas, 66
meters; Son of Beast, 66 meters — ed.
For nanotechnology to fulfill its promise of revolutionizing industry, the government-
funded National Nanotechnology Initiative needs to increase its support of long-term
research and promote more interdisciplinary effort, says a new report from the National
Academies’ National Research Council. Creative programs and long-term funding
commitments are needed to encourage the development of interdisciplinary research. An
independent advisory board composed of leaders from industry and academia should be
established to provide guidance to federal agencies on important R&D opportunities in
nanoscale science and technology, the report adds. Federal leaders of the initiative need to
develop an overarching strategic plan and outline goals and objectives for the long
term. Moreover, to move the science out of the laboratory and into society, continued
investment is necessary in the development of nanotechnology instruments. The initiative
will require continued collaboration at the local, national, and international level, the report
emphasizes. Partnerships with industry should be stimulated and nurtured to make the
science a commercial reality. The report is available at national-academies.org.

54 JOM • August 2002