Microstructure Evolution and Rheological Behavior of

Cooling Slope Processed Al-Si-Cu-Fe Alloy Slurry

PROSENJIT DAS, SUDIP K. SAMANTA, SUPRIYA BERA, and PRADIP DUTTA

In the present work, microstructure evolution during semi-solid slurry generation of

Al-Si-Cu-Fe alloy, using a cooling slope, was studied and the effect of microstructural
morphology of the slurry on its rheological behavior was investigated. Microstructure evolution
during melt flow along the slope was studied by extracting samples from various locations of the
slope and performing rapid oil quenching experiments. Quantitative investigation was
performed to evaluate primary phase shape and size for different process conditions of the
semi-solid slurry, and subsequently rheological investigations were performed to correlate slurry
morphology with its flow behavior. Three different types of rheological experiments were
performed: isothermal test, shear jump test, and shear time test, in order to investigate
rheological behavior of the semi-solid slurry. In addition, effect of melt treatment, by adding
modifier (0.1 wt pct of Al-10Sr) and grain refiner (0.15 wt pct of Al-5Ti-1B), on the
microstructure evolution during slurry generation, flow behavior of the slurry, and inter-
metallics formation was studied.

DOI: 10.1007/s11661-016-3356-3
Ó The Minerals, Metals & Materials Society and ASM International 2016

I. INTRODUCTION where a combination of mechanical and thermal prop-

erties is desirable. Moreover, it combines the benefits
AL-Si alloys offer a number of benefits such as associated with the binary Al-Si and Al-Cu alloys, where
excellent castability, high specific strength, low thermal Si increases strength, fluidity and Cu provides higher
expansion coefficient, and good corrosion resistance, ductility compared to binary Al-Si alloys.[7,8]
thus making them suitable for automotive applications. In order to further avoid casting defects associated
Moreover, these alloys reduce vehicle weight and with the conventional casting techniques, semi-solid
improve fuel economy in return.[1–3] However, in case casting has emerged as a potential manufacturing
of components subjected to high mechanical and ther- technique for automobile and aviation industries.[1–3]
mal cyclic stresses prevailing during service, Al-Si binary The primary goal behind semi-solid processing is to
alloys become unreliable mainly because of the presence achieve near-spherical grains in the cast parts, by
of solidification defects[4] in the cast parts such as macro preparing semi-solid slurry having independent islands
and micro pores, which become preferential sites for of spheroidal primary particles. Among the various
crack nucleation. Al-Si-Cu alloys have shown potential techniques reported in literature to prepare semi-solid
for usage in safety-critical applications pertaining to slurry,[9–11] the cooling slope method has gained popu-
automobile and aviation industries. In these alloys, larity in recent times owing to its lower costs and ease of
copper increases the mechanical strength of the cast operation. Although the cooling slope has been used to
parts by precipitating Cu-based intermetallic com- generate semi-solid slurry of several types of aluminum
pounds, both at room and elevated temperatures.[5,6] alloys,[12–18] very little work has been reported on its
On the other hand, addition of Fe as an alloying element application on Si-Cu-Fe (A380) alloy.[7,8] In view of this,
to these alloys improves their castability and generates the present work has been taken up to investigate the
thermally stable Fe-based intermetallics. In view of the semi-solid processability of Al-Si-Cu-Fe (A380) alloy
above, Al-Si-Cu-Fe (A380) alloy has emerged as poten- using cooling slope. The present study is an attempt to
tially useful for aerospace and automobile appliances understand the microstructure evolution and intermetal-
lic formation mechanism of the Al-Si-Cu-Fe (A380)
PROSENJIT DAS, Scientist, is with the NNMT Group, CSIR- alloy during cooling slope slurry generation, and the
Central Mechanical Engg. Research Institute, Durgapur 713209, India, effect of primary Al particles morphology on its flow
and also with the Department of Mechanical Engineering, Indian behavior at semi-solid state. These studies are critical to
Institute of Science, Bangalore 560012, India. SUDIP K. SAMANTA,
Scientist, is with the NNMT Group, CSIR-Central Mechanical Engg. establish process control during component develop-
Research Institute. SUPRIYA BERA, Assistant Professor, is with the ment using rheo pressure die casting system. Further-
Department of Metallurgical and Materials Engineering, National more, the effect of melt treatment in the form of modifier
Institute of Technology, Durgapur 713209, India. PRADIP DUTTA, (Al-10Sr) and grain refiner (Al-5Ti-1B) addition on
Professor, is with the Department of Mechanical Engineering, Indian
Institute of Science. Contact e-mail: pradip@[Link]
microstructure, intermetallics formation, and rheologi-
Manuscript submitted April 2, 2015. cal behavior of the alloy is also studied in the present
Article published online February 18, 2016 work.

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 47A, MAY 2016—2243

 II. EXPERIMENTAL PROCEDURE rheometer, graphite-made concentric cylinder geome-
tries were used to estimate non-Newtonian flow behav-
A380 Al alloy (composition shown in Table I) was ior of the semi-solid slurry. Dimensions of the
used for the present experiments. A schematic diagram measuring geometries were 24 mm cup (outer cylinder)
of the cooling slope rheocasting facility, used in the and grooved 20 mm bob (inner cylinder). The inner
present work for semi-solid slurry generation, is shown
in Figure 1. After preparing the melt at 1023 K (750 °C)
within a resistance heating furnace, it was degassed and
poured on the cooling slope at a desired pouring
temperature of 903 K (630 °C) using a bottom pouring
arrangement. Bottom pouring was performed using a
silicon carbide crucible having 10 mm hole at the
bottom. The cooling slope was adjusted at 60 deg slope
angle with respect to the horizontal plane, and it was
designed as a counter flow heat exchanger with oil
circulation underneath at 333 K (60 °C). In order to
study the effect of primary phase morphology on
rheological behavior of the said alloy, melt treatment
was performed prior to cooling slope processing in two
cases out of three different process conditions studied in
the present work. Cooling slope casting of the untreated
melt was performed, referred to as condition 1. In
condition 2, melt treatment was performed in the form
of modifier addition (0.1 wt pct of Al-10Sr master
alloy), while condition 3 refers to the combined addition
of grain refiner (0.15 wt pct of Al-5Ti-1B master alloy) Fig. 1—Schematic diagram of the cooling slope rheocasting set-up.
and modifier.
To ensure repeatability, the experiments reported in
the present work were repeated thrice. K-type thermo-
couples were placed at three different locations of the
slope to measure the melt temperature during flow along
the slope and recorded using rapid data acquisition
software (Scada). Melt samples were collected from the
top (100 mm from the contact zone), middle (250 mm
from the contact zone), and exit section (500 mm from
the contact zone) of the flow front and rapidly oil
quenched to study the morphological evolution of
primary Al, eutectic phases, and intermetallic particles.
The semi-solid slurry was allowed to solidify within a
mild steel mold under air cooling. Details on cooling
slope processing may be seen from earlier publications
of this group.[17,19]
After preparing metallographic samples from the oil
quenched samples, optical microscopy, scanning elec-
tron microscopy (SEM), and XRD analysis were per-
formed to investigate the morphology of primary Al and
intermetallic, eutectic phases, respectively.
Rheological investigation was performed on the
rheocast (using cooling slope) samples, using a
high-temperature Searle-type rheometer (Figure 2).
Prior to the commencement of rheological experiments,
the samples were heated upto a semi-solid temperature
of 843 K (570 °C), which was the average cooling slope
exit temperature of the melt for all three process Fig. 2—(a) Schematic diagram of the High-temperature rheometer.
conditions studied in the present work. In this (b) Zoomed view of the marked location.

Table I. Chemical Composition of A380 Aluminum Alloy Used in the Experiments

Si Mg Mn Cu Fe Ti Ni Cr Zn Pb Al
8.51 0.025 0.189 3.56 0.91 0.037 0.062 0.045 1.65 0.069 84.9

2244—VOLUME 47A, MAY 2016 METALLURGICAL AND MATERIALS TRANSACTIONS A

rotating cylinder is grooved to lessen the chances of wall B. Melt Temperature and Solid Fraction During Flow
slippage that can lead to incorrect rheological measure- Plots of temperature variation with time during melt
ments. Use of graphite in the measuring geometries flow down the cooling slope are shown in Figure 4.
nullifies the chances of chemical interactions with the The temperature values were measured using K-type
melt, while inert argon atmosphere within the heating thermocouples, placed along the melt stream at start,
chamber, maintained during experimentation, elimi- middle, and exit sections, as shown schematically in
nates the chances of melt oxidation at semi-solid Figure 1. The recorded slope exit slurry temperature in
condition. case of unrefined melt is 846 K (573 °C) (Figure 4(a)).
The rheological experiments performed in the present Hence, temperature drop of about 330 K (57 °C)
study are: isothermal tests, shear jump test, and steady temperature is observed during melt flow along the
state tests. Isothermal tests were carried out at 843 K slope from a pouring temperature of 903 K (630 °C).
(570 °C), where the shear rate was varied from 10 to On the other hand, modifier added melt shows a
700 s1. Ascending shear jump tests were performed, higher temperature drop (approximately 333 K
starting from 10 to 700 s1. To investigate time depen- (60 °C)) and slightly increased residence time within
dence of the slurry viscosity, it was allowed to shear at a the slope (Figure 4(b)). Although modifier addition
constant shear rate of 700 s1 for an hour after which decreases surface tension of the melt and reduces its
steady state was achieved. viscosity, a higher flow velocity results in faster
temperature drop due to enhanced heat transfer
coefficient, as evidenced by Das et al.[17,19] in their
III. RESULTS & DISCUSSIONS numerical study. Higher temperature drop increases
the solid fraction towards the latter part of the slope,
A. DTA Analysis leading to higher flow resistance and greater melt
Figure 3(a) represents the DTA curve of A380 residence time within the slope. Combined addition of
aluminum alloy, which shows the variation of heat grain refiner and modifier prior to cooling slope
flow vs temperature. Solidus and liquidus temperatures processing yields even lower slope exit temperature of
of the alloy were measured from the DTA curve. The the slurry i.e., 838 K (565 °C) (Figure 4(c)). Grain
values obtained for solidus and liquidus temperatures refiner facilitates nucleation of higher number of
are 795 K and 853 K (522 °C and 580 °C), respec- primary Al particles and subsequent increase in solid
tively. Figure 3(b) shows solid fraction vs temperature fraction, resulting in higher melt viscosity. On the
plot of the alloy. Rheocasting requires 30 to 40 pct other hand, modifier aids to decrease melt surface
solid fraction within the semi-solid slurry. Accordingly, tension. These two competing mechanisms ensure
843 K (570 °C) was found to be an appropriate highest melt residence time within the slope (7 seconds)
temperature for further processing of the slurry to and highest temperature drop (approximately 338 K
shape cast parts. At that temperature, the slurry (65 °C)). Table II shows temperature values measured
contains approximately 32 pct solid and the cooling at start, middle, and exit sections of the melt front and
slope causes approximately 333 K (60 °C) temperature corresponding solid fractions present within the melt
drop during melt flow along the slope. Hence, 903 K for different process conditions studied. Solid fraction
(630 °C) was chosen as the design pouring temperature for a particular temperature was measured from solid
of the said alloy to generate semi-solid slurry using fraction vs temperature plot of the alloy, as shown in
cooling slope. Figure 3(b).

Fig. 3—(a) DTA curve of Al-Si-Cu-Fe (A380) alloy during heating and (b) solid fraction vs temperature curve of A380 Al alloy.

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 47A, MAY 2016—2245

 nucleation of primary Al particles occurs at the
melt–slope interface due to the chilling effect on super-
heated melt. Initial crystals forming at the contact zone
get detached from the slope surface as a result of shear
stress caused by melt flow inertia. Hence, the free surface
of the slope promotes new nucleation and successive
detachment of dendrites from its surface, turning the
downstream liquid melt into a semi-solid slurry.
Solute segregation complements the shear-driven
fragmentation of primary Al particles due to the
erosion–corrosion mechanism attained by solute enrich-
ment at the roots of the solid crystals and subsequent
neck formation, as discussed in Robert et al.[20] Solute
enrichment occurs within the liquid ahead of the
solidification front, resulting in lower growing rate in
this region. Local retardation of the crystal growth
facilitates neck development at the dendrite root, thus
facilitating detachment by melt flow. Figure 5(a) shows
a schematic diagram explaining the aforementioned
mechanism, while Figure 5(b) shows the shear stress and
velocity distribution within the melt during flow along
the slope. Ohno and Motegi[21] have described this
mechanism as ‘‘separation theory’’ while explaining the
formation of equiaxial grains in castings. Further
insights on primary particle formation and their subse-
quent spheroidization during cooling slope processing of
the present alloy may be obtained from our published
numerical findings.[22]
In the present work, effect of melt treatment in the
form of modifier (Al-10Sr, 0.1 pct by weight) and grain
refiner (Al-5Ti-1B, 0.15 pct by weight) addition is
studied. Melt modification influences the primary Al
phase morphology by reducing melt surface tension,
leading to reduced viscosity and higher melt flow
velocity which, in turn, yields improved sphericity and
reduced particle size of primary Al particles. At the same
time, grain refiner addition facilitates early nucleation
sequence of primary Al particles by shifting the liquidus
temperature of the present alloy upwards, and Ti
particles (originated from Al-5Ti-1B master alloy), on
the other hand, increases the number of nucleation sites
within the melt, as evidenced by Nafisi et al.[23] During
cooling slope processing, simultaneous effect of the
above mechanisms results in reduced grain size and
improved sphericity of primary Al grains within the
semi-solid slurry.
Microstructure of the present alloy also consists of
eutectic Si and iron-based intermetallic particles. Nucle-
ation sequence of primary Al particles, Fe-based inter-
metallics, and eutectic Si in case of the present alloy are
schematically shown in Figure 5(c). At the initial stage
of slurry formation, primary Al phase forms at the
beginning as dendrites/equiaxed particles, followed by
the evolution of iron-based intermetallics between the
liquidus temperature and eutectic temperature of the
Fig. 4—Experimentally obtained temperature vs flow time plot dur- alloy (stage I of Figure 5(b)). Presence of Mn in the
ing melt flow through the cooling slope: (a) unrefined melt, (b) modi-
fier added melt, and (c) grain refiner and modifier added melt.
present alloy facilitates formation of polyhedron-shaped
a-Al15(Fe,Mn)3Si2 intermetallic particles, in spite of
commonly observed brittle intermetallics (Al8Fe2Si,
C. Mechanism of Microstructure Formation
Al5FeSi etc.,). When the melt reaches eutectic temper-
In the cooling slope method, as the liquid is poured ature, eutectic silicon nucleates on the Fe-based inter-
over the inclined slope cooled from underneath, copious metallics in the solute field ahead of the growing

2246—VOLUME 47A, MAY 2016 METALLURGICAL AND MATERIALS TRANSACTIONS A

 Table II. Slurry Temperature and Solid Fraction Present at Different Locations of the Slope

Temperature Solid Fraction

Processing Conditions Start Sample Middle Sample Exit Sample Start Sample Middle Sample Exit Sample
Unrefined 604 598 573 0 0.01 0.30
Modified 604 596 570 0 0.03 0.33
Refined + modified 600 593 565 0 0.05 0.51

Fig. 5—Schematic diagram of the (a) mechanism of neck formation of primary Al crystals, (b) microstructure formation mechanism during
shear-driven flow of the melt along the slope, and (c) Nucleation sequence of constituent phases and spherodization mechanism of primary Al
grains during cooling slope rheocasting process of the A380 Al alloy.

primary aluminum (stage II of Figure 5(c)). Grav-

ity-driven melt flow along the slope triggers shear-driven
rotational flow within the inter-particle spacing of
primary Al grains (Karagadde et al.[24]). In the present
study, diffusion-driven growth of primary Al particles
and rotational flow of the melt within the inter-particle
spacing leads to the spheroidization of microstructural
features within the semi-solid slurry of present Al-Si-
Cu-Fe alloy (stage III of Figure 5(c)).

D. Microstructural Investigation
1. Optical micrographs
Large dendrites are evident from the optical micro-
graph (Figure 6) of the as-received A380 Al alloy,
showing shrinkage porosities represented by the black
dots. Coarse acicular eutectic Si needles are observed in
the interdendritic regions. Figure 7 shows the optical Fig. 6—Optical micrograph of the as-received A380 ingot.

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 47A, MAY 2016—2247

Fig. 7—Optical micrographs of the evolution samples of untreated alloy: (a) Start, (b) Middle, (c) Exit, and (d) Mold.

micrographs of the oil quenched slurry samples of helps in improving their sphericity due to solute enrich-
untreated melt. Sample collected from the start position ment at their periphery. Improvement in the microstruc-
of the melt front (shown in Figure 1) shows finer tural morphology, in the form of rosettes and
dendrites (Figure 7(a)) with average secondary dendrite near-spherical primary Al particles, is observed in case
arm spacing (SDAS) of 20 lm and fragmented a-Al of the sample from the middle position of the slope
particles. Optical micrograph of the sample collected (Figure 8(b)). Further increase in sphericity and grain
from middle position of the melt front shows a size is evidenced in the exit-position sample because of
significantly different microstructural morphology. Coa- ripening-driven growth and abrasion between particles
lescence and ripening-driven growth of the degenerated and the slope wall during melt flow (Figure 8(c)). This
primary Al dendrites is evident from the micrograph phenomena is caused by decreased solidus temperature
(Figure 7(b)). Coalescence increases mean free path of of the alloy attained by modifier addition and because of
diffusion in solid phase until the grains attain spherical slow cooling within the metallic mold (Figure 8(d)).
shape (Guo et al.[25]). Microstructure of the slope exit Notable presence of polygonal a-Al15(Fe,Mn)3Si2 inter-
slurry sample show near-spherical particles, as evident metallic particles is observed in the case of mold sample
from the optical micrograph shown in Figure 7(c). of the modifier added slurry, the explanation of which is
Along with particle growth mechanisms, effect of provided in Section III–D–3 of the manuscript.
shear-driven fragmentation of the primary Al particles Highest number density of individual primary Al
is also evident from their broken connectivity and particles is observed in the start-position sample of the
distribution within the sample matrix in the form of grain refiner and modifier added slurry sample
islands. However, samples obtained from the solidified (Figure 9(a)). Grain refiner addition shifts the liquidus
billets show less sphericity and higher grain size temperature upwards, facilitating enhanced nucleation of
(Figure 7(d)) due to slow cooling within the mild steel primary particles and ensuring their finer morphology by
mold under ambient conditions. Moreover, slow cooling inhibiting their radial growth. Rosettes and isolated
facilitates formation of Fe-based intermetallics during fragments of the primary phase are observed in the
eutectic solidification (Figure 7(d)). middle-position sample (Figure 9(b)). Effect of coales-
Optical micrograph of the sample from the start cence is predominant over ripening-driven growth, as
position of the modifier added melt shows fragmented evident from the decreased number density of the primary
dendrite arms and rosettes (Figure 8(a)). Grain sizes of particles compared to that of the start-position sample
the individual particles are found to be comparable with (Figure 9(b)). Finally, optical micrograph (Figure 9(c))
those of unrefined melt, as modifier addition does not of the slope exit slurry sample shows isolated spheroids
facilitate enhancement of seed generation of primary Al and rosettes of a-Al phase and no evidence of fully
particles. However, broken connectivity of the particles formed dendrite arms. Addition of grain refiner leads to

2248—VOLUME 47A, MAY 2016 METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 8—Optical micrographs of the evolution samples of modifier added alloy: (a) Start, (b) Middle, (c) Exit, and (d) Mold.

Fig. 9—Optical micrographs of the evolution samples of combined melt-treated (grain refiner and modifier added) alloy: (a) Start, (b) Middle, (c)
Exit, and (d) Mold.

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 47A, MAY 2016—2249

greater number of primary Al particles resulting in higher
solid fraction. Higher solid fraction increases slurry
viscosity which, in turn, reduces the slurry velocity and
allows particles to agglomerate, coalesce, and sinter
together. Although individual grain sizes are smaller,
their connected nature is evident. On the other hand, the
slurry sample solidified within the metallic mold after
exiting cooling slope shows fully formed globular grains
(Figure 9(d)). This is because of the (i) higher presence of
isolated primary Al particles within the semi-solid slurry
while exiting cooling slope, which restricts their columnar
growth within the mild steel mold and (ii) turbu-
lence-driven disintegration of sintered primary Al parti-
cles during slurry pouring within the mold.

2. Image analysis of optical micrographs

Image analysis of the micrographs was performed
using ‘‘Image J’’ software to understand the effect of
process conditions on the shape and size of primary Al
grains (Figures 7, 8, and 9). Data obtained for grain size
and degree of sphericity (DOS) of primary Al phase for
different samples studied in the present work are shown
in Figures 10(a) and (b), respectively. The quantitative
results obtained are in line with the earlier discussion
presented in Section D–1. Slope exit slurry sample of the
combined melt-treated sample shows sphericity and
grain size of 0.86 and 25.23 lm, respectively, indicating
significant improvement compared to those of the exit
sample of the unrefined melt (sphericity: 0.74, grain size:
27.1 lm). The equations used to determine grain size
(GS) and sphericity (SP) are shown below, as also given
by Taghavi et al.[26]

pffiffiffiffiffiffiffiffiffiffiffi
GS ¼ 2 Aa =p ½1

  
SP ¼ 4pAa P2a ; ½2 Fig. 10—(a) Grain size and (b) degree of sphericity of the primary
Al particles for all three process conditions under consideration.
where Aa and Pa denotes the area and perimeter of
primary a-Al phase, respectively. solidification. SEM micrographs collected successively
The individual a-Al particles are treated as separate from the middle and exit positions of the melt front are
grains for grain size and sphericity analysis. shown in Figures 11(c) through (f), respectively.
Decrease in eutectic content with progressive nucleation
3. SEM micrographs, EDS analysis and growth of the primary Al phase with the advance-
SEM micrographs of the oil quenched slurry samples ment of semi-solid slurry towards slope exit is clearly
of the unrefined alloy are shown in Figure 11. Fig- evident. Significant change is observed in the morphol-
ures 11(a) and (b) show SEM micrographs of the sample ogy of Al2Cu particles, which are found to be dis-
collected from the initial section of melt front. Degen- tributed within the sample matrix at the grain
erated dendrite arms, along with coarse eutectic Si fibers boundaries, as against localized presence in case of the
and intermetallics, are observed at the interdendritic start-position sample. This is because of the shear-driven
locations. Figure 11(b) shows the morphology of the interdendritic melt flow along the slope, which redis-
intermetallic particles present within the sample, at high tributes the rejected solute from the growing primary Al
magnification, and marked locations where energy grains and Cu-based intermetallics.
dispersive spectroscopic (EDS) analysis (point and spot Effect of modifier addition in the slurry morphology
analysis) is performed. EDS analysis (Table III) shows can be observed from the SEM micrographs shown in
no evidence of brittle Fe-based intermetallics, perhaps Figure 12. Figures 12(a) and (b) show higher presence
because of the high cooling rate involved in case of oil of intermetallic Al2Cu particles in the microstructure
quenching. However, Cu-bearing intermetallic (Al2Cu) and finer eutectic Si fibers compared to that of untreated
particles are observed along with the eutectic Si parti- slurry samples. Sr addition in the melt provides larger
cles. Al2Cu particles are observed at the grain bound- number of nucleation sites for the eutectic Si particles,
aries, confirming their precipitation during eutectic which is responsible for their finer morphology, as also

2250—VOLUME 47A, MAY 2016 METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 11—SEM micrographs of the evolution samples of untreated alloy: (a) Start, (b) high magnification micrograph of (a), (c) middle, (d) high
magnification micrograph of (c), (e) exit, (f) high magnification micrograph of (e).

Table III. EDS Analysis Results of Marked Locations of Figures 11, 12, 13, 14

Point Al Si Ti Cr Mn Fe Ni Cu Zn Pb Possible Phase

pt1 72.35 26.18 0.15 0.16 0.00 0.14 0.09 0.89 0.04 0.00 Al-Si eutectic
pt2 64.41 3.96 0.00 0.00 0.00 0.75 1.16 29.52 0.00 0.20 Al2Cu intermetallic
pt3 63.12 0.52 0.16 0.01 0.00 0.21 0.92 34.56 0.00 0.49 Al2Cu intermetallic
pt4 68.30 30.14 0.00 0.00 0.00 0.05 0.00 0.94 0.30 0.26 Al-Si eutectic
pt5 75 24.63 0.01 0.00 0.00 0.16 0.00 0.15 0.05 0.00 Al-Si eutectic
pt6 73.78 0.57 0.00 0.00 0.26 0.16 1.26 22.80 0.91 0.27 Al2Cu intermetallic
pt7 62.15 0.55 0.14 0.02 0.00 0.24 1.10 35.42 0.00 0.38 Al2Cu intermetallic
pt8 75.70 4.02 0.27 1.15 3.52 13.23 0.00 1.79 0.22 0.11 a-Al15(Fe,Mn)3Si2 intermetallic
pt9 77.08 4.32 0.07 1.47 3.76 12.15 0.00 0.73 0.42 0.00 a-Al15(Fe,Mn)3Si2 intermetallic
pt10 62.85 0.39 0.00 0.00 0.00 0.26 0.63 32.85 1.30 0.17 Al2Cu intermetallic
pt11 60.66 38.36 0.01 0.02 0.00 0.14 0.00 0.54 0.25 0.00 Al-Si eutectic
pt12 98.14 0.51 0.16 0.00 0.00 0.05 0.11 0.37 0.66 0.00 Primary Al
pt13 75.70 4.02 0.27 1.15 3.52 13.23 0.00 1.79 0.22 0.11 a-Al15(Fe,Mn)3Si2 intermetallic
pt14 64.05 0.66 0.05 0.00 0.06 0.04 0.61 33.73 0.23 0.30 Al2Cu intermetallic

evidenced earlier by Hegde et al.[27] in case of Al-Si Higher number density of primary Al grains is evident
alloys. SEM micrographs of the sample, collected from from the micrograph of the start-section sample of the
the middle section of the flow front, are shown in combined melt-treated sample (Figure 13(a)), compared
Figures 12(c) and (d). Micrographs of the exit-section to the corresponding samples from the untreated and
sample (Figures 12(e) and (f)) show presence of Fe-based modifier added slurries. Eutectic Si morphology is found
intermetallics (Figure 12(e)). EDS analysis confirms that to be even finer (Figure 13(b)). Interestingly, the pres-
these Fe-based intermetallics are a-Al15(Fe,Mn)3Si2. ence of Fe-based intermetallic particle is observed within
This is perhaps due to the tendency of modifier to push the eutectic region of the start sample, which is perhaps
the solidus temperature of the given alloy downwards, due to Sr addition. In case of middle-section sample
which affects local thermosolutal convection and facil- (Figures 13(c) and (d)), higher sphericity of primary Al
itates nucleation of a-Al15(Fe,Mn)3Si2 particles at the grains and decreased eutectic content is evidenced, as
primary Al grain boundaries. expected. Micrographs observed in case of exit-section

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 47A, MAY 2016—2251

Fig. 12—SEM micrographs of the evolution samples of modified alloy: (a) Start, (b) high magnification micrograph of (a), (c) middle, (d) high
magnification micrograph of (c), (e) exit, (f) high magnification micrograph of (e).

Fig. 13—SEM micrographs of the evolution samples of grain refined and modified alloy: (a) Start, (b) high magnification micrograph of (a), (c)
middle, (d) high magnification micrograph of (c), (e) exit, (f) high magnification micrograph of (e).

samples (Figures 13(e) and (f)) appear to be ideal for solute rejection, which facilitates formation of primary
rheo pressure die casting, with fine eutectic Si and Si due to local hypereutectic composition of the melt. At
eutectic Al2Cu particles distributed uniformly within the the same time, it depletes the available Al content within
grain boundaries, and absence of iron-based inter- the melt to prevent formation of Fe-based intermetallics.
metallics. Enhanced progressive nucleation of primary In case of combined melt treatment, comparison of
Al particles due to modifier addition results in higher primary Al phase, Fe- and Cu-based intermetallics and

2252—VOLUME 47A, MAY 2016 METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 14—Comparative SEM micrographs of the (a) conventional cast
sample and (b) cooling slope cast sample.

eutectic phase morphology between the conventional

cast sample and cooling slope cast sample is shown in
Figure 14. Figure 14(a) shows the presence of Fe-based
intermetallic a-Al15(Fe,Mn)3Si2 particles (polygon
shaped), Cu-bearing intermetallic Al2Cu particles, den-
dritic primary Al grains, and eutectic Si in case of
conventional cast sample. Intermetallics observed in the
micrograph are mostly present in the grain boundaries,
which confirms their precipitation towards the end of Fig. 15—Comparative XRD patterns of the quenched slurry sam-
ples: (a) untreated slurry, (b) modifier added slurry, and (c) com-
solidification i.e., during eutectic solidification. On the bined melt-treated slurry.
contrary, in case of cooling slope cast combined melt-
treated sample (Figure 14(b)), significantly enhanced
number of primary Al grains are observed with signifi-
cantly reduced grain size and improved sphericity, with no the melt flow direction, is evident from the peak inten-
evidence of iron-based intermetallics. sities (Figure 15(a)). While a sample collected from the
starting section of the slope shows minimum peak
intensity of solutal silicon, maximum peak intensity is
E. XRD Study observed in case of slurry sample from the slope exit
X-ray diffraction analysis is performed in this study to (Figure 15(a)). This trend is found to be true for all three
complement the findings obtained from SEM micro- process conditions studied in the present work. However,
graphs and EDS analysis. X-ray diffraction analysis relative intensity of solutal silicon peak varies among the
confirms the rejection of solutal Si during progressive samples, because of enhanced volume fraction of primary
nucleation and growth of primary Al phase, while flowing Si in case of melt-treated slurry samples. Higher peak
down the cooling slope. Increasing silicon (solute) con- intensity of primary Si in case of modifier added sample is
tent within the slurry samples, collected successively in attributed to the widened solidification interval due to the

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 47A, MAY 2016—2253

tendency of modifier (Sr) to reduce the solidus temper-
ature, which facilitates enhanced progressive nucleation
and growth of primary Al particles and subsequent
rejection of solutal Si (Figure 15(b)). Further increase in
peak intensities of solutal Si in case of combined
melt-treated samples is attributed to the increased num-
ber of primary particles within the semi-solid slurry,
which results in even higher rejection of solute
(Figure 15(c)). The findings obtained from XRD inves-
tigations are in line with the evidence obtained from the
SEM investigations. Enhanced peak intensities of
a-Al15(Fe,Mn)3Si2 and Al2Cu particles may be seen from
Figure 15(b), as compared to those of untreated slurry
sample. However, further decrease in peak intensities of
intermetallics is evidenced in case of combined melt-
treated slurry samples (Figure 15(c)).

Fig. 16—Viscosity vs shear rate of cooling slope cast samples of

A380 Al alloy at semi-solid state.
F. Rheological Characterization
1. Isothermal tests
Isothermal test confirms shear thinning nature of all rate, power law relation between viscosity (g) and
the samples at semi-solid condition. Figure 16 shows the shear rate can be found as follows:
results of isothermal tests performed at 843 K (570 °C)
g ¼ C_cm ; ½4
by varying the shear rate from 0 to 700 s1. In the entire
shear regime, the highest viscosity values are recorded in where m is the power law index and equals to n1.
case of samples without melt treatment, owing to the
presence of primary dendritic particles which are 2. Shear jump tests
responsible for impeding the free movement of the Shear jump brings sudden change in flow behavior of
liquid matrix under application of shear. the semi-solid slurry of a given alloy by formation and
Melt treatment in the form of modifier addition is breakage of structural linkages i.e., particle clusters of
found to decrease slurry viscosity due to higher amount primary Al particles present within the slurry matrix, as
of spheroidal primary particles within the slurry matrix, discussed by Modigell et al.[29] and Koke et al.[30] Effect
and also because of the tendency of the modifier to of primary particle shape, size, and distribution i.e.,
decrease surface tension of the melt, leading to a particle morphology is found to have profound effect on
decrease in viscosity. Higher viscosity values are non-Newtonian flow behavior of the present alloy slurry
observed in the low shear regime upon grain refiner during sudden jump of shear rate. In the present study,
addition. Grain refinement facilitates copious nucleation ascending shear jump experiments were performed at
and enhances generation of spheroidal primary Al 843 K (570 °C) by varying the shear rate from 10 to
particles having smaller particle size within the cooling 700 s1 for all cases.
slope cast samples. Presence of more primary particles Shear jump experiments were performed in the case of
gives rise to the tendency to form agglomerates at low all three process conditions studied in the present work,
shear regime and enhances melt viscosity. With increas- and representative findings of combined melt-treated
ing shear, however, structural linkages between particle slurry are shown in Figure 17(a) and (b). Figure 17(a)
clusters disintegrate to decrease slurry viscosity. Starting shows the viscosity evolution of the sample cast using a
from 6.275 PaS at the shear rate of 50 s1, viscosity slope angle of 60 deg after combined melt treatment. As
values decrease to 0.0301 PaS at 700 s1 shear rate in evident from the corresponding micrograph (Figure 9(d)),
case of combined melt-treated slurry samples. This is the the sample contains mostly fully formed globular grains,
lowest among the samples studied and confirms the unlike rosettes and degenerated dendrites in case of mold
effectiveness of combined melt treatment in improving samples of untreated and modified slurry, which results in
slurry rheological behavior of the given alloy. lowest viscosity and shear stress values in this case. Sudden
Ostwald–De Wagle power law, used earlier by jump of shear rate from 100 to 400 s1 results in sharp
Brabazon et al.,[28] effectively expresses that the relation drop of viscosity and shear stress (Figures 17(a) and (b))
exists in the present case between shear stress,s, and values, and the curve flattens afterwards. It is interesting to
shear rate, c_ . The equation is shown below: note that both viscosity and shear stress values show
s ¼ C_cn ; ½3 overshoot immediately after the individual shear jump due
to melt flow inertia, agglomeration, and sudden increase in
where C is the power law constant and n is the shear shear rate, respectively. The slurry is found to attain steady
rate exponent. Using Eq. [3] and from the definition of or equilibrium state after a period of 5 minutes for each
Newtonian viscosity i.e., shear stress divided by shear an individual shear rate. The fluctuations observed in the

2254—VOLUME 47A, MAY 2016 METALLURGICAL AND MATERIALS TRANSACTIONS A

 Fig. 18—Shear time dependence of viscosity and shear stress for the
combined melt-treated sample.

IV. CONCLUSIONS
The conclusions obtained from the present study are
described below:
1. Copious nucleation of primary Al particles at the
melt contact zone of the cooling slope and subse-
quent shear-driven fragmentation of nucleated pri-
mary Al crystals governs the primary Al particles
shape, size, and morphology of the present Al-Si-
Cu-Fe alloy.
2. Combined addition of grain refiner and modifier
yields the lowest slope exit temperature, highest
solid fraction, smallest primary Al particle size, and
highest sphericity.
Fig. 17—Shear jump test results of grain refined+modified sample:
(a) viscosity and (b) shear stress.
3. Along with primary Al grains, shear-driven melt
flow along the slope is found to facilitate fragmen-
tation of Al2Cu-based intermetallics and redistribu-
tion of the rejected solute from growing primary Al
viscosity and shear stress values after 400 s1 shear rate are grains.
attributed to the continuous agglomeration and deag- 4. Modifier addition pushes the solidus temperature of
glomeration phenomena between primary particles due to the present alloy downwards and, in turn, affects
higher density of primary particles within the slurry local thermosolutal convection and facilitates nucle-
matrix, attained by combined grain refiner and modifier ation of a-Al15(Fe,Mn)3Si2 particles at the primary
addition. Similar findings have been reported in an earlier Al grain boundaries. On the other hand, combined
work by the present research group.[31] addition of grain refiner and modifier is found to
enhance formation of primary Si due to higher
3. Steady state tests amount of solute rejection in this case.
In the present study, A380 Al alloy is found to be 5. During isothermal test, grain refiner and modifier
thixotropic i.e., it behaves like solid when kept at rest added slurry samples show higher viscosity values at
and flows like liquid under application of shear. Steady the beginning compared to untreated slurry sample
state tests performed at a constant shear rate of 700 s1 due to agglomerate formation. However, with
and constant temperature of 843 K (570 °C) for a increasing shear, structural linkages between parti-
period of 60 minutes show decreasing trend in slurry cle clusters disintegrate to decrease slurry viscosity.
viscosity and stress developed within the slurry matrix
under application of shear (Figure 18). The mechanisms
responsible for this phenomenon are crystal fragmenta-
tion, breakage of structural linkages between primary Al
particles under shear-driven flow of the slurry within the ACKNOWLEDGMENTS
annular space of the rheometer. Continued agglomera-
tion and deagglomeration of primary particle chunks Authors would like to acknowledge the funding re-
during steady state tests are responsible for the fluctu- ceived from DST, New Delhi and CSIR-CMERI,
ations observed in the viscosity and shear stress curves. Durgapur.

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 47A, MAY 2016—2255

 REFERENCES 18. P. Das, S.K. Samanta, H. Chattopadhyay, P. Dutta, and B.R.K.
Venkatapathi: Trans. Indian Inst. Met., 2012, vol. 65 (6),
1. H.V. Atkinson: Prog. Mater Sci., 2005, vol. 50, pp. 341–412. pp. 669–72.
2. Z. Fan: Int. Mater. Rev., 2002, vol. 47 (2), pp. 49–85. 19. P. Das, S.K. Samanta, H. Chattopadhyay, and P. Dutta: Acta
3. D.H. Kirkwood: Int. Mater. Rev., 1994, vol. 39, pp. 173–89. Metall. Sin. (Engl. Lett.), 2012, vol. 25 (5), pp. 329–39.
4. L. Ceschini, I. Boromei, A. Morri, S. Seifeddine, and I.L. Svensson: 20. M.H. Robert, E.J. Zoqui, F. Tanabe, and T. Motegi: J. Achiev.
J. Mater. Process. Technol., 2009, vol. 209, pp. 5669–79. Mater. Manuf. Eng., 2007, vol. 20 (1-2), pp. 19–26.
5. H. Puga, S. Costa, J. Barbosa, S. Ribeiro, and M. Prokic: J. 21. A. Ohno, T. Motegi, and H. Soda: Trans. Iron Steel Inst. Jpn.,
Mater. Process. Technol., 2011, vol. 211, pp. 1729–35. 1971, vol. 18, pp. 11–14.
6. K. Labisz, M. Krupinski, and L.A. Dobrzanski: J. Achiev. Mater. 22. P. Das, S.K. Samanta, P. Kumar, and P. Dutta: ISIJ Int., 2014,
Manuf. Eng., 2009, vol. 37 (2), pp. 309–16. vol. 54 (7), pp. 1601–10.
7. E. Parshizfard and S.G. Shabestari: J. Alloy. Compd., 2011, 23. S. Nafisi, O. Lashkari, R. Ghomashchi, F. Ajersch, and A.
vol. 509, pp. 9654–58. Charette: Acta Mater., 2006, vol. 54, pp. 3503–11.
8. S.G. Shabestari and E. Parshizfard: J. Alloy. Compd., 2011, 24. S. Karagadde, A. Bhattacharya, G. Tomar, and P. Dutta: J.
vol. 509, pp. 7973–78. Comput. Phys., 2012, vol. 231, pp. 3987–4000.
9. R. Canyook, S. Petsut, S. Wisutmethangoon, M.C. Flemings, and 25. H.M. Guo, X.Q. Luo, A.S. Zhang, and X.J. Yang: Trans. Non-
J. Wannasin: Trans. Nonferrous Met. Soc. China, 2010, vol. 20, ferrous Met. Soc. China, 2010, vol. 20, pp. 1361–66.
pp. 1649–55. 26. F. Taghavi and A. Ghassemi: Mater. Des., 2009, vol. 30,
10. R. Canyook, J. Wannasin, S. Wisutmethangoon, and M.C. pp. 1762–67.
Flemings: Acta Mater., 2012, vol. 60, pp. 3501–10. 27. S. Hegde and K.N. Prabhu: J. Mater. Sci., 2008, vol. 43,
11. S. Gencalp and N. Saklakoglu: Mater. Manuf. Process., 2010, pp. 3009–27.
vol. 25, pp. 943–47. 28. D. Brabazon, D.J. Browne, and A.J. Carr: Mater. Sci. Eng. A,
12. T. Haga: J. Mater. Process. Technol., 2002, vols. 130–131, pp. 558–61. 2003, vol. 356, pp. 69–80.
13. T. Haga and S. Suzuki: J. Mater. Process. Technol., 2001, vol. 118, 29. M. Modigell and J. Koke: J. Mater. Process. Technol., 2001,
pp. 169–72. vol. 111, pp. 53–58.
14. Y. Birol: J. Mater. Process. Technol., 2008, vol. 207, pp. 200–203. 30. J. Koke and M. Modigell: J. Non Newtonian Fluid Mech., 2003,
15. Y. Birol: J. Mater. Process. Technol., 2007, vol. 86, pp. 94–101. vol. 112, pp. 141–60.
16. Y. Birol: J. Alloy. Compd., 2009, vol. 473 (1–2), pp. 133–38. 31. P. Das, S.K. Samanta, and P. Dutta: Metall. Mater. Trans. B,
17. P. Das, S.K. Samanta, H. Chattopadhyay, B.B. Sharma, and P. 2015, vol. 46B, pp. 1302–13.
Dutta: Mater. Sci. Technol., 2013, vol. 29 (1), pp. 83–92.

2256—VOLUME 47A, MAY 2016 METALLURGICAL AND MATERIALS TRANSACTIONS A