Materials Transactions, Vol. 44, No. 5 (2003) pp.

845 to 852
Special Issue on Solidication Science and Processing for Advanced Materials
#2003 The Japan Institute of Metals

Inuence of Copper and Iron on Solidication Characteristics

of 356 and 380-Type Aluminum Alloys
Piyada Suwanpinij1 , Usanee Kitkamthorn2 , Ittipon Diewwanit1 and Takateru Umeda1
1

Department of Metallurgical Engineering, Faculty of Engineering, Chulalongkorn University,

Phayathai Road, Pathumwan, Bangkok, 10330 Thailand
2
School of Metallurgical Engineering, Institute of Engineering, Suranaree University of Technology, Nakorn Rachsima, Thailand 30000
The inuence of copper and iron on the solidication characteristics of two major aluminum foundry alloys was investigated. The thermal
history during solidication of each sample was recorded and compared with the solidication path calculated from the multicomponent
equilibrium and Gulliver-Scheil solidication models. SEM/EDX analysis and optical microscope were used to examine the microstructure of
solidied samples. The amount of phases was also calculated from the latter model and compared with the observed microstructure. Binary
interaction parameters were used in calculation for their practicality. Results show that the high content of copper and iron suppresses the
liquiduses and nal solidication temperatures. Moreover, the crystallization of Al2 Cu and Al5 FeSi is very sensitive to copper and iron content
respectively; Al2 Cu increases signicantly when copper is added while Al5 FeSi does greatly when iron content is higher.
(Received November 21, 2002; Accepted March 7, 2003)
Keywords: aluminum foundry alloy, solidication, solidication microstructure, intermetallic compound, solidication path prediction

1.

Introduction

Commercial aluminum alloys contain several alloying

elements and impurities. Some of these elements are able to
form complex intermetallic compounds during solidication
while others may just dissolve in the aluminum or precipitate
from the solid solution as separate phases. The amount and
morphology of these intermetallic compounds are highly
crucial to the mechanical properties of castings. Some of
them may be prevented from crystallization by controlling
the amount of the alloying elements which are the main
components of the compound. However, this may not be the
solution for all cases because most of these compounds are
complex having more than two sublattices with possibility of
replacing some elements in the lattices with others.
Studies in the past attempted to identify the compositions
and crystallographic structures of these intermetallics in
order to understand their crystallization sequences during
solidication.1,2) It is important for making a proper alloy
design to study on the solidication paths. If the crystallization sequences are known, the crystallization of undesirable intermetallics can be prevented.
The microstructure of 356-type aluminum foundry alloy
consists of the common dendritic network of aluminum solid
solution with interdendritic silicon and intermetallic compounds. Major intermetallics in this type are Al5 FeSi, Mg2 Si
and Al8 FeMg3 Si6 . Table 1 lists the solid phases in A356.1
aluminum alloy which solidies at cooling rate between 0.2
and 5 Ks
1 .314) Because of the limited diusion during
solidication and the kinetics of nucleation and growth, the
amount and presence of some phases depend on the cooling
rate. Despite all these, the microstructure of 356-type alloy is
relatively not so complicated as that of other alloys with
higher content of alloying elements.
The properties of another aluminum alloy, 380-type alloy,
are highly dependent on the size and shape of intermetallics.
This is because the Aluminum Association (AA) Standard
allows higher content of iron and manganese in such alloy.

Table 1 Phases observed in A356.1 alloys solidied at cooling rate

between 0.2 and 5 Ks
1 .
Cooling rate (K/s)

Phase

Temperature (K)

Aluminum FCC

885883

0.2

Al15 (Fe,Mn)3 Si2

867844

Al5 FeSi

842841

Mg2 Si

826812

Aluminum FCC

886881

0.6

5.0

Al15 (Fe,Mn)3 Si2

865843

Al5 FeSi

836834

Mg2 Si

825815

Aluminum FCC

887884

Al15 (Fe, Mn)3 Si2

863847

Al5 FeSi

838835

Mg2 Si

821814

Al8 FeMg3 Si6

814778

These two elements can form several intermetallic phases

having plate-like or needle-shaped morphology. The sharp
edges of these phases give detrimental eect to castings by
debasing their strength. Among important intermetallics are
Al15 (Fe,Mn)3 Si2 and Al5 FeSi with the possibility of chromium, copper, and manganese replacing iron atoms in the
iron sublattice. Others are Al2 Cu and Al5 Cu2 Mg8 Si6 . The
amount of these intermetallics and the temperatures at which
they crystallize depend on cooling rate and alloy composition.15,16)
In this study, a thermodynamic phase equilibrium calculation software Thermo-Calc and a thermodynamic database
intended for the calculation of aluminum systems were used
to analyze the solidication paths of 356-type and 380-type
alloys. Binary interaction parameters instead of higher order
interaction parameters were used in the calculation in spite of
multicomponent alloys. To predict the solidication paths of
these alloys accurately, ternary and higher order interaction
parameters are essential. However, very little data are
available in the present and, even in the future, higher order

846

P. Suwanpinij, U. Kitkamthorn, I. Diewwanit and T. Umeda

interaction parameters may not be expected to be well

compiled. Therefore, binary interaction parameters are
expected to be far-reaching and adequately applied in
multicomponent alloy system calculation. The tools mentioned above were rstly applied to the eects of copper and
iron on the microstructure of these two alloys. Such
thermodynamic prediction was then compared to the observed microstructure. Secondly, the amount of two intermetallics, Al5 FeSi and Al8 FeMg3 Si6 , in A356 was discussed
by varying the content of iron, silicon and magnesium within
the composition range of the specication of A356.
2.

Experiment and Thermodynamic Modeling

Six 45-gram-weight samples were cut from commercial

alloy ingots and placed in graphite crucibles with dimension
shown in Fig. 1. A K-type thermocouple was placed in the
wall of each crucible 1.5 mm away from the molten alloy.
Melting operation was done in a resistance furnace lled with
inert argon gas. The melting/alloying temperatures for
reference samples 1 and 4, copper-added samples 2 and 5,
and iron-added samples 3 and 6 were 1023, 1123 and 1223 K
respectively. The chemical compositions of samples were
determined by spark emission spectrometer and were listed in

1.5

Table 2.17) All samples were allowed to cool and solidify in

the furnace while their temperatures were monitored by the
thermocouples connected to a data acquisition unit. At the
beginning, the thermocouple was calibrated to allow error
within 1 K. The average cooling rate for all samples was
0.07 Ks
1 . Solidied samples were sectioned and prepared
for optical and scanning electron microscope (SEM) observation. Phases were determined by EDX.
The thermodynamic analysis of solidication was done by
Thermo-Calc version M and aluminum database version 2.
Gibbs free energy calculation by this software was based on
the Redlich-Kister equation18)
X
X
Gm
xi Gi RT
xi ln xi
i

xi xj

j>i

vij xi
xj v ;

is a binary interaction parameter. The modeling of

where
silicon phase is based on pure substance with no solubility of
any element in the silicon phase. Details on the calculation
method were referred in19,20) while interaction parameters,
and other thermodynamic properties in the database were
referred in the Thermo-Calcs User Guide.
In addition to the equilibrium solidication model, the
Gulliver-Scheil solidication model, on the assumption of no
diusion in solid and complete mixing in liquid, was mostly
used in calculation. The calculation was based on the
following equation;
Cs kC0 1
fs k
1

Graphite Cover

5
24
Sample

30

Graphite Crucible

3.

45

3.1 Solidication of 380-type Alloys

The variation of calculated fraction solid with temperature
in sample 2 is shown in Fig. 2. The calculated values from

Schematic drawing of the graphite crucible.

Sample

where k is equilibrium partition ratio. The content of nine

alloying elements: silicon, copper, iron, manganese, magnesium, chromium, zinc, nickel, and titanium was input as
variables. Afterwards, the calculation was performed to
determine the phases crystallized and their amount as well as
crystallization temperatures at each temperature interval of
0.5 K. The eects of copper and iron were also investigated
by stepping corresponding content of such elements. Moreover, as a guide to control the amount of detrimental phases
in A356 alloy, the variations of Al5 FeSi and Al8 FeMg3 Si6
with some alloying elements were calculated by the GulliverScheil model.

30

Table 2
Alloy

vij

TC

Fig. 1

XX

Condition

Results and Discussion

Chemical composition (mass percent) of the samples in this study.

Si

Cu

Mg

Fe

Mn

Zn

Ni

Ti

Cr

number
380

356

Reference

8.21

2.37

0.03

1.01

0.13

0.08

0.05

0.05

0.03

Cu-added

7.87

4.62

0.29

1.01

0.13

0.86

0.05

0.04

0.03

Fe-added

7.96

2.41

0.29

1.32

0.11

0.88

0.05

0.05

0.03

Reference

7.23

0.02

0.48

0.16

0.002

<0:001

0.003

0.14

0.003

Cu-added

6.29

0.17

0.40

0.07

0.004

<0:001

0.001

0.13

0.002

Fe-added

6.43

0.01

0.47

0.99

0.006

0.017

<0:001

0.13

0.002

Inuence of Copper and Iron on Solidication Characteristics of 356 and 380-Type Aluminum Alloys

847

1000

Gulliver-Scheil fraction solid

880

aluminum FCC

equilibrium fraction solid

950

Temperature, T/K

Temperature, T/K

860

840

silicon
820

800

Solidification
completes (equilibrium)

780

Al15(Fe,Mn)3Si2 starts
Aluminum FCC starts
Al FeSi starts
5
Silicon starts

900

850

800

750

Solidification completes (Scheil)

760
0.2

0.0

0.4

0.6

0.8

700
10000

1.0

Fraction Solid

Temperature, T/K

900

Al (Fe,Mn) Si
15
3 2
aluminum FCC
Al FeSi
5

silicon
800

Al Cu
2

Al5Cu2Mg8Si6
Solidification
completes

750

700
2000

2500

3000

3500

4000

20000

25000

30000

Enthalpy of the Whole System, J/mol

Fig. 2 Calculated weight fraction versus temperature of sample number 2

(Cu-added 380-type alloy).

850

15000

4500

5000

Elapsed Time, t/s

Fig. 3 The observed cooling curve of sample 2 (Cu-added 380-type alloy)
with arrows indicating the calculated crystallization temperatures from the
Gulliver-Scheil model.

both the equilibrium and Gulliver-Scheil solidication

models are almost identical except the temperatures at which
solidication completes. Table 3 and Fig. 3 show the
calculated crystallization temperatures of major phases in

Fig. 4

The enthalpy change of the whole alloy system in sample 2.

the same sample by the Gulliver-Scheil model and compare

them with the experimental results. Due to the limit of
sensitivity in the measurement, only three temperatures, at
which relatively large enthalpy changes occur, could be
detected; they are the temperatures at which aluminum FCC
and silicon start to crystallize and the solidication completes. When Al15 (Fe,Mn)3 Si2 crystallized, abrupt change
could not be detected in the cooling curve, corresponding to
the small enthalpy change as shown in Fig. 4.
Discrepancies between the observed and calculated temperatures for the crystallization of aluminum FCC appear in
Fig. 3. The aluminum FCC crystallized starts to crystallize at
849 K which is 14 K lower than the calculated value. Such
inaccuracy other than the undercooling occurs from that the
thermocouples were not placed in molten alloys but in the
walls of crucibles. The other source of inaccuracy might be
attributed to the Gibbs free energy of the liquid and
aluminum FCC phases. The inaccuracy of the Gibbs free
energy is due to the scarcity of reliable experimental data on
higher order interaction parameters. The present thermodynamic was based on binary interaction parameters only.
Because of the high concentration of solutes in 380 alloy,
contribution from these higher order interactions leads to the
error in the calculated Gibbs free energy of the liquid and

Table 3 Solidication sequence of sample 2 (Cu-added 380-type alloy) and weight percentage of each phase at complete solidication
temperature.
Phase

Calculated

Observed

Crystallization
temperature (K)

Amount
(mass%)

Al15 (Fe,Mn)3 Si2

881.6

1.12

Aluminum FCC

863.1

82.83

Al5 FeSi

857.7

2.90

Silicon

835.0

6.06

Al7 Cu2 Fe

792.6

negligible

Al7 Cu4 Ni

786.9

negligible

Al2 Cu

784.7

4.94

Al5 Cu2 Mg8 Si6

778.2

0.54

Solidication completed

764.7

Y: Observed in solidication microstructure

N: Not observed in solidication microstructure

Crystallization
temperature (K)

Microstructure
Y

849.0

Y
Y

833.7

P. Suwanpinij, U. Kitkamthorn, I. Diewwanit and T. Umeda

Calculated Temperature from Scheil Model, T/K

848

900
890
880

y=x
fcc aluminum

870
860
850
840

silicon

830
820
820

830

840

850

860

870

880

890

900

Observed Temperature, T/K

Fig. 5 Comparison between the calculated temperatures from the GulliverScheil model and the observed temperatures of aluminum FCC and silicon
in 380-type from various literatures.

aluminum FCC phases. However, binary interaction parameters are practical while the higher order ones derive from
laborious work and consumed time.
The observed crystallization temperatures of aluminum
FCC and silicon from other literatures316,2022) were
collected and compared with the calculated thermodynamic
results of which chemical compositions were input. Compared in Fig. 5, almost all the calculated temperatures for
aluminum FCC are higher than the observed ones not more
than 15 degree. However, the calculated temperatures for
silicon deviate from the observed temperatures 5 degree,
corresponding to this study.
The microstructure of solidied 380-type alloys (sample 1,
2, and 3) contains aluminum FCC, silicon, Al15 (Fe,Mn)3 Si2 ,
Al5 FeSi, Al2 Cu, and Al5 Cu2 Mg8 Si6 . The quantity of phases
in sample 2 was also calculated and is compared with the
detection by SEM with EDX analysis in Table 3. Al7 Cu2 Fe
and Al7 Cu4 Ni, which were predicted to have negligible
amount, could not be detected in the solidied microstructure. The electron micrographs of sample 2 are shown in
Fig. 6. The Al15 (Fe,Mn)3 Si2 phase (usually called the alpha
phase) crystallizes in two dierent morphologies: dense
idiomorphic and Chinese script as in Figs. 6(a) and (b)
respectively. The former is the primary phase that crystallizes
and normally termed sludge. Certain sedimentation of the
sludge Al15 (Fe,Mn)3 Si2 phase was found in all samples in
380 series due to the slow cooling rate used in this study. The
Al5 FeSi phase (usually called the beta phase) appears as
platelets or needles which are very distinct (Fig. 6(c)). The
eutectic phase Al2 Cu seems agglomerated pink bubbles (Fig.
6(c)).
For sample 3, the variations of the predicted amount of
intermetallics and calculated crystallization temperature of
major phases with copper are revealed in Figs. 7(a) and (b)
respectively. The quantity of Al2 Cu increases greatly with

Fig. 6 Electron micrographs of sample number 2 (Cu-added 380-type

alloy); 6(a): Al15 (Fe,Mn)3 Si2 which appears in dense idiomorphic, 6(b):
Al15 (Fe,Mn)3 Si2 in Chinese script morphology, 6(c): Al15 FeSi in needle
shape and bubble-like Al2 Cu.

copper content while others varies slightly. The crystallization temperatures of aluminum FCC (liquidus temperature) and silicon decline moderately with copper content
while those of Al15 (Fe,Mn)3 Si2 and Al5 FeSi remain almost
constant. The crystallization temperature of Al5 Cu2 Mg8 Si6
goes down signicantly at copper content 1.03.0 mass% and
decreases slightly at 3.06.5 mass%. In the range of 0.0 to
1.0 mass% of copper, the nal solidication temperature
decreases signicantly and then rises up to some degree and
decreases slightly.
In the same sample, the relationships between the
predicted amount of intermetallics and calculated crystallization temperatures of major phases with iron are shown in
Figs. 8(a) and (b). With increasing iron content, the amount
of Al5 FeSi rises up remarkably but saturates at about
2 mass% Fe and then decreases moderately. Al8 Fe2 Si does
not crystallize until the iron content is added up to 2.1 mass%
and then increases abruptly. The crystallization temperatures
of aluminum FCC, silicon, Al5 Cu2 Mg8 Si6 , Al2 Cu and the
complete solidication temperature remain unchanged while
those of Al15 (Fe,Mn)3 Si2 and Al5 FeSi increase substantially
and saturated at 2.2 mass% Fe.

Inuence of Copper and Iron on Solidication Characteristics of 356 and 380-Type Aluminum Alloys

Al (Fe,Mn) Si
15

Al Cu

Al Fe Si

Al Cu Mg Si

Al (Fe,Mn) Si

Al FeSi
5

Al Cu

15

849

Al Cu Mg Si

Al FeSi

8.0

(a)
7

5.0
4.0
3.0
2.0
1.0
0.0
0.0

2.0

3.0

Al (Fe,Mn) Si
15

4.0

5.0

6.0

7.0

0
0

0.5

Al Cu

1.5

2.5

3.5

Fe Content, (mass%)

Al5 FeSi

Al5 Cu2Mg8 Si6

aluminum FCC
silicon

Solidification completed

(b)

Al8 Fe2 Si

Al5 FeSi

aluminum FCC

Al Cu Mg Si

silicon

Al2 Cu

Al15 (Fe,Mn) 3 Si2

Solidification completes

880

950 (b)
860

840

820

800

780

760
0

Cu Content, (mass%)
Fig. 7 Eects of copper content on the solidication characteristics of 380type alloy predicted from the Gulliver-Scheil model. (Fe 1.32%, Si 7.96%,
Mg 0.29%, Mn 0.11%, Zn 0.88%, Ni 0.05%, Ti 0.05%, Cr 0.03%); (a) The
relationship between the amount of intermetallics and copper content., (b)
The relationship between the crystallization temperatures of major phases
and copper content.

3.2 Solidication of 356 alloys

Similarly to 380 type, the calculated fraction solid of
sample 5 from the equilibrium and Gulliver-Scheil models in
Fig. 9 is very similar except near the end of solidication.
Table 4 and Fig. 10 compare the calculated crystallization
temperatures of major phases with the experimental result. In
the experiment, only the crystallization of aluminum FCC
and silicon could be observed.
The information about the discrepancy of crystallization
temperature, which was collected from literatures,314,2126)
is revealed in Fig. 11.
Figure 12 shows the solidication microstructure of 356
reference-sample 4. There are aluminum FCC, silicon,
Al5 FeSi, and Al8 Mg3 FeSi6 . The quantity of phases in sample
5 was calculated and is compared in Table 4 with the
detection by SEM with EDX analysis. In contrast, Mg2 Si can
be detected in sample 5 by SEM. The amount of interme-

Crystallization Temperature, T/K

Crystallization Temperature, T/K

1
1.0

Cu Content, (mass%)

900

(a)

6.0

Mass Percent of Phases

Mass Percent of Phases

7.0

900

850

800

750
0

0.5

1.5

2.5

3.5

Fe Content, (mass%)
Fig. 8 Eects of iron content on the solidication characteristics of 380type alloy calculated from the Gulliver-Scheil model. (Cu 2.41%,
Si 7.96%, Mg 0.29%, Mn 0.11%, Zn 0.88%, Ni 0.05%, Ti 0.05%, Cr
0.03%); (a) The relationship between the amount of intermetallics and iron
content., (b) The relationship between the crystallization temperatures of
major phases and iron content.

tallics and the crystallization temperatures of major phases

varying with copper for sample 5 are shown in Figs. 13(a)
and (b). Al5 FeSi increases slightly with copper. Al2 Cu does
not appear until 0.6 mass% Cu and then increases substantially. The crystallization temperatures of aluminum FCC,
silicon and Al5 FeSi go lower slightly with increasing copper
content. In contrast, the nal solidication temperature
declined substantially.
Figures 14(a) and (b) show the variation of the calculated
amount of intermetallics and crystallization temperatures of
major phases with iron content for sample 5. According to
Fig. 14(a), Al15 (Fe,Mn)3 Si2 crystallizes in so negligible
amount even at very high content of iron that can not be
shown, corresponding to the observed microstructure. The
amount of Al5 FeSi is very sensitive to iron content. Its

850

P. Suwanpinij, U. Kitkamthorn, I. Diewwanit and T. Umeda

900

950

Gulliver-Scheil fraction solid

equilibrium fraction solid

aluminum FCC

Temperature, T/K

Temperature, T/K

880

860

silicon
840

Solidification
completes (equilibrium)

900

aluminum FCC
silicon

850

800
820

Solidification
completes (Scheil)

800
0.0

0.2

0.4

0.6

0.8

750
2000

1.0

Fraction Solid

Al8FeMg3Si6

Al FeSi

2500

Mg Si
2
Solidification
completes (Scheil)

3000

3500

4000

4500

Elapsed Time, t/s

Fig. 9 Calculated weight fraction versus temperature of sample number 5

(Cu-added 356-type alloy).

Fig. 10 Observed cooling curve of sample 5 (Cu-added 356-type alloy)

with arrows indicating the calculated crystallization temperatures from the
Gulliver-Scheil model.

Table 4 Solidication sequence of sample 5 (Cu-added 356-type alloy) and mass percentage of each phase at complete solidication
temperature.
Phase

Calculated

Observed

Crystallization

Amount

Crystallization

temperature (K)

(mass%)

temperature (K)

894.4

93.67

885.2
836.4

Aluminum FCC

Microstructure
Y

Silicon

845.9

4.83

Al5 FeSi

836.3

0.12

Al8 FeMg3 Si6

831.4

0.35

Mg2 Si

826.2

0.20

Al15 (Fe,Mn)3 Si2

816.8

0.0003

Solidication completed

815.2

Y: Observed in solidication microstructure

N: Not observed in solidication microstructure

Temperature from Scheil Model, T/K

900

fcc aluminum
890
880

y=x

870
860

silicon

850
840
830
820
820

830

840

850

860

870

880

890

900

Temperature from Literature, T/K

Fig. 11 Comparison between calculated temperatures from the GulliverScheil model and observed temperature of aluminum FCC and silicon in
356-type in various literatures.

Fig. 12 The observed microstructure of sample 4 (reference 356-type

alloy).

Inuence of Copper and Iron on Solidication Characteristics of 356 and 380-Type Aluminum Alloys
Al5 FeSi

Al5 Cu2 Mg8 Si6

Mg Si

Al Cu

Al FeMg Si
8

Mg2Si

Al5FeSi

Al8FeMg3Si6

Al8Fe2Si

851

(a)

Mass Percent of Phases

Mass Percent of Phases

(a)
0.8

0.6

0.4

1
0.2

0
0

0
0

0.2

0.4

0.6

0.8

0.2

0.4

Mg Si

Al FeMg Si

Al Cu

Solidification completes
Al Cu Mg Si

Crystallization Temperature, T/K

900

880

(b)

860

840

820

800

780
0

0.4

0.6

0.8

1.2

1.4

1.6

Mg2Si
Solidification completed
Al8Fe2Si

(b)
880

860

840

820

800
0
0.2

900

Crystallization Temperature, T/K

Aluminum FCC
Silicon
Al5FeSi
Al8FeMg3Si6

0.8

Fe Content, (mass%)

Cu Content, (mass%)
aluminum FCC
silicon
Al FeSi

0.6

1.2

1.2

0.5

1.5

Fe Content, (mass%)

Cu Content, (mass%)
Fig. 13 Eects of copper content on the solidication characteristics of
356-type alloy predicted from the Gulliver-Scheil model. (Fe 0.07%, Si
6.286%, Mg 0.403%, Mn 0.004%, Zn < 0:001%, Ti 0.130%, Cr 0.002%);
(a) The relationship between the amount of intermetallics and copper
content., (b) The relationship between the crystallization temperatures of
major phases and copper content.

amount rises up 3.5 mass% per 1.0 mass% Fe increment in

the rst range. Beyond 1.2 mass% Fe, its amount saturates
and decreases slightly. In Fig. 14(b), the Al5 FeSi temperature
increases signicantly at rst but gradually decelerates and
nally saturates from 1.3 mass% Fe. On the other hand, the
liquidus (aluminum FCC temperature) and the complete
solidication temperature as well as the crystallization
temperature of silicon negligibly changed.
As a guide to control the amount of detrimental phases in
A356 alloy, which are needle-shaped Al5 FeSi and Chinese
script Al8 FeMg3 Si6 , their variations in amount with increasing content of silicon, magnesium and iron were calculated
by the Gulliver-Scheil model and are shown in Figs. 15(a)
and (b). The content of these alloying elements input in the
thermodynamic calculation was within a composition range
according to the specication of A356 alloy and other

Fig. 14 Eects of iron content on the solidication characteristics of 356type alloy predicted from the Gulliver-Scheil model. (Cu 0.17%, Si
6.286%, Mg 0.403%, Mn 0.004%, Zn < 0:001%, Ti 0.130%, Cr 0.002%);
(a) The relationship between the amount of intermetallics and iron
content., (b) The relationship between the crystallization temperatures of
major phases and iron content.

elements were set to be zero. At 0.25 mass% Mg, Fig. 15(a),

the needle-like Al5 FeSi starts to crystallize at 0.025 mass%
Fe and then increases substantially and linearly with iron
content. Its amount at 6.5 mass% Si is a slightly higher than
those at 7.0 and 7.5 mass% Si, which are almost exactly equal
along the calculation. The other phase, Al8 FeMg3 Si6 ,
increases linearly until 0.03 mass% Fe and then this phase
stays at about 0.2 mass%. For 0.45 mass% Mg, Fig. 15(b),
Al5 FeSi starts to crystallize at 0.04 mass% Fe and then
increases substantially and linearly, similarly to the case of
0.25 mass% Mg. The amount of Al5 FeSi at 6.5 mass% Si is
slightly higher than those at 7.0 and 7.5 mass% Si.
Al8 FeMg3 Si6 increases linearly from zero until 0.06 mass%
Fe. After that, it stays at about 0.4 mass%, larger amount than
those in 0.25% Mg alloy. The amount of Al8 FeMg3 Si6 is
higher in the order of 7.0 mass% Si, 7.5 mass% Si and
6.5 mass% Si.

852

P. Suwanpinij, U. Kitkamthorn, I. Diewwanit and T. Umeda

Al FeSi at Si 6.5%

Al8 FeMg3 Si6 at Si 6.5%

Al FeSi at Si 7.0%

Al FeMg Si at Si 7.0%

Al FeSi at Si 7.5%

Al FeMg Si at Si 7.5%

5
5

aluminum FCC due to the position of thermocouple and

the scarcity of experimental and thermodynamic data
on higher order interaction parameters of the thermodynamic modeling. However, in the present, higher
order interaction parameters can not be well compiled.
Binary interaction parameters are expected to be
adequate in multicomponent alloy system calculation.

(a) 0.6

Mass Percent of Phases

0.5

0.4

Acknowledgements

0.3

0.2

0.1

0
0

0.05

0.1

0.15

0.2

Fe Content, (mass%)
Al FeSi at Si 6.5%

Al8 FeMg3 Si6 at Si 6.5%

Al FeSi at Si 7.0%

Al FeMg Si at Si 7.0%

Al5 FeSi at Si 7.5%

Al8 FeMg3 Si6 at Si 7.5%

5
5

0.6

REFERENCES

(b)

Mass Percent of Phases

0.5

0.4

0.3

0.2

0.1

0
0

0.05

0.1

0.15

0.2

Fe Content, (mass%)
Fig. 15 The variation in amount of Al5 FeSi and Al8 FeMg3 Si6 phases with
iron and silicon content within A356 specication; (a) at magnesium
0.25 mass%, (b) at magnesium 0.45%.

4.

The authors would like to acknowledge the indispensable

support of MTEC (National Metal and Materials Technology
Center) and TJTTP (Thai-Japan Technology Transfer Project). We are also grateful to Asahi Somboon Aluminium
Co., Ltd. and Messrs Tetsuo Moriya and Tatsuo Kawara for
the chemical analyses of the alloys in this paper. We also
thank Mr. Suvanchai Pongsukitwat for his useful advice. In
addition, we owe a debt of gratitude for the experiment
preparation to Mr. Senee Maneepetch.

Conclusions

(1) In this study, crystallization sequence of 380-type

alloys was Al15 (Fe, Mn)3 Si2 ! aluminum FCC !
Al5 FeSi ! silicon ! Al2 Cu ! Al5 Cu2 Mg8 Si6 ;
crystallization sequence of 356-type alloys was aluminum FCC ! silicon ! Al5 FeSi ! Al8 FeMg3 Si6 !
Mg2 Si.
(2) When copper and iron content increases, liquidus and
eutectic temperature of 356 and 380 were suppressed.
(3) The quantity of Al2 Cu increased greatly with the
increasing copper and that of Al5 FeSi increased greatly
with iron content.
(4) Thermodynamic prediction provided an eective way
in studying the eects of alloying elements.
(5) There were small discrepancies between the observed
and calculated temperatures for the crystallization of

1) A. L. Dons: Z. Metallk. 75 (1984) 170174.

2) J. Lacaze, G. Lesoult, O. Relave, I. Ansara and J. P. Riquet: Z. Metallk.
78 (1987) 141150.
3) L. Backerud, G. Chai and J. Tamminen: Solidication Characteristics
of Aluminum Alloys, (American Foundrymens Society, Inc., 1990)
pp. 128130.
4) ibid., pp. 136138.
5) ibid., pp. 152.
6) ibid., pp. 159.
7) ibid., pp. 165.
8) ibid., pp. 171.
9) ibid., pp. 177.
10) ibid., pp. 183.
11) ibid., pp. 189.
12) ibid., pp. 193.
13) ibid., pp. 197.
14) ibid., pp. 203.
15) S. Gowri and F. H. Samuel: Metall. Mater. Trans. A 25A (1994) 437
448.
16) F. H. Samuel, G. Pucella, C. Villeneuve, A. M. Samuel, H. W. Doty and
S. Valtierra: Int. J. Cast Metals Res. 12 (1999) 197210.
17) In the experiment, the Mg and/or Cu content in sample 13 was not
corresponding to A380 specication while Mg and/or Si in sample 46
and Fe in sample 6 was not corresponding to A356 specication.
Therefore, they were called 380-type and 356-type alloys respectively.
18) N. Saunders and A. P. Miodownik: CALPHAD (Calculation of Phase
Diagram): A Comprehensive Guide, (Pergamon Press, 1998) pp. 91
126.
19) M. Hillert: Phase Equilibria, Phase Diagram and Phase Transformations; a Thermodynamic Basis, (Cambridge University Press, Cambridge, 1998) pp. 167202.
20) K. Ohsasa, M. Shoji and T. Narita: J. JFS 72 (2000) 525529.
21) L. Arnberg and L. Backerud: Solidication Characteristics of Aluminum Alloys Vol. 3 Dendrite, (American Foundrymens Society, Inc.,
1996) pp. 9596.
22) ibid., 115117.
23) S. Gowri: AFS Transactions 102 (1994) 503508.
24) E. N. Pan, C. N. Lin and C. P. Loper, Jr: AFS Transactions 98 (1990)
735746.
25) R. I. Mackay and J. E. Gruzleski: J. Cast Metals Res. 10 (1997) 131
145.
26) R. I. Mackay and J. E. Gruzleski: J. Cast Metals Res. 10 (1998) 255
266.