Materials Science and Technology

ISSN: 0267-0836 (Print) 1743-2847 (Online) Journal homepage: [Link]

Microstructure and microhardness in horizontally

solidified Al–7Si–0.15Fe–(3Cu; 0.3Mg) alloys

F. Souza, J. Lima, C. Rizziolli, I. Magno, A. Barros, A. Moreira & O. Rocha

To cite this article: F. Souza, J. Lima, C. Rizziolli, I. Magno, A. Barros, A. Moreira & O. Rocha
(2018): Microstructure and microhardness in horizontally solidified Al–7Si–0.15Fe–(3Cu; 0.3Mg)
alloys, Materials Science and Technology, DOI: 10.1080/02670836.2018.1444923

To link to this article: [Link]

Published online: 08 Mar 2018.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

[Link]
MATERIALS SCIENCE AND TECHNOLOGY, 2018
[Link]

Microstructure and microhardness in horizontally solidified

Al–7Si–0.15Fe–(3Cu; 0.3Mg) alloys
F. Souzaa,b , J. Limaa , C. Rizziollia , I. Magnoa,b , A. Barrosb , A. Moreirab and O. Rocha a,b

a Federal Institute of Education, Science and Technology of Pará, IFPA, Belém, Brazil; b Institute of Technology, Federal University of Pará,
UFPA, Belém, Brazil

ABSTRACT ARTICLE HISTORY

Transient horizontal directional solidification (THDS) experiments have been carried out Received 4 November 2017
with Al–7wt.%Si–0.15Fe, Al–7wt.%Si–3wt.%Cu–0.15wt.%Fe and Al–7wt.%Si–0.3wt.%Mg– Revised 15 January 2018
0.15wt.%Fe alloys, to identify experimental relationships between growth rates (GR ), cooling Accepted 16 February 2018
rates (C R ), tertiary dendrite arm spacings (λ3 ) and microhardness (HV). Optical microscopy KEYWORDS
and scanning electron microscopy/energy-dispersive spectrometry (SEM/EDS) were used to Cast aluminium alloys;
perform a comprehensive microstructural characterisation of the β-Al5 FeSi, ω-Al7 Cu2 Fe, θ- transient horizontal
Al2 Cu, π-Al8 Mg3 FeSi6 and α-Mg2 Si intermetallic phases. The addition of Cu and Mg to solidification; dendritic
the Al–7wt.%Si–0.15wt.%Fe alloy led to the precipitation of ω and π phases from the β microstructure; intermetallic
phase. It has been found for all analysed alloys that power experimental functions given by phases; microhardness
λ3 = constant.(GR )−1.1 and λ3 = constant.(C R )−0.55 best describe the variation of λ3 with cor-
responding thermal and microstructural parameters.

Introduction
intermetallic phases, and solidification defects [13]. It
Owing to excellent castability, good specific strength, is important to mention that despite the literature pre-
low coefficient of thermal expansion as well as good senting several theoretical and experimental papers
wear resistance, Al-based alloys of the ANSI series 319.1 focusing on the interconnections among solidification
and 333.1 (in which compositions lie mostly within thermal parameters, λ1 and λ2 of binary aluminium
the ranges 5.5–10%Si and 3–4%Cu) have a wide range alloys, there is still a large research gap to be filled
of applications in automotive and aeronautics indus- on the growth of tertiary dendritic branches, espe-
tries such as manufacture of cylinder heads, pistons cially for multicomponent aluminium alloys of great
and engine blocks [1–4]. Al–7Si–Mg cast alloys (typ- industrial interest [14]. According to Sá et al. [39], the
ically A356 and A357 alloys) are equally extensively spacing between secondary and tertiary arms can effec-
applied in both aerospace and automotive industries tively serve to isolate a complex dispersion of second
due to their excellent formability, high corrosion resis- phases significantly influencing the mechanical proper-
tance, and good comprehensive mechanical properties ties. Thus, since the dendritic microstructure has been
[5–7]. These characteristics of Al–Si–(Cu, Mg) alloys characterised and the tertiary ramifications observed,
have attracted the attention of investigators with a the dependencies of the mechanical properties must
view to analysing a mutual interaction among solid- always be reported as a function of λ3 .
ification processing conditions, alloying composition, It is well known that low Fe concentration may be
microstructure evolution and mechanical properties of present in Al–Si alloys and may induce the formation
such alloys. of hard brittle plates of the β-Al5 FeSi phase which is
A number of studies [2–46] have highlighted the particularly deleterious to the alloy’s mechanical prop-
effect of casting thermal parameters, such as growth erties [15]. The addition of Cu as an alloying element
rate (GR ) and cooling rate (CR ), on primary (λ1 ), to Al–Si alloys benefits some of the mechanical proper-
secondary (λ2 ) and tertiary (λ3 ) dendrite arm spac- ties, such as yield strength and ultimate tensile strength
ings (usually used to describe the fineness of den- (UTS) [15,16]. Copper forms an intermetallic phase
dritic structures) as a significant factor determining with Al which, depending on the cooling conditions,
the mechanical properties of Al–Si–(Cu, Mg) alloys can precipitate either as blocky Al2 Cu or in an eutec-
[8–12]. In general, this has been attributed to the tic form as (α-Al)eutectic + Al2 Cu [17,18]. According to
fact that microstructures characterised by smaller den- Emma and Salem [19], the addition of Cu increases
dritic spacings, provided by higher CR and GR values, considerably the strength of Al–Si alloys due to precip-
have a more homogeneous distribution of eutectic, itation of the Al2 Cu intermetallic phase (θ ) dispersed

CONTACT O. Rocha [Link]@[Link] Federal Institute of Education, Science and Technology of Pará, IFPA, Almirante Barroso Avenue
1155, 66093-020 Belém, PA, Brazil; Institute of Technology, Federal University of Pará, UFPA, Augusto Correa Avenue 1, 66075-110 Belém, PA, Brazil

© 2018 Institute of Materials, Minerals and Mining.

2 F. SOUZA ET AL.

in the Al-rich dendritic matrix during aging. On the describe the Vickers microhardness (HV) dependence
other hand, during solidification of Al–Si–Cu alloys, on the λ2 values, given by HV = 169(λ2 )−0.17 and
in the presence of iron, other phases may be formed HV = 61 + 183(λ2 )−0.5 respectively, for the Al–5.5wt.%
such as ω-Al7 Cu2 Fe. In the case of Al–Si–Mg alloys, Si–3wt.%Cu ternary alloy solidified horizontally under
Mg and Si are elements in Al alloys that are able to transient conditions. The authors have also com-
form Mg2 Si [20]. Al–(5–12)wt.%Si casting alloys with pared the obtained results with those achieved for
the addition of magnesium ranging from 0.2 to 0.5wt.% Al–3wt.%Cu and Al–3wt.%Si alloys solidified under
allow the hardening of the matrix by precipitation of the same conditions and have observed that the com-
second phases, with emphasis on the Mg2 Si phase, after bined addition of both Cu and Si in the aluminium
heat treatments of solution and artificial aging [20,21]. matrix allowed to obtain higher HV values in Al–5.5wt-
However, the presence of iron in Al–Si–Mg alloys may %Si–3wt-%Cu ternary alloy.
result in the appearance of the π-Al8 Mg3 FeSi6 phase Many investigations on the microstructural char-
which is difficult to dissolve during solid solution heat acteristics of casting alloys have been developed in
treatment [22]. the last decades; however, several aspects of phys-
With the application of the directional solidifica- ical nature related to the formation of products
tion (DS) technique, a single ingot can be solidified microstructure obtained by casting remain still to be
enabling to experience a wide spectrum of microstruc- effectively understood especially for Al-based multi-
tural length scales, morphologies and intermetallic par- component alloys solidification by the THDS tech-
ticles distribution [23–27]. The vast majority of stud- nique. With this in mind, one of the aims of this
ies on DS have been carried out considering the chill study is to analyse the growth of tertiary dendrite
placed on the bottom of the mould (upward direction) arm spacings of Al–7wt.%Si–0.15wt.%Fe, Al–7wt-
[28–34]. The main reason is that in the case of vertical %Si–3wt.%Cu–0.15wt.%Fe, and Al–7wt.%Si–0.3wt.
upward DS, the effect of convection flows can be min- %Mg–0.15wt.%Fe alloys as a function of both GR and
imised, promoting a better stability to the solidification CR values experienced during THDS. The effects of GR ,
process. On the other hand, when the chill is placed on CR and λ3 on the HV results of these alloys, consid-
the side of the mould (horizontal DS), convection flow ering different intermetallic phases distributed within
as a function of both thermal and composition gradi- interdendritic regions, are also investigated.
ents in the liquid occurs [35–38]. As a consequence, the
microstructural development during the transient hor-
Experimental
izontal directional solidification (THDS) of Al–Si–(Cu,
Mg) alloys may be considered as a complex and inter- Al–7wt.%Si–0.15wt-%Fe, Al–7wt.%Si–3wt.%Cu–0.15
esting phenomenon to be studied. wt.%Fe and Al–7wt.%Si–0.3wt.%Mg–0.15wt.%Fe
In the particular case of Al–Si–Mg and Al–Si–Cu alloys have been prepared from high-purity Al, Si, Cu,
alloy systems, the first one is highlighted by the fact and Mg elements ( > 99.7%). Quantitative and quali-
that the solutes’ combination (Si and Mg) is lighter tative chemical analyses were performed by Q4 TAS-
than the solvent, and its rejections cause an interden- MAN – CCD Based Optical Emission Spectrometer for
dritic liquid, close to the solidification interface, also metal analysis. It has been observed that the Fe content
lighter, when compared to the remaining global liquid in the investigated alloys was approximately 0.15wt.%,
volume. In the other case, the presence of Cu will form derived from primary aluminium. The alloys have been
an interdendritic liquid of higher density than the over- directionally solidified in a water-cooled horizontal
all liquid volume of the alloy. Despite the large number solidification device which has been designed in such
of solidification studies in the literature, there is still a a way that the heat has been extracted only through a
large investigation gap with respect to the influence of cooling system placed in the stainless steel lateral wall
thermo-solutal convection on the solidification thermal mould (3 mm thick). The schematic of the horizontal
parameters and, consequently, its effects on the scale of solidification device can be seen in Figure 1. Water flux
microstructure length. has kept the stainless steel plate at a constant tempera-
Vasconcelos et al. [35] have studied the ture of about 25°C during solidification. In addition, the
interconnection between solidification thermal and thermal contact condition at the metal–mould interface
microstructural parameters of horizontally solidified was also standardised with the heat-extracting surface
A–6wt.%Cu and A–6wt.%Cu-8wt.%Si alloys, whose being polished. Further details about the solidification
results have showed that the incorporation of Si system used in the present research have been also
in Al–6wt-%Cu alloy has affected significantly the described in previous studies [8,37].
microstructure and consequently the microhardness; The temperature measurements have been per-
however, λ1 growth laws as a function of GR and CR formed using fine type K thermocouples previously cal-
with exponents −1.1 and −0.55, respectively, have not ibrated, located at different positions along the casting
changed. Araújo et al. [37] have recently proposed length with respect to the metal–mould interface con-
power and Hall-Petch type experimental equations to sidering the central middle line, connected by coaxial
 MATERIALS SCIENCE AND TECHNOLOGY 3

Figure 1. The horizontal DS device applied in the present study.

cables to a data acquisition system controlled by a and mechanically polished with abrasive papers, pol-
computer. The registered time–temperature data dur- ished and etched with a solution of 5% NaOH in
ing THDS have been used to provide plots of position water during 15 s. Image processing system Olym-
from the metal–mould interface and the corresponding pus BX51 and ImageJ software were used to mea-
time of the solidification front passing by each thermo- sure λ3 values on the transverse sections by aver-
couple which have been represented by a power-type aging the distance between adjacent side branches
mathematical expression given by P = constant (t)n , [18,39–41]. At least 30 measurements were performed
where ‘P’ is the liquidus isotherm displacement and ‘t’ is for each selected position along the casting length.
the time corresponding to the passage of this isotherm Figure 2 shows the λ3 measurement technique applied
for each thermocouple. The derivative of this equation in this investigation. A scanning electron microscope
has allowed to obtain the GR profile, i.e. GR = dP/dt. (SEM Shimadzu, VEGA 3 SBU TESCAM) coupled to
The CR values have been determined by the deriva- an energy dispersion spectrum (EDS AZTec Energy
tive of each curve [T = f (t)] considering the thermal X-Act, Oxford) was also used during microstruc-
data recorded immediately after the passage of liquidus tural characterisation. The measured λ3 values, for
isotherm for each thermocouple. the Al–7wt.%Si–3wt.%Cu–0.15wt.%Fe, have been pre-
Thermo-Calc software has been used to generate sented in a recently published paper [44].
the solidification path for the analysed alloys con- Finally, microhardness measurements were carried
sidering the Scheil model. For micrograph exam- out using a Shimadzu HMV-2 model hardness mea-
ination, transverse sections (perpendicular to the suring test device using a 50 g load and a dwell time
growth direction) of the directionally solidified spec- of 10 s. The adopted hardness value of a representa-
imens have been extracted along the columnar zone tive position was the average of at least 20 different
4 F. SOUZA ET AL.

Figure 2. Schematic illustration of λ3 measurements at transverse sections, in which n is the number of tertiary arms.

Figure 3. Experimental cooling curves for Al–7wt.%Si–0.15wt.%Fe–(3wt.%Cu and 0.3wt.%Mg) alloys at different positions from the
heat-extracting surface.

regions on transverse sections. The measured λ3 values and the derivative of these functions with respect to
for the Al–7wt.%Si–3wt-%Si–0.15wt.%Fe, have been time has provided GR values, as shown in Figure 4(a).
presented in our recently published paper [44]. Figure 4(b) shows the evolution of the experimental CR
profiles that have been calculated by considering the
thermal data recorded immediately after the passing of
Results and discussion the liquidus front by each thermocouple. It is empha-
Temperature–time profiles for the Al–7wt.%Si– sised that the results of the thermal analysis for the
0.15wt.%Fe, Al–7wt.%Si–3wt.%Cu–0.15wt.%Fe, and Al–7wt.%Si–0.15wt.%Fe–(3wt.%Cu) alloys have been
Al–7wt.%Si–0.3wt.%Mg–0.15wt.%Fe alloys are pre- recently published [44,46], which have been used in this
sented in Figure 3. For all investigated alloys, each work to investigate the HV dependence on GR , CR , and
experimental cooling curve refers to a thermocou- λ3 values.
ple inserted into the horizontally solidified casting at The profiles’ evolution of thermal parameters dur-
a specific distance from the metal–mould interface. ing the horizontal solidification has allowed to observe
A curve fitting technique on experimental points of the decrease in the GR and CR values with the advance
position from the metal–mould interface as a func- of the solidification front. This can be attributed to the
tion of time corresponding to the liquidus front pass- heat-conducting heat resistance, which increases with
ing by each thermocouple generated a power function the formation of the solidification layer.
 MATERIALS SCIENCE AND TECHNOLOGY 5

Additionally, Barros et al. [42] have found exponen-

tial relationships given by λ3 = constant (GR )−1.1 and
λ3 = constant (CR )−0.55 to correlate the λ3 growth,
respectively, with GR and CR for the Al–3wt.%Cu alloy
produced by the THDS technique. Thus, the obtained
results in this investigation reinforce the proper appli-
cation of the exponents −0.55 and −1.1 to correlate the
λ3 evolution with CR and GR , respectively, for every
non-ferrous alloy systems independent of the growth
direction.
It can be evidenced in this work that index
equal to −0.55 found in the mathematical expres-
sion λ3 = f (CR ) is similar to that proposed by the
Bouchard–Kirkaldy’s model [43] for primary dendritic
growth (λ1 ), i.e. λ1(BK) α (CR ) −0.50 . The λ1 dependence
as a function of CR has also been characterised by the
index −0.55 in the experimental equations proposed by
several works in the literature [7,14,32,34,36,38,45,46],
whose index has been the same for all growth direc-
tions (horizontal, upward vertical, and downward ver-
tical). Among these works can be mentioned the ones
of Peres et al. [34] and Costa et al. [44] for binary
Al–Xwt-% Si (X = 3, 5, 7, and 9) alloys and the
Al–7wt.%Si–3wt.%Cu multicomponent alloy, respec-
tively. In this sense, considering the results of this
investigation and those observed in the literature,
it is possible to predict that once the tertiary den-
dritic branches nucleate from the secondary branches
they begin to grow as primary dendritic branches.
Figure 7 shows schematically the mechanism of tertiary
Figure 4. (a) GR and (b) C R as a function of position from the branches growth proposed in this work for horizontal
metal–cooling chamber interface for the three analysed alloys. DS. A similar characteristic has been observed by Costa
et al. [36] for the upward and horizontally solidified
A–6wt.%Cu-4wt.%Si alloy.
The mentioned solidification thermal variables per- It is well known that low Fe concentration can be
mit us to stress that a large range of cooling rates is present in commercially pure Al and Si metals induc-
possible to be obtained by DS. Consequently, a wide ing the formation of iron-containing phases even in low
spectrum of microstructural length scales and inter- fractions. Thus, it is possible to observe in the solidifi-
metallic phase distribution needs to be examined. Some cation paths calculated by the Scheil model (assump-
typical microstructures of transverse sections for the tion of complete mixing in the liquid but no diffu-
three analysed alloys are depicted in Figure 5. It can sion in the solid) in Thermo-Calc software consider-
be observed, for all investigated alloys, that the λ3 val- ing impurity of 0.15wt-%Fe for all investigated alloys
ues increase with the distance from the metal–mould (Figure 8), which at the end of the solidification pro-
interface, i.e. the coarsening of dendrite arms with the cess, the resulting microstructure will consist mainly of
progress of solidification may be seen. Al-rich (α-Al) primary phase and interdendritic eutec-
It can be observed in Figure 6 that −1.1 and −0.55 tic mixture containing the eutectic (α-Al) phase, Si
power law functions can be applied to characterise the particles and intermetallic phases such as β-Al5 FeSi
λ3 evolution versus GR and CR , respectively. This is phase in Al–7wt-%Si–0.15wt-%Fe alloy, θ-Al2 Cu and
in agreement with experimental observations reported ω-Al7 Cu2 Fe phases in Al–7wt-%Si–3wt-%Cu–0.15wt-
by Sá et al. [39] (Al–Cu and Sn–Pb alloys), Osório %Fe alloy as well as α-Mg2 Si and π-Al8 Mg3 FeSi6
et al. [40] (Pb–3.5wt.%Sb alloy) and Spinelli et al. ] phases in Al–7wt-%Si–0.3wt-%Mg–0.15wt-%Fe alloy.
(Sn–0.7wt.%Cu–Ag alloys) who have investigated the Obviously, Figure 8 presents a theoretical predic-
λ3 growth during upward vertical DS. Rosa et al. tion of the transformations of phases that can occur
[41] have also suggested −0.55 law to represent the in the solidification process of the investigated alloys.
λ3 variation with CR for Al–Cu and Al–Si hypoeu- However, in solidification practical conditions of Al–Si
tectic alloys, including Al–7wt.%Si alloy, considering alloys, Fe-rich intermetallis phases can be grouped into
vertical upward and downward directional systems. three kinds of morphologies: polyhedral or star-like,
6 F. SOUZA ET AL.

Figure 5. Evolution of λ3 and as-cast microstructures of transverse sections where ‘P’ is the position from the metal–mould interface
of the castings.

Chinese script and platelet [47]. The presence of cation process conditions. It is known that the for-
one or more of these morphologies depends on the mation of the β-Al5 FeSi phase is favoured for iron
nominal composition of the alloy and the solidifi- contents greater than 0.4wt.%Fe and in conditions of
 MATERIALS SCIENCE AND TECHNOLOGY 7

script morphology [48]. Awamo and Shimizu [49] per-

formed experiments with Al-XSi alloys (X = 6 and
11wt.%) temperatures ranging from 700 to 950°C, and
Fe content from 0.1 to 0.9%, whose results showed that
the formation of the α-Al8 Fe2 Si phase has been more
favoured by higher superheat.
Figure 9 shows representative SEM micrographs
of transverse sections of Al–7wt.%Si–0.15wt.%Fe(3wt.
%Cu and 0.3wt.%Mg) alloys revealing the as-cast
microstructure constituted of the α-Al-rich phase
of dendritic morphology, according to the primary
phase, with a eutectic mixture filling the interden-
dritic regions that contain the eutectic α phase, Si
particles as well as α -Al8 Fe2 Si intermetallic phase
in the Al–7wt.%Si–0.15wt.%Fe alloy, θ-Al2 Cu and
ω-Al7 Cu2 Fe intermetallic phases in the Al–7wt.%Si–
3wt.%Cu–0.15wt.%Fe alloy as well as α-Mg2 Si and
π-Al8 Mg3 FeSi6 intermetallic phases in the Al–7wt.%Si–
0.3wt.%Mg–0.15wt.%Fe alloy as indicated by EDS map-
ping analysis. In concomitant analysis of the phases
constituted in the solidification paths shown in Figure 8
with the SEM microstructures of Figure 9, it is veri-
fied that the addition of the Cu and Mg elements in
the AlSiFe alloy modifies the Fe-intermetallic phases,
i.e. the precipitation of ω and π phases from the
β phase is observed, through the following peri-
tectic reactions: Liq.+β-Al5 FeSi → ω-Al7 Cu2 Fe and
Liq.+β-Al5 FeSi → π-Al8 Mg3 FeSi6 , during solidifica-
tion of the Al–7wt.%Si–3wt.%Cu–0.15wt.%Fe and Al–
Figure 6. Dependence of λ3 on both GR and C R values for 7wt.%Si–0.3wt.%Mg–0.15wt.%Fe alloys, respectively.
Al–7wt.%Si–0.15wt.%Fe–(3wt.%Cu and 0.3wt%Mg) alloys. It is known that the peritectic reaction consists of
a liquid (Liq) reacting with a solid phase (β) resulting
in another solid phase (ω or π), that is, Liq. + β →
ω or π. In the case of Al–Si–Cu–Fe and Al–Si–Mg–Fe
low superheat [47–49]. Investigations have shown that investigated multicomponent alloys, it is obvious that
superheated Al–Si alloys above 750°C, for example, the formed primary phase is (α-Al), followed by the
inhibit the formation of β-Al5 FeSi particles and favours transformation of the first [(α-Al)eutectic + Si particles]
the nucleation of α’ -Al8 Fe2 Si particle with Chinese eutectic reaction, as shown in Figure 8(b,c). At this

Figure 7. Schematic representation of λ3 growth mechanism proposed in this investigation.

8 F. SOUZA ET AL.

Figure 8. Solidification paths: (a) Al–7wt-%Si–0.15wt-%Fe, (b) Al–7wt-%Si–3wt-%Cu–0.15wt-%Fe, and (c) Al–7wt-%Si–0.3wt-
%Mg–0.15wt-%Fe alloys.

moment of the solidification process, both alloys when temperatures of 530 and 561°C are reached,
are constituted of (α-Al) + (α-Al)eutectic + Si parti- promote the respective peritectic reactions: Liq. + β-
cles + Liq. With cooling, the remaining liquid, which Al5 FeSi → ω-Al7 Cu2 Fe and π-Al8 Mg3 FeSi6 , for
is fully miscible and rich in rejected solutes (Si, Cu, the Al–7wt.%Si–3wt.%Cu–0.15wt.%Fe and Al–7wt.%-
and Fe, and Si, Mg, and Fe, for the Al–Si–Cu–Fe and 0.3wt.%Mg–0.15wt.%Fe alloys. The ω and/or π phases
Al–Si–Mg–Fe alloys, respectively), due to their low nucleate by means of the corresponding peritectic reac-
solubility in the (α-Al) primary phase, reaches the β tions, from the β solid phase surfaces and/or isolated
phase-nucleation temperatures, 560 and 572°C for the from the Si, Cu, Mg, and Fe solute enriched liquid.
Al–Si–Cu–Fe and Al–Si–Mg–Fe alloys (Figure 8(b,c)), Wang et al. [50] and Gustafsson et al. [51] have shown
respectively. With continued cooling most of the that for Al–Si–Mg alloys, the π phase either nucleates
enriched liquid of rejected solutes plus the β phase, independently or grows from the surfaces of β platelets.
 MATERIALS SCIENCE AND TECHNOLOGY 9

Figure 9. Elemental mapping detailing the phases formed during the solidification process of Al–7wt.%Si, Al–7wt.%Si–3wt-
%Cu–0.15wt.%Fe, and Al–7wt.%Si–0.3wt.%Mg–0.15wt.%Fe alloys.

Figure 10 shows the dependence of the microhard- dependence on λ3 can be characterised by a Hall-Petch
ness with the distance from the metal–mould inter- equation given by HV = 84−229(λ3 )−1/2 . It is known
face, growth rates, cooling rates, and tertiary dendrite that the properties of Al–Si alloys are influenced by
arm spacings for the three investigated alloys. It is the eutectic Si morphologies, which can be observed as
observed that GR , CR , and λ3 have not affected on spheroidal, fibrous and lamellar (or plate) depending on
HV values, resulting in HV = 82 along the length the levels of cooling rates during solidification (ranging
of the solidified-as ingot. On the other hand, for the from higher to lower rates, respectively) [2,5,7].
Al–7wt.%Si–0.15wt.%Fe alloy, the HV values increase The lamellar Si particles, observed in this investi-
gradually with the increase of λ3 values, and that HV gation for low CR values in the Al–7wt-%Si–0.15wt-%
10 F. SOUZA ET AL.

Figure 10. Variation of HV as a function of: (a) P, (b) GR , (c) C R , and (d) λ3 −1/2 for Al–7wt.%Si–0.15wt.%Fe–(3wt.%Cu and 0.3wt.%Mg)
alloys.

Fe alloy, although of high hardness, have deleteri- formation of block or plate-type eutectic Si particles and
ous effects on ductility, acting as initiators for prop- coarse intermetallics in the microstructure, observed at
agation of cracks. It is obvious that high hardness low cooling rates. Brito et al. [9] have found a behaviour
has a degenerative effect on the plastic deforma- of HV on λ2 for upward solidified Al–Mg and Al–Mg-
tion of metals and alloys. This has influenced the Si alloys, similar and opposite to those observed in this
behaviour of HV = f (GR , CR , and λ3 ) observed in work for Al–Si–Fe and Al–Si–Fe– (Cu and Mg) alloys,
Figure 10(b–d), i.e. higher HV values were found respectively, that is, for both investigated alloys by the
for higher CR , GR , and λ3 values. However, higher authors, the HV values increase with the increase in the
HV values are evidenced for the Al–Si–Cu–Fe and secondary dendrite arm spacings, whose phenomenon
Al–Si–Mg–Fe multicomponent alloys, as shown in has been associated with the decrease in the fraction of
Figure 10(a). This should be attributed to the pres- intermetallic Mg2 Si particles of high hardness within
ence of θ-Al2 Cu and ω-Al7 Cu2 Fe phases in Al–7wt- the interdendritic regions, on the other hand, at the
%Si–3wt.%Cu–0.15wt.%Fe alloy as well as α-Mg2 Si and same time the fraction of Fe-based intermetallic par-
π-Al8 Mg3 FeSi6 phases in Al–7wt.%Si–0.3wt.%Mg– ticles increases, contributing to the increase in the HV
0.15wt.%Fe alloy between the dendritic branches. In the values for lower cooling rates.
multicomponent investigated alloys, the formation of Among the highlights of this investigation, it can
other intermetallic eutectic phases is reinforcing them be emphasised that there are no studies in the liter-
to the point of not allowing the lamellar Si morphol- ature that correlate microhardness and tensile prop-
ogy to be a dominant factor. Chen et al. [5] have found erties with λ3 for multicomponent alloys and for the
Al–Si–Mg alloys have higher UTS (σ US ) and elongation case of binary alloys only the work of Barros et al.
(E%) values for more refined as-cast microstructures, [42], recently published, presents results that relate HV
and the authors have associated the low E% values to the with λ3 . Barros et al. [42] have developed horizontal
 MATERIALS SCIENCE AND TECHNOLOGY 11

solidification experiments with the Al–3wt.%Cu alloy, most common of the Fe-intermetallic phases (β-
and the authors have characterised the HV depen- Al5 FeSi) was not been observed.
dence on λ3 by means of a Hall-Pecth expression given
by HV = 6 + 310(λ3 )−1/2 . In contrast to the observed Disclosure statement
behaviour in this work (Al–7wt.%Si), higher HV val-
No potential conflict of interest was reported by the authors .
ues were found for lower λ3 values for the binary
Al–3wt.%Cu alloy.
Funding
The authors acknowledge the financial support provided by
Conclusions IFPA – Federal Institute of Education, Science and Technol-
ogy of Pará, UFPA – Federal University of Pará, and CNPq
Al–7wt.%Si–0.15wt.%Fe, Al–7wt.%Si–3wt.%Cu–0.15
– The Brazilian Research Council (grants 308784/2014-6),
wt.%Fe, and Al–7wt.%Si–0.3wt.%Mg–0.15wt.%Fe FAPESPA – Amazon Foundation of Support to Study and
alloys grown by the THDS technique and solid- Research (grants ICAAF 064/2016) and CAPES - Coordina-
ification thermal variables (GR and CR ), tertiary tion of Superior Level Staff Improvement, Brazil .
dendrite arm spacings (λ3 ), phase constitution and
Vickers microindentation hardness (HV) have been ORCID
determined. The following major conclusions can be O. Rocha [Link]
drawn:
References
(1) The variation of λ3 values relating to the ther-
[1] Zeren M, Karakulak E. Study on hardness and micro-
mal variables, GR and CR , for the Al–7wt.%Si-
structural characteristics of sand cast Al-Si-Cu alloys.
0.15wt.%Fe-(3wt.%Cu and 0.3wt.%Mg) alloys has Bull Mater Sci. 2009;32:617–620.
been investigated and relationships among them [2] Costa TA, Dias M, Gomes LG, et al. Effect of solution
have been obtained. It has been observed that the time in T6 heat treatment on microstructure and hard-
growth direction (horizontal vs vertical) had not ness of a directionally solidified Al-Si-Cu alloy. J Alloys
affected the λ3 growth experimental laws as a func- Compd. 2016;683:485–494.
[3] Araújo RL, Kikuchi RH, Barros AS, et al. Influence of
tion of GR and CR for the Al–7wt-%Si-0.15wt-%Fe upward and horizontal growth direction on microstruc-
alloy, i.e. −1.1 and −0.55 exponents can be applied ture and microhardness of an unsteady-state direction-
to characterize the experimental variation of λ3 ally solidified Al-Cu-Si alloy. Matéria. 2016;21:260–269.
with corresponding cooling thermal variables for [4] Ursolino CH, Kikuchi RH, Barrros AS, et al. On the nat-
the mentioned multicomponent alloys. ural convection in the columnar to equiaxed transition
in directionally solidified aluminum-based binary and
(2) As noted for the growth direction, the addition
multicomponent alloys. Mater Res. 2015;18:1362–1371.
of both 3wt.%Cu and 0.3wt.%Mg alloying in the [5] Chen R, Xu Q, Guo H, et al. Correlation of solidifica-
Al–7wt.%Si-0.15wt.%Fe alloy has not affected the tion microstructure refining scale, Mg composition and
growth experimental law of λ3 as a function of heat treatment conditions with mechanical properties
GR and CR . However, a more complex arrange- in Al-7Si-Mg cast aluminum alloys. Mater Sci Eng A.
ment of phases in microstructure of alloys has been 2017;685:391–402.
[6] Zhu M, Jian Z, Yang G, et al. Effects of T6 heat treat-
observed. ment on the microstructure, tensile properties, and frac-
(3) Observations carried out in the interdendritic ture behavior of the modified A356 alloys. Mater Des.
eutectic mixture reveal clearly the eutectic (α- 2012;36:243–249.
Al) phase, Si particles and intermetallic phases [7] Chen R, Shi YF, Xu QY, et al. Effect of cooling rate on
such as θ-Al2 Cu and ω-Al7 Cu2 Fe in the Al–7wt- solidification parameters and microstructure of Al-7Si-
0.3Mg-0.15Fe alloy, trans. Nonferrous Met Soc China.
%Si–3wt.%Cu–0.15wt.%Fe alloy, as well as α-
2014;24:1645–1652.
Mg2 Si and π-Al8 Mg3 FeSi6 in the Al–7wt-%Si– [8] Brito C, Vida T, Freitas E, et al. Cellular/dendritic
0.3wt.%Mg–0.15wt.%Fe alloy which appear to jus- arrays and intermetallic phases affecting corrosion and
tify the higher HV values for the investigated alloys mechanical resistances of an Al-Mg-Si alloy. J Alloys
than those from the Al–7wt.%Si alloy. Compd. 2016;673:220–230.
(4) It was observed that the addition of Cu and Mg [9] Brito C, Reinhart G, Nguyen-Thi H, et al. High cool-
ing rate cells, dendrites, microstructural spacings and
alloying elements in the Al–7wt-%Si–0.15wt-%Fe microhardness in a directionally solidified Al-Mg-Si
alloy have induced the following peritectic reac- alloy. J Alloys Compd. 2015;636:145–149.
tions: Liq+β-Al5 FeSi → ω-Al7 Cu2 Fe and Liq+β- [10] Brito C, Costa TA, Vida TA, et al. Characterization
Al5 FeSi→ → π-Al8 Mg3 FeSi6. of dendritic microstructure, intermetallic phases, and
(5) Higher HV values were observed for the Al–7wt. hardness of directionally solidified Al-Mg and Al-Mg-Si
alloys. Metall Mater Trans A. 2015;46:3342–3355.
%Si–0.15wt.%Fe alloys containing Cu and Mg
[11] Gomes LG, Moutinho DJ, Ferreira IL, et al. The growth
alloying elements. of secondary dendritic arms in directionally solidified
(6) For the solidification conditions assumed in this Al-Si-Cu alloys: a comparative study with binary Al-Si
work, such as Fe content and superheat, the alloys. Appl Mech Mater. 2015;719–720:102–105.
12 F. SOUZA ET AL.

[12] Moutinho DJ, Gomes LG, Rocha OL, et al. Thermal [30] Gomes LF, Silva BL, Garcia A, et al. Dendritic
parameters, microstructure and porosity during tran- growth, solidification thermal parameters, and Mg
sient solidification of ternary Al-Cu-Si alloys. Mater Sci content affecting the tensile properties of Al-Mg-
Forum. 2013;730–732:883–888. 1.5WtPctFe alloys. Metall Mater Trans A. 2017;48:1841–
[13] Spinelli JE, Cheung N, Goulart PR, et al. Design of 1855.
mechanical properties of Al-alloys chill castings based [31] Rocha OL, Gomes LG, Moutinho DJ, et al. The colum-
on the metal/mold interfacial heat transfer coefficient. nar to equiaxed transition in the directional solidifica-
Int J Therm Sci. 2012;51:145–154. tion of aluminum based multicomponent alloys. Revista
[14] Spinelli JE, Silva BL, Garcia A. Assessment of tertiary Escola de Minas. 2015;68:85–90.
dendritic growth and its effects on mechanical prop- [32] Rocha OL, Siqueira CA, Garcia A. Heat flow parameters
erties of directionally solidified Sn-0.7Cu-xAg solder affecting dendrite spacings during unsteady state solid-
alloys. J Electron Mater. 2014;43:1347–1361. ification of Sn-Pb and Al-Cu alloys. Metall Mater Trans
[15] Moustafa MA, Samuel FH, Doty HW, et al. Effect of Mg A. 2003;34:995–1006.
and Cu additions on the microstructural characteris- [33] Jácome PAD, Fernandes MT, Garcia A, et al. Application
tics and tensile properties of Sr-modified Al-Si eutectic of computational thermodynamics to the evolution of
alloys. Int J Cast Metals Res. 2002;14:235–253. surface tension and Gibbs-Thomson coefficient during
[16] Li Z, Samuel AM, Samuel FH, et al. Effect of alloy- multicomponent aluminum alloy solidification. Mater
ing elements on the segregation and dissolution of Sci Forum. 2016;869:416–422.
CuAl2 phase in Al-Si-Cu 319 alloys. J Mater Sci. [34] Peres M, Siqueira C, Garcia A. Macrostructural and
2003;38:1203–1218. microstructural development in Al-Si alloys direction-
[17] Argo D, Gruzleski JE. Porosity in modified aluminum ally solidified under unsteady-state conditions. J Alloys
alloy castings. AFS Trans. 1988;96:65–73. Compd. 2004;381:168–181.
[18] Mondolfo LF. Aluminum alloys: structure and proper- [35] Vasconcelos AJ, Kikuchi RH, Barros AS, et al. Intercon-
ties. London: Butterworths; 1976. nection between microstructure and microhardness of
[19] Emma S, Salem S. The heat treatment of Al-Si-Cu- directionally solidified binary Al-6wt.%Cu and multi-
Mg casting alloys. J Mater Process Technol. 2010;210: component Al-6wt.%Cu-8wt.%Si alloys. An Acad Bras
1249–1259. Cienc. 2016;88:1099–1111.
[20] Mikolajczak P. Microstructural evolution in AlMgSi [36] Costa TA, Moreira AL, Moutinho DJ, et al. Growth
alloys during solidification under electromagnetic stir- direction and Si alloying affecting directionally solid-
ring. Met. 2017;7:89. ified structures of Al-Cu-Si alloys. Mater Sci Technol.
[21] Samuel AM, Doty HW, Valtierra S, et al. Relationship 2015;31:1103–1112.
between tensile and impact properties in Al–Si–Cu–Mg [37] Araújo EC, Barros AS, Kikuchi RH, et al. The role of
cast alloys and their fracture mechanisms. Mater Des. Si and Cu alloying elements on the dendritic growth
2014;53:938–946. and microhardness in horizontally solidified binary and
[22] Meyers CW, Hinton KH, Chou JS. Towards the opti- multicomponent aluminum-based alloys. Metall Mater
mization of heat treatment in aluminum alloys. Mater Trans A. 2017;48:1163–1175.
Sci Forum. 1992;72:102–104. [38] Barros AS, Magno IAB, Silva FP, et al. Measurements of
[23] Silva BL, Xavier MGC, Garcia A, et al. Cu and Ag addi- microhardness during transient horizontal directional
tions affecting the solidification microstructure and ten- solidification of Al-Rich Al-Cu alloys: effect of ther-
sile properties of Sn-Bi lead-free solder alloys. Mater Sci mal parameters, primary dendrite arm spacing and
Eng, A. 2017;705:325–334. Al2Cu intermetallic phase. Met Mater Int. 2015;21:
[24] Ourdjini A, Yilmaz F, Hamed QS, et al. Microstructure 429–439.
and mechanical properties of directionally solidified [39] Sá F, Rocha OL, Siqueira CA, et al. The effect of
Al–Si eutectic alloys with and without antimony. Mater solidification variables on tertiary dendrite arm spac-
Sci Technol. 2013;8:764–776. ing in unsteady-state directional solidification of Sn-
[25] Trivedi R, Liu S, Mazumder P, et al. Microstruc- Pb and Al-Cu alloys. Mater Sci Eng, A. 2004;373:
ture development in the directionally solidified Al- 131–138.
4.0 wt%Cu alloy system. Sci Technol Adv Mater. [40] Osório WR, Freitas ES, Peixoto LC, et al. The effects
2001;2:309–320. of tertiary dendrite arm spacing and segregation on
[26] Zhou YZ, Volek A. Influence of solidification conditions the corrosion behavior of a Pb-Sb alloy for lead-
on grain boundary cohesion in the mushy zone dur- acid battery components. J Power Sources. 2012;207:
ing directional solidification of nickel base superalloy. 183–190.
Mater Sci Technol. 2007;23:297–302. [41] Rosa D, Spinelli J, Garcia A. Tertiary dendrite arm
[27] Yuan SN, Jia LN, Ma LM, et al. Microstructure and spacing during downward transient solidification of
room temperature mechanical properties of hypereu- Al Cu and Al Si alloys. Mater Lett. 2006;60:1871–
tectic Nb–Si based alloy processed by directional solid- 1874.
ification. Mater Sci Technol. 2014;30:75–80. [42] Barros AS, Silva MAPS, Rocha OL, et al. Thermal
[28] Bertelli F, Freitas ES, Cheung N, et al. Microstructure, parameters, tertiary dendritic growth and microhard-
tensile properties and wear resistance correlations on ness of directionally solidified Al-3wt%Cu alloy. Mater
directionally solidified Al-Sn-(Cu; Si) alloys. J Alloys Sci Forum. 2016;869:452–457.
Compd. 2017;695:3621–3631. [43] Bouchard D, Kirkaldy JS. Prediction of dendrite arm
[29] Reyes RV, Bello TS, Kakitani R, et al. Tensile proper- spacings in unsteady and steady-state heat flow of uni-
ties and related microstructural aspects of hypereutectic directionally solidified binary alloys. Metall Mater Trans
Al-Si alloys directionally solidified under different melt B. 1997;28:651–663.
superheats and transient heat flow conditions. Mater Sci [44] Costa MO, Souza FVA, Magno IAB, et al. Interrelation
Eng, A. 2017;685:235–243. of solidification processing variables and microstructure
 MATERIALS SCIENCE AND TECHNOLOGY 13

of a horizontally solidified Al-based 319.1 alloy. Mater [48] Cao X, Campbell J. The nucleation of Fe-rich phases
Sci-Medzg. 2017;23:124–128. on oxide films in a Al-11,5Si-0,4Mg cast alloys. Metall
[45] Carvalho DB, Guimarães EC, Moreira AL, et al. Mater Trans A. 2003;34:1409–1420.
Characterization of the Al-3wt.%Si alloy in unsteady- [49] Awano Y, Shimizu Y. Non equilibrium crystallization
state horizontal directional solidification. Mater Res. of AlFeSi compound in melt-superheated Al-Si alloy
2013;16:874–883. casting. AFS Trans. 1990;176:889–895.
[46] Carvalho DLB, Costa TA, Moreira ALS, et al. Solidifica- [50] Wang, QG, Jones PE, Osborne M. Effect of iron on
tion thermal parameters and dendritic growth during the microstructure and mechanical properties of an Al-
the horizontal directional solidification of Al-7wt.%Si 7%Si-0.4%Mg casting alloy. SAE-Technical Publication,
alloy. Revista Escola de Minas. 2014;67:265–270. 2003-01-0823.
[47] Dinnis CM, Taylor JA, Dahle AK. As-cast morphology [51] Gustafsson G, Thorvaldsson T, Dunlopd GL. The influ-
of iron-intermetallics in Al-Si foundry alloys. Scripta ence of Fe and Cr on the microstructure of cast Al-Si-
Mater. 2005;53:955–958. Mg alloys. Metall Mater Trans A. 1986;17:45–52.