Kovove Mater.

57 2019 33–43 33
DOI: 10.4149/km 2019 1 33

Improving mechanical properties of AlSi10Mg aluminum alloy

using ultrasonic melt treatment combined with T6 heat treatment

F. Vatansever, A. T. Ertürk*, S. Karabay

Mechanical Engineering Department, Engineering Faculty, Kocaeli University, Izmit-Kocaeli, Turkey

Received 12 July 2018, received in revised form 8 October 2018, accepted 15 October 2018

Abstract
The purpose of this study is examining the effects of ultrasonic melt treatment on mi-
crostructure and mechanical properties of the AlSi10Mg alloy. In this study, ultrasonic melt
treatment with different frequencies and duration was applied to AlSi10Mg molten alloy. Also,
T6 heat treatment was applied to samples which were obtained by ultrasonic melt treatment
(UST) with different application parameters to investigate the combination of T6 heat treat-
ment and UST effect on the microstructure and mechanical properties. The microstructure
of cast samples was characterized by optical and scanning electron microscopy. Also, hard-
ness and tensile tests were carried out. The results indicate that primary α-Al phases are
transformed from coarse dendrites to smaller dendrites by the ultrasonic melt treatment. Be-
sides, secondary dendrite arm spacing of the alloy decreased and mechanical properties such
as hardness and tensile strength increased. UST reduces the secondary dendrite arm spacing
(SDAS) of the aluminum alloy to 50 %. Additionally, increasing treatment time contributes
to hardness up to 12 % and tensile strength up to 50 %.

K e y w o r d s : grain refining, ultrasonic melt treatment, aluminum alloys, mechanical prop-

erties

1. Introduction microstructure and mechanical properties of alloys in

metallurgy. Even though initial experiments of using
Aluminum is the most important secondary metal UST at the laboratory level for processing molten al-
that comes after steel, increasingly used in many sec- loys were based on the 1940s and first industrial exper-
tors such as automotive, aerospace, machinery, man- iments in the 1950–60s, the method has not yet been
ufacturing, transportation, construction, packaging, used in mass production. In the last 15 years, there
conductor wire production, and household appliances has been a great increase in the researches related to
due to its lightweight, mechanical, physical, corro- the UST depending on the technological developments
sion properties and many other reasons. Parallel to on modeling techniques and producibility of more ad-
the rapidly developing technology, expectations of the vanced manufacturing equipment [6, 7].
qualification of materials are also increasing. Over the Researchers implemented UST with mostly similar
several past decade’s grain refinement has been mostly parametric approaches. In melting stage, 0.2–1 kg of
applied for enhancing the microstructural and mecha- weight alloy in a crucible was heated in an electric oven
nical properties of aluminum alloys through physi- and melted. In the next stage, ultrasonic vibration was
cal and chemical methods such as severe plastic de- applied to melt in a pot for 10–1200 s at 580–750 ◦C.
formation, mechanical alloying, conventional thermo- In general, a molten alloy was subjected to ultrasonic
mechanical processing, and adding grain refiners [1– vibration in mainly two different cases as above the
5]. Ultrasonic vibration (20 kHz–1 GHz frequencies) liquidus temperature (liquid state) and during solidi-
application is a relatively new method compared to fication (nucleation of α-Al phase and eutectic trans-
the other methods used to improve the mechanical formation). At last stage, the melt was poured into
properties of materials. UST is used to improve the a crucible made by graphite, copper or steel molds
*Corresponding author: tel.: +90-262-3033402; fax: +90-262-3033003; e-mail address: tamererturk@gmail.com
34 F. Vatansever et al. / Kovove Mater. 57 2019 33–43

T a b l e 1. Previously studied conditions of UST for Al-Si alloys

Material Frequency Treatment time Amplitude Power Radiator

(kHz) (µm) (kW) (Sonotrode)

AlSi9Cu3(Fe) [8] 19.8 ± 0.25 From 680 to 580 ◦C – 0.2, 0.4, 0.6 Ti6Al4V

A390 [9] 20 0–1200 s – 0.84 –

Al-23%Si [10] 20 600 s 4 0.05 Indirect UST

Different Al-Si alloys [11] 20 120 s – 0.26 –

Al-7%Si and Al-16%Si [12] 20 From 750 ◦C to 25 – Ti6Al4V

solidification

Al-17%Si [13] 20 ± 1 During DC casting 48 2 Ceramic

Al-12.2Si-3.3Cu-2.4Ni-0.8Mg- 19 60 s 20 – Titanium
-0.1Fe [14]

Al-8%Si [15] 20 Liquid, nucleation and – 0.5 –

growth and eutectic
transformation states

A390 [16] 20 30 s – 1 –

LM6 [17] 20 From 750 ◦C to 25 – Ti6Al4V

solidification

Al-12%Si-2%Fe [18] 20 60, 120 s – 0.27 –

AlSi9Cu3 [19] 19.8 ± 0.25 120 s – 0.2, 0.4, 0.6 Ti6Al4V

A356 [20] 20 From 630 ◦C to 56.7 1.5 Ti6Al4V

solidification

Al-Si alloy (Etial-195) [21] 40 25, 30, 40, 50 s – 0.5 Indirect UST

Different Al-Si alloys [22] – 10 s, liquid state, 40 4 Niobium

solidification range

and solidified under atmospheric conditions. As seen increased, and grain size decreased with increasing ul-
in Table 1, many researchers conducted studies on the trasonic power [8]. The results have shown that ul-
effects of UST at a frequency of 19–40 kHz with an am- trasonic vibration application during solidification of
plitude of 4–56.7 µm and at a power of 0.05–4 kW con- aluminum alloys is more efficient when the transition
ditions to various alloys by immersing a preheated ti- metals such as Zr and Ti are present in the alloy [23].
tanium, ceramic or niobium radiator (sonotrode) into Different than extensively examined aluminum alloys,
the melt or applying ultrasonic vibration indirectly. the effects of ultrasonic vibration on the microstruc-
It has been reported at previous studies [8–33] that ture and mechanical properties on Mg-8Li-3Al [24],
grain refinement at the microstructure decreased seg- AS41 [25], Al-4Mg [26] alloys and 35CrMo steel [27]
regation, increased homogeneity, improved mechani- were investigated. Researchers reported that the UST
cal properties (tensile strength, stiffness, elongation positively affected the microstructure of these mate-
at break) and removal of dissolved gases (degassing) rials [24–27].
were obtained in tested materials by UST. Besides, Researchers have investigated the effects of ultra-
the ultimate tensile strength and elongation at break sonic vibration on different Al-Si alloys. Kotadia et al.
 F. Vatansever et al. / Kovove Mater. 57 2019 33–43 35

Fig. 1. (a) Schematic of experimental setup: 1 – piezoelectric transducer, 2 – horn, 3 – radiator, 4 – power amplifier, 5 –
oscilloscope, 6 – ultrasonic function generator, 7 – movable clamp, 8 – graphite crucible, 9 – electrical furnace, and (b)
UST application in the electrical furnace.

[12] have investigated the effect of UST during solid- ultrasonic vibration with T6 heat treatment of the al-
ification of Al-7%Si and Al-16%Si alloys. The results loy were investigated. For this purpose, ultrasonic vi-
show that dendritic grains of hypoeutectic alloy trans- bration was applied to the AlSi10Mg aluminum alloy
formed into equiaxed α-Al grains and better disper- melt with the frequency of 15, 17.5 and 20 kHz for 0,
sion of primary Si particles was obtained for hypereu- 180, 300 and 600 s and the microstructure and mecha-
tectic alloy. Wang et al. [15] applied UST to Al-8%Si nical properties of obtained samples were investigated.
alloy at different solidification stages including a fully
liquid state, nucleation, and growth of α-Al phase and
eutectic transformation period. They have shown that 2. Materials and methods
UST at all application stages changed eutectic mor-
phology and improved yield strength of the alloy, but In this study, the hypoeutectic AlSi10Mg alu-
the apparent morphological change was observed when minum alloy was used. The experimental setup used
UST was directly introduced in the eutectic transfor- in this study mainly consists of a 300 W power ampli-
mation period. Jian et al. [20] have shown that UST fier, a transducer which converts electrical energy to
modified the morphology of eutectic silicon of A356 mechanical energy in the form of ultrasonic vibration,
alloy from coarse acicular plate-like form to a finely a 20 mm diameter and 400 mm long Ti6Al4V radiator
dispersed rosette-like form and reduced the size of eu- which transfers ultrasonic energy to molten alloy, and
tectic silicon. Zhang et al. [22] applied UST to Al-Si an ultrasonic function generator which enables chang-
alloys with 5, 11 and 17 % Si content. According to the ing of ultrasonic vibration frequency and amplitude.
results, the grain size of hypoeutectic alloys and the Rest of the parts of the experimental setup is shown
size of primary Si particles of hypereutectic alloy de- in Fig. 1.
creased with UST. They also showed that while UST AlSi10Mg aluminum alloy was cut and prepared in
at solidification of hypoeutectic alloys refined grain 400 g using a precision scale for every experiment. The
structure, UST at the liquid phase of hypereutectic alloy was melted at 700 ◦C in a graphite crucible using
alloy was more effective than the application at solid- an electrical furnace. The temperature of the molten
ification. alloy was controlled using a K-type thermocouple. Af-
Although there are different studies in the litera- ter that, the Ti6Al4V radiator was put into the elec-
ture related to the UST during the solidification of al- tric furnace through a hole at the top of the furnace
loys, liquid state application of UST is limited. There and preheated at 700 ◦C to avoid thermal shock and
is a scant comprehensive study that examines the ef- temperature drop of the melt. After preheating the
fects of vibration frequency and duration on mecha- radiator was dipped into the molten alloy by 30 mm
nical properties. Also, there is not an extensive study below the melt surface. Different ultrasonic vibration
about how T6 heat treatment affects microstructure with 15, 17.5, 20 kHz frequency was applied for 180,
and mechanical properties of aluminum alloys when 300 and 600 s isothermally at 700 ◦C. The amplitude of
applied with UST. In this study, the effects of ultra- ultrasonic vibration is measured as 6 µm by Keyence
sonic vibration were investigated to improve the mi- LC-2420 laser displacement sensor. Experimental con-
crostructure and hence the mechanical properties of ditions are given in Table 2.
the AlSi10Mg alloy. Further, the combined effects of After the application of ultrasonic treatment, the
36 F. Vatansever et al. / Kovove Mater. 57 2019 33–43

T a b l e 2. Experimental conditions of the study

Testing & specimen conditions

Specimen number Mold temperature ( ◦C) Diameter (mm) Heat treatment

1 23 ø 15 –
2 23 ø 10 –
3 300 ø 15 –
4 300 ø 15 T6

Parameters of UST

Ultrasonic melt treatment time (s) 0, 180, 300, 600

Vibration frequency (kHz) 15, 17.5, 20
Vibration amplitude (µm) 6
Melting temperature ( ◦C) 700
Cooling rate ( ◦C s−1 ) 3.7 (cold mold)–1.53 (hot mold)
Weight of molten alloy (g) 400
Radiator diameter (mm) 20
Immersed length of the radiator (mm) 30

Parameters of T6 heat treatment [35]

Solutionizing: 530 ± 5 ◦C, 6 h Artificial aging:160 ± 5 ◦C, 12 h

T a b l e 3. Chemical composition, physical and mechanical properties of tested alloy [35]

Chemical composition of tested alloy

Si Fe Cu Mn Mg Ti Ni Zn Sn

10.7 0.285 0.001 0.343 0.287 0.06 0.03 0.01 0.004

Physical and mechanical properties

Density (g cm−3 ) 2.68

Melting range ( ◦C) 555–590
Modulus of elasticity (N mm−2 ) 74535
Thermal conductivity at 25 ◦C (W cm−1 K−1 ) 1.39–1.60

UTS (N mm−2 ) Elongation (%) Brinell hardness

Sand castings F 157 3.5 60

T6 215 2 85
Permanent mold castings F 176 3 70
T6 245 3 70

radiator was taken out from molten alloy. The alloy ual of the etchant. Finally, microstructure examina-
was poured into two different steel molds ø 10 and tion was characterized by optical and scanning elec-
ø 15 mm. This procedure was repeated without ul- tron microscopes and mechanical properties were mea-
trasonic vibration to observe the effect on the mi- sured by tensile and hardness tests. The tensile test
crostructure of the cast alloy. Solidified alloys were was performed according to ASTM B557M-10. The
cut to 10 mm height. All the samples were sanded tensile test specimens were manufactured by a CNC
with different SiC papers from 300 to 2400 grids. Af- lathe, having a gauge length of 30 mm and a diameter
ter that, obtained samples were polished with 6 and of 6 mm. Brinell hardness of the alloys was measured
1 µm diamond paste respectively followed by final pol- according to ASTM E-10 using a ø 2.5 mm diameter
ishing using 0.06 µm colloidal silica suspension. Pol- ball indenter under a load of 62.5 kgf. Physical and
ished specimens were immersed into Keller’s etchant mechanical properties of the tested alloy are listed in
for 30 s and washed with alcohol to neutralize resid- Table 3.
 F. Vatansever et al. / Kovove Mater. 57 2019 33–43 37

Fig. 2a–d. Microstructures after 15 kHz UST with diameter cold mold ø 15 mm: (a) untreated, (b) 180 s, (c) 300 s, and
(d) 600 s.

3. Results show that after 300 s UST almost all the prior den-
drites were disintegrated to the structure. Different
3.1. Microstructure frequency values caused a little difference in the dis-
integration of the prior dendrites (see Figs. 2f–h). It
More than 700 optical and scanning electron mi- is therefore not surprising that no important differ-
croscope images were analyzed to investigate the ef- ence was found between the mechanical properties of
fect of UST on the microstructure of AlSi10Mg alloy. these samples. Haghayeghi et al. [28] reported that
Selected microstructures of the ø 15 mm diameter cold the application at 17.5 and 20 kHz reduced the grain
mold samples produced by UST and as-cast are shown size and porosity, while 10 and 14 kHz did not cause
in Figs. 2a–d. The untreated alloy consists of a coarse- any change because these frequencies did not outnum-
grained microstructure containing α-Al dendrites with ber the cavitation threshold of the AA7075 aluminum
a maximum size of 100 µm as observed in Fig. 2a. It alloy. The presented research results show that grain
is also distinctly seen that in Figs. 2b–d the coarse refinement was achieved by applying all frequency val-
α-Al dendrites are fragmented and distributed rela- ues. Also, the rapid cooling condition of the ø 10 mm
tively homogeneously in the eutectic phase depending diameter mold results in thinner microstructure than
on the increasing application period from 180 to 600 s the ø 15 mm diameter mold.
at a frequency of 15 kHz. By the effect of the T6 heat treatment shown in
Relatively coarse-grained dendrites are distributed Fig. 3a, the microstructure consisting of fragmented
heterogeneously in the microstructure of the ø 10 mm and dispersed α-Al dendrites in the eutectic phase was
diameter cold mold sample obtained without ultra- unlikely to be detected in Figs. 2a,e. Besides, Figs. 3b–
sonic vibration in Fig. 2e. Microstructures obtained d show the microstructures of the ø 15 mm diameter
with different ultrasonic vibration frequencies (15, hot mold samples obtained by a combination of T6
17.5 and 20 kHz) of the ø 10 mm diameter cold mold heat treatment and UST with a frequency of 20 kHz.
samples can be seen in Figs. 2f–h. The micrographs The combined effect of coexistent UST and T6 heat
38 F. Vatansever et al. / Kovove Mater. 57 2019 33–43

Fig. 2e–h. Microstructures after 300 s UST with diameter cold mold ø 10 mm; (e) untreated, (f) 15, (g) 17.5, and (h)
20 kHz frequency.

treatment application on AlSi10Mg alloy is observed tween secondary dendrite arms, as a criterion, in eval-
as evolving time-dependent size reduction from 180 uating materials exhibiting dendritic microstructure
to 600 s (see Figs. 3b–d) of both α-Al and eutectic was also measured. SDAS values of the samples ob-
phases, exhibiting a close homogeneous distribution tained by different UST conditions are presented in
within each other. Fig. 5. It is clear that the UST application reduces
The resulting microstructures after UST with a SDAS with increasing treatment time for all frequen-
frequency of 15 kHz for 180 and 300 s are given in cies and samples. It is furthermore detected that the
Fig. 4. According to the micrographs, dendrites be- ø 15 and ø 10mm diameter cold mold samples have a
come smaller and spread better into the microstruc- similar, nearly linear decreasing characteristic whereas
ture of AlSi10Mg alloy by the increasing UST time. the ø 15mm diameter hot mold samples show a sudden
In addition to this, it has been found that the eutec- decrease for 180 s application period of UST and then
tic silicon phase shrinks similarly and spreads better a decrease of SDAS becomes relatively lower with in-
in the microstructure with the increasing application creasing treatment time. It is also observed that the
time. These microstructural results of the UST ap- SDAS values of ø 10 mm diameter size samples are
plication on AlSi10Mg alloy correspond with previous lower when compared to ø 15 mm in cold mold. It can
researches [11, 14, 16, 17, 20]. UST on hyper-eutectic be said that this case is due to the samples of ø 10 mm
Al-23%Si alloy reduced coarse primary silicon and α- diameter size have smaller cross-sectional area than
Al phase size and diffused silicon phase more homoge- the ø 15 mm diameter samples and therefore solidify
neously into the structure [10]. UST caused a trans- faster. Furthermore, owing to the lower cooling rate
formation of the primary α-Al phase from dendrite the distances between the secondary dendrite arms of
to fine globular form [17, 19, 22] and shortened the the hot mold samples are higher than those of cold
length of the eutectic silicon plates in hypo-eutectic mold samples.
Al-Si alloys [12, 17, 19, 20]. In the previous studies, many researchers found out
To evaluate the microstructure, the distance be- that UST modified the microstructure by providing
 F. Vatansever et al. / Kovove Mater. 57 2019 33–43 39

Fig. 3. Microstructures after 20 kHz UST with diameter hot mold ø 15 mm and T6 heat treatment: (a) untreated, (b)
180 s, (c) 300 s and (d) 600 s.

Fig. 4. SEM micrographs of the ø 15 mm diameter cold mold samples at 15 kHz frequency for (a) 180 s and (b) 300 s.

smaller dendrites, equiaxial and globular grains [8– 3.2. Mechanical properties
34]. In addition to this, Yao et al. [24], Patel et al. [25]
and Ruirun et al. [32] showed that grain size of Mg- It is known that the mechanical properties of mate-
8Li-3Al, AS41 and TiAl alloys decreased by increasing rials are directly affected by microstructure. While
treatment time. Our investigation results verify these fine-grained microstructure develops mechanical prop-
researchers’ findings in this manner. erties, coarse-grained microstructure affects them ad-
40 F. Vatansever et al. / Kovove Mater. 57 2019 33–43

defect in the tensile specimens. It has also been found

that the T6 heat treatment application increases the
tensile strength even without UST. In the case of com-
bining two applications the maximum value of the ten-
sile strength is obtained. Similar results were previ-
ously reported that ultrasonic vibration caused an in-
crease of the tensile strength of A390 [9], near-eutectic
Al-Si [14], AlSi9Cu3 [19], Mg-8Li-3Al [24], AA7075
[28] alloys.
All hardness values of the test samples were in-
creased by UST for all frequencies with increasing
treatment time as shown in Fig. 6b. Hardness incre-
ment changes similarly for all samples. As a numerical
example of the hardness increment, while the hardness
of the ø 10 mm diameter cold mold samples without
ultrasonic treatment was 84.9 HB, the UST for 600 s
increased hardness to 90.03 HB for 15 kHz, 89.37 HB
for 17.5 kHz and 90.03 HB for 20 kHz.
As hardness test results of the ø 10 and ø 15 mm
diameter samples with cold mold are analyzed, while
the hardness increases for 300 s at all three frequency
values, the increasing rate after 300 s reduces signifi-
cantly. Accordingly, it can be said that UST for more
than 300 s does not cause an important change in
the hardness of these samples. Although the hardness
value of ø 15 mm diameter samples with hot mold in-
creases linearly for 600 s with frequencies of 15 and
17.5 kHz, UST does not cause a severe change in hard-
Fig. 5. Secondary dendrite arm spacing values of samples ness after 300 s for 20 kHz frequency. In addition to
treated with different ultrasonic parameters. these results, it is seen that the hardness values of hot
mold are generally lower than the cold mold samples
since the 300 ◦C mold temperature causes slow cool-
versely. Hardness and tensile test measurements were ing. As a result, it was determined that the hardness
carried out to determine the effect of UST on the me- of all samples increased at different ratios by ultra-
chanical properties of the AlSi10Mg alloy. sonic vibration application. Both ultrasonic and T6
Figure 6(a) shows the tensile strength values of ø heat treated ø 15 mm diameter hot mold samples have
15 mm diameter hot mold samples with the effect of maximum hardness values about 114 HB. The rea-
UST combined with T6 heat treatment. The tensile son for the increase in hardness is thought to be the
strength value without UST and T6 heat treatment of modification of microstructure by UST. Similar results
ø 15 mm diameter hot mold sample was 133 MPa. This were reported by Unal et al. [21] that ultrasonic treat-
value is increased after UST for 600 s to 152.8 MPa ment favorably affected hyper-eutectic cast Al-Si alloy
for 15 kHz, 174.8 MPa for 17.5 kHz and 205.6 MPa for hardness values and provided an increase of 15–20 %.
20 kHz. It can be inferred from Fig. 6a that UST with Also, Zhai et al. [31] and Ruirun et al. [32] reported
all frequencies linearly increases the tensile strength that the application of ultrasonic vibration increased
up to 180 s. However, after this period UST causes a micro-hardness of aluminum alloys.
nearly linear decreasing trend of the tensile strength.
The tensile strength of the ø 15 mm diameter hot
mold sample subjected to T6 heat treatment without 4. Discussion
UST was 187.29 MPa. After performing UST for 600 s
and T6 heat treatment, the tensile strength of the According to the obtained results, the primary
ø 15 mm diameter hot mold sample rose to 216.3 MPa Al phase evolved from coarse dendrites into nearly
for 15 kHz, 258.2 MPa for 17.5 kHz and 293.59 MPa equiaxed grains. Meanwhile, UST application at the
for 20 kHz. When the combined effect of UST and T6 liquid state resulted in global – equiaxed grains in
heat treatment is examined, it can be expressed that previous researches [11, 15, 19]. Therefore, it has
the tensile strength values for all treatment frequen- been provided that the α-Al dendrites transform into
cies increase in a wavy pattern. The reason for fluctu- equiaxial grains when the ultrasonic vibration is ap-
ating increase in curves is thought to be due to some plied entirely in the liquid phase.
 F. Vatansever et al. / Kovove Mater. 57 2019 33–43 41

Fig. 6. Time and frequency dependent effect of UST: (a) tensile strength and (b) hardness.

It has been inferred that the primary sources local high undercooling which results in nucleation.
of refining primary Al phase and changes in micro- The alternating pressure of ultrasound above the cav-
structure concerning the phase transformation and itation threshold creates extra unstable and continu-
morphological evolution of alloys by UST are the ously growing gas bubbles [8]. The swollen gas bubbles
collapse of small cavities, dendritic fragmentation, at negative acoustic (tensile force) half-period (see in
and wetting due to the cavitation-induced nucleation Fig. 7) of the ultrasonic wave absorb the thermal en-
mechanism. As some researchers demonstrated, the ergy from the melt-bubble interface. Positive acoustic
improvements in microstructure and mechanical prop- pressure (compression force) in a half-period of ultra-
erties arise from cavitation induced heterogeneous nu- sonic wave causes instantaneous imploding of bubbles.
cleation [8, 12, 17, 19, 29–31, 33, 34] and fragmen- Arising of local melting temperature point due to a
tation of the dendritic structure [15, 22, 23, 25] due bubble collapse locally causes undercooling of the liq-
to ultrasonic vibration. As previously mentioned in uid. Thus, nuclei are created by increasing the melting
this study, ultrasonic vibration was applied at a fully point at a bubble surface. A cavitation bubble collapse
liquid state (at 700 ◦C) of AlSi10Mg alloy. At the liq- causes breaking of oxide layer due to the energetic
uid phase, no dendrites occurred inherently. For this shock waves (see Fig. 7) [29]. Nucleation is localized
reason, it is thought that cavitation induced hetero- in the vicinity of impurities. The role of bubbles is
geneous nucleation is the main mechanism for re- identical to that of chemical grain modifier elements.
fining the microstructure of AlSi10Mg alloy in this Cavitation-induced high undercooling and fragmenta-
study. tion creates refined primary aluminum phase.
Three possible different grain refinement mecha- II. The second one is cavitation-induced dendrite
nisms in UST are consequently explained by assem- fragmentation and grain multiplication. Bubble col-
bling information from previous studies and our in- lapsing generates shock waves and sharp transient
vestigation: pressure spikes (up to 5 GPa) [15, 30]. An acous-
I. The first one is cavitation induced heteroge- tic stream of the shock waves scatters the nucleation
neous nucleation, which is described below: Ultra- throughout the melt, breaks dendrite arms and sin-
sonic wave produces cavitation. Cavitation induces gle crystals [11]. On the other hand, shock waves can
42 F. Vatansever et al. / Kovove Mater. 57 2019 33–43

Fig. 7. Schematic diagram of grain refinement and nucleation mechanisms in UST.

result in homogeneous nucleation by promoting under- r is the secondary liquid droplet radius, ∂σ/∂T and
cooling (see in Fig. 7) [11]. Pressure pulse leads to a lo- ∂T /∂x are interfacial energy gradient and tempera-
cal elevation in melting temperature [15, 30]. Thus, lo- ture gradient, respectively.
cally freezing liquid phase causes globular grains in so-
lute fields. Changing liquidus temperature with pres-
sure PL can be calculated by the Clausius–Clapeyron 5. Conclusions
equation [15, 36]:
Following conclusions can be drawn:
Tmelting ΔV – Microstructural examination and experimental
Tliquidus = Tmelting + (Pliquidus − P 0 ) ,
ΔHm results show that UST improves microstructure and
(1) mechanical properties of the AlSi10Mg alloy. The rea-
where P0 is atmospheric pressure, ΔV and ΔHm are son for this improvement is the evolution of primary
the volume and enthalpy changes in the liquid-solid α-Al phase from coarse dendrites to smaller dendrites.
transformation [15, 36]. Cavitation induced dendrite Also, the acicular eutectic-Si phase became smaller by
fragmentation solely happens provided that the melt the effect of UST. It is concluded that the predom-
includes solid fraction volume. Contrarily to that cav- inant refinement mechanism was cavitation induced
itation enhanced heterogeneous nucleation occurs in heterogenous nucleation because of pressure and tem-
liquid alloys. perature fluctuations.
III. The third one is a vibration-stimulated sepa- – Modification of microstructure and mechanical
ration of wall crystals: According to this theory, UST properties of the alloy increased with increasing UST
increases the wettability of the mold wall with an alloy time for all frequencies and samples. However, no sig-
which promotes the nucleation of the alloy from mold nificant effect of UST frequency on microstructure and
to center [29]. Theoretical approach of Marangoni con- mechanical properties has been observed.
vection (MC) is that an imbalance of surface tension – The UST brings an increase in hardness and ten-
on a liquid interface causes capillary motion. The tem- sile strength until the application period of 300 s for
perature gradient of cooling melt can be driven by the all frequencies, but the vibration applied after 300 s
MC of a spherical liquid-liquid interface under reduced does not significantly influence hardness and tensile
gravity conditions with UST. The MC causes direc- strength of the material.
tional solidification from the mold wall to the center – Maximum hardness tensile strength is obtained
of the molten alloy. The motion (VM ) independently with the combined UST and T6 heat treatment appli-
of solidification direction can be described as [31]: cation.

−K1 r ∂σ ∂T
VM = , (2) Acknowledgement
(2K1 − K2 ) (2U1 − 3U2 ) ∂T ∂x
The authors gratefully acknowledge the support of the
where K values are the thermal conductivity, U values Scientific Research Coordination Foundation (BAP Unit,
are the viscosities of the parent and secondary liquids, Project No. 2017/002) of Kocaeli University.
 F. Vatansever et al. / Kovove Mater. 57 2019 33–43 43

References [19] Puga, H., Barbosa, J., Costa, S., Ribeiro, J. S., Prokic,
M.: J. Mater. Process. Technol., 211, 2011, p. 1729.
[1] Maung, K. N., Yoshida, T., Liu, G., Lwin, C. M., doi:10.1016/j.jmatprotec.2011.05.012
Muller, D. B., Hashimoto, S.: Resour. Conserv. Re- [20] Jian, X., Meek, T. T., Han, Q.: Scr. Mater., 54, 2006,
cycl., 126, 2017, p. 34. p. 893. doi:10.1016/j.scriptamat.2005.11.004
doi:10.1016/j.resconrec.2017.06.016 [21] Ünal, N., Çamurlu, H. E., Koçak, S., Düztepe, G.: Int.
[2] Karabay, S., Ertürk, A. T., Zeren, M., Yamanoglu, J. Cast. Met. Res., 25, 2012, p. 246.
R., Karakulak, E.: Eng. Fail. Anal., 83, 2018, p. 47. doi:10.1179/1743133612Y.0000000011
doi:10.1016/j.engfailanal.2017.09.007 [22] Zhang, L., Eskin, D. G., Miroux, A., Katgerman, L.:
[3] Feyzullahoğlu, E., Ertürk, A. T., Güven, E. A.: Light Metals. Hoboken, Wiley 2012.
Trans. Nonferrous Met. Soc. China, 23, 2013, p. 3575. doi:10.1007/978-3-319-48179-1 174
doi:10.1016/S1003-6326(13)62903-9 [23] Eskin, D. G., Atamanenko, T. V., Zhang, L., Katger-
[4] Karabay, S., Güven, E. A., Ertürk, A. T.: Eng. Fail. man, L.: Essent. Readings Light Met., 3, 2013, p. 415.
Anal., 31, 2013, p. 153. doi:10.1002/9781118647783.ch51
doi:10.1016/j.engfailanal.2013.02.005 [24] Yao, L., Hao, H., Ji, S. H., Fang, C. F., Zhang, X. G.:
[5] Ertürk, A. T., Şahin, M., Aras, M.: Trans. Indian Inst. Trans. Nonferrous Met. Soc. China, 21, 2011, p. 1241.
Met., 70, 2017, p. 1233. doi:10.1016/S1003-6326(11)60848-0
doi:10.1007/s12666-016-0915-7 [25] Patel, B., Chaudhari, G. P., Bhingole, P. P.: Mater.
[6] Eskin, D. G.: Mater. Sci. Forum, 828, 2015, p. 112. Lett., 66, 2012, p. 335.
doi:10.4028/www.scientific.net/MSF.828-829.112 doi:10.1016/j.matlet.2011.08.113
[7] Eskin, G. I., Eskin, D. G.: Ultrasonic Treatment of [26] Raghu, R., Nampoothiri, J., Kumar, T. S.: Measure-
Light Alloy Melts. Boca Raton, CRC Press 2015. ment, 129, 2018, p. 389.
[8] Puga, H., Barbosa, J., Costa, S., Ribeiro, J. S., Pinto, doi:10.1016/j.measurement.2018.07.056
A. M. P., Prokic, M.: Mater. Sci. Eng. A, 560, 2013, [27] Liang, G., Shi, C., Zhou, Y., Mao, D.: Metals, 6, 2016,
p. 589. doi:10.1016/j.msea.2012.09.106 p. 260. doi:10.3390/met6110260
[9] Youn, J. I., Kang, B. I., Ko, D. G., Kim, Y. J.: Int. J. [28] Haghayeghi, R., Heydari, A., Kapranos, P.: Mater.
Cast. Met. Res., 21, 2008, p. 135. Lett., 153, 2015, p. 175.
doi:10.1179/136404608X361828 doi:10.1016/j.matlet.2015.04.034
[10] Feng, H. K., Yu, S. R., Li, Y. L., Gong, L. Y.: [29] Tuan, N. Q., Puga, H., Barbosa, J., Pinto, A. M. P.:
J. Mater. Process. Technol., 208, 2008, p. 330. Met. Mater. Int., 21, 2015, p. 72.
doi:10.1016/j.jmatprotec.2007.12.121 doi:10.1007/s12540-015-1008-6
[11] Zhang, Y. B., Lu, Y. P., Jie, J. C., Fu, Y., Zhong, D., [30] Jiang, R. P., Li, X. Q., Zhang, M.: Met. Mater. Int.,
Li, T.: Trans. Nonferrous Met. Soc. China, 23, 2013, 21, 2015, p. 104. doi:10.1007/s12540-015-1012-x
p. 3242. doi:10.1016/S1003-6326(13)62859-9 [31] Zhai, W., Liu, H., Zuo, P., Zhu, X., Wei, B.: Prog.
[12] Kotadia, H. R., Das, A.: J. Alloys Compd., 620, 2015, Nat. Sci. Mater. Int., 25, 2015, p. 471.
p. 1. doi:10.1016/j.jallcom.2014.09.089 doi:10.1016/j.pnsc.2015.10.006
[13] Komarov, S., Ishiwata, Y., Mikhailov, I.: Metall. [32] Ruirun, C., Deshuang, Z., Jingjie, G., Tengfei, M.,
Mater. Trans. A Phys. Metall. Mater. Sci., 46, 2015, Hongsheng, D., Yanqing, S., Hengzhi, F.: Mater. Sci.
p. 2876. doi:10.1007/s11661-015-2829-0 Eng. A, 653, 2015, p. 23.
[14] Jung, J. G., Lee, S. H., Lee, J. M., Cho, Y. H., Kim, doi:10.1016/j.msea.2015.12.001
S. H., Yoon, W. H.: Mater. Sci. Eng. A, 669, 2016, p. [33] Lü, S., Wu, S., Lin, C., Hu, Z., An, P.: Mater. Sci. Eng.
187. doi:10.1016/j.msea.2016.05.087 A, 528, 2011, p. 8635. doi:10.1016/j.msea.2011.08.014
[15] Wang, J. Y., Wang, B. J., Huang, L. F.: J. Mater. Sci. [34] Atamanenko, T. V., Eskin, D. G., Katgerman, L.: Int.
Technol., 33, 2017, p. 1235. J. Cast. Met. Res., 22, 2009, p. 26.
doi:10.1016/j.jmst.2017.07.018 doi:10.1179/136404609X367254
[16] Haghayeghi, R., Paula, L. C., Zoqui, E. J.: J. Mater. [35] Geiser, W.: Aluminium Alloy Tables. Zurich, Alcan
Eng. Perform., 26, 2017, p. 1900. S.A. 1982.
doi:10.1007/s11665-017-2602-1 [36] Li, J. W., Momono, T., Fu, Y., Zheng, J., Tayu, Y.:
[17] Das, A., Kotadia, H. R.: Mater. Chem. Phys., 125, Trans. Nonferrous Met. Soc. China, 17, 2007, p. 691.
2011, p. 853. doi:10.1016/j.matchemphys.2010.09.035 doi:10.1016/S1003-6326(07)60158-7
[18] Zhang. Y., Jie. J., Gao. Y., Lu, Y., Li, T.: Inter-
metallics, 42, 2013, p. 120.
doi:10.1016/j.intermet.2013.05.018