h a n g e Vi h a n g e Vi

XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c .d o .c
c u -tr a c k c u -tr a c k
Ultrasonic grain refinement of die cast copper alloys
Puga H1*, Barbosa J1, Machado J. M.1, Vilarinho C2

1
CMEMS - Centre for Micro-Electro Mechanical Systems
2
MEtRiCS - Mechanical Engineering and Resource Sustainability Center

University of Minho, Campus of Azurém, 4800-058 Guimarães, Portugal

Tel.: +351 253 510 220

T
IP
Author Correspondence: puga@dem.uminho.pt

R
Abstract

SC
In this study, the effect of ultrasound as an alternative melt refinement technique of
commercially available brass alloys and its impact in the microstructural and mechanical

U
properties is analysed. A Cu-Zn alloy was cast in different conditions: without melt treatment,
N
chemically refined and ultrasonically refined at different temperatures. Ultrasonic treatment was
found to be more efficient than chemical refinement, promoting lower grain size and higher
A

roundness of the -phase and decreased the ratio of the / phases. The best results were
M

achieved for 940 ºC melt refinement temperature. The microstructure morphology improved the
alloy mechanical properties, namely ultimate tensile strength, yield stress and elongation, which
ED

attained 386 MPa, 192 MPa and 11.5%, respectively, for that melt refinement temperature.
PT

Keywords: Ultrasonic treatment; Copper alloys; Grain refinement; Casting

1. Introduction
E
CC

The casting of brass alloys traditionally leads to coarse and heterogeneous microstructures, even
when the die casting process is used, requiring suitable melt treatment to refine and modify the
A

microstructure morphology to promote an increase in the castings sanity and mechanical

behaviour. As already demonstrated by Sadayappan (2002), together with the pouring
temperature, grain refinement is a required treatment to enhance the mechanical performance of
the final product and the alloy fluidity/castability characteristics. Since high cooling rates are not
always achievable, particularly when high thickness sections are present in the casting or low
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c do . .c
c u -tr a c k c u -tr a c k
thermal conductivity moulding materials are used (sand, ceramic materials), the most frequently
used refinement technique is the addition of nucleating additives to the melt. As assessed by
Sadayappan (2002), the use of common chemical refiners exhibits a fading effect over time,
resulting in a loss of its effectiveness, which is particularly important in situations where the time
interval between the refinement operation and pouring is long. Moreover, chemical refinement
usually leads to significant amount of dross, which entails an important environmental impact,
and the process is difficult to control.

T
Grain refinement is crucial to achieve high mechanical properties, being imperative to find

IP
viable alternative routes, more efficient and clean, which may lead to the development of new
melt processing technologies. The ultrasonic treatment, successfully used with aluminium and

R
magnesium alloys, can be a promising alternative to promote brass grain refinement through the

SC
cavitation effect, as stated by Qian et al. (2009). Although the ultrasonic treatment is well
established in the treatment of the Al and Mg alloy melts, as demonstrated by Puga et al. (2011)
and Eskin et al. (2017), there are no studies of this application in the treatment of liquid brass
U
(Cu-Zn alloys). This work describes the development and evaluation of the ultrasonic refinement
N
technique applied to brass alloys, capable of significantly improving their mechanical properties
A
when compared to those obtained through traditional chemical refinement techniques.
M

2. Experimental Procedure
2.1 Cu-Zn alloy
ED

Table 1 presents the chemical composition of the Cu-Zn alloy obtained by X-ray fluorescence
PT

spectrometry (XRF).
Table 1. Chemical composition of the Cu-Zn alloy.
E

Element Cu Zn Fe Mn Ni Pb Sn others
Wt% 59.30 36.60 0.85 0.05 0.30 1.90 0.46 0.54
CC

Despite the increasing use of unleaded brass compositions, for most applications that do not have
direct contact with water for human consumption, lead is still used to facilitate the casting and
A

machining operations, as it is the case of the raw material used in this work, which was supplied
by an industrial company. Although the presence of high lead contents in the alloy suggests that
it may impair the microstructure, its presence promotes machinability enhancing due to smooth
chip breakage and lower tool insert wear, as referred by Christoph et al. (2014).
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c do . .c
c u -tr a c k c u -tr a c k
Fig. 1 presents the phase diagram of Cu-Zn alloy. The temperatures of isothermal ultrasonic melt
refinement (TUS_1=900 ºC, TUS_2=920 ºC and TUS_3=940 ºC) have been selected considering that
they are located between the solidus-liquidus range of temperature, thus, a slight in-situ
supercooling caused by cavitation can generate the formation of solidification nuclei. For the
sake of comparison, melt refinement without (TW/T) and with addition of chemical elements
(TChemical((Cu-B refiner)) at 1020 ºC were performed.

T
R IP
SC
U
N
Fig. 1. Cu-Zn phase diagram with the indication of the temperatures selected to carry out melt refinement, for
different techniques.
A
M

2.2 Experimental apparatus

ED

In Fig. 2 is shown the experimental apparatus. The device consisted in a crucible with a capacity
of 20 kg of Cu-Zn alloy in a resistance melting furnace, a positioning and holding device of the
ultrasound unit, which employs the multi-frequency-mode-modulated technology (MMM). The
PT

ultrasonic energy (3000 W) was driven from a transducer by an acoustic waveguide with 35 mm
diameter and 420 mm long (4 HWL) to the clamp-on where a sialon tube with 60 mm diameter
E

and 1200 mm length was attached in order to transfer the acoustic vibration to the liquid melt.
CC

When operating above the cavitation threshold, the ultrasonic device, based on MMM
technology, promotes the formation of near vacuum bubbles in the melt, as well as improving
A

bubble growth and expansion/shrinking cyclic regimes until their collapse. The volumetric
homogenous cavitation in the melt treatment of non-ferrous alloys can result in melt degassing,
microstructure refinement, and structure modification with a significant alloy density increase, as
demonstrated by Puga et al. (2017).
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c .d o .c
c u -tr a c k c u -tr a c k

T
IP
Fig. 2. Experimental apparatus: (1) Sialon tube; (2) MMM generator; (3) Feedback transducer/MMM generator; (4)

R
Thermocouple control; (5) Crucible with capacity for 20 kg of melt brass; (6) Electric furnace.

SC
The optimal operation parameters, considering both frequency and phase modulation, are
controlled by the MPI software. The vibration frequency and amplitude transmitted by the sialon

U
tube to the melt can be adjusted to promote the maximum acoustic intensity as a function of the
N
medium properties.
A
2.3 Cu-Zn alloy processing
M

Melting was carried out in a crucible with a capacity of 20 kg of Cu-Zn alloy within a resistance
ED

melting furnace where the furnace/crucible was preheated at 650° C for 30 min to remove
moisture. The melts were heated to 1050 ± 10 °C, held at that temperature for 30 min for
homogenization and cleaned by a solid flow (Coveral ®). Melt temperature was then reduced to
PT

1000 ºC and the terminal 150 mm of the acoustic radiator was introduced into the melt for pre-
heating for a period of 5 min. The melt temperature was then reduced to the established
E

processing temperatures for each test, i.e. 900, 920, 940 ± 5 ºC. Ultrasonic treatment was carried
CC

out for 4 min using a resonant frequency of 19.75 kHz ± 200 Hz and 600 W of electric power.
After ultrasound treatment, the alloys were poured into the metallic mould pre-heated to 400 ± 5
A

ºC, as shown in Fig. 3. This mould was supplied by an industrial company working in the field of
sanitary brass components. The mould cavity provides a diameter of 18 mm and 100 mm long
cast part.
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c .d o .c
c u -tr a c k c u -tr a c k

T
IP
Fig. 3. Section view of metallic mould for tensile testing samples and microstructural characterization. (1) Mould;
(2) Cast part; (3) Sprue; (4) Samples for microstructural and hardness characterization; (5) Tensile test specimens.

R
For the sake of comparison, samples were cast without refinement and refined with Cu-2%B

SC
master alloy. In both cases the same procedures of melting, homogenization, and cleaning of the
melt were carried out and the pouring temperature was 1020 ºC. Chemical refinement was

U
carried out using 20 min incubation time. For all conditions, there were 5 distinct casted samples.
N
A
2.4 Cu-Zn alloy characterization
M

Samples for microstructure and hardness characterization were taken from each cast part by
sectioning them perpendicularly to its longitudinal axis at half of length (40 mm from base of
ED

cast part), as presented in Fig. 3. They were ground using successively #320, #600, #800, #1200
SiC paper and polished up to 1 m. The polished samples used for optical microscopy and
PT

hardness characterization were etched using a chemical reagent composed by 5 g FeCl + 2 ml

HCl + 96 ml alcohol in order to reveal the  and -phases. Optical microscopy (OM) equipped
quantitative metallographic analysis capability to evaluate  and -phase shape were used.
E
CC

ImageJ Software was used to determine the shape and area of the phases. 20 measurements per
optical field (200×) were performed on 5 different fields. To simplify and allow the
quantification of  and -phase morphology in the polished samples, the measured area in the
A

analysis plane was considered equivalent to its fraction in volume. The  phase average diameter
and  phase roundness characteristics were estimated in accordance to ASTM Standard E112-10
and equations (1) and (2), where A and P are the  phase area and perimeter, respectively.
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c .d o .c
c u -tr a c k c u -tr a c k
𝐴
𝛼 𝑝ℎ𝑎𝑠𝑒 𝑑𝑖𝑎𝑚𝑒𝑡𝑒𝑟 = 2 × √𝜋 (1)
4×𝜋×𝐴
𝛼 𝑝ℎ𝑎𝑠𝑒 𝑟𝑜𝑢𝑛𝑑𝑛𝑒𝑠𝑠 = (2)
𝑃2

Vickers hardness measurement was performed on polished samples with a diamond square-based
pyramid under a load of 1 kgf for 15 s, using a Shimadzu HMV-2 equipment. The measurements
for and-phase hardness were repeated 10 times from which the mean value and standard
deviation were calculated.

T
IP
Five tensile test specimens for each processing condition were machined from the as-cast
samples with cross section of do of 10 ± 0.075 mm and a testing length of 50 ± 0.05 mm. Tensile

R
testing was performed according to ISO6892 in a displacement control with a strain rate of 0.25

SC
mm/min while monitoring the imposed reaction forces by the load cell. The characterization of
the different testing parameters, namely the yield stress at 0.2%, the tensile stress (rupture

U
stress), and the elongation of the specimens submitted to the stress tests were obtained using
video-extensometry technique.
N
A
3. Results and Discussion
M

3.1 Microstructure
Fig. 4 shows the microstructure of the castings produced without and with chemical refinement.
It is clear the presence of a stretched and dendritic -phase (L > 100 m) and an interdendritic
ED

-phase mainly in microstructures of the samples without treatment (Fig. 4 (a and b)).
Furthermore, the microstructural morphology of these samples is characterized by classical
PT

Widmanstäetten structures reported in bibliography by Chih-Chun et al. (2013) and Stanford and
Bate (2005). With the addition of Cu-2%B mater alloy refiner, the homogeneity of and -
E

phase becomes more evident and many Widmanstäetten structures can be observed with a
CC

random orientation (Fig. 4(c)). Overall, the effect of chemical refinement is quite clear as can be
observed by Fig. 4(c). The -phase is significantly refined, showing an elongated shape (average
A

length of ≈45 mm) with dimensions smaller than those observed in the castings that were not
refined. Additionally, due to the reduced solubility of lead in this type of alloys (0.04% at 800
°C), its presence is in the form of heterogeneously dispersed globules along the matrix, as shown
in Fig. 4. Although its presence has a beneficial effect in the machining operations (lubricating
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c do . .c
c u -tr a c k c u -tr a c k
effect), it has the disadvantage of contributing negatively to the mechanical properties, namely
the maximum tensile strength.

T
R IP
SC
U
N
A
M
ED

Fig. 4. Example of microstructures obtained without and with chemical refinement. (a) and (b) Without treatment;
(c) and (d) Chemical refinement.
PT

Fig. 5 shows an example of microstructures obtained with ultrasonic refinement for different
E

isothermal processing temperatures. By comparison with the microstructure of those samples

without and with chemical refinement, the ultrasonic treatment promoted a perceptible and very
CC

significant grain refinement, with the length of -phase grains decreasing by comparison with
both the non-refined and the chemically refined samples. Moreover, Fig. 5 suggests that the
A

refinement results, i.e. the reduction of the -phase grain size, are more effective for higher melt
treatment temperatures. Furthermore, with increasing ultrasonic treatment temperature, the
Widmanstäetten structures are less perceptible and present shorter plates and acicular
morphology with random orientation.
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c .d o .c
c u -tr a c k c u -tr a c k

T
R IP
SC
U
N
A
M
ED
E PT
CC

Fig. 5. Example of microstructures obtained with ultrasonic refinement. (a) 900 ºC; (b) 920 ºC; (c) 940 ºC.

Fig. 6 presents the size and morphology (Fig. 6(a)) and the hardness (Fig. 6(b)) of and-
A

phases observed in the microstructures presented in Figs. 4 and 5. From Fig. 6(a), it is evident
that the matrix refinement has a notorious role on the -phase morphology and the final
microstructure depends in the processing conditions as well as in the ultrasonic treatment
temperature.
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c .d o .c
c u -tr a c k c u -tr a c k

T
Fig. 6. Microstructural characteristics of processing conditions in the: (a) -phase shape; (b) samples hardness.

IP
For chemical grain refinement, the use of master alloy Cu-B promoted a reduction of the -phase

R
from an average size of 40 m to 25 m. Besides, the roundness of the -phase increased 10%

SC
when compared with the non-treated alloy. Concerning the phase hardness (Fig. 6(b)), both
microstructures (chemically refined and non-refined) present the same average hardness level,

U
but the hardness of the -phase is lower than that of the -phase, which agrees with the reports
N
of Haruhiko et al. (2010).
A
Regarding ultrasonic refinement, the acoustic energy promoted a reduction to 25, 21, and 18 m
M

in the average size of the -phase, while the average roundness increased to 0.79, 0.84, and 0.87,
for processing temperatures of 900, 920, and 940 ºC, respectively. According to Fig. 6(b) the
ED

hardness of both phases directly increased with the ultrasonic processing temperature. Moreover,
the hardness standard deviation suggests that the matrix homogeneity of ultrasound-treated
PT

samples is more uniform when compared to non-ultrasound-treated ones and tends to increase
with the increasing of treatment temperature. Furthermore, the experimental results show that the
area ratio of  phases decreased from 1.43 and 1.33 in the cases of non-refined and
E

chemically refined, respectively, to 1.20, 1.05, and 1.04 in the ultrasonic refined samples at 900,
CC

920, and 940 ºC, respectively. Indeed, the lattice structure of the -phase is face-centred cubic
(fcc) (as reported by Haquey and Khan (2008)) and therefore, brasses with high amount of -
A

phase tend to present low hardness, as stated by Christoph et al. (2014). Additionally, increasing
the -phase will favour the surface quality of the machined brass components.

To date, two mechanisms are suggested to justify the microstructure refinement using
ultrasound: dendritic fragmentation and heterogeneous nucleation induced by cavitation
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c . .c
c u -tr a c k
development. However, this heterogeneous nucleation seems to be a valid hypothesis and diso c u -tr a c k
defended by most investigators who have worked in this field. In this case, it is suggested that
the theory of heterogeneous nucleation induced by cavitation implies that the rapid adiabatic
expansion of the gas inside the cavitation bubbles causes, at the bubble-liquid interface, an
overcooling in the liquid metal that is sufficient to generate solid cores. In addition, the acoustic
fluxes resulting from the collapse of the bubbles promote their distribution throughout the liquid
mass, dispersing numerous solidification nuclei promoting heterogeneous nucleation and phase’s
homogenization.

T
IP
Concerning the brass melts treated by ultrasound when compared to those melts with chemical
treatment or without treatment, there is a contrast in the -phase volume fraction. This difference

R
in the amount of α-phase is related to the homogenization of the cooling rate of the mass to be

SC
solidified, contrary to what happens in the melts with and without chemical treatment in which,
according to the Cu-Zn equilibrium diagram, the volume fraction of β and β'-phases tend to

U
increase as the temperature increases. In fact, by comparing the microstructures of ultrasound-
treated samples at 900 ºC and 940 ºC, the volume fraction of the -phase increased from 45.5%
N
to 48.9%. Moreover, an additional mechanism can be suggested as responsible for the changes in
A
the α and β-phases. Khosro and Niroumand (2011) suggested that, in the half-cycle bubble
M

compression, there is an in-situ increase of melt temperature promoting local melting of the
phases. The proposed effect can be related to the increase in the number of nucleation particles
as well as promote their spheroidization, especially the α-phase, as can be observed in Fig. 6(a).
ED

This effect can be observed in Fig. 7, which presents the cooling curve and its 1st derivative
obtained for samples without refinement and with ultrasonic refinement at 940 ºC. It is clear that
PT

ultrasound increases the alloy liquidus temperature (promoting α-phase homogenization) and
decreases the cooling rate for lower temperatures.
E
CC
A
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c .d o .c
c u -tr a c k c u -tr a c k

T
Fig. 7. Cooling rate of samples without treatment (as-cast: red curve) and samples refined by ultrasound (case of
ultrasonic refinement at 940 ºC: blue curve).

R IP
3.2 Mechanical properties

SC
Table 2 presents the ultimate tensile strength (UTS), yield stress, and elongation of the Cu-Zn
alloy for different melt refinement techniques.

U
Table 2. Mechanical properties of cast brass samples for refinement techniques.
N
Refinement technique UTS Yield Stress Elongation
A
(MPa) (MPa) (%)
M

Not refined 282 (± 30) 178 (± 14) 4.3 (± 1.0)

Chemical refinement 310 (± 15) 181 (± 11) 5.0 (± 0.7)
Ultrasonic refinement at 900 ºC 345 (± 15) 183 (± 10) 6.9 (± 0.5)
ED

Ultrasonic refinement at 920 ºC 367 (± 12) 187 (± 7) 9.4 (± 0.6)

Ultrasonic refinement at 940 ºC 386 (± 10) 192 (± 7) 11.5 (± 0.6)
PT

In those samples without any treatment, the average UTS was 282 MPa which increased to 310
E

MPa in those samples chemically refined. In those samples ultrasonically isothermally refined
CC

for 4 minutes at 900, 920, and 940 ºC, UTS increased to 345, 367, and 386 MPa respectively. In
what concerns to the average yield stress it increased from 178 to 181 MPa with chemical
treatment and to 183, 187, and 192 MPa after isothermal ultrasonic refinement for 4 min at 900,
A

920, and 940 ° C, respectively. Average elongation increased from 4.3 to 5.0% with chemical
treatment and to 6.9, 9.4, and 11.5% after ultrasonic refinement for 4 minutes at 900, 920, and
940 ° C, respectively. The acoustic treatment enhanced the mechanical properties of the Cu-Zn
alloy, with particular relevance in the material ductility.
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c do . .c
c u -tr a c k c u -tr a c k
Fig. 8 shows that the stress-strain curve following the same trend, although the mechanical
properties of each sample significantly changed with the processing conditions. This trend can be
confirmed by equation (3), where the correlation between tensile strength (UTSCu-Zn) and the
strain () is linear.

T
R IP
SC
Fig. 8. Stress-strain curve of tensile testing for the different processing conditions.

𝑈𝑇𝑆𝐶𝑢−𝑍𝑛 = 1355.5𝜀 + 236.5

U (3)
N
A
From these results, it is clear that the chemical treatment traditionally used for microstructure
refinement of such alloys has relatively low effectiveness when compared to ultrasonic
M

refinement. Although some refinement can be observed (Fig. 4(c)), the columnar -phase
throughout the matrix reveals that the procedure was not effectiveness. Due to considerable
ED

amount of iron (0.85%) present in the Cu-Zn brass alloy used in this work, it can be suggested
that the boron has potential to work as a catalyst, thus being able to promote the nucleation of
PT

iron particles, resulting in a weak matrix refinement, as demonstrated by Sadayappan et al.

(2004).
E

Fig. 9 presents the correlation between elongation and -phase roundness for the different
CC

processing conditions. With the decrease of  phases area ratio (↓and ↑ the average
hardness of the samples {A()% × Hv() + A()% × Hv()} as well as the elongation
A

increased. This evidence can be attributed to changes in the morphology of the -phase, i.e.,
according to Figs. 4, 5, and 9, it is evident that with the increase of the ultrasonic treatment
temperature, the -phase morphology changed from large needle block-like shape to a more
globular shape.
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c .d o .c
c u -tr a c k c u -tr a c k

T
IP
Fig. 9. Correlation between Elongation and α-phase roundness for different processing conditions.

R
SC
According to the microstructural and mechanical results, it is possible to suggest a relation
between the yield stress and the  area ratio of the phases represented by the following
equation (4):
U
N
𝛼
𝑌𝑆𝐶𝑢−𝑍𝑛 = −29.8 × (𝛽) + 220.5 (4)
A
M

Through equation (4), it is clear that the yield stress (YSCu-Zn) has a linear
regression with the  phases area ratio. The strengthening mechanisms of the brass
ED

alloy treated by ultrasound may be attributed to: (i) homogenization of α and

β phase in the solid solution due to acoustic effects; (ii) decreasing of
PT

α/β phases area ratio.

E

Fig. 10 shows the fracture surface of the cast samples submitted to tensile testing for different
CC

processing conditions. It appears that ultrasonic vibration also has an important role in melt
cleaning, avoiding the presence of inclusions. In fact, the fracture surface of the non-treated and
the chemically treated samples reveals the presence of non-metallic inclusions, as demonstrated
A

by Fig. 10 (a and b). On the other hand, those samples treated by ultrasound reveal a cleaner
surface, showing no traces of non-metallic inclusions, as shown in Fig. 10 (c).
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c .d o .c
c u -tr a c k c u -tr a c k

T
R IP
SC
U
Fig. 10. Fracture surface of the cast samples submitted to tensile testing: (a) Without treatment; (b) Chemical
N
treatment; (c) Isothermal treatment for 4 minutes by ultrasound at 940 °C.
A
Although ultrasonic processing of light alloy melts is able to remove non-metallic impurities,
M

available studies presented until now have only been validated for light alloys, namely
aluminium and magnesium alloys. The experimental results of this work reveal that acoustic
ED

excitation of brass alloy melts promoted by ultrasound is favourable to the removal/decrease of

the content of non-metallic inclusions, as well as the reduction of the level of porosities in the
final castings. The removal of such inclusions occurs simultaneously with the refinement process
PT

due to the level of cavitation that is generated in the melts. This cavitation is able to agglomerate,
and consequently remove by flotation, those inclusions, as referred by some authors Eskin et al.
E

(2017) and Shao et al. (2012). Considering that the flows generated during the collapse of a
CC

certain number of bubbles (implosion) favour the transport of the inclusions to the surface of the
bubbles and induce fluctuation into the surface of the melt, it can be concluded that the melt
A

treatment of brass alloys by ultrasound promotes not only refinement of the castings
microstructure, but also melt cleaning. Thus, the improvement in the mechanical properties of
the alloy might be assigned to the effect integration of microstructure refinement and
reduction/elimination of porosities and non-metallic inclusions promoted by ultrasound.
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c .d o .c
c u -tr a c k c u -tr a c k
Conclusions

A study was carried out to evaluate the influence of ultrasonic melt treatment in the
microstructure and mechanical properties of the current commercially available brass alloys. The
following conclusions were drawn:

 Ultrasonic melt treatment is suggested to be an efficient and alternative technique to

control the shape and size of  and β-phases of Cu-Zn alloy.

T
 Ultrasonic melt treatment promotes the spheroidization of the -phase, whose diameter

IP
and roundness depend on the melt temperature.
 For a 940 ºC melt treatment temperature, the -phase changed its morphology from a

R
large needle block-like shape to a more globular shape when compared with non-treated Cu-Zn

SC
alloys.
 The best results were reached for Cu-Zn alloy ultrasonic treatment at 940 ºC, promoting

U
the microstructure modification and, as consequence, the enhancement of the alloy’s mechanical
properties.
N
 It is suggested that the strengthening mechanisms of the brass alloy
A
treated by ultrasound are a consequence of the homogenization of α and β
M

phases on solid solution and the decrease of the α/β phases area ratio.
ED

Acknowledgements

This work is supported by the research Post-Doctoral grant SFRH/BPD/76680/2011 and FCT
PT

with the reference project UID/EEA/04436/2013, COMPETE 2020 with the code POCI-01-
0145-FEDER-006941.
E
CC

References

Chistoph N., Fritz K., Dieter L., Sebastian W., (2014). Machinability enhancement of lead-free
A

brass alloy. Procedia CIRP. 14, 95-100. https://doi: 10.1016/j.procir.2014.03.018

Chih-Chun H., Jia-Siang W., Paxon Ti-Yuan W., Weite W., (2013). Microstructure development
of brass alloys with various Bi and Pb additions. Met Mater Int. 19:6, 1173-1179.
https://doi.org/10.1007/s12540-013-6002-2
 h a n g e Vi h a n g e Vi
XC e XC e
F- w F- w
PD

PD
er

er
!

!
W

W
O

O
N

N
y

y
bu

bu
to

to
k

k
lic

lic
C

C
w

w
m

m
w w
w

w
o

o
.d o .c do . .c
c u -tr a c k c u -tr a c k
Eskin G.I., Eskin D.G., (2017). Ultrasonic treatment of light alloy melts. Second Edition. CRC
Press.

Haque M.M., Khan A.A., (2008). Investigation on structure and properties of brass casting. J
Mater Sci Technol. 24:3, 299-301

Haruhiko A., Hisashi I., Shufeng L., Kousaka Y., Kojima A., Kondoh K., (2010). Microstructure
and Mechanical Properties of High Strength Brass Alloy with Some Elements. Mat
Sci Forum. 654-656, 2552-2555. https://doi:10.4028/www.scientific.net/MSF.654-
656.2552

T
Khosro Aghayani M., Niroumand B., (2011). Effects of ultrasonic treatment on microstructure
and tensile strength of AZ91 magnesium alloy. J Alloy Comp. 509, 114-122.

IP
https://doi:10.1016/j.jallcom.2010.08.139

R
Qian M., Ramirez A., Das A., (2009). Ultrasonic refinement of magnesium by cavitation:
Clarifying the role of wall crystals. J Cryst Growth. 311, 3708-3715.

SC
https://doi.org/10.1016/j.jcrysgro.2009.04.036

Sadayappan M., Cousineau D., Zavadil R., Sahoo M., Michels H., (2002). Grain refinement of

U
permanent mold cast copper-base alloys. FS Transactions 02:108, 505–514.
N
Sadayappan M., Thomson J.P., Elboujdaini M., Ping Gu G., Sahoo M., (2004). Grain refinement
A
of permanent mold cast copper base alloys. MTL Report 2004 - 6 (TR-R)

Stanford N., Bate P.S., (2005). Crystallographic variant selection in - brass. Acta Mater. 53,
M

859-867. https://doi.org/10.1016/j.actamat.2004.10.043

Shao W. Z., Le C. Q., Zhang Q. Z., Cui Z. J., (2012). Effects of various parameters on ultrasonic
ED

separation of inclusions from magnesium alloy melt. Materialwiss Werksy. 43:3,

220-225. https://doi:10.1002/mawe.201200843
PT

Puga H., Costa S., Barbosa J., Ribeiro S., Prokic M., (2011). Influence of ultrasonic melt
treatment on microstructure and properties of AlSi9Cu3 alloy. J Mater Process
Tech. 211, 1729-1735. https://doi: 10.1016/j.jmatprotec.2011.05.012
E
CC
A