J Therm Anal Calorim (2015) 119:1593–1601

DOI 10.1007/s10973-014-4367-1

Evaluation of the effect of grain refiners on the solidification

characteristics of an Sr-modified ADC12 die-casting alloy
by cooling curve thermal analysis
Saeed Farahany • Mohd. Hasbullah Idris •
Ali Ourdjini • Fadi Faris • Hamidreza Ghandvar

Received: 20 June 2014 / Accepted: 18 December 2014 / Published online: 4 February 2015
Ó Akadémiai Kiadó, Budapest, Hungary 2015

Abstract The influence of grain refinement—with the Introduction

addition of 6 mass% Al–5Ti–B and 10 mass% Al–3Ti–B
master alloys—on the temperature-based parameters of a The Al–Si–Cu alloy possesses exceptional fluidity and
commercial ADC12 die-casting alloy, modified with 0.04 resistance to hot tearing because of its high silicon content,
mass% Sr, was investigated using computer-aided cooling so that it is capable of producing intricate castings of thin
curve thermal analysis. Results show that the addition of sections. This alloy system is widely used in automotive
grain refiners caused the reaction area of the aluminium applications to produce critical components, such as engine
dendrite to shift up. With the addition of refiner reagents, blocks and cylinder heads. Eutectic or near eutectic alloys,
the nucleation temperature and growth temperature for the containing 10–12.7 mass% Si, have a small freezing range
evolution of the primary Al dendrite increased, while the and exhibit better casting features. Grain refiners, such as
recalescence temperature decreased. The dendrite coher- Al–Ti–B and Al–Ti–C, are commonly used to decrease the
ency temperature (TDCP), time (tDCP), temperature (TN– dendrite size and achieve fine, equiaxed Al grains. Si
TDCP), and time (tN–tDCP) interval for dendrite coherency modification is also a popular melt treatment widely
and solid fraction at coherency point increased with the applied in the Al–Si cast alloys to modify the morphology
addition of Al–3Ti–B and Al–5Ti–B grain refiners. Nev- of coarse Si and to maintain its material quality and
ertheless, the growth temperature of the eutectic alumin- engineering requirements. It is also well established that
ium–silicon phase remained at the modified range the addition of modifier elements causes change in the
(563.5–565 °C) even after the addition of grain refiners. A metallurgical characteristics of eutectic Si, whose structure
grain-refined and modified structure was obtained after the controls the mechanical properties of Al–Si castings [1].
addition of refiners, but no obvious mutual effect was Among the various modifiers, strontium has been used
found in terms of the coarsening of grain size or silicon extensively, as it is easier to handle, more resistant to
demodification in the range of concentrations used in this fading [2, 3] and possesses a recovery rate in the melt to the
study. order of 90 %.
Of late, the challenge for the aluminium industry is to
Keywords Aluminium alloy  ADC12  Grain improve the material properties of alloys in order to meet
refinement  Modification  Interaction  Thermal analysis the growing demands for lighter and stronger materials,
with the cost of production also taken into account. These
requirements have led to the use of melt treatments, such as
S. Farahany  Mohd. H. Idris (&)  A. Ourdjini  F. Faris  grain refinement and modification, to alter the micro-
H. Ghandvar structure and enhance the properties of aluminium cast
Department of Materials, Manufacturing and Industrial
alloys. To date, no single refiner or modifier can fully
Engineering, Faculty of Mechanical Engineering, Universiti
Teknologi Malaysia (UTM), 81310 Johor Bahru, Malaysia meet all customer requirements, especially in the automo-
e-mail: [email protected] tive and aerospace industries; each exerts its own positive
S. Farahany and negative influence on the final structure of the alloy. As
e-mail: [email protected]; [email protected] such, a combination of refiners and modifiers—to obtain

123
1594 S. Farahany et al.

the advantages of both simultaneously—has been used as a The molten alloy was then poured at a temperature of
solution to meet growing customer demands. Over the last 730 ± 5 °C into a preheated ceramic mould for thermal
few years, this combination of refiners and modifiers in analysis. To determine the behaviour and to detect char-
hypoeutectic Al–Si alloys has attracted the attention of acteristic temperatures during alloy solidification, two
many researchers [4–6]. This melt treatment benefits the K-type thermocouples were located in the centre and close
metal casting industry through lower overall cost, due to to the wall of the preheated ceramic mould (see Fig. 1)
combined addition, decreased foundry addition errors, which had solidified in slow conditions. The top and bot-
reduced hot tearing and shrinkage, as well as improved tom of the ceramic mould were insulated to control radial
mechanical properties. However, some interaction between heat flow, so as to accurately measure the coherency point.
Al–Ti–B refiners and Si modifiers has been reported [7–9]. The accuracy of the K-type thermocouple used is ±2.2 °C.
In order to assess the quality of the melting process, Prior to temperature measurement, the thermocouple was
optimise added elements, and detect levels of impurity, the calibrated by two points against the melting point of pure
thermal analysis technique has been widely accepted, since aluminium (660.0 °C) and the freezing point of water
it is a non-destructive and quantitative technique [10] that (0.00 °C). The temperature–time data were captured con-
can be used prior to casting [11]. Recently, computer-aided tinuously at a dynamic rate of 100 Hz/ch and resolution of
cooling curve thermal analyses (CA-CCTA) have been 0.01 °C using a high-speed data acquisition system, which
used extensively to examine the features of Al–Si alloys consists of a EPAD2-TH8-K module, EPAD-BASE and
[12–14]. In short, the development of industrial Al alloys Dewesoft-7-LT software. Then recorded data were impor-
depends heavily on knowing how the application of melt ted into Flexpro 8.0.31 for data analysis. The cooling
treatments, such as grain refinement and eutectic modifi- curves and corresponding derivative curves were plotted to
cation, will affect solidification and the resultant micro- determine nucleation, minimum temperature, and growth
structure. The aim of this study, therefore, is to investigate temperature for primary Al and eutectic formation, based
the effect of strontium modifiers and two types of grain on the first and second derivative cooling curve approach
refiners on the solidification behaviour of near eutectic Al– [12].
Si alloys through the use of CA-CCTA. The results In order to ensure the reproducibility of the testing
obtained can assist casting designers in modelling the results, a thermal analysis of each alloy was carried out at
solidification of Al alloys to meet product specifications. least three times. Samples for metallography were sec-
tioned horizontally at the tip of the thermocouples,
mounted, ground, and polished. The ground specimens
Experimental were then subjected to a final polishing with silica OPS
suspension (0.5 lm). The secondary dendrite arm spacing
A commercially available ADC12 alloy (see chemical (SDAS) was measured according to the linear intercept
composition in Table 1) was melted in a silicon carbide method using an i-Solution image analyser.
crucible using a resistance furnace (Nabertherm) with a
melt temperature of 750 ± 5 °C. Upon thermal equilib-
rium, a weighted Al–10Sr master alloy was added to ensure Results and discussion
that the Sr content in the melt reached 0.04 mass%. The
concentration of Sr was selected based on the author’s Thermal analysis
previously published paper [15]. After allowing 10 min for
homogenisation, the Sr-modified melt was treated by The influence of the addition of 0.04 mass% Sr and that of
introducing the 6 mass% Al–5Ti–B master alloy. The same its combination with the 6 mass% Al–5Ti–B and 10 mass%
process was repeated for the addition of the 10 mass% Al– Al–3Ti–B grain refiners on the solidification behaviour of
3Ti–B master alloy. To investigate the probable interaction the ADC12 alloy is plotted in Fig. 2a. The shape of the
of strontium with boron, the concentration of titanium was cooling curve is the result of heat lost to the surroundings
kept constant for both grain refiner additions. Prior to by the cooling metal and heat evolved in the ceramic cup
casting, both alloys were stirred and surface skimmed to during phase transformation [16]. It can be seen that the
remove dross and other impurities. combination of the individual effects of grain refinement

Table 1 Chemical composition of the ADC12 alloy

Element Si Cu Zn Fe Mn Mg Cr Ni Pb Al

mass% 11.72 2.02 0.805 0.823 0.254 0.220 0.037 0.025 0.009 Balance

123
Cooling curve thermal analysis 1595

and modification produces considerable change in cooling shown in Fig. 2b, c and d, respectively. The cooling curve
behaviour. The formations of primary Al dendrite, eutectic of alloys modified with individual Sr was also plotted for
Al–Si, and eutectic Al2Cu were evaluated through the CA- comparison with the alloys treated with a combination of
CCTA technique. The cooling curve portion associated Sr and two different Al–Ti–B grain refiners (Al–3Ti–B and
with these three phase reactions in the alloys treated with Al–5Ti–B). The influence of the addition of the 6 mass%
the combined strontium modifier and Al–Ti–B refiner is Al–5Ti–B and 10 mass% Al–3Ti–B grain refiners to the Sr-
modified alloys is clearly observed on the cooling curves of
aluminium dendrite formation and eutectic reaction. The
K-type thermocouples temperature-based parameters in terms of nucleation tem-
perature (TN), growth temperature (TG), and recalescence
Insulator plate
(DTR) temperatures were analysed to examine the evolu-
Ceramic
tion of the primary Al and Al–Si eutectic phase during
14.7 mm
mould cooling.
Figure 3 presents the typical cooling curve of 0.04
mass% Sr-modified ADC12 alloy at the centre of the
ceramic cup (TC), and its first and second derivative curves,
which were used to determine characteristic temperatures.
40 mm
20 mm

The recording on cooling curve on the wall of the cup

(TW), shown in Fig. 3, was based on the assumption that
Φ 30 mm
Φ 40 mm
the molten metal at the wall is cooler than that at the
centre; this shows that the aluminium dendrite phase
Insulator plate nucleates at the wall before moving towards the centre. The
Fig. 1 Schematics diagram of ceramic mould with embedded
temperature difference between the centre and wall of the
thermocouples for thermal analysis test ceramic cup (DT = TW-TC) was used to determine the

680
1 (a)
Sr
640 2
Aluminium Sr+Al3TiB
dendrite 3
Sr+Al5TiB
Temperature/°C

600

560
Eutectic
Al-Al2Cu
520
Eutectic Al-Si

480

440
0 200 400 600 800 1000
Time/s

594 576 500

2 (b) (c) (d)
588 572 498
Temperature/°C

Temperature/°C
Temperature/°C

582 3 568 496

2 2
576 1 564 494
1
570 560 492
1
3 3
564 556 490
80 120 160 200 240 270 360 450 540 630 850 860 870 880 890 900
Time/s Time/s Time/s

Fig. 2 a Cooling curves of alloys containing single and combined Sr with Al–Ti–B grain refiners. The magnification of area related to b onset of
aluminium dendrite, c Al–Si eutectic, and d Al2Cu formation

123
1596 S. Farahany et al.

Fig. 3 Thermal analysis of 0.04 640 0.08 2

mass% Sr alloy representing the
cooling curves of central (TC) 0
thermocouple and the 1st derivative
600 0.06

Second derivative/°C2 s–2

corresponding first derivative 0

First derivative/°C s–1

and second derivative curves as TDCP TC

Temperature/°C
well as the wall thermocouple –0.3

ΔT=TW –TC
(TW), and difference 560 0.04
temperature between TW –2
thermocouples (DT = TW-TC)
520 0.02
ΔT –0.6

–4
480 0

tDCP 2nd derivative –0.9

440 –0.02 –6
200 400 600 800 1000
Time/s

coherency point (DCP). The first valley is defined as the 596 2.5
DCP, which indicates the transition from a liquid to a (a) TNAl
liquid–solid state. The dendrite coherency temperature 592 2
(TDCP) and time (tDCP), which are the temperature and time
Temperature/°C

588 TGAl
during solidification at which the dendrite network begins
1.5

ΔTR/°C
to establish in the solidifying melt, were recorded. After
DCP, dendrites continued to grow and became thicker with 584
further melt cooling. 1
580

Aluminium phase
576
ΔTRAl 0.5

The characteristic temperatures extracted from the obtained 572 0

cooling curves for the formation of Al dendrite are plotted Sr Sr+Al3TiB Sr+Al5TiB
in Fig. 4a. Clearly, the nucleation event for the Al dendrite
takes place at higher temperatures with the addition of 580
(b) 112
grain refiners to the Sr-modified alloy. TAl N was 15
580.7 ± 0.2 °C for the Sr-modified alloy, which shifted up TDCP
578 TN –TDCP 104
to 594.9 ± 0.5 and 596 ± 0.3 °C after the additions of Al–
3Ti–B and Al–5Ti–B. Similar findings were reported by 12 TN–TDCP/°C

tN – tDCP/s
TDCP/°C

other researchers for A356 [17] and 319 [18] when the melt 576 86
was treated with an Al–5Ti–B grain refiner. Additionally,
9
the growth temperature of the primary Al dendrite (TAl G)
88
increased from 575.5 ± 0.4 to 587.9 ± 0.5 and 574
586.9 ± 0.5 °C with the addition of 10 mass% Al–3Ti–B 6
and 6 mass% Al–5Ti–B, respectively, as reported for an tN –tDCP
80
Al–Si–Mg alloy [4]. As is evident, the nucleation temper- 572
Sr Sr+Al3TiB Sr+Al5TiB
ature increased by approximately 15 °C, while the growth
temperature increased by approximately 11.5 °C with Fig. 4 a Effect of 10 mass% Al–3Ti–B and 6 mass% Al–5Ti–B grain
increasing B content up to 6 mass%. The increasing refiners on the variations of a nucleation (TAl Al
N ), growth (TG ) and
Al
nucleation temperature allows new crystals to form ahead recalescence (DTR ) temperatures and b coherency temperature
(TDCP), interval temperature (TN–TDCP) and time (tN–tDCP) for
of solidification. Furthermore, the rate at which nucleation formation of aluminium dendrite
temperature increased is greater than the rise in growth
temperature. In other words, there are more nuclei with less Figure 4a also shows that the magnitude of recalescence
potential for growth, and thus effective grain refinement (DTAl Al Al
R = TG -TMin) for Al dendrites decreases by about
should be expected [17]. 2.3 °C after the addition of the Al–5Ti–B grain refiner. A

123
Cooling curve thermal analysis 1597

comparative calculation of DTAl R for the addition of 10 and Al–5Ti–B, respectively. In addition, tN–tDCP for the Sr-
mass% Al–3Ti–B and 6 mass% Al–5Ti–B grain refiners modified was around 79.8 s, which increased to 111.6 s and
shows that there is no significant change in the recales- 112.2 s with the addition of Al–3Ti–B and Al–5Ti–B,
cence magnitude. It is well established that the low mag- respectively, which corresponds to previous studies [19, 20].
nitude of the recalescence of Al dendrite (DTAl R) In other words, the coherency parameters reveal that grain
corresponds to a good refinement [4]. The presence of refinement increases the temperature (TN–TDCP) and time
recalescence indicates that the heat generated at the start of (tN–tDCP) interval of coherency and can postpone dendrite
solidification could not be transferred out of the mould coherency. This improves the fluidity of molten metal dur-
completely; this heat balance leads to the appearance of ing solidification. In short, the DCP has a direct relationship
recalescence [17]. with fluidity; if DCP is postponed, better fluidity will be
The grain size in a casting is associated with the number obtained.
of nucleation sites available in the melt at the liquidus
temperature or nucleation temperature of primary Al. Many Eutectic Al–Si phase
grains can be nucleated with little recalescence, and a fine-
grained structure will result if the number of sites is large To extensively study the effect of Al–Ti–B grain refiners,
[18]. Nafisi et al. [17] report that in the case of decreasing the characteristic temperature of the eutectic Al–Si phase in
recalescence due to grain refinement, particles can preserve the Sr-modified alloy was examined by CA-CCTA. The
the heat released from surrounding liquid, which leads to variations of nucleation temperature, undercooling, growth
lower recalescence compared to the unrefined alloy. They temperature, and recalescence value as a function of alloy
also suggest that reduced recalescence with grain refine- conditions for the Al–Si eutectic reaction are plotted in
ment may be attributed to the lower growth rate in the Fig. 5a. It can be seen that the nucleation temperature of
refined alloys, which in turn results in less heat being eutectic Al–Si reaction (TAl–Si
N ) increased from 563.6 ± 0.3
evolved to balance out the heat extracted from the solidi- to 565.1 ± 0.5 °C after the additions of Al–3Ti–B and Al–
fying alloy. 5Ti–B. In fact, an increase in the TAl–SiN is the same for
The influence of different concentrations of individual these two different grain refiners. This may be attributed to
Al–5Ti–B grain refiners on DCP parameters was reported the ability of B to facilitate the nucleation of eutectic grain
for an A319 alloy [19]. To assess the effect of the com- at higher temperatures. Moreover, the undercooling mag-
bination of a Sr modifier and Al–Ti–B grain refiners on Al nitude decreased slightly from 3.1 to 2.9 °C, with the
dendrite size, the concept of DCP was used according to addition of Al–3Ti–B. However, it remained constant at
the two thermocouple technique. As visible above in 3.1 °C with the addition of Al–5Ti–B. As can be seen, the
Fig. 3, the DCP occurs at the first maximum difference eutectic growth temperature (TAl–Si G ) for 0.04 mass% Sr,
between the wall and the centre thermocouples which induces fully modified Si, was 562.6 ± 0.5 °C. This
(DT = TW-TC) during the formation of primary Al. This temperature increased slightly to 563.9 ± 0.5 and
depression in temperature is attributed to enhanced thermal 563.2 ± 0.3 °C with the addition of Al–3Ti–B and Al–
contact through the solid dendrite network [20]. The ther- 5Ti–B grain refiners, respectively. With regard to the map
mal conductivity of solid Al is twice that of liquid Al [21]. established for the ADC12 alloy [23], these values indicate
After this point, both the TC and TW thermocouples record a modified Si structure, even after the addition of grain
a similar rate of cooling, indicating that the conduction refiners. The recalescence magnitude for the Sr-modified
from the centre and the wall has become similar [21] as alloy is 2.1 °C, which decreased to 1 and 1.2 °C after the
shown in Fig. 3. addition of Al–3Ti–B and Al–5Ti–B. It has been shown
The difference between nucleation temperature and that the depression of the eutectic growth temperature is a
dendrite coherency temperature (TN–TDCP) corresponds to largely acceptable criterion for the evaluation of eutectic
the free dendritic growth of primary Al, while after DCP it modification level [3, 10, 24–26]. The larger magnitude of
refers to dendrite thickening, resulting in a fixed dendritic depression is associated with a higher level of modification.
network [19]. DCP is important as it marks the transition It can be said, therefore, that all studied alloys have a
from mass feeding to inter-dendritic feeding in the solidi- modified eutectic Si structure.
fication process [22]. The effect of different Al–Ti–B grain With regard to obtained results, the nucleation temper-
refiners on coherency temperature, temperature, and time ature of Al dendrite (TAl N ) increased from 580.7 ± 0.2 to
interval of coherency (TN–TDCP) is plotted in Fig. 4b. As is 596 ± 0.3 and 594.9 ± 0.5 °C after the addition of 6
evident, TDCP increased from 574 ± 0.3 to 578.8 ± 0.4 °C mass% Al–5Ti–B and 10 mass% Al–3Ti–B grain refiners
with the addition of 6 mass% Al–%Ti–B. Moreover, TN– to the Sr-modified alloy, respectively. At the same time, the
TDCP increased significantly from 6.7 °C for the Sr-modified nucleation temperature of eutectic Al–Si (TAl–SiN ) increased
alloy, to 16.5 and 16.8 °C after the addition of Al–3Ti–B slightly from 563.6 ± 0.3 to 565.1 ± 0.5 °C for both grain

123
1598 S. Farahany et al.

(a) 566 10 4 grain refiner did not substantially influence the onset of the
TNAl-Si
Al2Cu phase. This more likely indicates that the exact
565 temperature for the onset of the Al2Cu phase depends
8 3 mainly on the amount of Cu in the molten metal, which is
Temperature/°C

564 TGAl-Si constant in all treated alloys in the present study. It has
been reported that the precipitation temperatures of Cu-

ΔTU/°C

ΔTR/°C
563 6 2 enriched phases decrease when Cu increases from 1 to 4
mass% in an AlSi5Cu system [27].
ΔTRAl-Si
562
4 1 Fraction solid curve
561 ΔTUAl-Si
Figure 6a–c presents the variation of fraction solids (fS) for
560 2 0 the Sr-modified alloy, and the alloys containing a combi-
Sr Sr+Al3TiB Sr+Al5TiB nation of Sr and two different grain refiners. When calcu-
lating the fraction solid, the zero line that indicates a no
(b) 500
phase formation must be plotted on the first derivative of
the cooling curve. It follows the first derivative in the
498 TNAl2Cu single phase area above the liquidus or start of solidifica-
tion (Ts), and below solidus or end of solidification (Te)
Temperature/°C

496 temperatures [10]. The Newtonian zero line has been

estimated by curve fitting the base line data between the
onset and the end of solidification in the first derivative
494
curve, using a fourth order polynomial curve as follows:
ðdT=dtÞbl ¼ A0 þ A1 X þ A2 X 2 þ A3 X 3 þ A4 X 4 ð1Þ
492
According to the calculated fraction solid curve, the fS
can be measured directly at a certain temperature or time.
490
Sr Sr+Al3TiB Sr+Al5TiB The computation of fraction solid at DCP showed that fDCPS
increases from 12 % for the Sr-modified alloy, to 15 and
Fig. 5 Effect of Al–3Ti–B and Al–5Ti–B grain refiner additions on
17 % after the additions of 10 mass% Al–3Ti–B and 6
the variations of a nucleation (TAl–Si
N ), growth (TAl–Si
G ), undercooling
(DTAl–Si
U ), and recalescence (DTAl–Si
R ) temperatures of Al–Si eutectic mass% Al–5Ti–B grain refiners. This implies mass feeding
reaction, and b nucleation temperature of Al2Cu reaction is extended after the addition of grain refiners. However,
the fraction solid of the Al–Si eutectic phase did not
changed markedly and remained constant around *74 %
refiners. In fact, in the combination of 0.04 mass% Sr with
for all treated alloys. Figure 6d shows the fraction solid of
6 mass% Al–5Ti–B and 10 mass% Al–3Ti–B, the solidi-
0.04 mass% Sr addition only and combined Sr with 10
fication range increased by 13.8 and 12.7 °C, respectively,
mass% Al–3Ti–B and 6 mass% Al–5Ti–B as a function of
which is in line with the findings of a previous study on an
solidification time. It can be seen that when the Al–3Ti–B
A356 alloy [4]. This shows that the widening of the
and Al–5Ti–B are added to the Sr-modified alloy, the
solidification range can offers benefits to semi-solid metal
fraction solid curves are shifted up slightly during the
processes; however, it does increase the incident porosity
formation of primary Al dendrite. This means that the
in conventional casting.
volume fraction of the solid is completed somewhat faster
when grain refiners are introduced into the melt, which is
Al2Cu phase consistent with the increase of fraction solids at DCP.
Subsequently, the fraction solid curves overlapped each
Figure 5b shows the effect of individual Sr and its com- other and followed similar trends.
bination with two different Al–Ti–B grain refiners on the In summary, the investigation of primary Al dendrite
nucleation temperature of Al2Cu phase (TAl2Cu
N ). It can be formation by way of CA-CCTA shows that the recales-
Al2Cu
seen that the TN is almost constant: 497.1 ± 0.6 °C for cence of Al dendrite decreased, while tDCP, TN–TDCP and
the ADC12 cast alloy treated with Sr, 497.3 ± 0.4 °C that fDCP
S increased with the addition of grain refiners. In
when treated with a combination of Sr and 10 mass% Al– other words, dendrites become coherent later with higher
3Ti–B, and 497.0 ± 0.5 °C when treated with a combina- solid fraction. No other indication of negative interaction
tion of Sr and 6 mass% Al–5Ti–B. Indeed, the addition of a between Sr and two grain refiners, namely Al–3Ti–B and

123
Cooling curve thermal analysis 1599

(a) (b)
580 0.08 100 0.06 100
Cooling
90 curve 90

Second derivative/°C2 s –2

Second derivative/°C2 s –2
560 80 570 80
0.04 0.03
70 70

Fraction solid/%

Fraction solid/%
Temperature/°C

Temperature/°C
Fraction solid
540 60 60
0 50 540 0 50

520 Second derivative 40 40

Second derivative
30 30
–0.04 510 –0.03
500 20 20
Cooling curve Fraction solid
10 10
–0.08 0 –0.06 0
200 400 600 800 200 400 600 800
Time/s Time/s
(c) (d)
0.06 100 100
Cooling
90 Sr
curve
Second derivative/°C2 s –2

80 Sr+Al3TiB
80
570 0.03 Sr+Al5TiB
70
Fraction solid/%

Fraction solid/%
Temperature/°C

60 60
540 0 50
40 40
Second derivative
30
–0.03
510 20 20
Fraction solid
10
–0.06 0 0
200 400 600 800 200 400 600 800
Time/s Time/s

Fig. 6 Cooling curve, second derivative curve, and associated fraction solid curve for a individual 0.04 mass% Sr, b combined Sr and 10 mass%
Al–3Ti–B, and c combined Sr and 6 mass% Al–5Ti–B grain refiner. d A comparison of calculated fraction solid curves

Al–5Ti–B, was observed, indicating that refined dendrite is addition of Sr to the ADC12 alloy promotes the columnar
predictable. growth of Al dendrites, which is consistent with the find-
ings of Liao et al. [28]. However, the microstructures of the
Microstructure analysis Sr-modified alloy after the addition of Al–3Ti–B and Al–
5Ti–B grain refiners reveal the columnar to equiaxed
Figure 7a, b illustrates the microstructure of the ADC12 dendrite transformation, as shown in Fig. 7c, e. The cal-
die-casting alloy modified with 0.04 mass% Sr. It can be culation of secondary dendrite arm size (SDAS) of primary
seen that the microstructure is composed of Al dendrite and Al showed that SDAS decreased from 69.6 ± 5.21 to
the Al–Si eutectic phase. Moreover, fibrous eutectic Si was 44.07 ± 8.67 and 40.96 ± 7.96 lm after the additions of
clearly observed in higher magnification (Fig. 7b), which is 10 mass% Al–3Ti–B and 6 mass% Al–5Ti–B to the Sr-
typical of a modified structure. The microstructures of Sr- modified alloys, respectively.
modified alloy after the addition of 10 mass% Al–3Ti–B to A comparison of the microstructures shown in Fig. 7b,
the master alloy are shown in Fig. 7c, d. It can be observed d, f shows that at the levels of Al–3Ti–B and Al–5Ti–B
that the eutectic Al–Si phase is distributed among Al additions, there are no major observable changes to the
dendrites. Figure 7e, f, meanwhile, shows the microstruc- morphology of Si particles. The fibrous Si appearance
ture of the Sr-modified alloy after the addition of 6 mass% associated with the modified condition is seen in the
Al–5Ti–B, which is similar to the microstructures pre- eutectic Al–Si area between dendrites for all treated alloys.
sented in Fig. 7c, d. It can be seen from Fig. 7a that the The results of thermal analysis were supported by

123
1600 S. Farahany et al.

Fig. 7 Optical micrographs in

low (left) and high (right)
magnifications showing the
microstructure of samples
treated with a, b individual 0.04
mass% Sr, c, d combined 0.04
mass% Sr and 10 mass% Al–
3Ti–B and e, f combined 0.04
mass% Sr and 6 mass% Al–
5Ti–B

microstructure observation, as shown in Fig. 7. The with 0.04 mass% Sr. The obtained thermal analysis results
microstructures show that the dendrite became finer and Si correlated well with observed microstructures. The recales-
structure remained modified, while recalescence (DTAl R ) for cence magnitude of the Al dendrite (DTAl R ) decreased, tem-
primary Al and growth temperature (TAl–SiG ) for the Al–Si perature interval of coherency (TN–TDCP) increased, and the
eutectic phase decreased. No obvious mutual effect was solid fraction at dendrite coherency (fDCP
S ) increased after the
found in terms of the coarsening of dendrite size (increase addition of Al–3Ti–B and Al–5Ti–B to the Sr-modified
of SDAS) or Si demodification. This can be related to the alloy. Nevertheless, the eutectic growth temperature (TAl–Si
G )
low amount of B addition in the present study. In other remained in the modified range. No sign of any interaction
words, with low levels of grain refiners, the effect of Sr as a between Al–5Ti–B and Al–3Ti–B and Sr was detected over
modifier is unhindered. But with high levels of additions, the range of the concentrations investigated in this study. A
Sr loses its modifying efficiency [7]. The formation of a well-refined grain and modified structure was obtained.
Sr–B compound (SrB6) consumes a large amount of Sr,
which reduces the amount of free Sr, and in turn reduces Acknowledgements The authors would like to thank Universiti
Teknologi Malaysia (UTM) for their financial support of this study
the modification efficiency of Sr. Previous studies have under the Vote No. Q.J130000.2524.08H36.
reported that a mutual poisoning effect occurs when the
content of Sr and B exceed a certain level [8, 9].

References
Conclusions
1. Knuutinen A, Nogita K, McDonald SD, Dahle AK. Modification of
Al–Si alloys with Ba, Ca, Y and Yb. Light Met. 2001;1:229–40.
CA-CCTA was used to investigate the influence of 6 mass% 2. Liu L, Samuel AM, Samuel FH, Doty HW, Valtierra S. Influence
Al–5Ti–B and 10 mass% Al–3Ti–B grain refiners on the of oxides on porosity formation in Sr-treated Al–Si casting alloys.
solidification behaviour of an ADC12 cast alloy modified J Mater Sci. 2003;38:1255–67.

123
Cooling curve thermal analysis 1601

3. Djurdjevic M, Jiang H, Sokolowski JH. On-line prediction of 16. Barlow JO, Stefanescu DM. Computer aided cooling curve ana-
aluminum–silicon eutectic modification level using thermal ana- lysis revisited. AFS Trans. 1997;97–04:349–54.
lysis. Mater Charact. 2001;46:31–8. 17. Nafisi S, Ghomashchi R. Grain refining of conventional and semi-
4. Nafisi S, Ghomashchi R. Combined grain refining and modifi- solid A356 Al–Si alloy. J Mater Process Technol. 2006;174:371–83.
cation of conventional and rheo-cast A356 Al–Si alloy. Mater 18. Shabestari SG, Malekan M. Assessment of the effect of grain
Charact. 2006;57:371–85. refinement on the solidification characteristics of 319 aluminum
5. Samuel AM, Doty HW, Valtierra S, Samuel FH. Effect of grain alloy using thermal analysis. J Alloy Compd. 2010;492:134–42.
refining and Sr-modification interactions on the impact toughness 19. Malekan M, Shabestari SG. Effect of grain refinement on the
of Al–Si–Mg cast alloys. Mater Des. 2014;56:264–73. dendrite coherency point during solidification of the A319 alu-
6. Prasada Rao AK. Das K, Murty BS, Chakraborty M. Micro- minium alloy. Metall Mater Trans A. 2009;40(13):3196–203.
structural features of as-cast A356 alloy inoculated with Sr, Sb 20. Arnberg L, Chai G, Bäckerud L. Determination of dendritic
modifiers and Al–Ti–C grain refiner simultaneously. Mater Lett. coherency in solidifying melts by rheological measurements.
2008;62:273–5. Mater Sci Eng A. 1993;173:101–3.
7. Lu L, Dahle AK. Effects of combined additions of Sr and AlTiB 21. MacKay RI, Djurdjevic MB, Sokołowski JH. Effect of cooling
grain refiners in hypoeutectic Al–Si foundry alloys. Mater Sci rate on fraction solid of metallurgical reactions in 319 alloy. AFS
Eng A. 2006;435–436:288–96. Trans. 2000;00–25:521–30.
8. Liao H, Sun GX. Mutual poisoning effect between Sr and B in 22. Veldman NLM, Dahle AK, StJohn DH, Arnberg L. Dendrite coher-
Al–Si casting alloys. Scr Mater. 2003;48:1035–9. ency of Al–Si–Cu alloys. Metall Mater Trans A. 2001;32:147–55.
9. Golbahar B, Samuel FH, Samuel AM, Doty HW, Valtierra S. Effect of 23. Farahany S, Ourdjini A, Abu Bakar TA, Idris MH. A new
grain refiner–modifier interaction on the microstructure and approach to assess the effects of Sr and Bi interaction in ADC12
mechanical properties of A356.2 alloys. AFS Trans. 2007;115:7–12. Al–Si die casting alloy. Thermochim Acta. 2014;575:179–87.
10. Emadi D, Whiting LV, Nafisi S, Ghomashchi R. Application of 24. Apelian D, Sigworth GK, Whaler KR. Assessment of grain
thermal analysis in quality control of solidification process. refinement and modification of Al–Si foundry alloys by thermal
J Therm Anal Calorim. 2005;81:235–42. analysis. AFS Trans. 1984;92:297–307.
11. Heusler L, Schneider W. Influence of alloying elements on the thermal 25. Chen X, Geng H, Li Y. Study on the eutectic modification level
analysis results of Al–Si cast alloys. Light Met. 2002;2:17–26. of Al–7Si Alloy by computer aided recognition of thermal ana-
12. Farahany S, Ourdjini A, Idris MH, Shabestari SG. Computer- lysis cooling curves. Mater Sci Eng A. 2006;419:283–9.
aided cooling curve thermal analysis of near eutectic Al–Si–Cu– 26. Bäckerud L, Chai G, Tamminen J. Foundry alloys. In: Solidifi-
Fe alloy—effect of silicon modifier/refiner and solidification cation characteristics of aluminum alloys, Vol 2. Stockholm,
conditions on the nucleation and growth of dendrites. J Therm Sweden: AFS/Skanaluminium; 1990.
Anal Calorim. 2013;114(2):705–17. 27. Djurdjevic MB, Odanovic Z, Talijan N. Characterization of the
13. Malekan M, Shabestari SG. Computer-aided cooling curve ther- solidification path of AlSi5Cu (1–4 wt%) alloys using cooling
mal analysis used to predict the quality of aluminum alloys. curve analysis. JOM. 2011;63(11):51–7.
J Therm Anal Calorim. 2011;103:453–8. 28. Liao H, Sun Y, Sun GX. Correlation between mechanical prop-
14. Mahfoud M, Emadi D, Prasada Rao AK. The role of thermal erties and amount of dendrite Al phase in as-cast near-eutectic
analysis in detecting impurity levels during aluminum recycling. Al-11.6 % Si alloys modified with strontium. Mater Sci Eng A.
J Therm Anal Calorim. 2010;100:847–51. 2002;335:62–6.
15. Farahany S, Ourdjini A, Idris MH. The usage of computer-aided
cooling curve thermal analysis to optimize eutectic refiner and mod-
ifier in Al–Si alloys. J Therm Anal Calorim. 2012;109(1):105–11.

123