Professional Documents
Culture Documents
a r t i c l e i n f o a b s t r a c t
Article history: The T6 heat treatment is commonly used for gravity cast Al–Si–Cu–Mg alloys. The influence of the alloy-
Received 14 February 2011 ing elements Cu and Mg and the artificial ageing temperature on the age hardening response were
Received in revised form 13 June 2011 investigated. Artificial ageing was conducted at 170 ◦ C and 210 ◦ C for various times for three alloys,
Accepted 14 June 2011
Al–7Si–0.3Mg, Al–8Si–3Cu and Al–8Si–3Cu–0.5Mg, cast with three different solidification rates (sec-
Available online 21 June 2011
ondary dendrite arm spacing of about 10, 25 and 50 m). The coarseness of the microstructure has a
small influence on the yield strength, as long as the solution treatment is adjusted to obtain complete
Keywords:
dissolution and homogenisation. The peak yield strength of the Al–Si–Mg alloy is not as sensitive to the
Mechanical characterization
Aluminium alloys
ageing temperature as the Al–Si–Cu and Al–Si–Cu–Mg alloys are. The ageing response of the Al–Si–Cu
Casting alloy is low and very slow. When 0.5 wt% Mg is added the ageing response increases drastically and a
Age hardening peak yield strength of 380 MPa is obtained after 20 h of ageing at 170 ◦ C for the finest microstructure,
but the elongation to fracture is decreased to 3%. The elongation to fracture decreases with ageing time
in the underaged condition as the yield strength increases for all three alloys. A recovery in elongation
to fracture of the Al–Si–Cu–Mg alloy on overageing is obtained for the finest microstructure, while the
elongation remains low for the coarser microstructures. The quality index, Q = YS + Kε, can be used to
compare the quality of different Al–Si–Mg alloys. This is not true for Al–Si–Cu–Mg alloys, as K depends
on the alloy composition. Overageing of the Al–Si–Mg alloy results in a decrease in quality compared to
the underaged condition.
© 2011 Elsevier B.V. All rights reserved.
1. Introduction as well as the peak hardness are unaffected [5]. A change of age-
ing temperature within the range 170–210 ◦ C does not influence
The T6 heat treatment is commonly used to increase the strength the peak yield strength strongly [1,7], while a decrease is obtained
of gravity cast Al–Si components containing Cu and/or Mg. The T6 for higher temperatures. There is a change in precipitate from the
heat treatment consists of solution treatment at a high tempera-  (Mg2 Si) phase to the  (Mg2 Si) phase in Al–Cu–Mg–Si alloys at
ture to dissolve large intermetallic particles and homogenize the around 200 ◦ C according to Eskin [8], which could be the reason for
alloying elements, quenching at a high cooling rate to retain a high lower peak yield strengths at high ageing temperatures.
concentration of vacancies and solute in solid solution and artifi- Few ageing curves exist in the literature for Al–Si–Cu alloys as
cial ageing at a lower temperature to form fine precipitates. The they are of less importance because of their lower age hardening
age hardening response is the increase in yield strength during response compared to Al–Si–Mg and Al–Si–Cu–Mg alloys. Ageing
artificial ageing relative the yield strength in the natural aged condi- curves from three articles [9–11] show good agreement, see Fig. 2.
tion. Alloy composition, artificial ageing temperature and thermal A long ageing time, around 100 h at 160 ◦ C, is needed to reach
history (solution treatment, quench, natural ageing) are the main peak hardness for Al–Si–Cu alloys. Some differences are seen in
parameters influencing the age hardening response. reported ageing curves [9–14], which could be because low con-
A large number of ageing curves are available in the literature centrations of Mg increase the age hardening response drastically
[1–7] for Al–Si–Mg alloys. Some examples for ageing at around and decrease the time to peak hardness [9]. When the ageing tem-
170 ◦ C are shown in Fig. 1. The ageing curves show two different perature is increased the peak condition is achieved faster, but the
behaviours. One set of curves shows a fast initial ageing response, peak hardness is decreased [9].
while the other set shows a much slower initial age hardening Precursors of the (Al2 Cu) phase form both in Al–Si–Cu casting
response. The slower age hardening response appears when an alloys and in Al–Cu wrought alloys at ageing, but the age harden-
alloy is natural aged prior to artificial ageing, but time to peak ing response per added wt% of Cu is lower in casting alloys [15].
The high dislocation density formed during quenching in Al–Si–Cu
alloys may be a reason for the low age hardening response [9,15].
∗ Corresponding author. Tel.: +46 36 101669; fax: +46 36 166560. The phase forms on dislocations, while the phase forms far
E-mail address: emma.sjolander@jth.hj.se (E. Sjölander). from the Si particles, where the dislocation density is low. The for-
0921-5093/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.msea.2011.06.036
E. Sjölander, S. Seifeddine / Materials Science and Engineering A 528 (2011) 7402–7409 7403
Table 1
Alloy composition in wt%.
Si Cu Mg Fe Ti Al Sr (ppm)
Table 2
Average SDAS for the three solidification rates; standard deviations within brackets.
various times. The times needed for heating the samples to the solu- spectroscopy (EDS). The concentrations of Mg and Si in the ␣-Al
tion treatment temperature (10–15 min) and the artificial ageing dendrites were measured using WDS on tensile test bars show-
temperature (20 min) are excluded from the recorded times. ing deviating yield strengths for the Al–Si–Mg alloy. Samples aged
The concentration of Cu, Mg and Si in the ␣-Al matrix after 20 min at 210 ◦ C were used for the concentration measurements.
solution treatment was measured using wavelength dispersive Eight points in the centre and eight points on the side of secondary
spectroscopy (WDS) to control that dissolution and homogeniza- dendrite arms were measured for each sample.
tion were completed. Four points in the centre and six points on
the side of secondary dendrite arms were measured for each sam- 3. Results and discussion
ple. The length of Si particles, Fe rich particles, Q particles and Al2 Cu
particles was measured after solution treatment for the three alloys 3.1. The Al–Si–Mg alloy
and the three coarsenesses of the microstructure. A fixed magnifi-
cation was used for each coarseness of the microstructure to be able The evolution of yield strength, UTS and elongation to fracture
to compare the different alloys. Ten frames were used and the three during ageing at 170 ◦ C and 210 ◦ C are presented in Fig. 3. The yield
longest particles of each type were measured in each frame. The strength curves for the Al–Si–Mg alloy aged at 170 ◦ C show good
number of measurements is small and is only used as an indication agreement with the ageing curves taken from the literature shown
of the relative size of the particles. in Fig. 1a. The strength is somewhat lower compared to the litera-
Tensile tests were performed at a constant strain rate of ture ageing curves as expected due to the lower Mg concentration
0.5 mm min−1 using a Zwick/Roell Z100 machine equipped with of the alloy. The ageing curves for 210 ◦ C show good agreement
a 100 kN load cell and a clip-on 25 mm gauge length extensome- with ageing curves reported by Tiryakioglu et al. [19] for sand cast
ter. Samples were tested until fracture, using two tensile test Al–Si–Mg alloys having 0.2 and 0.4 wt% Mg aged at 200 ◦ C. The
bars for each ageing condition. Both results are presented for the yield strengths for ageing at 170 ◦ C are about the same for the
yield strength, while UTS and elongation to fracture are for the three coarsenesses of the microstructure, see Fig. 3a. The differ-
sample having the highest UTS. Fracture surfaces were studied ence between the finest and the coarsest microstructure is 7 MPa
using scanning electron microscopy (SEM) and energy dispersive for the peak aged condition. This is in agreement with data reported
Fig. 3. Mechanical properties of the Al–Si–Mg alloy aged at 170 ◦ C (open symbols) and 210 ◦ C (solid symbols). a) Yield strength (results from two occasions are shown for the
finest structure aged at 210 ◦ C), b) UTS, c) elongation to fracture and d) correlation between yield strength and elongation to fracture. The lines are added to guide the eye.
E. Sjölander, S. Seifeddine / Materials Science and Engineering A 528 (2011) 7402–7409 7405
by Caceres et al. [20] for an Sr modified A356 alloy aged at 170 ◦ C Most of the samples having the finest microstructure reached
for 6 h having SDAS in the range 17–65 m. For ageing at 210 ◦ C the necking, while some fractured much earlier. Samples of the finest
yield strength shows some discrepancies with SDAS. Similar yield microstructure fracturing much before necking were excluded
strengths were obtained for the two coarser microstructures, while from the results presented in Fig. 4, as the aim of this investigation
the finest microstructure had lower strength and a large scatter in is to study the influence of the artificial ageing on the elongation to
the data (rhombus in Fig. 3a). Repeated heat treatments and tensile fracture and not the influence of defects. For the coarser microstruc-
tests were performed for ageing at 10 min, 20 min and 1 h at 210 ◦ C tures necking was not obtained and all results (except for natural
for the finest microstructure, resulting in higher yield strengths ageing of the intermediate microstructure which fractured much
(crosses in Fig. 3a). earlier than expected) are shown in Fig. 4. The elongation to frac-
Concentration measurements were performed on the four sam- ture decreases in the underaged condition as the yield strength
ples aged 20 min at 210 ◦ C having the finest microstructure to increases, see Fig. 4d. A tendency of an increase in elongation
investigate the reason for the variation in yield strength. All samples to fracture at overageing as the yield strength decreases can be
showed comparable Mg concentrations, 0.31 ± 0.02 wt% Mg, while observed. Ageing at 170 ◦ C gives a better combination of yield
the Si concentrations showed some variation. Test bars having strength and elongation to fracture than ageing at 210 ◦ C, for the
lower yield strength of 194 and 203 MPa had lower Si concentra- two finer microstructures. The scatter in elongation to fracture and
tions of 0.91 ± 0.10 wt% Si and 0.96 ± 0.10 wt% Si compared to test UTS indicates a large content of defects, which was confirmed by
bars having a higher yield strength of 231 MPa which had Si con- EDS measurements on the fracture surface. Defects having a much
centrations of 1.07 ± 0.03 wt% Si and 1.13 ± 0.04 wt% Si. It was also higher oxygen concentration than areas experiencing ductile frac-
noted that the Si concentration was especially low at the edge of ture were identified.
the dendrites close to the Si eutectic, 0.85 ± 0.05 wt% Si, for samples
having low yield strength. A too slow quench is therefore thought 3.3. The Al–Si–Cu–Mg alloy
to be the reason for the lower yield strength obtained for two of the
test bars. Due to lack of material tensile tests could not be repeated The evolution of yield strength and UTS during ageing at 170 ◦ C
for all ageing times and coarsenesses of the microstructure. and 210 ◦ C are presented in Fig. 5a and b. The time to peak
In Fig. 3c the elongation to fracture is seen to decrease with strength is a little longer for the Al–Si–Cu–Mg alloy compared to
ageing time as the yield strength increases. The test bars of the the Al–Si–Mg alloy, 20 h compared to 10 h at 170 ◦ C, indicating
intermediate microstructure showing low elongation at 170 ◦ C the presence of the Q phase in the Al–Si–Cu–Mg alloy. Natural
(ageing times of 20 h and 50 h) fractured outside the extensome- ageing was used in this investigation which may have influenced
ter, indicating the likely presence of defects. These samples did not the precipitation kinetics, favouring the formation of the Q phase
reach necking, which the rest of the samples with the interme- [17]. As for the Al–Si–Cu alloy, the coarsest microstructure shows
diate microstructure did. Excluding these results, the elongation lower yield strength than the two finer microstructures due to an
to fracture of the intermediate microstructure seems to follow the ineffective solution treatment. The Cu concentration in the matrix
same behaviour as the other two microstructures. However no rea- after solution treatment was 2.2 wt% for the coarsest microstruc-
son for the low elongation at 210 ◦ C (2 h) was found. It can be ture, while it was 3.3 wt% for the two finer microstructures. The Mg
noted that the elongation to fracture continues to decrease on over- and Si concentrations on the other hand were the same for the dif-
ageing when the yield strength decreases, see Fig. 3d. The ageing ferent coarsenesses of the microstructure, 0.24 wt% and 0.81 wt%,
times of the present investigation are too short to observe a clear respectively. Far from all Mg is dissolved into the ␣-Al matrix, due
increase in elongation to fracture at overageing. Tiryakioglu et al. to the presence of the Q-Al5 Mg8 Cu2 Si6 phase, which does not dis-
[19] has however earlier observed a clear increase in elongation solve during solution treatment at 495 ◦ C. The peak yield strength
to fracture for longer ageing times. The total elongation to frac- is on average 25 MPa lower for ageing at 210 ◦ C compared to 170 ◦ C.
ture is lower at overageing than at underageing and the uniform A decrease in peak yield strength for Al–Si–Cu–Mg alloys with
elongation is observed to be a lower fraction of the total elonga- increasing ageing temperature has earlier been observed by others
tion. This behaviour has earlier been observed by Alexopoulos and [9,13,23–25]. A higher fraction of Q phases at the higher ageing
Tiryakioglu [21] for an Al–7Si–0.55Mg alloy. temperature [8,17] is a possible explanation for the reduction in
peak yield strength with increasing ageing temperature.
The elongation to fracture reaches a minimum when the yield
3.2. The Al–Si–Cu alloy strength is at its maximum, see Fig. 5d. For the finest microstruc-
ture the elongation is seen to increase upon overageing. An increase
The ageing curves of the Al–Si–Cu alloy are shown in Fig. 4a and in elongation on overageing for 319-type alloys has earlier been
b. The ageing curves show the same behaviour for the three coarse- reported in the literature [14,23,24]. The elongation to fracture for
nesses of the microstructure, but the yield strength of the coarsest the three coarsenesses of the microstructure of the Al–Si–Cu–Mg
microstructure is lower. Ineffective solution treatment might be alloy have the same trend for ageing times shorter than 10 h and
the reason for the lower yield strength of the coarsest microstruc- longer than 100 h for ageing at 170 ◦ C, see Fig. 5c. The two coarser
ture. The solution treatment was 6 h at 495 ◦ C for the coarsest microstructures have a low elongation to fracture for ageing times
microstructure as this was thought to be enough to achieve com- in between. These ageing times corresponds to high yield strengths.
plete dissolution and homogenisation. However a later study [22] The low elongation to fracture of the two coarser microstructures
has shown that more than 10 h at 495 ◦ C is needed. After 6 h of is probably caused by a combination of the high yield strength and
solution treatment at 495 ◦ C the Cu concentration in the matrix is larger brittle phases and defects present in the coarser microstruc-
about 2.4 wt% for the coarsest microstructure, compared to 3.2 wt% tures compared to in the finest microstructure.
for the two finer microstructures after 1 h and 3 h respectively. A
lower concentration of Cu in solid solution after solution treatment 3.4. Age hardening response
results in a lower yield strength after artificial ageing, as can be seen
in Fig. 4a–b. A long time is needed to reach peak strength, about The time to peak tpeak , the age hardening response (the differ-
12 h at 210 ◦ C and 170 h or more at 170 ◦ C. The peak yield strength ence in yield strength between the peak aged condition and the
is influenced by the ageing temperature, being about 31 MPa lower natural aged condition) YS and the yield strength and elonga-
at 210 ◦ C compared to 170 ◦ C. tion to fracture at the peak aged condition for the intermediate
7406 E. Sjölander, S. Seifeddine / Materials Science and Engineering A 528 (2011) 7402–7409
Fig. 4. Mechanical properties of the Al–Si–Cu alloy. Yield strength (open symbols) and UTS (solid symbols) at a) 170 ◦ C and b) 210 ◦ C. c) Elongation to fracture and d)
correlation between yield strength and elongation to fracture at 170 ◦ C and 210 ◦ C. The lines are added to guide the eye.
microstructure, are summarized in Table 3. A low age hardening 3.5. Elongation to fracture
response was observed for the Al–Si–Cu alloy, 76 MPa at 170 ◦ C
and 44 MPa at 210 ◦ C. An addition of 0.5 wt% Mg changes the The elongation to fracture of the two finest microstructures of
age hardening response completely. The time to peak strength the Al–Si–Mg alloy are comparable, while it is significantly lower for
is decreased from 170 h to 20 h at 170 ◦ C and the age hardening the coarsest microstructure, see Fig. 3c. This observation is in agree-
response is increased to 168 MPa at 170 ◦ C and 138 MPa at 210 ◦ C, ment with results presented by Caceres et al. [20] for a Sr modified
see Table 3. The yield strength of the Al–Si–Cu–Mg alloy is more Al–7.5Si–0.4Mg alloy. The elongation to fracture depends strongly
sensitive to the ageing temperature compared to the Al–Si–Mg on the ability of Si particles in the Al–Si eutectic to stop the motion
alloy and the difference in age hardening response between the of dislocations according to Caceres et al. [20]. The Si particles are
two alloys is observed to be smaller for ageing at 210 ◦ C compared more homogeneous distributed in a fine microstructure after solu-
to 170 ◦ C. tion treatment and the dislocations interact with the Si particles
Table 3
Age hardening response and peak properties for the three alloys.
Fig. 5. Mechanical properties of the Al–Si–Cu–Mg alloy aged at 170 ◦ C (open symbols) and 210 ◦ C (solid symbols). a) Yield strength, b) UTS, c) elongation to fracture and d)
correlation between yield strength and elongation to fracture. The lines are added to guide the eye.
individually. Subsequently the dislocations pass the Si particles and during solidification, transforms into short -Fe phases and Mg in
pile up at grain boundaries where the fracture takes place, result- solid solution during solution treatment [26,27], while the length of
ing in a high elongation to fracture [20]. In a coarser microstructure the Fe-rich phases in the Cu containing alloys is unchanged during
the Si particles form a more continuous barrier for the dislocations solution treatment.
which pile up at the Al–Si eutectic, where the fracture takes place, To summarize, when Fe-rich phases are of comparable size
resulting in lower elongation to fracture [20]. The situation for the as the Si particles the theory for the Al–Si–Mg alloy derived by
Cu containing alloys is however different. The finest microstruc- Caceres et al. seems to be valid with a high elongation to fracture
ture has a significantly higher elongation to fracture compared to independent of the coarseness of the microstructure for fine
the coarser ones, see Figs. 4c and 5c. microstructures having SDAS below about 35 m [20]. However,
That is, the intermediate microstructure has an elongation to when longer Fe-rich phases are present, as in the Cu containing
fracture comparable to the finest microstructure for the Al–Si–Mg alloys, the elongation to fracture is more sensitive to the coarseness
alloy, while it is comparable to the coarsest microstructure for the of the microstructure also for fine microstructures.
Cu containing alloys. The size and distribution of the Si particles The much lower elongation to fracture of the Al–Si–Cu–Mg alloy
are similar for the different alloys and cannot be the cause of the compared to the Al–Si–Cu alloy is probably due to the higher yield
difference in behaviour and the cause is instead likely due to the strength of the Al–Si–Cu–Mg alloy. A strong matrix transfers a high
length of Fe-rich phases after solution treatment. The Cu contain- load on brittle phases already at small strains, which makes the
ing alloys have longer Fe-rich phases, 20 ± 10 m, compared to the brittle phases crack at low strains, resulting in a low elongation
Al–Si–Mg alloy, 7 ± 2 m, for the intermediate microstructure. The to fracture [28]. The Q phase remains stable after solution treat-
length of the Fe-rich phases of the Cu containing alloys is thereby ment for the Al–Si–Cu–Mg alloy. It is however smaller than the
about twice the length of the Si particles. One reason for the dif- Fe-rich phases and will probably not have a large influence on the
ference is that the -Fe phase, which forms in the Al–Si–Mg alloy elongation to fracture.
7408 E. Sjölander, S. Seifeddine / Materials Science and Engineering A 528 (2011) 7402–7409
data obtained from underaged and peak aged conditions for data
from the literature [1,19,31] and from the present investigation. The
slope has a value of around 10–13 for most data sets, but values up
to 27 were obtained.
For Al–Si–Cu–Mg alloys a decrease in quality with increasing
ageing temperature is observed, which is in agreement with data
reported in the literature [14,24,25,32]. A change in precipitates
could be the reason for the decrease in properties. The slope, K,
of the yield strength – elongation curves was calculated and val-
ues ranging from 19 to 60 were obtained. K is alloy dependent
and decreases with increasing Mg concentration and decreasing Cu
concentration. The Q index can therefore not be used to compare
different Al–Si–Cu–Mg alloys.
An increase in yield strength generally leads to a decrease in
ductility for a material that exhibits necking [33]. The expected
behaviour is observed for the Al–Si–Cu–Mg alloy for the finest
microstructure, where the elongation increases directly after the
peak aged condition when the yield strength decreases, see Fig. 6b.
For the Al–Si–Mg alloy the behaviour is different and the elon-
gation remains low at initial overageing. The formation of small
Si particles in the matrix at initial overageing [2] can be spec-
ulated to be the cause of the lower ductility of the overaged
condition compared to the underaged condition for the Al–Si–Mg
alloy.
The Al–Si–Mg alloy in the underaged condition has a higher
quality than the Al–Si–Cu–Mg alloy, see Fig. 6b. Al–Si–Mg alloys are
therefore preferable to Al–Si–Cu–Mg alloys, as long as the Al–Si–Mg
alloys can reach the required strength. The maximum yield strength
that can be reached for Al–Si–Mg alloys is around 300–320 MPa
[1,31]. The difference between the alloys however decreases with
increasing yield strength.
4. Conclusions
[2] Zhang, D.L., D. H. StJohn, (1995). “The effect of composition and heat treatment [18] S. Seifeddine, PhD, Department of Mechanical Engineering /Component Tech-
on the microstructure and mechanical properties of cast Al-7wt%Si-Mg alloys”. nology - Castings, Jönköping University, Jönköping, Sweden, 2006.
Casting and Solidification of Light Alloys, IMMA, 30-31 Aug 1995, Gold Coast, [19] M. Tiryakioglu, J. Campbell, J.T. Staley, Mater. Sci. Eng. A 368 (2004) 205–211.
pp. 21-25. [20] C.H. Caceres, C.J. Davidson, J.R. Griffiths, Q.G. Wang, Metall. Mater. Trans. A 30
[3] Y.J. Li, S. Brusethaug, A. Olsen, Scripta Mater. 54 (2006) 99–103. (1999) 2611–2618.
[4] G. Wang, Q. Sun, L. Feng, L. Hui, C. Jing, Mater. Des. 28 (2007) 1001–1005. [21] N.D. Alexopoulos, M. Tiryakioglu, Mater. Sci. Eng. A 507 (2009)
[5] H. Möller, G. Govender, W.E. Stumpf, Int. J. Cast Met. Res. 20 (2007) 340–346. 236–240.
[6] D.L. Zhang, Mater. Sci. Forum 217–222 (1996) 771–776. [22] E. Sjölander, S. Seifeddine, Mater. Des. 31 (2010) S44–S49.
[7] P.A. Rometsch, G.B. Schaffer, Mater. Sci. Eng. A 325 (2002) 424–434. [23] E. Cerri, E. Evangelista, S. Spigarelli, P. Cavaliere, F. DeRiccardis, Mater. Sci. Eng.
[8] D.G. Eskin, J. Mater. Sci. 38 (2003) 279–290. A 284 (2000) 254–260.
[9] H.G. Kang, M. Kida, H. Miyahara, K. Ogi, AFS Trans. 27 (1999) 507–515. [24] C.A. Cloutier, J.W.J.E. Jones, Allison in: SAE 2000 World Congress, Detroit, MI,
[10] W. Reif, J. Dutkiewicz, R. Ciach, S. Yu, J. Krol, Mater. Sci. Eng. A 234–236 (1997) USA, 2000, pp. 327–338.
165–168. [25] G. Geier, H. Rockenschaub, T. Pabel, M. Hopfinger, Giessereiforschung 58 (2006)
[11] G. Wang, X. Bian, X. Liu, J. Zhang, J. Mater. Sci. 39 (2004) 2535–2537. 32–48.
[12] P. Ouellet, F.H. Samuel, J. Mater. Sci. 34 (1999) 4671–4697. [26] A.L. Dons, L. Pedersen, S. Brusethaug, Aluminium 76 (2000) 294–297.
[13] R.X. Li, R.D. Li, Y.H. Zhao, L.Z. He, C.X. Li, H.R. Guan, Z.Q. Hu, Mater. Lett. 58 (2004) [27] J.A. Taylor, D.H. St John, J. Barresi, M.J. Couper, Mater. Sci. Forum 331–337 (2000)
2096–2101. 277–282.
[14] F.J. Tavitas-Medrano, J.E. Gruzleski, F.H. Samuel, S. Valtierra, H.W. Doty, Mater. [28] C.H. Cáceres, I.L. Svensson, J.A. Taylor, Int. J. Cast Met. Res. 15 (2003)
Sci. Eng. A 480 (2008) 356–364. 531–543.
[15] G. Wang, L. Yan, G. Ren, Z. Zhao, Adv. Mater. Res. 146–147 (2011) [29] T. Din, A.K.M.B. Rashid, J. Campbell, Mater. Sci. Technol. 12 (1996) 269–273.
1667–1670. [30] M. Drouzy, S. Jacob, M. Richard, Int. Cast Met. J. 5 (1980) 43–50.
[16] X. Wang, S. Esmaeili, D.J. Lloyd, Metall. Mater. Trans. A 37 (2006) 2691–2699. [31] S. Shivkumar, C. Keller, D. Apelian, AFS Trans. 98 (1990) 905–911.
[17] X. Wang, W.J. Poole, S. Esmaeili, D.J. Lloyd, J.D. Embury, Metall. Mater. Trans. A [32] J. Gauthier, P.R. Louchez, F.H. Samuel, Cast Met. 8 (1995) 107–114.
34 (2003) 2913–2924. [33] C.H. Cáceres, Int. J. Cast Met. Res. 10 (1998) 293–299.