Materials Transactions, Vol. 43, No. 11 (2002) pp. 2789 to 2795
c
2002
The Japan Institute of Light Metals
Effects of Cu and Transition Metals on the Precipitation Behaviors of
Metastable Phases at 523 K in Al–Mg–Si Alloys
Kenji Matsuda1 , Shohei Taniguchi1, ∗ , Kosuke Kido1,∗ , Yasuhiro Uetani2 and Susumu Ikeno1
1
2
Faculty of Engineering, Toyama University, Toyama 930-8555, Japan
Toyama Prefectural University, Kosugi, Toyama 939-0398, Japan
Morphology and crystal lattice of precipitates formed in the 6000 series Al–Mg–Si alloys containing Cu and transition metals of Cr
and Fe (TM) after aging at 523 K were investigated by high-resolution transmission microscopy (HRTEM). The precipitates in the Cu-bearing
alloys are more finely and densely distributed than those in the Cu-free alloys, whereas those in the TM-bearing alloys were coarser and formed
inhomogeneously at the interfaces between the dispersoids and the matrix. Three kinds of metastable phases were detected by HRTEM as
follows: β -phase in the balanced alloy, the TYPE-B precipitate in the excess-Si alloy, and Q -phase in the Cu-bearing alloys. No significant
difference in between the precipitates in the TM-free alloys and in the TM-bearing alloys was observed. However, the Type-B precipitate in
the excess-Si alloy was replaced with the β -phase by the addition of TM. This suggests that the chemical composition of the excess-Si alloy
changes to the balanced composition, because the excess Si in the matrix has been consumed by the formation of the AlSi(Fe, Cr) and AlSiFe
dispersoids.
(Received March 13, 2002; Accepted August 26, 2002)
Keywords: aluminum–magnesium–silicon alloys, copper, transition metals, precipitation, high-resolution transmission electron microscopy
1. Introduction
The addition of copper to the Al–Mg–Si alloys results in
the formation of the Q-phase,1) or the metastable Q -phase
as recently reported.2, 3) However, these reports of the Q or
Q-phase formation are mostly associated with the addition
of Si to 2000 series Al–Cu–Mg alloys, and there are relatively few reports about the addition of small amounts of
Cu to the Al–Mg–Si alloys.4) It is known that the addition
of Cr to the Al–Mg–Si alloys results in the formation of the
AlCrSi or AlFeCrSi dispersoids, and also in the decrease in
age-hardenability.5, 6) There are no reports that precipitates in
the Al–Mg–Si alloys including Cu and Cr were observed by
high-resolution transmission electron microscopy (HRTEM),
and were classified into kinds of precipitates based on characteristics of their HRTEM images. We have reported the
precipitation sequence of Al–Mg–Si alloys without Cu and
Cr by HRTEM,7, 8) and recently reported the Q -phase in the
Al–Mg–Si–Cu alloy.9) We have also reported on the effect
of transition metals of Cr and Fe (TM) and Cu on the agehardening behavior of the Al–1.1 mass%Mg2 Si–X mass%Si
alloys (X = 0, 0.34) aged at 423 K.10) The dispersoids
of AlSi(Cr, Fe) and AlSiFe were found to exist in the TMbearing alloys, and it was expected that the formation of these
dispersoids causes the decrease of Si contents in the matrix.
Figure 1 shows changes in hardness of alloys as-quenched
and of alloys peak-hardened at 423 K with Mg2 Si contents.
Hardness of both as-quenched and peak-aged alloys increase
with Mg2 Si content. The as-quenched TM-bearing alloys (●,
▲, ■) are harder than the TM-free alloys (, , ). This is
attributed to grain refinement in the TM-bearing alloys and
solid-solution hardening of the matrix by Cr.11) The mean
grain size of the TM-bearing alloys is smaller than 60 µm,
which is 1/3 that of the TM-free alloys. On the contrary, hard∗ Graduate
Student, Toyama University.
ness of the peak hardened TM-bearing alloys is clearly lower
than that of TM-free alloys. This phenomenon was found
in our most recent study,10) and has been also explained in
the previous studies12–15) by the formation of dispersoids and
by the increase in inhomogeneous precipitation at the interface between dispersoids and the matrix. The formation of
the dispersoids reduces the amount of solute Si in the matrix,
and the interface between dispersoids and the matrix acts as a
sink for super-saturated quenched-in vacancies. Finally, the
inhomogeneous precipitation is caused by this mechanism.
The precipitates in alloys aged at 423 K were observed by
HRTEM, however no remarkable difference were observed
in that study.
The precipitation behavior of Al–Mg–Si alloys at 523 K
aging has been reported in our studies in detail.8, 9) 523 K
is a suitable temperature to classify each metastable phase,
because precipitates are large, have the periodic structures
in contrast to precipitates in alloys aged below 473 K, and
are shown in clear HRTEM images. Several new metastable
phases have been found out by HRTEM.8, 9)
In this study, metastable phases in Al–Mg–Si alloys at
523 K containing Cu and TM are observed by HRTEM, and
the kinds of precipitates are classified in order to investigate
the effects of the addition of these minor elements and examine the validity of the precipitation mechanisms previously
suggested by the authors.
2. Experimental
Eight alloys were chosen from the 16 alloys provided by
the Research Group for Properties Near Grain Boundaries of
the Japan Institute of Light Metals. The chemical compositions of the 8 alloys were shown in Table 1. This table also
shows the mean grain sizes of alloys after solution heat treatment. Types of alloys were also described in this table. The
2790
K. Matsuda, S. Taniguchi, K. Kido, Y. Uetani and S. Ikeno
50
No.1
No.2
Micro vickers hardness, H / HV
40
20
No.8
60
d)Cu-bearing (peak hardness)
120
30
No.C6
No.C7
No.C8
No.C1
No.C2
No.C3
90
150
b)Cu-bearing (as-quenched)
No.6
No.7
No.1
No.2
No.3
120
No.C1
No.C2
No.C3
40
c)Cu-free (peak hardness)
No.3
30
50
150
a)Cu-free (as-quenched)
No.6
No.7
No.8
90
No.C6
60
20
0.6 0.8 1 1.2 1.4 1.6 1.8 0.6
0.9
1.2
No.C7
No.C8
1.5
Mg2Si (mass%)
Fig. 1 Changes in micro Vickers hardness of alloys containing Mg2 Si for 423 K aging. (a) and (b) as-quenched Cu-free and Cu-bearing
alloys, and (c) and (d) peak-hardened Cu-free and Cu-bearing alloys.
Table 1 Chemical compositions (mass%) and mean grain sizes of the samples.
Samples
Types
Si
Fe
Cu
Mg
Mn
Zn
Cr
Mg2 Si
Excess Si
Grain size
No. 2
No. 4
No. 7
No. 9
No. C2
No. C4
No. C7
No. C9
Balanced
ex.Si
Balanced + Cu
ex.Si + Cu
Cr + balanced
Cr + ex.Si
Cr + balanced + Cu
Cr + ex.Si + Cu
0.44
0.76
0.45
0.76
0.45
0.76
0.45
0.76
0.02
0.03
0.03
0.030
0.20
0.20
0.20
0.200
0.01
0.01
0.33
0.34
0.00
0.01
0.34
0.34
0.70
0.71
0.68
0.70
0.72
0.73
0.73
0.72
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0
0
0
0
0.21
0.21
0.21
0.21
1.10
1.12
1.07
1.10
1.14
1.15
1.15
1.14
0.04
0.35
0.06
0.36
0.03
0.34
0.03
0.34
139 µm
139 µm
168 µm
210 µm
54 µm
31 µm
66 µm
45 µm
ingots of all alloys were prepared by DC casting and homogenization at 813 K for 28.8 ks. The homogenized ingots were
then hot- and cold-rolled into 1 mm (for hardness tests) and
0.2 mm thick (for TEM observation) sheets. The sheets were
solution heated at 813 K for 3.6 ks and quenched in water at
room temperature. 423 K aging was performed for quenched
sheets in an oil bath for the time achieving peak hardness
of each alloy, and Micro-Vickers hardness was measured by
Akashi MVK-E II hardness tester (load: 0.98 N, holding time:
15 s). Samples for TEM were aged at 523 K for 12 ks in a salt
bath. Thin foils for TEM were prepared by electrolytic polishing using a solution of perchloric acid and ethanol. TEM
used was Topcon EM-002B, equipped with energy dispersive
X-ray spectroscopy (EDS).
3. Results and Discussion
3.1 TEM observation
3.1.1 Matrix precipitates
An upper column of Fig. 2 shows bright field images of
TM-free alloys. The number density of precipitates increases
in the turn of alloy 2 (balanced), alloy 4 (ex.Si), alloy 7
(balanced + Cu) and alloy 9 (ex.Si + Cu). Remarkable difference of microstructures can be seen in TM-bearing alloys
of the lower column in Fig. 2. Coarse precipitates can be
observed at the interface between the dispersoids and the matrix. This tendency was common to all the TM-bearing alloys. The number density of precipitates per unit area (N )
measured from these TEM images is shown in Fig. 3. The
value N in ex.Si alloy 4 is higher than that of balanced alloy
2. The same tendency can be seen for Cu-bearing alloys 7
and 9 in Fig. 3(b). TM-bearing alloys C2, C4, C7 and C9 also
indicated the same behavior as TM-free alloys. Cu-bearing
alloys, particularly alloy C7, have a remarkably low N . It
means that the solute atoms, which are consumed to form dispersoids and contribute to the formation of precipitates in Cubearing alloys compared with those in Cu-free alloys.
3.1.2 Dispersoids
α-Al(CrFe)Si (bcc/sc, a = 1.25–1.27 nm) and α -AlCrSi
(Al13 Cr4 Si4 , fcc, a = 1.09 nm) in aluminum alloys have
been reported5, 16) as the dispersoids including Cr in Al alloys.
Figure 4 shows TEM images of a dispersoid in alloy C9. Al,
Effects of Cu and Transition Metals on the Precipitation Behaviors of Metastable Phases at 523 K in Al–Mg–Si Alloys
2791
Fig. 2 TEM images of alloys aged at 523 K for 12 ks. (a) Alloy 2, (b) alloy 4, (c) alloy 7, (d) alloy 9, (e) alloy C2, (f) alloy C4, (g) alloy
C7 and (h) alloy C9.
Fe, Cr and Si peaks were detected by EDS analysis. The mean
ratio of Al : (Cr + Fe) : Si is about 9.5 : 1 : 1. According to
the results of EDS and electron diffraction patterns, this dispersoid is probably α-Al(CrFe)Si.5) A large dispersoid was
also seen (Fig. 5). It was identified as α-Al8 Fe2 Si16) based on
the strong Fe, Al and Si peaks in the EDS profile and selected
area diffraction pattern (SADP) analysis. It is considered that
the decrease of Si content in the matrix by the formation of
two kinds of dispersoids10) causes the remarkable decrease in
the number density of precipitates in TM-bearing alloy, especially alloy C7.
3.2 HRTEM observation
3.2.1 TM-free alloys
According to the equilibrium phase diagram of Al–Mg–
Si alloys, Mg–Si (β) precipitate forms in Cu-free alloys,
and Mg–Si (β) and AlMgSiCu (Q) precipitates form in Cubearing alloys. As the precipitates in both alloys incorporate
Si, it follows that the number density of precipitates will decrease through the consumption of Si by the formation of dispersoids during casting or homogenizing. This reduction in
the abundance of precipitates will be more prominent in Cubearing alloys, because there are two kinds of the Si-bearing
metastable phases. In our recent studies, the Si-content affects
kinds of precipitates formed in Al–Mg–Si alloys.8, 9) Figure 6
shows HRTEM images of alloys. There is a cross-section of
the rod-shaped precipitate perpendicular to its longitudinal direction in {200} lattice fringes of 0.203 nm shown in each image. These are referred as “the rod-section”, in the present
study. The precipitation sequence of the balanced type alloy
is as follows:9)
GP zones → transition phases [random-, parallelogram-types] → β -phase → β-phase
The precipitation sequence for the excess-Si and Cu-bearing alloys differ from that of the balanced alloy as follows:8)
[excess Si]
GP zones → transition phases [random-, parallelogram-types, β -phase∗ ]
→metastable phases [β -phase, Type-A∗ , Type-B∗ , Type-C∗ ] → β-phase, Si-phase
∗ β -phase, Type-A, Type-B and Type-C precipitates have higher Si contents than β -phase.
[Cu-bearing alloy]
GP zones → transition phases [random-, parallelogram-types] → β -phase, Q -phase → β-phase, (Q-phase).
The β -phase, which is a typical phase in the balanced alloy, was detected in the alloy 2 as a hexagonal network of
0.71 nm as shown in Fig. 6(a). Figure 6(b) shows the Type-
B precipitate in alloy 4, which is one of typical precipitate
in the excess-Si alloys. The Type-B precipitate appears as a
rectangular network of 0.68 nm and 0.79 nm. In our previous
K. Matsuda, S. Taniguchi, K. Kido, Y. Uetani and S. Ikeno
150
(a)
Number of precipitates,N /
m
2
2792
(b)
100
TM-free
No.9
No.4
No.7
TM-free
50
No.2
TM-bearing
No.C9
No.C4
TM-bearing
No.C7
No.C2
0
0
0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4
excess Si(mass%)
Fig. 3 Changes in number of precipitates per unit area in each alloy with
excess Si contents. The aging treatment was at 523 K for 12 ks. (a) Cu-free
and (b) Cu-bearing alloys.
Fig. 5 TEM images of a large dispersoids in alloy C9. (a) Bright-field
image, (b) and (c) EDS profile and a selected area diffraction pattern of
the dispersoid marked by an arrow, and (d) a schematic illustration of (c).
Incident beam direction in (c) is parallel to [243̄] of AlFeSi. The aging
treatment was at 523 K for 12 ks.
Fig. 4 TEM images of dispersoids in alloy C9. (a) Bright-field image, (b)
and (c) EDS profile and a selected area diffraction pattern of the dispersoid
marked by an arrow, and (d) a schematic illustration of (c). Incident beam
direction in (c) is parallel to [2̄01] of Al(Cr, Fe)Si. The aging treatment
was at 523 K for 12 ks.
study,8) the population of several kinds of metastable phases
in the Al–1.0 mass%Mg2 Si–0.4 mass%Si alloy aged at 523 K
for 12 ks consisted of 60% Type-B precipitate, 30% Type-A
precipitate, and 10% β -phase. A small amount of the TypeA precipitate was also observed in the alloy 4 as shown in
Fig. 7. The Type-A precipitates are also detected at grain
boundaries as shown in Fig. 8. This is in good agreement
with our previous study,17) because the grain boundary precipitates in Al–1.0 mass%Mg2 Si–0.4 mass%Si alloy aged at below 523 K were mainly the Type-A precipitates. These results
support that the alloy 4 has the same precipitation sequence as
that of the typical excess Si alloy we studied before.8, 17)
Precipitates with elongated cross-sections were observed in
alloys 7 and 9. Figure 9 shows a typical bright field of the alloy 9. Many elongated cross-sections can be seen in Fig. 9(b).
A HRTEM image of the precipitate with a hexagonal network
of 1.04 nm is shown in Fig. 9(c). The direction marked C in
this figure makes an angle of 10 degrees, which is in good
agreement with the Type-C and the Q phase.9) The TypeC precipitate is a Si–Mg–Al ternary metastable phase, and
the Q -phase is a quaternary metastable phases in Cu-bearing
alloys.
3.2.2 TM-bearing alloys
EDS profiles of elongated cross-section in alloys C2, C4,
C7 and C9 are shown in Fig. 10. Al, Cu, Mg and Si peaks
were detected in alloys C7 and C9 of Figs. 10(a) and (b). The
Cu peak was not present for alloys C2 and C4. It is known
that the chemical composition of the Q phase is the same as
that of the Q-phase; Mg : Al : Si : Cu = 8 : 5 : 6 : 2.1, 3)
Apparently, a stronger Al peak than expected is detected in
the present study, because of the effect of the Al matrix. The
ratio of Mg : Si : Cu is 4 : 2 : 1, which is close to 4 : 3 : 1
for the metastable Q - or equilibrium Q-phases. Thus, this
precipitate having elongated cross-section is classified as the
Q -phase reffering to our previous study.9) The precipitates
in the TM-bearing alloy C4 differ remarkably from those in
the TM-free alloy 4. In the C4 alloy, the β -phase is present,
which is the same phase as observed in alloys 2 and C2. In
contrast, the Type-B precipitate is observed in the alloy 4.
The Mg/Si ratio of the precipitate in the alloy C4 is about 2.0,
which is characteristic of the β -phase. This means that the
excess Si contents in the alloy C4 decreases significantly. It
can be concluded that excess Si in TM-bearing alloys is consumed to form dispersoids, Si content in the matrix decreases,
and then the precipitation sequence of the TM-bearing alloy
changes from that typical one of the excess Si alloy to that of
the balanced alloy.
3.3 Effect of Cu addition to excess Mg alloy
Addition of TM decreases Si contents, therefore it causes
the decrement of Mg2 Si contents, number density of precipitates, and a change of the precipitation sequence from that of
the balanced alloy to the excess Mg alloy. Figure 11 shows the
results obtained for the Al–1.0 mass%Mg2 Si–0.4 mass%Mg
alloy aged at 523 K. Figure 11(a) shows a bright-field image of rod-shaped precipitates, as seen in balanced alloy 2.
HRTEM observation of the cross-sections reveals β -phase as
Effects of Cu and Transition Metals on the Precipitation Behaviors of Metastable Phases at 523 K in Al–Mg–Si Alloys
alloy2
alloyC2
alloy4
alloy7
2793
alloy9
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
alloyC4
alloyC7
alloyC9
Fig. 6 HRTEM images of typical precipitates in alloys aged at 523 K for 12 ks. (a) Alloy 2, (b) alloy 4, (c) alloy 7, (d) alloy 9, (e) alloy
C2, (f) alloy C4, (g) alloy C7 and (h) alloy C9.
Fig. 8 TEM image of the grain boundary precipitates in the alloy 4 aged
at 523 K for 12 ks. (a) Bright-field image, (b) selected area diffraction
pattern, and (c) EDS profile of the precipitate marked by an arrow.
Fig. 7 TEM image of the Type-A precipitate in alloy 4 aged at 523 K for
12 ks.
shown in Fig. 11(b). Figure 11(c) shows changes in variation
of the type of the precipitates in this alloy with aging time.
In this figure, the relative frequency is given with respect to
aging time at 523 K. The β -phase is the dominant precipi-
tate in the sample aged for 12 ks and there is no difference
from the balanced alloy.9) Thus, the precipitation sequence of
the metastable phase in the excess-Mg alloy is similar to that
of the balanced alloy. This suggests that Cu addition to the
excess-Mg and balanced alloys does not cause any changes
to the kind of metastable phases, when the compositions of
alloys C7 and C9 became the excess-Mg and balanced alloys
with the addition of TM.
The results of the present HRTEM observation are summarized in Table 2. The results for the TM-bearing C4 alloy differ remarkably from those for the TM-free alloy, in contrast to
2794
K. Matsuda, S. Taniguchi, K. Kido, Y. Uetani and S. Ikeno
Table 2
TM-free alloys
Aging temp. (K)
Balanced
Excess Si
Cu-bearing
Excess Si+Cu
423
Random-type
Parallelogram-type
β
Random-type
Random-type
Random-type
Type-B
(Type-A, β )
Random-type
β
Q
Q
Random-type
Q
Random-type
Q
523
TM-bearing alloys
Summary of TEM observation of various precipitates.
423
523
Random-type
Parallelogram-type
β
Fig. 10 EDS profiles obtained from precipitates in (a) alloy C9, (b) alloy
C7, (c) alloy C4 and (d) alloy C2. Precipitates in alloys C9 and C7 have
clear Cu peaks. The aging treatment was at 523 K for 12 ks.
Fig. 9 TEM images of alloy C9 aged at 523 K for 12 ks. (a) Bright-field
and (b) dark-field images. (c) Typical HRTEM image of Q -phase in this
sample.
the previous results obtained for 423 K aging.10) The change
in types of precipitate with the decrease in Si content by the
formation of dispersoids was observed in our previous studies on Fe-bearing and Mn-bearing Al–Mg–Si alloys,18, 19) and
those results are also supported the present mechanism. As
the addition of TM to Al–Mg–Si alloys causes decrease of Si
contents, it can be suggested that the addition of Si in excess
and Cu is useful to save or improve hardness of TM-bearing
alloys.
4. Conclusions
Precipitates in the eight Al–Mg–Si alloys, which varied
Mg and Si contents, containing Cu and TM after 523 K aging were observed by HRTEM. Results are summarized as
follows.
(1) Precipitates in the Cu-bearing alloys are smaller than
those in the Cu-free alloys. Coarse dispersoids of AlFeCrSi
and AlFeSi existed in the TM-bearing alloys. Inhomogeneous
precipitation was also observed at interfaces between dispersoids and the matrix. Fewer precipitates were present in the
TM-bearing balanced alloy (alloy C7) than in the TM-free
balanced alloy (alloy 7).
(2) The β -phase and the Type-B precipitate exist in the
balanced alloys (alloy 2) and excess-Si alloy (alloy 4), respectively. The Q -phase exists in Cu-bearing alloys (alloy 7 and
9). The precipitates in the TM-bearing alloys are almost the
same as in TM-free alloys. However, the precipitate in the
TM-bearing excess-Si alloy (alloy C4) was the β -phase. This
means that the Type-B precipitate in the excess-Si alloy (alloy
4) was replaced to the β -phase in the balanced alloy by the
addition of TM.
(3) The TM-bearing alloys contained dispersoids of
AlSi(Fe, Cr) and AlSiFe, and it caused the decrease of Si content in the matrix.
(4) The precipitation sequence of metastable phases in
the excess-Mg alloy aged at 523 K was similar to those in
the balanced alloy.
(5) The age-hardenability of TM-bearing alloys at 423 K
aging can be controlled via the amount of excess-Si and Cu
addition.
Acknowledgements
The Authors are grateful to Prof. G. Ito of Ibaragi
University, and the Research Group for “Properties Near
Grain Boundaries” in the Japan Institute of Light Metals, for
the supply of alloys used in the present study.
Effects of Cu and Transition Metals on the Precipitation Behaviors of Metastable Phases at 523 K in Al–Mg–Si Alloys
REFERENCES
(b)
(a)
Relative frequencies of precipitates (%)
200nm
100
80
(c)
Parallelogram-type
’-phase
60
Random-type
40
20
”-phase
0
-1
asQ. 10
2795
100
101
Aging time,t /ks
Fig. 11 Precipitates in Al–1.0 mass%Mg2 Si–0.4 mass%Mg alloy aged at
523 K. (a) Bright-field image of microstructure aged at 523 K for 1.2 ks and
(b) a typical HRTEM image of the β -phase in this sample. (c) Changes
in relative frequencies of metastable precipitates in this alloy with aging
time.
1) D. L. Collins: J. Inst. Metals 86 (1957-58) 325–336.
2) D. D. Perovic, G. C. Weatherly and D. J. Lloyd: Scr. Mater. 41 (1999)
703–708.
3) C. Cayron, L. Sagalowicz, O. Beffort and P. A. Buffat: Philos. Mag. A
79 (1999) 2833–2851.
4) G. A. Edwards, K. Stiller, G. L. Dunlop and M. J. Couper: Acta Mater.
46 (1998) 3893–3904.
5) L. Lodgaad and N. Ryum: Mater. Sci. Tech. 16 (2000) 599–604.
6) Microstructures and Properties of Aluminum, (Japan Inst. Light Metals,
Tokyo, Japan, 1991) pp. 280.
7) K. Matsuda, H. Gamada, K. Fujii, Y. Uetani, T. Sato, A. Kamio and S.
Ikeno: Metall. Mater. Trans. 29A (1998) 1161–1167.
8) K. Matsuda, Y. Sakaguchi, Y. Miyata, Y. Uetani, T. Sato, A. Kamio and
S. Ikeno: J. Mater. Sci. 35 (2000) 179–189.
9) K. Matsuda, Y. Uetani, T. Sato and S. Ikeno: Metall. Mater. Trans. 32A
(2001) 1293–1299.
10) K. Matsuda, K. Kido, S. Taniguchi, Y. Uetani and S. Ikeno: J. Japan
Light Metals, (to be submitted).
11) Microstructures and Properties of Aluminum, Japan Inst. Light Metals,
Tokyo, Japan, (1991) pp. 356–357.
12) Y. Baba and A. Takashima: J. Japan Inst. Light Metals 19 (1969) 90–98.
13) H. Suzuki, M. Kanno, Y. Shiraishi and K. Hanawa: J. Japan Inst. Light
Metals 29 (1979) 575–581.
14) T. Uno and Y. Baba: J. Japan Inst. Metals 45 (1981) 405–410.
15) R. J. Livak: Metall. Trans. A. 13A (1982) 1318–1321.
16) S.-W. Lai and D. D. L. Cung: J. Mater. Sci. 29 (1994) 2998.
17) Y. Uetani, M. Isurugi, K. Matsuda, S. Tada and S. Ikeno: J. Japan Light
Metals 42 (1992) 578–584.
18) H. Tanihata, K. Matsuda and S. Ikeno: Mate. Sci. Forum 217–222
(1996) 809–814.
19) K. Matsuda, R. Fujii, T. Kawabata, Y. Uetani and S. Ikeno: J. Japan
Light Metals 51 (2001) 279–284.