Materials Transactions, Vol. 46, No. 12 (2005) pp.

2602 to 2608
Special Issue on Growth of Ecomaterials as a Key to Eco-Society II
#2005 The Japan Institute of Metals

High-Speed Twin-Roll Strip Casting of Al–Mg–Si Alloys with High Iron Content
Kenta Suzuki1 , Shinji Kumai1 , Yuichi Saito1; * and Toshio Haga2
1
Department of Materials Science and Engineering, Tokyo Institute of Technology, Yokohama 226-8502, Japan
2
Department of Mechanical Engineering, Osaka Institute of Technology, Osaka 535-8585, Japan

Thin strips of the 6063 aluminum alloy and the alloys with increased nominal Fe contents (0.7–6 mass%Fe) were fabricated directly from
the molten alloys using a vertical-type high speed twin-roll strip caster equipped with a pair of water-cooled pure copper rolls. The estimated
cooling rates from the DAS measurement were about 4500 and 100
C/s at the near the surface region and the mid-thickness region of the strip
respectively. Refinement of Al–Fe–Si intermetallic compound particles was also successfully achieved, however, segregated coarse particles
were also observed at the mid-thickness region. The cast strips were cold-rolled and heat-treated to form 0.5 mm-thick thin sheets. They were
either naturally aged at room temperature (T4) or artificially aged (T6), and then subjected to bending, tensile and hardness tests. No detrimental
effect of Fe was appeared concerning the bendability even in the alloy containing 3 mass%Fe. No cracking took place even in the 180
bending
(hemming) test. The reduction of age-hardenability was evident for the alloy with 1 mass%Fe and more. The alloy sheet containing 0.7 mass%Fe
exhibited not only a good bendability in the T4 condition but also the larger proof stress and UTS than those of the master alloy in the T6
condition. This means that the capacity of Fe impurity in a 6063 alloy (0.35 mass% according to JIS) can be doubled by using the present roll
caster. Strip casting at a high cooling rate using the present roll caster is considered to be a promising method for reducing the detrimental effect
of impurity iron from the scrap melt.

(Received July 4, 2005; Accepted October 18, 2005; Published December 15, 2005)
Keywords: strip casting, rapid solidification, aluminum–magnesium–silicon aluminum alloys, iron, bendability, recycling

1. Introduction insoluble elements, and (iii) fine and homogeneous distribu-

tion of secondary particles.
In order to achieve closed product-to-product recycling, it Because of these benefits, fabrication of the alloy strip at a
is essential to thoroughly reduce the harmful impurities from high cooling rate is a promising method for reducing the
scrap melt. However, purification of scrap melt to the original detrimental effect of iron impurity in the scrap melt. Relevant
level of the virgin metal and alloy is quite costly and is often research has been carried out by Shibue et al.4) using a
accompanied by additional environmental problems and horizontal-type twin roll caster. The objective of their
increased energy consumption. Therefore, we must look for research was to increase the Fe content in a 6016 aluminum
another method for fabricating from scrap metal a product of alloy sheet from the present 0.2 to 0.4 mass%Fe.
as high a quality as the original. In contrast, we have developed a vertical-type twin-roll
The use of wrought aluminum alloy for inner and outer caster. This caster achieved a cooling speed about 10 times
skins to reduce body weight and consequently improve fuel higher than that of the conventional horizontal-type twin-roll
efficiency is expected to increase significantly in the next caster.5)
decade.1) This means that the amount of wrought aluminum In this study, the cast strips of 6063 aluminum alloy and
alloy scrap is also expected to increase in the near future. alloys with increasing Fe content were fabricated by the high-
However, the demand for aluminum scrap for cast and die- speed twin-roll strip caster. The rapidly solidified cast strips
cast products in open recycling systems is currently saturated. were cold-rolled and heat-treated to form thin sheets.
Therefore, closed recycling of aluminum alloy body sheets Microstructure, hardness, tensile properties and bendability
should be promoted in order to use these precious natural of the resultant sheets were measured and the effectiveness of
resources effectively. rapid cooling is discussed.
The introduction of additional iron (Fe) to the recycled
aluminum alloy is of primary concern since increased iron 2. Experimental Procedures
levels are known to reduce formability.1,2) The iron combines
with other alloying elements to form second-phase particles. 2.1 Alloy preparation
They act as void-nucleation sites during deformation.3) A commercial 6063 aluminum alloy and alloys with
Strip casting is as a method that produces aluminum alloy increasing Fe content were prepared. The chemical compo-
sheets directly from the melt. However, the practical sition of the 6063 alloy ingot is shown in Table 1. Fe content
applications of strip casting have been limited because the of the 6063 alloy was less than 0.2 mass%. We will refer to
quality of cast strips is generally inferior to that of conven- the 6063 alloy as the master alloy from now on.
tional DC cast and hot-rolled strips. However, the recent The alloys with different Fe content (0.7, 1.0, 2.0, 3.0, 4.0,
strong moves towards recycling of wrought aluminum alloy 5.0 and 6.0 mass%) will be referred to as 0.7%Fe alloy, 1%Fe
reminded us of the particular advantages of strip casting. The
high cooling rate of strip casting is expected to lead to (i)
grain refinement, (ii) supersaturation of the relatively Table 1 Chemical composition (mass%) of the 6063 alloy ingot.

Si Mg Fe Cu Mn Ti Cr Zn
*Graduate Student, Tokyo Institute of Technology. Present address: Fuji 6063 0.44 0.48 0.17 0.01 0.01 0.01 0.00 0.01
Heavy Industries Ltd, Oura-gun 370-0531, Japan
 High-Speed Twin-Roll Strip Casting of Al–Mg–Si Alloys with High Iron Content 2603

Table 2 Melting temperature of each alloy. the cooling slope. The volume fraction of solid formed at this
Fe content (mass%) Melting temperature ( C)
stage was estimated to be less than 10%. The partially
0.35 (6063 alloy) 670
solidified melt flowed down, filled the casting nozzle and
0.7 670
then came into contact with the rotating roll surfaces. The
1.0 670
peripheral speed of the rotating roll was 60 m/min (1 m/s).
2.0 680
The melt rapidly cooled at the roll gap and the solid grew
3.0 730
from both roll surfaces. The initial roll gap was 1 mm.
4.0 755
However, the solidified alloy forced open the initial roll gap.
5.0 785
Therefore, the resultant strip thickness was greater than the
6.0 810
initial roll gap (1 mm). Strip thickness varied depending on
the alloy composition and the pouring temperature.

2.4 Annealing and cold rolling

(a) (b) The resultant cast strip was cut into several pieces and
Casting nozzle
Cooling slope annealed at 540
C for 2 h. The annealed strip was cold-rolled
into a 1 mm thick sheet parallel to the longitudinal direction
Springs of the cast strip. No surface treatment was performed. The
(Load)
sheet was annealed again at 540
C for 1 h and cold-rolled into
a 0.5 mm thick sheet.

2.5 Aging treatment, bending, hardness and tensile tests

The cold-rolled sheet was solution-treated at 540
C for 2 h,
water-quenched and either naturally aged at room temper-
ature (T4) or artificially aged at 160
C for 6 h (T6). Small
Water-cooled
Pure copper rolls rectangular strips (10 mm  30 mm) were taken from the
sheet both parallel and perpendicular to the rolling direction
for the 180
bending tests. Micro-Vickers hardness tests were
performed under a load of 100 g for 10 s. Tensile specimens
Fig. 1 Appearance of the high-speed twin-roll strip caster (a) and the
enlarged view of the place where rapid solidification of the melt occurs (b). with shoulders (gauge length: 26 mm, gauge width: 10 mm)
were machined from the sheet parallel to the rolling direction.
Tensile tests were carried out using an Instron-type machine
at a constant crosshead speed of 1 mm/min.
alloy, etc. from now on. The amount of Fe impurity in the
0.7 mass%Fe was twice that of the allowable Fe content 2.6 Microstructural observations
(0.35 mass%) of the 6063 aluminum alloy designated in the Optical microscopy was carried out on transverse sections
JIS (Japanese Industrial Standards). The master alloy was of the cast strip and the cold-rolled sheet. Polished samples
melted and a pure Fe wire was dissolved slowly into the melt. for metallography were anodized at 25 V and a current
The melting temperature of each alloy was 15
C above the density of 0.2 Amm2 for 90 s in a 2% solution of HF.
liquidus line temperature of the Al–Fe phase diagram, as Secondary dendrite arm spacing (DAS) of the master alloy
listed in Table 2. It should be mentioned that the Fe content was measured using an intersection line method.6) SEM
of the alloys are only nominal values. observation of the samples was carried out after the bending
test.
2.2 Roll caster
Figure 1 shows the vertical-type twin-roll caster used in 3. Results
this study. Figure 1(a) is what the roll caster actually looks
like. Figure 1(b) is an enlargement of the place where rapid 3.1 Macroscopic appearance of the cast strip
solidification of the melt occurs. The roll caster was equipped A 100 mm wide 2 to 3 m long continuous cast strip was
with a pair of water-cooled pure copper rolls with a diameter fabricated for all the alloys. The thickness of the strip
of 300 mm and a face-width of 100 mm. No lubricant was decreased as the Fe content increased, as shown in Fig. 2.
used on the roll surfaces. A casting nozzle consisting of steel The thickness of the 6%Fe alloy was about a half that of the
plates coated with a heat-insulating material was mounted on master alloy. No large difference in the surface condition of
the rolls. It should be mentioned that, in the figure, the side the strip was observed. However, the surface glossiness was
walls of the nozzle have been removed in order to show the slightly reduced for the alloy strips containing more than
appearance of the roll gap. One of the rolls can move 1%Fe. The reduction in thickness and glossiness of the alloy
independently since the roll is not fixed but supported by a strips with higher Fe content is thought to be due to the
series of springs. A cooling slope is located above the rolls. increased pouring (melting) temperature.
In the high-speed twin-roll caster, the sum of the solid
2.3 Casting procedures thickness growing from the both roll surfaces determined
The molten alloy (about 2.5 kg for one charge) was poured the strip thickness. As the melt temperature increased, the
onto the cooling slope. The melt was partially solidified on efficiency of heat extraction from the melt to the copper roll
2604 K. Suzuki, S. Kumai, Y. Saito and T. Haga

4.0
Near-surface Mid-thickness
(a) (g)

3.0
Thickness, t / mm

2.0
(b) (h)

1.0

0.0 (c) (i)

0 1 2 3 4 5 6 7

Fe content, C / mass%

Fig. 2 Change in thickness of the strips with increasing Fe content.

(a)
(d) (j)

(b)

(e) (k)

(c)

(f) (l)

200µm

Fig. 3 Anodized microstructures of the transverse cross section of the cast

strips. Master alloy (a), 1%Fe (b) and 3%Fe (c). 10µm 10µm

Fig. 4 Change in as-polished microstructure of the near-surface and mid-

thickness regions with increasing Fe content. Master alloy (a) (g), 0.7%Fe
was reduced. This may result in a reduced growth rate of the
(b) (h), 1%Fe (c) (i), 3%Fe (d) (j), 4%Fe (e) (k) and 6%Fe (f) (l).
solids before the growing solid-fronts finally meet at the
center line of the roll gap to form a strip. Consequently, the
thickness of the strip is reduced. Another possible reason for 3.2 Grain structure of the cast strip
the reduction in thickness is that the addition of Fe may Figure 3 shows the anodized microstructure of the trans-
change solidification from a skin-type to a mushy-type verse cross section of the cast strip for the master alloy (a),
solidification. The strip produced from the high temperature 1%Fe (b) and 3%Fe (c). Differences in sheet thickness are
melt (high Fe content) looked relatively soft when it came clearly shown in Fig. 3. In spite of the different thicknesses,
out of the roll gap. This means that the strip was still hot. they exhibited a common microstructural feature. Columnar
The tarnished surface condition of the high Fe content grains grew from both surfaces and they were replaced
alloys is thought to be due to a relatively slow cooling gradually by equiaxed grains. At the mid-thickness region, a
process in air. band of granular grains was observed. Dispersion of small
 High-Speed Twin-Roll Strip Casting of Al–Mg–Si Alloys with High Iron Content 2605

(a) (b)
Top-view Side-view
(a) (g)

(c) (d)
(b) (h)

(e) (f)
(c) (i)

100µm
(d) (j)
Fig. 5 Anodized microstructure of the transverse cross section of the
0.5 mm thick cold-rolled and T4-treated sheet. Master alloy (a), 0.7%Fe
(b), 1%Fe (c), 3%Fe (d), 4%Fe (e) and 6%Fe (f).

(a) (b)
(e) (k)

(c) (d) (f) (l)

100µm 100µm

(e) (f)
Fig. 7 SEM micrographs of the top- and side-view of the T4-treated sheet
after the 180
bending (hemming) test. Master alloy (a) (g), 0.7%Fe
(b) (h), 1%Fe (c) (i), 3%Fe (d) (j), 4%Fe (e) (k) and 6%Fe (f) (l).

10µm 3.3 Distribution of the intermetallic compound particles

in the cast strip
Fig. 6 Intermetallic particles at the mid-thickness region of the sheet for Figure 4 shows the change in as-polished microstructure
the master alloy (a), 0.7%Fe (b), 1%Fe (c), 3%Fe (d), 4%Fe (e) and 6%Fe of the near-surface and mid-thickness regions with an
(f). increase in Fe content. Coarse needle-like particles were
frequently observed for the alloy with 3% and higher Fe
content. They were large both in terms of size and volume
granular grains was also observed in the columnar and fraction at the mid-thickness region. Various kinds of Al–Fe–
equiaxed grain regions. It is thought that such granular grains Si have been reported for the commercial 6063 alloy.7) In this
arose from the crystallized
-Al dendrites on the cooling study, these particles were confirmed to be Al–Fe–Si base
slope. For 3%Fe, coarse needle-like intermetallic compound intermetallics by EDX analysis, but no further specific
particles were also observed at the mid-thickness region. identification was made.
2606 K. Suzuki, S. Kumai, Y. Saito and T. Haga

250
Top-view Side-view
(a)
(a) (g)

Proof stress and UTS, P / MPa

200

UTS
T6
150
T4
(b) (h)
100
Proof stress T6

50 T4

(c) (i)
0
0 1 2 3 4 5 6 7
Fe content, C / mass%

30
(b)
(d) (j) 25
T6
Elongation, E / %

20

T4
15

(e) (k)
10

0
(f) (l) 0 1 2 3 4 5 6 7
Fe content, C / mass%

Fig. 9 Effect of Fe content on proof stress, UTS and elongation.

100µm 100µm

3.5 Bendability of the T4- and T6-treated sheet

Fig. 8 SEM micrographs of the top- and side-view of the T6-treated sheet Figure 7 shows SEM micrographs of the top- and side-
after the 180
bending (hemming) test. Master alloy (a) (g), 0.7%Fe
view of the T4-treated sheet after the 180
bending (hem-
(b) (h), 1%Fe (c) (i), 3%Fe (d) (j), 4%Fe (e) (k) and 6%Fe (f) (l).
ming) test. No cracking occurred in the alloys containing 3%
or less Fe. Results for the T6-treated sheet are shown in
Fig. 8. Small cracks were detected for the bent specimen of
3.4 Microstructure of the cold-rolled sheet the master alloy. However, no cracking occurred in alloys
Figure 5 shows the anodized microstructure of the trans- containing 0.7, 1, 2 and 3%Fe although they were T6-treated.
verse cross section of the 0.5 mm thick cold-rolled and T4-
treated sheet. Recrystallized grain structure was clearly 3.6 Tensile properties of the T4- and T6-treated sheet
observed. Grain size reduced significantly with increasing Fe Figure 9 shows the effect of Fe content on proof stress,
content up to 3 mass%. Dark spots or banded areas in the UTS, and elongation. Large increases in proof stress and UTS
mid-thickness region corresponded to the segregated coarse were obtained for the T6-treated master alloy and 0.7%Fe
Al–Fe–Si intermetallic particles. Figure 6 shows the particles alloy. The proof stress and UTS for the 0.7%Fe alloy were
in the mid-thickness region of the sheet. Needle-like particles higher than those of the master alloy. In contrast, there was
were reduced and replaced by irregular-shaped blocky little or no increase in UTS and proof stress for the T6-treated
particles. alloys containing 1% or more Fe. Reduced age-hardening is
evident for the alloys containing 1% or more Fe. The
 High-Speed Twin-Roll Strip Casting of Al–Mg–Si Alloys with High Iron Content 2607

elongation of the T4-treated sheets gradually decreased with examination revealed that the cooling rate is more than
an increase in Fe content. For the T6-alloy sheets, elongation 100
C/s even at the mid-central region of the strip. It is
became larger with an increase in Fe content up to 3%Fe and unlikely that these coarse intermetallic particles formed at
then decreased when the Fe content exceeded 3%. such high cooling rates. The cooling slope is not considered
to be a likely reason for the formation of such particles either.
4. Discussion From a morphological viewpoint, these coarse needle-like
particles are -AlFeSi. The particles were more than 10 mm
4.1 Estimated cooling speed achieved by the twin-roll long. Particles of this size were reported in the furnace cooled
casting process 6063 alloy at a cooling rate of 0.06
C/s by Tanihara et al.7)
Secondary dendrite arm spacing (DAS) of the as-solidified In this study, the pure Fe wire was dissolved slowly into
strip can be used as a useful guide for estimating the cooling the melt of the master alloy. The melt was held for one hour
rate in the course of solidification. The relationship between before pouring. However, referring to the present exper-
DAS and the cooling rate has been examined by many imental results, it is thought that the original molten alloys
researchers for various alloys. Horiuchi and Kaneko exam- with high Fe contents (more than 2%) were not homogeneous
ined the relationship for 6063 alloy and obtained the and included these coarse particles.
following equation.8)
4.3 Effects of Fe on age-hardening of the sheet
DAS ¼ 85V0:38
As shown in Fig. 9, a significantly reduced age-hardening
where DAS is the dendrite arm spacing (mm) and V is the effect was observed for alloys with 1% or higher Fe. In order
cooling rate (
C/s). The measured DAS values for the master to confirm this supposition, micro-Vickers hardness tests
alloy were 3.5 mm at the near-surface region and 14.5 mm at were performed for both the T4- and T6-treated sheets. The
the mid-thickness region. By using the equation, the cooling results are shown in Fig. 10. Although a large increase in the
rates were estimated at about 4500
C/s at the near-surface hardness was obtained by T6 treatment of the master and
and about 100
C/s at the mid-thickness region. The 0.7%Fe alloy, only a moderate hardness increase was
estimated cooling rate near the surface was the same as that obtained for alloys with Fe content of 1 or higher mass%.
obtained by the present authors for the twin-roll cast A356 Age-hardening of the alloy mainly occurs by the precip-
alloy.9) However, the estimated cooling rate at the mid- itation of Mg2 Si in the matrix. The decreased hardness is
thickness region was lower than that of A356 (500
C/s). This thought to be due to the reduction in the quantity of
may result from the complicated microstructure in this precipitates. When the molten alloy contains a large amount
region. Solidified microstructure at the mid-thickness region of Fe, Si atoms are used to form the Al–Fe–Si base
is a mixture of equiaxed and granular grains formed on the intermetallic compound during cooling. Consequently, the
cooling slope. The DAS of granular grains is coarser than that number of Si atoms remaining in the
-aluminum matrix
of equiaxed grains. We need to measure the DAS of the decreases and so the quantity of precipitates also decreases.
equiaxed grains in order to obtain the local cooling rate. With the cooling conditions used in this study, 1 mass% of Fe
However, it is generally hard to distinguish one from the is considered to be enough to trap the Mg atoms and
other in such a bi-modal DAS size distribution. Therefore, the eliminate the hardening effect of artificially aging.
average DAS value tends to be the overestimated one. This
may result in an underestimated cooling rate at the mid-
thickness region.
100
4.2 Effects of Fe content and cooling rates on the
formation of Al–Fe–Si base intermetallic compound
Mulazimoglu et al. found the Chinese-script-type
- 80
AlFeSi and the plate-like and sharp needle-like -AlFeSi
Vickers hardness, Hv / HV

particles in the Al–0.58%Mg–0.58%Si alloy containing

0.29%Fe.10) Tanihara et al. examined the relationship T6
60
between the kind of Al–Fe–Si intermetallic compound and
the cooling rate for the 6063 alloy at various Fe contents
using a TEM technique.7) -AlFeSi is mainly formed when
an alloy with low Fe content (0.01 mass%) is solidified at a 40
T4
low cooling rate (0.06
C/s). It was reported that -AlFeSi is
replaced by
- and
0 -AlFeSi as the cooling rate is increased
to 5
C/s. The Chinese-script-type
-AlFeSi increases with an 20
increase in Fe content. The main phase in the 0.5%Fe as-cast
ingot is the
-AlFeSi at a casting cooling rate of 50
C/s.
As mentioned in the experimental section, the Fe contents 0
described in this paper are nominal values. Quite large 0 1 2 3 4 5 6 7
intermetallic compounds were observed even for the near- Fe content, C / mass%
surface region of the 3%Fe. Their size and morphology were
similar to those of alloys with higher Fe content. The DAS Fig. 10 Micro-Vickers hardness for T4 and T6-treated sheet.
2608 K. Suzuki, S. Kumai, Y. Saito and T. Haga

4.4 Effects of Fe on bendability of the sheet bendability but also a higher proof stress and UTS than
In the automotive industry, hemming is often used to join those of the T6-treated master alloy. This means that,
outer skins to inner panels. The very tight radius of curvature although there are many problems left to be solved, the
in a hem is an extreme test of the bendability of sheet content of Fe impurity in a 6063 alloy can be doubled by
material, and tearing or cracking of the outer tensile bend using the high-speed twin-roll caster. Strip casting at a high
surface is common. Under these conditions, the combined cooling rate using the high-speed twin-roll caster is a
role of void damage nucleated from Fe-based constituent promising method for reducing the detrimental effect of iron
particles and shear bands emanating from outer surface impurity from the scrap melt.
undulations is very important.3)
Even the alloy sheet containing 3%Fe could be bent Acknowledgements
successfully without cracking. Such good bendability of the
T4-treated sheet is thought to result from the following: The authors would like to acknowledge the financial
(1) Intermetallic compound particles which act as initiation support provided by RISTEX of JST (Japan Science and
sites for cracking are mainly dispersed at the mid- Technology Agency) and the Light Metal Education Foun-
thickness region in the twin roll cast strip. The cast strip dation. They also wish to acknowledge Miss Akiko Ishihara
was rolled into a sheet directly without removing the and Mr. Yusuke Nagata for their assistance in experiments.
surface. Therefore, the intermetallic particles had little The Japan Aluminum Association is also acknowledged by
chance of appearing on the sheet surface. one of the authors (K. Suzuki) for the financial support.
(2) Decrease in grain size was observed with an increase in
Fe content (see Fig. 6). Refined grain structure is
REFERENCES
generally beneficial to homogeneous deformation.
In the present study, even the T6-treated sheets could be 1) J. Sarkar, T. R. G. Kutty, D. S. Wilkinson, J. D. Embury and D. J.
bent without cracking, as shown in Fig. 8. However, Figs. 9 Lloyd: Mater. Sci. Eng. A 369 (2004) 258–266.
and 10 indicate that the
-Al matrix of the alloy with 1% and 2) A. Hibino, T. Muramatsu, M. Saga and K. Takata: J. Jpn. Inst. Light
higher Fe was not actually age-hardened. Therefore, the Met. 53 (2003) 534–541 (in Japanese).
3) W. B. Lievers, A. K. Pilkey and D. J. Lloyd: Mater. Sci. Eng. A 361
bending test results for the T6-treated sheet should be treated (2003) 312–320.
with caution. 4) K. Shibue, H. Uto and K. Ito: 107th Abstracts of the Japan Inst. of Light
Metals (2004) 9–10.
5. Conclusions 5) T. Haga, K. Takahashi, M. Ikawa and H. Watari: J. Mater. Proc.
Technol. 140 (2003) 610–615.
6) Japan Institute of Light Metals, Committee of casting and solid-
A 6063 aluminum alloy and seven other alloys with ification: J. Jpn. Inst. Light Met. 38 (1988) 54–60 (in Japanese).
increasing Fe content were fabricated in a high-speed twin- 7) H. Tanihara, T. Sugawara, K. Matsuda and S. Ikeno: J. Mater. Sci. 34
roll strip caster. The rapidly solidified 1.5–2.7 mm thick cast (1999) 1205–1210.
strips were cold-rolled and heat-treated to form 0.5 mm thick 8) R. Horiuchi and J. Kaneko: J. Jpn. Inst. Met. 35 (1971) 156–161 (in
Japanese).
sheets. There appeared to be no detrimental effect of Fe on
9) K. Suzuki, S. Kumai, Y. Saito, A. Sato and T. Haga: Mater. Trans. 45
the bendability, even in the alloy containing 3 mass%Fe. (2004) 403–406.
However, a reduction of age-hardenability was evident for 10) M. H. Mulazimoglu, A. Zaluska, J. E. Gruzleski and F. Paray: Metall.
the alloy with 1 or higher mass%Fe. The T4-treated alloy Mater. Trans. A 27A (1996) 929–936.
sheet containing 0.7 mass%Fe exhibited not only good