Intermetallics 86 (2017) 51e58

Contents lists available at ScienceDirect

Intermetallics
journal homepage: www.elsevier.com/locate/intermet

An investigation on selective laser melting of Al-Cu-Fe-Cr quasicrystal:

From single layer to multilayers
Yingqing Fu a, b, Nan Kang b, *, Hanlin Liao b, Yang Gao a, Christian Coddet b
a
Department of Materials Science and Engineering, Dalian Maritime University, 116026, Dalian, China
b
University Bourgogne Franche-Comte, IRTES EA7274, F-90100 Belfort, France

a r t i c l e i n f o a b s t r a c t

Article history: In this study, the effects of processing parameters on the microstructure of Al-Cu-Fe-Cr quasicrystalline
Received 16 November 2016 (QC) coatings fabricated by selective laser melting (SLM) are investigated. A qualitative analysis on the
Received in revised form XRD patterns indicates that the phase composition for the SLM processed coating mainly consisted of Al-
27 February 2017
Cu-Fe-Cr quasicrystals and a-Al (CuFeCr) solid solution, and with increasing laser energy input or coating
Accepted 14 March 2017
thickness, the volume fraction of QC i-Al91Fe4Cr5 reduced and those of QC d-Al65Cu20Fe10Cr5 and crys-
talline q-Al2Cu increased. The formation of cracks during the coating building procedure from single layer
to multilayers is also discussed. For the coatings with the same layer number, the pores and balling
Keywords:
Selective laser melting
particles diminish as laser power increases, due to the growth of melting degree. At the early stage of
Al-Cu-Fe-Cr quasicrystal fabrication, with increment of layer number (or coating thickness), pores and balling particles decrease
Microstructure considerably because the molten pool solidified more “slowly”. However, after the layer number in-
Marangoni convection creases continuously from 10 to 20, the porosity no longer decreases, and some big size pores, micro-
cracks and fractures appear, especially for the sample obtained at lower laser power. A wavy-like pattern
composed chiefly of Al and QC phases, is formed at the interfacial region between substrate and coating
due to Marangoni effect.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction matrix/polymer-matrix composite materials as reinforcement

As intermetallics with long-range order but no translational
periodicity, quasicrystals have aroused more and more interests in
the intense theoretical and experimental studies, since firstly
discovered by Schatman et al. in the case of rapid solidified Al-Mn
alloy in 1984 [1]. Over one hundred kinds of quasicrystals were
developed up to date, they all are generally known for their high
strength, high corrosion and wear resistance, such as quasicrys-
talline (QC) alloys based on aluminum [2], copper [3], magnesium
[4], nickel [5], titanium [6] and zinc [7]. Among metallic glasses, Al-
Cu-Fe system QC alloys are known to present a very low cost and
high forming ability [8]. The QC phases can be obtained using
conventional solidification, which allows us to produce large
single-grain samples with a relatively low cooling rate [9]. How-
ever, due to their high room-temperature brittleness, the applica-
tions of QC alloys are confined to the coating area [10] or metal-
* Corresponding author. LERMPS - University of Technology BelforteMontbeliard,
venans, 90010 Belfort Cedex, France.
Site de Se
E-mail addresses: Yingqing_fu@aliyun.com (Y. Fu), nan.kang@utbm.fr (N. Kang).
particles [11,12]. Thermal spraying technologies were considered as
an effective way to produce the QC coatings [13] and quasicrystal-
reinforced composites [14], the high ductility of substrate/matrix
material could compensate for the fragile mechanical properties.
Recently, spray forming technology was used to produce the bulk
Al-Cu-Fe based quasicrystals and composites [15,16]. They indi-
cated that the spray formed sample presents a high hardness and
low ductility, leading to low workability.
Selective Laser Melting (SLM), a novel high accuracy powder-
bed fusion additive manufacturing process, has been more and
more used in the past decades [17]. And it is particularly attractive
to produce near net-shaped components with customized geom-
etries and properties, without resorting to moulds, tools or dies.
The 3D parts are constructed layer by layer, using a laser that melts
selectively the material particles at the surface of a powder bed to
construct the part cross-sections. Moreover, the high cooling rate
(103e108 K/s) could result in an ultrafine microstructure or even
metastable phases [18,19]. Attar et al. produced a dense composite
part with a nanostructure and high mechanical properties using
SLM [20]. Dubois et al. confirmed that using Al-based QC particles

http://dx.doi.org/10.1016/j.intermet.2017.03.012
0966-9795/© 2017 Elsevier Ltd. All rights reserved.
52 Y. Fu et al. / Intermetallics 86 (2017) 51e58

allows extending the selective laser sintering technology to pro-

duce Al-based parts of any complex shape [21]. Therefore, devel-
oping Al-based QC or quasicrystal-reinforced composite materials
via SLM may be interesting and valuable, for the production of
lightweight metallic components, especially in the automotive and
aviation industry.
The objective of this work is concentrated on the effects of SLM
process parameters, such as laser power and laser scanning speed
etc., on the melting of QC material, during which the QC parts
present numerous cracks and low formability, as layer number
increases. Thus “from single layer to multilayers” method was
employed. In this work, Al65Cu20Fe10Cr5 (all compositions are in
atom percent unless otherwise stated) QC alloy was melted on a
pure Al substrate using SLM under argon atmosphere. The in-
fluences of laser power and scanning speed on the microstructure
and phase composition were investigated. Additionally, the entire
building process from 1 layer (50 mm) to 20 layers (1 mm) was also
studied. Fig. 2. XRD patterns of gas-atomized powders.

2. Experimental morphology and cross-section of the gas-atomized powders.

An Al65Cu20Fe10Cr5 powder produced in the laboratory by gas
atomization in argon atmosphere (Nanoval process) was used in
this study. The surface morphology and cross-section of the QC
powders were shown in Fig. 1(a) and (b). Particles size distribution
was determined by laser light scattering (Malvern Masterizer
2000). The raw atomized metal glass powder exhibited an average
value of 45 mm (d (50)). The size distribution of the QC powders was
shown in Fig. 1(c). QC particles appear free of porosity as shown in
Fig. 1(b). The small grains could be observed in both the surface
X-ray diffraction (XRD) pattern of the gas-atomized Al65Cu20-
Fe10Cr5 powders is shown in Fig. 2, which indicates that a decagonal
QC Al65Cu20Fe10Cr5 (i.e. d-phase) is the major phase; and an
orthorhombic Al65Cu20Fe10Cr5 (i.e. O-phase) as an approximant of
the d-phase whose XRD patterns are quite similar to those of the d-
phase, also exists in the powders. The peaks of d-phase and O-
phase can be well indexed according to the scheme proposed by
Dong et al. [22]. Additionally, the other three minor phases detected
include two icosahedral QC phases, I-Al65Cu20Fe10Cr5 (i.e. I-phase)

Fig. 1. (a) Morphology and (b) cross-section of the Al65Cu20Fe10Cr5 QC powders.

 Y. Fu et al. / Intermetallics 86 (2017) 51e58 53

thickness of 1 mm were fabricated on the sandblasted pure Al

substrate, which was heated to 400 K during the SLM process.
Microstructures of coating samples were observed by an optical
microscopy (EPIPHOT, NIKON, Japan) and a scanning electron mi-
croscopy (JEOL JSM 7800F, Japan) equipped with X-ray energy
dispersive spectroscopy (EDS). The SLM coating surfaces were
analyzed in an X-ray diffractometer with Co Ka radiation
(l ¼ 1.78897 Å) operated at 35 kV and 40 mA.

3. Results and discussion

3.1. Densification and phase composition analysis

In the present study, laser beam focus position, hatch-spacing,

Fig. 3. Processing windows for selective laser melting of quasicrystalline powder
(Zone I-zone of non-forming, Zone II-Zone of poor formation, Zone III-Zone of forming)
versus laser power and scanning speed.
and Al91Fe4Cr5 [23] (i.e. i-phase), and a tetragonal crystalline phase
q-Al2Cu (i.e. T-phase).
A commercial SLM machine MCP-realizer SLM 250 equipped
with YLR-100-SM single-mode CW ytterbium fiber laser (MCP-HEK
Tooling GmbH, Germany) was employed. Argon was employed as
protective gas to avoid the oxidation of QC powder. Coatings with
scan distance and layer thickness were kept constant. The laser
power (P) as well as scanning speed (v) was varied in order to study
the effects of the energy input E ¼ P/v on the sample microstruc-
ture. According to our previous work [24,25], the laser power and
scanning speed were set ranging from 150 W to 200 W and 1 m/s to
4 m/s respectively. For simplicity, each sample is identified by the
values of energy input E used to prepare it. For instance, the coating
sample manufactured using P ¼ 150 W and v ¼ 1 m/s, which cor-
responds to an energy input E ¼ 150 J/m is referred to as S150. Ac-
cording to the formation ability of single layer and coating samples,
the process window was could be divided into three zones shown
in Fig. 3: Zone I-zone of non-forming, Zone II-Zone of poor for-
mation and Zone III-Zone of forming.
Analyzing the XRD patterns of the SLM samples shown in Fig. 4,
their phase assemblage includes not only the same phases as d-

Fig. 4. XRD patterns of SLM coatings' surfaces prepared with different (a) laser power (b) laser scanning speed and (c) layer number (d: decagonal QC Al65Cu25Fe10Cr5; O:
Orthorhombic Al65Cu25Fe10Cr5; I: icosahedral QC Al65Cu25Fe10Cr5; i: icosahedral QC Al91Fe4Cr5; T: Tetragonal q-Al2Cu; A: a-Al (CuFeCr)).
54 Y. Fu et al. / Intermetallics 86 (2017) 51e58

Table 1
XRD peaks intensity ratio of d/O, i/O and T/O for the SLM coating surfaces (d: decagonal QC Al65Cu25Fe10Cr5; O: Orthorhombic Al65Cu25Fe10Cr5; I: icosahedral
QC Al65Cu25Fe10Cr5; i: icosahedral QC Al91Fe4Cr5; T: Tetragonal q-Al2Cu).

Energy input E (J/m) 50 75 87.5 100 200

Coating sample No. S50 S75 S87.5 S100 S200
Coating thickness (layer) 20 20 20 20 5 20 5
d(13312)/O(852) peak intensity ratio (%) 15.7 17.2 18.5 19.6 2.7 24.8 3.2
i(100000)/O(852) peak intensity ratio (%) 9.3 7.8 5.8 4.9 62.7 1.7 85.7
T(310)/O(852) peak intensity ratio (%) 1.2 2.1 6.1 8.0 3.8 15.9 4.2

phase, O-phase, I-phase, i-phase and Al2Cu, but also two new
.
phases, a-Al (CuFeCr) and Fe-Cr, compared with that of the PIRi=O ¼ І
 І
 (2)
Al65Cu20Fe10Cr5 feedstock powders (Fig. 2). The order of increasing i 100000 O 852

melting point (low to high) is Al, Cu, Fe and Cr among these alloying
ingredients, when the feedstock powders were laser-heated and
fused, Al was firstly melted and liquid Al tended to float to the
molten pool surface because of its lower density, resulting in the
formation of a-Al (CuFeCr) solid solution. Moreover, Cr and Fe with
higher fusion points would solidify earlier than Al and Cu in the
following rapid solidification, leading to the Fe-Cr phase formation.
For a qualitative treatment, XRD peak intensity (area) ratios
(PIR) of d/O, i/O and T/O shown in Table 1, Figs. 5 and 6, were
calculated by the following equations:
. speed and keeping the other process parameters unchanged,
PIRd=O ¼ І
 І
 (1)
d 13312 O 8 52
..
PIRT=O ¼ І
 І
 (3)
T 310 O 852
As confirmed by Sordelet et al. [27,26], the value of PIR is roughly
related to the volume fraction of each phase in the coating sample,
so some phase's variation in volume fraction can be derived from its
change of PIR. Hence, it can be concluded that, with the increment
of laser power, the volume fractions of QC d-phase and T-phase (q-
Al2Cu) increased; whilst that of i-phase (Al91Fe4Cr5) reduced, as
shown in Table 1 and Fig. 5a. However, increasing laser scanning
resulted in an inverse effect on the volume fractions of these three
phases described before. The volume fraction of i-phase increased
but those of d-phase and T-phase decreased, as illustrated in Fig. 5b.

Fig. 5. XRD peak intensity ratio of d/O, i/O and T/O in SLM coatings prepared with different (a) laser power, (b) laser scanning speed and (c) energy input E.
 Y. Fu et al. / Intermetallics 86 (2017) 51e58
Fig. 6. XRD peak intensity ratio of d/O, i/O and T/O in S200 and S100 with different layer
numbers (5 and 20).
The effect of laser power (Figs. 4a and 5a) or scanning speed
(Figs. 4b and 5b) on the coating phase composition factually was
the influence of the laser energy input E, and it affected the XRD
peak intensity ratios of d/O, i/O and T/O as shown in Fig. 5c. It
implied that the volume fraction of i-phase reduced and those of d-
phase and T-phase increased with increasing the laser energy input
E.
Furthermore, the deposited layer number (or coating thickness)
played a significant role in the phase evolution, as shown in Figs. 4c
and 6. In the early stage of SLM processing (thickness  5 layers),
the molten pool was formed in the feedstock powders layer which
deposited directly on Al substrate (if it was No.1 layer) or on the
previously deposited layer. As mentioned before, the deposited
coating contained a significant number of QC phases such as d-
Al65Cu20Fe10Cr5, i-Al91Fe4Cr5 and I-Al65Cu20Fe10Cr5, which had very
low thermal conductivities (almost two orders of magnitude lower
than that of Al) [29,28]. But the coating was diluted by Al substrate
with high heat conductivity and coating thickness was too thin to
take dominant effect, the powders melt would solidify very rapidly
with a considerably high cooling rate. Holland-Moritz et al.'s
studies [30] revealed that the maximum undercoolings achieved
with respect to the melting temperature of the icosahedral QC
Fig. 7. Surface morphologies of the SLM processed samples with several layer numbers under different laser powers.
55
phase are significantly lower than the maximum undercoolings
typically observed for crystalline phases, while the undercoolability
of the decagonal QC phase lies in between. Moreover, an activation
threshold for the nucleation in undercooled melts is expected,
being smallest for the icosahedral QC phase (e.g. i-Al91Fe4Cr5),
medium for the decagonal QC phase (e.g. d-Al65Cu20Fe10Cr5), and
the Frank-Kasper phases (e.g. O-Al65Cu20Fe10Cr5) with large crys-
talline unit cells containing polytetrahedral symmetry elements,
and largest for non-polytetrahedral crystals (e.g. q-Al2Cu). There-
fore, i-phase nucleated in preference to the d-phase and T-phase,
from the significantly undercooled powders melts, and for both
S100 and S200, the volume fraction of i-phase was much higher than
those of d-phase and T-phase. With increasing the layer number,
the influence of low thermal conductive coating previously
deposited could not be neglected, and its thickness was high
enough (e.g.19 layers) to make cooling rate and undercooling de-
gree of the melt decreased dramatically. Then, the undercooling of
20th layer melt became so low and it solidified so “slowly”, that
some decagonal quasicrystal or even non-polytetrahedral crystals
were prone to form [13]. Hence, the volume fractions of d-phase
and T-phase increased whilst that of i-phase decreased massively
with increasing the layer number.
3.2. Surface morphologies
The surface morphologies of the SLM processed coating sam-
ples, in function of laser power and layer number, are shown in
Fig. 7. The pores and balling particles could be observed for all the
sample surfaces. It is obvious that an increase of laser power
resulted in a decrease of porosity, and the size and quantity of pores
reduced because the low fusion porosity decreased due to the
increasing degree of melting. The first layer deposited directly on
high thermal conductive Al substrate, so the melt solidified very
rapidly with considerably high cooling rate and undercooling,
many balling particles and pores being formed. With the accumu-
lation of layers (e.g. from 1 to 5 and 10), the existing time of the
molten pool increased and it solidified more “slowly”, and the
porosity due to the entrapment of gases by turbulent flow of liquid
metal in molten pools diminished [31e35]. Therefore, pores and
balling particles decreased significantly, and the laser traces could
be observed obviously on the surface with extremely low rough-
ness. However, after the layer number increased from 10 to 20, the
coating porosity no longer reduced, but on the contrary increased,
56 Y. Fu et al. / Intermetallics 86 (2017) 51e58
Fig. 8. Micrographs of cross-sectional quasicrystalline coatings for different laser powers: (a) 150 W, (b) 175 W and (c) 200 W.
and some big size pores and microcracks appeared, especially for
the sample obtained at lower laser power (e.g. 150 W). This pores
could be ascribed to the formation of metal (e.g. Al) vapor cavity
entrained by turbulent flow of molten pool, and with increasing the
degree of melting, the increase of thermal distortion and residual
thermal stresses resulted in the initiation and growth of micro-
cracks [36]. Additionally, it is obvious that the cracks directions in
the 20-layer samples fabricated at laser powers of 175 W and
200 W are very different, as indicated by arrows in Fig. 7. For the
sample with low laser power (175 W), the cracks directions are
random due to randomly distributed porosities, and most crack
paths are between pores. As laser power increases to 200 W, the
porosity could be almost eliminated. Thus the bonding strength of
inter region between two laser traces is relatively weak, where the
cracks are liable to develop and grow, and consequently the most
cracks directions are parallel or vertical to the laser traces.
3.3. Microstructure
Fig. 8 presents the microstructure of SLM processed QC coating
(cross-section) with several laser power (scanning speed fixed at
1 m/s). For the sample obtained at laser power of 150 W, the coating
presented a fragmental structure. As the laser power increased to
175 and 200 W, a stable coating with several big cracks could be
formed. In the meantime, a thin interaction layer with the wavy-
liked morphology appeared between substrate and coating for
the sample obtained at laser power of 175 W. As the laser power
continuously increased to 200 W, the thickness of interaction layer
increases from several micrometers to 50 mm. The cracks ran
through the coating from surface to interaction layer. The formation
of the interaction layer would be discussed in detail in the following
section.
In order to investigate the interaction layer between QC coating
and Al substrate, an EDS analysis with higher magnification was
carried out and corresponding results were shown in Fig. 9.
From Figs. 8c and 9a, a wavy-liked interaction layer with a dense
structure between substrate and coating could be observed clearly
in the sample obtained at high laser power (200 W). The EDS im-
ages of Al, Cu, Fe, Cr elements were presented in Fig. 9b, c, d and e. It
indicated that the dark region in Fig. 9a was mainly Al, and the gray
or light gray region consisted of Al-Cu-Fe-Cr, which presented a
similar composition with the QC feedstock powders. This could be
attributed to Marangoni convection [37e39]. It is known that when
interaction between laser radiation and metal powder occurs, the
energy is absorbed by powder particles through bulk coupling and
powder coupling mechanisms. The exposure region (spherical
surface with diameter about 40 mm) is firstly heated and then
melted. As the result of thermal conduction, the frontier region of
substrate is also melted. In general, a high energy input leads to a
high temperature rise and a large amount of liquid formation. As
reported by Dai et al. [40] and Kang et al. [41], during laser melting
process, a steep temperature gradient is formed, leading to surface
tension gradient as well as the resultant shear stress and convective
movement of the melt pool which is known as Marangoni effect
[37e39]. The melted QC powders and Al substrate flowed in the
molten pool, and the diffusion between them was significantly
limited, due to the extremely high cooling rate. Therefore, a wavy-
liked pattern, which consisted mainly of Al and Al-Cu-Fe-Cr qua-
sicrystals, was formed after rapid solidification.
4. Conclusion
Al-Cu-Fe-Cr quasicrystalline (QC) coatings were prepared by
selective laser melting. The effects of layer number, laser power and
scanning speed on the phase constituent and microstructure of the
coatings were investigated. The results are summarized as follows:
1 With increasing layer number or laser power or decreasing
scanning speed, the volume fraction of QC i-Al91Fe4Cr5 reduced
whilst those of QC d-Al65Cu20Fe10Cr5 and crystalline q-Al2Cu
increased, because the higher energy input or coating thickness
decreased the cooling rate and undercooling degree of the melt,
 Y. Fu et al. / Intermetallics 86 (2017) 51e58
Fig. 9. (a) SEM and (bee) EDS mapping of cross-sectional interface region for the QC coating produced at laser power of 200 W.
and the decagonal quasicrystal and even non-polytetrahedral
crystal were prone to form rather than icosahedral quasicrystals.
2 At low laser power or layer number, large pores and balling
particles can be observed in the coating surface, owing to the
insufficient melting of powders. Moderate increase of laser po-
wer or layer number can decrease the coating porosity and
balling particles due to the increasing fusion degree and the
slower solidification rate. However, with further increasing layer
number (e.g. from 10 to 20), the coating porosity no longer
decreased but increased; and some big size pores, microcracks
and fractures appeared, because of the increases of metal (e.g.
Al) vapor cavities, thermal distortion and residual thermal
stresses.
3 A wavy-like structure, mainly composed of Al and Al-Cu-Fe-Cr
quasicrystals, was formed at the interfacial region between
substrate and coating, due to Marangoni convection
Acknowledgements
The authors are grateful to the financial supports by the National
Natural Science Foundation of China (NSFC-No. 50805012) and the
China Scholarship Council (CSC-No. 201506575018 & CSC-No.
201206290187) and Marie Curie [Grant No. FP7-IPACT-
268696(EU)].
57
References
[1] D. Shechtman, I. Blech, D. Gratias, J.W. Cahn, Phys. Rev. Lett. 53 (1984)
1951e1953.
[2] C.J. Jenks, P.A. Thiel, J. Mol. Catal. A131 (1998) 301e306.
[3] Z.F. Li, B.X. Liu, Nucl. Instrum. Methods Phys. Res. B 178 (2001) 224e228.
[4] S. Yi, E.S. Park, J.B. Ok, W.T. Kim, D.H. Kim, Mater. Sci. Eng. A300 (2001)
312e315.
[5] B. Rubinstein, S.I. Ben-Abraham, Mater. Sci. Eng. A294e296 (2000) 418e420.
[6] M.A. Shaz, R.K. Mandal, O.N. Srivastava, Mater. Sci. Eng. A304e306 (2001)
860e862.
[7] T.J. Sato, E. Abe, A.P. Tsai, Mater. Sci. Eng. A304e306 (2001) 867e870.
[8] E.H. Saarivirta, J. Alloys Compd. 363 (2004) 150e174.
[9] M.F. Besser, T. Eisenhammer, Mater. Res. Soc. Bull. 10 (1997) 59e63.
[10] J.M. Dubois, A. Proner, B. Bucaille, Ph. Cathonnet, C. Dong, V. Richard,
A. Pianelli, Y. Massiani, S. Ait-Yaazza, E. Belin-Ferre , Ann. Chim. 19 (1994)
3e25.
[11] A.P. Tsai, K. Aoki, A. Inoue, T. Masumoto, J. Mater. Res. 8 (1993) 5e7.
[12] P.D. Bloom, K.G. Baikerikar, J.U. Otaigbe, V.V. Sheares, Mater. Sci. Eng. A
294e296 (2000) 156e159.
[13] Y.Q. Fu, L.T. An, F. Zhou, Y.Y. Zhao, D.M. Yang, Y. Gao, Surf. Coat. Technol. 202
(2008) 4964e4970.
[14] X.P. Guo, J.F. Chen, H.L. Yu, H.L. Liao, C. Coddet, Surf. Coat. Technol. 268 (2015)
94e98.
[15] V.C. Srivastava, E. Huttunen-Saarivirta, C. Cui, V. Uhlenwinkel, A. Schulz,
N.K. Mukhopadhyay, J. Alloys Compd. 597 (2014) 258e268.
[16] N.K. Mukhopadhyay, V. Uhlenwinkel, V.C. Srivastava, J. Mater. Eng. Perf. 24
(2015) 2172e2178.
[17] D.D. Gu, W. Meiners, K. Wissenbach, R. Poprawe, Int. Mater. Rev. 57 (2012)
133e164.
[18] X.P. Li, X.J. Wang, M. Saunders, A. Suvorova, L.C. Zhang, Y.J. Liu, M.H. Fang,
58 Y. Fu et al. / Intermetallics 86 (2017) 51e58
Z.H. Huang, T.B. Sercombe, Acta Mater. 95 (2015) 74e82.
[19] D.D. Gu, Y.C. Hagedorn, W. Meiners, G. Meng, R.J.S. Batista, K. Wissenbach,
R. Poprawe, Acta Mater. 60 (2012) 3849e3860.
[20] H. Attar, M. Bonisch, M. Calin, L.C. Zhang, S. Scudino, J. Eckert, Acta Mater. 76
(2014) 13e22.
[21] S. Kenzari, D. Bonina, J. Dubois, V. Fournee, J. Mater. Process. Technol. 214
(2014) 3108e3111.
[22] C. Dong, J.M. Dubois, J. Mater. Sci. 26 (1991) 1647e1654.
[23] R. Manaila, V. Florescu, A. Jianu, A. Badescu, Phys. Status Solidi A109 (1988)
61e66.
[24] N. Kang, P. Coddet, L. Dembinski, H.L. Liao, C. Coddet, J. Alloys Compd. 691
(2017) 316e322.
[25] N. Kang, P. Coddet, H.L. Liao, T. Baur, C. Coddet, Appl. Surf. Sci. 378 (2016)
142e149.
[26] D.J. Sordelet, M.J. Kramer, O. Unal, J. Therm. Spray. Technol. 4 (1995) 235e244.
[27] D.J. Sordelet, M.F. Bresser, I.E. Anderson, J. Therm. Spray. Technol. 5 (1996)
161e174.
[28] C. Dong, Quasicrystal Materials, National Defense Industry Press, Beijing,
1998.
[29] Y. Fu, T. Peng, D. Yang, C. Sun, Y. Chen, Y. Gao, Surf. Coat. Technol. 252 (2014)
29e34.
[30] D. Holland-Moritz, J. Schroers, D.M. Herlach, B. Grushko, K. Urban, Acta Metall.
46 (1998) 1601e1615.
[31] R.H. Morgan, A.J. Papworth, C. Sutcliffe, P. Fox, W. O'neill, J. Mater. Sci. 37
(2002) 3093e3100.
[32] K.W. Dalgarno, C.S. Wright, Powder Metall. Prog. 1 (2001) 70e79.
[33] M. Ma, Z. Wang, M. Gao, X. Zeng, J. Mater. Process. Technol. 215 (2015)
142e150.
[34] P.A. Kobryn, E.H. Moore, S.L. Semiatin, Scr. Mater. 43 (2000) 299e305.
[35] E. King Wayne, D. Barth Holly, M. Castillo Victor, F. Gallegos Gilbert, W. Gibbs
John, E. Hahn Douglas, Chandrika Kamath, M. Rubenchik Alexander, J. Mater.
Process. Technol. 214 (2014) 2915e2925.
[36] E.O. Olakanmi, R.F. Cochrane, K.W. Dalgarno, Prog. Mater. Sci. 74 (2015)
401e477.
[37] S.L. Sing, J. An, W.Y. Yeong, F.E. Wiria, J. Orthop. Res. 34 (2016) 369e385.
[38] Y. Wang, Y. Shen, Z. Wang, J. Yang, N. Liu, W. Huang, Mater. Lett. 64 (2010)
674e676.
[39] L.J. Pawlowski, Therm. Spray. Technol. 8 (1999) 279e295.
[40] D.H. Dai, D.D. Gu, Int. J. Mach. Tool. Manuf. 88 (2015) 95e107.
[41] N. Kang, P. Coddet, H.L. Liao, C. Coddet, J. Alloys Compd. 662 (2016) 259e262.