Ceramics International 42 (2016) 1573915746

Contents lists available at ScienceDirect

Ceramics International
journal homepage: www.elsevier.com/locate/ceramint

Ceramic protection plates brazed to aluminum brake discs

L. Gorjan a,n, M. Boretius b, G. Blugan a, F. Gili c, D. Mangherini c, X. Lizarralde d, M. Ferraris e,
T. Graule a, A. Igartua f, G. Mendoza f, J. Kuebler a
a
Empa Swiss Federal Laboratories for Materials Science and Technology, Laboratory for High Performance Ceramics, berlandstrasse 129, 8600 Dbendorf,
Switzerland
b
Listemann Technology AG, Wirtschaftspark 34, P.O. Box 247, LI-9492 Eschen, Liechtenstein
c
CRF Centro Ricerche Fiat, Strada Torino 50, Orbassano, Italy
d
Fagor Ederlan Edertek, Garaia Berrikuntza Gunea, Isasi Kalea, 6 Posta K. 19, 20500 Arrasate-Mondragn, Gipuzkoa, Spain
e
POLITO Politecnico di Torino, DISAT Department of Applied Science and Technology, Corso Duca degli Abruzzi 24, I-10129 Torino, Italy
f
IK4 TEKNIKER, Parke Teknologikoa, Calle Iaki Goenaga 5, 20600 Eibar, Gipuzkoa, Spain

art ic l e i nf o a b s t r a c t

Article history: Aluminum alloys are light-weight and one of the most interesting material solutions to optimize the
Received 3 May 2016 strength/weight ratio to reduce car weight; however they are also relatively soft and therefore cannot be
Received in revised form used for intensive wear applications. We developed an aluminum alloy part combined with hard and
5 July 2016
wear-resistant Al2O3-based ceramic plates on the surface for demanding mechanical parts for automotive
Accepted 5 July 2016
industry such as disc brakes.
Available online 7 July 2016
Tribological tests of various engineering ceramic materials were performed in order to nd a ceramic
Keywords: material with a combination of coefcient, wear resistance and thermal energy dissipation for the car
Ceramic brakes. Al2O3-based ceramic showed promising properties, as well as being cost effective.
Wear resistant
Two different approaches to braze ceramic on aluminum were investigated. A two-step brazing
Soldering
process using Cu-Sn-Ti-Zr ller alloy and a single step ultrasonic active soldering with Sn-Ag-Ti ller
Brazing
Aluminum brake disc alloy. Larger areas of aluminum could be covered with a segmented brake design in which many ceramic
plates were joined surface. Comparable tribological properties to those of the bulk ceramic material were
achieved.
& 2016 Elsevier Ltd and Techna Group S.r.l. All rights reserved.

1. Introduction Brazing and soldering are preferred joining techniques when a

There is a strong interest in the automotive industry to obtain
lighter components in order to improve energy efciency and
emission output. The amount of aluminum, used in passenger cars
has risen signicantly in the past decades, which resulted in im-
proved energy efciency. It is used in many different components
like wheels, engines, heat exchangers, transmission and others.
This is possible, because aluminum alloys have several good
properties like high strength to weight ratio, excellent thermal
conductivity, corrosion resistance and they are easy to machine
and recycle. However, they are relatively soft and offer insufcient
wear resistance for applications where intensive friction is pre-
sent. In this paper, an idea to combine a tough and light-weight
aluminum body with a hard and wear-resistant ceramic, layer
joined at the surface, is explored and evaluated for applicability to
automotive disc brakes.
n
Corresponding author.
E-mail addresses: lovro.gorjan@empa.ch, lovro.gorjan@gmail.com (L. Gorjan).
strong, thermal conductive, temperature and solvent resistant
joint is required. This approach has an additional advantage re-
garding the surface protection of soft aluminum, because it en-
ables to attach a material which was made in a separate process
to have well-dened and optimized properties for the desired
application.
In the past other ways to improve the aluminum surface have
been explored. Typically the surface is improved by applying hard
and wear resistant coating on the surface. For example ferrous
thermal spray coating [1]. Especially with micro-arc oxidation a
very hard and strong layer of Al2O3 coating can be prepared [25].
Although very good properties were achieved, these promising
technologies are not the focus of this paper.
Brazing and soldering of ceramic to aluminum alloys is not a
well-established and proven technology. There are two major
reasons for that. First, brazing ceramic to aluminum is difcult
because molten ller metal does not spontaneously form a strong
bonding to the surfaces, especially at the relatively low tempera-
tures which are required to join aluminum alloys. Filler alloys with

http://dx.doi.org/10.1016/j.ceramint.2016.07.035
0272-8842/& 2016 Elsevier Ltd and Techna Group S.r.l. All rights reserved.
15740 L. Gorjan et al. / Ceramics International 42 (2016) 1573915746

a so called active element (most commonly Ti or Zr) are used to

braze ceramics. Only moderate bonding strengths (15
30 MPa)
were reported by using hot pressing of Al2O3 to aluminum at
600 C in vacuum using alloys without an active element such as
Al-Si and Al-Cu [6]. When ller alloy contains an active element it
can react with the ceramic surface and forms strong chemical
bonding [717]. The modied surface then enables the wetting of
the molten ller. The addition of rare earth elements, such as Ce
and Ga also improves wettability, while it also improves strength
[18] and the oxidation resistance of the metal in molten state
[19,20]. Relatively high temperature is typically needed for the
activation of the active element. For example a temperature in
excess of 700 C is needed to achieve good wetting of the Sn-Ag-Ti
alloys on Al2O3 ceramic [13,14,20,21]. However, high temperature
presents a problem for joining aluminum alloys since cast alumi-
num alloys have relatively low melting points (o600 C) and their
properties can be compromised at higher temperatures.
The second major challenge in brazing or soldering ceramic to
aluminum originates from the different shrinkages of both mate-
rials after the cooling to ambient temperature. Large thermal Fig. 1. Sketch of an aluminum braze disc protected by wear-resistant hard material.
stresses are generated which can compromise the mechanical
properties of the composite product. In order to avoid this in- Table 1
herent problem a sophisticated graded material structure or spe- List of selected materials used in the research.
cial design in size and shape could potentially avoid the critical
Material Composition Density [g/ CTE Thermal con-
issues.
cm3] [10
6/K] ductivity [W/mK]
For applications like brake discs, a high precision surface at-
ness and parallelism is required, which can most likely be only A380 Al 8.5Si 3.5Cu 2.71 22 96
achieved by nal grinding, after the ceramic parts have already A356 Al 8.5Si 0.3Mg 2.69 22 160
been soldered on the aluminum frame. During such a mechanical Al2O3 99.7% Al2O3 3.93 7.0 27
Al2O3-SiC 99.99% pure 3.84 6.8 24
operation possible defects are introduced on the surface of stres- Al2O3 5 vol% SiC
sed ceramic. An additional important aspect of the research was TZ3Y ZrO2 5% Y2O3 6.1 10 2
also the price of the nal composite. Only cost effective products Mullite 3Al2O3 2SiO2 3.1 5 6
would be acceptable for automotive mass application. SiC SiC 3.2 4 120
ZTA Al2O3 8 vol% TZ3Y 4.2 8 20
In this work the development of composite materials for disc
brakes and possibly other friction intensive applications are pre-
sented. First a material with the desired friction characteristics Table 2
was selected for the outer layer of the composite. Then different List of materials used for joining.
joining techniques were considered to prepare aluminum/ceramic
composite. Material Type Composition [wt%] Melting temperature [C]

GS220 Solder alloy 92Sn 3Ag 3Ti (Ce, Ga) 220
230
Diabraze Braze alloy 74Cu 14.5Sn 10Ti 1.5Zr 868
925
2. Experimental

2.1. Concept design
The idea behind the project was to use an aluminum alloy to
make a lightweight brake disc body and cover the wear area with a
hard and wear-resistant material. The concept is schematically
shown in Fig. 1. Cast aluminum alloy is used for a body onto which
harder and wear resistant ceramic material is brazed.
2.2. Materials
Materials used in our research are listed in Table 1 together
with their basic properties. Two different aluminum alloys were
used - A380 and A356, which are widely used in automotive and
other industries. Solid sample parts were cast at Fagor Ederlan,
Spain, using a low-pressure aluminum casting process.
Several technical ceramic materials were considered for the top
layer. After the tribological tests of different bulk ceramic samples
we focused on Al2O3 ceramic. Material marked as Al2O3 was pre-
pared from commercial 99.7% alumina powder (Granalox NM9922,
Nabaltec, Germany), pressed and sintered at 1620 C for 2 h in air.
Materials used for joining wear resistant layers to aluminum
alloy substrates are listed in Table 2.
GS220 (MBR Electronics, Switzerland) is an active soldering
alloy with the composition of 92Sn3Ag3Ti ( Ce, Ga). It has a
melting range of 220230 C and it contains titanium as an active
element.
Diabraze is Cu-Sn-based active ller metal and was chosen
because of its high strength and ability to wet and react with
materials which are typically difcult to bond by brazing [2224].
The good bonding is due to the large amount of the active ele-
ments Ti and Zr. Diabraze has been already successfully used to
join diamond and boron nitride ceramics with steel substrate
[22,24].
2.3. Joining and machining the samples
Two different approaches to bond wear resistant material on
top of aluminum part were considered and are sketched in Fig. 2.
a) Two-step vacuum brazing was made using the Diabraze
CuSnTiZr braze metal (Fig. 2a). First, the alloy slurry was applied to
the ceramic surface and processed at 900 C in vacuum in order to
melt the ller metal and make a strong bonding with the ceramic
surface. After the end of this metallization step, the samples were
put on the top of aluminum alloy substrate so that the metalli-
zation layer was in direct contact with the aluminum alloy. The
samples were then heated in vacuum to 555 C for 15 min. This
 L. Gorjan et al. / Ceramics International 42 (2016) 1573915746 15741

Fig. 2. Different approaches to apply hard material on the surface of aluminum alloy part by brazing or soldering.

temperature is actually far below the solidus temperature of the
braze alloy (868 C) because we are limited by the melting point of
the aluminum alloy substrate, which has a melting range of 555
615 C. However holding at the solidus temperature of the sub-
strate, the braze alloy reacts with the aluminum, forming a bond.
The joining procedure was developed and made by Listemann
Technology AG, Liechtenstein.
The braze alloy has a relatively high melting point and thus
guarantees temperature resistance of the joint. On the other hand
the process involves relatively high processing temperatures
which causes large thermal stresses upon cooling to room
temperature.
b) Direct soldering was performed using the SnAgTi(Ce, Ga)
solder alloy. Soldering was made with the ultrasonic soldering
system (USS-9510, MBR Electronics, Switzerland) using a 6 mm
diameter tip. Such joining was successfuly used in our recent re-
search on joining ceramic to aluminum alloy [17]. Cleaned samples
(10 min in acetone lled ultrasound bath) were heated to 260 C
and solder alloy was melted on both surfaces, ceramic and alu-
minum. The wetting of the solder was promoted with the sol-
dering tip with 15 W of ultrasound power at 40 kHz. The tem-
perature of the tip was 260 C and the time of ultrasonic assisted
wetting was 15 s. Both parts were then joined together, positioned
and cooled down in air to ambient temperature. Since the sol-
dering is done at a relatively low temperature the thermal stresses
are smaller than in the case of brazing. However, the thermal
resistance of the joint is lower, due to the relatively low melting
point of the solder alloy. To further reduce the thermal stresses we
also explored the so called segmented design, where large surface
of aluminum is covered by multiple smaller pieces of ceramic.
After joining the surfaces were ground by diamond wheel in
order to achieve the proper atness required for the tribological
tests. Final surface roughness Ra was 1.1 mm as measured by a
Hommel tester T500 (Hommelwerke, Germany).
2.4. Characterization
Apparent shear strength was determined by a single-lap ap-
parent shear test sketched in Fig. 3a. Samples were prepared from
6  6  2 Al2O3 ceramic plates joined on 12  12  13 mm alumi-
num blocks. The test was performed with a Zwick Z005 (Zwick,
Germany), universal testing machine with a constant deformation
rate of 0.5 mm/min and with the recording of the applied force.
Apparent shear strength was simply calculated as a maximum
force divided by the joint surface area.
Tribological characterization was carried out with the pin on
disc laboratory scale testing apparatus at Centro Ricerche Fiat
(CRF), Italy as shown in Fig. 3b. The applied force was 200 N and
the rotational speed of sample was 850 rpm. Total time of the test
lasted 900 s. Air cooling system was used to simulate car brakes
during running vehicle. The temperature on the sample surface
was measured with a pyrometer. Two different counterpart pad

Fig. 3. Sketch of an apparent shear test (a) and tribological test (b).
15742 L. Gorjan et al. / Ceramics International 42 (2016) 1573915746
materials were used for the tests; a) a reference pad material ty-
pical for cast iron disc brakes supplied by CRF and b) a harder type
of pad material supplied by Edertek.
3. Results and discussion
3.1. Bulk materials characterization
Tribological tests of bulk materials were used as a rst test to
select the most appropriate candidate materials for the top layer.
Cast iron was also tested as a reference; in combination with soft
pads it is used in the industrial production of car brakes. According
to the industrial experience in CRF the optimal coefcient of fric-
tion should be between 0.35 and 0.55. The surface temperature
during breaking should be as low as possible. The tribological re-
sults of bulk ceramic materials are shown in Fig. 4. The region of
acceptable conditions is marked in Fig. 4. Tribological tests showed
that the soft pads exhibit too low coefcient of friction. Relatively
high temperature was measured at the surface of mullite and
zirconia. This is due to the low thermal conductivity of the two
materials. The heat generated during friction cannot quickly dis-
sipate through the ceramic layer. Hard pads in combination with
Al2O3, Al2O3-SiC and ZTA showed good properties.
According to the tribological results made with bulk materials,
our further research focused on Al2O3 because of the best overall
properties and also relative low price. In combination with hard
pads it showed the most promising tribological behavior for the
brake application. The microstructure of Al2O3 ceramic is shown in
Fig. 5.
3.2. Two-step direct vacuum bonding of ceramic to aluminum
Metallization of the ceramic surface and subsequent joining at
550 C in vacuum resulted in apparently good joints. However,
after metallographic analysis it was revealed, that all ceramic parts
with 20  10 mm  2 mm in size were cracked. A typical micro-
graph of a joint is shown in Fig. 6. A distinct layer ( E50 mm thick)
near the ceramic surface was made of mixed Al and Si2Ti phases.
ZrO2 grains were found at the ceramic surface. The titanium was
spread all over the thin layer, but it did not diffuse into the alu-
minum substrate. However, the layer was almost depleted of
copper which diffused deeply into the substrate, forming a CuAl
phase between the aluminum grains. In some areas the Cu
Fig. 4. Coefcient of friction and surface temperature during tribological tests for
different ceramic materials and iron as a reference. Red rectangular show the re-
gion in which both the coefcient of friction and the temperature are good for the
car brake discs.
Fig. 5. Microstructure of sintered, polished and thermally etched Al2O3 ceramic.
diffusion layer was 2 mm thick. We could not nd any con-
centrated areas of Sn.
Smaller ceramic samples (6 mm  6 mm  2 mm) could be
joined without cracks. Samples were tested for apparent shear
strength evaluation and in all cases the ceramic part did not crack
during the test. The typical fracture surfaces after the test are
shown in Fig. 7. The joint failed quite far away from the ceramic
surface, in the Cu-diffusion zone as can be seen from the side view
of ceramic part after the test. Thick chunks of residual metal were
still attached on the surface.
Cohesive failure inside the reaction zone was identied as a
mechanism of failure. Average apparent shear strength was
42 MPa with a standard deviation of 8 MPa (see Fig. 10). This is
considerably lower than the reported values in the literature for
the yield strength of such an alloy which should be 550 MPa [22].
The failure of the joint actually goes through the Cu-diffusion zone
with unknown properties that also contains large cavities. Cavities
can act as a strength determining aws. Moreover, large residual
thermal stresses after cooling to ambient temperature, can con-
siderably reduce the strength of the alloy by introducing micro-
cracks [25]. All the mentioned factors can explain the low strength,
although the precise contribution of each factor is difcult to
quantify.
However, the results also show that the wetting and the ad-
hesion of ller metal to the ceramic is relatively strong. Optimizing
the process would possibly lead to a diffusion zone with smaller
voids and higher shear strength of the diffusion zone. But the
cracking of large ceramic parts is not expected to be solved just by
optimizing the processing conditions. In order to join larger parts
without cracking a ller alloy with a lower melting point is re-
quired. That is the reason to investigate the joining with Sn-Ag-Ti
active solder with a melting range of 220230 C. It is also well
known that the amplitude of residual stresses signicantly depend
on the dimension of the parts [25,26]. Joining parts with larger
interface areas result in larger residual stresses. Therefore, by
choosing a smaller ceramic piece, the cracking due to thermal
stresses can be potentially avoided.
3.3. Directly joined ceramic to aluminum by ultrasonic soldering
Al2O3 ceramic was successfully joined to the aluminum alloy
substrate by ultrasonic soldering with Sn-Ag-Ti active ller alloy. A
typical microstructure of the cross section is shown in Fig. 8. The
thickness of solder layer varied from 30 to 50 mm for different
samples and no pores or other larger defects were found in the
joint layer. A very sharp and well dened interface between the
ller metal and aluminum substrate with no diffusion region was
observed. Silver-rich components were mostly concentrates at the
 L. Gorjan et al. / Ceramics International 42 (2016) 1573915746 15743

Fig. 6. Low magnication cross section micrograph of Al2O3 joint to aluminum alloy in a two-step vacuum process using CuSnTiZr braze alloy is shown in g. a). Elements
were identied by EDX analysis. Figure b) shows a magnied interface area between ceramic and ller metal.

Fig. 7. A shear fracture surfaces on the aluminum side (a) and on the ceramic side (b). A rough metallic surface left over was found on both sides. The residual metal is also
clearly seen from the side view of ceramic part (c).

solder/aluminum interface. Tin/titanium intermetallic grains were
found through the braze layer.
Single-lap apparent shear strength was measured on samples
with 6  6  2 mm3 Al2O3 plates soldered on a 12  12  13 mm3
aluminum alloy block. Two mechanisms of fracture were ob-
served: a) the cohesive failure of the ller alloy and b) the
detachment of ller alloy from the ceramic surface. Typical shear
fracture surfaces are shown in Fig. 9. Large areas of exposed
ceramic were found; a consequence of the adhesion failure during
the shear test. In other areas residual solder alloy was found on the
ceramic surface. It was conrmed by EDX analysis that the residual
metal is solder alloy and not aluminum alloy which is in
15744 L. Gorjan et al. / Ceramics International 42 (2016) 1573915746
Fig. 8. Micrograph of polished cross section of soldered Al2O3 to Aluminum alloy
sample with. Elements identied by the EDX analysis are shown.
agreement with cohesive failure of the solder mechanism. Average
apparent shear strength was 50 MPa with a standard deviation of
12 MPa (Fig. 10). The large scattering was connected to the scat-
tering of the percentage of cohesive failure area the larger the
area, the higher the strength.
For tribological testing larger parts were required. Attempts to
make a tribological sample with 54 mm diameter single ceramic
disc plate, soldered onto 5 mm thick aluminum alloy disc substrate
failed due to the cracking of the ceramic after cooling to ambient
temperature. Even though the temperature difference between the
solidus of the solder and the ambient temperature is only around
200 C, thermal stresses are large enough to cause cracks in
ceramic (Fig. 11b).
A solution was found in the segmented design where the sur-
face is covered with many ceramic plates. In this way we could
obtain a crack-free 54 mm diameter discs for tribological tests
(Fig. 11b). For additional reliability a design with side support rim
was chosen. The ceramic plated surface was ground with a dia-
mond blade before the tribological testing to achieve proper at-
ness and parallelism for the tribological testing. Surface roughness
parameter Ra after the grinding was 1.1 mm.
Tribological properties of the bulk Al2O3 was compared with
the segmented sample under continuous pin on disc mode,
Fig. 9. Fracture surfaces on the aluminum side (a) and on the ceramic side (b). Mixed failure mechanism was observed. Cohesive failure of solder layer and detachment of
solder from ceramic surface.
showing higher coefcient of friction and a relatively long initial
phase in which coefcient of friction is not stable (Fig. 12). After
analysis of the friction surface of the sample, it was determined
that some edges were slightly sticking out of the at plane. Also
some solder material was spread from the junctions between the
ceramic plates (Fig. 11d). The topological irregularity must have
occurred after the grinding, due to a stress-releasing deformation
in the aluminum alloy substrate. This caused an additional friction
resulting in an increased measured coefcient of friction. During
the tribological test the temperature at the surface reached 140 C,
which is relatively high for a solder with a low-temperature
melting point. Therefore, thermal stresses might have caused de-
formation and slight repositioning of the ceramic plates and con-
sequently the increase in the coefcient of friction.
Another sample (marked as the sample 2) with smoother
transitions between the segments, showed similar friction prop-
erties as those of the monolithic single piece of Al2O3 ceramic. This
demonstrates that similar friction behavior as in the case of bulk
material can be made with the segmented design if the nish of
the surface is properly done. The surface quality can be a critical
issue for the plated design in applications where coefcient of
friction must be precisely dened. No wear of ceramic or change in
surface roughness could be detected; similar surface as produced
by the diamond grinding process remained after the test.
3.4. Demonstrational brake disc
In order to demonstrate the possibility of scaling-up and cov-
ering even larger areas of aluminum surface, a half-scale alumi-
num brake disc covered with Al2O3 ceramic plates was made by
ultrasonic soldering with Sn-Ag-Ti active alloy. The photograph of
the part is shown in Fig. 13.
4. Conclusion
Different technical ceramic materials were investigated as a
wear protection layer for aluminum alloy car brake discs and
possibly other friction intensive wear application. Al2O3 proved to
have good tribological and thermal properties for the application;
however the joining of Al2O3 to aluminum substrate presents
 L. Gorjan et al. / Ceramics International 42 (2016) 1573915746 15745

Fig. 10. Apparent shear strengths of 6 mm  6 mm Al2O3 samples soldered by Sn-Ag-Ti solder alloy or brazed by Cu-Sn-Ti-Zr alloy are presented in a). Bars present average
value and vertical line present standard deviation. Typical stress-strain curves are shown in b). Black lines are for Sn-Ag-Ti soldered samples and grey-dashed lines are for
Cu-Sn-Ti-Zr brazed samples.

major challenges. Two different approaches for joining were
investigated.
Although the two steps vacuum brazing with Diabraze Cu-Sn-
Ti-Zr braze forms a good bonding to the ceramic surface, the
thermal stresses are so large that when brazing a large ceramic (4
ca. 6  6 mm2) part to aluminum, the ceramic is cracked during
the joining process.
Using the lower melting point Sn-Ag-Ti active solder a crack
free bonding can be achieved for ceramic plates with size of
20 mm  20 mm. Larger areas can be covered by multiple plates;

Fig. 11. Photograph of cracked disc after soldering made from one ceramic piece (a) and crack-free disc with segmented ceramic cover soldered on aluminum alloy substrate
(b). The disc after the tribological test is show on photo c) with clearly visible tribo-layer deposited from the friction with the pad. The detail of tribolayer at the edge between
two ceramic plates is shown in gure g).
15746 L. Gorjan et al. / Ceramics International 42 (2016) 1573915746
Fig. 12. Results of the tribological testing with a hard pad against ceramic surface.
Fig. 13. 150 mm diameter aluminum brake disc with segmented Al2O3 plates sol-
dered on the two wear surfaces.
55 mm diameter tribological sample disc and 140 mm diameter
small-scale car brake disc were successfully made by this techni-
que. Tribological tests showed that similar sliding properties can
be achieved with the segmented design as those of the bulk sin-
tered Al2O3. Highly wear resistant aluminum body brake disc can
be made, but for this solution the working temperature must be
kept low, because of low melting point of the ller metal. Fur-
thermore, if application demands very at surfaces with well-de-
ned coefcient of friction a special attention must be taken at
grinding the surface in order to achieve a proper atness and
alignment of individual ceramic plates.
Acknowledgements
The research was done in the scope of JOLIE ERA-NET project
(FP6, MA-TERA/HPE-2227), nanced by the Swiss Commission of
Technology and Innovation (CTI) under contract no. 12128.1 PFIW-
IW and MIUR (Ministero dellIstruzione, dellUniversit e della
Ricerca) Italy under the contract Decreto n. 698/ric17/10/2012, by
the CDTI Spanish (Contract IDI-20120133) and SPRI Basque re-
gional research Agencies.
Website of the Jolie project: http://www.composites.polito.it/?
p projects_european_jolie.
References
[1] A. Edrisy, T. Perry, Y.T. Cheng, A.T. Alpas, Wear of thermal spray deposited low
carbon steel coatings on aluminium alloys, Wear 251 (2001) 10231033.
[2] P.A. Dearnley, J. Gummersbach, H. Weiss, A.A. Ogwu, T.J. Davies, The sliding
wear resistance and frictional characteristics of surface modied aluminium
alloys under extreme pressure, Wear 225 (1999) 127134.
[3] M. Sieber, T. Mehner, D. Dietrich, G. Alisch, D. Nickel, D. Meyer, I. Scharf,
T. Lampke, Wear-resistant coatings on aluminium produced by plasma ano-
dizing a correlation of wear properties, microstructure, phase composition
and distribution, Surf. Coat. Technol. 240 (2014) 96102.
[4] H. Wu, J. Wang, B. Long, Z. Jin, W. Naidan, F. Yu, D. Bi, Ultra-hard ceramic
coating fabricated through microarc oxidation on aluminium alloy, Appl. Surf.
Sci. 252 (2005) 15451552.
[5] G. Sundararajan, L.R. Krishna, Mechanisms underlying the formation of thick
alumina coatings through the MAO coating technology, Surf. Coat. Technol.
167 (2003) 269277.
[6] M. Kobashi, T. Ninomiya, N. Kanetake, T. Choh, Effect of alloying elements in
the brazing sheet on the bonding strength between Al2O3 and aluminum, Scr.
Mater. 34 (1996) 415420.
[7] A.P. Xian, Joining of sialon ceramics by Sn-5 at% TiX ternary active solders, J.
Mater. Sci. 25 (1994) 3946.
[8] W.R. Smith, Process of using an active solder alloy, US Patent number 6047876,
Apr. 11, 2000.
[9] N. Honglong, G. Zhiting, M. Jusheng, H. Fuxiang, Q. Zhiyong, H. Zhongde,
Joining of sapphire and hot pressed Al2O3 using Ag70.5Cu27.5Ti2 brazing ller
metal, Ceram. Int. 29 (2003), 68994.
[10] G.L. Jian, Wetting of ceramic materials by liquid silicon, aluminium and me-
tallic melts containing titanium and other reactive elements, Ceram. Int. 20
(1996) 91412.
[11] R.R. Kapoor, T.W. Eagar, Tin-based reactive solders for ceramic/metal joints,
Metall. Mater. Trans. B 20 (1989) 919924.
[12] X. Chen, J. Yan, S. Ren, Q. Wang, J. Wei, G. Fan, Microstructure, mechanical
properties and bonding mechanism of ultrasonic-assisted brazed joints of SiC
ceramics with ZnAlMg ller metals in air, Ceram. Int. 40 (2014) 683689.
[13] W.R. Smith, Active solder joining of metals ceramics and composites, Weld. J.
10 (2001) 3035.
[14] L. Gremillard, E. Saiz, V.R. Radmilovic, A.P. Tomsia, Role of titanium on the
reactive spreading of the lead-free solders on alumina, J. Mater. Res. 21 (2006)
32223233.
[15] W.L. Wang, Y.C. Tsai, Microstructural characterization and mechanical prop-
erty of active soldering anodized 6061 Al alloy using Sn-3.5Ag-xTi active
solders, Mater. Charact. 68 (2012) 4248.
[16] Y. Tang, G. Li, Thermodynamic study of Sn-Ag-Ti active ller metals, Phys.
Procedia 25 (2012) 3035.
[17] L. Gorjan, G. Blugan, M. Boretius, S. De La Pierre, M. Ferraris, V. Casalegno,
S. Rizzo, T. Graule, J. Kuebler, Fracture behavior of soldered Al2O3 ceramic to
A356 aluminum alloy and resistance of the joint to low temperature exposure,
Mater. Des. 88 (2015) 889896.
[18] C.M.L. Wu, D.Q. Yu, C.M.T. Law, L. Wang, Properties of lead-free solder alloys
with rare earth element additions, Mater. Sci. Eng. R 44 (2004) 144.
[19] R. Kolenak, P. ubor, Soldering of ceramic materials using ultrasonic energy,
Weld. World 49 (2005), 546553.
[20] R. Kolenak, P. ebo, M. Provaznik, M. Kolenakova, K. Ulrich, Shear strength and
wettability of active Sn3.5Ag4Ti(Ce, Ga) solder on Al2O3 ceramics, Mater. Des.
32 (2011) 39974003.
[21] Y.H. Chai, W.P. Weng, T.H. Chuang, Relationship between wettability and in-
terfacial reaction for Sn10Ag4Ti on Al2O3 and SiC substrates, Ceram. Int. 24
(1998) 273279.
[22] S. Buhl, C. Leinenbach, R. Spolenak, K. Wegener, Microstructure, residual
stresses and shear strength of diamond-steel-joints brazed with a Cu-Sn-
based active ller alloy, Int. J. Refract. Met. Hard Mater. 30 (2012) 1624.
[23] A. Rabinkin, A.E. Shapiro, M. Boretius, 6 Brazing of diamonds and cubic
boron nitride, in: D.P. Sekuli (Ed.), Advances in Brazing, Woodhead Pub-
lishing, Cambridge, UK, 2013, pp. 160193.
[24] A.B. Spierings, C. Leinenbach, C. Kenel, K. Wegener, Processing of metal-dia-
mond-composites using selective laser melting, Rapid Prototyp. J. 21 (2015)
130136.
[25] K. Suganuma, Joining ceramic and metal, in: S. Somiya (Ed.), Handbook of
Advanced Ceramics, Elsevier, Amsterdam, 2013,
pp. 775788, http://dx.doi.org/10.1016/B978-0-12-385469-8.01001-7.
[26] K. Sagunama, T. Okamoto T, M. Koizumi, Inuence of shape and size on re-
sidual stress in ceramic/metal joining, J. Mater. Sci. 22 (1987) 35613565.