- Email: [email protected]
Local transformation of amorphous hydrogenated carbon coating induced by high contact pressure
Accepted Manuscript Local transformation of amorphous hydrogenated carbon coating induced by high contact pressure N.K. Fukumasu, C.F. Bernardes, M.A. Ramirez, V.J. Trava-Airoldi, R.M. Souza, I.F. Machado PII:
S0301-679X(18)30193-2
DOI:
10.1016/j.triboint.2018.04.006
Reference:
JTRI 5185
To appear in:
Tribology International
Received Date: 29 December 2017 Revised Date:
31 March 2018
Accepted Date: 7 April 2018
Please cite this article as: Fukumasu NK, Bernardes CF, Ramirez MA, Trava-Airoldi VJ, Souza RM, Machado IF, Local transformation of amorphous hydrogenated carbon coating induced by high contact pressure, Tribology International (2018), doi: 10.1016/j.triboint.2018.04.006. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
2
Local Transformation of Amorphous Hydrogenated Carbon Coating induced by High Contact Pressure
3
N.K. Fukumasua,∗, C.F. Bernardesa , M.A. Ramirezb , V.J. Trava-Airoldib , R.M. Souzaa , I.F. Machadoa
5
Abstract
SC
The development of coatings presenting adaptive characteristics based on mechanical loads would promote improved local characteristics. In this work, the surface modification of an a-C:H coating was analyzed based on the contact pressure using two tribological tests: lubricated reciprocating ball-on-plane and dry micro-scratch tests. Local transformation were evaluated based on Raman spectroscopy and indentation hardness. Results indicated regions presenting a red-shift of the G band and higher indentation hardness. High compressive stress field developed under the contacting region (numerical simulations) coupled to the red-shift of the G band and the increase on indentation hardness inside the tested regions are compatible with the nucleation of sp3 carbon bond sites derived from sp2 bonds,
EP
TE
D
indicating a possible modification of the coating.
AC C
6
Phenomena Laboratory, Polytechnic School of the University of Sao Paulo, Sao Paulo, Brazil b National Institute for Space Research, Brazil
RI PT
a Surface
4
M AN U
1
author. Tel.: +55-11-3091-9865 Email address: [email protected] (N.K. Fukumasu )
∗ Corresponding
Preprint submitted to Elsevier
April 10, 2018
ACCEPTED MANUSCRIPT
7
1. Introduction Automotive and aerospace industries present several examples where an improvement in performance is required
9
to maximize energy efficiency and durability. According to Holmberg et al. [1], on regular passenger vehicles, power
10
generation and transmission systems present total mechanical losses due to friction responsible for approximately 28%
11
of fuel consumption. In this context, several alternatives may be applied to reduce friction losses, such as the use of
12
low friction coatings and improved surface finishing, reducing surface roughness or developing surface topographies
13
to improve lubrication conditions.
RI PT
8
The development and use of low friction coatings became an attractive technology with the use of modern reactors
15
that employ Plasma Enhanced Chemical Vapor Deposition (PECVD) techniques to produce carbon based coatings on
16
regular or complex geometries [2, 3, 4], such as engine taps, helical gears and bearings. High cyclic loads and relative
17
velocities are achieved during the operation of power train components. Those conditions may lead to significant wear
18
and failure of the coatings, decreasing the overall performance of the system. The understanding of coating behavior
19
under extreme conditions is vital to prevent this scenario, and tribological tests are usually conducted before coating
20
the real component. These tests allow exploring issues such as coating/substrate interaction under tribological loads.
M AN U
SC
14
Usually, carbon-based coatings present satisfactory performance in tribological applications, in which high hard-
22
ness and low coefficient of friction (COF) are commonly required [2, 3, 5, 6, 7, 8]. Literature reports an increasing use
23
of diamond like carbon (DLC) coatings, in which tribological properties can be tuned depending on the environment.
24
Under dry sliding condition, DLC coated systems may present a reduction of friction force based on the graphitization
25
of the contacting surfaces, as observed by Liu et al. [9]. This phenomenon is related to the rearrangement of the sp3
26
and sp2 carbon bonds by energy transferred from the mechanical movement to chemical bond kinetics. Polaki et al.
27
[10] reported the influence of high contacting surfaces relative speed of the reduction of COF and wear resistance
28
by the conversion of the sp3 to sp2 bonds. This change on the chemical bonds indicates the formation of a graphitic
29
phase that presents lower COF, due to the lamellar atomic structure, and lower wear resistance associated with lower
30
hardness of this phase.
EP
TE
D
21
Local phase change promoted by localized contact pressure in ductile materials are well reported in the literature
32
[11, 15, 13] and leads to gradients of local microstructure, mechanical and fracture properties. This behavior is not
33
commonly seen for carbon brittle materials, but Gogotsi et al. [14] showed a phase transformation of diamond after
34
an indentation test, in which the deformation of the original material produced nanocrystalline and graphitic carbon
35
localized structures. In terms of carbon-based coatings, Wang et al. [15] showed the transformation of hydrogenated
36
amorphous carbon (a-C:H) coating into graphitic nanocrystallites at the surface of the coating. These nanocrystals
37
were identified as small regions presenting graphite lamellar crystal arrangement, observed using the High Resolution
38
Transmission Electron Microscopy (HRTEM) technique. The formation of these nanostructures was associated to the
39
increase of sp2 carbon bond (G band) peak intensity of the Raman spectrum, observed inside of the deformed region.
40
As presented by Gogotsi et al. [14], Wang et al. [15] and others in the literature [9, 10, 16], the formation
AC C
31
2
ACCEPTED MANUSCRIPT
41
of graphitic structure arrangements from other carbon-based structures is widely known. On the other hand, phase
42
transformation from graphite or amorphous carbon into diamond structures is rare. Usually, stable diamond structures
43
are obtained by Chemical Vapor Deposition (CVD) or thermomechanical systems with high temperature and high
44
pressure (HTHP) techniques [17, 18, 19]. The equilibrium phase diagram of pure carbon systems [20, 30] and experimental observation [21, 22] indicates
46
the possibility of formation of diamond structures for pressures above 10 GPa at room temperature. This process
47
is possible given high hydrostatic pressures, which induces a stable carbon phase with diamond-like microstructure
48
(higher sp3 bonds) and a metastable phase of graphite-like microstructure (sp2 associated bonds).
RI PT
45
For dynamic cases, in which the phase transformation of carbon occurs under continuosly increasing pressures at
50
room temperature was reported by Bai et al. [22]. In that work, para-xylene was subjected to pressures as high as 30
51
GPa at room temperature. The phase transformation from liquid to carbon crystal structures was measured by XRD,
52
Raman and Infrared spectroscopy techniques. Results showed that the increase of the load compacting the para-xylene
53
produced a crystal structure with increasing sp3 (D band) Raman peak proportional to the applied load. Those authors
54
have also indicated that the pressure induced carbon phase transformation was an irreversible process, in which the
55
unloaded sample presented XRD and Raman spectrum peaks not initially presented for the para-xylene liquid.
M AN U
SC
49
Tribological applications usually present elevated contacting loads, resulting in contact pressures on the order of
57
tenths of GPa. Then, the use of a carbon-based coating presenting a dynamic mechanical behavior would promote
58
improved wear resistance while maintaining lower coeficient of friction. In this work, Raman spectroscopy and nano-
59
indentation measurements were conducted to monitor the structure and phase transformations of the materials inside
60
the wear track of high-contact-load tribological analyses. The system under study was composed of an a-C:H coating
61
deposited onto a ductile substrate subjected to high mechanical contact pressures, on the order of 2 GPa.
62
2. Coating deposition and characterization
EP
TE
D
56
To minimize the influence of substrate on the analyses of the coating, the substrate was made of AISI H13 steel,
64
which was water quenched from austenitizing temperature of 1000o C and tempered at 200o C. Prior to the deposition,
65
substrate surface was polished up to 1 µm (20 nm average roughness).
AC C
63
66
The coating and the interlayer in this work were deposited using a pulsed Direct Current Plasma Enhanced Chem-
67
ical Vapor Deposition (DC PECVD) reactor modified to include an additional cathode. This arrangement confines the
68
plasma near the surface of the samples, improving the deposition rate, the quality of the coating and allows a lower
69
deposition temperature (under 100o C). Details of this reactor and the deposition parameters were presented on the
70
literature [2, 23, 24].
71
A silicon interlayer was deposited to improve the adhesion of the carbon-based coating and the substrate. Silane
72
was used to grow this interlayer with 3 sccm gas flow, a bias of -950 V and a working pressure of 0.3 Pa. After this
73
process, the coating was grown using acetylene as precursor with an 8 sccm gas flow, at a pressure of 0.1 Pa and a 3
ACCEPTED MANUSCRIPT
74
constant bias of -800 V. According to Lugo et al. [2], these deposition parameters resulted in a silicon interlayer and a carbon-based
76
coating with, approximately, 500 nm and 1000 nm in thickness, respectively. The composition of the coating was
77
measured by Elastic Recoil Detection Analysis (ERDA) [2] as being 25 wt% of hydrogen and 75 wt% of carbon for
78
the first 500 nm and similar deposition results were obtained by Capote et al. [23, 24].
RI PT
75
81
1800 grating per mm difraction mesh, a 50 µm slit and a 100x objective lens, allowing measurement of D and G bands
82
of the Raman spectra in regions up to 1 µm in diameter with a spectral resolution of 2 cm−1 . All acquired Raman
83
spectra were processed in the Origin v9.5 by Origin Labs Corp., using a Gaussian Non-linear curve fit with a R2
84
higher than 0.95 to deconvolute the D and G shift positions.
SC
80
The characterization of the carbon bonds and the ID /IG intensity ratio at the surface was obtained by using a R confocal Horiba Xplora OneT M Raman system. This system was configured with a 532 nm laser wavelenght, a
79
Mechanical properties of the a-C:H coating were evaluated using the Hysitron Ti950 triboindenter. Several nano-
86
indentation curves were measured to estimate both hardness and elastic modulus, using the Oliver and Pharr method
87
[25]. Load to penetration depth curves were acquired using a Berkovich tip and a maximum load of 700 µN. This
88
load assured the measurement of coating properties with no significant influence of the substrate, once the maximum
89
penetration depth was kept under 50 nm. Additionally, critical loads for coating failure were measured using a cono-
90
spherical tip with 10 µm in diameter and maximum normal load of 360 mN. These critical loads were characterized
91
as the first pop-in in the load-penetration depth indentation curves and are directly linked to the failure of the coating
92
[26]. Also, this equipment can be used in a Scanning Probe Microscopy (SPM) configuration, producing a topography
93
image of selected regions. In this configuration, the Berkovich tip was used with a normal load of 2 µN to acquire
94
images of 10 µm x 10 µm.
95
3. Tribological analyses
EP
TE
D
M AN U
85
Components subjected to pure sliding conditions and elevated contacting loads may present high wear and friction.
97
One way to prevent such critical tribological conditions is the use of protective coatings. In this work, the tribological
98
behavior of a hydrogenated amorphous carbon coating was analyzed based on two sets of experiments: macroscale
99
lubricated reciprocating tests and microscale dry scratch tests.
AC C
96
100
R Macroscale lubricated reciprocating tests were conducted on an Optimol SRV v4 device with a ball on plane
101
configuration (Figure 1), considering that both the plane and the sphere were coated. Tests were based on a step-
102
increase in load from 5 N to 150 N with 5 minutes in each step (total duration of 30 minutes), 25 Hz of reciprocating
103
frequency and a stroke of 2 mm. A droplet of additive free Poly-Alpha-Olefin (PAO) synthetic lubricant in a mixture
104
of PAO4/6 was used in the contact region. The tested spheres were made of AISI 52100 material, coated in a similar
105
process used for the substrate.
106
Microscale scratch tests were carried on a Hysitron Ti950 triboindenter using a cono-spherical tip with 10 µm 4
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 1: Macroscale reciprocating test configuration analyzed in this work, in which both sphere and disk were coated.
in diameter and loads ranging from 25 mN to 250 mN. The total scratch length was 500 µm and constant load
108
configuration were used. In microscale analyses, no liquid lubricant was used (dry condition).
109
4. Finite element analysis
M AN U
107
The microscale scratch test was also analyzed using the TriboCODE numerical package, which is in development
111
at the “Laborat´orio de Fenˆomenos de Superf´ıcie” (LFS) of the University of S˜ao Paulo, Brazil. This package is based
112
on several modules, which are used to simulate and analyze microscale tribological phenomena presenting complex
113
material microstructural configurations and physical mechanisms, such as abrasion, adhesion and thermal fatigue. In
D
110
116
The numerical model was solved using an explicit time evolving algorithm in quasi-static regime, in which no time
117
derivatives nor inertial forces were considered.
EP
TE
115
this work, the scratch test was analyzed as a tridimensional numerical model using three TriboCODE modules (TC R Abrasion, TC Bulk Material Properties and TC Coatings), coupled to the ABAQUS finite element environment.
114
The geometry of the numerical model is presented in Figure 2. In this figure, the configuration consists of the
119
cono-spherical tip (dark gray), the coating (red) and the substrate (light gray). The coating thickness was 2 µm and
120
the substrate was defined as a parallelepiped of 100 µm x 50 µm x 10 µm. The scratch length was 50 µm and the
121
normal load was increased continuously from 0 to 170 mN along the track. No lubricant was considered and two
122
failure models were employed: cohesive zone model for the coating/substrate interface and brittle failure model for
123
the coating. The substrate was considered elastic-perfectly plastic with mechanical properties representative of AISI
124
H13 steel. Mechanical properties of the coating were obtained from the experimental portion of this work (Section
125
2). The failure modeling and property measurements were analyzed following the procedure presented by Fukumasu
126
et al. [27].
AC C
118
5
SC
RI PT
ACCEPTED MANUSCRIPT
M AN U
Figure 2: Numerical model configuration consisting of a cono-spherical tip used to scratch the coated (red) substrate (light gray)
a)
D
b)
Figure 3: Coefficient of friction in the reciprocating tests: a) evolution during the ramping load and b) along the stroke for the last
TE
reciprocating movement of the sphere.
5. Results and discussions
128
5.1. Macroscale reciprocating test results
EP
127
Figure 3 presents the evolution of the coefficient of friction (COF) during the reciprocating tests. The COF curves
130
are presented based on the average of 3 repetition tests. Figure 3a indicates a stabilized value after 50 s and a minimum
131
COF of of 0.091±0.002 achieved at 350 s. This stabilized value of COF is compatible with peak values observed in
132
the reversion position of the reciprocating movement, presented in Figure 3b. In this figure, both reversal points of
133
the sphere movement present a surface to surface contact condition with no lubricant due to the zero velocity. Also, a
134
COF reduction to 0.080±0.001 at the middle of the stroke is observed. This reduction is related to a better lubrication
135
condition compared to the reversal points, given by the occurrence of the maximum relative velocity between the
136
contacting surfaces during the reciprocating cycle.
AC C
129
137
Figure 4 shows a SEM image of the macroscale wear track at the middle of the stroke, after the reciprocating test.
138
This image is representative of the tests and indicates that the coating was well adhered to the substrate, except for
139
isolated coating failures as indicated inside the marked region. 6
SC
RI PT
ACCEPTED MANUSCRIPT
M AN U
Figure 4: Representative wear track after the reciprocating test. The traced red rectangle indicates the region analyzed near the failure of the coating
Raman spectroscopy analyses inside the wear track indicated spatial variations on the G band, which may be
141
related to the fraction of sp2 bonds in the carbon coating. To analyze the change in the Raman shift of the G band, an
142
intensity ratio between two spectral ranges IGr /IGb was analysed. The IGr was defined in the spectral range from 1500
143
cm−1 to 1550 cm−1 , while IGb was defined from 1500 cm−1 to 1600 cm−1 . This selection allows the indication of the G
144
peak position inside the IGr spectrum range for higher values of IGr /IGb , while lower values indicate a G peak in 1550
145
cm−1 to 1600 cm−1 range.
TE
D
140
Figure 5 presents the IGr /IGb ratio image superimposed to the image of a failed region of the coating. In this figure,
147
a red-shift of the G band (higher IGr /IGb ) was observed at the coating surface inside and at the borders of the failed
148
region. The spatial variation of the non-processed G band peak position along the wear track suggests load-induced
149
surface changes at the coating during the reciprocating test. The deconvolution of the Raman spectrum are presented
150
in this figure. For positions presenting higher IGr /IGb ratio (red regions), the G peak position was 1534.4±2.6 cm−1 and
151
a FWHMG value was 197.9±7.8 cm−1 while, for regions presenting lower IGr /IGb (blue regions), the G peak position
152
was 1552.3±1.8 cm−1 and FWHMG was 175.4±8.7 cm−1 .
AC C
EP
146
153
Nano-indentation measurements (Figure 6a), with the corresponding calculation of the hardness and reduced
154
elastic modulus (Figure 6b), were applied to analyze the coating surface in regions presenting variations on the G
155
band position. In this figure, gray color represents measured regions presenting low values of reduced elastic modulus
156
and hardness, which may be a result of coating adhesive or cohesive failure. Nevertheless, Red and Black color
157
measurements presented similar reduced elastic modulus, in the range of 200 GPa to 230 GPa. Black squares in Figure
158
6b represent the original condition (outside the wear track) of the coating, while red measurements are representative
159
of the worn surface presenting a higher IGr /IGb ratio. Comparing both results, the worn surface presented higher 7
EP
TE
D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
Figure 5: Raman shift change of G peak position by mapping the intensity ratio IGr /IGb near the failed coating region: spatial variation of the non-processed G band peak position inside the wear track. In the graphs, gray squares indicate typical spectra obtained for similar regions, while lines indicate the deconvolution of the Raman spectra into D (blue) and G (red) bands.
8
ACCEPTED MANUSCRIPT
7 0 0
In s id e
5 0 0
O u ts id e 4 0 0 3 0 0
In s id e w ith in te r f a c e f a ilu r e
2 0 0 1 0 0 0 0
5
1 0
1 5
2 0
2 5
3 0
3 5
4 0
P e n e tr a tio n D e p th ( n m )
4 5
b)
SC
a)
RI PT
I n d e n t a t i o n L o a d ( µm )
6 0 0
Figure 6: Nano-indentation measurements of the coating: a) nano-indentation curves for inside (red and gray lines) and outside (black lines) of the wear track; b) results for hardness and reduced elastic modulus of the coating for inside (circles) and out-
M AN U
side (squares) of the wear track. Red circles indicate similar reduced elastic modulus but higher hardness compared to outside measurements (black squares), while gray circles indicate a reduction on both characteristics.
160
hardness (in the range of 25 GPa to 40 GPa), suggesting a change in the mechanical behavior that could be associated
161
to the Raman red-shift of the G band. For comparisson, the measured indentation hardness of the substrate was,
162
appoximately, 6.2±0.2 GPa.
163
As reported in the next section, dry microscale scratch tests reproduced the G band peak position variation, indicating no contribution from the additive free lubricant on the modification of the worn coating surface.
165
5.2. Microscale scratch test results
TE
D
164
The groove marks, produced inside the wear track of the macroscale reciprocating test, present similarities to
167
the deformation induced by a single abrasive particle scratching the coating surface. Therefore, microscale scratch
168
tests could be able to reproduce the groove formation, inducing localized changes of hardness and G band position
169
promoted by the abrasive particle in contact with the coating surface.
EP
166
In this test, the influence of the normal load on the coefficient of friction is displayed on Figure 7. Load values up
171
to 100 mN produce similar COF of, approximately, 0.05. Increasing the load led to an increase of the COF, in which
172
the intermediary load of 150 mN produced a COF of 0.08 while the highest analyzed load (250 mN) led to a COF of
173
0.15. In all cases, the tangential force was stable along the entire scratch distance.
AC C
170
174
Figure 8 shows a representative portion of the wear track obtained with the intermediary load case (150 mN
175
case). In this figure, the coating presents three failure modes: adhesive with small cohesive failures of the coating (a),
176
complete spallation of the coating (b) and adhesive failure of the coating without visible cohesive failure (c).
177
The measured map of the IGr /IGb ratio is presented in Figure 9, superimposed to an optical image of the microscale
178
scratch track, in a region presenting no spallation of the coating (Figure 8). In this figure, the higher intensity of
179
IGr /IGb (red colored regions) can be directly associated to regions in which the coating presented some type surface 9
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
Figure 7: Coefficient of friction along the scratch length during the test for analyzed normal loads
Figure 8: Microscale scratch track presenting the local typical observed failure modes: a) adhesive and cohesive failures of the coating; b) complete spallation of the coating and c) adhesive failure of the coating/substrate interface.
10
TE
D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
EP
Figure 9: Superimposed Raman spectroscopy map of IGr /IGb ratio on the scratch track of figure 8. Higher ratio values (red colored regions) indicate a red-shift of the G band.
modification. This localized increase on the IGr /IGb ratio indicates a red-shift of the G band position induced by the
181
deformation of the coating, subjected to high localized contact stresses.
AC C
180
182
Figure 10 shows typical Raman spectra for point measurements of the coating surface inside and outside of the
183
scratched region. Also, this figure presents the deconvolution of the Raman spectrum inside and outside the scratch
184
track. A total of 10 point measurements in each region were analyzed, leading to a deconvoluted G band peak position
185
of 1544.7 ± 1.6 cm−1 inside and 1557.1 ± 0.7 cm−1 outside the scratched region. Also, the inside region presented a
186
full width at half maximum (FWHMG ) value of 183.9 ± 4.1 cm−1 while the outside region presented a value of 171.4
187
± 1.4 cm−1 .
188
Figure 11 was acquired using the Scanning Probe Microscopy technique, in which the indentation tip scans the 11
In s id e
In te n s ity ( a .u .)
-1
M AN U
1 3 5 5 c m
1 5 5 7 c m
1 3 7 3 c m
TE
1 0 0 0
1 2 0 0
EP
8 0 0
-1
-1
D
O u ts id e
-1
SC
1 5 4 4 c m
RI PT
ACCEPTED MANUSCRIPT
1 4 0 0
R a m a n s h ift ( c m
1 6 0 0 -1
1 8 0 0
2 0 0 0
)
AC C
Figure 10: Raman spectroscopy analysis of a-C:H coatings after the scratch test. Gray squares indicate typical spectra obtained for inside and outside the scratched regions, while lines indicate the deconvolution of the Raman spectra into D (blue) and G (red) bands.
12
ACCEPTED MANUSCRIPT
1 0 0
H e ig h t ( n m )
-1 0 -2 0 -3 0
-5 0 -6 0 0
a)
b)
2
RI PT
-4 0
4
6
8
1 0
L a te r a l D is ta n c e ( µ m )
Figure 11: Scanning Probe Microscopy of the scratch track presenting cohesive failure of the coating: a) planar view indicating the indentation analyses markers (black circles) and profile measurement across the track (green dotted line); b) measurement of
M AN U
SC
surface topography across the scratch track.
D
Figure 12: Indentation hardness measurements at positions indicated on Figure 11a and outside the scratch track (original condi-
TE
tion).
surface to obtain an image of the topography of the analyzed region. Figure 11a shows the top view of the wear
190
track. In this figure, three cohesive cracks at the coating surface are observed and recognized as the curved localized
191
variation on the topography. Figure 11b shows the depth profile of the scanned region, indicated in Figure 11a by the
192
green dotted line. The maximum residual depth of the scratched region was 60 nm or 6% of a-C:H coating thickness.
193
Also, hardness measurements (Figure 12) were taken at three points between two cohesive cracks of Figure 11a, as
194
indicated by points 1, 2 and 3.
AC C
EP
189
195
Figure 12 indicates higher hardness for the three measured points on the wear track compared to the initial values
196
of the coating. The difference on hardness measured for point 1 compared to points 2 and 3 may be related to the
197
proximity to the crack, in which the stifness of the coating may be reduced by the coating failure. Nevertheless,
198
these results suggest an increase of coating hardness coupled to an increase of IGr /IGb ratio, both promoted by the
199
deformation of the system induced by the scratch test, similar to the results obtained inside the grooves of macroscale
200
reciprocating tests.
201
Literature indicates that for amorphous carbon coatings subjected to elevated temperatures (annealing or rubbing 13
a)
RI PT
ACCEPTED MANUSCRIPT
b)
indentation curve and b) first pop-in presented on the loading curve.
SC
Figure 13: Coating cohesive failure critical loads measured by nano-indentation technique: a) typical load/penetration depth
processes), the sp3 carbon bonds tend to be converted to sp2 bonds and a blue-shift of the G band is observed [28, 29].
203
In this work, the deconvolution of the Raman spectrum indicated an opposite trend, in which a red-shift of the G band
204
and an increase of the FWHMG value inside the scratched region were observed. According to Ferrari and Robertson
205
[29], this red-shift is linked to an increase of sp3 content for an a-C:H coating and an increase in hardness. Also, the
206
increase of the FWHMG indicate a decrease of the in-plane correlation length (La ) and an increase of carbon disorder,
207
which is an indicative of the nucleation of small sp3 sites derived from sp2 bonds [29].
M AN U
202
One possible cause for this coating modification would be the high contact pressure developed during the scratch
209
test. To better evaluate the contact conditions, a finite element model was developed to simulate the coating/substrate
210
mechanical behavior under these test conditions.
211
5.3. Finite element analysis of the scratch test
212
5.3.1. Mechanical and fracture properties for the numerical simulations
EP
TE
D
208
The initial condition of Young’s modulus of the coating is obtained using the reduced elastic modulus (black
214
square symbols) plotted in Figure 6. The critical load for coating cohesive failure was obtained by nano-indentation
215
technique using a cono-spherical tip. A typical load/penetration depth curve for this test is presented on Figure 13a,
216
in which pop-ins are related to the cohesive failure of the coating, as previously described by Li [26] and considered
217
by Fukumasu [27]. According to those authors, the critical load for coating failure is assumed as the load in which
218
the first pop-in occurs. Figure 13b, which presents a series of pop-ins observed during the indentation of the coated
219
sample by the cono-spherical tip, indicates that the first pop-in occurs in a range from 139 mN to 158 mN normal load,
220
presenting an average value for this coating as 148±5 mN. This value is compatible with the scratch test transitioning
221
load, in which an increase in the COF and the coating failure were observed.
AC C
213
222
Data from Figure 13b were used to evaluate mode I failure stress and fracture energy for the analyzed coating
223
using the TC Coatings module of the TriboCODE. The critical failure stress was calculated as 2.5 GPa and the fracture 14
ACCEPTED MANUSCRIPT
226
used to model cohesive failure of the coating in this work.
227
5.3.2. Numerical scratch test results
RI PT
225
energy as 0.0002 J/µm. These values coupled to the failure strain of 0.0025, which is associated to the characteristic R mesh size, are the input parameters for the brittle failure model [31] implemented in the Abaqus environment and
224
Numerical simulations of the microscale scratch test provided the contact pressure and stress distributions of the
229
coating/substrate system. Figure 14 shows the contact pressure before cohesive or adhesive failure of the coating.
230
In this figure, the maximum contact pressure was calculated as 12 GPa distributed in a half moon geometry towards
231
the direction of the movement of the indenter (Figures 14a and 14b). The numerical simulation reproduces the
232
experimental pressure distribution given the similarities to the geometry of the spalled region of the coating, indicated
233
by the failure mode b presented in Figure 8.
SC
228
Figure 14c shows the evolution of the contact pressure with the increase of the normal load. In this figure, the
235
maximum contact pressure (12 GPa) is achieved for a load of 100 mN and is kept constant for higher loads, promoted
236
by the plastic deformation of the substrate. Once the coating damage threshold is achieved, cohesive cracks nucleate
237
at the surface of the coating and the contact pressure starts to oscilate, as noted in this figure. The numerical simulation
238
predicts the nucleation of cohesive cracks that promote the oscilations of the contact pressure at a scratch distance of
239
about 40 µm or 140 mN (Figure 14c), compatible with experimental critical normal load for the failure of the coating.
240
Figure 15 presents the distribution of the minimum principal stresses in the coating/substrate system. The high
241
contact pressure developed at the coating surface promoted an elevated internal stress distribution in the coating,
242
achieving compressive values of 20 GPa near the surface and decreasing to 12 GPa at coating/substrate interface.
D
M AN U
234
As reported by Bai et al. [22], this high level of compressive stresses developed in the coating is enough to induce
244
a phase transformation of the carbon material at room temperature, in which compressive stresses on the order of 20
245
GPa or higher promotes the rearrangement of carbon bonds, becoming sp3 dominant (red-shift of the G band, higher
246
FWHMG values and higher hardness).
EP
TE
243
After the removal of the load, the induced carbon transformation inside the wear grooves and the scratch track
248
remains stable. This induced phase was observed by the red-shift of the G band (Figures 5 and 10) and by higher
249
hardness measurements (Figures 6 and 12) for both macro and microscale analyzed tracks. This phase stability was
250
also reported by Bai et al. after the removal of the load, in which a solid phase of the carbon material was observed,
251
presenting crystalline material identified by the X-Ray diffraction technique.
AC C
247
15
RI PT
ACCEPTED MANUSCRIPT
M AN U
SC
a)
b)
c)
D
Figure 14: Contact Pressure at the coating promoted by the indenter movement: a) Instantaneous spatial distribution of the contact
AC C
EP
TE
pressure; b) detail of the contact region and c) evolution of the contact pressure with the ramping load during the scratch test.
Figure 15: Minimum Principal Stresses distribution developed inside of the material during the scratch test.
16
ACCEPTED MANUSCRIPT
252
6. Conclusions Coatings that present adaptive features, such as microstructure and mechanical properties, have the ability to
254
improve the performance of components by tuning specific characteristics during the process. In this work, local
255
transformation of an amorphous hydrogenated carbon coating (a-C:H) was analyzed after high load lubricated re-
256
ciprocating ball-on-plane tests and dry micro-scratch tests using Raman spectroscopy and indentation hardness, in
257
which:
260
261
262
263
264
265
regions under high contacting loads on both test configurations;
SC
259
• the analyzed coating presented low coeficient of friction, good adherence to the substrate and small spalled
• Raman spectroscopy analyses of the macroscale wear track indicated regions presenting a red-shift of the G band compared to the coating original condition (outside the track);
• indentation hardness analyses indicated a spatial correlation of higher hardness on regions with higher IGr /IGb
M AN U
258
RI PT
253
intensity ratio;
• microscale dry scratch tests presented similar Raman red-shift compared to macroscale reciprocating tests, indicating a negligible significant contribution from the additive free lubricant on surface modification; • good agreement was obtained between experimental and numerical analysis of coating critical load for first
267
failure. This result was enhanced by the good measurement of both mechanical and fracture properties of the
268
coating, used as input to the numerical solver;
271
272
TE
270
• the experimental spalled region shape, observed on the transitional load of the scratch test, was compatible with the spatial distribution of contact stresses computed by the numerical simulation; • numerical simulation indicates high contact pressure (>12 GPa) developed at the surface and high internal
EP
269
D
266
stresses, raging from 20 GPa to 12 GPa, are developed along coating thickness; • high internal stresses, red-shift of the G band, higher values of FWHMG and the increase on indentation hardness
274
inside the scratched region are compatible with the nucleation of sp3 carbon bond sites derived from sp2 bonds,
275
according to the literature.
276
AC C
273
Acknowledgements
277
The authors recognize the National Council for Scientific and Technological Development (CNPq), the Sao Paulo
278
Research Foundation (FAPESP) and the National Bank for Economic and Social Development (BNDES) for financial
279
support.
17
ACCEPTED MANUSCRIPT
280
References [1] K. Holmberg, P. Andesson, A. Erdemir, Global Energy consumption due to friction in passenger cars. Tribol. Int., 47, 221-234 (2012).
282
[2] D.C. Lugo, P.C. Silva, M.A. Ramirez, E.J.D.M. Pillaca, C.L. Rodrigues, N.K. Fukumasu, E.J. Corat, M.H. Tabacniks, V.J. Trava-Airoldi,
283
Characterization and tribologic study in high vacuum of hydrogenated DLC films deposited using pulsed DC PECVD system for space
284
applications, Surf. Coat. Technol., 332, 135-141 (2017).
287 288 289 290 291 292 293 294
[3] J. Solis, H. Zhao, C. Wang, J.A. Verduzco, A.S. Bueno, A. Neville, Tribological performance of an H-DLC coating prepared by PECVD, Appl. Surf. Sci., 383, 222-232 (2016).
[4] D. Lusk, M. Gore, W. Boardman, T. Casserly, K. Boinapally, M. Oppus, D. Upadhyaya, A. Tudhope, M.Gupta, Y. Cao, S.Lapp, Thick DLC films deposited by PECVD on the internal surface of cylindrical substrates, Diam. Relat. Mat., 17, 1613-1621 (2008).
[5] A. Erdemir, C. Donnet, Tribology of diamond-like carbon films: recent progress and future prospects, J. Phys. D-Appl. Phys., 39, 311-327 (2006).
SC
286
[6] D. Sheeja, B.K. Tay, S.M. Krishnan, L.N. Nung, Tribological characterization of diamond-like carbon (DLC) coatings sliding against DLC coatings, Diam. Relat. Mat., 12, 1389-1395 (2003).
[7] H.A. Tasdemir, T. Tokoroyama, H. Kousaka, N. Umehara, Y. Mabuchi, Friction and wear performance of boundary-lubricated DLC/DLC contacts in synthetic base oil, Procedia Engineering, 68, 518-524 (2013).
M AN U
285
RI PT
281
295
[8] S.J. Bull, Tribology of carbon coatings: DLC, diamond and beyond, Diam. Relat. Mat., 4, 827-836 (1995).
296
[9] Y. Liu, A. Erdemir, E.I. Meletis, An investigation of the relationship between graphitization and frictional behavior of DLC coatings, Surf.
299 300 301 302 303 304 305
[10] S.R. Polaki, N. Kumar, K. Ganesan, K. Madapu, A. Bahuguna, M. Kamruddin, S. Dash, A.K. Tyagi, Tribological behavior of hydrogenated DLC film: Chemical and physical transformations at nano-scale, Wear 339-339, 105-113 (2015). [11] F.P. Marques, C. Scandian, A.C. Bozzi, N.K. Fukumasu, A.P. Tschiptschin, Formation of a nanocrystalline recrystallized layer during microabrasive wear of a cobalt-chromium based alloy (Co-30Cr-19Fe), Tribol. Int., 116, 105-112 (2017). [12] S.J. Wang, H. Wang, K. Du, W. Zhang, M.L. Sui, S.X. Mao, Deformation-induced structural transition in body-centred cubic molybdenum,
D
298
Coat. Technol., 86-87, 564-568 (1996).
Nat. Commun., 5, 3433 (2014).
[13] A. Das, Contribution of deformation-induced martensite to fracture appearance of austenitic stainless steel, Mater. Sci. Technol., 32, 1366-
TE
297
1373 (2016).
306
[14] Y.G. Gogotsi, A. Kailer, K.G. Nickel, Pressure-induced phase transformations in diamond, J. Appl. Phys., 84, 1299-1304 (1998).
307
[15] D.S. Wang, S.Y. Chang, Y.C. Huang, J.B. Wu, H.J. Lai, M.S. Leu, Nanoscopic observations of stress-induced formation of graphitic nanocrystallites at amorphous carbon surfaces, Carbon, 74, 302-311 (2014).
EP
308
[16] Y.G. Gogotsi, A. Kailer, K.G. Nickel, Transformation of diamond to graphite, Nature 401, 663-664 (1999).
310
[17] B.V. Spitsyn, L.L. Bouilov, B.V. Derjaguin, Vapor growth of diamond on diamond and other surfaces, J. Cryst. Growth, 52, 219-226 (1981).
311
[18] S. Matsumoto, Y. Sato, M. Tsutsumi, N. Setaka, Growth of diamond particles from methane-hydrogen gas, J. Mater. Sci., 17, 3106-3112
312
AC C
309
(1982).
313
[19] H. Kanda, P. Capper, Bulk Crystal Growth in Electronic, Optical and Optoelectronic Materials, John Wiley and Sons, (2005).
314
[20] Q. Jiang, J.C. Li, G. Wilde, The size dependence of the diamond-graphite transition, J. Phys.-Condes. Matter, 12, 5623-5627 (2000).
315
[21] T.B. Shiell, D.G. McCulloch, J.E. Bradby, B. Haberl, R. Boehler, D.R. McKenzie, Nanocrystalline hexagonal diamond formed from glassy
316 317 318 319 320 321 322
carbon, Sci. Rep., 6, 37232 (2016).
[22] Y. Bai, Z. Yu, R. Liu, N. Li, S. Yan, K. Yang, B. Liu, D. Wei, Lin Wang, Pressure-Induced Crystallization and Phase Transformation of Para-xylene, Sci. Rep., 7, 5321 (2017). [23] G. Capote, G.C. Mastrapa, V.J. Trava-Airoldi. Influence of acetylene precursor diluted with argon on the microstructure and the mechanical and tribological properties of a-C:H films deposited via the modified pulsed-DC PECVD method, Thin Solid Films, 589, 286-291 (2015). [24] G. Capote, M.A. Ramirez, P.C. Silva, D.C. Gonc¸alez, V.J. Trava-Airoldi. Improvement of the properties and the adherence of DLC coatings deposited using a modified pulsed-DC PECVD technique and an additional cathode. Surf. Coat. Technol., 308, 70-79 (2016).
18
ACCEPTED MANUSCRIPT
323 324
[25] W.C. Oliver, G.M. Pharr, An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments, J. Matter. Res., 7, 1564-1583 (1992).
325
[26] X. Li, B. Bhushan, Measurement of fracture toughness of ultra-thin amorphous carbon films, Thin Solid Films 315, 214-221 (1998).
326
[27] N.K. Fukumasu, R.M. Souza, Numerical evaluation of cohesive and adhesive failure modes during the indentation of coated systems with
327
compliant substrates, Surf. Coat. Technol. 260, 266-271 (2014). [28] J. Robertson, Mechanical properties and structure of diamond-like carbon, Diam. Relat. Mat., 1, 397-406 (1992).
329
[29] A.C. Ferrari, J. Robertson, Raman spectroscopy of amorphous, nanostructured, diamond-like carbon, and nanodiamond, Phil. Trans. R. Soc.
SC
Finite Elements, Cem. Concr. Res., 6, 773-782 (1976).
M AN U
334
(1993).
[31] A. Hillerborg, M. Modeer, P.E. Petersson, Analysis of Crack Formation and Crack Growth in Concrete by Means of Fracture Mechanics and
D
333
TE
332
EP
331
¨ ı2512 (2004). Lond. A, 362, 2477 A` ’ [30] H.O. Pierson, Handbook of carbon, graphite, diamond and fullerenes: properties, processing and applications, Noyes Publication, New Jersey
AC C
330
RI PT
328
19
ACCEPTED MANUSCRIPT
• High contact pressure tribology measurements of Diamond like Carbon coatings
336
• High correlation of higher indentation hardness on regions with increased sp3 carbon bonds inside wear track
337
• Damage calibrated model to reproduce realistic cohesive and adhesive coating failure by FEM model
338
• Contact pressures and internal coating stresses were induced in the range of equilibrium formation of diamond-
EP
TE
D
M AN U
SC
like carbon phase
AC C
339
RI PT
335
20
ACCEPTED MANUSCRIPT
• High contact pressure tribology measurements of Diamond like Carbon coatings
316
• High correlation of higher indentation hardness on regions with increased sp3 carbon bonds inside wear track
317
• Damage calibrated model to reproduce realistic cohesive and adhesive coating failure by FEM model
318
• Contact pressures and internal coating stresses were induced in the range of equilibrium formation of diamond-
EP
TE D
M AN U
SC
like carbon phase
AC C
319
RI PT
315
18
S0301-679X(18)30193-2
DOI:
10.1016/j.triboint.2018.04.006
Reference:
JTRI 5185
To appear in:
Tribology International
Received Date: 29 December 2017 Revised Date:
31 March 2018
Accepted Date: 7 April 2018
Please cite this article as: Fukumasu NK, Bernardes CF, Ramirez MA, Trava-Airoldi VJ, Souza RM, Machado IF, Local transformation of amorphous hydrogenated carbon coating induced by high contact pressure, Tribology International (2018), doi: 10.1016/j.triboint.2018.04.006. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
2
Local Transformation of Amorphous Hydrogenated Carbon Coating induced by High Contact Pressure
3
N.K. Fukumasua,∗, C.F. Bernardesa , M.A. Ramirezb , V.J. Trava-Airoldib , R.M. Souzaa , I.F. Machadoa
5
Abstract
SC
The development of coatings presenting adaptive characteristics based on mechanical loads would promote improved local characteristics. In this work, the surface modification of an a-C:H coating was analyzed based on the contact pressure using two tribological tests: lubricated reciprocating ball-on-plane and dry micro-scratch tests. Local transformation were evaluated based on Raman spectroscopy and indentation hardness. Results indicated regions presenting a red-shift of the G band and higher indentation hardness. High compressive stress field developed under the contacting region (numerical simulations) coupled to the red-shift of the G band and the increase on indentation hardness inside the tested regions are compatible with the nucleation of sp3 carbon bond sites derived from sp2 bonds,
EP
TE
D
indicating a possible modification of the coating.
AC C
6
Phenomena Laboratory, Polytechnic School of the University of Sao Paulo, Sao Paulo, Brazil b National Institute for Space Research, Brazil
RI PT
a Surface
4
M AN U
1
author. Tel.: +55-11-3091-9865 Email address: [email protected] (N.K. Fukumasu )
∗ Corresponding
Preprint submitted to Elsevier
April 10, 2018
ACCEPTED MANUSCRIPT
7
1. Introduction Automotive and aerospace industries present several examples where an improvement in performance is required
9
to maximize energy efficiency and durability. According to Holmberg et al. [1], on regular passenger vehicles, power
10
generation and transmission systems present total mechanical losses due to friction responsible for approximately 28%
11
of fuel consumption. In this context, several alternatives may be applied to reduce friction losses, such as the use of
12
low friction coatings and improved surface finishing, reducing surface roughness or developing surface topographies
13
to improve lubrication conditions.
RI PT
8
The development and use of low friction coatings became an attractive technology with the use of modern reactors
15
that employ Plasma Enhanced Chemical Vapor Deposition (PECVD) techniques to produce carbon based coatings on
16
regular or complex geometries [2, 3, 4], such as engine taps, helical gears and bearings. High cyclic loads and relative
17
velocities are achieved during the operation of power train components. Those conditions may lead to significant wear
18
and failure of the coatings, decreasing the overall performance of the system. The understanding of coating behavior
19
under extreme conditions is vital to prevent this scenario, and tribological tests are usually conducted before coating
20
the real component. These tests allow exploring issues such as coating/substrate interaction under tribological loads.
M AN U
SC
14
Usually, carbon-based coatings present satisfactory performance in tribological applications, in which high hard-
22
ness and low coefficient of friction (COF) are commonly required [2, 3, 5, 6, 7, 8]. Literature reports an increasing use
23
of diamond like carbon (DLC) coatings, in which tribological properties can be tuned depending on the environment.
24
Under dry sliding condition, DLC coated systems may present a reduction of friction force based on the graphitization
25
of the contacting surfaces, as observed by Liu et al. [9]. This phenomenon is related to the rearrangement of the sp3
26
and sp2 carbon bonds by energy transferred from the mechanical movement to chemical bond kinetics. Polaki et al.
27
[10] reported the influence of high contacting surfaces relative speed of the reduction of COF and wear resistance
28
by the conversion of the sp3 to sp2 bonds. This change on the chemical bonds indicates the formation of a graphitic
29
phase that presents lower COF, due to the lamellar atomic structure, and lower wear resistance associated with lower
30
hardness of this phase.
EP
TE
D
21
Local phase change promoted by localized contact pressure in ductile materials are well reported in the literature
32
[11, 15, 13] and leads to gradients of local microstructure, mechanical and fracture properties. This behavior is not
33
commonly seen for carbon brittle materials, but Gogotsi et al. [14] showed a phase transformation of diamond after
34
an indentation test, in which the deformation of the original material produced nanocrystalline and graphitic carbon
35
localized structures. In terms of carbon-based coatings, Wang et al. [15] showed the transformation of hydrogenated
36
amorphous carbon (a-C:H) coating into graphitic nanocrystallites at the surface of the coating. These nanocrystals
37
were identified as small regions presenting graphite lamellar crystal arrangement, observed using the High Resolution
38
Transmission Electron Microscopy (HRTEM) technique. The formation of these nanostructures was associated to the
39
increase of sp2 carbon bond (G band) peak intensity of the Raman spectrum, observed inside of the deformed region.
40
As presented by Gogotsi et al. [14], Wang et al. [15] and others in the literature [9, 10, 16], the formation
AC C
31
2
ACCEPTED MANUSCRIPT
41
of graphitic structure arrangements from other carbon-based structures is widely known. On the other hand, phase
42
transformation from graphite or amorphous carbon into diamond structures is rare. Usually, stable diamond structures
43
are obtained by Chemical Vapor Deposition (CVD) or thermomechanical systems with high temperature and high
44
pressure (HTHP) techniques [17, 18, 19]. The equilibrium phase diagram of pure carbon systems [20, 30] and experimental observation [21, 22] indicates
46
the possibility of formation of diamond structures for pressures above 10 GPa at room temperature. This process
47
is possible given high hydrostatic pressures, which induces a stable carbon phase with diamond-like microstructure
48
(higher sp3 bonds) and a metastable phase of graphite-like microstructure (sp2 associated bonds).
RI PT
45
For dynamic cases, in which the phase transformation of carbon occurs under continuosly increasing pressures at
50
room temperature was reported by Bai et al. [22]. In that work, para-xylene was subjected to pressures as high as 30
51
GPa at room temperature. The phase transformation from liquid to carbon crystal structures was measured by XRD,
52
Raman and Infrared spectroscopy techniques. Results showed that the increase of the load compacting the para-xylene
53
produced a crystal structure with increasing sp3 (D band) Raman peak proportional to the applied load. Those authors
54
have also indicated that the pressure induced carbon phase transformation was an irreversible process, in which the
55
unloaded sample presented XRD and Raman spectrum peaks not initially presented for the para-xylene liquid.
M AN U
SC
49
Tribological applications usually present elevated contacting loads, resulting in contact pressures on the order of
57
tenths of GPa. Then, the use of a carbon-based coating presenting a dynamic mechanical behavior would promote
58
improved wear resistance while maintaining lower coeficient of friction. In this work, Raman spectroscopy and nano-
59
indentation measurements were conducted to monitor the structure and phase transformations of the materials inside
60
the wear track of high-contact-load tribological analyses. The system under study was composed of an a-C:H coating
61
deposited onto a ductile substrate subjected to high mechanical contact pressures, on the order of 2 GPa.
62
2. Coating deposition and characterization
EP
TE
D
56
To minimize the influence of substrate on the analyses of the coating, the substrate was made of AISI H13 steel,
64
which was water quenched from austenitizing temperature of 1000o C and tempered at 200o C. Prior to the deposition,
65
substrate surface was polished up to 1 µm (20 nm average roughness).
AC C
63
66
The coating and the interlayer in this work were deposited using a pulsed Direct Current Plasma Enhanced Chem-
67
ical Vapor Deposition (DC PECVD) reactor modified to include an additional cathode. This arrangement confines the
68
plasma near the surface of the samples, improving the deposition rate, the quality of the coating and allows a lower
69
deposition temperature (under 100o C). Details of this reactor and the deposition parameters were presented on the
70
literature [2, 23, 24].
71
A silicon interlayer was deposited to improve the adhesion of the carbon-based coating and the substrate. Silane
72
was used to grow this interlayer with 3 sccm gas flow, a bias of -950 V and a working pressure of 0.3 Pa. After this
73
process, the coating was grown using acetylene as precursor with an 8 sccm gas flow, at a pressure of 0.1 Pa and a 3
ACCEPTED MANUSCRIPT
74
constant bias of -800 V. According to Lugo et al. [2], these deposition parameters resulted in a silicon interlayer and a carbon-based
76
coating with, approximately, 500 nm and 1000 nm in thickness, respectively. The composition of the coating was
77
measured by Elastic Recoil Detection Analysis (ERDA) [2] as being 25 wt% of hydrogen and 75 wt% of carbon for
78
the first 500 nm and similar deposition results were obtained by Capote et al. [23, 24].
RI PT
75
81
1800 grating per mm difraction mesh, a 50 µm slit and a 100x objective lens, allowing measurement of D and G bands
82
of the Raman spectra in regions up to 1 µm in diameter with a spectral resolution of 2 cm−1 . All acquired Raman
83
spectra were processed in the Origin v9.5 by Origin Labs Corp., using a Gaussian Non-linear curve fit with a R2
84
higher than 0.95 to deconvolute the D and G shift positions.
SC
80
The characterization of the carbon bonds and the ID /IG intensity ratio at the surface was obtained by using a R confocal Horiba Xplora OneT M Raman system. This system was configured with a 532 nm laser wavelenght, a
79
Mechanical properties of the a-C:H coating were evaluated using the Hysitron Ti950 triboindenter. Several nano-
86
indentation curves were measured to estimate both hardness and elastic modulus, using the Oliver and Pharr method
87
[25]. Load to penetration depth curves were acquired using a Berkovich tip and a maximum load of 700 µN. This
88
load assured the measurement of coating properties with no significant influence of the substrate, once the maximum
89
penetration depth was kept under 50 nm. Additionally, critical loads for coating failure were measured using a cono-
90
spherical tip with 10 µm in diameter and maximum normal load of 360 mN. These critical loads were characterized
91
as the first pop-in in the load-penetration depth indentation curves and are directly linked to the failure of the coating
92
[26]. Also, this equipment can be used in a Scanning Probe Microscopy (SPM) configuration, producing a topography
93
image of selected regions. In this configuration, the Berkovich tip was used with a normal load of 2 µN to acquire
94
images of 10 µm x 10 µm.
95
3. Tribological analyses
EP
TE
D
M AN U
85
Components subjected to pure sliding conditions and elevated contacting loads may present high wear and friction.
97
One way to prevent such critical tribological conditions is the use of protective coatings. In this work, the tribological
98
behavior of a hydrogenated amorphous carbon coating was analyzed based on two sets of experiments: macroscale
99
lubricated reciprocating tests and microscale dry scratch tests.
AC C
96
100
R Macroscale lubricated reciprocating tests were conducted on an Optimol SRV v4 device with a ball on plane
101
configuration (Figure 1), considering that both the plane and the sphere were coated. Tests were based on a step-
102
increase in load from 5 N to 150 N with 5 minutes in each step (total duration of 30 minutes), 25 Hz of reciprocating
103
frequency and a stroke of 2 mm. A droplet of additive free Poly-Alpha-Olefin (PAO) synthetic lubricant in a mixture
104
of PAO4/6 was used in the contact region. The tested spheres were made of AISI 52100 material, coated in a similar
105
process used for the substrate.
106
Microscale scratch tests were carried on a Hysitron Ti950 triboindenter using a cono-spherical tip with 10 µm 4
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 1: Macroscale reciprocating test configuration analyzed in this work, in which both sphere and disk were coated.
in diameter and loads ranging from 25 mN to 250 mN. The total scratch length was 500 µm and constant load
108
configuration were used. In microscale analyses, no liquid lubricant was used (dry condition).
109
4. Finite element analysis
M AN U
107
The microscale scratch test was also analyzed using the TriboCODE numerical package, which is in development
111
at the “Laborat´orio de Fenˆomenos de Superf´ıcie” (LFS) of the University of S˜ao Paulo, Brazil. This package is based
112
on several modules, which are used to simulate and analyze microscale tribological phenomena presenting complex
113
material microstructural configurations and physical mechanisms, such as abrasion, adhesion and thermal fatigue. In
D
110
116
The numerical model was solved using an explicit time evolving algorithm in quasi-static regime, in which no time
117
derivatives nor inertial forces were considered.
EP
TE
115
this work, the scratch test was analyzed as a tridimensional numerical model using three TriboCODE modules (TC R Abrasion, TC Bulk Material Properties and TC Coatings), coupled to the ABAQUS finite element environment.
114
The geometry of the numerical model is presented in Figure 2. In this figure, the configuration consists of the
119
cono-spherical tip (dark gray), the coating (red) and the substrate (light gray). The coating thickness was 2 µm and
120
the substrate was defined as a parallelepiped of 100 µm x 50 µm x 10 µm. The scratch length was 50 µm and the
121
normal load was increased continuously from 0 to 170 mN along the track. No lubricant was considered and two
122
failure models were employed: cohesive zone model for the coating/substrate interface and brittle failure model for
123
the coating. The substrate was considered elastic-perfectly plastic with mechanical properties representative of AISI
124
H13 steel. Mechanical properties of the coating were obtained from the experimental portion of this work (Section
125
2). The failure modeling and property measurements were analyzed following the procedure presented by Fukumasu
126
et al. [27].
AC C
118
5
SC
RI PT
ACCEPTED MANUSCRIPT
M AN U
Figure 2: Numerical model configuration consisting of a cono-spherical tip used to scratch the coated (red) substrate (light gray)
a)
D
b)
Figure 3: Coefficient of friction in the reciprocating tests: a) evolution during the ramping load and b) along the stroke for the last
TE
reciprocating movement of the sphere.
5. Results and discussions
128
5.1. Macroscale reciprocating test results
EP
127
Figure 3 presents the evolution of the coefficient of friction (COF) during the reciprocating tests. The COF curves
130
are presented based on the average of 3 repetition tests. Figure 3a indicates a stabilized value after 50 s and a minimum
131
COF of of 0.091±0.002 achieved at 350 s. This stabilized value of COF is compatible with peak values observed in
132
the reversion position of the reciprocating movement, presented in Figure 3b. In this figure, both reversal points of
133
the sphere movement present a surface to surface contact condition with no lubricant due to the zero velocity. Also, a
134
COF reduction to 0.080±0.001 at the middle of the stroke is observed. This reduction is related to a better lubrication
135
condition compared to the reversal points, given by the occurrence of the maximum relative velocity between the
136
contacting surfaces during the reciprocating cycle.
AC C
129
137
Figure 4 shows a SEM image of the macroscale wear track at the middle of the stroke, after the reciprocating test.
138
This image is representative of the tests and indicates that the coating was well adhered to the substrate, except for
139
isolated coating failures as indicated inside the marked region. 6
SC
RI PT
ACCEPTED MANUSCRIPT
M AN U
Figure 4: Representative wear track after the reciprocating test. The traced red rectangle indicates the region analyzed near the failure of the coating
Raman spectroscopy analyses inside the wear track indicated spatial variations on the G band, which may be
141
related to the fraction of sp2 bonds in the carbon coating. To analyze the change in the Raman shift of the G band, an
142
intensity ratio between two spectral ranges IGr /IGb was analysed. The IGr was defined in the spectral range from 1500
143
cm−1 to 1550 cm−1 , while IGb was defined from 1500 cm−1 to 1600 cm−1 . This selection allows the indication of the G
144
peak position inside the IGr spectrum range for higher values of IGr /IGb , while lower values indicate a G peak in 1550
145
cm−1 to 1600 cm−1 range.
TE
D
140
Figure 5 presents the IGr /IGb ratio image superimposed to the image of a failed region of the coating. In this figure,
147
a red-shift of the G band (higher IGr /IGb ) was observed at the coating surface inside and at the borders of the failed
148
region. The spatial variation of the non-processed G band peak position along the wear track suggests load-induced
149
surface changes at the coating during the reciprocating test. The deconvolution of the Raman spectrum are presented
150
in this figure. For positions presenting higher IGr /IGb ratio (red regions), the G peak position was 1534.4±2.6 cm−1 and
151
a FWHMG value was 197.9±7.8 cm−1 while, for regions presenting lower IGr /IGb (blue regions), the G peak position
152
was 1552.3±1.8 cm−1 and FWHMG was 175.4±8.7 cm−1 .
AC C
EP
146
153
Nano-indentation measurements (Figure 6a), with the corresponding calculation of the hardness and reduced
154
elastic modulus (Figure 6b), were applied to analyze the coating surface in regions presenting variations on the G
155
band position. In this figure, gray color represents measured regions presenting low values of reduced elastic modulus
156
and hardness, which may be a result of coating adhesive or cohesive failure. Nevertheless, Red and Black color
157
measurements presented similar reduced elastic modulus, in the range of 200 GPa to 230 GPa. Black squares in Figure
158
6b represent the original condition (outside the wear track) of the coating, while red measurements are representative
159
of the worn surface presenting a higher IGr /IGb ratio. Comparing both results, the worn surface presented higher 7
EP
TE
D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
Figure 5: Raman shift change of G peak position by mapping the intensity ratio IGr /IGb near the failed coating region: spatial variation of the non-processed G band peak position inside the wear track. In the graphs, gray squares indicate typical spectra obtained for similar regions, while lines indicate the deconvolution of the Raman spectra into D (blue) and G (red) bands.
8
ACCEPTED MANUSCRIPT
7 0 0
In s id e
5 0 0
O u ts id e 4 0 0 3 0 0
In s id e w ith in te r f a c e f a ilu r e
2 0 0 1 0 0 0 0
5
1 0
1 5
2 0
2 5
3 0
3 5
4 0
P e n e tr a tio n D e p th ( n m )
4 5
b)
SC
a)
RI PT
I n d e n t a t i o n L o a d ( µm )
6 0 0
Figure 6: Nano-indentation measurements of the coating: a) nano-indentation curves for inside (red and gray lines) and outside (black lines) of the wear track; b) results for hardness and reduced elastic modulus of the coating for inside (circles) and out-
M AN U
side (squares) of the wear track. Red circles indicate similar reduced elastic modulus but higher hardness compared to outside measurements (black squares), while gray circles indicate a reduction on both characteristics.
160
hardness (in the range of 25 GPa to 40 GPa), suggesting a change in the mechanical behavior that could be associated
161
to the Raman red-shift of the G band. For comparisson, the measured indentation hardness of the substrate was,
162
appoximately, 6.2±0.2 GPa.
163
As reported in the next section, dry microscale scratch tests reproduced the G band peak position variation, indicating no contribution from the additive free lubricant on the modification of the worn coating surface.
165
5.2. Microscale scratch test results
TE
D
164
The groove marks, produced inside the wear track of the macroscale reciprocating test, present similarities to
167
the deformation induced by a single abrasive particle scratching the coating surface. Therefore, microscale scratch
168
tests could be able to reproduce the groove formation, inducing localized changes of hardness and G band position
169
promoted by the abrasive particle in contact with the coating surface.
EP
166
In this test, the influence of the normal load on the coefficient of friction is displayed on Figure 7. Load values up
171
to 100 mN produce similar COF of, approximately, 0.05. Increasing the load led to an increase of the COF, in which
172
the intermediary load of 150 mN produced a COF of 0.08 while the highest analyzed load (250 mN) led to a COF of
173
0.15. In all cases, the tangential force was stable along the entire scratch distance.
AC C
170
174
Figure 8 shows a representative portion of the wear track obtained with the intermediary load case (150 mN
175
case). In this figure, the coating presents three failure modes: adhesive with small cohesive failures of the coating (a),
176
complete spallation of the coating (b) and adhesive failure of the coating without visible cohesive failure (c).
177
The measured map of the IGr /IGb ratio is presented in Figure 9, superimposed to an optical image of the microscale
178
scratch track, in a region presenting no spallation of the coating (Figure 8). In this figure, the higher intensity of
179
IGr /IGb (red colored regions) can be directly associated to regions in which the coating presented some type surface 9
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
Figure 7: Coefficient of friction along the scratch length during the test for analyzed normal loads
Figure 8: Microscale scratch track presenting the local typical observed failure modes: a) adhesive and cohesive failures of the coating; b) complete spallation of the coating and c) adhesive failure of the coating/substrate interface.
10
TE
D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
EP
Figure 9: Superimposed Raman spectroscopy map of IGr /IGb ratio on the scratch track of figure 8. Higher ratio values (red colored regions) indicate a red-shift of the G band.
modification. This localized increase on the IGr /IGb ratio indicates a red-shift of the G band position induced by the
181
deformation of the coating, subjected to high localized contact stresses.
AC C
180
182
Figure 10 shows typical Raman spectra for point measurements of the coating surface inside and outside of the
183
scratched region. Also, this figure presents the deconvolution of the Raman spectrum inside and outside the scratch
184
track. A total of 10 point measurements in each region were analyzed, leading to a deconvoluted G band peak position
185
of 1544.7 ± 1.6 cm−1 inside and 1557.1 ± 0.7 cm−1 outside the scratched region. Also, the inside region presented a
186
full width at half maximum (FWHMG ) value of 183.9 ± 4.1 cm−1 while the outside region presented a value of 171.4
187
± 1.4 cm−1 .
188
Figure 11 was acquired using the Scanning Probe Microscopy technique, in which the indentation tip scans the 11
In s id e
In te n s ity ( a .u .)
-1
M AN U
1 3 5 5 c m
1 5 5 7 c m
1 3 7 3 c m
TE
1 0 0 0
1 2 0 0
EP
8 0 0
-1
-1
D
O u ts id e
-1
SC
1 5 4 4 c m
RI PT
ACCEPTED MANUSCRIPT
1 4 0 0
R a m a n s h ift ( c m
1 6 0 0 -1
1 8 0 0
2 0 0 0
)
AC C
Figure 10: Raman spectroscopy analysis of a-C:H coatings after the scratch test. Gray squares indicate typical spectra obtained for inside and outside the scratched regions, while lines indicate the deconvolution of the Raman spectra into D (blue) and G (red) bands.
12
ACCEPTED MANUSCRIPT
1 0 0
H e ig h t ( n m )
-1 0 -2 0 -3 0
-5 0 -6 0 0
a)
b)
2
RI PT
-4 0
4
6
8
1 0
L a te r a l D is ta n c e ( µ m )
Figure 11: Scanning Probe Microscopy of the scratch track presenting cohesive failure of the coating: a) planar view indicating the indentation analyses markers (black circles) and profile measurement across the track (green dotted line); b) measurement of
M AN U
SC
surface topography across the scratch track.
D
Figure 12: Indentation hardness measurements at positions indicated on Figure 11a and outside the scratch track (original condi-
TE
tion).
surface to obtain an image of the topography of the analyzed region. Figure 11a shows the top view of the wear
190
track. In this figure, three cohesive cracks at the coating surface are observed and recognized as the curved localized
191
variation on the topography. Figure 11b shows the depth profile of the scanned region, indicated in Figure 11a by the
192
green dotted line. The maximum residual depth of the scratched region was 60 nm or 6% of a-C:H coating thickness.
193
Also, hardness measurements (Figure 12) were taken at three points between two cohesive cracks of Figure 11a, as
194
indicated by points 1, 2 and 3.
AC C
EP
189
195
Figure 12 indicates higher hardness for the three measured points on the wear track compared to the initial values
196
of the coating. The difference on hardness measured for point 1 compared to points 2 and 3 may be related to the
197
proximity to the crack, in which the stifness of the coating may be reduced by the coating failure. Nevertheless,
198
these results suggest an increase of coating hardness coupled to an increase of IGr /IGb ratio, both promoted by the
199
deformation of the system induced by the scratch test, similar to the results obtained inside the grooves of macroscale
200
reciprocating tests.
201
Literature indicates that for amorphous carbon coatings subjected to elevated temperatures (annealing or rubbing 13
a)
RI PT
ACCEPTED MANUSCRIPT
b)
indentation curve and b) first pop-in presented on the loading curve.
SC
Figure 13: Coating cohesive failure critical loads measured by nano-indentation technique: a) typical load/penetration depth
processes), the sp3 carbon bonds tend to be converted to sp2 bonds and a blue-shift of the G band is observed [28, 29].
203
In this work, the deconvolution of the Raman spectrum indicated an opposite trend, in which a red-shift of the G band
204
and an increase of the FWHMG value inside the scratched region were observed. According to Ferrari and Robertson
205
[29], this red-shift is linked to an increase of sp3 content for an a-C:H coating and an increase in hardness. Also, the
206
increase of the FWHMG indicate a decrease of the in-plane correlation length (La ) and an increase of carbon disorder,
207
which is an indicative of the nucleation of small sp3 sites derived from sp2 bonds [29].
M AN U
202
One possible cause for this coating modification would be the high contact pressure developed during the scratch
209
test. To better evaluate the contact conditions, a finite element model was developed to simulate the coating/substrate
210
mechanical behavior under these test conditions.
211
5.3. Finite element analysis of the scratch test
212
5.3.1. Mechanical and fracture properties for the numerical simulations
EP
TE
D
208
The initial condition of Young’s modulus of the coating is obtained using the reduced elastic modulus (black
214
square symbols) plotted in Figure 6. The critical load for coating cohesive failure was obtained by nano-indentation
215
technique using a cono-spherical tip. A typical load/penetration depth curve for this test is presented on Figure 13a,
216
in which pop-ins are related to the cohesive failure of the coating, as previously described by Li [26] and considered
217
by Fukumasu [27]. According to those authors, the critical load for coating failure is assumed as the load in which
218
the first pop-in occurs. Figure 13b, which presents a series of pop-ins observed during the indentation of the coated
219
sample by the cono-spherical tip, indicates that the first pop-in occurs in a range from 139 mN to 158 mN normal load,
220
presenting an average value for this coating as 148±5 mN. This value is compatible with the scratch test transitioning
221
load, in which an increase in the COF and the coating failure were observed.
AC C
213
222
Data from Figure 13b were used to evaluate mode I failure stress and fracture energy for the analyzed coating
223
using the TC Coatings module of the TriboCODE. The critical failure stress was calculated as 2.5 GPa and the fracture 14
ACCEPTED MANUSCRIPT
226
used to model cohesive failure of the coating in this work.
227
5.3.2. Numerical scratch test results
RI PT
225
energy as 0.0002 J/µm. These values coupled to the failure strain of 0.0025, which is associated to the characteristic R mesh size, are the input parameters for the brittle failure model [31] implemented in the Abaqus environment and
224
Numerical simulations of the microscale scratch test provided the contact pressure and stress distributions of the
229
coating/substrate system. Figure 14 shows the contact pressure before cohesive or adhesive failure of the coating.
230
In this figure, the maximum contact pressure was calculated as 12 GPa distributed in a half moon geometry towards
231
the direction of the movement of the indenter (Figures 14a and 14b). The numerical simulation reproduces the
232
experimental pressure distribution given the similarities to the geometry of the spalled region of the coating, indicated
233
by the failure mode b presented in Figure 8.
SC
228
Figure 14c shows the evolution of the contact pressure with the increase of the normal load. In this figure, the
235
maximum contact pressure (12 GPa) is achieved for a load of 100 mN and is kept constant for higher loads, promoted
236
by the plastic deformation of the substrate. Once the coating damage threshold is achieved, cohesive cracks nucleate
237
at the surface of the coating and the contact pressure starts to oscilate, as noted in this figure. The numerical simulation
238
predicts the nucleation of cohesive cracks that promote the oscilations of the contact pressure at a scratch distance of
239
about 40 µm or 140 mN (Figure 14c), compatible with experimental critical normal load for the failure of the coating.
240
Figure 15 presents the distribution of the minimum principal stresses in the coating/substrate system. The high
241
contact pressure developed at the coating surface promoted an elevated internal stress distribution in the coating,
242
achieving compressive values of 20 GPa near the surface and decreasing to 12 GPa at coating/substrate interface.
D
M AN U
234
As reported by Bai et al. [22], this high level of compressive stresses developed in the coating is enough to induce
244
a phase transformation of the carbon material at room temperature, in which compressive stresses on the order of 20
245
GPa or higher promotes the rearrangement of carbon bonds, becoming sp3 dominant (red-shift of the G band, higher
246
FWHMG values and higher hardness).
EP
TE
243
After the removal of the load, the induced carbon transformation inside the wear grooves and the scratch track
248
remains stable. This induced phase was observed by the red-shift of the G band (Figures 5 and 10) and by higher
249
hardness measurements (Figures 6 and 12) for both macro and microscale analyzed tracks. This phase stability was
250
also reported by Bai et al. after the removal of the load, in which a solid phase of the carbon material was observed,
251
presenting crystalline material identified by the X-Ray diffraction technique.
AC C
247
15
RI PT
ACCEPTED MANUSCRIPT
M AN U
SC
a)
b)
c)
D
Figure 14: Contact Pressure at the coating promoted by the indenter movement: a) Instantaneous spatial distribution of the contact
AC C
EP
TE
pressure; b) detail of the contact region and c) evolution of the contact pressure with the ramping load during the scratch test.
Figure 15: Minimum Principal Stresses distribution developed inside of the material during the scratch test.
16
ACCEPTED MANUSCRIPT
252
6. Conclusions Coatings that present adaptive features, such as microstructure and mechanical properties, have the ability to
254
improve the performance of components by tuning specific characteristics during the process. In this work, local
255
transformation of an amorphous hydrogenated carbon coating (a-C:H) was analyzed after high load lubricated re-
256
ciprocating ball-on-plane tests and dry micro-scratch tests using Raman spectroscopy and indentation hardness, in
257
which:
260
261
262
263
264
265
regions under high contacting loads on both test configurations;
SC
259
• the analyzed coating presented low coeficient of friction, good adherence to the substrate and small spalled
• Raman spectroscopy analyses of the macroscale wear track indicated regions presenting a red-shift of the G band compared to the coating original condition (outside the track);
• indentation hardness analyses indicated a spatial correlation of higher hardness on regions with higher IGr /IGb
M AN U
258
RI PT
253
intensity ratio;
• microscale dry scratch tests presented similar Raman red-shift compared to macroscale reciprocating tests, indicating a negligible significant contribution from the additive free lubricant on surface modification; • good agreement was obtained between experimental and numerical analysis of coating critical load for first
267
failure. This result was enhanced by the good measurement of both mechanical and fracture properties of the
268
coating, used as input to the numerical solver;
271
272
TE
270
• the experimental spalled region shape, observed on the transitional load of the scratch test, was compatible with the spatial distribution of contact stresses computed by the numerical simulation; • numerical simulation indicates high contact pressure (>12 GPa) developed at the surface and high internal
EP
269
D
266
stresses, raging from 20 GPa to 12 GPa, are developed along coating thickness; • high internal stresses, red-shift of the G band, higher values of FWHMG and the increase on indentation hardness
274
inside the scratched region are compatible with the nucleation of sp3 carbon bond sites derived from sp2 bonds,
275
according to the literature.
276
AC C
273
Acknowledgements
277
The authors recognize the National Council for Scientific and Technological Development (CNPq), the Sao Paulo
278
Research Foundation (FAPESP) and the National Bank for Economic and Social Development (BNDES) for financial
279
support.
17
ACCEPTED MANUSCRIPT
280
References [1] K. Holmberg, P. Andesson, A. Erdemir, Global Energy consumption due to friction in passenger cars. Tribol. Int., 47, 221-234 (2012).
282
[2] D.C. Lugo, P.C. Silva, M.A. Ramirez, E.J.D.M. Pillaca, C.L. Rodrigues, N.K. Fukumasu, E.J. Corat, M.H. Tabacniks, V.J. Trava-Airoldi,
283
Characterization and tribologic study in high vacuum of hydrogenated DLC films deposited using pulsed DC PECVD system for space
284
applications, Surf. Coat. Technol., 332, 135-141 (2017).
287 288 289 290 291 292 293 294
[3] J. Solis, H. Zhao, C. Wang, J.A. Verduzco, A.S. Bueno, A. Neville, Tribological performance of an H-DLC coating prepared by PECVD, Appl. Surf. Sci., 383, 222-232 (2016).
[4] D. Lusk, M. Gore, W. Boardman, T. Casserly, K. Boinapally, M. Oppus, D. Upadhyaya, A. Tudhope, M.Gupta, Y. Cao, S.Lapp, Thick DLC films deposited by PECVD on the internal surface of cylindrical substrates, Diam. Relat. Mat., 17, 1613-1621 (2008).
[5] A. Erdemir, C. Donnet, Tribology of diamond-like carbon films: recent progress and future prospects, J. Phys. D-Appl. Phys., 39, 311-327 (2006).
SC
286
[6] D. Sheeja, B.K. Tay, S.M. Krishnan, L.N. Nung, Tribological characterization of diamond-like carbon (DLC) coatings sliding against DLC coatings, Diam. Relat. Mat., 12, 1389-1395 (2003).
[7] H.A. Tasdemir, T. Tokoroyama, H. Kousaka, N. Umehara, Y. Mabuchi, Friction and wear performance of boundary-lubricated DLC/DLC contacts in synthetic base oil, Procedia Engineering, 68, 518-524 (2013).
M AN U
285
RI PT
281
295
[8] S.J. Bull, Tribology of carbon coatings: DLC, diamond and beyond, Diam. Relat. Mat., 4, 827-836 (1995).
296
[9] Y. Liu, A. Erdemir, E.I. Meletis, An investigation of the relationship between graphitization and frictional behavior of DLC coatings, Surf.
299 300 301 302 303 304 305
[10] S.R. Polaki, N. Kumar, K. Ganesan, K. Madapu, A. Bahuguna, M. Kamruddin, S. Dash, A.K. Tyagi, Tribological behavior of hydrogenated DLC film: Chemical and physical transformations at nano-scale, Wear 339-339, 105-113 (2015). [11] F.P. Marques, C. Scandian, A.C. Bozzi, N.K. Fukumasu, A.P. Tschiptschin, Formation of a nanocrystalline recrystallized layer during microabrasive wear of a cobalt-chromium based alloy (Co-30Cr-19Fe), Tribol. Int., 116, 105-112 (2017). [12] S.J. Wang, H. Wang, K. Du, W. Zhang, M.L. Sui, S.X. Mao, Deformation-induced structural transition in body-centred cubic molybdenum,
D
298
Coat. Technol., 86-87, 564-568 (1996).
Nat. Commun., 5, 3433 (2014).
[13] A. Das, Contribution of deformation-induced martensite to fracture appearance of austenitic stainless steel, Mater. Sci. Technol., 32, 1366-
TE
297
1373 (2016).
306
[14] Y.G. Gogotsi, A. Kailer, K.G. Nickel, Pressure-induced phase transformations in diamond, J. Appl. Phys., 84, 1299-1304 (1998).
307
[15] D.S. Wang, S.Y. Chang, Y.C. Huang, J.B. Wu, H.J. Lai, M.S. Leu, Nanoscopic observations of stress-induced formation of graphitic nanocrystallites at amorphous carbon surfaces, Carbon, 74, 302-311 (2014).
EP
308
[16] Y.G. Gogotsi, A. Kailer, K.G. Nickel, Transformation of diamond to graphite, Nature 401, 663-664 (1999).
310
[17] B.V. Spitsyn, L.L. Bouilov, B.V. Derjaguin, Vapor growth of diamond on diamond and other surfaces, J. Cryst. Growth, 52, 219-226 (1981).
311
[18] S. Matsumoto, Y. Sato, M. Tsutsumi, N. Setaka, Growth of diamond particles from methane-hydrogen gas, J. Mater. Sci., 17, 3106-3112
312
AC C
309
(1982).
313
[19] H. Kanda, P. Capper, Bulk Crystal Growth in Electronic, Optical and Optoelectronic Materials, John Wiley and Sons, (2005).
314
[20] Q. Jiang, J.C. Li, G. Wilde, The size dependence of the diamond-graphite transition, J. Phys.-Condes. Matter, 12, 5623-5627 (2000).
315
[21] T.B. Shiell, D.G. McCulloch, J.E. Bradby, B. Haberl, R. Boehler, D.R. McKenzie, Nanocrystalline hexagonal diamond formed from glassy
316 317 318 319 320 321 322
carbon, Sci. Rep., 6, 37232 (2016).
[22] Y. Bai, Z. Yu, R. Liu, N. Li, S. Yan, K. Yang, B. Liu, D. Wei, Lin Wang, Pressure-Induced Crystallization and Phase Transformation of Para-xylene, Sci. Rep., 7, 5321 (2017). [23] G. Capote, G.C. Mastrapa, V.J. Trava-Airoldi. Influence of acetylene precursor diluted with argon on the microstructure and the mechanical and tribological properties of a-C:H films deposited via the modified pulsed-DC PECVD method, Thin Solid Films, 589, 286-291 (2015). [24] G. Capote, M.A. Ramirez, P.C. Silva, D.C. Gonc¸alez, V.J. Trava-Airoldi. Improvement of the properties and the adherence of DLC coatings deposited using a modified pulsed-DC PECVD technique and an additional cathode. Surf. Coat. Technol., 308, 70-79 (2016).
18
ACCEPTED MANUSCRIPT
323 324
[25] W.C. Oliver, G.M. Pharr, An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments, J. Matter. Res., 7, 1564-1583 (1992).
325
[26] X. Li, B. Bhushan, Measurement of fracture toughness of ultra-thin amorphous carbon films, Thin Solid Films 315, 214-221 (1998).
326
[27] N.K. Fukumasu, R.M. Souza, Numerical evaluation of cohesive and adhesive failure modes during the indentation of coated systems with
327
compliant substrates, Surf. Coat. Technol. 260, 266-271 (2014). [28] J. Robertson, Mechanical properties and structure of diamond-like carbon, Diam. Relat. Mat., 1, 397-406 (1992).
329
[29] A.C. Ferrari, J. Robertson, Raman spectroscopy of amorphous, nanostructured, diamond-like carbon, and nanodiamond, Phil. Trans. R. Soc.
SC
Finite Elements, Cem. Concr. Res., 6, 773-782 (1976).
M AN U
334
(1993).
[31] A. Hillerborg, M. Modeer, P.E. Petersson, Analysis of Crack Formation and Crack Growth in Concrete by Means of Fracture Mechanics and
D
333
TE
332
EP
331
¨ ı2512 (2004). Lond. A, 362, 2477 A` ’ [30] H.O. Pierson, Handbook of carbon, graphite, diamond and fullerenes: properties, processing and applications, Noyes Publication, New Jersey
AC C
330
RI PT
328
19
ACCEPTED MANUSCRIPT
• High contact pressure tribology measurements of Diamond like Carbon coatings
336
• High correlation of higher indentation hardness on regions with increased sp3 carbon bonds inside wear track
337
• Damage calibrated model to reproduce realistic cohesive and adhesive coating failure by FEM model
338
• Contact pressures and internal coating stresses were induced in the range of equilibrium formation of diamond-
EP
TE
D
M AN U
SC
like carbon phase
AC C
339
RI PT
335
20
ACCEPTED MANUSCRIPT
• High contact pressure tribology measurements of Diamond like Carbon coatings
316
• High correlation of higher indentation hardness on regions with increased sp3 carbon bonds inside wear track
317
• Damage calibrated model to reproduce realistic cohesive and adhesive coating failure by FEM model
318
• Contact pressures and internal coating stresses were induced in the range of equilibrium formation of diamond-
EP
TE D
M AN U
SC
like carbon phase
AC C
319
RI PT
315
18