materials

Article
Effect of Hardening Exponent of Power-Law
Hardening Elastic-Plastic Substrate on Contact
Behaviors in Coated Asperity Contact
Xiqun Lu, Hanzhang Xu and Bin Zhao *
College of Power and Energy Engineering, Harbin Engineering University, Harbin 150001, China;
luxiqun@hrbeu.edu.cn (X.L.); xuhanzhang@hrbeu.edu.cn (H.X.)
* Correspondence: donglizhaobin@hrbeu.edu.cn; Tel.: +86-451-8258-9199




Received: 20 September 2018; Accepted: 11 October 2018; Published: 12 October 2018


Abstract: The contact between a rigid flat and a coated asperity is studied using the finite element
method. The substrate is assumed as the power-law hardening elastic–plastic material. The effect of
the hardening exponent of the substrate (n) on the contact behaviors including contact load, area,
coating thickness variation and stress in the coating, is investigated. It shows larger hardening
exponent results in larger contact loads and larger maximum stresses in the coating at a given
interference, and leads to smaller contact area at a specific contact load. The coating thickness
becomes smaller monotonically as the interference increases for larger hardening exponents, while it
recovers gradually after reaching the minimum value for the smaller n cases. This work will give
some universal guidance to improve the contact performance for coatings by adjusting the hardening
exponent of the substrate and by optimizing the coatings parameters.

Keywords: asperity contact; coatings; power-law hardening; contact behaviors; contact mechanics

1. Introduction
Contact between surfaces plays a very important role in the industry field [1]. Coatings have
been widely applied on the contact surfaces [2–4], since it is an effective means to improve the
mechanical performance such as the wear resistance, corrosion resistance and so on, where tribological
performance should be specially considered among them. To select optimal coatings efficiently and
achieve best tribology performance for different substrate materials is very meaningful but full of
challenges. Until now, the trial and error method is still the usual approach since there seems to be
little effective and perfect theories to guide to achieve the coating optimization, as the authors know.
Therefore, to develop the related theory should be the direction which guides the efforts going forward.
All contact surfaces are rough on the microscale and contain many asperities. When two coating
surfaces are in contact with each other, the contact happens at the top of the coated asperity in fact.
Therefore, the contact between coated asperities should be the first problem to be considered, which is
always simplified as the contact between a sphere and a flat [5]. Generally, when an asperity and a flat
contact with each other, the asperity could be flattened by the flat or indent it [6]. Most studies about the
coating contact so far considered the indentation, i.e., a rigid spherical indenter contact with a coated
deformable flat, since they focused on the mechanical properties of the coating. Pogrebnjak [7] studied
the structural and mechanical properties of NbN and Nb–Si–N films with the experimental method
and the molecular dynamics simulation. Coy et al. [8] considered the topographic reconstruction
and mechanical properties of atomic layer deposited Al2 O3 /TiO2 nanolaminates by nanoindentation.
However, when the tribological performance is the focus of attention, the flattening that is associated
with mild adhesive friction and wear should be considered. In this work, the flattening process of the
coated asperity by a rigid flat is explored.

Materials 2018, 11, 1965; doi:10.3390/ma11101965 www.mdpi.com/journal/materials

Materials 2018, 11, 1965 2 of 13

Chang [9] developed a modified elastic–plastic asperity contact model on basis of

Greenwood–Williamson (GW) model [10] to study the contact between surfaces with soft coatings.
The effect of the coating thickness and plasticity index on the contact area and normal load was
investigated. McCool [11] described a micro-contact model by extending the GW model [10] to
investigate the coating contact, and explored the effect of the coating thickness and compliance on
the contact load, area and mean real pressure. Liu et al. [12] presented extended Hertzian formulae
to predict contact behaviors for circular and elliptical point contact problems with the presence of
coating. They expressed the extended equivalent modulus of the coated asperity as the function of the
coating thickness along with Young’s modulus and Poisson’s ratio of the coating and the substrate.
Garjonis et al. [13] presented the analytical and finite element (FE) models by means of Hertz theory
and FE methods respectively, to study the normal contact between two layered spheres in the elastic
phase. They obtained the force-displacement relation in the process of contact between the layered
asperities, and validated these relations by FE results. Yeo et al. [14] used an improved elastic contact
model to investigate the contact between a coated asperity and a rigid flat, and to study the asperity
contact stiffness. The results were also verified by the FE results. Goltsberg and Etsion [15] used the
FE method to explore the elastic contact between a coated asperity and a rigid flat. They studied the
effect of the geometrical and mechanical properties of the coating and substrate on contact behaviors,
and got universal dimensionless formulae about the relation of load and displacement based on many
simulation results. Subsequently, Goltsberg and Etsion [16] continued to study the contact area and
maximum equivalent stress using the FE method and obtained the corresponding dimensionless
expressions, constituting a complete universal contact model for the elastic coated asperity contact
along with their previous work presented in [15].
All these models about the coated asperity contact introduced above considered the elastic case,
i.e., both the substrate and the coating were assumed as the purely elastic materials. However, for many
metallic materials, their stress-strain relations actually show elastic–plastic characteristics. Therefore,
it should be necessary to consider the behavior of the elastic–plastic asperity contact. Eid et al. [17]
proposed a coated asperity contact model considering the effect of adhesion. The material properties
of the asperity including the substrate and coating are both assumed to be elastic–plastic with 2%
linear hardening of the elastic modulus. The effect of the coating thickness on the pull-off force and
contact radius in the loading and unloading processes was studied. Goltsberg et al. [18] investigated
the onset of the plastic yielding in a coated sphere when it contacted with a rigid flat by means of the
FE software. The asperity materials were also assumed as elastic–plastic with 2% linear hardening of
the elastic modulus. The influence of the asperity material properties and the coating thickness on
the critical contact parameters was investigated, such as critical normal load, area and interference
when the plastic yielding first appears. Chen et al. [19] continued to study the plasticity evolution in
the coated asperity which is 2% linear hardening elastic–plastic material during the process of contact
with a rigid flat, and obtained empirical expressions about the critical interferences at the onset of
plastic yielding. In addition, they found that as the interference increased, the elastic core (low-stress
zone) appeared under the contact area, leading to the decrease of the mean pressure.
Most of the FE models mentioned above about the coated asperity contact focused on the elastic
or 2% linear hardening elastic–plastic asperity materials. However, for many materials, the power
law is more precise to describe their strain hardening behavior [20]. The influence of the power-law
hardening material properties on the coated asperity contact has not been studied in detail, which
will be considered in this work where the substrate is set as the power law hardening elastic–plastic
materials. Then, the effect of the substrate-hardening exponent on the contact parameters such as
contact load, area and stress is focused on in this work. On basis of the numerical analysis in this work,
the effect of the material property and geometrical characteristics of coating/substrate system on its
tribological performance could be predicted preliminarily and accurately in the future, and then it
would be meaningful to help choose the optimal coating and obtain the best tribological performance.
Materials 2018, 11, x FOR PEER REVIEW 3 of 13
Materials 2018, 11, 1965 3 of 13
like carbon (DLC) film, which could be simplified and assumed as the purely elastic materials as
revealed in Refs. [15,16].
It should be noted that this work focuses on the very hard coating such as diamond-like carbon (DLC)
2. Construction
film, which couldofbe
the Finite Element
simplified Modelas the purely elastic materials as revealed in Refs. [15,16].
and assumed

2. Construction
2.1. of the Finite
Power-Law Hardening Element Model
Property
In this work,
2.1. Power-Law the substrate
Hardening is assumed as the power law hardening elastic–plastic material. The
Property
relation between the stress and strain at first are linear in the elastic stage, while after the yield
In this work, the substrate is assumed as the power law hardening elastic–plastic material.
strength Y0, i.e., in the hardening stage, the relation changes to be power law and the plastic behavior
The relation between the stress and strain at first are linear in the elastic stage, while after the yield
satisfies J2 flow theory and isotropic hardening model, as shown in Figure 1. The relation is descripted
strength Y0 , i.e., in the hardening stage, the relation changes to be power law and the plastic behavior
as Equation (1).
satisfies J2 flow theory and isotropic hardening model, as shown in Figure 1. The relation is descripted
as Equation (1). σ(/ E ， σ ≤ Y0
ξ= σ ≤ 1/Yn0 (1)
(Y0 / E )(σ / Y0 ) ,σ1/n
> Y0
σ/E,
ξ= (1)
(Y0 /E)(σ/Y0 ) , σ > Y0
where
where ξξisisthe
thestrain,
strain,σ σisisthe
the stress,
stress, E the
E is is the Young’s
Young’s Modulus,
Modulus, Y0 isYthe
0 is yield
the yield strength,
strength, and nand n
is the
is the hardening
strain strain hardening
exponent exponent ranging
ranging from 0 to from
1. For0two
to 1. For two
extreme extreme
cases, cases,
n = 1 or n the
n = 0, = 1 materials
or n = 0,
the materials
change change
to be the purelyto be the
elastic or purely elastic or elastic
elastic perfectly-plastic perfectly-plastic case.
case.

Figure 1.
Figure The constitutive
1. The constitutive law
law of
of power-law
power-law hardening
hardening elastic–plastic
elastic–plastic materials.
materials.

2.2. Finite Element Model

2.2. Finite Element Model
A two-dimensional axisymmetric finite element model was developed to consider the contact
A two-dimensional axisymmetric finite element model was developed to consider the contact
between a rigid flat and a coated asperity using the software ANSYS 16.1, as presented in Figure 2.
between a rigid flat and a coated asperity using the software ANSYS 16.1, as presented in Figure 2.
The asperity was modeled as a quarter of a circular plane, consisting of the coating part whose thickness
The asperity was modeled as a quarter of a circular plane, consisting of the coating part whose
was t and the substrate part whose radius was R. The rigid flat was represented by a line. Considering
thickness was t and the substrate part whose radius was R. The rigid flat was represented by a line.
the contact region is of most interest, the coating part and the substrate part close to the contact region
Considering the contact region is of most interest, the coating part and the substrate part close to the
(Zone I in Figure 2) were meshed fine. The radius of Zone I is 0.1 R. Zone II of the substrate part was
contact region (Zone I in Figure 2) were meshed fine. The radius of Zone I is 0.1 R. Zone II of the
discretized much more coarsely since it was far away from the contact region. The eight-node PLANE
substrate part was discretized much more coarsely since it was far away from the contact region. The
183 element was used to mesh the asperity, while the contact element CONTA 172 and target element
eight-node PLANE 183 element was used to mesh the asperity, while the contact element CONTA
TAEGET 169 were utilized to consider the contact between the asperity and rigid flat. After meshing,
172 and target element TAEGET 169 were utilized to consider the contact between the asperity and
the total number of the elements are 17,810 and the nodes are 48,825.
rigid flat. After meshing, the total number of the elements are 17,810 and the nodes are 48,825.
Two boundary conditions were assumed as follows: First, nodes on the symmetry axis (y axis in
Two boundary conditions were assumed as follows: First, nodes on the symmetry axis (y axis in
Figure 2) were limited to move in the radial direction; secondly, nodes at the bottom of the asperity
Figure 2) were limited to move in the radial direction; secondly, nodes at the bottom of the asperity
were constrained in all directions. The frictionless contact condition between the asperity and the flat
were constrained in all directions. The frictionless contact condition between the asperity and the flat
was used, while for the contact between coating and substrate, the bonded condition was adopted.
was used, while for the contact between coating and substrate, the bonded condition was adopted.
The coating part was set as the elastic homogenous material without residual stresses, while the
The coating part was set as the elastic homogenous material without residual stresses, while the
substrate part was assumed as the elastic power-law hardening plastic material. Both parts were
substrate part was assumed as the elastic power-law hardening plastic material. Both parts were
isotropic. A displacement load was applied to the coated asperity through the rigid flat, and the
isotropic. A displacement load was applied to the coated asperity through the rigid flat, and the
reaction force of the rigid flat as an output could be obtained as the contact force. The von Mises
reaction force of the rigid flat as an output could be obtained as the contact force. The von Mises
Materials 2018, 11, 1965 4 of 13
Materials 2018, 11, x FOR PEER REVIEW 4 of 13

yielding criterion was adopted to consider the behavior of substrate

substrate changing
changing from
from elastic
elastic to
to plastic.
plastic.
The augmented Lagrangian method was selected as the solution method.
method was selected as the solution method.

Figure 2. The two-dimensional finite element model.

The developed
developedFE FEmodel
model was verified
was withwith
verified two following
two followingapproaches. First, assuming
approaches. the coating
First, assuming the
and substrate
coating materialsmaterials
and substrate were identical and elastic,
were identical andtheelastic,
contacttheforces were
contact calculated
forces by the developed
were calculated by the
FE model inFE
developed this workinand
model thiscompared
work andwith the analytical
compared with the Hertz solution.
analytical Thesolution.
Hertz differences
Thewere shown
differences
to be less
were shown than
to2.8%.
be lessSecondly,
than 2.8%. theSecondly,
mesh density was doubled
the mesh density wasiteratively
doubled until it had little
iteratively untileffect
it hadonlittle
the
results between successive iteration, in order to ensure the mesh convergence.
effect on the results between successive iteration, in order to ensure the mesh convergence.
It should be noted that some ideal assumptions were used in this FE model making the
complicated contact
contactproblem
problemsimplified
simplifiedasas follows:
follows: (1)(1)
TheThe contact
contact between
between the asperity
the asperity andrigid
and the the
rigid is frictionless;
is frictionless; (2) the(2) bonded
the bonded condition
condition is adopted
is adopted between
between thethe coatingand
coating andthethesubstrate;
substrate; (3)
(3) the
substrate part
part is
is assumed
assumedas asthetheelastic
elasticpower-law
power-lawhardening
hardening plastic
plasticmaterial.
material.Typical
Typicalresults could
results be
could
obtained
be obtained to preliminarily
to preliminarily explore the effect
explore of theofpower-law
the effect the power-law hardening elastic–plastic
hardening materials
elastic–plastic of the
materials
substrate on the on
of the substrate contact behaviors.
the contact As the As
behaviors. work
thegoes
work on, the on,
goes idealtheassumptions will be relaxed
ideal assumptions step by
will be relaxed
step in
bythe
stepfuture.
in the future.

3. Results and
3. Results and Discussion
Discussion
The results about
The results about the contact behaviors
the contact behaviors obtained
obtained with
with the
the developed
developed FE FE model
model are are shown
shown in this
in this
section. The geometrical and material parameters for the coated asperity
section. The geometrical and material parameters for the coated asperity are set as follows: The are set as follows: The asperity
radius
asperity R radius
changes R from
changes5 mm fromto 20 mm;tothe
5 mm 20 dimensionless coating thickness
mm; the dimensionless t/R variest/R
coating thickness in varies
the rangein theof
0.001 ≤ t/R ≤ 0.02 with the interval of 0.001; the Poisson’s ratios of the coating
range of 0.001 ≤ t/R ≤ 0.02 with the interval of 0.001; the Poisson’s ratios of the coating νco and the ν co and the substrate
νsu are both
substrate νsuset
areasboth
0.32,set
a common
as 0.32, avalue
common for metallic
value for materials;
metallic the Young’sthe
materials; modulus
Young’s ofmodulus
the coating of E co
the
and the substrate E su changes in a wide range of 1 ≤ E co /E su ≤ 10; the yield
coating Eco and the substrate Esu changes in a wide range of 1 ≤ Eco/Esu ≤ 10; the yield strength of thestrength of the substrate
Y 0_sub is assumed
substrate as a constant
Y0_sub is assumed as a 210 MPa;210
constant the MPa;
hardening exponent exponent
the hardening of the substrate n ranges from
of the substrate 0 to
n ranges
1. The geometrical and material parameters could be chosen arbitrarily
from 0 to 1. The geometrical and material parameters could be chosen arbitrarily from the range from the range mentioned
above,
mentioned covering
above,a large number
covering a largeof realistic
number materials
of realistictomaterials
considerto the effect ofthe
consider theeffect
substrate
of thehardening
substrate
exponent on the contact behaviors. Some typical cases are selected
hardening exponent on the contact behaviors. Some typical cases are selected and shown and shown as follows based on the
as follows
condition
based on the thatcondition
asperity radius R is 10radius
that asperity mm. R is 10 mm.
Figure 3 presents the influence of
Figure 3 presents the influence of hardening hardening exponents
exponents of of the substrate n
the substrate on the
n on the relation
relation between
between
the dimensionless contact load
the dimensionless contact load P/Pc_sub P/P and interference w/w
c_sub and interference w/wc_sub
for the contact between a
c_subfor the contact between a rigid flat rigid flat
and
and aa coated
coated asperity
asperity having
having the the power
power law law hardening
hardening elastic–plastic
elastic–plastic substrate. Some typical
substrate. Some typical results
results
are
are selected to be shown for the cases where the hardening exponent of the substrate n rangesfrom
selected to be shown for the cases where the hardening exponent of the substrate n ranges from 0
to 1 at two selected Young’s modulus
0 to 1 at two selected Young’s modulus ratios ratios E /E
co Ecosu = 1, 4. The dimensionless coating thickness
/Esu = 1, 4. The dimensionless coating thickness t/R t/R is
set as 0.01. The critical contact area, load and interference
is set as 0.01. The critical contact area, load and interference of the substrate,
of the Ac_sub
substrate, , P,c_sub
Ac_sub andwwc_sub
Pc_suband , at
c_sub, at
the yielding inception under the normal load were given by
the yielding inception under the normal load were given by Jackson and Green [21]: Jackson and Green [21]:

C
CνvYY00_sub
_sub RR2
2
Ac_sub π 33((
Ac_sub== π )) (2)(2)
Esusu
22E
Materials 2018, 11, x FOR PEER REVIEW 5 of 13
Materials 2018, 11, 1965 5 of 13

π 3Y0_sub Y
Pc_sub = π 3 Y Cv 3 ( R(1 − v 2 )( Y0_sub ))2 2 (3)
Pc_sub = 60_sub Cν3 ( R(1 − ν2 )( E0_sub
su )) (3)
6 Esu
π (1 − v ) Y 2
2
wc_sub = (Cvπ (1 − ν2() 0_sub 2
Y )) R (4)
wc_sub = (Cν 2 (Esu0_sub )) R (4)
2 Esu
where
where E su and
Esu and Y
Y0_sub is the Young’s modulus and yield strength of the substrate, Cν is a parameter
0_sub is the Young’s modulus and yield strength of the substrate, Cν is a parameter
related
related to
to the
the Poisson’s
Poisson’sratio
ratioas
as C = 1.234
Cνν = 1.234++1.256 ν.
1.256ν.

Figure
Figure 3.3. The
The relation
relation between
between the the dimensionless
dimensionlesscontact
contactload
load P/P
P/P c_sub and
c_sub andinterference
interference w/w
w/w for
c_subc_sub

for some
some typical
typical cases
cases where
where dimensionless
dimensionless coating
coating thicknesst/R
thickness t/R== 0.01,hardening
0.01, hardeningexponents
exponentsof of the
the
substratenn==0.1,
substrate 0.1,0.5,
0.5, 0.7,
0.7, 0.90.9
andand Young’s
Young’s modulus
modulus ratios
ratios Eco
Eco/E su /E
= 1,su4.=The
1, 4.results
The results
for the for the elastic
purely purely
elastic
(n (n = elastic
= 1) and 1) and elastic fully-plastic
fully-plastic (n = 0)(n = 0) are
cases cases aregiven
also also given and compared
and compared with with existing
existing literature.
literature. (a)
(a) E
Eco/Esu /E = 1; (b) E
co = 1;su(b) Eco/Esu co /E
= 4. su = 4.

Figure 3a
Figure 3ashows
showsthe theresults
resultsofofthe the contact
contact loadload EcoE
forfor /Ecosu/E
= 1. = 1. extreme
su Two Two extremecases are cases aregiven
also also
given as a comparison. One is the case of the substrate hardening exponent
as a comparison. One is the case of the substrate hardening exponent n = 1.0, which means the asperity n = 1.0, which means the
asperity transforms
transforms into the homogenous
into the homogenous and purelyand purely
elastic elastic
material. Inmaterial.
this case, Inthethis case,load
contact the is contact load
calculated
is calculated and compared with the Hertz solution. The differences
and compared with the Hertz solution. The differences are less than 2.8% as stated in Section 2. The are less than 2.8% as stated in
Section 2. The other extreme case is that n = 0, which means the substrate
other extreme case is that n = 0, which means the substrate changes to be the elastic perfectly-plastic changes to be the elastic
perfectly-plastic
material. material.results
The numerical The numerical
for 2% linear results for 2% linear
hardening substratehardening
case aresubstrate
also shown casein areFigure
also shown
3a, to
in Figure 3a, to simulate the elastic perfectly-plastic material under the small
simulate the elastic perfectly-plastic material under the small interference considering its stress-strain interference considering
its stress-strain
relation relation
as revealed as revealed
in Ref. [22]. Thein Ref. [22].
results showThe results
a good show a good
coincidence coincidence
besides besides a small
a small difference. The
difference.gets
difference The larger
difference
as the gets larger as the
interference interference
increases. This increases.
is due to the Thisreason
is duethat
to thethereason that the
difference of
difference
material of material
properties properties
between between
the real 2% linear thehardening
real 2% linear hardeningmaterial
elastic–plastic elastic–plastic
and thematerial
ideal elasticand
the ideal elastic perfectly–plastic
perfectly–plastic material gets larger. material gets larger.for
The comparison The comparison
these two extreme for these
cases two
proves extreme
that the cases
FE
proves that the FE model is accurate and can be used to consider
model is accurate and can be used to consider the contact behavior under other conditions. Figure 3bthe contact behavior under other
conditions.
shows Figure for
the results 3b shows
Eco/Esuthe results
= 4. for Eco /E
The elastic (n su = 4.and
= 1) Theelastic
elasticfully-plastic
(n = 1) and elastic
(n = 0)fully-plastic
substrate of (n =
the0)
substrateare
asperity of also
the asperity
considered, are also
which considered,
appears atwhich two sidesappears
of the at different
two sidesexponent
of the different
cases (0exponent
< n < 1).
cases (0 < n < 1). This also can prove the accuracy of this model
This also can prove the accuracy of this model to some extent. The curves in Figure 3 reveal to some extent. The curves in that
Figure the3
reveal that the load increases with the increase of interferences.
load increases with the increase of interferences. In addition, for the same Young’s modulus ratio In addition, for the same Young’s
modulus
case, ratiobecomes
the load case, thelarger
load becomes larger as the
as the hardening hardening
exponent exponent
increases fromincreases
0 to 1 at from
a given 0 tointerference.
1 at a given
interference. This is because as the hardening exponent increases, the
This is because as the hardening exponent increases, the substrate becomes more elastic, which means substrate becomes more elastic,
which
the meansforce
reaction the reaction force ofisthe
of the asperity asperity
larger underisthe larger
sameunder the same interference.
interference.
The contact area plays a very important role in the
The contact area plays a very important role in the tribological behaviors tribological behaviors suchsuch as as wear,
wear, friction
friction
and so on, and thus the effect of the hardening exponent of the
and so on, and thus the effect of the hardening exponent of the substrate on the contact area of substrate on the contact area of the
the
coated asperity
coated asperity are are also
also studied
studied in in this
this work.
work. Figure
Figure 44 shows
shows the the relation
relation between
between the the dimensionless
dimensionless
contact area A/A
contact area A/Ac_sub and load P/P
c_suband load P/Pc_sub for some typical cases where dimensionless
for some typical cases where dimensionless coating thickness
c_sub coating thickness
t/R = 0.01, Young’s modulus
t/R = 0.01, Young’s modulus ratios Eco ratios E /E
co/Esusu = 4 at some selected hardening exponents
= 4 at some selected hardening exponents of the substrate of the substrate
nn ==0.1,
0.1, 0.5,
0.5, 0.7
0.7 and
and 0.9. In addition,
0.9. In addition, the thecases
casesfor forthe
thepurely
purelyelastic
elastic(n(n= =1)1)and
andelastic
elasticfully-plastic
fully-plastic (n (n
= 0)
=
Materials 2018, 11, 1965 6 of 13
Materials 2018, 11, x FOR PEER REVIEW 6 of 13

substrate areare
0) substrate also considered
also considered andand
shown
shownin Figure 4. It4.could
in Figure It couldbe be
seen
seenthat thethe
that contact area
contact increases
area increasesas
the value of n decreases at a specific contact load. It is because the substrate with smaller
as the value of n decreases at a specific contact load. It is because the substrate with smaller n is more n is more
plastic,
plastic, and
and the
the same
same contact
contact load
load will
will lead
lead to
to larger
larger interferences.
interferences. ThusThus the
the contact
contact area
area increases
increases
accordingly.
accordingly. This tendency can also be seen from Figure 5, where the change of the profiles of
This tendency can also be seen from Figure 5, where the change of the profiles of the
the
asperities with different substrate materials before and after the normal loading is
asperities with different substrate materials before and after the normal loading is given. The contactgiven. The contact
radii
radii are
areshown
shownasasthe theflat parts
flat in the
parts center
in the of the
center of profiles, whichwhich
the profiles, also represents the sizethe
also represents of the
sizecontact
of the
area. It could also be found that as the hardening exponent of the substrate n decreases,
contact area. It could also be found that as the hardening exponent of the substrate n decreases, the the contact
radii become
contact larger, which
radii become larger,accords
which with thewith
accords conclusions given in given
the conclusions Figurein4.Figure 4.

Figure 4.4. The relation

relation between
betweenthe thedimensionless
dimensionlesscontact
contactarea areaA/AA/A c_sub and
c_sub loadP/P
andload c_sub for some
P/Pc_sub
typical
typical cases
cases where
wheredimensionless
dimensionlesscoating thicknesst/R
coatingthickness t/R==0.01,
0.01,Young’s
Young’smodulusmodulusratiosratiosEco /E
Eco su = 4 at
/Esu
different
differenthardening
hardeningexponents
exponentsofofthe substrate
the substraten =n 0.1, 0.5,0.5,
= 0.1, 0.7,0.7,
0.9.0.9.
TheThe results for for
results the the
purely elastic
purely (n = (n
elastic 1)
and
= 1) elastic fully-plastic
and elastic (n = 0)(ncases
fully-plastic are also
= 0) cases aregiven. (a) 0 ≤(a)
also given. P/P ≤ P/P≤
0 c_sub 3000;
c_sub ≤ P/P
(b) 0(b)
≤ 3000; 0 ≤ c_sub ≤ 600.
P/Pc_sub ≤ 600.

The effect of the Young’s modulus ratios and coating thicknesses on the dimensionless contact
behaviors was already considered in detail for the purely elastic or 2% linear hardening elastic–plastic
asperity materials in the existing studies [15,16,19], and the corresponding variation trend has been
gotten for these materials. While for the power-law hardening substrate considered in this work,
the trend should be verified. Therefore, the effect of the Young’s modulus ratios and dimensionless
coating thicknesses on the dimensionless contact load P/Pc_sub and area A/Ac_sub are investigated,
and the results for some typical cases where the typical hardening exponents n = 0.1 and 0.9 are
revealed in Figure 6. Figure 6a,b shows that the variation of contact parameters for the cases where
the Young’s modulus ratios Eco /Esu = 1, 4, 8 when the dimensionless coating thickness t/R = 0.01.
It could be seen that for the coated asperity having a given n value, the contact load increases as
the Eco /Esu value increases at a given interference. In other words, the interferences were larger for
the coating system with smaller Eco /Esu values to get the same contact load. It means this kind of
Figure 5. The profiles of the outer surface before and after the normal loading process for some typical
cases where dimensionless coating thickness t/R = 0.01, Young’s modulus ratios Eco/Esu = 4 at different
Materials 2018, 11, 1965 7 of 13

coating system has higher load bearing capacity to resist plowing, which will lead to smaller friction
coefficient. This could help explain the experimental results shown in Figure 2 in Ref. [23], where the
TiN(coating)/WC-Co(substrate) system presented a lower friction coefficient in the running-in stage
since it had the lower elastic modulus ratio Eco /Esu than the TiN(coating)/Cu(substrate) system [23].
While at a given contact load, the contact area decreases with the increase of the Eco /Esu values.
This trend is in accord with the conclusions given in Ref. [16]. Figure 6c,d show the effect of the
dimensionless coating thickness t/R on the contact parameters. Some typical dimensionless coating
thicknesses t/R = 0.001, 0.007, 0.01, spanning an order of magnitude, are considered for the Young’s
modulus ratios Eco /Esu = 4. As shown in Figure 6c,d, the contact load increases as the coating thickness
becomes larger when the substrates are the same material. It is expected because the coating is harder
and more elastic than the substrate, and the same interference will lead to larger contact load for the
Figure 4. The relation between the dimensionless contact area A/Ac_sub and load P/Pc_sub for some
asperity with larger coating thickness. The contact area decreases with the increase of the coating
typical cases where dimensionless coating thickness t/R = 0.01, Young’s modulus ratios Eco/Esu = 4 at
thickness. It is probably
different hardeningbecause
exponents theof same contactn load
the substrate = 0.1, will leads
0.5, 0.7, 0.9. to
Theless yielding
results for thefor the asperities
purely elastic (n with
the thicker coating.
= 1) and elastic These results
fully-plastic (n =above accord
0) cases with
are also the(a)conclusions
given. 0 ≤ P/Pc_sub ≤given in 0Ref.
3000; (b) [15].
≤ P/P c_sub ≤ 600.

Figure
Figure 5. 5.
TheTheprofiles
profilesofofthe
theouter
outersurface
surfacebefore
before and after the
the normal
normalloading
loadingprocess
processfor forsome
sometypical
typical
cases
cases where
where dimensionless
dimensionless coatingthickness
coating thicknesst/R
t/R==0.01,
0.01,Young’s
Young’smodulus
modulus ratios
ratios E co/E
Eco /Esusu==4 4atatdifferent
different
hardening exponents of the substrate n = 0.1, 0.5, 0.9 under selected dimensionless contact loads
P/Pc_sub = 500, 1000. (a) P/Pc_sub = 500; (b) P/Pc_sub = 1000.

The effect of the hardening exponent of substrate on the change of the coating thickness during
the normal loading process is studied. Figure 7 shows the ratio of coating thickness before and after
loading, (t − ∆t)/t, changes with the dimensionless interference w/wc_sub for different hardening
exponent n = 0.1, 0.5, 0.7, 0.9 cases. The decrease of the coating thickness on the axis of symmetry
∆t could be calculated as the difference value between the interference of rigid flat w and the axial
displacement of the interface point on the axis of symmetry wsub , i.e., ∆t = w − wsub . The elastic (n = 1)
and elastic fully-plastic (n = 0) cases are also considered. The dimensionless coating thickness t/R are
set as 0.01 and Young’s modulus ratios Eco /Esu = 4. It could be seen from Figure 7 that for the asperity
with the elastic substrate (n = 1) or with the substrate whose hardening exponents are large and close
to 1 (such as n = 0.9), the coating thickness becomes smaller monotonically as the interference increases
in the loading process. For these cases, the rate of the variation of coating thickness (absolute value)
is very large at the very beginning of the contact, and then decreases gradually as the interference
increases until it approaches zero (see the case n = 0.9 in Figure 7). While for the asperity with the
substrate whose hardening exponents are smaller (such as n ≤ 0.7), the coating thickness also reduces
considerably in the early contact stage until reaches a minimum value, then it recovers gradually.
For smaller hardening exponent cases, the transition point appears earlier and the recovery of the
thickness is more obvious and rapid. Especially for the small n cases such as n = 0 and 0.1, the coating
thickness can recover to 99.9% of the initial value.
Materials 2018, 11, 1965 8 of 13
Materials 2018,
Materials 2018, 11,
11, xx FOR
FOR PEER
PEER REVIEW
REVIEW 88 of
of 13
13

Figure6.6.
Figure
Figure 6.The
Theeffect
The effectof
effect ofthe
of theYoung’s
the Young’s modulus
Young’s modulus ratios
modulus ratios and
ratios and dimensionless
and dimensionless coating
dimensionlesscoating thicknesses
coatingthicknesses
thicknesses onon
on the
thethe
dimensionless
dimensionless contact
contact
dimensionless contact load P/Pc_subload
load P/P
P/P c_sub
c_sub and
and area
area A/A
A/A at
at different
different hardening
hardening exponents
exponents of
of
A/Ac_sub at different hardening exponents of the substrate
c_sub
c_sub the
the substrate
substrate
n ==0.1,
n n= 0.1,0.9,
0.1, 0.9,where
0.9, where (a)
where (a) P/P
P/Pc_sub vs. w/w
c_sub vs. w/w and and
c_sub and (b) A/A
A/A vs. w/w
w/wc_sub are for
for the
the cases
cases where the Young’s
Young’s
P/P c_sub vs. w/w c_sub (b) A/A
(b)c_sub vs.
c_sub vs. are
w/w c_sub are forwhere the
the cases where the
c_sub c_sub c_sub

modulus
modulus
Young’s ratios
modulus E /E = 1,
co/Esu =E1, 4,
ratios Eratios
co su 4, 8 and
co /E8suand
the
= 1,the dimensionless
4, dimensionless coating thickness
coating thickness
8 and the dimensionless t/R
coatingt/R = 0.01; (c)
= 0.01; (c)
thickness t/RP/P
P/P vs. w/w
0.01;vs.(c)w/w
c_sub
= c_sub c_sub
P/Pc_sub
c_sub
and
vs.and
w/w (d) A/A
(d) A/Ac_sub vs. w/w
c_sub vs. w/wc_sub are for
c_sub are for the
the cases
cases where
where the the dimensionless
dimensionless coating
coating thickness
thickness t/R t/R == 0.001,
0.001,
c_sub and (d) A/Ac_sub vs. w/wc_sub are for the cases where the dimensionless coating thickness
t/R0.007,
0.007, 0.010.007,
0.01
= 0.001, and the
and the
0.01 Young’s
Young’s
and the modulus
modulus ratios
Young’sratios modulus /Esu
EEcoco/E = 4.
su = 4. E /E = 4.
ratiosco su

Figure7.7.
Figure
Figure 7.The
Theratio
The ratio of
ratio of coating
of coating thickness
coating thickness before
thickness before and
and after
after loading
after loading (t
loading (t ∆t)/tchanging
(t −−−Δt)/t
Δt)/t changing
changingwithwith
with the
thethe
dimensionless interference w/wc_sub for typical cases where hardening exponent n = 0.1, 0.5,for
dimensionless
dimensionless interference
interference w/w
w/w c_sub
c_sub for
for typical
typical cases
cases where
where hardening
hardening exponent
exponent nn =
= 0.1,
0.1, 0.5,
0.5, 0.7,
0.7, 0.9
0.9 for0.7,
dimensionless
0.9dimensionless
for dimensionlesscoating
coating thickness
thickness
coating t/R == 0.01
t/R
thickness 0.01 and
t/R and
= Young’s
Young’s
0.01 modulus
modulus
and Young’s ratios
modulus co /Esu
ratios EEratios
co/E =E4.
su = 4. The
The
co /E
elastic
su elastic
= 4. The (nelastic
(n == 1)
1)
and
(nand elastic
= 1)elastic fully-plastic
fully-plastic
and elastic (n == 0)
(n
fully-plastic 0)(n
cases
cases are
= 0)are alsoare
also
cases considered.
considered.
also considered.
Materials 2018, 11, 1965 9 of 13
Materials 2018, 11, x FOR PEER REVIEW 9 of 13

The distribution and intensity of the contact stress is closely related to the possible failure modes
in the coated asperity contact, and the effect of the substrate hardening exponent on the stress should
investigated. Figure
be investigated. Figure 88shows
showsthe therelation
relationbetween
betweenthe thedimensionless
dimensionless maximum
maximum vonvon Mises
Mises stress
stress in
in the
the coating
coating σmax_co
σmax_co /Y/Y0_sub the dimensionless
0_sub and the dimensionlessinterference
interferencew/w w/wc_sub
c_sub for
for some
some selected hardening
exponents of the the substrate
substrate nn==0.1, 0.1,0.5,
0.5,0.7,
0.7,0.9.
0.9.The The dimensionless
dimensionless coating
coating thickness
thicknesst/R t/R
= 0.01, and
= 0.01,
Young’s
and modulus
Young’s modulusratiosratios
Eco/Esu = /E
Eco 4. It
sucould
= 4. Itbecould
seen from Figure
be seen from8 Figure
that the8maximum stress increases
that the maximum stress
as the interference
increases w/wc_sub becomes
as the interference w/wc_sub larger. Under
becomes the small
larger. Under interference, the effect of the
the small interference, thesubstrate
effect of
withsubstrate
the differentwith
n values on the
different maximum
n values on thestress in the coating
maximum stress in is the
negligible.
coatingThen the dimensionless
is negligible. Then the
interference w/w
dimensionless c_sub gets larger
interference ranging
w/wc_sub getsfrom
larger12ranging
to 40, thefrom maximum
12 to 40, the stress still shows
maximum to still
stress be slightly
shows
influenced
to be slightlybyinfluenced
the n values. by While the interference
the n values. While thew/w c_sub is above
interference w/w40, the
c_sub effect
is above of the
40, n
the value
effect on
of the
maximum
n value on stress could notstress
the maximum be omitted,
could not and belarger n values
omitted, andoflarger
the substrate
n valuesresultof theinsubstrate
larger maximum
result in
stresses
larger in the coating.
maximum stresses This is expected
in the coating. Thissinceisthe larger since
expected n values the means
larger nmore values elastic
means substrate, and
more elastic
the same and
substrate, interference
the samewill lead to larger
interference will lead maximum
to larger stress
maximum for the
stressasperities with these
for the asperities withkind of
these
substrates.
kind of substrates.

8. The
Figure 8. The relation
relation between
betweenthe
thedimensionless
dimensionlessmaximum
maximumvon vonMises
Misesstress
stress σmax_co/Y
σmax_co /Y0_sub
0_sub in the

coating and
and the
the dimensionless interference w/w
dimensionless interference for some
c_sub for
w/wc_sub some typical
typical cases where dimensionless
dimensionless
thickness t/R
coating thickness t/R == 0.01, Young’s
Young’s modulus ratioEEcoco/Esu
modulusratio = 44 at
su = at selected
selected hardening
hardening exponents of the
substrate n = 0.1, 0.5, 0.7, 0.9.

An
An inflection
inflection point,
point, where
where the the slope
slope of the increase
of the increase ofof maximum
maximum stresses
stresses changes,
changes, appears
appears when when
the value of w/w
the value of w/wc_sub approximately equals to 9 for all n cases as revealed in
approximately equals to 9 for all n cases as revealed in Figure 8. All curves for
c_sub Figure 8. All curves
for different
different n cases
n cases consolidate
consolidate to a single
to a single curvecurve
whenwhen the interferences
the interferences arethan
are less lessthe
than the inflection
inflection point.
point. After the inflection point, the curves separate from each other, and
After the inflection point, the curves separate from each other, and the maximum stress in the coating the maximum stress in the
coating for larger n cases increases more quickly. To study the reason of the appearance
for larger n cases increases more quickly. To study the reason of the appearance of the inflection point, of the inflection
point, the distribution
the distribution of theofequivalent
the equivalent von vonMises Mises stresses
stresses co /Y0_sub
σco/Yσ0_sub in thein coating
the coatingalongalongthe the
axisaxis
of
of symmetry (y axis) was considered at different interferences near the
symmetry (y axis) was considered at different interferences near the inflection point in Figure 9. It inflection point in Figure 9.
It shows that the inflection point appears approximately at dimensionless
shows that the inflection point appears approximately at dimensionless interference w/wc_sub interference w/w =
= 9 in
c_sub 9
in Figure
Figure 8, thus
8, thus thethe stress
stress distributions
distributions are are studied
studied for for
the the cases
cases wherewhere interferences
interferences range range
fromfromw=
w = 6w
6wc_sub c_sub to w = 25w
to w = 25wc_sub. c_sub . The geometrical and material parameters were
The geometrical and material parameters were set as those in Figure 8. The set as those in Figure 8.
The dimensionless
dimensionless vertical
vertical distance distance y/t changes
y/t changes from 0from
to 1, 0meaning
to 1, meaning the position
the position from the from the contact
contact surface
surface to the interface between the substrate and the coating. It could be
to the interface between the substrate and the coating. It could be seen from Figure 9 that for the caseseen from Figure 9 that for
the case w/w
w/wc_sub = 6 which = 6 which is below the inflection point, the maximum stress
c_sub is below the inflection point, the maximum stress appears approximately at the appears approximately
at the dimensionless
dimensionless vertical vertical
distance distance y/t a=position
y/t = 0.28, 0.28, a position
close toclose to the contact
the contact surface.surface. For the
For the case w/wcase
c_sub
w/w
= 12 and = 12 and 25 which is above the inflection point, the maximum stress
c_sub25 which is above the inflection point, the maximum stress appears at the interface between appears at the interface
the substrate and the coating (y/t = 1). While for the case w/wc_sub = 9 which is almost the inflection
point, the maximum stress appears at two positions: one is at the position y/t = 0.28 close to the contact
Materials 2018, 11, 1965 10 of 13

between the substrate and the coating (y/t = 1). While for the case w/wc_sub = 9 which is almost
Materials 2018, 11, x FOR PEER REVIEW
the
10 of 13
inflection point, the maximum stress appears at two positions: one is at the position y/t = 0.28 close to
the contact
area; area; is
the other theatother is at the interface.
the interface. As revealed
As revealed in Ref. in Ref.the
[16], [16], the equivalent
equivalent von Mises
von Mises stress
stress in
in the
the coating is influenced by two factors: One is the applied contact load and
coating is influenced by two factors: One is the applied contact load and the other is mismatch ofthe other is mismatch
of elastic
elastic modulus.
modulus. At the
At the position
position closeclose tocontact
to the the contact surface,
surface, the factor
the factor of contact
of contact load isload is crucial,
crucial, while
while
at theatinterface
the interface
betweenbetween the substrate
the substrate and and coating,
coating, the the factor
factor of the
of the elastic
elastic modulus
modulus mismatch
mismatch is
is dominant. The inflection point appears at the interference where the two
dominant. The inflection point appears at the interference where the two factors have the same factors have the same
contributions,
contributions, leading
leading toto two
two equal
equal maximum
maximum stressstress at
at different
differentpositions.
positions.

Figure 9. 9. The
The distribution
distribution of
ofthe
thedimensionless
dimensionlessequivalent
equivalentvon
vonMises
Misesstresses
stressesσcoσco/Y
/Y0_sub
0_sub in the coating

along the axis of symmetry at at the

the different
different interferences
interferences near
nearthe
theinflection
inflectionpoint
pointw/w w/w c_sub = 6,=9,
c_sub 6,12,
9, 12,
25
25
forfor some
some typical
typical caseswhere
cases wheredimensionless
dimensionlesscoating
coatingthickness t/R==0.01,
thicknesst/R 0.01, Young’s
Young’s modulus ratios ratios
E /Esu
co/E
Eco su==44at
atselected
selectedhardening
hardeningexponents
exponentsof ofthe substratenn==0.1,
thesubstrate 0.1,0.5,
0.5,0.7,
0.7,0.9.
0.9.

It could be
It could be seen
seenfrom
fromFigure
Figure99that thatthe
thedistribution
distribution and
and maximum
maximum value
value of the
of the stresses
stresses along
along the
the
axisaxis of symmetry
of symmetry is little
is little affectedaffected
by theby the hardening
hardening exponent
exponent when thewhen the interference
interference is small,issuch
small,as
such as w/w
w/wc_sub less than less than 9. While as the dimensionless interference
c_sub9. While as the dimensionless interference w/wc_sub increases from w/w increases from
c_sub 12, the distribution 12,
the distribution
of the of the
stress shows to stress showsby
be affected to the
be affected
values of bythe
thesubstrate
values ofhardening
the substrate hardening
exponents to exponents
some extent. to
some extent.
A notable A notable phenomenon
phenomenon is that as the is that as the interference
interference increases,increases,
the stresstheatstress at the position
the position close toclose
the
to the contact surface decreases to zero, which means a low stress zone appears.
contact surface decreases to zero, which means a low stress zone appears. This phenomenon accords This phenomenon
accords
with thewith the conclusion
conclusion given
given in Ref. in To
[19]. Ref. [19].this
show To phenomenon
show this phenomenon
more clearly, more
Figure clearly, Figurethe
10 presents 10
presents the stress distribution
stress distribution along the axis along the axis ofinsymmetry
of symmetry the coating in for
thethe
coating
casesfor
w/w the = 40w/w
cases
c_sub and c_sub = 40
70. From
and 70. From
Figures 9 andFigures 9 andalso
10, it could 10, be
it could
foundalsothatbeasfound that as the increases,
the interference interferencetheincreases,
low stress the lowmoves
zone stress
zone
from moves from very
the position the position
close to very close tosurface
the contact the contact surface
(e.g., w/w (e.g.,
c_sub = w/w
12) to thec_sub
deeper= 12) to the deeper
position. Under
position.
the same interference, the low stress zone is closer to the contact surface for the larger n cases. larger
Under the same interference, the low stress zone is closer to the contact surface for the
n cases.
 Ma teria
Materials ls 2018,
2018, 11, x FOR PEER REVIEW
11, 1965 1113
11 of of 13
Materials 2018, 11, x FOR PEER REVIEW    11 of 13 

F 

Figure
Figure 10. 10.
TheThe stredistribution
stress ss distribution σco0_sub
σco /Y /Y0_sub
ininthethe coating
coating along
along thetheaxisaxis of symmetry
of symmetry in the
in the coating
coating
Figure 10. The stress distribution σ
nearnear
the the
lowlow stress
stress zonezone
at the co/Y0_sub in the coating along the axis of symmetry in the coating 
at dimensionless
the dime nsionlessinterferences w/ww/w
interferences = 40
= 40
c_sub andand70 70 n =n0.1,
forfor = 0.1,
0.5,0.5,
0.7,0.7,
0.9 0.9
c_sub
near the low stress zone at the dimensionless interferences w/w
cases
cases whenwhe n dime nsionless
dimensionless coating
coating thickne
thickness t/Rss= t/R = 0.01,
0.01, Young’s c_sub = 40 and 70 for n = 0.1, 0.5, 0.7, 0.9 
Young’s
modulus modulus Eco /EsuEco=/E
ratiosratios = 4.w/w
4.su(a) w/wc_sub
(a) c_sub
cases when dimensionless coating thickness t/R = 0.01, Young’s modulus ratios E
= 40;
= 40; (b)(b)
w/ww/w c_sub==70.
70. co/Esu = 4. (a) w/wc_sub 
c_sub
= 40; (b) w/wc_sub = 70. 
4. Conclusions
4. Conclusions
4. Conclusions 
The Thenormal
normal loading
loading process
process of of
thethecontact
contact between
between a rigid
a rigid flatflat
and a coated
and a coated asperity
asperity was was
The 
considered normal 
by theloading 
FE process 
method in of 
thisthe 
work.contact 
The between 
asperity a  rigid 
consisted flat 
considered by the FE method in this work. The asperity consisted of the purely elastic coating and ofand 
the a  coated 
purely asperity 
elastic was 
coating
considered by the FE method in this work. The asperity consisted of the purely elastic coating and 
and thepower-law
the power-lawhardening hardeningelastic–plastic
elastic–plastic substrate.
substrate. The The effect
effect ofofthe
thehardening
hardening exponent
exponent of of
thethe
the  power‐law 
substrate n on thehardening 
contact elastic–plastic 
behaviors was substrate. 
studied in a The 
wide effect 
range of 
of the  hardening 
interferences.
substrate n on the contact behaviors was studied in a wide range of interferences. It revealed that It exponent 
revealed of 
that the 
thethe
substrate n on the contact behaviors was studied in a wide range of interferences. It revealed that the 
behaviors,
behaviors, including
including contact
contact load,
load,contact
contact area, coating
area, coatingthickness
thicknessvariation,
variation,von Mises
von stress
Mises stress in the
in the
behaviors, including contact load, contact area, coating thickness variation, von Mises stress in the 
coating, were significantly influenced by the hardening exponent of the
coating, were significantly influenced by the hardening exponent of the substrate. Some conclusions substrate. Some conclusions
coating, were significantly influenced by the hardening exponent of the substrate. Some conclusions 
areare
listed as follows:
listed as follows:
are listed as follows: 
(1) (1)The The contact
contact loaddecreases
load decreaseswith withthe the decrease
decrease of of the
thehardening
hardeningexponentexponentofofthe substrate
the substrate n from
n
(1) The contact load decreases with the decrease of the hardening exponent of the substrate n from 
from1 to 0 at a given interference, meanwhile the contact radius and area enlarges at a special contact
1 to 0 at a given interference, meanwhile the contact radius and area enlarges at a special
1 to 0 at a given interference, meanwhile the contact radius and area enlarges at a special contact 
load. It
contact is because
load. the substrate
It is because (i.e., the
the substrate asperity)
(i.e., becomes
the asperity) more plastic
becomes more plastic n value
as the as the n becomes
value
load. It is because the substrate (i.e., the asperity) becomes more plastic as the n value becomes 
smaller.
becomes smaller.
(2) (2)smaller. 
AsAs thethe interference
interference increases,
increases, thethe coating
coating thickness
thickness becomes
becomes smaller
smaller monotonically
monotonically forfor larger
larger
(2) As  the  interference 
hardening
hardening exponents
exponents increases, 
such
such nthe 
as as =coating 
=n0.9 0.9
oror thickness becomes 
larger.
larger. While
While forfor
thethe smaller 
smaller
smaller nmonotonically for 
cases,
n cases, thethe coating
coating larger 
thickness
thickness
hardening exponents such as n = 0.9 or larger. While for the smaller n cases, the coating thickness 
recovers gradually after the minimum value. The mechanism should be studied in detail in the
recovers gradually after the minimum value. The mechanism should be studied in detail in the 
future.
future. 
Materials 2018, 11, 1965 12 of 13

recovers gradually after the minimum value. The mechanism should be studied in detail in
the future.
(3) The maximum von Mises stress increases as the interference increases. A inflection point appears
at the interference where two equal maximum stress values occur at different positions. As the n
value gets larger, the maximum stress enlarges especially at the moderate and large interference.
In addition, the low stress zone appears in the coating during the loading process, and the zone
becomes larger as the interference increases. For larger n value cases, the low stress zone appears
closer to the contact area.

Author Contributions: Data curation, X.L. and H.X.; Project administration, B.Z.; Writing—original draft, X.L.
Funding: This work was supported by the National Natural Science Foundation of China (51809057) and Marine
Low Speed Engine Project—Phase I (Grant No. CDGC01-KT11).
Acknowledgments: The authors would like to acknowledge the helpful discussions with Xuan Ma and
Xiujiang Shi.
Conflicts of Interest: The authors declare no conflicts of interest.

Nomenclature
A contact area
Ac_sub critical contact area of the substrate
Cν parameter related to ν, Cν = 1.234 + 1.256ν
E Young’s Modulus
Eco Young’s Modulus of the coating
Esu Young’s Modulus of the substrate
n strain hardening exponent of the substrate
P normal contact load
Pc_sub critical load for yield inception of the substrate
R radius of the sphere
t Coating thickness
w interference under normal load
wc_sub critical interference for yield inception of the substrate
Y0_sub yield strength of the substrate
ν Poisson’s ratio
ξ strain
σ stress
σmax_co maximum von Mises stress in the coating

References
1. Sonntag, R.; Feige, K.; dos Santos, C.B.; Kretzer, J.P. Hard Chrome-Coated and Fullerene-Doped Metal
Surfaces in Orthopedic Bearings. Materials 2017, 10, 1449. [CrossRef] [PubMed]
2. Pogrebnjak, A.D.; Bagdasaryan, A.A.E.; Pshyk, A.; Dyadyura, K. Adaptive multicomponent nanocomposite
coatings in surface engineering. Phys. Uspekhi 2017, 60, 586. [CrossRef]
3. Grau, M.; Matena, J.; Teske, M.; Petersen, S.; Aliuos, P.; Roland, L.; Grabow, N.; Escobar, H.M.; Gellrich, N.C.;
Haferkamp, H.; et al. In vitro evaluation of PCL and P (3HB) as coating materials for selective laser melted
porous titanium implants. Materials 2017, 10, 1344. [CrossRef] [PubMed]
4. Zhuang, Q.; Zhang, P.; Li, M.; Yan, H.; Yu, Z.; Lu, Q. Microstructure, wear resistance and oxidation behavior
of Ni-Ti-Si coatings fabricated on Ti6Al4V by laser cladding. Materials 2017, 10, 1248. [CrossRef] [PubMed]
5. Brizmer, V.; Kligerman, Y.; Etsion, I. The effect of contact conditions and material properties on the elasticity
terminus of a spherical contact. Int. J. Solids Struct. 2006, 43, 5736–5749. [CrossRef]
6. Jackson, R.L.; Kogut, L. A comparison of flattening and indentation approaches for contact mechanics
modeling of single asperity contacts. J. Tribol. 2006, 128, 209–212. [CrossRef]
Materials 2018, 11, 1965 13 of 13

7. Pogrebnjak, A.D.; Bondar, O.V.; Abadias, G.; Ivashchenko, V.; Sobol, O.V.; Jurga, S.; Coy, E. Structural
and mechanical properties of NbN and Nb-Si-N films: Experiment and molecular dynamics simulations.
Ceram. Int. 2016, 42, 11743–11756. [CrossRef]
8. Coy, E.; Yate, L.; Kabacińska, Z.; Jancelewicz, M.; Jurga, S.; Iatsunskyi, I. Topographic reconstruction and
mechanical analysis of atomic layer deposited Al2 O3 /TiO2 nanolaminates by nanoindentation. Mater. Des.
2016, 111, 584–591. [CrossRef]
9. Chang, W.R. An elastic-plastic contact model for a rough surface with an ion-plated soft metallic coating.
Wear 1997, 212, 229–237. [CrossRef]
10. Greenwood, J.A.; Williamson, J.P. Contact of nominally flat surfaces. Proc. R. Soc. Lond. Ser. A 1966, 295,
300–319. [CrossRef]
11. McCool, J.I. Extending the Capability of the Greenwood Williamson Microcontact Model. J. Tribol. 2002, 122,
496–502. [CrossRef]
12. Liu, S.B.; Peyronnel, A.; Wang, Q.J.; Keer, L.M. An extension of the Hertz theory for three-dimensional coated
bodies. Tribol. Lett. 2005, 18, 303–314. [CrossRef]
13. Garjonis, J.; Kačianauskas, R.; Stupak, E.; Vadlūga, V. Investigation of contact behaviour of elastic layered
spheres by FEM. Mechanika 2009, 3, 5–12.
14. Yeo, C.D.; Katta, R.R.; Polycarpou, A.A. Improved elastic contact model accounting for asperity and bulk
substrate deformation. Tribol. Lett. 2009, 35, 191–203. [CrossRef]
15. Goltsberg, R.; Etsion, I. A universal model for the load–displacement relation in an elastic coated spherical
contact. Wear 2015, 322, 126–132. [CrossRef]
16. Goltsberg, R.; Etsion, I. Contact area and maximum equivalent stress in elastic spherical contact with thin
hard coating. Tribol. Int. 2016, 93, 289–296. [CrossRef]
17. Eid, H.; Chen, L.; Joshi, N.; McGruer, N.E.; Adams, G.G. Contact and Adhesion of Elastic-Plastic Layered
microsphere. In Proceedings of the STLE/ASME International Joint Tribology Conference, San Francisco,
CA, USA, 17–20 October 2010; pp. 341–343.
18. Goltsberg, R.; Etsion, I.; Davidi, G. The onset of plastic yielding in a coated sphere compressed by a rigid flat.
Wear 2011, 271, 2968–2977. [CrossRef]
19. Chen, Z.; Goltsberg, R.; Etsion, I. Plasticity evolution in a coated sphere compressed by a rigid flat. Tribol. Int.
2016, 98, 116–124. [CrossRef]
20. Zhao, B.; Zhang, S.; Wang, P.; Hai, Y. Loading–unloading normal stiffness model for power-law hardening
surfaces considering actual surface topography. Tribol. Int. 2015, 90, 332–342. [CrossRef]
21. Jackson, R.L.; Green, I. A finite element study of elasto-plastic hemispherical contact against a rigid flat.
J. Tribol. 2005, 127, 343–354. [CrossRef]
22. Zhao, B.; Zhang, S.; Keer, L.M. Spherical Elastic–Plastic Contact Model for Power-Law Hardening Materials
Under Combined Normal and Tangential Loads. J. Tribol. 2017, 139, 021401. [CrossRef]
23. Huang, X.; Etsion, I.; Shao, T. Effects of elastic modulus mismatch between coating and substrate on the
friction and wear properties of TiN and TiAlN coating systems. Wear 2015, 338, 54–61. [CrossRef]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).