Wear 450-451 (2020) 203262

Contents lists available at ScienceDirect

Wear
journal homepage: http://www.elsevier.com/locate/wear

Experimental analysis of contact fatigue damage using

fractal methodologies
Deepak K. Prajapati *, Mayank Tiwari
Department of Mechanical Engineering, Indian Institute of Technology, Patna, 801106, India

A R T I C L E I N F O A B S T R A C T

Keywords: Gears are used in high power density (power throughput/weight) machines for continuous power transmission.
Micropitting The efficiency and reliability of gears significantly depend on tribological performance. Micropitting is a type of
Macropitting rolling contact fatigue damage mechanism that occurs at the asperity scale in lubricated contacts and highly
Gear failure
prevalent in gears. A detailed surface characterization study is essential to get an insight into the incident of
Slide-to-roll ratio
Boundary lubrication
micropitting. In this work, factors affecting resistance to micropitting are investigated by performing series of
Fractal signature twin disk experiments. The effect of disc geometry and slide-to-roll ratio on resistance to micropitting is discussed
in detail. It is observed that micro pits coalesces occur when components run for very large number of fatigue
cycles. A positive correlation is found between autocorrelation length and surface damage. Rose plot and area-
scale fractal analysis is employed on damaged surfaces, and it is concluded that there exists a certain scale, below
which fractal theory is applicable.

the continuing increase in operational cycles leads to growth in micro

pits, which causes high vibration and noise in machines, and finally,
1. Introduction breakdown occurs. Micropitting can initially appear on gears and grow,
but it also sometimes stagnates after the running-in or even disappears
Gears are used to transmit power through rolling/sliding motion [7]. The first incidence of micropitting in non-conformal concentrated
between meshing gear teeth. For continuous transmission of power, contacts was addressed by Way [8]. He showed the relation between
gears are expected to withstand for longer life with minimum friction micropitting and the surface finish [8]. Al-Tubi et al. [9] performed
and negligible wear. The tribological conditions in lubricated and highly rolling contact fatigue tests using gear test rig to study the micropitting
loaded case-hardened gears have been researched for several decades initiation and propagation under various loading and constant speed
[1]. Mostly, case-hardened gears are ground so that the grinding grooves conditions. They concluded that micropitting initiates at the pinion
dominate the surface characteristics. Case-hardened and ground gears dedendum but escalates at the addendum. A design of experiments
are supposed to operate in elastohydrodynamic lubrication (EHL) based study has been performed by Oila and Bull [10] to study the effect
regime to achieve minimum wear. But what is achieved is boundary or of variables (contact load, slide-to-roll ratio, oil temperature, lubricant
mixed-elastohydrodynamic lubrication regimes [2–4], which is a strong type, hardness) on the resistance to micropitting. They concluded that
reason for micropitting to occurs in gears. Micropitting is a surface the initiation of surface/sub-surface cracks depends on contact load
contact fatigue failure that occurs on the scale of roughness due to re­ whereas, the propagation of surface/sub-surface cracks depends on the
petitive cyclic stresses during the rolling/sliding motion of slide-to-roll ratio [10]. Warren and Guo [11] performed rolling contact
non-conformal concentrated contacts. Micropitting is readily observed fatigue tests to demonstrate the effect of the machining process on
in all types of gear, but it can also occur in rolling element bearings [5]. bearing life and found that turned surface has a longer life (>84%) than
Due to the increasing demand for high mechanically efficient machines, the ground one for equivalent roughness. Al-Mayali et al. [12] per­
loads are increasing and surface speed is decreasing resulting in thinner formed series of twin disc tests under the mixed-lubrication regime to
fluid film, which leads to severe asperity-to-asperity contacts [6]. Due to study the micropitting initiation, and it was found that initial micro pits
direct metallic contacts, the severity of micropitting increases signifi­ form nearly at one million cycles of the slower disc. The effect of
cantly in such high power density machines. The initial appearance of slide-to-roll ratio (SRR) on the incidence of micropitting has been
micropitting on gears does not indicate permanent damage. However,

* Corresponding author.
E-mail address: [email protected] (D.K. Prajapati).

https://doi.org/10.1016/j.wear.2020.203262
Received 14 October 2019; Received in revised form 24 February 2020; Accepted 7 March 2020
Available online 12 March 2020
0043-1648/© 2020 Elsevier B.V. All rights reserved.
D.K. Prajapati and M. Tiwari Wear 450-451 (2020) 203262

Nomenclature hmin minimum film thickness, μm

λin specific film thickness or lambda ratio
ASFA Area-scale fractal analysis Sq composite RMS roughness, μm
SRC Smooth-rough crossover Sq1 RMS roughness of surface 1, μm
H Fractal signature or Hurst coefficient Sq2 RMS roughness of surface 2, μm
M (%) percentage of micro pit area Us sliding speed, m/s
FN normal applied load, N Uhigh speed of the lower disc, m/s
pmax maximum Hertzian pressure, GPa Ulow speed of the upper disc, m/s
a semi-major axis, μm Str texture aspect ratio
b semi-minor axis, μm Ssc mean summit curvature, μm 1
SRR slide-to-roll ratio, % Sal autocorrelation length, μm
Tin oil-inlet temperature, 0C Sku kurtosis
Ur entrainment or rolling speed, m/s

investigated by Cen et al. [13]. They observed lesser micropitting method are used to analyze the surface texture and fractal nature of
damage at higher SRR due to wiped off of asperities at higher SRR. Asi contact fatigue damaged surfaces.
[14] reported that fatigue damage in helical gears occurs primarily due
to fatigue crack initiation. Pariente et al. [15] utilized X-ray diffractive 2. Experimental methodology
techniques on rolling contact fatigue tested shot-peened gear to measure
the intensity of residual stresses by measuring the intensity of the X-ray First two sections (Sec. 2.1 and 2.2) describe the sample preparation
beam and width of the peak. Dimitrov et al. [16] obtained higher surface and the working of the set-up used for surface contact fatigue tests. The
durability for thermally treated shot-peened gears. An increase in rationale behind selecting test variables and detail experimental pro­
micropitting life with an increase in the percentage of retained austenite cedure is given in Sec. 2.3. The detail on surface characterization tech­
(RA) has been reported by Roy et al. [17]. The effect of the content of niques utilized in this work for micropitting observations are given at
dissolved water on wear, corrosion, and contact fatigue behavior was the last of this section (Sec. 2.4).
addressed by Soltanahmadi et al. [18]. They concluded that dissolved
water levels exhibit a negative effect on micropitting life. 2.1. Twin disc test rig
Morales-Espejel et al. [19] showed from the numerical simulation as
well as from triple-disc experiments that slide-to-roll ratio significantly Micropitting tests are performed using a power-recirculating Plint
affects micropitting and surface crack propagation increases with an 74S two roller machine (Phoenix Tribology Ltd., UK), which is designed
increase in the slide-to-roll ratio [19]. They also concluded that both to perform rolling/sliding tests in the line contact configuration with
fatigue wear and mild wear occur in the condition of low specific film simplified kinematics. Previously, it has been used by various re­
thickness, and the outcome in the form of surface damage significantly searchers to study wear and rolling contact fatigue mechanisms [31,32].
depends on specific film thickness, load, surface topography, and sliding Fig. 1 shows a schematic of the main components of the 74S two roller
speed (or slide-to-roll ratio). An exhaustive review of the mechanism of machine and disc contact configuration. The machine consists of two
rolling contact fatigue has been presented by Liu et al. [20]. They geometrically identical discs driven by two independently motors
concluded that the mechanism of rolling contact fatigue significantly (master and slave) to impose the fixed slide-to-roll (SRR) ratio between
depends on surface topography, running-in, contact load, specific film discs. The speed of the upper disc is lower than the speed of the lower
thickness, friction modifiers, and the slide-to-roll ratio [20]. It is known disc, which results in more damage on the upper disc [32]. The loading
that micropitting frequently occurs in off-highway vehicles (OHV’s) due is achieved by using a pneumatic jack, and friction force is measured
to low-speed and high-load conditions. Recently, Terrin and Meneghetti during the test from the axial torque of the lower shaft. A lubricant
[21] performed four square tests (FST) for studying pitting life of reservoir is used for heating of the lubricant, and the temperature of the
sun-planet gearbox of off-highway vehicles (OHV’s) on both lubricant reservoir is measured using a K-type thermocouple. The
shot-peened and un-peened 17NICrMo6-4 steel and no significant lubricant oil passes to the oil shaft bearings and the inlet of the discs, and
improvement found in fatigue strength of the gear material. Due to the the temperature at the inlet of the disc is measured using a K-type
development of surface characterization tools, the current trend in thermocouple. More details on the working on this machine can be
tribological research has been directed towards the establishment of the found elsewhere [31,32]. The twin-disc test is chosen as the experi­
correlation between surface topography parameters and contact fatigue mental method to simulate conditions which are found in real gear
damage mechanisms [22–24]. Various attempts have earlier been made testing.
to study the evolution of topography parameters with fatigue cycles.
However, those studies were restricted to the evolution of topography 2.2. Material and sample preparation
parameters during the running-in of lubricated rolling/sliding contacts
[25–30]. In this work, a detail investigation is carried to study the The discs used for micropitting tests are made of AISI 9310 alloy steel
change in surface topography during surface contact fatigue tests. Ex­ extensively used for the manufacturing of gears [10]. The mechanical
periments are performed using a twin-disc tribometer under boundary properties of AISI 9310 steel are shown in Table 1. The sample discs
lubrication regime. The topography of the surface is measured using a (upper and lower) consist of 40 mm diameter and lower disc thickness is
white light interferometer (WLI), and essential texture parameters are 12 mm. Upper disc is crowned to a radius of 20 mm or 50 mm. It should
determined. A brief description of topography parameters used in this be noted that in the basic design of twin disc machine, the standard
work can be found elsewhere [24]. Surface damage quantifies in terms specimen is two flat discs, which results in line contact configuration. In
of the percentage of micro pit area (ratio of the total area of micro pits to this work, upper disc geometry is modified, which produces the elliptical
the nominal area). The effect of geometry and slide-to-roll ratio on contact configuration (see Fig. 2). The lower disc is flat, and no modi­
surface damage and topography parameters is also discussed in detail. In fication is made in the geometry of lower disc. It can be seen from Fig. 2
this work, the power spectral density of the surface and area-scale fractal that the minor axis (b) is aligned in the rolling/sliding direction which is

2
D.K. Prajapati and M. Tiwari Wear 450-451 (2020) 203262

Fig. 1. Schematic of Plint 74S two roller machine and disc contact configuration.

2.3. Experimental procedure

Table 1
Mechanical properties of AISI 9310 steel.
The test procedure is designed to study the resistance to micropitting,
Properties Value keeping variables as few as possible to simplify micropitting tests. To
Yield strength, σy 931 MPa achieve this, contact load (FN), oil-inlet temperature (Tin), entrainment
Elastic modulus, E 200 GPa
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Poisson ratio, ν 0.29 speed (Ur), surface roughness (Sq ¼ S2q1 þ S2q2 ), minimum film thick­
Heat capacity, α 0.472 (J/kg 0C) ness (hmin), and lambda ratio (λ ¼ hmin =Sq ) are kept constant for each
Thermal conductivity, k 51.9 (W/mK)
test. The test variables and their values are shown in Table 2. Test 1 and
2 are performed to see the effect of the disc geometry on the incident of
micropitting. Test 3 is performed to see the effect of slide-to-roll ratio on
the incident of micropitting. At the end of each test, slower disc (upper)
completes 6 million cycles, and faster disc (lower) completes 6.12
million cycles and 6. 62 million cycles for the value of the slide-to-roll
ratio (SRR) ¼ 0.05 and 0.1 respectively. The definition of SRR
(slide-to-roll ratio) and minimum film thickness (hmin) is given in Ap­
pendix A and B. After completion of each test, discs are allowed to cool
for 8 h before surface measurements. It is done to avoid any error due to
the thermal expansion of the test discs. Micropitting experiments are
repeated twice to ensure the repeatability in topography parameters and
the average value of topography parameters is reported. The repetition
of micropitting tests is shown in Appendix C (see Table C).

2.4. Surface characterization techniques

It is known that systematic characterization of worn surfaces can

provide a better understanding of wear mechanisms. In this work, worn/
unworn surfaces are characterized by utilizing instruments such as 3D
non-contact optical profilometer, light and scanning electron
microscope.

2.4.1. 3D optical profilometer

For topography measurement, 3D optical profilometer (model: MFT
500, Rtec Instruments, USA) with a 10x objective is used. The surface
topography is measured on three equally spaced locations on the disc
(see Fig. 3), and the average value of topography parameters is reported.
A surface measured using a 10x objective consists of 1.9 mm� 1.6 mm
Fig. 2. Schematic of elliptical contact on disc surface (green arrow indicates scan area. The saved raw data file is post-processed in MountainMap®
the rolling/sliding direction). (For interpretation of the references to color in
this figure legend, the reader is referred to the Web version of this article.) Table 2
Test variables and their values.
consistent in this work. Cylindrical grinding produces a longitudinal lay Test Load SRR a b Tin Ur λin Rcr
along the circumferential direction of the disc are used to finish disc No. (N) (μm) (μm) (0C) (m/s)
samples. Sample discs are case-hardened up to a case-depth of 0.92 mm 1 500 0.1 350 220 60 3 0.16 20
(as per manufacturer certification). The Vickers hardness of mm
case-hardened discs is found to be 749 Hv. Micropitting tests are per­ 2 500 0.1 690 240 60 3 0.16 50
formed using API Group I base oil. Further detail on sample preparation, mm
3 500 0.05 690 240 60 3 0.16 50
lubricant properties, and heat-treatment can be found elsewhere [24].
mm

3
D.K. Prajapati and M. Tiwari Wear 450-451 (2020) 203262

and crowning is done only on upper discs. The experimental procedure

and test conditions are previously described in Sec. 2.3. Fig. 6a- 6b
represent the evolution of surface damage after 6 million cycles for two
different values of crown radius. It can be seen from Fig. 6a–b that
number of pits increases with an increase in crown radius (Rcr). As
illustrated in Fig. 6a maximum pit depth is 10 μm for the disc with 20
mm crown radius, which is much more than the maximum pit depth
(6.4 μm) observed for the disc with 50 mm crown radius (see Fig. 6b).
This indicates the formation of deep craters due to very high plastic
deformation of asperities for lower crown radius. Initially, the maximum
contact pressure (pmax) is very high about 1.9 GPa for the disc with lower
Fig. 3. Schematic for surface topography measurement [24]. crown radius (Rcr ¼ 20 mm) resulting in very high plastic deformation.
Very high plastic deformation of asperities facilitates more cracks
7.0 [33] to determine the essential topography parameters and the nucleation sites as a result very deep craters are formed resulting in
percentage of the micro pit area. It is expected to obtain a cylindrical higher surface damage for lower crown radius disc as can be seen in
feature during the roughness measurement due to the geometry of the Fig. 6a. It can also be seen from Fig. 6 that for the disc with lower crown
sample disc. A fourth-order polynomial is used to remove the cylindrical radius, roughness peaks heights reduce significantly (color map of
feature [24]. Table 3 represents the calculated surface topography pa­ Fig. 6a) in comparison to the disc with higher crown radius indicating
rameters for unworn upper and lower discs. Fig. 4a–b represent surface the significant removal of asperities. Image at the right side of
map of unworn upper and lower disc respectively. The percentage of Fig. 6aandb is the greyscale image in which white regions represent
micro pit area (M) of unworn upper and lower discs is also shown in undamaged area, whereas, black regions represent the damaged area.
Fig. 4c–d. The procedure for calculating the percentage of micro pit area The surface damage that is quantified in term of percentage of micro pit
is given in Appendix D. Micro pits determine within the measured sur­ area (M) is found to decrease from 6.1% to 4.5% with an increase in the
face is highlighted with black dots in Fig. 4c–d. It can be seen that micro crown radius (Rcr) from 20 mm to 50 mm.
pits area is very less for unworn upper and lower disc. This initial micro To recognize the region of micropitting and macropitting, a corre­
pit area is subtracted from the micro pit area obtained after micropitting lation between the micro pits depth and micro pits diameter is estab­
tests, which gives the correct area of micro pits due to surface contact lished and represent in Fig. 7a- 7b. In this work, a pit is said to be a
fatigue. macro pit when it has pit diameter more than 50 μm. Fig. 7a–b represent
the correlation between pit depth and pit diameter for different values of
2.4.2. Light and filed-electron scanning electron microscope crown radius (Rcr). It can be seen from Fig. 7a–b that the region of
After completion of surface contact fatigue tests, test surfaces are micropitting increases with an increase in the crown radius (from 20 mm
prepared for metallographic investigations. The test surfaces are to 50 mm) due to an increase in initial contact width or contact area. As
observed periodically utilizing light (model: STM 6, Olympus, USA) and illustrated in Fig. 7b the depth of pits for the disc with higher crown
scanning electron microscope (SEM) without any special preparation. radius is limited to 7 μm. Whereas, for the disc with lower crown radius
However, prior to the observation, test surfaces are cleaned in ultrasonic (20 mm) maximum pit depth is 11 μm (see Fig. 7a) which is much more
bath for 15 min and dried by using hot air. The test surfaces under FE- than 50 mm crown radius disc. The region of micro and macro pits is also
SEM (Model-Gemini SEM 500, Zeiss, Germany) are observed in sec­ shown in Fig. 7a–b. As illustrated in Fig. 7a–b the number of macro pits
ondary electron and backscattered electron emission modes. Commonly, are less but deeper for 20 mm crown radius disc than macro pits
secondary emission modes have been used to inspect the test surfaces observed for 50 mm crown radius disc. For crown radius, Rcr ¼ 20 mm,
[10]. Fig. 5a–b represent SEM micrographs of unworn lower and upper macro pits diameter is found in the range of 50 μm–80 μm whereas, for
discs. Grinding ridges on unworn upper and lower disc can be seen in crown radius Rcr ¼ 50 mm, macro pit diameter lies in the range of 50
Fig. 5c–d. μm–225 μm. From Figs. 6 and 7 it can be inferred that surface damage
decreases with an increase in crown radius (Rcr).
3. Result and discussions
3.1.1. Surface morphology
3.1. Effect of disc geometry on resistance to micropitting Fig. 8a- 8b represent optical and SEM micrographs of slower (upper)
disc for 20 mm crown radius. Fig. 8c–d represent optical and SEM mi­
Gear teeth profiles are frequently modified by making crowning on crographs of slower disc after 6 million cycles for 50 mm crown radius.
gear teeth to reduce the edge stress concentration. In previously re­ SEM micrographs are taken in secondary electron emission mode at 100
ported works [19,34,35], different values of crown radius (Rcr) were μm scale. It can be seen from Fig. 8a–c that surface damage decreases
chosen to study the severity of micropitting in different operating con­ with an increase in crown radius. More intense surface damage is
ditions. In this section, the effect of geometry (crown radius) on surface observed for the disc with lower crown radius (Rcr ¼ 20 mm) as can be
damage is discussed in detail. This is examined in this section by per­ seen in Fig. 8c. Due to higher surface damage at lower crown radius, pits
forming micropitting tests on sample discs which are crowned to a of large depth (up to 12 μm) are observed as previously shown in Fig. 7a.
radius of 20 mm and 50 mm. It should be noted that lower discs are flat It can also be seen from Fig. 8d that sliding track width over which
surface damage is occurred also increases with an increase in crown
radius (Rcr) from 20 mm to 50 mm (see Fig. 8c–d). Fig. 9a–d represents
Table 3 SEM micrographs of the slower (upper) disc after 6 million cycles for 20
Calculated topography parameters for unworn upper and lower disc. mm and 50 mm crown radius at different scales (20 μm and 2 μm). As
Parameters Unworn upper disc Unworn lower disc illustrated in Fig. 9aandb the severity of surface damage is more in
Fig. 9a. Some pits in Fig. 9c appear deeper than pits observed in Fig. 9d.
Sq 0.457 μm 0.468 μm
Ssk 0.033 0.15
This may happen due to higher pits depth for the disc with lower crown
Sku 3.65 4.12 radius (see Fig. 7a) than pits depth for the disc with higher crown radius
Sal 4 μm 6 μm (see Fig. 7b). As illustrated in Fig. 9a–d surface cracks occurs for both 20
Str 0.03 0.098 mm and 50 mm crown radius discs.
Ssc 516 μm 1 521 μm 1

4
D.K. Prajapati and M. Tiwari Wear 450-451 (2020) 203262

Fig. 4. Surface map of (a) upper unworn disc (b) lower unworn disc.

Fig. 5. SEM micrographs of (a–b) upper and lower disc at 100 μm scale (c–d) upper and lower discs at 20 μm scale.

5
D.K. Prajapati and M. Tiwari Wear 450-451 (2020) 203262

Fig. 6. Surface map of (a) upper disc after 6 million cycles for crown radius (Rcr) ¼ 20 mm (b) upper disc after 6 million cycles for crown radius (Rcr) ¼ 50 mm, FN ¼
500 N, SRR ¼ 0.1, Tin ¼ 60 � C, λin ¼ 0.16, Ur ¼ 3 m/s.

Fig. 7. Correlation between micro pit diameter and micro pit depth after 6 million cycles (a) for slide-to-roll ratio (SRR) ¼ - 0.1 and crown radius (Rcr) ¼ 20 mm, pmax
¼ 1.96 GPa, λin ¼ 0.16, Tin ¼ 60 0C, Ur ¼ 3 m/s (b) for slide-to-roll ratio (SRR) ¼ - 0.1 and crown radius (Rcr) ¼ 50 mm, pmax ¼ 1.5 GPa, λin ¼ 0.16, Tin ¼ 60 0C, Ur ¼ 3
m/s.

3.2. Effect of slide-to-roll ratio (SRR) on resistance to micropitting 0.05 to 0.1. It can be seen from Figs. 10a–b that the density of micro
pits increases significantly with an increase in SRR. The reason is
It has been mentioned that slide-to-roll ratio (SRR) significantly af­ shearing between asperities which increases with an increase in SRR
fects micropitting [10,19,34,35]. However, limited discussion has been resulting in an increase in coalesces of micro pits. Coalesces of micro pits
provided on the effect of SRR on surface topography (how the lubrica­ lead to the formation of macro pits of larger size (depth and diameter)
tion performance of the surface changes during the incidence of than the size of the micro pits. The formation of macro pits is also ex­
micropitting) [10,19,34,35]. In this section, the effect of SRR on surface pected because tests are performed on case-hardened discs for 6 million
damage is discussed in detail. Fig. 10a- 10b represent the surface dam­ fatigue cycles at high contact load (1.5 GPa) and high SRR. It has been
age after 6 million cycles of slower disc with an increase in SRR from mentioned by Clarke et al. [5] that the diameter of micro pits typically

6
D.K. Prajapati and M. Tiwari Wear 450-451 (2020) 203262

Fig. 8. (a–b) Optical and SEM micrographs of upper disc for 20 mm crown radius after 6 million cycles (c–d) optical and SEM micrographs of upper disc for 50 mm
crown radius after 6 million cycles.

Fig. 9. SEM micrographs of upper disc (a) for 20 mm crown radius at 20 μm scale (b) for 50 mm crown radius at 20 μm scale (c) for 20 mm crown radius at 2 μm scale
(d) for 50 mm crown radius at 2 μm scale.

varies from 10 μm to 30 μm, whereas, the depth of micro pits typically experiences more sliding resulting in more surface damage. The region
varies from 5 μm to 10 μm. The percentage of micro pit area (M) is also of macro pits is also shown in Fig. 11aandb. As illustrated in
presented at the right side of Fig. 10a–b. It can be seen that micro pit Figs. 11aandb macropitting occurs for every value of SRR ( 0.05 and
area (M) increases from 3.5% to 4.5% as SRR increase from 0.05 to 0.1). The region of macropitting is also shown in Fig. 11a–b. It can be
0.1. Fig. 11aandb represent the correlation between micro pit depth seen from Fig. 11a that diameter of macro pits varies from 50 to 175 μm
and micro pit diameter for the different slide-to-roll ratio (SRR ¼ 0.05 for SRR ¼ - 0.05 and from 50 to 225 μm for SRR ¼ - 0.1. It can be said
and 0.1). It can be seen from Figs. 11a b that the region of micro­ from Fig. 11 that both macropitting and macropitting regions exist when
pitting increases with increase in side-to-roll ratio from 0.05 to 0.1. It discs run for high fatigue cycles (6 million cycles in this work). From
can also be observed from Figs. 11a–b that within micropitting region, Figs. 10 and 11 it can be inferred that surface damage increases with an
the number of micro pits significantly increase at a higher value of SRR increase in SRR.
( 0.1). An increase in the number of micro pits at higher SRR clearly
shows the effect of sliding on micropitting. At higher SRR, asperity

7
D.K. Prajapati and M. Tiwari Wear 450-451 (2020) 203262

Fig. 10. Evolution of surface damage on the upper disc after 6 million cycles (a) for slide-to-roll ratio (SRR) ¼ 0.05 (b) for slide-to-roll ratio (SRR) ¼ 0.1, (pmax ¼
1.96 GPa, λin ¼ 0.16, Tin ¼ 60 � C, Ur ¼ 3 m/s).

Fig. 11. (a) Correlation between micro pit diameter and micro pit depth after 6 million cycles, pmax ¼ 1.5 GPa, SRR ¼ -0.05, λin ¼ 0.16, Tin ¼ 60 0C, Ur ¼ 3 m/s (b)
correlation between micro pit diameter and micro pit depth after 6 million cycles, pmax ¼ 1.5 GPa, SRR ¼ -0.1, λin ¼ 0.16, Tin ¼ 60 0C, Ur ¼ 3 m/s.

3.2.1. Surface morphology asperities at slower rate. However, severity of micropitting is more for
Fig. 12a- 12d represent the optical and SEM micrographs of the higher SRR as can be seen in Fig. 12a. The reason is frequently breaking
slower disc for different slide-to-roll ratio (SRR) ¼ 0.1 and 0.05. It of roughness peaks breaks due to high sliding resulting in the formation
can clearly be seen from Figs. 12a–c that surface damage decreases as of dense micro pits on contacting surfaces. The similar observation has
slide-to-roll ratio decrease from 0.1 to 0.05. The occurrence of mild been reported in Ref. [10,19]. The severity of micropitting can also be
wear can also be clearly observed in Fig. 12a–c. The reason for the observed from SEM micrographs (see Fig. 12b–d), which shows the same
occurrence of mild wear is running of discs under boundary lubrication observations as seen from optical micrographs (Fig. 12a–c). SEM
regime. It can also be seen that after 6 million fatigue cycles, the severity micrograph of the upper disc after micropitting tests at higher magni­
of mild wear is more for lower SRR due to large removal of contacting fications is shown in Fig. 13. Figs. 13a–b represent the SEM micrograph

8
D.K. Prajapati and M. Tiwari Wear 450-451 (2020) 203262

Fig. 12. (a–b) Optical and SEM micrographs of upper disc after 6 million cycles for SRR ¼ 0.1 (c–d) optical and SEM micrographs of upper disc after 6 million
cycles for SRR ¼ 0.05.

Fig. 13. (a–b) SEM micrographs of upper disc after 6 million cycles for SRR ¼ 0.1 at 20 and 2 μm scale respectively (c–d) SEM micrographs of upper disc after 6
million cycles for SRR ¼ 0.05 at 20 and 2 μm scale respectively.

of the upper disc for SRR ¼ 0.1 observed under secondary electron d.
emission modes at 20 μm and 2 μm scales respectively. As illustrated in
Fig. 13a micro pits are formed within sliding track after 6 million fatigue
3.3. Change in topography parameters
cycles. It can also be seen from Fig. 13b that mild wear also occurs
resulting in the removal of material during rolling/sliding motion. Sig­
Fig. 14a-b represent the change in topography parameters after 6
nificant asperity-to-asperity contacts take place under boundary lubri­
million fatigue cycles for the different slide-to-roll ratio (SRR ¼ 0.05
cation regime resulting in the removal of material in form of mild wear.
and 0.1) and crown radius (Rcr ¼ 20 mm and 50 mm). The change in
Damaged surface region within the sliding track can be seen from
the percentage of micro pit area, M (%) is also shown in Fig. 14a for
Fig. 13a. At higher magnification (2 μm scale), the occurrence of micro
comparison with topography parameters. As illustrated in Fig. 14a root
and macro pits can be observed from Fig. 13b. Decrease in the
mean square (Sq) roughness increases with an increase in SRR. RMS
slide-to-roll ratio (SRR) from 0.1 to 0.05 results in decrease in
roughness is related to the statistical distribution of roughness peaks,
micropitting damage and surface cracks as can be seen from Fig. 13c and
and the change in value of RMS roughness provides only information

9
D.K. Prajapati and M. Tiwari Wear 450-451 (2020) 203262

Fig. 14. Change in surface topography parameters for upper disc before and after micropitting tests.

asperities resulting in an increase flatness of roughness peaks. The value

of kurtosis (Sku) gives information about ‘peakedness’ of roughness
peaks. It can be seen from Fig. 16b that kurtosis (Sku) decreases with an
increase in SRR. The highest value of kurtosis is obtained for the disc
with lower crown radius (Rcr ¼ 20 mm) indicates an adequate removal
of roughness peaks above the mean plane of surface. It can be said from
Fig. 16b that and mean summit curvature (Ssc) are negatively correlated
with micro pit area (M). It can be inferred from Fig. 16 that topography
parameters change significantly with the change in geometry and
operating conditions.

3.4. Use of fractal theory on damaged surfaces

3.4.1. Relation between fractal signature and surface damage

Hurst coefficient or fractal signature (H) has been considered as a
vital descriptor of fractal geometry, which indicates the variation in the
randomness of an object [36,37]. In this work, the Hurst coefficient is
determined for unworn and damaged surfaces according to the method
reported in Refs. [38]. Fig. 15 represents the bar diagram of fractal
Fig. 15. Change in micro pit area and fractal signature (H) before and after
signature (H) and micro pit area M (%) for different SRR and crown
micropitting tests.
radius. It can be seen from Fig. 15 that before micropitting test, the value
of fractal signature (H) is 0.31, indicating the presence of sharp rough­
about the change in asperity heights. It does not provide any idea about ness peaks. Sharp roughness peaks are removed during micropitting test
slope and curvature of the asperity that is highly important for better increasing the flatness of contacting surfaces. As a result, fractal signa­
understanding of surface contact fatigue. It is shown in Sec. 3.2 and 3.3 ture (H) increases with an increase in SRR from 0.05 to 0.1 for the
that micro pits of different sizes and depths are found which may change same value of the crown radius, Rcr ¼ 50 mm, as can be seen in Fig. 15.
the texture of the surface. As illustrated in Fig. 14a texture aspect ratio As illustrated in Fig. 15 for the same value of SRR ( 0.1), the fractal
(0 < Str <1), which is the ratio of shortest correlation length to longest signature (H) increases with a decrease in crown radius (Rcr) from 50
autocorrelation length increases with an increase in SRR indicates an mm to 20 mm. It can be inferred from Fig. 15 that Hurst coefficient (H) is
increase in isotropic nature of the damaged surface. It is also shown in positively correlated with surface damage.
Fig. 14a that the percentage of micro pit area increases with an increase
in SRR. As illustrated in Fig. 14a that for maximum surface damage is 3.4.2. Areal-scale fractal analysis (ASFA) of damaged surfaces
occurred for the disc with 20 mm crown radius. It can be said from Firstly, Brown et al. [39] proposed the theory of scale sensitive
Fig. 16a that texture aspect ratio (Str) and RMS roughness (Sq) are fractal analysis. They analyzed various rough surfaces over many scales
positively correlated with surface damage. The value of mean summit using the tilting method. ASFA were extensively applied on to determine
curvature (Ssc) provides an idea about sharpness or flatness of asperity the fractal nature of the unworn rough surfaces [40–42]. ASFA is based
peaks. As illustrated in Fig. 16b that for unworn disc, the value of mean on the hypothesis that real surfaces are not exactly flat and transition
summit curvature (Ssc) is highest, indicating the presence of sharp from smooth to rough is possible by varying the scale. Recently,
asperity peaks. Asperity removal during micropitting test results in in­ area-scale fractal analysis (ASFA) has been used by Prajapati et al. [43]
crease in the flatness of roughness peaks due to which value of mean to determine the complexity of the surfaces under dry sliding conditions.
summit radius (Ssc) decreases (see Fig. 16b). Autocorrelation length (Sal) In this section, ASFA is used to determine the complexity of the surfaces
provides spatial information about surfaces. It can be seen from Fig. 16b before and after micropitting tests. More detail on ASFA method can be
that autocorrelation length increases with an increase in SRR. A higher found in Prajapati et al. [43]. For area-scale fractal analysis, the
value of autocorrelation length indicates more flattened rough peaks. At maximum patch area is 987042 μm2, and the minimum patch area
higher sliding speed, surface experiences more shearing between (1.939 μm2) is the square of the sampling interval divided by two.

10
D.K. Prajapati and M. Tiwari Wear 450-451 (2020) 203262

Fig. 16. Area-scale plot of upper disc (a) before the test (unworn) (b) after 6 million cycles for SRR ¼ 0.05, Rcr ¼ 50 mm (c) after 6 million cycles for SRR ¼ 0.1,
Rcr ¼ 50 mm (d) after 6 million cycles for SRR ¼ 0.1, Rcr ¼ 20 mm.

Figs. 16a–d represent area-scale plots of slower (upper) disc before test,
for SRR ¼ 0.05 and Rcr ¼ 50 mm after 6 million cycles, for SRR ¼ 0.1
and Rcr ¼ 50 mm after 6 million cycles, and for SRR ¼ 0.1 and Rcr ¼ 20
mm after 6 million cycles respectively. It can be seen that the relative
area in Figs. 16a–d decreases with an increase in scale, which indicates a
decrease in the complexity of the surface with an increase in scale. It can
be seen from Fig. 16a that there is a steep change in area scale curve at a
particular instant when the relative area becomes nearly one. This point
in the area-scale plot is called smooth-rough crossover (SRC). A value
higher than SRC indicates that surfaces can be described by Euclidean
geometry. Whereas, a value lower than SRC indicates applicability of the
fractal theory. It can also be seen in Fig. 16a that the value of SRC before
the micropitting test is approximately 18.3 μm2. After micropitting test,
the value of SRC increases with an increase in SRR (see Fig. 16b–c). More
surface damage occurs at higher SRR resulting in increase in complexity
of the damaged surface. The highest value of SRC is obtained SRR ¼
0.1 and Rcr ¼ 20 mm as shown in Fig. 16d. It happens due to maximum
damage on the surface of less crowned disc (upper). Previously, the
fractal theory has been extensively used to analyze the self-similarity or
self-affine nature of worn surfaces and wear scars [39–43]. However,
none of those discussed the selection of scale over which the fractal
theory was applied. It can be inferred from Fig. 16 that for different
operating conditions, there exists a minimum scale below which fractal
theory is applicable, and it is true for both unworn and damaged
surfaces.

3.4.3. Analysis of surface texture using power spectral density Fig. 17. Rose plot of upper disc (a) before the test (unworn) (b) after 6 million
Fig. 17a- 17d represent rose plot slower (upper) disc, before test, cycles for SRR ¼ 0.05, Rcr ¼ 50 mm (c) after 6 million cycles for SRR ¼ 0.1,
after 6 million cycles for SRR ¼ 0.05, Rcr ¼ 50 mm, after 6 million Rcr ¼ 50 mm (d) after 6 million cycles for SRR ¼ 0.1, Rcr ¼ 20 mm.

11
D.K. Prajapati and M. Tiwari Wear 450-451 (2020) 203262

cycles for SRR ¼ 0.1, Rcr ¼ 50 mm, and after 6 million cycles for SRR ¼ micropitting is observed for discs with a lower crown radius and SRR.
0.1, Rcr ¼ 20 mm respectively. To construct a rose plot, a sample area However, discs with higher crown radius have shallow micropitting.
of 993 μm � 993 μm is selected, and digitized data points of each surface It is found that the density of macro pits depends on the operating
are saved in ASCII file format and post-processed in Matlab 2017b to condition such as slide-to-roll ratio, and magnitude of the crown
determine the rose plot for each surface. The method used to construct a radius.
rose plot can be found elsewhere [38]. It can be seen from Fig. 17a that 2. It is shown that topography parameters change significantly with the
surface texture is oriented along 900 (or 2700) before micropitting test change in geometry and operating conditions. The texture aspect
indicates the anisotropic nature of the unworn slower disc. After 6 ratio (Str), autocorrelation length (Sal), and RMS roughness (Sq) are
million cycles, deterioration in surface texture occurs, leading to an positively correlated with surface damage. Whereas, a negative
increase in the value of texture aspect ratio (Str). It can be seen from correlation is found between mean summit curvature (Ssc) and sur­
Figs. 17a–b that the texture aspect ratio increase from 0.045 to 0.146, face damage.
indicating an increase in isotropy of the surface. An increase in isotropy 3. It is observed that surface damage is positively correlated with fractal
occurs due to the formation of craters on the surface. It can also be seen signature (H). For unworn/damaged surfaces, there exists a mini­
from Figs. 17b–c that deterioration in surface texture increases with an mum scale only below which fractal theory is applicable.
increase in SRR. It is interesting to see in Fig. 17d that surface texture is 4. Rose plots of damaged surfaces reveal the change in surface texture,
almost isotropic (Str > 0.3) for the surface with lower crown radius (Rcr and the highest deterioration in surface texture is found for the disc
¼ 20 mm). Deterioration in surface texture can also be verified from the with a lower crown radius. It is also observed that geometrical
distribution of slope (magnitude of the slope on a log-log plot of power isotropy increases with an increase in SRR.
spectral density vs. wave vector) in angular directions. It can be seen 5. It believes that the application of fractal methodologies for analyzing
from Figs. 17b–d that slope is not concentrated along particular direc­ pitted surfaces adds a new dimension to the quantitative study of
tion indicating the distribution of power spectral density (PSD) of surface topography. Understanding surface topography evolution
damaged surfaces in angular directions. over time may also help to improve the reliability and life of tribo­
logical components. It is part of ongoing work and will be discussed
4. Conclusions in further research.

The main findings of the present work can be summarized as follows:

Declaration of competing interest
1. Micropitting tests are performed under the boundary lubrication
regime to study the change in topography parameters and surface The authors declare that they have no known competing financial
damage for different operating conditions. It is concluded that within interests or personal relationships that could have appeared to influence
the range of fatigue cycles (6 million in this work), the severe the work reported in this paper.

Appendix

A. Definition of slide-to-roll ratio (SRR)

Slide-to-roll ratio (SRR) is defined as the ratio of sliding speed (Us) to mean rolling speed (Ur). In this work, SRR is mathematically defined as given
in Eq. (A1) Equations for determining sliding speed and mean rolling speed is given in Eq. (A2) and Eq. (A3).
sliding speedðUs Þ
SRR ¼ (A1)
mean rolling speedðUr Þ

sliding speed ðUs Þ ¼ Uslow Uhigh (A2)

Uhigh þ Uslow
mean rolling speed ðUr Þ ¼ (A3)
2

where, Uhigh is the speed of the lower disc, Uslow is the speed of the upper disc, Us is the sliding speed, Ur is the mean rolling speed or entrainment speed.
It should be noted that according to Eq. (A1) slide-to-roll ratio (SRR) is negative. Due to negative SRR, the upper disc which is slower experiences
negative slip, whereas, the lower disc which is faster experiences positive slip. The schematic of the negative and positive slip is shown in Fig. A1 [44].

12
D.K. Prajapati and M. Tiwari Wear 450-451 (2020) 203262

Fig. A1. Schematic of positive and negative slip [44

B. Calculation of minimum film thickness (hmin) for elliptical contact

The expression for determining the minimum film thickness (hmin) for elliptical contact configuration under heavily loaded condition is developed
by Hamrock [ [45], Eq. 19.16] and given in Eq. B1
hmin �
Hmin ¼ ¼ 3:63 � U0:68 � G0:49 � W 0:073
� 1 e 0:68k
(B1)
Rx

where, Hmin is the dimensionless minimum film thickness, hmin is the dimensional minimum film thickness, Rx is the reduced radius of curvature in x
direction (Rx < Ry), U is the dimensionless speed parameter, G is the dimensionless material parameter, W is the dimensionless load parameter, k (¼ a/
b) is the elliptic ratio. a and b are the semi-major and the semi-minor axis respectively.

C. Repetition of micropitting tests

Fig.C represent surface map after 6 million cycles of the slower disc for each micropitting test (Test 1, Test 2, Test 3). Test conditions of Test 1, Test
2, and Test 3 are given in Table 2 (see Sec. 2.3). Run 1 and Run 2 in Fig. Ca-Cb indicating that micropitting test is performed twice for the same
experimental conditions (Test 1). Similar can be seen in Fig. Cc-Cd, and Fig. Cd–Cf for Test 2 and Test 3, respectively. A satisfactory repetition in
surface topography parameters, percentage of micro pit area, and Hurst coefficient (H) can be seen from Table C.

13
D.K. Prajapati and M. Tiwari Wear 450-451 (2020) 203262

Fig.C. Surface map of (a) slower (upper) disc after 6 million fatigue cycles for Test 1, Run 1 (b) slower (upper) disc after 6 million fatigue cycles for Test 1, Run 2 (c)
slower (upper) disc after 6 million fatigue cycles for Test 2, Run 1 (d) slower (upper) disc after 6 million fatigue cycles for Test 2, Run 2 (e) slower (upper) disc after 6
million fatigue cycles for Test 3, Run 1 (f) slower (upper) disc after 6 million fatigue cycles for Test 3, Run 3

Table C
Repeatability in calculated topography parameters for each micropitting tests (Test 1, Test 2, Test 3)

Parameters Test No. 1 Test No. 2 Test No. 3

Run 1 Run 2 Run 1 Run 2 Run 1 Run 2

Sa 0.69 μm 0.73 μm 0.53 μm 0.65 μm 0.44 μm 0.42 μm

Sq 0.84 μm 0.86 μm 0.64 μm 0.62 μm 0.61 μm 0.55 μm
Sku 49.35 51.65 4.98 6.14 13.26 14.16
Sal 50.41 μm 52.79 μm 61.3 μm 55.5 μm 46.12 μm 44.08 μm
Ssc 154 μm 1 152 μm 1 168 μm 1 162 μm 1 151.2 μm 1
146.8 μm 1

Str 0.401 0.421 0.305 0.319 0.137 0.155

H 0.812 0.868 0.68 0.74 0.601 0.579
M (%) 5.89 6.31 3.95 5.05 3.8 3.2

D. Calculation of percentage of micro pit area, M (%)

To calculate the percentage of micro pit area (M), surface topography at particular load stage is measured using a white light interferometer (Rtec
Instruments, USA) and post-processed in MountainsMap® 7 [33] which is a surface characterization software. A surface defect is said to be a micro pit
if it has depth >1 μm and diameter in the range of 10–50 μm, below the mean plane of the surface [17,23]. To simplify the present analysis, it is
assumed that the shape of all existing micro pits is circular and the diameter of micro pits is calculated by using the area of a circle. An expression for
calculating the percentage of micro pit area is given in Eq. (D1)
total ​ area ​ of ​ micro ​ pit
Mð%Þ ¼ (D1)
sample ​ area

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.wear.2020.203262.

14
D.K. Prajapati and M. Tiwari Wear 450-451 (2020) 203262

References [22] W.P. Dong, K.J. Stout, An integrated approach to the characterization of surface
wear I: qualitative characterization, Wear 181–183 (1995) 700–716.
[23] S. Roy, D. White, S. Sundararajan, Correlation between evolution of surface
[1] G. Stachowiak, Engineering Tribology, second ed., Butterworth-Heinemann, UK,
roughness parameters and micropitting of carburized steel under boundary
2013 (Chapter 1).
lubrication condition, Surf. Coating. Technol. 350 (2018) 445–452.
[2] K. Holmberg, P. Andersson, A. Erdemir, Global energy consumption due to friction
[24] D.K. Prajapati, M. Tiwari, Assessment of topography parameters during running-in
in passenger cars, Tribol. Int. 47 (2012) 221–234.
and subsequent rolling contact fatigue tests, ASME J. Tribol. 141 (2019), 051401.
[3] K. Holmberg, P. Kivikyto-Reponen, P. Harkisaari, K. Valtonen, A. Erdemir, Global
[25] S. Akbarzadeh, M.M. Khonsari, On the optimization of running-in operating
energy consumption due to friction and wear in the mining industry, Tribol. Int.
conditions in applications involving EHL line contact, Wear 303 (2013) 130–137.
115 (2017) 116–139.
[26] J.J. Wagner, A.D. Jenson, S. Sundararajan, The effect of contact pressure and
[4] K. Holmberg, P. Andersson, N.O. Nylund, K. Makela, A. Erdemir, Global energy
surface texture on running-in behavior of case carburized steel under boundary
consumption due to friction in truck and buses, Tribol. Int. 78 (2014) 94–114.
lubrication, Wear 376–377 (2017) 851–857.
[5] A. Clarke, H.P. Evans, R.W. Snidle, Understanding micropitting in gears, Proc. Inst.
[27] L. Xiao, B.G. Rosen, N. Amini, P.H. Nilsson, A study on the effect of surface
Mech. Eng. C Mech. Eng. Sci.: J. Eng. Tribol. 230 (2016) 1276–1289.
topography on rough friction in roller contact, Wear 254 (2003) 1162–1169.
[6] G.E. Morales-Espejel, A. Gabelli, A model for gear life with surface and subsurface
[28] S. Akbarzadeh, M.M. Khonsari, On the prediction of running-in behavior in mixed-
survival: tribological effects, Wear 404–405 (2018) 133–142.
lubrication line contact, ASME J. Tribol. 132 (2010), 0321021-11.
[7] M. Weibring, L. Gondecki, P. Tenberge, Simulation of fatigue failure on tooth
[29] S. Akbarzadeh, M.M. Khonsari, Experimental and theoretical investigation of
flanks in consideration of pitting initiation and growth, Tribol. Int. 131 (2019)
running-in, Tribol. Int. 44 (2011) 92–100.
299–307.
[30] Y.R. Jeng, Z.W. Lin, S.H. Shyu, Change of surface topography during running-in
[8] S. Way, Pitting due to rolling contact, J. Appl. Mech. 57 (1935) A49–A58.
process, ASME J. Tribol. 126 (2004) 620–625.
[9] I.S. Al-Tubi, H. Long, J. Zhang, B. Shaw, Experimental and analytical study of gear
[31] P. Rabaso, T. Gauthier, M. Diaby, F. Ville, Rolling contact fatigue: experimental
micropitting initiation and propagation under varying loading conditions, Wear
study of the influence of sliding, load, and material properties on the resistance to
328–329 (2015) 8–16.
micropitting of steel discs, Tribol. Trans. 56 (2013) 203–214.
[10] A. Oila, S.J. Bull, Assessment of factors influencing micropitting in rolling sliding
[32] T. Ahlroos, H. Ronkainen, A. Helle, R. Parikka, J. Virta, S. Varjus, Twin disc
contacts, Wear 258 (2005) 1510–1524.
micropitting tests, Tribol. Int. 42 (2009) 1460–1466.
[11] W. Warren, Y.B. Guo, The impact of surface integrity by hard turning versus
[33] MountainsMap® 7–surface imaging and analysis software. www.digitulsurf.com.
grinding on rolling contact fatigue – part I: comparison of fatigue life and acoustic
[34] A. Clarke, I.J.J. Weeks, R.W. Snidle, H.P. Evans, Running-in and micropitting
emission signals, Fatig. Fract. Eng. Mater. Struct. 30 (2007) 698–711.
behavior of steel surfaces under mixed lubrication conditions, Tribol. Int. 101
[12] M.F. AL-Mayali, S. Hutt, K.J. Sharif, A. Clarke, H.P. Evans, Experimental and
(2016) 59–68.
numerical study of micropitting initiation in real rough surfaces in a micro-
[35] S. Hutt, A. Clarke, H.P. Evans, Generation of acoustic emission from running-in and
elastohydrodynamic lubrication regime, Tribol. Lett. 66 (2018) 150.
subsequent micropitting of mixed-elastohydrodynamic contacts, Tribol. Int. 119
[13] H. Cen, A. Morina, A. Neville, Effect of slide to roll ratio on the micropitting
(2018) 270–280.
behaviour in rolling-sliding contacts lubricated with ZDDP-containing lubricants,
[36] Y. Xu, R.L. Jackson, Statistical models of nearly complete elastic rough surface
Tribol. Int. 122 (2018) 210–217.
contact-comparison with numerical solutions, Tribol. Int. 105 (2017) 274–291.
[14] O. Asi, Fatigue failure of a helical gear in a gearbox, Eng. Fail. Anal. 13 (2006)
[37] B.N. Persson, On the fractal dimension of rough surfaces, Tribol. Lett. 54 (2014)
1116–1125.
99–106.
[15] F. Pariente, M. Guagliano, Contact fatigue damage analysis of shot peened gears by
[38] D.K. Prajapati, M. Tiwari, The relation between fractal signature and topography
means of X-ray measurements, Eng. Fail. Anal. 16 (2009) 964–971.
parameters: a numerical and experimental study, Surf. Topogr. Metrol. Prop. 6
[16] L. Dimitrov, D. Michalopoulos, Ch Alk Apostolopoulos, T.D. Neshkov, Investigation
(2018), 045008.
of contact fatigue of high strength steel gears subjected to surface treatment,
[39] C.A. Brown, P.D. Charles, W.A. Johnsen, S. Chesters, Fractal analysis of
JMEPEG 18 (2009) 939–946.
topographic data by the patchwork method, Wear 161 (1993) 61–67.
[17] S. Roy, G.T.C. Ooi, S. Sundararajan, Effect of retained austenite on micropitting
[40] K. Zak, Areal field and fractal based characterization of hard surfaces produced by
behavior of carburized AISI 8620 steel under boundary lubrication, Materialia
different machining operations, J. Mach. Eng. 16 (2016) 24–32.
(2018) 1–10, 00.
[41] M. Bigerelle, D. Najjar, A. Lost, Multiscale functional analysis of wear a fractal
[18] S. Soltanahmadia, A. Morinaa, M.C.P. van Eijkb, I. Nedelcub, A. Neville,
model of the grinding process, Wear 258 (2005) 232–239.
Tribochemical study of micropitting in tribocorrosive lubricated contacts: the
[42] G.L. Goic, M. Bigerelle, S. Samper, H. Favreliere, M. Pillet, Multiscale roughness
influence of water and relative humidity, Tribol. Int. 107 (2017) 184–198.
analysis of engineering surfaces: a comparison method for the investigation of
[19] G.E. Morales-Espejel, P. Rycerz, A. Kadiric, Prediction of micropitting damage in
functional correlations, Mech. Syst. Signal Process. 66–67 (2016) 437–457.
gear teeth contacts considering the concurrent effects of surface fatigue and mild
[43] D.K. Prajapati, M. Tiwari, Tribological investigation of copper and AISI 304 steel
wear, Wear 398–399 (2018) 99–115.
during dry sliding wear, Surf. Topogr. Metrol. Prop. 7 (2019), 015021.
[20] H. Liu, H. Liu, C. Zhu, Y. Zhou, A review on micropitting studies of steel gears,
[44] D.K. Prajapati, M. Tiwari, Experimental investigation on evolution of surface
Coatings 9 (2019) 42.
damage and topography parameters during rolling contact fatigue tests, FFEMS 43
[21] A. Terrin, C. Dengo, G. Meneghetti, Experimental analysis of contact fatigue
(2020) 355–370.
damage in case hardened gears for off-highway axles, Eng. Fail. Anal. 76 (2017)
[45] B.J. Hamrock, S.R. Schmid, B.O. Jacobson, Fundamentals of Fluid Film
10–26.
Lubrication, McGraw- Hill Press, USA, 1994.

15