Anda di halaman 1dari 13

Wear 322-323 (2015) 238250

Contents lists available at ScienceDirect

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

Analysis of the effect of different types of additives added to a low


viscosity polyalphaolen base on micropitting
E. de la Guerra Ochoa a,n, J. Echvarri Otero b, E. Chacn Tanarro b, J.M. Munoz-Guijosa b,
B. del Ro Lpez c, Cristina Aln Cordero d
a

Talgo R&D, Patentes Talgo, Paseo del Tren Talgo 2, Las Rozas, 28290 Madrid, Spain
Grupo de Investigacin en Ingeniera de Mquinas, Universidad Politcnica de Madrid, Jos Gutirrez Abascal 2, 28006 Madrid, Spain
c
Fsica Aplicada e Ingeniera de los Materiales, Universidad Politcnica de Madrid, Jos Gutirrez Abascal 2, 28006 Madrid, Spain
d
Teora de la Seal y Comunicaciones, Escuela Politcnica Alcal de Henares, Plaza de San Diego s/n, Alcal de Henares, 28801 Madrid, Spain
b

art ic l e i nf o

a b s t r a c t

Article history:
Received 11 August 2014
Received in revised form
10 November 2014
Accepted 13 November 2014
Available online 21 November 2014

This article explores the inuence of the appearance of micropitting in different types of additives added
to a polyalphaolenic low-viscosity base oil, namely a polyalphaolen PAO6. Additives commonly used
in mechanical transmissions by gears were used, i.e., extreme pressure, anti-wear and friction modier.
For the study, we have conducted a series of tests on a disc machine under various operating
conditions. The temperature and surface roughness of the test specimens have been changed in order to
study different lubricant specic lm thicknesses, as the last is one of the most inuential parameters on
the appearance of micropitting.
Tests have shown the important inuence of the additives, and their concentration, in the
development of micropitting and their associated effects, depending on the operating conditions of
the contact. Along with the study of the effect of additives, the inuence of specic lm thickness on the
friction coefcient, surface appearance, conditions of lubrication, and wear was also studied.
& 2014 Elsevier B.V. All rights reserved.

Keywords:
PAO
Rolling contact fatigue
Sliding wear
Sliding friction
Lubricant additives
Gears

1. Introduction
Surface fatigue is a phenomenon associated with mechanical
contacts under cyclic loading. There are different scales depending
on the size of the defect: micropitting, (macro)pitting and spalling
[1]. This article studies micropitting in lubricated line contacts.
Micropitting is dened by ASTM [2] as a form of surface
damage in rolling contacts consisting of numerous pits and
associated cracks on a scale smaller than that of the Hertz elastic
contact half-width. An example of this phenomenon is shown in
Fig. 1, after a test with an approximated value of 100 m of the
Hertz half-width.
The appearance of micropitting in mechanical contacts is
inuenced by multiple factors [3,4], such as contacting materials,
their roughness, types of lubricants (both base oil and additives)
and factors associated with operating conditions, namely load,
temperature, average velocity, and sliding velocity.
Of all these possible causes, specic lm thickness is the key
parameter [5,6]. It compares the lubricant lm thickness to the
n

Corresponding author.
E-mail addresses: edelaguerra@talgo.com (E. de la Guerra Ochoa),
jechavarri@etsii.upm.es (J.E. Otero), bdelrio@etsii.upm.es (B. del Ro Lpez).
http://dx.doi.org/10.1016/j.wear.2014.11.014
0043-1648/& 2014 Elsevier B.V. All rights reserved.

combined surface roughness of the contact surfaces and serves to


nd the lubrication regime in which the contact operates. In the
case of low specic lm thicknesses, high stresses are concentrated on the surface of the material. This leads to an early, and
localized, failure on the surface. Therefore, all parameters that
improve lm formation improve resistance to micropitting. These
parameters can be related to the lubricant (high viscosity, high
viscositypressure coefcient, etc.), operating conditions (low
temperature of the lubricant, high average velocity, low slide-toroll ratios, reduced loads, etc.) or materials and manufacturing
process (reduced roughness, etc.).
Once uid lm lubrication is reached the risk of micropitting is
reduced. However, the rheological behavior of the lubricant plays
an important role in the onset of macropitting [7]. On one hand,
through the coefcient of friction attained: increasing the coefcient of friction rises the risk of surface fatigue by increased
stresses, as well as the undesired effect of reducing the machine's
performance. On the other hand, through the pressure distribution
due to the elastohydrodynamic contact: the occurrence of pressure
peak at the outlet accelerates the onset of pits. Besides operating
conditions, pressureviscosity coefcient and compressibility are
some parameters which inuence the pressure peak, friction
coefcient and lm thickness.

E. de la Guerra Ochoa et al. / Wear 322-323 (2015) 238250

239

MPR

32m

100m

Rings

Temperature
probe
Roller

67m

Fig. 1. An example of surface with micropitting damage.

Therefore, the behavior of surface fatigue is closely related with


the base oil. In this sense it seems synthetic lubricants perform
better than mineral lubricants with similar characteristics [8,9],
because they achieve higher lm thickness, have better stability,
lower operating temperature, lower coefcient of friction and
lower viscositypressure coefcient. These aspects have been
analyzed in [7] developing predictive behavior models to predict
macropitting and micropitting. However, the effect of additives,
although complex to model, is crucial to operation in regard to
wear, friction and surface fatigue, especially when the contacts are
working under mixed or boundary regimes, where the properties
of additives become more important [10,11].
The base oil used in this article is a PAO6, usually used in gears,
with a kinematic viscosity of 6 cSt at 100 1C. The effect of
temperature, surface roughness and several types of additives
were studied: a friction modier (FM), an anti-wear (AW), and an
extreme-pressure (EP). For the FM, we chose olein, for the AW a
ZnDTP and for the EP a mixture of amine phosphates. These kinds
of additives are very common in mechanical transmissions.
In general, locally induced high temperatures in the contact
asperities activate the EP additives to promote chemical etching at
the points of contact resulting in a less rough surface. However, the
AW generates a protective layer which adheres to the surfaces. The
asperities slide over the layer without metallic contact, and therefore,
the roughness is not reduced by chemical etching as in the case of EP.
The crucial difference between the EP/AW and FM additives are
their mechanical properties. The layer formed by the EP/AW
additives is semi-plastic, and therefore offers certain resistance
to shearing so that the coefcient of friction is generally high.
However, the FM additive lms are composed of multiple layers of
ordered and closely packed molecules. These layers are loosely
adhered to one another, with the polar part of the molecule
remaining anchored to the surface. The outer layers can be easily
sheared which allow low friction [12,13].
The danger of these additives, especially EP and AW, is that if
chemical activity is uncontrolled, its effects may not be localized to
the asperities of the contact surface and may extensively attack the
surface in high temperature and high pressure conditions.
In addition to studying the effect on surface fatigue, the inuence
of additives on friction and wear of the contact were also analyzed.
For this analysis, a commercial test machine Micro Pitting Rig (MPR)
by PCS Instruments was used (www.pcs-instruments.com).

2. Testing plan
2.1. Testing bench and test specimens
A MPR disc machine, shown in Fig. 2, was used for the tests.
This machine is for testing surface fatigue and also allows the
measurement of the friction coefcient and wear in lubricated line

Initial contact width=1mm


Fig. 2. Overview of MPR and a detail of the test zone; contact width of the roller.

Table 1
Geometrical dimensions and surface nish of specimens.

Roller
Rings

Radius r (mm)

Contact width L (mm)

Roughness Ra (m)

6
27.075

1
8

0.1
0.1 or 0.4

contacts between discs. The MPR is currently used by the Industry


to look at both macro- and micro-pitting resistance of lubricating
oils on gears and bearings [14].
The equipment consists of three rings which rotate at the same
velocity and in contact with a central roller connected to a separate
axis, which permits rotating at different velocity from the rings. This
way you can set the average (rolling) velocity um and the slide-to-roll
ratio (SRR) of the contact. SRR is dened as the ratio of the sliding
velocity u to the average velocity um, expressed as a percentage.
The zone of interest is the rolling track of the testing roller, due to
the higher strength of the outer rings [15]. The rolling track of the
roller has an initial contact width of 1 mm, as can be seen in Fig. 2.
This geometry of the roller produces widening of the rolling track in
case of wear.
The normal load W of the contact is applied on the upper ring
against the roller and lower rings. The arrangement is such that
the load is evenly distributed among them. The contact area is
lubricated by a thermostatic bath with controllable temperature. A
temperature probe is located into the test chamber with the tip of
the probe close to the contact region, as shown in Fig. 2.
In addition, it is equipped with a commercial ICP accelerometer
(PCB, model M353B16) mounted on the test head, which is used to
measure the vibration in the contact and provide real time
information on the condition of the specimen's surface.
The rings and roller are made of steel 16MnCr5, equivalent to
F1516, which is used in the construction of gears, pinions, and
cemented parts. It has undergone carburizing and quenching
treatments to achieve hardness of around 780 HV in the rings
and 680 HV in the roller. This material is the same as that of the
gears in the FZG gear testing machine [16,17].
The geometry and surface nish of the test specimens are
shown in Table 1. To observe the effect of specic lm thickness,
and study the behavior of the base and additives, rings of two
roughnesses (Ra) are used: 0.1 and 0.4 m.
This geometry of the specimens achieves a reduced contact
radius of 4.91 mm, similar to that found in the contact inlet in the
standard type C gears for testing micropitting in the FZG machine
[1820] as seen in Fig. 3. The FZG is used in standard surface

240

E. de la Guerra Ochoa et al. / Wear 322-323 (2015) 238250

 EP-mixture of amine phosphates: secondary amine salt of a

Reduced contact radius (Gear Type C)


9

C1114-branched alkyl, monohexyl and dihexyl phosphates.

In this way, the effect on surface fatigue of the additives and


their concentrations can be studied individually.

R(mm)

2.3. Test prole

6
MPR Reduced
contact radius

50

100

150

200

250

300

Points in line of action

Fig. 3. Evolution of the reduced contact radius of the standard gear type C of
the FZG.
Table 2
Physical properties of the PAO6 at different temperatures.
T (1C)

0 (mPa s)

(GPa  1)

30
40
60
80
100

37.95
25.00
12.57
7.36
4.78

12.3
11.5
10.1
9.0
8.2

fatigue testing, with the disadvantage of requiring a long time to


achieve a high number of cycles. Keeping in mind that a gear can
be modelled by a contact between rolling and sliding discs [21] the
MPR reproduces these test conditions much more quickly and
cheaply. The central roller has three contacts per cycle and a small
radius shortens the cycle duration and therefore more fatigue
cycles are achieved in a shorter time.
2.2. Lubricant
The lubricants used are as follows: a PAO6 oil base without
additives; the base with an FM additive at 2 wt%; the base with an
AW additive at 1 and 0.3 wt%; the base with an EP additive at 1 wt%
and 0.3 wt%. The highest percentages are achieved at base saturation,
and in the case of AW and EP, have also been used at a lower
percentage for lower reactivity as these additives are consumed during
chemical etching.
Viscous properties of the PAO6 oil base were measured in
laboratory and are given in Table 2, namely the viscosity at
ambient pressure 0 and the pressureviscosity coefcient .
The addition of the different types of additives has shown nonsignicant variations in the results.
The non-Newtonian behavior of the PAO6 has been modelled
by means of the Carreau equation [22], since references [2325]
reveal that the Carreau model properly characterizes the behavior
of shear-thinning lubricants like the PAO6.
The Carreau equation is dened by two additional parameters of
lubricant, namely Carreau's exponent n and the shear modulus G. In
the case of PAO6, the values taken at 40 1C are n0.53, G 1.59 MPa
and at 90 1C are n0.81, G 0.1 MPa, in line with references [7,26].
The types of additives used in the experiments are as follows:

 FM-olein: Glyceride of oleic acid (C17H33COO)3C3H5.


 AW-ZnDTP: Zinc dialkyldithiophosphate primary alkyl groups
with 8 carbon atoms.

Several test methods and proles have been developed for


analyzing micropitting in the MPR, taking into account the
experience of companies like PCS Instruments (www.pcs-instru
ments.com), Powertrib (www.powetrib.com) or the Society SAE
International (www.sae.org).
Bearing in mind this know-how, the test prole performed
consists of an initial running-in step, followed by four steps, as
reected in Table 3.
Until step 3 included, this test prole corresponds with the
DGMK GFKT-C/8.3/90 Micropitting short test on the FZG machine
(www.dgmk.de), although it is more severe in load because it
works with higher pressures (0.85 GPa to 1.7 GPa in respect to the
standard 0.51 GPa to 1.4 GPa), with less cycles (3E6 in the test in
respect to 4.33E6 which is the standard) and with a lower SRR than
that achieved in the prole of the type C gear, which reaches up to
120% in the beginning of the line of action. The duration is
200 min, whereas the duration of the DGMK GFKT-C/8.3/90 Micropitting short test is 2000 min approximately.
Additionally, we have introduced step 4 to evaluate the
endurance of the test specimen and the effect of additives in the
long run. Therefore, the total duration of the test is 21 million
cycles (1400 min) with stops to inspect the roller at the end of
each step.
During the test, the friction coefcient , the bath temperature
T, the specimen wear level and the acceleration/vibration of the
contact are continuously measured. The vibration level displayed
by this rig is a dimensionless result which compares the instantaneous acceleration with an initial reference value. The control
system was set to automatically stop the test in case of reaching a
peak-to-peak vibration level of 1000, which corresponds to an
acceleration of 90 m/s2, since this indicates severe worsening of
the roller surface by wear or the appearance of pits. Moreover,
after each step a micrograph is taken of the condition of the roller
surface and the contact width is measured.
This test prole is performed for two different temperatures
(40 and 90 1C) in the case of the base without additives. The effect
of the additives in the proportions indicated is only analyzed in the
worst case, i.e., at 90 1C. The tests are performed for two initial ring
roughnesses, 0.1 and 0.4 m.

2.4. Estimation of specic lm thickness


The specic lm thickness is dened [1] as the ratio of the
lubricant lm thickness hc to the combined surface roughness
Table 3
Test prole: step, average velocity um, slide-to-roll ratio (SRR), load W and Hertz
pressure p0, time t and number of cycles.
Step

um (m/s)

SRR (%)

W (N) and p0

t(min) and cycles

Running-in
Step 1
Step 2
Step 3
Step 4

3.15
3.15
3.15
3.15
3.15

1-20
20
20
20
20

100 (0.85 GPa)-230


230(1.3 GPa)
305(1.5 GPa)
390(1.7 GPa)
390(1.7 GPa)

20 (3E5 cycles)
60 (9E5 cycles)
60 (9E5 cycles)
60 (9E5 cycles)
1200(18E6 cycles)

E. de la Guerra Ochoa et al. / Wear 322-323 (2015) 238250

(RMS) of the contact specimens.

hc

hc
q
21 22

where 1 and 2 are the surface roughnesses (RMS) of the roller


and the rings, respectively.
Lubricant lm thickness is approached as constant and equal to
the central value [27]. Shear thinning effects in lm thickness are
considered by means of the following equation obtained by
Bair [28]:
"


 1 #3:61  n1:7
hNc
SRR um 0 1 0:002SRR
1 0:79
1
100 hNc G
hc

Eq. (2) takes into account the lubricant parameters n and G of the
Carreau rheological model [22], the low-shear viscosity at ambient
pressure 0 and the slide-to-roll ratio SRR. Finally, hNc represents
the lm thickness for the Newtonian approach, which is determined with the Hamrock equation [29] for line contacts.

0:692 0:110 0:308  0:332
E'
R
p0
3
hNc 2:1540:47 0 um
where is the viscositypressure coefcient, E0 Young's reduced
modulus, R the reduced contact radius and p0 the maximum Hertz
pressure.
The results of specic lm thickness calculated according to
Eqs. (1)(3) are presented in Table 4, together with the Hertz
contact half-width a.
Thermal effects are neglected [30] due to the low values of
slide-to-roll ratio considered in the test prole. Otherwise, a
thermal correction factor should be used for lm thickness
estimation, e.g., the formulas proposed in references [29,31].

3. Results obtained
3.1. Effect of temperature
In Fig. 4, the micropitting test micrographs are shown for PAO6
without additives and using rings of lowest roughness (0.1 m).
The results at 40 and 90 1C are compared at the end of each step of
the test.
In both cases, micropitting has begun to appear in step 2,
where the akes caused by surface fatigue are appreciable. The
results of the test at 90 1C, as expected, were more severe than
those at 40 1C due to the decrease in the viscous properties of the
lubricant and consequent reduction of lm thickness. As shown in
Table 4, for step 1 the specic lm thickness is approximately 1.17
for 40 1C and 0.33 for 90 1C. In steps 3 and 4, specic lm
thicknesses were reduced to 1.06 and 0.29 for 40 1C and 90 1C,
respectively.
The severity of the 90 1C test is revealed in the more affected
surface, mainly by the increase in pit size. In addition, greater pit

241

depth is seen in steps 2 and 3 in respect to the test at 40 1C, see


Fig. 4. In step 4 of the test at 90 1C, a coalescence phenomenon has
occurred and micropitting has resulted in incipient macropitting
given the size of the indentations formed (approximately 100 m
in diameter). The larger size of the indentations corresponds to a
higher recorded acceleration signal, as can be seen in Fig. 5a.
In both cases, the 1 mm contact width has remained in the
original dimensions of the specimen, without signicant expansion of the original contact width, despite the low specic lm
thickness, in contrast to reference [5]. This may be due to polishing
during running-in and rst step, so that the original machining
marks are not appreciable, with the consequent increase of
specic lm thickness. Fig. 5b shows wear quickly grows to step
4 where it stabilizes to a level of approximately 140 m.
The blue areas on the surface of the contact, seen in Fig. 4, are
the result of tempering steel by high temperatures reached locally
at the contact between asperities, which could be in a typical
range of 250290 1C.
Fig. 5c and its measures of friction coefcient at 40 1C, demonstrates friction increases as steps progress: in step 1 a value of
0.046 is reached, in step 2 it increases to 0.055, step 3 continues to
0.063, and nally reaches 0.070 in step 4. Steps 1 and 2 present a
low specic lm thickness which diminishes when gradually
increasing the load and operating temperature during successive
steps, leading to a higher coefcient of friction.
For the test performed at 90 1C, the friction coefcient slightly
decreases at the beginning of step 4, whereas lower uctuations
are found for the tests at 40 1C. Fig. 5d shows a drop in
temperature at the beginning of this step 4, from 100 1C (approximately) to the commanded value of 90 1C. The reduction in
temperature is associated with a rise in lm thickness and explains
the decrease observed in the friction coefcient according to
Stribeck's curve [1]. This point is supported by the results
presented in Table 4, concerning very low values of specic lm
thickness in the test at 90 1C. Once working conditions are
stabilized in step 4, the average friction coefcient is about 0.065.
A continuous increase in temperature is observed in the test at
40 1C because of the heat generated by friction. Thus, a refrigeration system would be required to keep the temperature in a
controlled range. In contrast, a simpler layout is required at 90 1C,
i.e. the temperature selected for the following tests, because the
heat generated in the contact is lower than is required to attain the
commanded temperature. Then, the use of an electric cartridge
that heats up the lubricant bath in a controlled way is enough to
control the temperature in the tests at 90 1C.
3.2. Effect of roughness
To study the effect of increased roughness, the previously used
rings of Ra 0.1 m are exchanged for rings of Ra 0.4 m. This
implies specic lm thicknesses of the previous case are approximately divided by three, taking into account the results of Table 4.

Table 4
Specic lm thickness calculation for the test prole considered.
Case

Step

W (N)

p0 (GPa)

a (lm)

hc (lm)

Temperature: 40 1C, Ra1 0.1 mm; Ra2 0.1 mm, (RMS) 0.156 mm

Step 1
Step 2
Steps 3 and 4
Step 1
Step 2
Steps 3 and 4
Step 1
Step 2
Steps 3 and 4

230
305
390
230
305
390
230
305
390

1.3
1.5
1.7
1.3
1.5
1.7
1.3
1.5
1.7

112
129
146
112
129
146
112
129
146

0.183
0.174
0.165
0.051
0.049
0.046
0.051
0.049
0.046

1.17
1.12
1.06
0.33
0.31
0.29
0.11
0.11
0.10

Temperature: 90 1C, Ra1 0.1 mm; Ra2 0.1 mm, (RMS) 0.156 mm

Temperature: 90 1C, Ra1 0.1 mm; Ra2 0.4 mm, (RMS) 0.454 mm

242

E. de la Guerra Ochoa et al. / Wear 322-323 (2015) 238250

Running in+step 1

Step 2

Step 3

Step 4

40C
200m

200m

100m

100m

200m

200m

100m

50m

90C

Fig. 4. Appearance of the roller at the end of each step. Experiment without additives at 40 and 90 1C with 0.1 m rings. (For interpretation of the references to color in this
gure, the reader is referred to the web version of this article.)

Comparison of wear

Comparison of vibration level (Peak-to-Peak)


200

700

180
600

160
140

400
PAO6-40C
PAO6-90C

300

(m)

500

120
100

PAO6-40C
PAO6-90C

80
60

200

40
100

20

0
0

10000 20000 30000 40000 50000 60000 70000 80000 90000


t(s)

t(s)

Comparison of temperature
120
100
80
PAO6-40C
PAO6-90C

T()

Comparison of friction coefficient


0.1
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0

10000 20000 30000 40000 50000 60000 70000 80000 90000

60

PAO6-40C
PAO6-90C

40
20
0

10000 20000 30000 40000 50000 60000 70000 80000 90000


t (s)

10000 20000 30000 40000 50000 60000 70000 80000 90000


t(s)

Fig. 5. Comparison of friction coefcient, wear, vibration level and temperature between PAO6 at 40 1C and 90 1C.

Fig. 6 shows the micrographs at the end of each testing step for
the PAO6 at 90 1C, without additives and roughest rings, compared
with the results of rings with a lower roughness.
As seen in Fig. 6, step 1 shows considerable wear and polishing
of the surface of the roller which has grown from an initial width
of 1 to 1.2 mm. For these conditions, the specic lm thickness is
around 0.11, according to Table 4.
Fig. 7 sheds light on the pits formation process, since subsurface cracks are clearly shown under the worn surface of the roller.
A main crack is observed in the subsurface, propagating through
the material in the sliding direction. Secondary cracks originated
close to the surface and near this main crack reach the surface. In
this way, pits are formed which can be easily removed by wear,
visually eliminating the appearance of a fatigued surface.

Thus, wear has probably masked the appearance of micropitting during the test, unlike the previous case with the rings of
0.1 m, Fig. 4. However, primary cracks can produce larger and
deeper pits (macropitting) more difcult to be masked by wear.
In authors' opinion, despite the loss of material at the surface
level, the inside continues to accumulate high stresses and contribute to originate primary cracks (step 2). This phenomenon
maintained for a high number of cycles can lead to severe
macropitting [7] with pits from 300 to 500 m in diameter, like
those seen in step 4. Testing was stopped without nishing step
4 because maximum acceleration was reached after only 250,000
cycles in step 4 (or 13,000 s from the start of the test), see Fig. 8a.
Also, being specimens with higher roughness, the friction coefcient in the rst steps (running-in and step 1) is approximately 0.08

E. de la Guerra Ochoa et al. / Wear 322-323 (2015) 238250

Step 1

Ra=0.1m

Step 2

200m

Step 3

Ra=0.1m

200m

243

Ra=0.1m

200m
Ra=0.1m

Step 4

0.25E6 cycles

200m

Fig. 6. Appearance of rollers at the end of each step. Test without additives at 90 1C with 0.4 m rings. In miniature, image under the same conditions with 0.1 m rings.

(Fig. 8b), which is slightly higher than those obtained previously.


This increased friction coefcient also increases the risk of macropitting [7].
The friction and wear (Fig. 8b and c) have not clearly increased
from step 2 due to the decrease in effective pressure by the
widening of the rolling track during the test. Using Hertzian
formulae for line contact [32], there is approximately a 10%
reduction of the maximum Hertz pressure at the end of the test
when the track width is 1.2 mm.
On the other hand, lubricant temperature remains stable
during the test, with slight oscillations around the commanded
value of 90 1C (Fig. 8d).
3.3. Effect of additives
The next phase is the introduction of additives up until the
maximum weight ratio: FM at 2%, and AW and EP at 1%, i.e.,
reaching the base saturation at the temperature of 90 1C. Test rings
of 0.1 m are used.
Fig. 9 shows the micrographs obtained at the end of each test
for the base without additives and for each of the additives. Fig. 10
includes the results of the parameters measured during the new
tests, where a stable temperature of 90 1C is observed in the
lubricant bath (Fig. 10a).
3.3.1. Results for FM additive at 2 wt%
In the case of FM, signicant differences in the appearance of
the surface are observed when compared to the base without
additives. In the base without additives, surface pits are larger
because surface fatigue has led to macropitting. Meanwhile, for
the base with FM additive, the surface has more micropitting
marks, but they have not increased as much in size.
If you examine Fig. 10b, the graphic showing wear with FM
additive, it is somewhat higher than that obtained with the base
alone. Wear has also given the surface a polished appearance,
which has made micropitting less noticeable. This is justied [7]
since the reduction of roughness increases specic lm thickness
and reduces the degree of micropitting. Furthermore, Fig. 10c
shows that the effect of FM in reducing the friction coefcient is
clear, from 0.060 in the base without additives to 0.045 with FM.

Thus helping to improve the surface appearance of the specimen


in line with the lower level of peak-to-peak vibration level
measured, as shown in Fig. 10d.
3.3.2. Results for AW additive at 1 wt%
As for the AW, the ZnDTP additive forms a protective reaction
layer on the surfaces leading to a reduction of mild wear, as
described in reference [10]. This slows the running-in process and
causes persistent roughness. A very low specic lm thickness and
a high concentration of AW promote an uncontrolled continuous
chemical reaction on the surface (Fig. 9). This leads to an increasing material loss, as observed in Fig. 10b, in agreement with
reference [12]. Due to chemical etching, signs of surface fatigue
in step 4 can be masked and not clearly visible.
Thus, if the time evolution of wear and friction in Fig. 10b and c
is analyzed for AW additive, it has achieved higher levels than
PAO6 without additives where a more appropriate running-in
process may occur. It can be seen that wear is almost double than
in the case of the base without additives, and also does not
stabilize over time but rather follows a rising trend. The increased
friction is due to deterioration of the surface by etching.
The persistent roughness induces high local stresses that
increase the degree of micropitting, in line with the results of
reference [10], which explains the high vibration level observed.
The acceleration measurement increases in the rst steps to later
descend and stabilize in the rst quarter of step 4, where the
surface state of the roller does not continue to worsen. The
decrease in acceleration may be due to the combination of two
effects: the decrease of pressure (due to the increase of the contact
width) and mild wear (which eliminates roughness).
3.3.3. Results for EP additive at 1 wt%
As with AW, the EP additive has produced widespread attack of
the surface as shown in Fig. 9. The EP additive can promote
corrosion in the contact [12,13] and has resulted in early failure of
the roller, which did not nish step 4. It reached the maximum
allowed acceleration at 4.22 million cycles of step 4 due to the
damage in the surface of the specimen.
Wear levels are maintained at the same values as in the case
without additives until step 2, where they begin to increase and

244

E. de la Guerra Ochoa et al. / Wear 322-323 (2015) 238250

25m

50m

10m

50m

25m

Fig. 7. Detail of the subsurface of the roller at the end of the test without additives at 90 1C with 0.4 m rings.

equal wear obtained with the AW additive. The friction coefcient


is slightly higher than without additives, maintaining a value close
to 0.070. Unlike the AW additive, mild wear has not been able to
compensate for the surface deterioration of the roller and the
signal of acceleration has increased signicantly in step 4.
Due to the poor behavior of the AW and EP additives, and
motivated by their high reactivity, the proportion is reduced to
0.3 wt% in both cases, yielding at the end of the test the results shown
in Figs. 11 and 12. The lubricant temperature remains stable in the new
tests and close to the commanded value of 90 1C (Fig. 12a and e).

3.3.4. Results for AW additive at 0.3 wt%


If Fig. 11 is compared to Fig. 9 for the cases with AW, further
deterioration of the specimen is observed for lower proportion of
additives. This deterioration begins at step 1, where pits begin to
appear. Due to the scarce concentration of AW, an insufcient
running-in process may have caused higher roughness than in the

previous case of higher proportion of additive. This higher roughness has led to an earlier and more prominent appearance of
micropitting.
The worse appearance and widespread presence of pits have
caused the roller to fail in step 3, at 600,000 load cycles when the
maximum acceleration allowed was reached (Fig. 12b), unlike the
concentration of 1% which survived to step 4.
The base without additives shows faster increase of wear
(Fig. 12c), but remains constant after achieving a certain level;
this causes the specimens to better withstand the test. However,
additives which cause chemical etching of the surface tend to
increase wear throughout the whole test. Furthermore, in some
cases, is has been observed a worsening of the surface of the roller,
i.e. increase of acceleration, that leads to an early failure, as has
been the case of AW additive at 0.3% by weight.
Regarding friction (Fig. 12d), it is slightly higher than results
obtained with the higher concentration due to the worse specimen
surface appearance.

E. de la Guerra Ochoa et al. / Wear 322-323 (2015) 238250

Vibration level (Peak-to-Peak)

1200
1000

800
600
400
200
0
0

2000

4000

6000

8000

10000

12000

Friction Coefficient

0.1
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
0

14000

245

2000

4000

6000

Wear

300
250

T()

(m)

200
150
100
50
0
0

2000

4000

6000

8000

8000

10000

12000

14000

10000

12000

14000

t(s)

t(s)

10000

12000

14000

Temperature

100
90
80
70
60
50
40
30
20
10
0
0

2000

t(s)

4000

6000

8000
t(s)

Fig. 8. Signals of friction, wear, vibration level and temperature for the test without additives at 90 1C with 0.4 m rings.

Step 4

18E6 cycles

Step 4 FM 2% wt.

50m

Step 4 AW 1% wt.

18E6 cycles

18E6 cycles

200m

Step 4 EP 1% wt.

200m

4.22E6 cycles

200m

Fig. 9. Comparison of the surface appearance of the 0.1 m rings in step 4 with different additives.

3.3.5. Results for EP additive at 0.3 wt%


Comparing both tested EP additive concentrations (Figs. 11 and 9)
it can be seen that for step 3 the concentration of 0.3% is not able to
mask akes that appear, showing signicant deterioration. However,
for the concentration of 1%, beginning from step 3 the appearance of
etching on the roller has mitigated signs of fatigue.
For the concentration of 0.3% in weight, the specimen failed
at 760,000 cycles in step 3 because of the acceleration peak
(Fig. 12f).

With regard to wear (Fig. 12g), the lower proportion of the EP


additive caused a decrease in the wear rate.
As in the case of AW, friction is higher than that obtained with the
highest concentration because of surface deterioration (Fig. 12h).

3.3.6. Summary of results for rings with 0.1 m roughness


Therefore, for low roughness it seems the use of very reactive
additives does not improve (or even worsens, as in the case of EP and

246

E. de la Guerra Ochoa et al. / Wear 322-323 (2015) 238250

Comparison of wear
300

100

250

80

200
(m)

T()

Comparison of temperature
120

60

100

40

50

20
0

150

0
10000

20000

30000

40000

50000

60000

70000

80000

90000

10000

20000

30000

40000

PAO6-90C

PAO6+FM-90C

PAO6+EP-90C

PAO6-90C

PAO6+AW-90C

60000

70000

80000

90000

PAO6+FM-90C

PAO6+EP-90C

PAO6+AW-90C

Comparison of vibration level (Peak-to-Peak)

Comparison of friction coefficient

0.1
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0

50000
t(s)

t (s)

1200
1000
800
600
400
200
0

10000

20000

30000

40000

50000

60000

70000

80000

90000

10000

20000

30000

40000

PAO6-90C

PAO6+FM-90C

50000

60000

70000

80000

90000

t(s)

t(s)
PAO6+EP-90C

PAO6+AW-90C

PAO6-90C

PAO6+FM-90C

PAO6+EP-90C

PAO6+AW-90C

Fig. 10. Comparison of the signals of wear, friction, vibration level and temperature for PAO 6, PAO6 FM at 2%, PAO6 AW at 1%, and PAO6 EP at 1%.

Step 3 EP 0.3% wt.

Step 3 AW 0.3% wt.

0.59E6 cycles

200m

0.76E6 cycles

200m

Fig. 11. Surface of roller for test using PAO6 with AW and EP additives at 0.3 wt% with 0.1 m rings at 90 1C.

AW) the behavior of the base without additives, in regard to surface


fatigue strength. In addition, lower concentrations of additive give
even worse results: a combination of wear and deterioration of surface
appearance.
On using less reactive additives, the FM additive has delivered
improved operating conditions due to the reduction of vibration
level and signicant improvement of friction in the contact, as
exposed above.

3.4. Combined effect of roughness and additives


After checking the effects of additives with the less rough discs,
rings of Ra 0.4 m, instead of Ra 0.1 m previously used, are
selected to analyze the effect when roughness is increased. In this
case the additives are introduced in the proportion of saturation.
The results are shown in Figs. 13 and 14. A stable temperature of
90 1C is also attained in the new test (Fig. 14a).

3.4.1. Results for FM additive at 2 wt%


For the FM additive, as with no additives, the high roughness
and associated wear of the specimens has concealed the appearance of micropitting, although in this case akes are seen on the
surface (Fig. 13).
With the Ra 0.4 m rings the friction modier has been
ineffective and has not caused the friction coefcient to decrease
as was observed with the 0.1 m discs. With such low specic lm
thicknesses, the shearing effect of the FM layers is negligible
compared to the metal-on-metal contact. Thus, friction coefcients come close to 0.075 (Fig. 14b), which are almost double
those of the same test with 0.1 m rings.
As for wear (Fig. 14c), it slightly decreased compared to the
base without additive, but obviously it was far superior to the
similar test with 0.1 m rings. The acceleration signal reects the
poor state of the roller surface (Fig. 14d). The test stopped after
11,000 load cycles in step 2 because the resulting pits caused peak
of acceleration higher than the xed maximum.

E. de la Guerra Ochoa et al. / Wear 322-323 (2015) 238250

Comparison of Temperature

100

Comparison of vibration level (Peak -to-Peak)

1200

90

247

1000

80
T()

800
70
600

60
50

400

40

200

30
0

10000

20000

30000

40000

50000

60000

70000

80000

90000

10000

20000

30000

40000

t(s)
PAO6-90C

PAO6+AW 1%

PAO6+AW 0.3%

PAO6-90C

Comparison of wear

300
250

150

(m)

200

100
50
0
0

10000

20000

30000

40000

50000

60000

70000

80000

90000

0.1
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0

PAO6+AW 1%

PAO6+AW 1%

10000

PAO6+AW 0.3%

20000

30000

PAO6-90C

90000

PAO6+AW 0.3%

40000

50000

60000

70000

80000

90000

PAO6+AW 1%

PAO6+AW 0.3%

Comparison of vibrartion level (Peak -to-Peak)


1000

80
T()

80000

1200

90

800

70
600

60
50

400

40

200

30

0
0

10000

20000

30000

40000

50000

60000

70000

80000

90000

10000

20000

30000

40000

t(s)
PAO6-90C

PAO6+EP-90C

PAO6+EP 0.3%-90C

PAO6-90C

150
100
50
0
10000

20000

30000

40000

50000

60000

70000

80000

90000

0.1
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0

10000

20000

t(s)
PAO6-90C

PAO6+EP-90C

60000

70000

80000

90000

PAO6+EP-90C

PAO6+EP 0.3%-90C

Comparison of friction coefficient

200

50000
t(s)

Comparison of wear

250

(m)

70000

t (s)

Comparison of temperature

100

60000

Comparison of friction coefficient

t(s)
PAO6-90C

50000

t(s)

30000

40000

50000

60000

70000

80000

90000

t (s)

PAO6+EP 0.3%-90C

PAO6-90C

PAO6+EP-90C

PAO6+EP 0.3%-90C

Fig. 12. Comparasion of wear, friction, vibration level and temperature of PAO6, PAO6 EP and PAO6 AW in concentrations of both 1 and 0.3%.

3.4.2. Results for AW additive at 1 wt%


With the roughest specimens, the AW additive can produce a
running-in more appropriate than in the case without additives.
This causes a reduction of the degree of micropitting (Fig. 13), as
reported in reference [33]. Behavioral differences of 0.4 m specimens in regards to those of 0.1 m seem clear with the AW
additive: apparently chemical etching has not been as severe, as
pits have not appeared until step 4. For the 0.1 m rings they
appeared during step 1.

In this case, one can still see the akes that indicate fatigue and
with a smooth appearance. However, if the evolution of wear is
observed (Fig. 14c), it increases continuously throughout the
whole test, leading to a rolling track close to 1.8 mm at the end
of the test. This represents a drop of 25% in maximum Hertzian
pressure.
As in the case with 0.1 m rings, the friction coefcients are
higher than for the base without additives due to deterioration of
surface appearance, see Fig. 14b.

248

E. de la Guerra Ochoa et al. / Wear 322-323 (2015) 238250

200m

Step 2 FM 2% wt.

11.000 cycles

Ra=0.1m

4.5E6 cycles

Step 4 EP 1% wt.

Ra=0.1m

7.2E6 cycles
Step 4 AW 1% wt.
200m

Step 4 PAO without


additives

0.25E6 cycles
Ra=0.1m

200m
Fig. 13. State of roller surface after tests with PAO6 FM, PAO6 EP, and PAO6 AW in proportion of saturation with 0.4 m rings.

Comparison of temperature

Comparison of friction coefficient

0.1

80

0.08

60

0.06

T()

100

40

0.04

20

0.02

0
0

5000

10000 15000 20000 25000 30000 35000 40000 45000 50000

10000

20000

PAO6-90C

PAO6+FM-90C

PAO6+AW-90C

40000

50000

PAO6+FM-90C

PAO6+AW-90C

PAO6+EP-90C

Comparison of vibration level (Peak-to Peak)

1200
1000

400
(m)

PAO6-90C

PAO6+EP-90C

Comparison of wear

500

30000
t(s)

t(s)

800

300

600
200

400

100

200
0

0
0

10000

20000

30000

40000

50000

5000

10000

15000

20000

PAO6-90C

PAO6+FM-90C

PAO6+AW-90C

25000

30000

35000

40000

45000

50000

t(s)

t(s)
PAO6+EP-90C

PAO6-90C

PAO6+FM-90C

PAO6+AW-90C

PAO6+EP-90C

Fig. 14. Comparasion of friction, vibration level, wear and temperature for the base with and without additives, using the roughest rings.

The AW test was stopped in step 4 when the maximum


acceleration was reached after 7.2 million cycles (Fig. 14d). Therefore, the specimen outlasted the base without additive.

3.4.3. Results for EP additive at 1 wt%


With the EP additive, the attack seems less severe when
compared with the specimens with less roughness (Fig. 13).
However, you can see widespread surface aking and a clear
increase in the width of the rolling track due to wear. At the end of
step 4 the rolling track measurement is 1.4 mm, this translates into
a reduction of 15% in maximum Hertzian pressure.
Wear has a tendency to grow throughout all testing steps and
with a slope greater than that observed for the test with 0.1 m

rings, but similar to that in the base with no additives test


(Fig. 14c).
The friction coefcient follows the tendency of the test with
0.1 m discs, and is superior over the base without additives with
a value close to 0.085 (Fig. 14b).
It has managed to prolong the life of the roller up until
4.5 million cycles in step 4, before reaching the maximum
acceleration (Fig. 14d).

3.4.4. Summary of results for rings with 0.4 m roughness


Therefore, for conditions of more roughness, the EP and AW
additives have proved effective to improve resistance to micropitting of the base without additives. However, due to the very low

E. de la Guerra Ochoa et al. / Wear 322-323 (2015) 238250

249

Table 5
Summary table of the results for the PAO6 with roller roughness of 0.1 m.
Test

Additive (wt%)

T (1C)

Ring Ra roughness (m)

Failure step

Cycles until failure (  1E6)

Wear track (mm)

Aspect

1
2
3
4
5
6
7
8
9
10
11

FM 2
AW 1
EP 1
AW 0.3
EP 0.3
FM 2
AW 1
EP 1

40
90
90
90
90
90
90
90
90
90
90

0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.4
0.4
0.4
0.4

4
4
4
4
4
3
3
4
2
4
4

16.24
16.24
16.24
16.24
4.22
0.59
0.76
0.25
0.01
7.25
4.49

1
1.02
0.98
1.32
1.22
1.03
1.03
1.2
1.02
1.73
1.38

Light micropitting
Large pits
Micropitting
Chemical etching
Chemical etching
Severe micropitting
Severe micropitting
Macropitting
Severe micropitting
Severe micropitting and chemical etching
Severe micropitting and chemical etching

specic lm thickness, a high wear level has been seen in all tests
for the base containing additives due to etching on the metal
surfaces. In the case of the FM additive, it proves less effective than
in the case of higher specic lm thickness.

Acknowledgments

3.5. Final summary

References

Table 5 briey presents the results obtained in all the cases


described above.

[1] B. Bhushan, Introduction to Tribology, John Wiley & Sons, USA, 2002.
[2] ASTM International. ASTM G-40-13. Standard Terminology Relating to Wear
and Erosion, ASTM Book of Standards, vol. 03.02, 2013.
[3] J.F. Brando, Gear micropitting prediction using the Dang Van high-cycle
fatigue criterion (Master's thesis), Faculdade de Engenharia da Universidade
do Porto, Oporto, Portugal, 2007.
[4] W.J. Bartz, Tribologische Aspekte bei ZahnradgetriebenSpeziell fr Fahrzeuge, 5. Internationales CTI Symposium.
[5] E. Lain, A.V. Olver, T.A. Beveridge, Effect of lubricants on micropitting and
wear, Tribol. Int. 41 (2008) 10491055.
[6] ISO (The International Organization for Standardization). ISO/TR 15144-1:
calculation of micropitting load capacity of cylindrical spur and helical gear
Part 1: introduction and basic principles, 2010.
[7] J. Echvarri, E. de la Guerra, E. Chacn, P. Lafont, A. Daz, J.M. Munoz-Guijosa,
J.L. Muoz, Inuence of the rheological behaviour of the lubricant on the
appearance of pitting in elastohydrodynamic regime, Fatigue Fract. Eng. Mater.
Struct. 35 (2012) 10471057.
[8] J.E. Fernandez Rico, A. Hernandez Battez, D. Garcia Cuervo, Rolling contact
fatigue in lubricated contacts, Tribol. Int. 36 (1) (2003) 3540.
[9] R. Errichello, Selecting and applying lubricants to avoid micropitting of gear
teeth, Mach. Lubr. 2 (6) (2002) 3036.
[10] C. Benyajati, A.V. Olver, The Effect of a ZnDTP Anti-wear Additive on
Micropitting Resistance of Carburised Steel Rollers, AGMA, Alexandria, USA,
2004.
[11] T. Ahlroos, H. Ronkainen, A. Helle, R. Parikka, J. Virta, S. Varjus, Twin disc
micropitting tests, Tribol. Int. 42 (10) (2009) 14601466.
[12] R.M. Mortier, M.F. Fox, S.T. Orszulik, Chemistry and Technology of Lubricants,
Springer, Dordrecht, Netherlands, 2010.
[13] L.R. Rudnick, Lubricant Additives, Chemistry and Applications, CRC Press, Boca
Raton, USA, 2009.
[14] T. Mang, K. Bobzin, T. Bartels, Industrial Tribology: Tribosystems, Friction,
Wear and Surface Engineering, Lubrication, Wiley-VCH, Germany, 2011.
[15] H.A. Spikes, A.V. Olver, P.B. Macpherson, Wear in rolling contact, Wear 112 (2)
(1986) 121144.
[16] L. Winkelmann, O. El-Saeed, M. Bell, The effect of supernishing on gear
micropitting, Gear Technol. (2009) 6065.
[17] C. Benyajati, A.V. Olver, C.J. Hamer, An experimental study of micropitting,
using a new miniature test-rig, Tribol. Ser. 43 (2003) 601610.
[18] G. Niemann, H. Winter, Maschinen elemente Band II Getriebe allgemein,
Zahnradgetriebe-Grundlagen, Stirnradgetriebe, Springer-Verlag, Berlin, 1985.
[19] Deutsches Institut fr Normung (DIN), DIN 3990 Teil 1. Tragfhigkeitsberechnung von Stirnrdern, Einfhrung und allgemeine Einusfaktoren, Germany,
1987.
[20] International Organization for Standardization (ISO), ISO/DIS 6336: Calculation of Load Capacity of Spur and Helical Gears, 2006.
[21] P. Lafont Morgado, A. Daz Lantada, J. Echvarri Otero, Diseo y clculo de
transmisiones por engranajes, Seccin de Publicaciones de la ETSI Industriales
de Madrid, Madrid, 2009.
[22] P.J. Carreau, Rheological equations from molecular network theories, Trans.
Soc. Rheol. 16 (1) (1972) 99127.
[23] A.D. Chapkov, S. Bair, P. Cann, A.A. Lubrecht, Film thickness in point contacts
under generalized Newtonian EHL conditions: numerical and experimental
analysis, Tribol. Int. 40 (2007) 14741478.
[24] Y. Liu, Q.J. Wang, S. Bair, P. Vergne, A quantitative solution for the full shearthinning EHL point contact problem including traction, Tribol. Lett. 28 (2007)
171181.

4. Conclusions
Through the tests, differences have been detected in resistance
to surface fatigue, wear, and friction with the base without
additives, as well as, the base with different types of additives.
The conditions set for the test have shown themselves effective to
study the behavior of the contact from the viewpoint of the
inuence of the lubricant, using a test prole which presents a
high degree of equivalence with the standard micropitting short
test on the FZG gear testing machine.
As for the results of less severe contact conditions (lower
roughness, lower supply temperatures, etc.) it has been found
that signs of surface fatigue are reduced and wear is low, even
with low specic lm thickness. As operating conditions toughen,
by increasing the roughness and temperature, surface fatigue
becomes clearer and the endurance of specimens is reduced. This
conrms the enormous inuence specic lm thickness has in the
appearance of micropitting.
The use of very reactive additives, such as those for extreme
pressure (EP) or anti-wear (AW), can either have no effect or even
worsen surface fatigue when the system is operating under the
less severe conditions associated with low roughness. For low
concentrations of additive, there is a combination of wear and
surface deterioration due to chemical etching. Besides not improving fatigue life, the tests with EP and AW have shown a signicant
increase in the friction coefcient. On using less reactive additives,
the FM additive has improved operating conditions by reducing
acceleration and signicantly improving friction in the contact.
However, for more severe conditions, i.e. higher roughness, the
EP and AW additives have been able to improve the duration of the
specimen, in contrast to when the base without additive is used. In
return for longer duration, more wear is seen on the roller. This
increase of wear on the rolling track leads to signicant reductions
in contact pressure because the initial contact width increases as
the wear rises during the test. The combined effect of lower
pressure, reducing surface roughness, and removal of the layer of
surface fatigue through wear helped prolong the life of the
specimen.

The authors would like to thank the valuable collaboration of


the Lubricants Laboratory of Repsol.

250

E. de la Guerra Ochoa et al. / Wear 322-323 (2015) 238250

[25] J.Y. Jang, M.M. Khonsari, S. Bair, On the elastohydrodynamic analysis of shearthinning uids, Proc. R. Soc. A 463 (2007) 32713290.
[26] P. Lafont, J. Echvarri, J.B. Snchez-Peuela, J.L. Muoz, A. Daz, J.M. MunozGuijosa, H. Lorenzo, P. Leal, J. Muoz, Models for predicting friction coefcient
and parameters with inuence in elastohydrodynamic lubrication, Proc. Inst.
Mech. Eng. Part J: J. Eng. Tribol. 223 (J7) (2009) 949958.
[27] J. Echvarri, P. Lafont, E. Chacn, E. de la Guerra, A. Daz, J.M. Munoz-Guijosa, J.
L. Muoz, Analytical model for predicting the friction coefcient in point
contacts with thermal elastohydrodynamic lubrication, Proc. Inst. Mech. Eng.
Part J: J. Eng. Tribol. 225 (4) (2011) 181191.
[28] S. Bair, Shear thinning correction for rolling/sliding elastohydrodynamic lm
thickness, Proc. Inst. Mech. Eng. Part J: J. Eng. Tribol. 219 (2005) 6974.
[29] B.J. Hamrock, Fundamentals of Fluid Film Lubrication, McGraw-Hill, New York,
USA, 1994.

[30] P. Anuradha, P. Kumar, New minimum lm thickness formula for EHL rolling/
sliding line contacts considering shear thinning behaviour, Proc. Inst. Mech.
Eng. Part J: J. Eng. Tribol. 227 (3) (2012) 187198.
[31] P.K. Gupta, H.S. Cheng, N.H. Forster, J.B. Schrand, Viscoelastic effects in MIL-L7808-type lubricant. I: analytical formulation, Tribol. Trans. 35 (2) (1992)
269274.
[32] E. de la Guerra, J. Echvarri, E. Chacn, P. Lafont, J.M. Munoz-Guijosa, A. Daz,
J.L. Muoz, Analysis of different multiaxial fatigue criteria in the prediction of
pitting failure in spur gears, Int. J. Surf. Sci. Eng. 8 (4) (2014) 356379.
[33] V. Brizmer, H.R. Pasaribu, G.E. Morales-Espejel, Micropitting performance of
oil additives in lubricated rolling contacts, Tribol. Trans. 56 (5) (2013)
739748.

Anda mungkin juga menyukai