‘The Effect of Thickener
on EP Additive Response
Gaston A. Aguilar; Joseph P Kaperick, Michael Lennon and Grant Pollard
Afton Chemical Corporation
500 Spring Street. Richmond, Virginia 23219
N0
Abstract
It is recognized in the industry that thickener used in
lubricating greases can interfere and negatively impact
additive function. Extreme pressure and antiwear
additives that function at the interface of the lubricant
and mating surfaces are particularly affected. This
study measured the effect of thickener concentration
on load-carrying ability of lithium complex grease
system. Specifically, Timken EP tests were conducted on
grease formulation at three consistencies treated with a
constant amount of a sulfur-phosphorus (S-P) gear oil
additive package. The testing showed that Timken OK
loads were not affected by soap thickener concentration.
Timken EP tests were also performed on corresponding
base oil formulation over a range of passing loads that
encompassed the highest passing load and ensuing
failure load of the greases. The effect of the thickener
was demonstrated by comparing differences in test
temperatures, wear scar widths and tribo-film coverage
and composition. Another objective of this study was
to elucidate mechanism by which thickener interferes
with S-P additive system. For this effort, oil separation
experiments were conducted and the resulting oils were
analyzed to measure decreases in additive concentration.
These experiments showed that sulfur concentrations
in the grease bleed oils were not significantly different
than the formulated oil. On the other hand, phosphorus
was not detected in the separated oil indicating that
phosphorus antiwear additives were completely retain in
the bulk of the grease, probably due strong interaction
with soap thickener. These results correlated well and
helped explain differences in test temperatures, wear scar
widths and tribo-flim compositions observed between the
greases and the corresponding base oil formulation.
Introduction
Greases are typically described as lubricating
compositions composed of base oil and additives
entrapped in a network of solid thickener particles,
platelets, clumps and/or fibers. Thus, the physical and
chemical interactions of thickener with additives are
recognized by many in the industry as having a significant
effect on additive performance 1-2. However, there are few
published studies that have detected these interactions
and/or measured the effect of these interactions on
lubrication performance. This does not imply that
grease lubrication, specifically of rolling bearing, has not
been studied. Numerous models have been published
to explain the mechanism by which grease lubricate
roller bearings in elastohydrodynamic lubrication
(EHL) regime. For the most part, bleeding, which is
the separation of base fluid, additives and/or thickener
from the bulk grease is recognized as an important
factor in these models2-3. In one such study, Cousseau
and co-workers used a ball-on disc test rig and optical
interferometry to measure the film thickness produced
by different formulated greases, their bleeds and base oils.
They found film thickness of grease resembled that of
their bleeds more than their base oils. This observation
underscores the importance of understanding how
thickener-additive interaction and surface competition
effect load-carrying ability of greases.
12 VOLUME 79, NUMBER 3
NLGI
In one study focused on thickener-additive interactions,
Sivik and co-workers utilized phosphorus Nuclear
Resonance Spectroscopy (31P NMR) to analyze the
interaction of zinc dithiophosphate (ZDDP) with lithium
1 2-hydroxystearate grease4. Their analysis revealed
that ZDDP underwent chemical changes in grease
containing residue amounts of lithium hydroxide raw
material. The changes involved the formation of lithium
dithiophosphate along with a decrease in neutral ZDDP
and an increase in basic ZDDP levels. Sivik proposed a
mechanism in which neutral ZDDP reacts with lithium
L
hydroxide to form lithium dithiophosphate, zinc oxide
and water. The resulting ZnO then reacts with neutral
ZDDP to produce additional basic ZDDP. More pertinent
to effect of these interactions was the analysis of the basic
lithium 12-hydroxystearate grease after “exhaustive”
hexane washing. The result of 13P NMR study showed
that a significant amount of neutral ZDDP was retained
in grease, which Sivik and co-workers proposed was due
to polar-polar “association” of the ZDDP with the lithium
soap fibers.
In another study, Kaperick used Timken test methods
to compare the load-carrying capacity of a sulfurphosphorus gear oil additive package treated into
lithium complex grease and the corresponding base oil
formulation5. Since the rate of lubricant delivery for the
grease and oil test methods are different, Kaperick first
adjusted the delivery systems to investigate the effect
of lubricant delivery rates. It was found that differences
in delivery rate of the lubricant did not affect Timken
response. The major finding was a large decrease in
Timken response for the S-P gear oil package in the
grease formulation. It was concluded that decrease
response was due to thickener-additive interactions
that reduced the ability of additives to get to the mating
surfaces or interfered with additive function.
This paper is an extension of the Kaperick’s work.
Specifically, the Timken test methods were again used
to compare the response of an S-P additive package in
lithium complex grease and the corresponding base
oil formulation. The study was broadened to include
different grease consistency grades to determine if
additive response was affected by thickener/base oil
concentration. The thought was that softer grease
formulations containing less thickener would behave
more like the base oil formulation. The study also
incorporated an extended test protocol to further
demonstrate differences in the lubricating mechanism
of grease versus its corresponding base oil formulation.
These measurements and techniques are briefly described
below:
• Temperature readings were taken to measure rate and
extent of temperature increase generated by frictional
forces during Timken runs.
• Wear widths on test blocks were measured on some
partial grease Timken runs and all complete Timken
runs.
• Scanning Electron Microscope-Energy-Dispersive
X-Ray Spectroscopy (SEM-EDX) was used to get a
measure of tribo-film coverage and composition.
• Oil separation experiments were performed on greases
to compare additive content of bleed oils against base
oil formulation.
The investigation showed that grease consistency
did not affect the load-carrying capacity. All grease
consistencies were judged to pass the Timken test at a 40
lb. load and to fail at a 50 lb. load due to metal pulling or
scoring after complete 10 minute runs. As expected, the
base oil formulation passed the Timken test at 30, 40, 50
and 60 lb. loads. Additional measurements showed that
the base oil formulation at all test loads had rapid rates
of temperature increase that plateau at approximately
2 minutes and produced relatively constant wear scar
widths covered with tribo-films rich in phosphorus
content. On the other hand, grease formulations had
much slower rates temperature increase, generated wider
wear widths that increased with test time and produced
sparser tribo-films rich in sulfur content and containing
none detectable to trace amounts of phosphorus. These
results are consistent with oil separation experiments that
revealed that phosphorus components in additive package
were unable to bleed away from greases regardless of
consistency while sulfur levels matched that of the oil
– 13 NLGI SPOKESMAN, JULY/AUGUST 2015
NLGT~
formulation.
Experimental
Timken Testing
Tests were carried out using “Standard Test Method
for Measurement of Extreme-Pressure Properties of
Lubricating Fluids (Timken Method)”, ASTM D-2782
and “Standard Test Method for Measurement of LoadCarrying Capacity of Lubricating Grease (Timken
Method)”, ASTM D 2509 on oil and greases respectively.
Both Timken test methods consist of passing lubricant
through an interface that consists of steel cup rotating
against steel block at a static spindle speed of 800 ± 5
rpm and static load for a 10 minute test period. For each
test, loads are increased until film rupture is detected by
either excessive wear or scoring on the test block. For this
study, temperature readings were collected at one minute
intervals using Control Company Traceable® infrared
thermometer gun. Temperature readings were taken on
test cups at approximately 180 degree position from ring
on block contact. Wear scar widths were measured using
low power microscope (4X) with filar micrometer to
measure with accuracy of ±0.05 mm (±0.002 in.). Contact
pressures were calculated using measured scar widths on
blocks and the equation provided in ASTM D 2509.
1)
2)
3)
4)
5)
6)
7)
Heptane rinse
Wiping with heptane-soaked Kim wipes
Heptane rinse
Isopropyl alcohol rinse
Wiping with isopropyl alcohol-soaked Kim wipes
Isopropyl alcohol rinse
Nitrogen blow dry
Oil Separation Experiments
Greases were subjected to a modified IP 121 standard
test method. In the standard method, grease is placed
in a 240 mesh woven wire cloth cone and loaded with
100 gram weight. The cone is placed on stainless steel
cup and bleed is collected for 168 hours at 40 °C. For
this study, the collection period was decreased to 8 hours
and collection temperature was increased to 140 °C. The
140 °C temperature was selected based on maximum
temperatures measured during Timken tests. The test
time was decreased to eliminate any oxidation effects.
Collected bleed samples were analyzed for phosphorus
and sulfur content. Phosphorus content was measured by
inductively coupled plasma atomic emission spectroscopy
(ICP) using a modified version of ASTM D 4951 standard
test method. Sulfur content was measured using LECO
Corporation SC-432 analyzer following ASTM D 1552
standard test method.
Surface Analysis
Lubricant Compositions
Test block surfaces were analyzed FET Quanta 650
Scanning Electron Microscope (SEM) and Oxford
Instruments X-Max 150 Energy-Dispersive X-Ray (EDX)
detector. Analyses were performed on the same 4 mm
by 3.5 mm area of each block containing part of the
wear scar. The beam energy was kept at 5 kV to increase
surface sensitivity. AZtec Version 2.2 software by Oxford
Instruments was used to create images of the mapped
area that consisted of three phases: 1) block background,
2) tribo-film and 3) carbon deposits. The AZtec software
was also used to perform elemental quants of EDX
spectra.
All lubricant compositions, oil and greases, were based
on ISO 220 mineral oil blend and were formulated with 3
mass percent of a proprietary gear oil package consisting
of sulfur based extreme pressure agents and phosphorus
based antiwear additives. The grease formulations
were made using unadditized, lithium complex base
grease with penetration (60 x worked) of 185 m-1. All
compositions were blended with standard laboratory
equipment with mild heating (60 ± 5 °C) for time judged
sufficient to thoroughly incorporate the additive package
and achieve homogeneous mixtures (1-2 hours). A
summary of lubricant compositions are provided in Table
1.
Blocks were clean before putting them in the SEM by
the following procedure:
– 14 VOLUME 79, NUMBER 3
NLGI
Table 1: Lubricant Compositions:
S-P Package (Weight %)
97
45
35
22
Lithium Complex Base Grease
(Weight %)
52
62
75
Penetration (60 x worked)
ASTM D 1403, (mm-I)
318
284
244
Iso 220 Base Oil Blend
(Weight %)
Results and Discussion
As summarized by Figure 1, Timken testing of the oil
formulation showed no significant loss of load-carrying
capacity of over a range of loads (30, 40, 50, and 60
lb.). All wears scars showed no visible signs of scoring
and wear scar widths were narrow and did not change
significantly with load. Average scar width for the four
loads was 1.04mm with a range 0.14 mm. The grease
2.5
S ore
formulations were able to carry 40 lb. without scoring
but wear widths were significantly larger than the oil
formulation with average and range of 1.45 mm and 0.26
mm respectively. Not unexpectedly, the greases were
unable carry 50 lb. load without scoring of the Timken
blocks and wear widths jumped to average of 2.21 mm
with a range of 0.39 mm.
x
I
2
• QuA
I
c_) ~
~ 4—’
I Grease A, NLGt I
1.5
Grease A, NLGI 2
X Grease A, NLG~ 3
a a a. a a sass S5.
I
*
Law~#~_ _ *
•5SSS
a
0.5
30
40
50
60
Load, lb.
Figure 1: Graph showing poorer anti-scoring and antiwear performance of the S-P package in grease.
Data points are an average two widths per scar.
– 15 NLGI SPOKESMAN, JULY/AUGUST 2015
N~jL
Wear rates were investigated by stopping Timken tests at 3 and 4.75 minutes for the NLGI 1 and 3 greases
at 50 lb. test load. As per Figure 2, wear scar widths increased with test time and were equal or exceeded
the widths of oil formulation at 3 minutes. Interestingly, the thickener concentration did not appear to
significantly affect test results. In other words, no improvement in performance was apparent with the softer
more oil like grease.
2.5
(
Scored
E
E
‘~
1.875
S cored
•
GreaseA, NLG~1
U GreaseA,NLGI3
1-
~u
1.25
~u
i
0,625
Figure 2:
Graph showing
wear increase
ofgrease
formulation
with time.
Data points
are an average
two widths per
scar.
0
0.00
2.30
5.00
7,50
10,00
Test TIme. minutes
As summarized Table 2, surface analysis of block
wear scars revealed that tribo-film compositions for
grease formulations are indeed very different than the
oil formulations. The oil formulation at all four loads
produced tribo-films high on phosphorus and oxygen
content, likely signifying the presence of phosphate
films that are known to provide excellent antiwear
protection6. Greases regardless of test load produced
films with undetectable to trace amounts of phosphorus
but four times more sulfur than the corresponding oil
formulation, possibly signifying the presence iron sulfide
and/or sulfate tribo-films generally generated by sulfur
EP agents7. As per mapped tribo-film area data in Table
2, the sulfur rich films generated by the greases increased
in area coverage with increasing grease consistency. This
observation is opposite the idea that softer more oil like
grease would perform more like oil counterpart. A more
important observation is that the greases were unable to
generate uniform tribo-film coverage at the higher 50 lb.
load. SEM-EDX images in Table 3 visually depict this
data.
The SEM-EDX data correlates well with wear scar width
data. Both data sets suggest that thickener is interfering
with phosphorus components and that load-carrying
capacity for grease formulations is mostly attributed
to sulfur EP additives, which function by producing
sacrificial tribo-films that shear away to smooth mating
surfaces and prevent metal-metal adhesion. However,
sulfur EP compounds that are designed for lubricant
applications typically function at high loads and
temperatures. Examination of temperature data (Figures
3 and 4) reveals that grease formulations produced very
different temperature profiles than the oil formulation.
– 16VOLUME 79, NUMBER 3
NLGI
Table 2: SEM-EDX Data on 3 mmX 4.5 mm Area of Timken Block
1.432
8.93
27.39
6.78
2.12
53.32
40
10
2.709
9.71
22.66
5.80
2.31
57.67
50
10
2,048
9,55
22,74
6.20
1.85
57.42
60
10
2,644
9.94
19.47
5.21
2.55
60.84
NLGI 1,
40
10
2.106
11,08
5.36
0.00
9.48
72.21
NLGI 1
50
10
1.559
12.06
9.32
0.00
9.57
67.19
NLGI 2
40
10
3.824
10.42
4.72
0.14
9.13
73.73
NLGI 2
50
10
1.610
11.35
7.95
0.18
10.58
67.46
NLGI3
40
10
4,354
10,70
7.33
0.00
8.46
71.66
NLGI 3
40
10
5.197
11.25
4.88
0.19
18.16
64.18
NLGI3
50
10
2.341
21.23
12.53
0.14
9.72
54.50
Grease
Table 3: SEM-EDX Images of Timken Blocks for Grease Formulations and 10 minute Runs
40 lb. Load
Grease
50 lb. Load
NLGI I
r
NLG1 2
J
NLGI 3
Aqua, red and blue area represents block background, tribo-film and high cathon content deposit respectively.
– 17 NLGI SPOKESMAN, JULY/AUGUST 2015
170
133
.J.
0
a
E
x
x
….
• Temp. ~C vs. Run Time, minutes Comments Wear Scar Dimensions
~ OilA
Grease A, NLGI 1
X Grease A, NLGI 2
Grease A, NLGI 3
95
x
58
20
0
3
5
Times minutes
10
8
Figure 3: Graph summarizing temperature profiles for 40 lb. Timken test runs
170
133
*
e
0
a
E
• Temp. ~C vs. Run Time, minutes Comments Wear Scar Dimensions
~ QuA
Grease A, NLGI I
• Grease A, NLGI 2
Grease A, NLGI 3
Grease A, NLGI 1
Grease A, NLGI 1
Grease A, NLGI 3
Grease A, NLGI 3
*
95
A
58
i~m~
20
0
3
5
Time, minutes
8
10
_________
Figure 4: Graph summarizing temperature profiles for 40 lb. Timken test runs.
The oil formulation at all loads produced rapid
rates of temperature increase that plateau quickly at
approximately 2 minutes. On the other hand, grease
formulations produced much slower and linear
rates of temperature increase that start to plateau at
approximately 6 minutes. As per Figure 5, the test
temperatures for the oil formulation were proportional
to contact pressures. In this case, contact pressures were
solely dependent on load since wear areas remained
constant at all test loads. This relationship is indicative
of steady and uniform tribo-films at the mating surfaces.
Conversely, contact wear areas for the grease formulations
were not steady and increased with time (Figure 2).
The wider scars led to reduction in contact pressures
that explained the slower rates of temperature increase
generated by the grease formulations but they are also
indicative of thin and un-uniform film formation. As
per Figure 5, the test temperature for grease formulations
tested at the 50 lb. load were inversely proportional
indicating temperature increase is partly due was to
friction produced by metal to metal contact.
This assumption is supported by SEM-EDX examination
of Timken blocks for partial Timken tests as shown in
Table 4 and Table 5. The tribo-film images for 3 minute
tests show almost no tribo-film formation but the wear
scars had no visible scoring. These observations indicate
that the thickener participated in carrying the test load.
– 18 VOLUME 79, NUMBER 3
NLGI
For the 4.75 minute tests, tribo-flim coverage increased slightly for the NLGI 1 grease and significantly
for NLGI 3 grease. Sulfur contents of the films were higher for 4.75 minute test than 10 minute test.
Nonetheless, the soap thickener appears to block the sulfur EP compounds from the surface, which
with time and pressure lead to unsustainable tribo-films as is the case for 50 lb. 10 minute tests. A
better understanding of how soap lubricates and interferes with sulfur EP additive could be gained
by performing this study’s test protocol on oil formulation lacking phosphorus components and by
running rheology and tribology tests with optical interferometry on grease bleeds following similar test
methodology as Cousseau and co-workers.
Figures Plots showing contact pressure and temperature relationship at end of test (10 minutes).
170
R2
=
0.9988
158
•
C-,
QuA
Grease, 40 lb.
• Grease, 50 lb.
0
ii. 145
133
120
100
50
250
200
150
Contact Pressure, MPa
Table 4: SEM-EDX Data for Partial Timken Runs
NLGI 1
50
3
0.023
22.64
20.91
0.00
6.88
49.57
NLGI 1
50
4.75
0.641
11.01
6.33
0.27
12.79
67.51
NLGI 1
50
10
1.559
12.06
9.32
0.00
9.57
67.19
NLGI3
50
3
0.117
30.18
14.23
0.00
3.83
50.63
NLGI 3
50
4.75
3.808
10.90
4.67
0.22
15.90
66.43
NLGI 3
50
10
2.341
21.23
12.53
0.14
9.72
54.50
– 19 NLGI SPOKESMAN, JULY/AUGUST 2015
Table 5: SEM-EDX Imagefor Grease Formulation at 50 lb. Test Load
TestTime, minutes
Grease, NLGI I
Grease, NLGI 3
3
I’
4.75
10
Aqua, red and blue area represents block background, tribo-flIm and high carbon content deposit respectively.
In regards to lack of phosphorus content in grease
tribo-flims, oil separation experiments were especially
insightful. As per Table 6, bleed oil that separated
from greases contained equal amount of sulfur content
as the oil formulation but no detectable amount
of phosphorus. Thus, it is concluded that a strong
interactions with soap thickener inhibited phosphorus
compounds from adhering and reacting metal surfaces.
To conclude, the study demonstrated that lithium
soap thickeners can inhibit the function of EP and
antiwear additives. Two basic mechanisms were
proposed based on experimental observations. One
mechanism involves soap blocking additives from
metal surfaces as was the case with sulfur EP additives
of this study. The second mechanism involves chemical
interactions between the soap and additives that
prevent the additives from adhering and reacting with
metal surfaces. This mechanism would be more in
effect with polar and ionic additives such as phosphorus
compounds that are commonly used in gear oil additive
packages.
– 20 VOLUME 79, NUMBER 3
NLGI
Table 6: Phosphorus and Sulfur Contents for Oil Formulation and
Grease Bleed Collected at 140 °Cfor 8 h.
330
O~96
NLGI 1
0
1.02
NLGI2
0
1.08
NLGI 3
0
1.15
Oil
Grease Bleed
References
Canter, N. Grease additives: Important contributors not be overlooked. Tribology &
Lubrication Technology, December 2013, 28-29.
Lugt, P.M., Grease Lubrication in Rolling Bearing. John Wiley & Sons; 2013.
Cousseau, T., Bjorling, M., Graca, B., Campos, A., Larsson, R. Film thickness in
a ball-on-disc contact lubricated with greases, bleed oils and base oils. Tribology
International 2012; 53: 53-60.
Sivik, M.R., Zeitz, J.B., Bayus, D., Interactions of Zinc Dithiophosphates with
Lithium 1 2-Hydroxystearate Grease. Presented at the Annual Meeting of the NLGI, at
West Palm Beach Buena, Florida, (2001).
Kaperick, J., Timken OK Load Media Bias? A Comparison of Timken Response
to Similar Additive Systems in Both Grease and Oil Formulation. Presented at the
Annual Meeting of the NLGI, at Lake Buena Vista, Florida, (2006).
—
Phillips, WD., Ashless Phosphorus-Containing Lubricating Additives in Lubricant
Additives: Chemistry and Applications, Second Edition, CRC Press,Rudnick, L.R. ed.
(2009).
Bovington, C.H., Friction, wear and role of additives in their control in Chemistry
and Technology of Lubricants, 3rd Edition, Springer, Mortier, R.M., Fox, M.F,
Orszulik, S. T. ed. (2010).
-21 NLGI SPOKESMAN, JULY/AUGUST 2015