NLGN

Fr~ctäon Mod~flers for Lubr~cat~ng Greases
Gareth Fish, PhD CLS CLGS, The Lubrizol Corporation, Wickliffe, Ohio, USA
Presented at NLGI’s 80th Annual Meeting, June, 2013, Tucson, Arizona, USA

Dynamic friction, also called kinetic friction in many
texts, is the component of resistance to motion of the
contacting bodies as they move relative to one another.
Both need to be considered and understood in order
to reduce the friction between the surfaces. Both play
a significant role in friction measurements determined
using the SRV test. Static friction is caused by the
interaction of the surfaces trying to start moving across
one another and has two components: interfacial
adhesion and roughness interaction. Dynamic friction
is caused by the interaction of the surfaces moving
across one another and is made up of three compo
nents: interfacial adhesion, roughness interaction, and

Introduction

F

riction
onefully
of the
most important
in nature,
but
it isisnot
understood.
It is theforces
reaction
force
that opposes motion of two bodies in contact (1).
Friction forces prevent people from falling over when
they walk. Friction losses in the driving mechanism
of automobiles account for approximately 25% of the
energy losses, but without friction the wheels would
slip and the vehicle would be undriveable. The friction
coefficient is typically defined as the ratio of the normal
force to the tangential force and is dependent on the
local tribological system.
From his preserved notebooks, Leonardo da Vinci
was known to have studied friction, but the first laws of
friction were devised by G. Amontons (1699) and later
modified by C-A de Coulomb (1 785) and L. Euler (2).
Amontons’ first law of friction states that “The force
of friction is directly proportional to the applied load”.
This can be expressed as follows:
Friction Force

=

F=p.N

Equation 1

Normal force

J,

Driving or
TangenFor~

or object weight
Friction Force
-,

Normal force x friction coefficient

Motion
*-

Figure 1

In the case of an object being pulled along, the normal
force is the weight of the object.
Amontons’ second law of friction states that “The
force of friction is independent of the apparent area of
contact” and assumes that the contacting materials are
perfectly rigid and inelastic. This is clearly not true for
real world applications, a good example being auto
mobile wheel tires, where increasing the width provides
increasing traction. Friction is dependent on the real
area of contact which is different form the apparent
area of contact. There are essentially two types of fric
tion: static and dynamic (3), Static friction, also called
adhesive friction, is the component of resistance to
motion as the contacting bodies try to move from rest.
It is not the same as inertia.

Tribological system

flrft,rfl.rr,r#,~,n≠#4flwrr..

$fl7~4?4~

if ~‘

flW$~CWU1

Load

w = load on each asperity
a = asperity contact patch
Figure 2 — Asperity interaction
-~

rfln~~~

—23—

NLGI SPOKESMAN, MARCH/APRIL 2014

NLG~

the shearing of the lubricant or surface layer separating
the contacting loaded surfaces. Coulomb’s law of
friction states that “Dynamic friction is independent of
the sliding velocity.” For dry sliding without interfacial
transfer this can be true but for most lubricated con
tacts this is not true. For an unlubricated steel on steel
couple, the adhesive portion of static friction coefficient
can be estimated as follows:
When the load or weight is applied, the surfaces
elastically deform until the contact area is large
enough to support the load. The true area of contact,
A (A = ta), can be determined from the load (L = ~w)
and the hardness of the material (H).
A

=

LI H

Equation 2

The asperities undergo localized adhesion and each
adhered junctidn has a shear strength, S. To produce
sliding, these junctions need to be sheared by an
applied shear force F, where F = A.S. Taking standard
material properties, the hardness is three times that of
the yield strength, Y (H= 3Y) and the shear strength is
Y12. From this, we have the following:
Shear Force I Load = Shear Strength I Hardness
Coefficient of Friction or
p=FIL=SIH=Y1213Y=116=0.167

=

Equation3

This figure of p= 0.167 is typical for most industrial and
common engineering steels. In order for rough surfaces
to slide over each other, mating asperities need to slide
or bend each other out of the way. This is typically
referred to as the ploughing term of static friction.
The more asperities interacting the greater the size
of the ploughing term. The ploughing term depends on
the material properties of the surface (Hardness etc.)
and on the roughness of the surface i.e. the asperity
heights, shapes and population density. As static fric
tion is typically greater in magnitude than is dynamic
friction, more effort is typically needed to reduce asper
ity adhesion and interaction at the turning points of
reciprocating sliding friction testers such as SRV. With
dynamic friction, the magnitude of the forces involved
depends on the amount of surface interaction, If the
surfaces are sliding together under boundary lubrica
tion, the shear properties of the surface film and any

Motion

Motion

1w>
Friction
Force

________

C

Friction
Force

~

Figure 3 — Asperity interaction – Ploughing

chemical additives present become of high importance
along with the roughness of the surface. If films can
be generated that partially separate the surfaces, the
asperity interaction and additive chemistry becomes
gradually less important. If, when in motion, the sur
faces can be sufficiently separated so that there is no
surface asperity interactions, the friction will depend
solely on the viscometric properties of the lubricant
(viscosity, viscosity-temperature, pressure-viscosity and
traction coefficients).

Friction Issues
Under pressure from governmental legislation, OVMs
and OEMs are looking to reduce energy losses
through friction. Small reductions in friction translate
as improved fuel efficiency. As most of the losses on
a vehicle are in the engine and transmission, for the
last few years OVMs were focused on improving the
efficiency of engines and drivelines with little or no
focus on other components. All accessory equipment
and some transmission components have grease
lubricated rolling element bearings. However, the big
gest grease usage on a modern passenger car or light
truck is in the constant velocity joints (CVJ5), which
transmit power and motion from the gearbox, differ
ential or transfer case to wheels (4). Based on recent
figures, more than 85% of all passenger cars and light
trucks have CVJ transmissions (5). The ball type rolling
element bearings used are very efficient and there is
little scope for energy savings with applying different
greases where these are used. However with tapered
roller bearings there is some scope for reducing friction
on the sliding contacts, but no increase in wear can
be tolerated.

—24—
VOLUME 78, NUMBER 1

NLG~

‘V

The only significant area for improving fuel efficiency
related to grease is in the CV jointed transmission.
Ball type CVJs, both fixed and plunging, have a lot
of sliding in their contacts and this coupled with high
stresses make them theoretically good candidates for
optimization. However the mode of operation coupled
with the complexity of the mechanism, make this a dif
ficult target. If the friction coefficient is made too low,
the internal components which rely on frictional forces
to maintain control of the cage and balls, will allow the
balls to skid and bounce, resulting in noise whenever
the steering wheel is operated to make right or left
hand turns.
Conversely, tripod type plunging CVJs need low fric
tion greases to prevent the generation of noise, vibra
tion and harshness (NVH) issues in the vehicle. When
the joints are installed at low angles, axial forces arising
from friction in the contacts are small, and typically
insufficient to cause audible noise or detectable vibra
tion. As the installation angle of the sideshaff and joint
increases, the generated forces become significant.
By using lower friction greases, the working angle of
the joint can be increased without compromising driver
comfort and without having to utilize more complex and
costly wide angle joints (4).
Historically, the internal components of fixed ball
type CVJ had very smooth tracks from the fine grind
ing operation used to finish the components after heat
treatment. As control of manufacturing became better,
improved heat treatments meant less distortion and the
fineness of grinding was reduced or eliminated entirely
resulting in the increased roughness of surfaces found
on today’s components. The rolling elements used have
not changed so this increased roughness on the tracks
and cages has caused the tribological contact condi
tions to change.

Historical Additive Issues
Since molybdenum disulfide (M0S2) (1960) and organic
molybdenum (M0DTP, M0DTC) friction modifiers (1985)
were first used in CV greases, there have be multiple
issues with the price and availability of molybdenum
compounds (6). In early Austin Mini cars, the original
ball type CVJ grease was replaced with a clay thickened

grease containing 20% Mo52. At that time the price
of molybdenum was low and the heavy grease solved
a noise and vibration (NVH) issue that was causing
customer dissatisfaction. In 1970, lithium soap greases
with 3% became the standard for ball type CVJ and
calcium complex greases for tripod type joints. Both
types of grease contained anti-oxidants, antiwear and
extreme pressure additives and rust and corrosion
inhibitors (4).
Since around 1990, organic molybdenum complexes
have been widely used in CVJ greases to reduce
lubricated sliding friction coefficients from a range of
0.12 to 0.15 to between 0.04 and 0.08 as measured
by the SRV tester. However most of the early organic
molybdenum complex containing CVJ greases gave
significant issues with wear and durability of the metal
components and caused seal compatibility issues, some
of which were described in US patent 6,656,890 (7).

Friction testing
The first friction and wear machine was developed
around 1800 by Charles Hatchett (2), who also found
fame as the discoverer of the rare earth Niobium.
Working at the British Museum in London, Hatchett
used his apparatus to test the wear properties of
gold coins.
Historically the sliding 4-ball wear tester was used
to measure friction. It was standardized for greases
in 1964 as ASTM D2266 (8) but the friction measur
ing system was added later. One important issue with
using the 4-ball machine to measure friction is the
bearing used in the apparatus influences the friction
values. Some machines have air bearings and others
have tapered roller bearings. The latter have higher
motion resistance and this makes comparative data
difficult to interpret.
Cameron-Plint instruments were developed in the
1960s to measure friction and wear. The main two
instruments used are the TE7O Micro Friction machine
(commonly known as the HFRR) and the TE77 High
Frequency Friction Machine (commonly known as
the Cameron-Plint). Both machines were designed to
have sinusoidal reciprocating motion. Improvements
in instrumentation led to the development and use of

—25—
NLGI SPOKESMAN, MARCH/APRIL 2014

NLGII

electrical resistance / capacitance methods for deter
mining film thickness. Around 1990, PCS instruments
introduced their version of the HFRR. It had technical
enhancements over the older HFRR and was subse
quently used for running the ASTM D6079 diesel fuel
lubricity test (9),
The SRV test method was developed in the 1980s
by Optimol Olwerke in Germany. The test instrument
company manufacturing arm was spun off as a sepa
rate company in the 1990s. The latest instruments have
many features that older machines did not have and
the upgraded drive and control systems allow higher
loads and better data acquisition. The general principle
is a “Square” wave motion driven forward and spring
ing back. The tester measures the reaction force from
the motion under load and from this the maximum,
minimum and average values of friction during the test
run are calculated. The test method was standardized
as ASTM D5707 (10). The software produces a fric
tion coefficient-time trace. The output from the force
transducer can be captured and digitized to give static
and dynamic coefficients. A comparison of the test
ing parameters of the four types of test machines is
included in Table 1 below.

Test rig
Contact Geometry
Contact shape
Contact motion
Rpm / Frequency
(Hz)
Speed (m/s)
Duration (mins)
Sliding distance (m)
Temperature (°C)
Load (N)
Stress (GPa)
~

Many newer requirements for low friction greases
specify SRV test results to be reported or targets
needed to be met. For many industrial greases, values
of <010 are required but most companies specifying
targets for automotive or CVJ greases want <0.08
and in some cases <0.06. The choice of SRV test rig
seemed to be almost unanimous amongst CVJ grease
developers, but the test conditions are an area of vari
ance. Some companies rely on DIN / ASTM standard
tests whilst others used different loads, lower frequen
cies and longer strokes. Conditions need to be chosen
based on the capabilities of the SRV instrument and
the operating conditions of the application. The SRV
only looks at sliding and it is generally assumed that
rolling friction, being an order of magnitude lower than
sliding friction, will not be significant. Correct specimen
geometries also need to be selected to try to mimic the
contact conditions of the application under study.
Looking at the contact conditions inside a CVJ on
a mid-sized front wheel drive passenger car, a driving
speed of 70 miles per hour (mph) or 110 kilometers
per hour (kmh) equates to a rotation speed of about
1800 rpm. This equates to a reciprocating speed
of 30 Hz. For cars driving at 100 mph (160 kmh), a

Table 1
Comparison of testing parameters
Sliding 4-ball
HFRR
Cameron-Punt
1 rotating ball loaded Ball on smooth
Ball on smooth
on three fixed balls
flat plate
flat plate
3 points
Point
Point
Unidirectional
Reciprocating
Reciprocating
sinusoidal motion
sinusoidal motion
1200
20 to 50
20

Ball on smooth
flat plate
Point
Reciprocating
square wave
50

0.4
60
1436
75
392
3.05

0.2
120
1440
50 or 80 or higher
200
2.71

0.04 or 0.1
0.4
75
120
200 or 450
3110
No set temperature 100 or 120
1.96 or 4.91
200
0.81 or 1.10
0.1

~

SRV

r<, ~‘v

—26—
VOLUME 78, NUMBER 1

~fl

NLG~

10% added PAO to give a final ISO VG of 100. The
frequency of 40 Hz is more typical and at a top auto
lithium complex was made in a laboratory grease kettle.
bahn speed of 155 mph (260 kmh) the driveshafts
It used azelaic acid as the complexing agent along with
would be rotating close to 3000 rpm which equates to
1 2-hydroxystearic acid in a straight cut 600N group I
a frequency of 50 Hz. Typical contact stresses in ball
paraffinic oil with a viscosity of 112 mm2/s at 40°C.
type CVJs are 1.0 to 4.0 GPa (11) and those in tripod
For subsequent testing, commercially manufactured
types are 1.0 to 2.0 GPa (12). These stresses equate to
simple lithium and lithium complex base greases were
loads of 100 to 500 N. At 100 N many older SRV test
used with the same base oil as the original laboratoryrigs can be somewhat bouncy, but will run satisfactorily
manufactured greases.
at 200 N. The two loads typically chosen are 200 N for
standard testing and 500 N for high
load tests. At high driving speeds, the
Table 2 Friction modifier additives
amplitude of the motion of the internal
Additive
Commercial or Additive type or
Number Descriptor developmental basic chemistry
components of a CVJ is small, but
this depends on the installation and
1
Zinci
Commercial
zinc dithiophosphate
running angles on the vehicle. The
2
Molyl
Commercial
molybdenum complex
standardized stroke length used in
SRV tests is 1 .0 mm, but based on
3
Moly2
Commercial
molybdenum complex
published data (13) a longer stroke
4
Moly3
Commercial
molybdenum complex
at lower frequency would be more
5
Phosi
Commercial
phosphorus compound
representative of actual service use.
6
Phos2
Developmental
phosphorus compound
However for the purposes of this
7
Phos3
Commercial
phosphorus compound
study to gather general data, the
D5707 conditions listed in Table 1
8
Deti
Commercial
calcium detergent
would be used.
9
Det2
Developmental
calcium detergent
10
Det3
Developmental
calcium detergent
Base greases
11
Det4
Commercial
fatty detergent
Based on published data (4), the
Disl
Commercial
fatty acid
most common globally used thickener 12
systems for low friction greases are
13
Dis2
Commercial
fatty acid
urea derivatives and simple lithium
14
Dis3
Developmental
fatty ester
soaps. For some higher temperature
15
£01
Commercial
engine oil FM (Mo-free)
applications, lithium complex greases
16
£02
Commercial
engine oil FM (Mo-free)
are also used so it was decided, to
17
£03
Developmental
engine oil FM (Mo-free)
use initially lithium complex and urea
thickened greases. The first selection
18
£04
Commercial
engine oil FM (Mo-free)
issue was type and viscosity of the
19
£05
Developmental
engine oil FM (Mo-free)
base fluid. Many low friction greases
20
Nanol
Commercial
nano particle
have base oil viscosities between ISO
21
DL1
Commercial
ATF additive
VG 68 and 150. The laboratory urea
grease had an MDI-fatty amine thick
22
DL2
Developmental
gear oil additive
ener (—12% wt in the finished grease)
23
DL3
Developmental
gear oil additive
and was made in a 3 liter resin flask.
24
DL4
Developmental
gear oil additive
The base oil was a mixture of naph
25
DL5
Developmental
gear oil additive
thenic and paraffinic mineral oils with

_________

_________

_________

_________

_________

_________

________

_________

—27—
NLGI SPOKESMAN, MARCH/APRIL 2014

NLGI

Additives were tested at 3wt% on their own and at
2wt% on top of a model non-friction modified pack
age, which included zinc and phosphorus antiwear,
sulfurized EP, anti-oxidant, corrosion inhibitor additives
at normal treatment levels. These levels were chosen
based on published data (6) and on recent testing in
our own laboratory, where friction modifiers were typi
cally found to be unstable when tested at less than
2%wt and no more effective when tested at above
3%wt.
We procured 15 additives sold commercially as fric
tion modifiers, and these, along with another 10 new
compounds, were incorporated into the four base
greases, two neat bases and two formulated bases
and tested in the first round. Subsequently, additional
combinations were put together and tested.

a third run was carried out and if one of the three runs
was distant from the other two then it was discarded
as an outlier. If the third result was in between those of
the first two then the three results were averaged.
It was initially attempted to run the greases for
1 hour only, but initial testing showed that some
greases had not stabilized after 1 hour and so it was
decided to run everything for 2 hours so that all the
wear scars had seen the same total sliding distance at
1440 m. The test results for the reference greases run
in the two hour tests are included in Table 3.
The results of the more than 250 grease tests were
reviewed in detail. Many components tested showed
very high friction, high wear or a combination of both.
Many of the additives tested had no significant effect
on friction or wear with results similar to the baselines.
The best pairs of results from the first round of testing
are included in Table 4.
As seen from the data table molybdenum com
pounds gave the best results of any of the additives
tested. MoIy2 worked best in the fully formulated urea
grease but took some time to fully run in. Looking at
the lithium complex base, the Molyl additive gave the
lowest friction of the single components tested.
Combinations of molybdenum additives with zinc
dithiophosphates were then studied in the fully formu
lated lithium complex and urea bases, along with a
series of combinations of molybdenum-free additives.
These results are included in Table 5.
It was then decided to look at combinations of the
three molybdenum complexes in conjunction with the
EO1 and E02 commercially available additives to deter
mine if a mixture of compounds would act synergisti

Test results
The initial testing was carried out on the base greases
and all subsequent results were compared to those
results. The sampling rate was 1 data point per second
giving 7230 friction coefficient values per test, which
were transferred to a spreadsheet file to enable data
sets to be calculated and friction graphs to be plot
ted. The wear scar on the ball was measured in two
directions 900 apart. The key data sets calculated and
recorded for every test were average wear scar diam
eter, and the end of test (EOT, averaged over the last
30s of the test) and average (after reaching the test
load, approximately 60 s into the test) friction coef
ficients. Every test was run in at least duplicate and the
data was averaged, If the results did not match then

Table 3
Test results for the base greases
Grease
Base Urea
Formulated urea

Average friction
coefficient
0.161
0.150

Base LiX

0.134

0.124

095

Formulated LiX

0.149

0.149

0.80

*

EOT friction
coefficient
0.164
0.145

Average wear scar Comments
diameter (mm)
0.99
very high wear
0.82
high wear

~a

~

—28—
VOLUME 78, NUMBER 1

very high wear; lower
than expected friction
high wear

NLGD

cally. It is known that molybdenum additives are among
the most expensive additives available and the target
was to try to get the coefficient of friction as close to
0.04 with minimum running-in, low wear and good sta
bility when using the minimum amount of molybdenum
in the grease. Greases containing 10 new combinations
of additives were screened and of these three (LF4,
LF9 and LF1 3) were selected for further evaluations.
It was also decided to put the combinations into the
commercially manufactured greases urea, lithium, and
lithium complex bases that contained a similar baseline
package to that used in the earlier experiments. These
results are included in Table 6.
Discussion
The base line tests all showed two runs that were
very similar in terms of the values of the friction coef
ficients and wear scar diameters, but the magnitude

Grease
Number
15
18
20
38
97
112
37
50
94
63
105
110
117
61
74
100

of the results were slightly unexpected. The base urea
gave higher friction and wear than expected. When
the baseline additive package was included the friction
coefficient was reduced to the 0.14 to 0.15 range but
the wear was still higher than expected. For the lithium
complex base, the friction without the additive pack
age was slightly lower than expected but the wear was
much higher. When fully formulated the fiction coef
ficient increased to the expected level of 0.14 to 0.15,
but the wear was higher than expected. However it
was similar to that of the fully formulated urea grease.
The 100 combinations tested showed that It is very
difficult to achieve low levels of friction desired with
most components, and have little or no effect on fric
tion and some increasing it above the levels seen for
the base grease. Only 4 single components tested
gave values below the desired 0.10 and of these,
two development additives (E03 and DL5) gave it in

Table 4
Test results for selected FM additives
Average friction
EOT friction
Grease + additives
coefficient
coefficient
Base Urea + Zinci
0.142
0.138
Base Urea + EO3
0.099
0.097
Base Urea+ EO1
0.105
0.105
Base Urea + Nanol
0.106
0.101
Base Urea+ Dis3
0.106
0.105
Base Urea+DL5
0.094
0.094
Formulated urea + Moly2
0.071
0.046
Formulated urea+ DL3
0.116
0.113
Formulated urea+DL4
0.112
0.111
Base LIX + E03
0.123
0.119
Base LIX + DL1
0.121
0.119
Base LiX+DL5
0.115
0.113
Base LiX+ EQS
0.118
0.119
Formulated LiX + Molyl
0.053
0.043
Formulated L1X+E02
0.112
0.112
Formulated L1X+ Dis3
0.116
0.122

~~~ ~

~..

~

—29—
NLGI SPOKESMAN, MARCH/APRIL 2014

Average Wear Scar
diameter (mm)
0.48
0.48
0.46
0.52
0.48
0.41
0.49
0.47
0.43
0.47
0.45
0.41
0.53
0.45
0.51
0.53
~-

I

NLGI

the urea base grease. Neither component worked as
well in either fully formulated grease, but both looked
better than the average in the lithium complex base.
Further work to understand the behavior of these two
additives is underway. As discussed earlier, the two
molybdenum complexes did not work on their own in
the base grease even at 3% treat rate. From previously
published work, molybdenum complexes typically need
to be added in conjunction with zinc dithiophosphate
(6) to perform well. Looking at the results for greases
14, 15, 56, and 57 in Table 5, Molyl additive did not

work very well in the urea grease but did very well in the
lithium complex base. Moly 2 in the urea base behaved
better than the Molyl but not so in the lithium complex.
Further comparisons with the two additives in the fully
formulated base greases showed that Moly 1 is able to
give lower friction but seems to work best in the lithium
complex, but only when zinc is present. In the presence
of other additives it takes longer to stabilize suggest
ing that it is competing for surface sites along with the
other additives present. Moly2 gives reduced friction
when used with zinc in the urea grease, but only gives

Table 5
Test results for combinations of additives

14
15
16
17
56
57
58
59
115

Base Urea +Zincl +Molyl
Base Urea + Zinci + Moly2
Base Urea + Combination 1
Base Urea + Combination 2
Base LiX + Zinci + Molyl
Base LiX + Zinci + Moly2
Base LiX + Combination 1
Base LiX+ Combination 2
Base LiX+ Combination 3

~

combination

LF4

LF9

LF13
~

0.139
0.081
0.149
0.105
0.049
0.092
0.147
0.135
0.123

0.118
0.105
0.151
0.109
0.047
0.074
0.148
0.135
0.122

0.81
0.56
0.78
0.49
0.50
0.41
0.52
0.52
0.53
~W

Table 6
Test results for selected FM additives
Average friction
EOT friction
Base Grease
coeffici~nt
coeffici~~t.
Urea
0,053
0.053
Lithium
0.057
0.059
Lithium complex
0.046
0.050
Urea
0.052
0.046
Lithium
0.061
0.061
Lithium complex
0.062
0.058
Urea
0.052
0.055
Lithium
0.055
0.057
Lithium complex
0.061
0.058
~—~—

-~—

—30 —
VOLUME 78, NUMBER 1

7’~~#~

Average Wear Scar
~di~rneter (mm)
0.53
0.48
0.47
0.50
0.46
0.47
0.43
0.47
0.46

NLG~

Summary

low friction when it interacts with the other additives in
the package. It is suspected that this particular com
pound needs not only zinc but active sulfur to function
as a low friction additive. Additional work is ongoing
to study these effects and elucidate its mechanism of
operation. In the lithium complex grease, none of the
additives apart from the molybdenum complex gave
friction coefficients below 0.10 and only E02 gave
values below 0.12 when incorporated into the fully
formulated base.
It is suspected that the reason the non-molybdenum
friction modifiers do not do very well in grease is
because the additives get caught up in the thickener.
Many of the commercial and developmental nonmolybdenum additives had fatty chain lengths similar to
the 18 carbon atoms on the lithium 1 2-hydroxystearate
chain. It would appear that only molecules that have
different organic tails to the thickener stand any chance
of freeing themselves from the thickener and getting
to the surface to adsorb or to react to form low fric
tion layers. One potential reason for the molybdenum
additives to work better is that they have much shorter
carbon chains (4, 8, or 13 or a mixture of 8 and 13)
attached to them than do the fatty acids and amines
of the thickener molecules. Work investigating this
is ongoing.
Further testing on the combinations of the three
molybdenum additives with the best commercial nonmolybdenum friction modifiers showed that LF9 in the
urea grease gave the lowest value of friction coefficient
but the friction trace showed significant running-in and
instability. The best overall combination was LF4, which
was a mixture of Molyl, Moly2 and EO1, and it gave
the best all round friction performance in all three of
the greases tested. The molybdenum content of this
combination was lower than that of other two pack
ages tested in the final round and typically lower than
that seen in commercial packages. This would suggest
that non-molybdenum friction modifiers do have a role
to play in reducing friction and in reducing the cost of
the package by reducing the high cost molybdenum
content of the grease.

Work has shown that typically, fatty long chain fric
tion modifiers that work in liquid lubricants such as
crankcase / engine or driveline oils do not work so well
in greases because they appear to get trapped in the
thickener.
Individual molybdenum compounds do not work very
well on their own and need other additives to help build
their low friction films.
Combinations of at least two molybdenum additives
with other components have led to the development of
new packages for low friction greases.

Conclusions
This work has shown that using molybdenum additives
continues to be the best way to achieve low friction
coefficients in grease, and that their combination with
other molybdenum-free additives provides an effective
way to formulate optimized low friction additive packages.

Acknowledgements
The author wishes to acknowledge many co-workers
and departments within The Lubrizol Corporation for
their contribution to this work.

References
(1) Bowden, FR and Tabor, D., “The Friction and
Lubrication of Solids” Oxford University Press (1950)
(2) Dowson, D., “History of Tribology (Second Edition)”
Professional Engineering Publishing (1997), ISBN
1 -86058-070-X
(3) ASTM G40-12 “Standard Terminology Relating to
Wear and Erosion” (2012) ASTM International, West
Conshohocken, PA
(4) Fish, G. “Constant Velocity Joint Greases” NLGI
Spokesman 1999, Vol 63(9), p14-29
(5) GKN plc. Annual Report 2012
(6) Fish, G and F, J, “The Effect of friction Modifier
Additives on Constant velocity joint performance”,
NLGI Spokesman, 66, No. 4, pp. 19-29, (2002)

—31—
NLGI 5POKESMAN, MARCH/APRIL 2014

NLG~

(7) Fish, G et al “Grease composition for Constant
Velocity Joints” US patent 6,656,890

ABOUT THE AUTHOR
Gareth Fish, PhD CLS CLGS
The
Lubrizol Corporation
Gareth Fish,

(8) ASTM Standard D2266-08 “Standard Test Method
for 4-Ball Test Machine,” (2008) ASTM International,
West Conshohocken, PA

Ph.D. B.Sc. (Honours), 1984, Ph.D. in
Tribology, Imperial College of Science,
Technology and Medicine, University of
London. Dr. Fish joined the UK Ministry
of Defence, Fuels and Lubricants
Branch in 1988. He then was appointed
Senior, then Principal Tribologist at
GKN Technology Ltd., Wolverhampton,

(9) ASTM D6079-1 1 “Standard Test Method for
Evaluating Lubricity of Diesel Fuels by the HighFrequency Reciprocating Rig (HFRR)”, (2011) ASTM
International, West Conshohocken, PA
(10) ASTM Standard D5707-05 “Standard Test Method
for Measuring Friction and Wear Properties of
Lubricating Grease Using a High-Frequency,
Linear-Oscillation (SRV) Test Machine,” (2011)
ASTM International, West Conshohocken, PA

England in 1990. In 1999, he became the Global Technical
Coordinator for Tribology and Grease. He relocated to Auburn
Hills, MI, USA, with GKN Automotive, Inc. in 2002. In June
of 2007, he became the Technology Manager for Grease for
The Lubrizol Corp., Wickliffe, OH, USA. He has authored 12
papers on tribology and grease, has two granted patents and
one pending. He is a recipient of the Chevron Products Award
(NLGI) for Outstanding Technical Paper and is an NLGI Fellow.
Member Royal Society of Chemistry, STLE, Energy Institute,
NLGI and is a Chartered Scientist (CSci).

(11) Cole, S. J, and Fish, G. “The Tribology of the Ball
Type CVJ”, (1997) WTC I. Mech. E. London
(12) Fish, G. and Cole, S.J. “The Tribology of the Tripod
Type CVJ”, (1997) WTC I. Mech. E. London
(13) Schmelz, F., Count Scherr-Thos, H.-Ch., and
Auktor, F, “Universal joints and Driveshafts” (1992),
Springer and Verlag

—32

VOLUME 78, NUMBER 1