Tribol Lett (2012) 48:355–365

DOI 10.1007/s11249-012-0030-9

ORIGINAL PAPER

Lubrication Properties of Polyalphaolefin and Polysiloxane

Lubricants: Molecular Structure–Tribology Relationships
Thomas Zolper • Zhong Li • Changle Chen •
Manfred Jungk • Tobin Marks • Yip-Wah Chung •

Qian Wang

Received: 10 April 2012 / Accepted: 10 August 2012 / Published online: 24 August 2012
Ó Springer Science+Business Media, LLC 2012

Abstract This study investigates the rheological proper- lubrication regime (boundary to full film) by optimal tuning
ties, elastohydrodynamic film thickness, and friction coef- of the molecular mass and chain branching.
ficients of several commercially available polyalphaolefin
(PAO) and polydimethylsiloxane (PDMS)-based lubricants Keywords Silicones  Synthetic base stocks 
to assess relationships between molecular structure and EHL with non-Newtonian lubricants  EHL friction 
lubricant performance. Molecular structures and masses Oxidation resistance
were determined by nuclear magnetic resonance spectros-
copy and gel permeation chromatography, respectively.
Density and viscosity are measured from 303 to 398 K, while
elastohydrodynamic lubricant film thickness and friction 1 Introduction
measurements were made at temperatures, loads, and speeds
that are representative of boundary, mixed, and full-film The development of synthetic lubricants in recent decades
lubrication regimes. The results show that PDMS-based has been the result of concerted efforts to formulate
lubricants are thermally and oxidatively more stable than lubricants with optimal rheological–tribological properties
PAOs, while the viscosity of PDMS-based lubricants is for diverse applications. The currently available lubricants
generally less temperature sensitive than PAOs, except for range from natural and petroleum-derived hydrocarbons
highly branched polysiloxanes. In particular, this study (mineral oils) to synthetic hydrocarbon- and silicone-based
provides quantitative insight into the use of PDMS-based polymers.
lubricants to obtain low friction through the entire Polyalphaolefins (PAOs) are synthetic hydrocarbons
with low side branch densities and comparatively short
molecular lengths. They were developed as high perfor-
mance base-stock oils for automotive and industrial
T. Zolper (&)  Y.-W. Chung  Q. Wang applications including motor oils, transmission fluids, and
Department of Mechanical Engineering, Northwestern hydraulic fluids. PAOs are composed of hydrogenated
University, Evanston, IL 60208, USA olefin (alkene) oligomers having the general formula of
e-mail: [email protected]
CnH2n?2. Alpha-olefins are a class of olefins having a
Z. Li  C. Chen  T. Marks carbon–carbon double bond between the first and second
Department of Chemistry, Northwestern University, Evanston, carbon atoms of the hydrocarbon chain. PAOs in this study
IL 60208, USA
are mixtures of dimers, trimers, and tetramers of 1-octene,
M. Jungk 1-decene, and 1-dodecene (Fig. 1a). PAOs have a greater
Dow Corning GmbH, Rheingaustr. 34, 65201 Weisbaden, consistency in molecular structure and molecular mass
Germany distribution resulting in more consistent lubrication prop-
erties than their mineral oil counterparts. PAOs generally
Y.-W. Chung
Department of Materials Science and Engineering, Northwestern have greater viscosity indices and vapor pressure than the
University, Evanston, IL 60208, USA mineral oils in current use [1].

123
356 Tribol Lett (2012) 48:355–365

Fig. 1 Molecular structures of a PAO, b PDMS, and c PPMS

Siloxanes are derived from the controlled hydrolysis of non-Newtonian shear-thinning fluids [11, 12]. Equation 1
dialkylchlorosilanes to produce silanols, which polymerize is a simplification of the Hamrock–Dowson film thickness
to form polydialkylsiloxanes, as exemplified by poly- equation made by combining the reduced parameters into a
dimethylsiloxane (PDMS; Fig. 1b). Siloxanes have constant k [10, 13–15].
exceptional thermal and oxidative stability in addition to
h ¼ kU 0:67 g0:67
0 a
0:53
ð1Þ
large viscosity indices [2]. They are also water repellent,
biologically inert, and have low surface tension that allows Measurements of film thicknesses have been previously
them to spread more evenly on surfaces than most hydro- made for PAOs over a broad viscosity range [1, 11, 15–21].
carbons [3]. Siloxanes are amenable to several different Over the range of molecular masses studied here, PAO
branch architectures. lubricants exhibit Newtonian flow characteristics, but shear
Siloxanes have silicon–oxygen backbones instead of the thinning occurs at higher molecular masses [19, 20]. The
carbon–carbon backbones of hydrocarbons. The strength of literature data on siloxane film formation are more limited
the Si–O bond (*460 kJ/mol) exceeds that of the C–C bond [16, 22–25], especially in the correlation of molecular
(*348 kJ/mol) [4], and siloxanes are more thermally stable structure to tribological performance. Dyson and Wilson
than hydrocarbons. In addition, siloxane molecules are more [22] used capacitive methods to measure the film
flexible than the corresponding hydrocarbons due to the thicknesses of PDMS in several viscosity grades. They
longer Si–O bonds (0.164 vs. 0.153 nm for C–C), and hence found that the fluids do not maintain Newtonian film-
have lower steric hindrance to rotation about the Si–O bonds forming behavior at viscosities above 100 mPa s [12, 22].
and greater, more flexible Si–O–Si bond angles (*143° Bair and Winer [11] developed correction factors to
vs. *110° for C–C–C). This enhanced flexibility affords estimate film reduction due to shear thinning for high
increased macromolecular compactness, lower melting molecular mass PAOs and some siloxanes [25].
temperatures [4], and lower glass transition temperatures This study presents a comparative analysis of how the
[5]. The high structural flexibility of PDMS provides a molecular mass and structure, including branch mass and
baseline reference for more heavily branched siloxanes. structure, affect rheological and lubrication properties of a
The high Hertzian pressures in many tribological inter- series of PAO, PDMS, and PPMS lubricants. A comparison
faces can cause elastic flattening of asperities, thus allowing of film thickness and friction provides new insight into the
lubricant films to form and to separate surfaces. This is the friction performance of these lubricants.
regime of elastohydrodynamic lubrication (EHL). Gohar
and Cameron [6] first suggested the use of optical interfer-
ometry to study EHL film thicknesses, and Foord et al. [7] 2 Experimental Procedures
used an early optical interferometry system to measure the
film thicknesses of several lubricants including paraffinic, The fluids examined in this research were characterized in
naphthenic, and aromatic oils. Molimard et al. [8] examined terms of the molecular mass and structure, density, vis-
many new developments in the optical measurement of EHL cosity, film thickness, and friction coefficient. These results
films including light intensity and hue processing methods, allow cross-comparison of the performance based on sim-
while Spikes and Cann [9] described the improved resolu- ilarities and differences in the molecular structures.
tion attainable using a thin spacer layer and proper spec-
trometric analysis of interfered light.
Hamrock and Dowson [10] derived an equation for EHL 3 Molecular Mass, Structure, and Oxidative Stability
film thickness h using the atmospheric viscosity g0 and
pressure–viscosity index a of a lubricant and the entrain- Gel permeation chromatography (GPC) was used to obtain
ment speed U for a given test configuration. The Hamrock– the molecular mass distributions of the lubricant samples.
Dowson equation often overestimates the film thickness for Verification of the PAO, PDMS, and PPMS molecular

123
Tribol Lett (2012) 48:355–365 357

structures was obtained through a combination of manu- Sliding speed jU1  U2 j

facturer’s information and 1H/13C NMR spectroscopic data. R¼ ¼ ð2Þ
Entrainment speed ðU1 þ U2 Þ=2
The PDMS and PPMS branch contents were assayed on
Varian INOVA 400 or Mercury 400 NMR spectrometers. 3.3 Friction
Differential scanning calorimetry was used to determine the
thermal-oxidative stability or breakdown temperature of the Measurements of friction coefficients l were made on the
lubricant samples by performing measurements in air and same PCS instrument used to measure film thickness from
nitrogen at a gas flow rate of 60 mL/min and temperature 303 to 398 K with temperature controlled to ±1 K for each
ramp rate of 10 K/min. test in the temperature sequence. Friction tests were con-
ducted using a 19.050 mm diameter AISI 52100 steel balls
3.1 Density and Viscosity placed under a load of 20 N in Hertzian contact with a steel
disk. The Young’s moduli of the steel disk and steel ball
Density q and kinematic viscosity t were measured are both 210 GPa, giving a maximum Hertzian pressure of
simultaneously from 303 to 398 K using a Cannon 0.82 GPa. The surface roughness of the ball and disk was
CT-2000 constant temperature bath with microprocessor about 5 and 30 nm, respectively, yielding a composite
control. The density was determined by precision mea- surface roughness of approximately 30 nm. The disk
surements of the mass and volume of each sample. The velocity was varied from 0.025 to 5.00 m/s. The friction
kinematic viscosity was measured using Cannon–Fenske coefficient was measured at R = 0.50.
capillary viscometers. Multiple measurements were taken
to determine the average viscosity and its reproducibility.
The absolute viscosity was obtained from the kinematic 4 Results
viscosity and density.
4.1 Molecular Mass and Structure
3.2 Film Thickness
Table 1 lists the lubricant type and molecular mass for the
Elastohydrodynamic lubricant film thickness was measured 15 samples that were tested. The NMR data established
with a PCS thin-film tribometer from 303 to 398 K. The that all of the PDMS samples have 100 % methyl
system uses a polished AISI 52100 steel ball of 19.050 mm branching within the accuracy of the instrument, while the
diameter, which is pressed against an optically transparent phenylmethyl branch content of the PPMS samples ranges
glass disk with a 500-nm-thick silica spacer layer under a from 10 to 90 %. Two of the PPMS samples have 10 %
20 N load. The respective Young’s moduli of the glass disk phenylmethyl incorporation, but different degrees of
and steel ball are 75 and 210 GPa, giving a maximum polymerization, hence different viscosities. The siloxanes
Hertzian pressure of 0.54 GPa [15]. The rms roughness are labeled by the approximate viscosity (cSt) of the
values of the glass disk and steel ball are 5 and 14 nm, sample at 303 K. The PPMS samples have a prefix to
respectively, giving a composite roughness of about 15 nm. denote phenylmethyl content. For example, PPMS 10-500
This assembly is able to measure lubricant film thicknesses means a 10 % phenylmethyl branch content and viscosity
with a precision up to 1 nm for films under 30 nm, and near 500 mPa s. PAOs are alphabetized to avoid incon-
within 5 % for film thicknesses [30 nm [14, 17]. The sistent notation and potential confusion in adopting the
temperatures were held constant to ±1 K for each test in siloxane standard. NMR established the identities of the
the temperature sequence. Before each test, the lubricant PAO samples with details of their molecular masses being
reservoir, disk, ball, and carriage were thoroughly cleaned more accurately measured by GPC.
with solvent and allowed to dry. During the test, the ball
was partly immersed in the fluid sample to allow fluid
transfer to the ball–disk interface. 4.2 Density and Viscosity
Measurements of film thickness were undertaken in
nominally pure rolling conditions with the ball completely Density and viscosity of the present lubricants at 303, 348,
driven by the disk, with the disk velocity U1 varying from and 398 K are also shown in Table 1. The density increases
0.020 to 4.35 m/s. Additional measurements were made on from PAO to PDMS to PPMS, generally increasing with
several fluids to study shear thinning in which the ball was increasing molecular mass for lubricants with similar
attached to a separate motor-driven shaft to allow inde- molecular structures. The density of PDMS has an inter-
pendent variation of the ball velocity U2. This allowed us mediate density, reaching a room temperature value of
to vary the slide-to-roll ratio R from pure rolling (R = 0) to about 0.97 g/cm3 at the highest molecular mass [26]. The
pure sliding (R = 2) as defined in Eq. 2. density of PPMS increases with phenyl content [27].

123
358 Tribol Lett (2012) 48:355–365

Table 1 Molecular mass, density, viscosity, and activation energy for PAOs and polysiloxanes with varying branch structures and contents
Sample Molecular Density (g/cm3) Viscosity (mPa s) Activation
mass (g/mol) energy (kJ/mol)
303 K 348 K 398 K 303 K 348 K 398 K

PAO A 433 0.78 0.75 0.71 5.6 1.8 0.8 20

PAO B 757 0.80 0.78 0.74 23.1 5.8 2.2 25
PAO C 956 0.81 0.79 0.75 39.2 9.0 3.1 27
PAO D 1090 0.81 0.78 0.75 62.1 13.0 4.2 29
PAO E 1260 0.82 0.79 0.76 103.0 16.5 4.6 33
PDMS 10 1760 0.88 0.85 0.81 8.3 4.1 2.1 15
PDMS 20 3270 0.92 0.88 0.84 17.3 8.1 4.2 15
PDMS 50 6090 0.93 0.89 0.86 43.1 20.1 10.9 15
PDMS 100 9930 0.94 0.91 0.87 102.3 46.6 23.4 15
PDMS 300 19900 0.95 0.92 0.87 261.6 127.2 63.7 14
PDMS 1000 32000 0.96 0.92 0.88 937.0 423.3 213.2 15
PPMS 10-100 8180 0.97 0.93 0.89 91.8 40.3 19.8 16
PPMS 10-500 26600 0.99 0.94 0.90 443.3 202.4 98.0 16
PPMS 50-125 2690 1.05 1.01 0.98 125.7 38.5 15.8 23
PPMS 90-500 1990 1.07 1.04 1.00 472.1 68.7 21.8 33

Figure 2 plots the variation of viscosity as a function of

temperature for various lubricants. Note the expected
increase of viscosity with molecular mass for lubricants of
the same basic molecular structure. Because of the more
flexible Si–O–Si backbone, PDMS lubricants with the
same molecular mass as PAOs have lower viscosity. The
effect of the molecular structure on molecular mass and
consequently viscosity is most evident in PPMS, where
similar viscosities can be obtained for different molecular
masses by variation of the macromolecular length and
phenyl branch content. The increased viscosity with
increasing branch content can be attributed to the addi-
tional attractive intermolecular forces between the poly-
mers of the highly branched PPMS. Fig. 2 Variation of the PAO, PDMS, and PPMS viscosity as a
By means of the Andrade–Eyring equation (Eq. 3) to function of molecular mass at T = 303 K
describe the temperature dependence of viscosity, where
gR is the low shear viscosity at infinite temperature, E is the exhibits increasing temperature fragility with increasing
activation energy, R is the universal gas constant, and T is phenylmethyl content.
the temperature [28], we can calculate the activation
energy for various lubricants. Table 1 lists the activation 4.3 Oxidative Stability
energy calculated from the viscosity over the whole tem-
perature range. Table 2 compares the oxidation onset temperatures of the
g0 ð T Þ ¼ gR  e
E
RT ð3Þ representative lubricants, showing that PAO C begins to
undergo oxidation at *473 K, while PDMS lubricants
The general trend is that PDMS lubricants have lower begin to undergo oxidation at *573 K. The addition of
activation energy (E * 15 kJ/mole) than PAOs (E * phenyl branches to the siloxane polymers substantially
25 kJ/mole), meaning that the viscosity of siloxanes does increases the oxidation onset temperature. High phenyl
not decrease as much with increasing temperature as PAOs content PPMSs remain stable up to *649 K, a substantial
[26]. The temperature dependence of viscosity of PPMS increase over PAOs and other hydrocarbons. Therefore,
increases with phenyl content. For example, with 90 % increasing the phenyl content increases the oxidative sta-
phenyl content, E = 33 kJ/mole. While the PAO and PDMS bility of siloxanes, but also increases the viscosity sensi-
exhibit strong Arrhenius behavior, the PPMS sample tivity to temperature.

123
Tribol Lett (2012) 48:355–365 359

Table 2 Oxidation onset temperatures (OOT) for representative

PAO, PDMS, and PPMS samples
Sample OOT (K)

PAO C 473
PDMS 50 573
PPMS 10-100 597
PPMS 50-125 629
PPMS 90-500 649

4.4 Film Thickness

Figures 3, 4, 5 show that the measured lubricant film

thickness as a function of the entrainment speed at three
temperatures for representative PAO, PDMS, and PPMS
samples. Corresponding calculated film thicknesses from
the Hamrock–Dowson equation are also shown. The pres-
sure–viscosity index values of the lubricants required to
perform these calculations are derived from the literature
and are included in the ‘‘Appendix’’ section. For all lubri-
cants, because of the dependence of the viscosity and pres-
sure–viscosity index on the temperature, the lubricant film
thickness decreases markedly with increasing temperature.
For PAOs, the film thickness appears to be accurately
modeled by the Hamrock–Dowson equation and follows a
logarithmic slope of 0.67 versus entrainment speed, sug-
gesting Newtonian behavior. On the other hand, as shown
in Figs. 4 and 5, the high molecular mass siloxanes
(Mw = 9930, 32200 and 26600) attain film thicknesses that
are significantly less than those calculated from the Ham-
rock–Dowson equation with a logarithmic slope less than
0.67. This deviation may be attributed to temporary shear
thinning [12, 25, 29] in which molecular alignment induced
by shear stress creates a pathway of lower resistance for the
motion of these lubricant molecules, thus lowering the
effective viscosity [24]. Support for this model comes both
theoretically from the kinetic theory and experimentally
from flow birefringence studies [30]. Molecular dynamics
simulations provide further support for the role of molec- Fig. 3 Measured (symbols) and calculated (lines) film thickness
ular alignment in shear thinning [31]. versus entrainment speed for a PAO B, b PAO C, and c PAO E at
The literature states that the approximate molecular mass 303 K (triangles), 348 K (squares), and 398 K (diamonds)
for the onset of shear thinning in PDMS is *Mw =
16,000–28,000 g/mol, depending on the method of deduc- (Fig. 5). The higher molecular mass sample (PPMS
tion [32–34]. This is in the same range as observed here. 10-500: Mw = 26,600 g/mol) produces a thinner film than
Hadjistamov [35] noted that the onset of shear thinning the other samples (PPMS 90-500: Mw = 1,990 g/mol) of
occurs at a nearly constant shear stress for high molecular similar viscosity (Fig. 5). This illustrates that the film-
mass PDMS at atmospheric pressure. Yamada [36] observed forming ability of siloxanes with similar viscosities and
that PDMS and several other polymer fluids obey a universal branch types may exhibit divergent behavior based on the
shear-thinning curve under confined conditions. details of the lubricant molecular structure. Thus, by vari-
This non-Newtonian behavior depends on high molec- ation of molecular length and branch content, phenyl
ular mass rather than on high viscosity as is illustrated by branched siloxanes may be designed to exhibit Newtonian
the two PPMS samples with viscosities near 500 mPa s behavior or temporary shear thinning.

123
360 Tribol Lett (2012) 48:355–365

Fig. 4 Measured (symbols) and calculated (lines) film thickness Fig. 5 Measured (symbols) and calculated (lines) film thickness
versus entrainment speed for a PDMS 20, b PDMS 100, and c PDMS versus entrainment speed for a PPMS 10-100, b PPMS 10-500, and
1000 at 303 K (triangles), 348 K (squares), and 398 K (diamonds) c PPMS 90-500 at 303 K (triangles), 348 K (squares), and 398 K
(diamonds)

Traditionally, shear thinning has been viewed as unde-

sirable. However, modern requirements for higher effi- 4.5 Friction Coefficient
ciency lubricants have caused researchers to reconsider its
potential benefits [37, 38]. High viscosity Newtonian fluids Figure 6a shows the variation of PAO D lubricant film
produce thick films and good surface protection, but they thickness as a function of entrainment speed at 303, 348,
also contribute to energy inefficiency through viscous and 398 K, while Fig. 6b shows the variation of friction
dissipation at high speeds. Thus, a high viscosity fluid that coefficient as a function of entrainment speed at 303, 348,
shear thins in the full-film lubrication regime will increase and 398 K using the same lubricant at R = 0.5. By com-
energy efficiency [37, 39, 40]. bining these two datasets, we can plot friction coefficient

123
Tribol Lett (2012) 48:355–365 361

The transition from boundary to full-film lubrication is

clearly observed from Fig. 6c. For example, at 398 K, the
friction coefficient is *0.1 at zero film thickness, dropping
rapidly to *0.01 at *30-nm film thickness, which is
about the same order of magnitude as the composite
roughness of the ball–disk interface. At the same film
thickness within the full-time lubrication regime, operating
at higher temperature results in lower viscosity and hence a
lower friction coefficient as observed. The data trend for
other PAOs explored in the present study is similar.
Figure 7 shows the variation of friction coefficient as a
function of film thickness for PDMS lubricant at
T = 303 K and R = 0.5. The trend is similar to that shown
in Fig. 6c for PAO. For example, the transition from high
(boundary lubrication) to low friction coefficient (full-film
lubrication) occurs at around the composite roughness of
the ball–disk interface; the friction behavior in the full-film
regime is consistent with the bulk viscosity of the lubricant.
Figure 8 shows the variation of the friction coefficient as
a function of film thickness for PPMS lubricants at

Fig. 7 Friction coefficient versus film thickness for PDMS at

T = 303 K and R = 0.5

Fig. 6 PAO D: a variation of lubricant film thickness versus

entrainment speed; b variation of friction coefficient versus entrain-
ment speed at R = 0.5; c variation of friction coefficient versus film
thickness at R = 0.5 and 303 K (triangles), 348 K (squares), and
398 K (diamonds)

versus lubricant film thickness as shown in Fig. 6c. At

303 K, the friction coefficient shows a weak dependence
on the lubricant film thickness. This suggests that an
adsorbed layer is providing sufficient protection for
asperities. Such an adsorbed layer is weakly bound as
inferred from the substantially larger friction coefficient at
zero film thickness when the temperature increases to 348 Fig. 8 Friction coefficient versus film thickness for PPMS at
and 398 K. T = 303 K and R = 0.5

123
362 Tribol Lett (2012) 48:355–365

T = 303 K and R = 0.5. These results illustrate the critical fluids, and film thickness measurements [7, 15, 16, 23, 42],
role of shear thinning in controlling friction behavior in the as well as high pressure viscometric data [24, 43, 47–53].
full-film lubrication regime. In spite of PPMS 10-500 Calculations of pressure–viscosity indices can be a pre-
having higher bulk viscosity than PPMS 50-125, Fig. 8 carious procedure if the influences of fragility [53],
shows that it gives the lowest friction coefficient in the full dynamic crossover [53, 54], scaling effects [55], com-
lubrication regime. As illustrated earlier (Fig. 5b), this is pressibility, and non-Newtonian characteristics [16] are not
due to temporary shear thinning in which the applied shear taken into account.
stress causes alignment of PPMS 10-500 lubricant mole- Two pressure–viscosity indices commonly calculated
cules and reduces the effective viscosity. The research from high pressure data are the isothermal aOT and reci-
demonstrates that similar viscosities, but different film procal asymptotic a indices. The latter is calculated from
formation abilities and friction performances can be atmospheric viscosity as well as the integral of viscosity to
attained by rational variation of the molecular chain length infinite pressure, and generally correlates with the logarithm
and branch content. of viscosity. The reciprocal asymptotic pressure–viscosity is
used in these calculations because it represents a greater
portion of the pressure–viscosity curve and reduces the
5 Conclusions effects of anomalies that give isothermal pressure–viscosity
indices unrealistic values [20, 24, 50, 56].
This research was conducted to explore and compare Several researchers have successfully approximated
lubrication properties of siloxane-based lubricants with temperature-dependent pressure–viscosity indices using the
those of PAOs through an investigation of the relationships kinematic viscosity at the temperatures under analysis
among molecular structure, viscosity, film formation, and (Eq. 4) [57–59]. The technique provides accurate approx-
friction performance. The results of this study indicate that imations for several lubricants when logarithmic viscosity
a siloxane macromolecule can exhibit a diverse lubrication and temperature are linearly related and sufficient data in
performance based on chain length, branch content, and the proximity of the temperature are available [57–59].
branch structure. The major findings are as follows: Equation 4 is only used with reference data from high
pressure viscometers. In all calculations, the measured
(1) Siloxanes have much higher oxidative onset temper-
viscosity at the test temperatures was used to include the
atures than PAOs, which increase with the phenyl
effect of temperature fragility in the approximation of
content. However, high phenyl contents in siloxanes
pressure–viscosity indices.
also increase the sensitivity of both the viscosity and
pressure–viscosity index to temperature variations. a ¼ b þ a  logðmÞ ð4Þ
(2) PDMS and PPMS lubricants may exhibit structure- The Hamrock–Dowson equation can be used to approach
dependent non-Newtonian shear-thinning behavior the pressure–viscosity index of a test fluid at from that of a
which may be remediated or exacerbated by adjust- reference fluid ar by means of the ratio of the measured
ment of the polymer chain length or introduction of film thicknesses of the test fluid ht and the reference fluid hr
branches. [7, 15, 16]. A modification of the aforementioned method
(3) The siloxanes with the highest phenyl content (PPMS from Yokoyama and Spikes [42] was used to calculate the
50-125 and PPMS 90-500) generally retain Newto- effective pressure–viscosity indices of the PAOs in this
nian flow characteristics, resulting in thick films, as study (Eq. 5).
well as high EHD friction. The EHD friction  1:89  1:26
increases with increasing the phenyl content and ht gr
at ¼ ar  ð5Þ
reaches a maximum in PPMS 90-500. hr gt
The effective pressure–viscosity index calculated from
Acknowledgments The authors thank the Dow Corning Corpora- film thickness can be subjected to errors when scaling
tion for support of this research.
effects, compressibility, and non-Newtonian shear-thinning
behavior occur. In this study, high pressure viscometric
data from Bair and Kottke [60] for PAO B were used in
Appendix: Pressure–Viscosity Indices Eq. 4 to calculate the pressure–viscosity indices at different
temperatures. Then, the pressure–viscosity indices for other
A variety of methods have been proposed to approximate PAOs were calculated at the same temperatures as the
the pressure–viscosity indices of different lubricants reference data by means of Eq. 5. The molecular masses of
[41–44]. Pressure–viscosity indices have been calculated the PAOs were within the range of Newtonian flow
from bulk materials properties [41, 45, 46], reference outlined by Liu et al. [20], so the method of Yokoyama and

123
Tribol Lett (2012) 48:355–365 363

Table 3 Pressure–viscosity coefficient for PAO, PDMS, and PPMS PDMS 1000 (a* = 15 GPa-1 at T = 303 K). The effect of
at 303, 348, and 398 K temperature on the pressure–viscosity index is slightly
Sample Pressure–viscosity index (GPa-1) lower than that of the reference sample to reflect the more
subtle decrease observed in the bulk of the PDMS data.
303 K 348 K 398 K
The reciprocal asymptotic pressure–viscosity coeffi-
PAO A 12.7 10.8 9.4 cients of the 10 % branched PPMS were calculated from
PAO B 15.7 12.6 10.3 the viscometric measurements of Kuss [51]. For the higher
PAO C 17.0 13.9 11.7 mass, shear-thinning sample, invariance of pressure–vis-
PAO D 18.0 14.9 12.6 cosity with degree of polymerization was applied as the
PAO E 19.3 15.6 13.4 best approximation from the data at hand. Aderin et al. [16]
PDMS 10 14.1 14.0 13.9 and Johnston et al. [23] used film thickness to calculate the
PDMS 20 14.1 14.0 13.9 effective pressure–viscosity index for PDMS and PPMS
PDMS 50 14.1 14.0 13.9 and noted that it was lower than that measured by Kuss
PDMS 100 14.1 14.0 13.9 [51]. The discrepancy is attributed to shear thinning.
PDMS 300 15.0 14.8 14.6 The pressure–viscosity indices calculated for PPMS
PDMS 1000 15.0 14.8 14.6 50-125 were obtained from the viscometric measurements
PPMS 10-100 15.5 14.1 13.0 of Jakobsen et al. [50] and supported by the measurements of
PPMS 10-500 15.5 14.1 13.0 Kuss [51], Bridgman [47], and the ASME pressure–viscos-
PPMS 50-125 23.7 17.8 12.8 ity report [48]. The PPMS 90-500 sample was also based on
PPMS 90-500 27.0 20.2 14.5 data from Jakobsen et al. [50]. The authors used the mea-
sured viscosity of the high phenyl content siloxanes at each
test temperature to include the effect of temperature fragility
Spikes [42] is applicable (Eq. 5). Table 3 lists the in the interpolations of the pressure–viscosity indices.
pressure–viscosity coefficients for the lubricants at the High phenyl contents significantly increase the pres-
temperatures studied. sure–viscosity index, but leave it susceptible to greater
Bridgman [47], Winer [49], ASME [48], Jakobsen et al. temperature-dependent viscosity and pressure–viscosity
[50], Kuss [51], King et al. [61], Bair and Qureshi [24], and variations. While the pressure–viscosity coefficients of
Liu et al. [20] have reported pressure–viscosity indices for PAO and PDMS were relatively stable over the tempera-
PDMS and PPMS that were measured on high pressure ture range investigated, that of PPMS fall nearly 50 % over
viscometers. The reciprocal asymptotic pressure–viscosity the temperature range from 303 to 398 K (Table 3). This is
indices from the data of Jakobsen et al. [50] and Winer [49] in line with the broad correlation between activation energy
were used to interpolate the pressure–viscosity indices for and pressure–viscosity index observed by Roelands et al.
PDMS (fluids 12 and 15) and PPMS (fluids 13 and 14), [63], Jakobsen et al. [50], Spikes [64], Aderin et al. [16],
respectively. Table 3 lists the siloxane pressure–viscosity and Gunsel et al. [15].
data at three temperatures (303, 348, and 398 K) that are
interpolated from four temperature points (297, 311, 372,
and 422 K) by means of Eq. 4.
Jakobsen et al. [50], Winer [49], and Kuss [51] corre-
References
lated the pressure–viscosity characteristics of several
siloxanes with the nature of the side radical (branch) and 1. Sharma, S., Snyder, C., Gschwender, L.: Tribological behavior of
found them to be independent of the degree of polymeri- some candidate advanced space lubricants. Tribol. Trans. 36(2),
zation. This observation is used to extrapolate the pressure– 321–325 (1993)
viscosity index to other siloxanes of the same species and 2. Ziegler, M., Fearon, F.: Silicon-based polymer science: a com-
prehensive resource. In: Advances in Chemistry Series, vol. 224.
similar molecular masses. American Chemical Society, Washington, DC (1990)
Two samples of PDMS from Jakobsen et al. [50] were 3. De Jaeger, R., Gleria, M.: Inorganic Polymers. Nova Science
used for the low and high mass PDMS pressure–viscosity Publishers Inc., New York (2007)
index. The low mass PDMS reference sample was inter- 4. Lewis, G.: The atom and the molecule. J. Am. Chem. Soc. 38(4),
762–785 (1916)
mediate to our low mass group and nearly identical to one 5. Brandup, J., Immergut, E., Grulke, E.: Polymer Handbook. In-
sample. The high mass pressure–viscosity data were taken terscience, New York (1975)
from a sample that is intermediate with respect to our high 6. Gohar, R., Cameron, A.: The mapping of elastohydrodynamic
mass group. The pressure–viscosity indices of the high contacts. ASLE Trans. 10(3), 215–225 (1967)
7. Foord, C., Hammann, W., Cameron, A.: Evaluation of lubricants
mass sample were updated to reflect the data more recently using optical elastohydrodynamics. Tribol. Trans. 11(1), 31–43
reported by Bair and Qureshi [62] for a sample similar to (1968)

123
364 Tribol Lett (2012) 48:355–365

8. Molimard, J., Querry, M., Vergne, P.: New tools for the experi- 32. Orrah, D., Semlyen, J., Ross-Murphy, S.: Studies of cyclic and
mental study of EDH and limit lubrications. Tribol. Ser. 36, linear poly(dimethylsiloxanes): 27. Bulk viscosities above the
717–726 (1999) critical molar mass for entanglement. Polymer 29(8), 1452–1454
9. Spikes, H., Cann, P.: The development and application of the (1988)
spacer layer imaging method for measuring lubricant film 33. Kissi, N., Piau, J., Attané, P., Turrel, G.: Shear rheometry of
thickness. Proc. Inst. Mech. Eng. J 215(3), 261–277 (2001) polydimethylsiloxanes. Master curves and testing of Gleissle and
10. Hamrock, B., Dowson, D.: Ball bearing lubrication: the elas- Yamamoto relations. Rheol. Acta 32(3), 293–310 (1993)
tohydrodynamics of elliptical contacts. National Aeronautics 34. Verdier, C., Longin, P., Piau, M.: Dynamic shear and compres-
Space Administration, Lewis Research Center, Cleveland (1981) sional behavior of polydimethylsiloxanes: ultrasonic and low
11. Bair, S., Winer, W.: A simple formula for EHD film thickness of frequency characterization. Rheol. Acta 37(3), 234–244 (1998)
non-Newtonian liquids. Tribol. Ser. 32, 235–241 (1997) 35. Hadjistamov, D.: Determination of the onset of shear thinning of
12. Chapkov, A., Bair, S., Cann, P., Lubrecht, A.: Film thickness in polydimethylsiloxane. J. Appl. Polym. Sci. 108(4), 2356–2364
point contacts under generalized Newtonian EHL conditions: (2008)
numerical and experimental analysis. Tribol. Int. 40, 1474–1478 36. Yamada, S.: General shear-thinning dynamics of confined fluids.
(2007) Tribol. Lett. 13(3), 167–171 (2002)
13. Guangteng, G., Spikes, H.: Boundary film formation by lubricant 37. Coy, R.: Practical applications of lubrication models in engines.
base fluids. Tribol. Trans. 39(2), 448–454 (1996) Tribol. Int. 31(10), 563–571 (1998)
14. Smeeth, M., Spikes, H., Gunsel, S.: The formation of viscous 38. Bronshteyn, L.A., Kreiner, J.H.: Energy efficiency of industrial
surface films by polymer solutions: boundary or elastohydrody- oils. Tribol. Trans. 42(4), 771–776 (1999)
namic lubrication? Tribol. Trans. 39(3), 720–725 (1996) 39. Martini, A., Zhu, D., Wang, Q.: Friction reduction in mixed
15. Gunsel, S., Korcek, S., Smeeth, M., Spikes, H.: The elastohy- lubrication. Tribol. Lett. 28(2), 139–147 (2007)
drodynamic friction and film forming properties of lubricant base 40. Akbarzadeh, S., Khonsari, M.: Performance of spur gears con-
oils. Tribol. Trans. 42(3), 559–569 (1999) sidering surface roughness and shear thinning lubricant. J. Tribol.
16. Aderin, M., Johnston, G., Spikes, H., Caporiccio, G.: The elas- 130, 021503 (2008)
tohydrodynamic properties of some advanced non hydrocarbon- 41. Wu, C., Klaus, E., Duda, J.: Development of a method for the
based lubricants. Lubr. Eng. 48(8), 633–638 (1992) prediction of pressure–viscosity coefficients of lubricating oils
17. Guangteng, G., Spikes, H.: The control of friction by molecular based on free-volume theory. J. Tribol. 111, 121–128 (1989)
fractionation of base fluid mixtures at metal surfaces. Tribol. 42. Yokoyama, F., Spikes, H.: Film-forming properties of polyol
Trans. 40(3), 461–469 (1997) esters, polyphenyl ethers and their mixtures over a wide range of
18. Bair, S.: Elastohydrodynamic film forming with shear thinning temperature. Tribol. Trans. 43(1), 130–136 (2000)
liquids. J. Tribol. 120, 173–178 (1998) 43. Bair, S., Qureshi, F.: Accurate measurements of pressure–vis-
19. Bair, S., Vergne, P., Querry, M.: A unified shear-thinning treat- cosity behavior in lubricants. Tribol. Trans. 45(3), 390–396
ment of both film thickness and traction in EHD. Tribol. Lett. (2002)
18(2), 145–152 (2005) 44. Bair, S., Liu, Y., Wang, Q.: The pressure–viscosity coefficient for
20. Liu, Y., Wang, Q., Krupka, I., Hartl, M., Bair, S.: The shear- Newtonian EHL film thickness with general Piezoviscous
thinning elastohydrodynamic film thickness of a two-component response. J. Tribol. 128, 624–631 (2006)
mixture. J. Tribol. 130, 021502 (2008) 45. Johnston, W.: A method to calculate the pressure–viscosity
21. Biresaw, G., Bantchev, G.: Effect of chemical structure on film- coefficient from bulk properties of lubricants. Tribol. Trans.
forming properties of seed oils. J. Synth. Lubr. 25(4), 159–183 24(2), 232–238 (1981)
(2008) 46. So, B., Klaus, E.: Viscosity–pressure correlation of liquids. Tri-
22. Dyson, A., Wilson, A.: Film thicknesses in elastohydrodynamic bol. Trans. 23(4), 409–421 (1980)
lubrication by silicone fluids. Prof Eng Publishing, Suffolk (1965) 47. Bridgman, P.: Viscosities to 30,000 kg/Cm. Proc. Am. Acad. Arts
23. Johnston, G., Wayte, R., Spikes, H.: The measurement and study Sci. 77(4), 117–128 (1949)
of very thin lubricant films in concentrated contacts. Tribol. 48. ASME: Viscosity and Density of Over 40 Lubricating Fluids of
Trans. 34(2), 187–194 (1991) Known Composition at Pressures to 150,000 Psi and Tempera-
24. Bair, S., Qureshi, F.: The generalized Newtonian fluid model and tures to 425 F. American Society of Mechanical Engineers, New
elastohydrodynamic film thickness. J. Tribol. 125(1), 70–75 (2003) York (1953)
25. Bair, S.: A rough shear-thinning correction for EHD film thick- 49. Winer, W.O.: A Study of the Elastohydrodynamic Lubrication
ness. Tribol. Trans. 47(3), 361–365 (2004) and High Pressure Rheological Behavior of a Series of Silicone
26. Kataoka, T., Ueda, S.: Viscosity–molecular weight relationship Fluids. Georgia Institute of Technology, Atlanta (1972)
for polydimethylsiloxane. J. Polym. Sci. C 4(5), 317–322 (1966) 50. Jakobsen, J., Sanborn, D., Winer, W.: Pressure viscosity char-
27. Fytas, G., Dorfmueller, T., Lin, Y., Chu, B.: Photon correlation acteristics of a series of siloxane fluids. ASME Trans. F 96,
spectroscopy of bulk siloxane polymers. Macromolecules 14(4), 410–417 (1974)
1088–1091 (1981) 51. Kuss, E.: Viskositats-Druckverhalten von flussigen Methyl-
28. Van Krevelen, D., Hoftyzer, P.: Newtonian shear viscosity of Phenylmethyl- und Cyclohexylmethyl-Siloxeanen. Erdol und
polymeric melts. Die Angew. Makromol. Chem. 52(1), 101–109 Kohle-Erdgas-Petrochemie vereinigt mit Brennstoff-Chemie
(1976) 27(8), 416–422 (1974)
29. Greenwood, J., Kauzlarich, J.: Elastohydrodynamic film thickness 52. Höglund, E.: Influence of lubricant properties on elastohydrody-
for shear-thinning lubricants. Proc. Inst. Mech. Eng. J 212(3), namic lubrication. Wear 232(2), 176–184 (1999)
179–191 (1998) 53. Bair, S., Roland, C., Casalini, R.: Fragility and the dynamic
30. Hutton, J.F.: Interdisciplinary Approach to Liquid Lubricant crossover in lubricants. Proc. Inst. Mech. Eng. J 221(7), 801–811
Technology. NASA Technical Document SP-318187 (1972) (2007)
31. Cui, S., Gupta, S., Cummings, P., Cochran, H.: Molecular 54. Bair, S., Laesecke, A.: Normalized Ashurst–Hoover scaling and a
dynamics simulations of the rheology of normal decane, hexa- comprehensive viscosity correlation for compressed liquids.
decane, and tetracosane. J. Chem. Phys. 105(3), 1214–1220 (1996) J. Tribol. 134, 021801 (2012)

123
Tribol Lett (2012) 48:355–365 365

55. Krupka, I., Bair, S., Kumar, P., Khonsari, M., Hartl, M.: An 60. Bair, S., Kottke, P.: Pressure–viscosity relationships for elas-
experimental validation of the recently discovered scale effect in tohydrodynamics. Tribol. Trans. 46(3), 289–295 (2003)
generalized Newtonian EHL. Tribol. Lett. 33(2), 127–135 (2009) 61. King, H., Herbolzheimer, E., Cook, R.L.: The diamond-anvil cell
56. Kumar, P., Bair, S., Krupka, I., Hartl, M.: Newtonian quantitative as a high-pressure viscometer. J. Appl. Phys. 71(5), 2071–2081
elastohydrodynamic film thickness with linear Piezoviscosity. (1992)
Tribol. Int. 43(11), 2159–2165 (2010) 62. Bair, S., Qureshi, F.: The high pressure rheology of polymer–oil
57. Schrader, R.: Zur Schmierfilmbildung von Schmierolen und solutions. Tribol. Int. 36(8), 637–645 (2003)
Schmierfetten im elastohydrodynamischen Walzkontakt, Disser- 63. Roelands, C., Vlugter, J., Watermann, H.: The viscosity tem-
tation, Hannover (1988) perature pressure relationship of lubricating oils and its correla-
58. Gold, P., Schmidt, A., Dicke, H., Loos, J., Assmann, C.: Vis- tion with chemical constitution. ASME J. Basic Eng. 85, 601–606
cosity–pressure–temperature behaviour of mineral and synthetic (1963)
oils. J. Synth. Lubr. 18(1), 51–79 (2001) 64. Spikes, H.: A thermodynamic approach to viscosity. Tribol.
59. Ohno, N., Mia, S., Morita, S., Obara, S.: Friction and wear Trans. 33(1), 140–148 (1990)
characteristics of advanced space lubricants. Tribol. Trans. 53(2),
249–255 (2010)

123