Tribology International 57 (2013) 195–201

Contents lists available at SciVerse ScienceDirect

Tribology International
journal homepage: www.elsevier.com/locate/triboint

A mixed lubrication model of a rotary lip seal using flow factors

Fei Guo a, Xiaohong Jia a, Shuangfu Suo a, Richard F. Salant b,n, Yuming Wang a
a
State Key Laboratory of Tribology, Tsinghua University, Beijing, China
b
Georgia Institute of Technology, Atlanta, GA, USA

a r t i c l e i n f o abstract

Article history: A mixed lubrication model of a rotary lip seal using flow factors has been developed. The model consists
Received 28 June 2012 of coupled fluid mechanics, contact mechanics and deformation mechanics analyses. The fluid
Received in revised form mechanics analysis is described by a Reynolds equation that takes into account the surface roughness
15 August 2012
effect using flow factors. The contact mechanics analysis uses the Greenwood and Williamson model to
Accepted 18 August 2012
compute contact pressure. The deformation mechanics analysis utilizes the influence coefficient
Available online 28 August 2012
approach to compute deformation of the seal. Results for a typical seal show how the operation
Keywords: parameters and the surface roughness affect seal behavior.
Rotary lip seal & 2012 Elsevier Ltd. All rights reserved.
Mixed lubrication
Flow factors

1. Introduction film pressure. However, both the macrodeformation of the lip and
microdeformation of the asperities were not considered.
In a successful lip seal, the existence of a thin lubricant film Salant and Flaherty [10] developed an elastohydrodynamic
between the lip and the shaft was first verified by Jagger [1]. This analysis of a rotary lip seal under full film lubrication conditions
film has been identified as the result of hydrodynamic generation by considering microundulations, incorporating the hydrody-
of elevated pressures within the film [2–4]. Gabelli and Poll [5], namic analysis of the flow in the lubricant film, coupled with
through experimental and theoretical methods, considered the the elastic deformation of the lip material. Subsequently, Shi and
hydrodynamic effects produced by asperities on the lip surface, Salant [11] developed a mixed elastohydrodynamic lubrication
which play a dominant role in the sealing mechanism. Qian [6], model through considering the effect of contact pressure between
Kammüller [7] and Müller [8] have shown that shear stress in the asperities and shaft. This effect is necessary to be included, for it
lubricant film will deform the lip surface in the circumferential will change the microstructure of the lip surface, which influences
direction, which causes the asperities to act like the vanes of a the reverse pumping action.
centrifugal pump, tending to pump the liquid from the air side of In the above models, the simulation of asperities on the lip
the lip seal back to the liquid side of the lip seal, thereby surface is deterministic. These models can provide detailed
counteracting the natural leakage from the liquid side toward variations and localized maximum and/or minimum values of
the air side. all the physical parameters that govern lip seal behavior. How-
In order to construct an approximate model to represent the ever, these deterministic models are not very practical as design
complex processes occurring in a lip seal, several significant tools, as they are very complicated and time-consuming. A
numerical approaches have been developed. The first numerical statistical approach can greatly reduce computation time by using
studies were conducted by Gabelli [9] and Gabelli and Poll [5]. the flow factor method [12,13], since it is not necessary to solve
They modeled the lubricant film formation between two parallel the Reynolds equation around each asperity. The first significant
rough surfaces, using the Reynolds equation and finite difference numerical studies in a rotary lip seal using flow factors were
techniques. In their model, they took into account the effect of performed by Salant and Rocke [14] and Rocke and Salant [15].
asperities, cavitation by using JFO boundary conditions, and the They developed an elastohydrodynamic analysis of a seal under
visco-elastic bulk effects of the elastomeric lip on the lubricant full film lubrication conditions. Their results indicate that the flow
factor method is a very practical approach. Most important, the
method is much simpler and faster than computations based on a
n
Corresponding author. Tel.: þ1 404 894 3176. deterministic approach, in the order of 30 times faster. However,
E-mail addresses: [email protected] (F. Guo),
[email protected] (X. Jia), [email protected] (S. Suo),
a major shortcoming of their statistical model is not considering
[email protected] (R.F. Salant), the effect of asperity contact, which will inevitably occur when
[email protected] (Y. Wang). the ratio of film thickness to surface roughness is less than 3.

0301-679X/$ - see front matter & 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.triboint.2012.08.008
196 F. Guo et al. / Tribology International 57 (2013) 195–201

Nomenclature Q^ y dimensionless pumping rate in y direction

R asperity radius
Ae total contact area x^ dimensionless circumferential coordinate, x=Lx
An nominal contact area ys the distance between the means of asperity and
Di shaft diameter surface heights
E Young’s modulus y^ dimensionless axial coordinate, y=Ly
f friction coefficient g Peklenik number, lx =ly
f(d) probability density function of surface heights dn normal deformation of lip surface
F cavitation index ds circumferential deformation of lip surface
F tol the total contact load Z asperity density
h nominal film thickness, separation between lx autocorrelation length in x direction
surface means ly autocorrelation length in y direction
h^ dimensionless film thickness, h=s m viscosity
h^ T dimensionless local film thickness, truncated if in n Poisson’s ratio
contact, hT =s x dimensionless number, mULx =pref s2
^
hT average local film thickness, equals h^ if not in contact rf density of the full film (uncavitated)
hs static undeformed film thickness r^ dimensionless lubricant density, r=rf
In influence coefficient for normal deformation s the standard deviation of surface heights
Is influence coefficient for normal deformation ss the standard deviation of asperity heights
K aspect ratio of solution space, Lx =Ly t^ avg average dimensionless viscous shear stress in the x
Lx length of solution domain in x direction direction, tavg/E
Ly length of solution domain in y direction tf shear stress due to contacting asperities
pa ambient pressure fc:c dimensionless density flow factor
pc contact pressure ff ,ff ss ,ff pp dimensionless shear stress factors
pcav cavitation pressure fs:c:x ,fs:c:y dimensionless shear flow factors
pref characteristic reference pressure fxx ,fxy dimensionless pressure flow factors
ps sealed pressure fyx ,fyy dimensionless pressure flow factors
psc static contact pressure F variable representing pressure/average density,
p^ f dimensionless fluid pressure, pf =pref defined by Eq. (1)
q^ y dimensionless pumping rate in y direction per unit o shaft rotating speed
length in the x direction

In this study, a mixed elastohydrodynamic lubrication model rough, elastomeric and stationary. Since the film thickness is
of a rotary lip seal is presented by using the flow factor method, in much smaller than the seal radius, a Cartesian coordinate system
which the hydrodynamic lubrication analysis, the asperity contact is used, and the coordinate system is fixed to the shaft in order to
analysis, and the deformation analysis are included. The goal of make the problem steady. The x direction represents the circum-
this study is to predict such seal characteristics as minimum film ferential direction, and the y direction represents the axial
thickness, contact area, reverse pumping rate and torque. While direction. In order to calculate the reverse pumping rate, it is
the rotary seal behavior can be affected by the surface roughness necessary to make an assumption that the air side of the seal is
of the shaft [16,17], vibration, dynamic runout and viscoelasticity flooded with liquid. Another assumption for the mixed lubrication
[18], such effects are beyond the scope of the present study. model is that microdeformation of the asperities on the lip
surface does not affect macrodeformation of the lip, which is
axisymmetric.
2. Numerical model and analysis
2.1. Fluid mechanics analysis
Fig. 1 shows a schematic diagram of the sealing zone, at a
given circumferential location. The shaft is assumed to be per- The fluid mechanics of the lubricant film is governed by
fectly smooth, rigid and rotating, while the seal lip is treated as Reynolds equation. Since cavitation occurs when the film pressure
falls below the cavitation pressure of dissolved gases in the
lubricant, the cavitation index F and the universal variable F are
used to account for such effect. The form of the universal, average
Reynolds equation with flow factors is
.…
.….
.….
.….
..
lip
..
 
@ ^3 @ðF FÞ @ðF FÞ
  
@ ^3 @ðF FÞ @ðF FÞ

…
…Ps..
.…… .. @x^
h fxx
@x^
þ K fxy
@y^
þK
@y^
h fyx
@x^
þ K fyy
@y^
b .……
.……
……..
……..  
Pa ..
……………… ……..
.. asperity ¼ 6x
@ n ^
o
½1 þð1FÞFðhT F fc:c Þ þ 6xF
@fs:c:x
þK
@fs:c:y
……………………….
……………………….. @x^ @x^ @y^
ð1Þ
………………………..
x
y In the liquid region,
shaft
FZ0 F ¼ 1
film pf pcav r ð2Þ
p^ f ¼ ¼ FF r^ ¼ ¼1
Fig. 1. Schematic of sealing zone. pref pcav rf
 F. Guo et al. / Tribology International 57 (2013) 195–201 197

In the cavitation region, are given by

Z 1
Fo0 F ¼ 0
Ae ¼ pZAn R GðzÞðzdÞdz ð9Þ
pf pcav r d
p^ f ¼ ¼0 r^ ¼ ¼ 1 þ ð1FÞF ð3Þ
pref pcav rf Z 1
F tol 4
pc ¼ ¼ ZR1=2 E0 GðzÞðzdÞ3=2 dz ð10Þ
The boundary conditions are An 3 d
F ¼ p^ f ¼ 0:1 at y^ ¼ 0 (air side) and y^ ¼ 1 (liquid side). Cyclic where Ftol represents the total contact load, and Z is the asperity
boundary condition on F at x^ ¼ 0,1.
density over the nominal contact area An. E0 represents the com-
The average truncated film thickness is given by
bined Young’s modulus, which is computed from
Z 1
^ ^ dÞf ðdÞdd
hT ¼ ðhþ ð4Þ 1 1u21 1u22
h^ 0 ¼ þ ð11Þ
E E1 E2
where f(d) is the probability density function of the surface
Since the shaft is treated as rigid, E0 is just the ‘plane-stress
roughness. For a Gaussian distribution, as a first approximation,
modulus’ for the elastomeric lip,
the form is
! E
h^ h^ h^ 1 ^2
E0 ¼ ð12Þ
^ 1v2
hT ¼ þ erf pffiffiffi þ pffiffiffiffiffiffi eðh =2Þ ð5Þ
2 2 2 2p E, v are Young’s modulus and Poisson’s ratio of the lip, respec-
In Eq. (1), the pressure flow factors fxx , fxy , fyx and fyy reflect tively. z is the asperity height, and d is the separation of the
the impedance of flow due to roughness when the flow is driven contact surfaces, which is obtained from
by a pressure gradient. The shear flow factors fsUcUx and fsUcUy d ¼ hys ð13Þ
represent additional flow transport due to the sliding of the rough
ys represents the distance between the means of the asperity
surfaces. The density flow factor fcUc accounts for the reduction in
and surface heights. As a first approximation, for a Gaussian dis-
effective density of the lubricant film due to the existence of air in
tribution,
the cavitation region. These seven flow factors are functions of the
ratio of the film thickness to the surface roughness h=s, the aspect 1 2
=2s2s
GðzÞ ¼ pffiffiffiffiffiffi ez ð14Þ
ratio of the asperities g, and the asperity orientation angle y. 2pss
Lo [19] developed expressions for the above flow factors for an ss represents the standard deviation of asperity heights, which is
arbitrary y by using the flow factors for y ¼01and 901, which are different from the standard deviation of surface heights s. The
obtained from Harp and Salant [20]. relation between s and ss is
A finite volume scheme is used to discretize Eq. (1) over the
solution domain to obtain a system of linear algebraic equations, 3:717  104
s2 ¼ s2s þ ð15Þ
which is solved for F by using the alternating direction, implicit Z2 R2
method (ADI). The ADI method alternates between the tri-
Due to the existence of contact pressure, there is also a
diagonal matrix algorithm (TDMA) in the axial direction and the
frictional shear stress, which is computed from the following
cyclic tri-diagonal matrix algorithm (CTDMA) in the circumfer-
equation for o 40
ential direction. A 136  136 mesh is used, validated by a mesh
refinement study. tf ¼ f pc ð16Þ
Once F is obtained, the dimensionless volumetric pumping
rate can be found from
Z 1 2.3. Deformation mechanics analysis
Q^ y ¼ q^ y dx^ ð6Þ
0
The deformation mechanics analysis is used to compute the
where deformed shape of the seal and the asperities, including the
 
3 @ðF FÞ @ðF FÞ aspect ratio of the asperities g and the asperity orientation angle
q^ y ¼ h^ fyx þK fyy þ6xfs:c:y ð7Þ
@x^ @y^ y as in [15], which allows one to determine the fluid film
thickness distribution. To obtain the deformation of the seal, an
The viscous shear stress on the lip can be computed from
influence coefficient method has been used. The method’s validity
tavg mU h @ðF FÞpref is based on two assumptions. First, during the entire computation
t^ avg ¼ ¼ ðff þ ff ss Þ þ ff pp ð8Þ
E hE 2E @x process, the seal’s stiffness is regarded as invariable. Second,
ff , ff ss and ff pp represent the shear stress factors, which are according to the small deformation theory, it is assumed that in
obtained from [15]. the sealing zone, the deformation at any location is proportional
to the forces applied at every location. So, the normal deformation
of the lip surface at the ith axial node can be expressed as
2.2. Contact mechanics analysis
X
m
ðdn Þi ¼ ðIn Þik ½pf þ pc psc k ð17Þ
As mentioned above, when the ratio of film thickness to
k¼1
surface roughness is less than 3, asperity contact will take place
and the contact pressure effect is significant and must be and the circumferential deformation at the ith axial node is given by
considered. This contact pressure is computed by using the X
m

Greenwood and Williamson surface contact model (the GW ðds Þi ¼ ðIs Þik ðtavg þ tf Þ ð18Þ
k¼1
model) [21].
In the GW model, all the contacting asperities are treated as a The influence coefficient matrices In and Is, and the static
collection of spherical caps having the same radius of curvature R, contact pressure psc are obtained from an off-line finite element
with a certain statistical distribution of their heights. The contact structural analysis, utilizing a commercial package, ANSYS. To
pressure pc and the total contact area Ae, according to the model obtain (In)ik, a unit normal load is applied to the lip surface at the
198 F. Guo et al. / Tribology International 57 (2013) 195–201

Initial input design and operating parameters

Initial input influence cofficient matrices and static contact pressure

Initial guess for film pressure distribution,film thickness and circumferential deformation

Calculate all flow factors

Fluid mechanics analysis

(Calculate film pressure by solving Reynolds equation)

Contact mechanics analysis

(Calculate contact pressure by using GW model) Relax film thickness

Defoumation mechanics analysis

(Calculate normal and circumferential deformations)

No
Film thickness
convergence?

Yes
Auxilliary calculations and output results

Fig. 2. Computational procedure.

kth axial node, and the normal displacement at the ith axial node
0.6
is calculated. This procedure is repeated for m times for all values
of i and k, to generate an m  m matrix In. A similar procedure is 0.5
used to obtain Is, only the load applied to the lip surface is a unit
Pressure,MPa

shear force. The analysis treats the seal as linearly elastic, as 0.4
nonlinear effects are negligible except at very high temperatures
0.3 static contact pressure Psc
and pressures [22]. fluid pressure Pf
contact pressure Pc
0.2 Pf + Pc
2.4. Computational scheme
0.1
In the solution scheme for the mixed lubrication model, the
fluid mechanics, contact mechanics, and deformation mechanics 0.0
0.0 0.2 0.4 0.6 0.8 1.0
are strongly coupled. This coupling is handled through an itera- y/Ly
tive computational procedure, illustrated in the flow diagram of
Fig. 2. It should be noted that although only a single iteration loop Fig. 3. Pressure distributions.

is shown explicitly in the figure, in fact two nested iteration loops

are used. The inner loop (not shown) is required for the solution of the latter two pressure distributions is similar to the distribu-
of Reynolds equation, for the film pressure distribution. In the tion of the static contact pressure. It is seen that the axial location
outer loop (shown), the film thickness is adjusted and the fluid of maximum pressure is closer to the liquid side of the seal than
mechanics, contact mechanics and deformations are recomputed. to the air side, which occurs in all successful lip seals. Thus, there
After all the computations are converged, auxiliary calculations is a net pumping of liquid from the air side toward the liquid side,
are performed for such characteristics as reverse pumping rate which prevents seal leakage. When the speed is changed, the
and friction torque. distribution of fluid and contact pressure will also change.
Fig. 4 shows the viscosity as a function of the shaft speed,
which was obtained from measurements of film temperature and
3. Results and discussion an empirical viscosity–temperature relationship [23]. As the
speed is increased, the temperature increases while the viscosity
The mixed lubrication model is used to evaluate the performance decreases.
of a representative rotary lip seal. The following parameter values Fig. 5 shows a plot of torque versus speed, holding all the base
serve as a base for the present study: E¼6.2  106 Pa, v¼0.499, parameters constant except speed. From Eq. (8), it is seen that the
ps ¼pa ¼1.02  105 Pa, pcav ¼0 Pa, Di ¼80 mm, o ¼1000 rpm, s ¼ viscous torque is primarily proportional to the product of viscos-
1 mm, lx ¼ 1.667 mm, ly ¼ 5.0 mm, R¼4 mm, f¼0.15, hs ¼1.4 mm. The ity times speed. At low speed, the influence of speed on the torque
contact width Ly is obtained from finite element analysis by using is dominant, so the torque increases with speed. However, as the
ANSYS. In addition, Lx is set equal to Ly. speed increases, the viscosity decreases, so the torque reaches a
Fig. 3 shows the static contact pressure, fluid pressure and maximum at approximately 3000 rpm and then decreases
contact pressure for the base case. As would be expected, the sum slightly. Above 5000 rpm, the torque is substantially constant.
 F. Guo et al. / Tribology International 57 (2013) 195–201 199

0.07
0.12

0.06 0.10

0.05 0.08

Contact Area Ratio

Viscosity,Pa.s

0.04 0.06

0.04
0.03
0.02
0.02
0.00
0.01
0 1000 2000 3000 4000 5000 6000 -0.02
0 1000 2000 3000 4000 5000 6000
Speed,rpm
Speed,rpm

Fig. 4. Viscosity versus speed. Fig. 7. Contact area ratio versus speed.

0.65
50
0.60
40
0.55

Power Loss,W
Torque,N/m

0.50 30

0.45
20
0.40

0.35 10

0.30
0
0 1000 2000 3000 4000 5000 6000 0 1000 2000 3000 4000 5000 6000

Speed,rpm Speed,rpm

Fig. 5. Torque versus speed. Fig. 8. Power loss versus speed.

24.0
8
Hydrodynamic Load Support,N
Reverse Pumping Rate,g/h

7
23.5
6

5
23.0
4

3
22.5
2

1
22.0
0 0 1000 2000 3000 4000 5000 6000
0 1000 2000 3000 4000 5000 6000
Speed,rpm
Speed,rpm
Fig. 9. Hydrodynamic load support versus speed.
Fig. 6. Reverse pumping rate versus speed.

Fig. 6 shows the reverse pumping rate as a function of the shaft Fig. 9 shows the hydrodynamic load support as a function of
speed. There is almost a linear increase of pumping rate speed. At low speed, where the seal operates in the mixed
with speed. lubrication regime, the hydrodynamic load support increases with
Fig. 7 shows the variation of contact area ratio with speed. As speed while the contact load decreases. Above the critical speed,
one might expect, the contact area decreases with speed until the the fluid film is fully established and there is no asperity contact,
area becomes nearly zero when the speed is higher than a critical so the hydrodynamic load support no longer changes with
value, 2500 rpm, as shown in the figure. Above this speed, the seal increasing speed.
operates with full film lubrication, while below it, mixed lubrica- Figs. 10–14 show the effect of surface roughness height on seal
tion occurs. Referring back to Fig. 6, it is interesting to note that performance, holding all other base parameters constant. It can be
the behavior of the reverse pumping rate does not change as the seen from Fig. 10 that, as the RMS roughness is increased, the
seal operation changes from mixed lubrication to full film minimum film thickness increases, which indicates that the
lubrication at 2500 rpm. hydrodynamic effect increases. Consequently, the reverse pump-
The power consumption is equal to the product of torque ing rate increases with roughness, as shown in Fig. 11. As one
times shaft speed. Fig. 8 shows that the power consumption might expect, the contact area ratio will increase with roughness,
increases almost linearly with speed. As in the case of the reverse as shown in Fig. 12. Although the friction will increase with
pumping rate, the behavior of the power consumption does not roughness due to the increasing contact area, yet the viscous
change as the seal operation changes from mixed lubrication to shear stress within the fluid decreases with roughness due to the
full film lubrication. increasing film thickness, with the result that the torque
200 F. Guo et al. / Tribology International 57 (2013) 195–201

1.60 24.4
Minimum Film Thickness,µm

Hydrodynamic Load Support,N

24.2
1.55 24.0

23.8
1.50 23.6

23.4
1.45
23.2

23.0
1.40
22.8
0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
RMS Roughness,µm RMS Roughness,µm

Fig. 10. Minimum film thickness versus RMS roughness. Fig. 14. Hydrodynamic load support versus RMS roughness.

4.5
support with RMS roughness. However, this is not to say that the
4.0 higher the RMS roughness, the better the seal performance. When
Reverse Pumping Rate,g/h

3.5 the roughness increases, the wear of the lip surface also increases,
which seriously affects the seal performance.
3.0

2.5
4. Conclusions
2.0
The results of this study demonstrate that a mixed lubrication
1.5
model can produce reasonable predictions of the key performance
1.0 characteristics of a rotary lip seal, such as minimum film thickness,
0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 contact area, reverse pumping rate and torque. It is important to note
RMS Roughness,µm that the current model using the flow factors results in faster com-
putations than comparable models based on a deterministic approach.
Fig. 11. Reverse pumping rate versus RMS roughness.

0.14 Acknowledgments

0.13
The work described in this paper was supported by National
Contact Area Ratio

Science Funding of China (51175283) and the National Science

0.12
and Technology Support Program of China (2011BAF09B06).

0.11
References
0.10
[1] Jagger ET. Rotary shaft seals: the sealing mechanism of synthetic rubber seals
running at atmospheric pressure. Proceedings of the Institution of Mechan-
0.09 ical Engineers 1957;171:597–616.
[2] Hirano F, Ishiwata H. The lubricating condition of a lip seal. Proceedings of
0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 the Institution of Mechanical Engineers 1965;180:187–96.
RMS Roughness,µm [3] Jagger ET, Walker PS. Further studies on the lubrication of synthetic rubber
rotary shaft shaft seals. Proceedings of the Institution of Mechanical Engi-
Fig. 12. Contact area ratio versus RMS roughness. neers 1966;9:191–204.
[4] Johnston DE. Using the fraction torque of rotary shaft seals to estimate the
film parameters and the elastomer surface characteristics. In: Proceedings of
0.427 the eighth international conference on fluid sealing. BHRA; 1978.
[5] Gabelli A, Poll G. Formation of lubricant film in rotary sealing contacts: part
0.426 I—lubricant film modeling. ASME Journal of Tribology 1992;114:280–9.
0.425
[6] Qian DS. The sealing mechanism and design factors of radial lip seals for
crankshafts. Neiranji Gongsheng (Chinese Internal Combustion Engine Engi-
Torque,N/m

0.424 neering) 1984;5:10–3.

[7] Kammüller M. Zur abdichtwirkung von radial-wellendichtringen. Dr. Ing.
0.423 Thesis. Germany: University of Stuttgart; 1986.
[8] Müller HK. Concepts of sealing mechanism of rubber lip type rotary shaft
0.422
seals. In: Proceedings of the 11th international conference on fluid sealing.
0.421 BHRA; 1987. p. 698–709.
[9] Gabelli A. The effect of operating parameters upon radial lip seal perfor-
0.420 mance. SAE paper 841145; 1985.
[10] Salant RF, Flaherty AL. Elastohydrodynamic analysis of reverse pumping in
0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
rotary lip seals with microundulations. ASME Journal of Tribology 1994;116:
RMS Roughness,µm 56–62.
[11] Shi F, Salant RF. Numerical study of a rotary lip seal with a quasi-random
Fig. 13. Torque versus RMS roughness. sealing surface. ASME Journal of Tribology 2001;123:517–24.
[12] Patir N, Cheng HS. An average flow model for determining effects of three-
dimensional roughness on partial hydrodynamic lubrication. ASME Journal of
decreases with RMS roughness, as shown in Fig. 13. Fig. 14 shows Lubrication Technology 1978;100:12–7.
[13] Patir N, Cheng HS. Application of the average flow model to lubrication
the hydrodynamic load support as a function of RMS roughness. It between rough sliding surfaces. ASME Journal of Lubrication Technology
is seen that there is almost a linear increase of hydrodynamic load 1979;101:220–9.
 F. Guo et al. / Tribology International 57 (2013) 195–201 201

[14] Salant RF, Rocke AH. Hydrodynamic analysis of the flow in radial lip seal [19] Lo SW. On the effects of roughness orientation—a mapping and double scale
using flow factors. Transactions of the ASME Journal of Tribology 2004;126: analysis of flow factors. ASME Journal of Tribology 1992;114:747–54.
156–61. [20] Harp SR, Salant RF. An average flow model of rough surface lubrication with
[15] Rocke AH, Salant RF. Elastohydrodynamic analysis of a rotary lip seal using inter-asperity cavitation. ASME Journal of Tribology 2001;123:134–43.
flow factors. Tribology Transactions 2005;48:308–16. [21] Greenwood JA, Williamson JBP. Contact of nominally flat rough surfaces.
[16] Jia XH, Jung S, Haas W, Salant RF. Numerical simulation and experimental Proceedings of the Royal Society of London Series A 1966;295:300–19.
study of shaft pumping by plunge ground shafts with rotary lip seal. [22] Nikas GK, Sayles RS. Nonlinear elasticity of rectangular elastomeric seals and
Tribology International 2012;48:155–61. its effect on elastohydrodynamic numerical analysis. Tribology International
[17] Shen D, Salant RF. Elastohydrodynamic analysis of the effect of shaft surface 2004;37:651–60.
finish on rotary lip seal behavior. Tribology Transactions 2003;46:391–6. [23] Jia XH, Jung S, Haas W, Salant RF. Numerical simulation and experimental
[18] Prati EA. A theoretical–experimental method for analyzing the dynamic study of shaft pumping by laser structured shafts with rotary lip seal.
behavior of elastomeric lip seals. Rubber Chemistry and Technology
Tribology International 2011;44:651–9.
1987;60:176–89.