Professional Documents
Culture Documents
13 - Friction in Fluid
13 - Friction in Fluid
culation. While this approach is useful for research purposes, it does not
suffice for use in the determination of ball and roller bearing perform-
ance in practical engineering applications.
This chapter describes an approach that synthesizes state-of-the art
models forlubricant film thickness and asperity load sharing into a prac-
ticable, analytical description of a real, rolling element-raceway contact.
For the contact of real surfaces Greenwood and Williamson [13.1] devel-
oped one of the first models that specifically accounted for the random
nature of interfacial phenomena. Themodel applies tothe contact of two
flat plastic planes, one rough and the othersmooth. It is readily adapted
to the case of two rough surfaces as discussed further below. In the GW
model the rough surface is presumed to be covered with local high spots
or asperities whose summits are spherical. The summits are presumed
to have the same radius R, but randomly variable heights, and to be
uniformly distributed over the rough surface witha known density DSUM
of summitshnit area.
The mean heightof summits liesabove the mean heightof the surface
as a whole by the amount Z, indicated in Fig. 13.2. The summit heights
x, are assumed to follow a Gaussian probability law with a standard
deviation u-,.Figure 13.3 shows the assumedform for the summit height
distribution or probability density function (pdf) f(z,). It is symmetrical
about the mean summit height. The probability that a summit has a
height, measured relative to the summit mean planein the interval(z,?
z, + dz,) is expressed in termsof the pdf as f(z,) dz,. The probability that
a randomly selected summit has a height in excess of some value d is
the area under thepdf to the right of d. The equation of the pdf is
(13'.1)
S U ~ ~ HEIGHT
I T
~IST~I~UTIO~
~ . ~ -
(13.3)
where F&) is the area under the standard normal curve to the right of
the value t. Values Fo(t)for t ranging from 1.0 to 4.0, are given in column
2 of Table 13.1.
It is assumed that when large flat surfaces are pressed together,their
mean planes remain parallel. Thus, if a rough surface and a smooth
surface are pressed against each other until the summit mean plane of
the rough surface and the mean plane of the smooth surface are sepa-
rated by an amount d, the probability that a randomly selected summit
will be a microcontact is
(13.7)
where E' = [(l- v?)lE, + (1 - ~i4)lE~I-l and Ei, vi (i = 1,2) are "Young's
moduli and Poisson's ratios for the two bodies. The maximum Hertzian
pressure in the microcontact is
(13.8)
(13.9)
which transforms to
(13.10)
where
(13.11)
(13.12)
By the same argument the total load per unit area supported by asper-
ities is
(13.13)
where
(13.14)
s2ie e2
A contacting summit will experience some degree of plastic flow when
the maximum shear stressexceeds half the yield stress insimple tension.
In thecontact of a sphere and aflat, the maximum shear stressis related
to the maximum Hertzian stress p o by
or
(13.17)
z>d+wp (13.18)
(13.19)
where
(13.20)
en the mean plane of a rough surface with this rms value is held at
ight h above a smooth plane, the rms value of the gap width is the
same as shown in Fig. 13.3, where both surfaces are rough. It is in this
sense that the surface contact of two rough surfaces may be translated
into the e~uivalentcontact of a rough surface and a smooth surface. As
2, the summit and surface mean planes
- 40-
x, = - (13.22)
G
The quantity a, h o w n as the bandwidth parameter, is defined by
where m,, m2,and m4 are known as the zeroth, second, and fourth spec-
tral moments of a profile. Theyare equivalent to the mean square height,
slope, and second derivative of a profile in an arbitrary direction; that is
m, = E (x2) = u2
(13.25)
(13.26)
(13.27)
d=h-Z, (13.28)
(13.29)
Equation (13.29) shows that dIus is linearly related to hlu. The ratio
h l u is also referred to as the lubricant film parameter A. When A > 3,
contacts are few and the surfaces may be considered to be well lubri-
cated.
For a specified or calculated value of A, dlcr, is computed from equa-
tion (13.29) for use in the GVV model. For an isotropic surface the two
SUM and 23, the average radius of the spherical caps of as-
perities, may be expressed as (Nayak C13.81):
(13.30)
(13.31)
For an anisotropic surface, the value of m, will vary with the direction
in which the profile is taken on the surface. The ma~imumand minimum
values occur in two orthogonal “principal”~irections.Sayles and T h o ~ a s
[13.9] recommend the use of an equivalent isotropic surface for whichm2
is computed as the harmonic mean of the m2 values found along the
principal directions. The value of m4 is similarly taken as theharmonic
mean of the m4 values in these two directions.
For a specified contact with semiases a and b, under a load P, with pla-
teau lubricant film thickness h and given values of m,, m,, and m4, the
a is determined by first computing PlA, from equation
[13.13] and using
(13.32)
If &, > P, the implication is that the lubricant film thickness is larger
than computed under smooth surface theory. Inthis case, equation
(13.13) could be solved iteratively until Qa= P.
The separation of the surface and summit mean plane is, by equations
(13.22) and (13.23),
(13.22)
(13.23)
= 2.006
- =
2, 4 (~n=) 1/2
(13.31)
a, = [(1- ~) 1/2
(13.27)
d - hlcr - 4/(wa)lf2
"
(13.29)
CrB (I - 0.8968Ia)'l2
(I - 0.8968/2.006)1f2
0.544
F, (0.544) = 0.2935
F , (0.544) = 0.1850
F3/2 (0.544) = 0.1812
From equation (14.13) the ratio of mean real contact area A, to nom-
inal contact area A, is
(13.12)
The actual contact area thus averages only 1.25%of the nominal con-
tact area. The mean actual pressure PIA, is
31.6
-
- "
0.0125
- 2528 N/mm2 (3.665 X lo5 psi)
(13.5)
w: = 6.4 ( ~( )~ ) 2
(13.20)
= 6.4 X
= 0.740
(13.19)
inter pol at in^ in Table 13.1 gives F, (1.284) = 0.100 and np = 177/
mm2 ( 114,000/in.2).
If the macroeontact is elliptical with semiaxes a = 3 mm (0.01181
in.) and b = 0.33 mm (0.01299 in.) under a load of P = 3500 N (786.5
lb), the mean asperity-supported load is
aU
7 - q- (12.1)
ax
( 13.34)
(13.34)
(13.35)
ean contact
pressure
rachman and Cheng [13.19] and Tevaarwerk and Johnson [13.20] in-
vestigated traction in rolling-sliding contacts and found that equation
(12.1) pertains only to a situation involving a relatively low slide-to-roll
ratio; for example, less than 0.003 as shown in Fig. 13.5. Notethat trac-
tion refers to the net frictional effect in the rolling direction. Similar to
Trachman and Cheng, for a given temperature and pressure, it is pos-
sible to define local contact friction as follows:
(13.36)
shear rate
F I ~ U R E13.6. Schematic illustration of equation (13.36).
ities on the component surfaces becomes morefrequent. The frictionthat
occurs due to sliding motions between asperities can be characterized as
Coulomb friction, such that
7
, = Pap (13.37)
where EA, is the Coulomb coefficientof friction and p is the local pressure.
On an average basis, this frictional stress may be assumed to apply to
the portion of the overall contact area associated with asperity-asperity
contact. If the contact area of the smooth components is defined as A,,
then, according to equation (13.12), the portion of the contact associated
with Coulomb friction is AJA, * A,.
(13.38)
13.1. J. Greenwood and J. Williamson, “Contact of Nominally Flat Surfaces,” Proc. RoyaZ
SOC.London A295, 300-319 (1966).
13.2. A. Bush, R. Gibson, and T. Thomas, “TheElastic Contact of a Rough Surface,” Wear
35, 87-111 (1975).
13.3. M. O’Callaghan and M. Cameron, “Static Contact under Load Between Nominally
Flat Surfaces,’’ Wear 36, 79-97 (1976).
13.4. A. Bush, R. Gibson, and G. Keogh, “Strongly Anisotropic Rough Surfaces,” ASME
Paper 78-LUB-16 (1978).
13.5. J. McCool and S. Gassel, “The Contactof Two Surfaces Having Anisotropic Rough-
ness Geometry,”ASLE Special ~ubzication(SP-71,29-38 (1981).
13.6. J, McCool, “Comparison of Models for the Contact of Rough Surfaces,” Wear 107,
37-60 (1986).
13.7. A. Bush, R. Gibson, and G. Keogh, “The Limitof Elastic Deformation in the Contact
of Rough Surfaces,” Mech. Res. Cornrn. 3, 169-174 (1976).
13.8. P. Nayak, “Random Process Model of Rough Surfaces,” Trans. ASME, J Lub. Tech-
nology 93F, 398-407 (1971).
13.9. R. Sayles and T. Thomas, “ThermalConductances of a Rough Elastic Contact,”AppZ.
Energy 2, 249-267 (1976).
13.10 T. Sasaki, H. Mori, and N. Okino, “Fluid LubricationTheory of Roller Bearings Parts
I and 11,” ASME Trans., J Basic Eng. 166, 175 (1963).
13.11 J. Bell, “Lubrication of Rolling Surfaces by a Ree-Eyring Fluid,” ASLE Trans. 5,
160-171 (1963).
13.12. F. Smith, “Rolling Contact Lubrication-The Application of Elastohydrod~amic
Theory,” ASME Paper 64-Lubs-2 (April 1964).
13.13 B. Gecim and W. Winer, “A Film Thickness Analysis for Line Contacts under Pure
Rolling Conditions with a Non-Newtonian Rheological Model,” ASME Paper 80C2/
LUB 26 (August 8, 1980).
FERE~CES
13.14. L. Houpert, “New Results of Traction Force Calculations in EHD Contacts,” ASME
Trans, J; Lub. Technology l07(2), 241 (1985).
13.15. C. Evans and IC. Johnson, “The Rheological Properties of EHI) Lubricants,” Proc.
Inst. Mech. Eng. 200(C5),303-312 (1986).
13.16. S. Bair andW. Winer, “A Rheological Model for Elastohydrodynamic Contacts Based
on Primary Laboratory Data,” ASME Trans., J. Lub. Tech. 101(3), 258-265 (1979).
13.17. IC.Johnson and R. Cameron, Proc. Inst. Mech. Eng. 182(1), 307 (1967).
13.18. D. Schipper, P. Vroegop, A. DeGee, and R. Bosma, ‘“Micro-EHL in Lubricated Con-
centrated Contacts,’’A5”E Trans., J. Tribology 112, 392-397 (1990).
13.19. E. Trachman and H. Cheng, “Thermal and Non-Newtonian EEects on Traction in
Elastohydrodynamic Contacts,” Proc. Inst. Mech. Eng. 2nd Symposium on Elasto-
hydrodynamic Lubrication, Leeds, 142-148 (1972).
13.20. J. Tevaamerk and IC. Johnson, “A Simple Non-Linear Constitutive Equation for
EHD Oil Films,” ‘Wear 35, 345-356 (1975).
13.21. T. Harris and R. Barnsby, “Tribological Performance Prediction of Aircraft Turbine
Mainshaft Ball Bearings,” Tribology Trans. 41(1), 60-68 (1998).