See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/43154651

Rolling Contact Fatigue and Wear Fundamentals for Rolling Bearing

Diagnostics

Article  in  ARCHIVE Proceedings of the Institution of Mechanical Engineers Part J Journal of Engineering Tribology 1994-1996 (vols 208-210) · April 2010
DOI: 10.1243/13506501JET656 · Source: OAI

CITATIONS READS

63 984

2 authors, including:

Jari Halme
VTT Technical Research Centre of Finland
16 PUBLICATIONS   117 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Jari Halme on 29 September 2014.

The user has requested enhancement of the downloaded file.

 REVIEW PAPER 377

Rolling contact fatigue and wear fundamentals for

rolling bearing diagnostics – state of the art
J Halme∗ and P Andersson
VTT Technical Research Centre of Finland, Espoo, Finland

The manuscript was received on 15 April 2009 and was accepted after revision for publication on 27 October 2009.
DOI: 10.1243/13506501JET656

Abstract: Rolling bearing operation is affected by friction, wear and lubrication mechanisms,
fluid dynamics and lubricant rheology, material properties, and contact mechanics. Changes
in rolling surfaces occur due to plastic deformation, rolling contact wear, and rolling contact
fatigue. Wear particles can be formed and mixed into the lubricant. Increased levels of vibra-
tions due to surface degradation can be monitored by sensors. Rolling contact wear and rolling
contact fatigue during rolling bearing operation can be diagnosed by combining measured and
interpreted condition monitoring data with theory, and conclusions drawn thereof can support
a continuous prognosis for the remaining bearing life. In the present work, connections between
bearing diagnostics and tribological mechanisms are outlined.

Keywords: rolling bearings, wear, rolling contact fatigue, condition monitoring, diagnostics

1 INTRODUCTION expresses the number of millions revolutions that 90

per cent of a bearing population is expected to sur-
Rolling bearings have a great influence on the indus- vive under conventional operating conditions before
trialized world. The bearing life theory has been devel- the first incidence of fatigue pit formation at any of the
oped more than a century ago, starting from the testing rolling surfaces [9].
of rolling bearings by Stribeck in 1896, followed by The lifetime of a rolling bearing is nowadays usually
work by Goodman in 1912 [1, 2]. A fatigue limit con- estimated not only by calculations of the basic rating
cept was introduced by Palmgren in 1924 [3]. In 1939, life, but rather by calculations of the modified rating
Weibull presented his statistical theory of the strength life Lnm [9] using the formula
of materials as a mathematical tool for analyses on the
bearing life scatter [4]. In the years 1947–1952, Lund- Lnm = a1 × aISO × L10 (2)
berg and Palmgren [5, 6] published their life theory of
rolling bearings, which was adopted as the formula for
the basic rating life of rolling bearings, L10 , in millions where a1 is the modification factor for reliability other
of revolutions, or than 90 per cent and aISO represents parameters such
 p as load, speed, oil viscosity, material fatigue limit, and
C content of contamination particles in the oil. From
L10 = (1)
P a statistical point of view, failures can be initiated in
10 per cent of a bearing population (at a1 = 1) before
where C is the basic dynamic load rating for the reaching the Lnm limit. The exact prediction of the first
bearing, P is the equivalent bearing load, p is an initiation of a failure in a specific bearing is practically
exponent with the value p = 3 for ball bearings, and not possible. For fatigue damages that occur before the
p = 10/3 for roller bearings [1, 7, 8]. The L10 value calculated number of revolutions is achieved, bearing
diagnostics is a powerful tool for avoiding undesired
consequences of bearing failures. Furthermore, if a
∗ Corresponding author:VTT Technical Research Centre of Finland,
bearing becomes subjected to wear instead of rolling
P. O. Box 1000, MK6, FIN-02044 VTT (Espoo), Finland. contact fatigue, the lifetime cannot be mathematically
email: [email protected] predicted on a statistical basis.

JET656 Proc. IMechE Vol. 224 Part J: J. Engineering Tribology

378 J Halme and P Andersson

2 ROLLING CONTACT MECHANICS Third, in the surface (z = 0) at the centre of contact

(r = 0), the radial stress σr is 0.8 × p0 [11, 12].
2.1 Loading of elastic solid bodies
When two non-conforming solids of geometrically 2.2 Non-Hertzian normal contact of elastic bodies
ideal shapes are brought into contact, the initial con- under rolling/sliding contact
tact is a point or a line, which can be analysed with
the Hertzian contact theory [10]. The contact pres- When a rolling surface is compressed in one direction,
sure from the normal load causes principal stresses its material tends to expand in the transverse direc-
and shear stresses beneath the contact surface (Fig. 1). tions according to Hooke’s law and Poisson’s ratio, and
Three features of the contact mechanics can be seen this creates tangential forces and local slip between
in Fig. 1. First, under normal static loading, the angle contact surfaces at the trailing edge of the rolling
of principal or maximum shear stress is ±45◦ in the surface [10, 13].
middle of the normal pressure (however, in a rolling In deep groove ball bearings, the raceway geometry
and sliding contact, tangential forces affect the angle of is a semi-conforming groove, and the contact area is a
the principal shear stress). Second, under the Hertzian curved ellipse with three slip zones (see Fig. 2); a cen-
contact pressure, the shear stress reaches its maxi- tral zone with a micro-slip in the direction of rotation
mum below the surface at a depth of 0.48a and the of the ball and two outer zones with a micro-slip in
maximum shear stress value is 0.31 × p0 with steel. the opposite direction [10]. Similarly, micro-slip zones

Fig. 1 A parabolic Hertzian contact pressure (upper image) acting on a circular point contact
surface of radius a. Contours of constant principal shear stress (τ1 ) and trajectories of
principal stress (σ1 , σ2 ) directions (left image in the middle) and sub-surface stress dis-
tribution along the axis of symmetry (right image in the middle) under elastic Hertzian
contact pressure in steel with ν = 0.3 [10, 11], and principal stress planes at points A and B
(lower image)

Proc. IMechE Vol. 224 Part J: J. Engineering Tribology JET656

 Rolling contact fatigue and wear fundamentals for rolling bearing diagnostics 379

2.4 Dynamical impacts and vibrations in

structures surrounding rolling contacts
Rolling contacts comprise dynamically varying stress
fields from surface roughness and contaminant par-
ticles. A wave motion initiated from any point of
dynamic stress propagates as spherical fronts of pres-
sure and shear waves [10, 15]. The interaction of pres-
sure and shear waves with the free surface gives rise to
the Rayleigh wave, which is a surface wave. In rolling
contact, fatigue crack formation and contaminant par-
ticle crushing introduce high-frequency vibrations.
Figure 4 shows the principle for the propagation of
high-frequency vibration waves.
During the progress of the wave fronts, their energy
Fig. 2 Micro-slip zones in a deep-groove ball bearing, remains almost constant, with the exception of minor
the Hertzian contact pressure p(x, y) and the trac- losses due to hysteresis. In rolling contacts, the
tion q(x, y) between a ball and a plane when a Rayleigh surface waves dominate and expand circu-
ball is rolling on a flat plane under a coefficient larly on the material surface. Partly the waves propa-
of friction μ without any influence of a liquid gate through the interface to the bearing housing and
lubricant partly the energy reflects back to the bearing [16].
In terms of the response properties, a theoretical
impulse of infinitely small duration contains equal
occur in spherical roller bearings but not in cylindrical
energy at all frequencies and a linear system is com-
roller bearings.
pletely specified by its impulse response [17, 18]. In a
mechanical system, like a rolling bearing, an incom-
2.3 Ideal and real surfaces ing impact excites all nominal vibration frequencies of
the system below the frequency of the impact, and as a
Real bodies have form errors, surface waviness, and consequence of this secondary excitation of higher fre-
surface roughness due to the surface manufacturing quencies within the system occurs. Most of the energy
and previous operation. In contacts between rolling of the vibration is bound to the nominal frequencies
elements and raceways, the real contact area is smaller of the structure and the displacement amplitude is at
than the apparent one and the border of the real con- its highest at the lower frequencies. The sustaining
tact area may expand outside that of the theoretical frequencies are typically measured with appropriate
one (see Fig. 3) [10, 14]. vibration acceleration sensors and the propagating
wave fronts with acoustic emission (AE) sensors. The
analyses of these are discussed in chapter 6.

3 LUBRICATION OF ROLLING BEARINGS

3.1 Lubrication regimes

Entrapment of a liquid into a converging gap at a slid-
ing or rolling contact creates a hydrodynamic pressure
and a lubricant film to separate the contact surfaces
and reduce the contact pressure. The ratio between the
oil film thickness and the combined surface roughness
in the contact, or the lambda (λ) value, determines
the lubrication mechanism and its effect [19–22]. An
overview of lubricating mechanisms and the Stribeck
curve are shown in Fig. 5 [2, 22, 23]. Transitions from
one lubrication mechanism to another may occur for
reasons of operation, running-in or wear [20, 24, 25].
Fig. 3 The contact of an ideally smooth elastic sphere
with an ideally smooth (broken line and contact 3.2 Boundary lubrication
radius rs ) and a randomly rough (solid line and
contact radius rr ) flat surface, and the respective If the hydrodynamic action is too weak for separat-
normal pressures [10] ing the contact surfaces from each other (λ < 1), and

JET656 Proc. IMechE Vol. 224 Part J: J. Engineering Tribology

380 J Halme and P Andersson

Fig. 4 Generation and propagation of a high-frequency pulse from crack initiation. The illustra-
tions in the centre represent acoustic wave types in the material and on its surface

the surfaces are only wetted by oil, the load is carried rolling elements and the raceways, or λ > 1. In
mainly by solid-to-solid contacts and boundary films practice, details in the texture of the contact surfaces
on the surface asperities [14, 23, 26, 27]. Boundary and the kinematics of the contact may cause large
lubrication relies on lubricant and surface chemistry variations in the λ-value requirement for appropriate
and the ambient atmosphere, on the formation rate of EHD lubrication [20, 29]. The minimum oil film thick-
the boundary films and on their mechanical proper- ness h under viscous–elastic EHD conditions can be
ties [14]. As a result of the solid-to-solid contact con- calculated using the formula [24]
ditions, some wear is to be expected by tribochemical
or mechanical wear [27]. In rolling bearings, bound- h = 3.63 × U 0.68 × G 0.49 × W −0.073 × (1 − e−0.68k )
ary lubrication can occur locally in micro-slip zones (3)
and at the end planes of axially loaded rollers. Bound-
ary lubricated contacts lack any damping effect, and where U is the dimensionless speed parameter (vis-
boundary lubricated contacts can act as a vibration cosity × velocity/elasticity modulus × radius), G is the
sources. dimensionless materials parameter (elasticity modu-
lus × pressure − viscosity coefficient), W is the dim-
ensionless load parameter (load/elasticity modulus ×
3.3 Mixed lubrication radius), and k is the ellipticity parameter for the con-
tact junction area. From formula (1), it is obvious that
Mixed lubrication is characterized by boundary lubri-
the oil film thickness under conditions of EHD lubri-
cated conditions as part of Elastohydrodynamic (EHD)
cation is only slightly dependent on the load and the
or hydrodynamic lubrication [28]. In mixed lubrica-
elasticity modulus, while the speed and oil viscosity
tion, the solid-to-solid proportion of the contact can
have a strong effect [24, 32].
be lubricated by a solid-like or viscous-like boundary
Micro-elastohydrodynamic lubrication means an
film [22].
elastic smoothening of the original surface rough-
ness and an improvement in the EHD lubrication
3.4 EHD lubrication conditions in comparison with the nominal λ-value
calculated for the unstressed surfaces [20].
EHD lubrication (EHL) refers to lubrication conditions The rolling elements eject some of the oil from the
based on Reynolds’ equation for hydrodynamic oil raceways and some of the oil is lost because of the
pressure build-up [24], and a combination of elastic centrifugal action. Starvation occurs in rolling bear-
deformation at contact surfaces of low geometric con- ings if the time between subsequent events of contact
formity, and an increase in oil viscosity under high between rolling elements and raceways is too short
pressure in the contacts. The slip-to-roll ratio in a for ensuring a sufficient replenishment of lubricant
rolling contact influences the oil film formation [29, to support a fully flooded EHD film formation [20,
30]. Conditions of EHL is the basic assumption behind 23, 33, 34]. In unfavourable cases, debris particles
bearing life calculations [9, 20, 23, 24, 31]. can agglomerate at the inlet zone of an EHD con-
In the EHL regime, the oil film thickness is slightly tact and cause oil inlet blockage and starvation [35].
higher than the combined surface roughness of the As the severity of starvation increases, the pressure

Proc. IMechE Vol. 224 Part J: J. Engineering Tribology JET656

 Rolling contact fatigue and wear fundamentals for rolling bearing diagnostics 381

Fig. 5 Schematic representation of a Stribeck curve [2] and its relation to lubrication regimes,
vibration response, and wear mechanism schematics; top: vibration response descriptions
and schematics, in which the dashed line indicates rotational vibration (e.g. unbalance, bent
shaft, and misalignment) and the solid line indicates transient vibration (e.g. impacts and
stress waves); middle: wear mechanism schematics; bottom: Stribeck curve with coefficient
of friction μ, velocity v, viscosity η, and load P

spike at the outlet zone of the EHD contact (Fig. 6) bearings. Starvation reduces the oil film thickness,
becomes smaller, the film becomes thinner and flat- which increases the stress and vibration level.
ter over most of the contact surface [23], the area
of load-carrying oil pressure becomes smaller, and
3.5 Hydrodynamic lubrication
the load-carrying capacity is reduced [24]. In grease-
lubricated rolling bearings, fully flooded conditions In hydrodynamic lubrication, sliding surfaces are
can occur in freshly lubricated bearings, while after completely separated by a fluid film, which is thick
a short time the bearing is normally running under in comparison with the surface roughness and
starved conditions [19, 36]. thicker than in EHD lubricated conditions. In rolling
Operation under EHL or μ-EHL conditions causes bearings, hydrodynamic lubrication occurs in con-
numerous dynamic stress field elements, which act tacts between rolling elements and their cages, and
as weak sources of vibrations. The thin oil film between roller ends and bearing race flanges in
provides only weak damping properties in rolling roller bearings with axially loaded rollers. The thick

JET656 Proc. IMechE Vol. 224 Part J: J. Engineering Tribology

382 J Halme and P Andersson

Fig. 6 Contact pressure and film thickness profiles [23]

oil film in hydrodynamically lubricated tribosystems 4.1 Fatigue wear

provides a significant degree of damping of dynamic
Fatigue wear can occur in surfaces that are dynami-
loads.
cally loaded. Under elastic contacts, the fatigue pro-
cess usually requires high number of cycles, while
under plastic contact, a low-cycle fatigue mechanism
3.6 Influence of contaminated lubricants can be expected [42].
Lubricants contaminated by hard, sharp particles
cause abrasive contact conditions similar to (or more 4.2 Rolling contact fatigue
severe than) those of boundary lubricated contacts.
Dynamic contact conditions, together with possible
Each particle that is larger than the oil film thick-
tangential stresses and hoop stresses, and intensified
ness can form a contact spot, which transmits part
by local asperity or contaminant particle contacts,
of the load and becomes surrounded by a local
cause local surface and subsurface shear stress fields,
stress field. Lubricants contaminated by solid particles
plastic deformation and dislocations, and eventually
are known to cause abrasive wear both in hydro-
the nucleation of rolling contact fatigue cracks at
dynamically lubricated thick-film contacts [37] and
defects or inclusions in the micro structure. The initial
EHL contacts [38–40]. Lubricant cleanliness data have
fatigue cracks are usually inclined to the surface [23,
been included as a parameter in rolling bearing life
43]. Intersection of fatigue cracks leads to pit forma-
calculations [9].
tion and wear. The rolling contact fatigue process is
usually self-propagating through surface roughening,
dynamic loads, and a decrease in the load-carrying
4 WEAR MECHANISMS AND SURFACE QUALITY area. The formation of dislocations and cracks, and
IN ROLLING BEARINGS the subsequent relative motion between fractured
surfaces, gives rise to vibrations in terms of AE.
Several wear mechanisms [41] can be active in rolling The life of rolling bearings, in terms of rolling con-
bearings. Additionally, the surface texture can be tact fatigue, can be estimated on statistical basis,
changed by plastic deformation. from the bearing design, load, speed, and lubrication

Proc. IMechE Vol. 224 Part J: J. Engineering Tribology JET656

 Rolling contact fatigue and wear fundamentals for rolling bearing diagnostics 383

conditions [1, 7, 9, 23, 24]. Present calculation pro- 5.1 Running-in wear of rolling bearings
cedures do not consider uncontrolled factors such
as electric currents, water in the lubricating oil and Under favourable running-in conditions, a ‘roll
hydrogen embrittlement of the rolling surfaces, tri- polishing’ effect occurs in rolling bearings, mainly
bochemical wear or corrosion of the bearing, strong through plastic deformation of surface asperities.
external vibrations during operation or standstill, or The roll polishing effect gradually disappears as the
any influence of intermittent operation [9, 44–46]. surface roughness decreases and a transition from
In roller bearings, the roller profile and its pres- mixed lubrication into EHD lubrication conditions
sure distribution can affect the bearing life [9, 47]. A occurs. The running-in process can be re-activated by
stress-life method, which takes into consideration the changed operational conditions [24, 25, 29, 43, 50].
entire stress situation in a rolling contact, has been Each incident of plastic deformation of a surface
developed in recent years [1]. asperity during running-in forms a dynamic stress
Under less severe rolling contact conditions, micro- field and a vibration source.
pitting can occur as the dominating surface deteriora- Micro-abrasive and sliding fatigue wear in the
tion effect [43]. Below the threshold for fatigue crack micro-slip regions (Fig. 2) of rolling contacts are res-
formation, surface changes in rolling contacts is limi- ponsible for part of the surface alterations, for rolling
ted to roll polishing by a mild plastic deformation or contact wear, and for wear particle formation in the
micro-abrasion. running-in of rolling bearings [51]. The plastic defor-
mation caused by abrasion and the micro cracking
caused by fatigue are weak sources of vibrations.
4.3 Adhesive wear Abrasive wear and surface indentations are caused
by particles in the oil during the running-in stage of
Under poor lubrication, rolling bearings may suffer
the bearing [52]. Abrasive wear by contaminant parti-
from adhesive wear at roller ends and in micro-
cles during the running-in stage of a bearing usually
slip zones. In these sliding contacts, strong adhesive
roughens the rolling surfaces and produces more wear
junctions between surface asperities may be formed
particles, and therefore weakens the preconditions for
due to frictional heating, and adhesive wear may
subsequent operation of the bearing [25]. Crushing of
occur [14, 42].
particles, abrasion by particles, and contact dynam-
ics due to surface roughening contribute as sources of
4.4 Abrasive wear vibration in a rolling contact.
Under running-in at stress levels leading to local
Solid, hard particles can cause local stress peaks and rolling contact fatigue at areas of asperity interaction,
shorten the life of the bearing [35, 38, 39, 48, 49]. surface roughening by micropitting can eventually
Even particles smaller than the mean film thickness occur. The small depth of the micropits gives a low
may cause abrasive wear. Furthermore, abrasive wear volume loss, although an abundance of wear debris of
may take place in contacts between rough surfaces, for small size. Micropitting increases the vibration level of
instance as micro-abrasion (or sliding micro-fatigue) the bearing, through contact dynamics, due to the sur-
between rolling elements and their cages. The addi- face roughening and the increasing contents of wear
tional solid-to-solid contact spots in the EHD or mixed particles in the contacts.
lubricated contact form local stress risers [1], which Depending on the rolling conditions during run-
under dynamic loading act as vibration sources. ning-in and the initial smoothness of the surfaces,
the surfaces smoothness can be either enhanced or
4.5 Tribochemical wear decreased. The wide band root mean square (RMS)
values of the vibration acceleration and the noise gen-
In tribochemical or corrosive wear, chemical or elec- erated directly from the surface asperities decrease if
trochemical reactions in the tribological contacts the surface asperities are smoothened during running-
result in the formation of soft or brittle surface lay- in. Whether this is detectable or not depends on the
ers, which are easily removed by subsequent contact background noise and other vibration sources of the
events. bearing. In addition to the changes in the RMS values,
the frequency content of vibrations and noise origi-
5 STAGES OF WEAR AND VIBRATIONS DURING nating from the surface roughness can change during
ROLLING BEARINGS SERVICE LIFE running-in.

The processes of rolling contact wear and rolling con-

5.2 Wear and fatigue in rolling bearings under
tact fatigue in rolling bearings during different stages
steady-state conditions
of the operational life of the bearings are described in
the following, and a schematic illustration is presented Most of the operational lifetime of a rolling bearing
in Fig. 7. is supposed to occur under steady-state conditions

JET656 Proc. IMechE Vol. 224 Part J: J. Engineering Tribology

384 J Halme and P Andersson

Fig. 7 Upper image: rolling bearing wear volume accumulation from running-in towards major
rolling contact wear or towards rolling contact fatigue spall formation. Lower image: the
corresponding evolution from running-in to rolling contact fatigue failure at rolling bearing
raceway surface and subsurface (inserted photograph: rolling contact fatigue failure in roller
bearing)

after the running-in stage. The run-in surfaces are fractured surfaces give rise to vibrations in terms of AE
the smoothest and the vibrations are at their lowest from the surfaces subjected to the fatigue process.
level. The operating conditions during the steady-state When the calculated rating life for a bearing has
phase will finally determine whether the bearing will been reached, the probability for a first occurrence
be subjected to rolling contact wear only or if the con- of pitting is 10 per cent (factor a1 = 1), by definition,
tact mechanism will develop into a process of rolling although the likelihood for fatigue pitting in practice
contact fatigue (Fig. 7). is reduced by built-in safety margins in the calculation
As in the running-in stage, micro-slip on part of the procedure. Upon the onset of pitting, the steady-state
rolling contact is responsible for part of the abrasive of the bearing operation is terminated and the bearing
wear and sliding fatigue wear as well as part of the moves into a stage of accelerated fatigue wear.
wear debris formation in steady-state operation [51].
Abrasive wear and surface indentations can roughen
5.3 Accelerated rolling contact fatigue and
the rolling surfaces and shorten the bearing life
rolling bearing failures
[1, 35, 38–40, 48]. Crushing of particles, abrasion by
particles, and increased contact dynamics due to sur- Choi and Liu [53] have divided the last stage of the
face roughening contribute as sources of vibration rolling contact fatigue process of a bearing into two
under steady-state operation, usually with increasing periods. The first period shows no significant change
intensity. in the vibration amplitude, as the crack initiation
When a bearing operates above fatigue threshold and propagation occur below the surface. The second
conditions, the bearing will be subjected to rolling period shows a significant increase in the vibration
contact fatigue cracking. The dislocation and crack for- amplitude, due to the formation and progress of a spall
mation and the subsequent relative motion between on the surface.

Proc. IMechE Vol. 224 Part J: J. Engineering Tribology JET656

 Rolling contact fatigue and wear fundamentals for rolling bearing diagnostics 385

Once initiated and developed into a first local pit- certain intervals, the frequencies of which depend on
ting failure, the running conditions of the bearing the bearing geometry features, the rotational veloc-
evolve, through the formation of fatigue wear parti- ity, and the location of the defect or defects [55–57].
cles, surface roughening, and dynamic loads, into a Figure 8 shows the theoretical generation of cyclically
stage of progressive pitting and failure development. repeated periods of vibrations due to bearing defects,
The acceleration of the fatigue and pitting process is and their envelope curves under constant load.
accompanied by an increase in the vibration level, rate Envelop curves are in this context visualizations
of wear particle formation, and wear particle size [54]. of the dynamics within vibrations originating from
The surface roughening increases the tangential point defects on rolling surfaces. The envelope curves
forces on the rolling elements in the bearing. The are determined from vibration acceleration data in
increase in the vibration level, surface roughness, and the time domain, which are filtered using a band
tangential forces in the bearing can lead to secondary pass filter in the expected natural bearing frequency
damages such as raceway fractures, typically originat- range, which typically lies between 500 and 3000 Hz.
ing from fatigue pits, and roller cage breakages. In the The purpose of the filtering is to remove the back-
most severe scenario, cage breakages block the motion ground noise around the natural frequencies, which
of the rolling elements. are modulated due to periodic impacts from point
defects at the rolling contacts. After performing a
Hilbert transformation of the filtered vibration accel-
6 ROLLING BEARING MONITORING AND
eration data, an envelope signal is obtained. Finally,
DIAGNOSTICS
the spectrum of the envelope signal is analysed in
order to determine the characteristic, statistically rele-
6.1 Vibration analysis vant frequency of the cyclic variations of the envelope
Analysis of vibrations from rolling bearings is typically curve.
based on the results of vibration acceleration and AE In practical rolling bearing applications, the fre-
measurements, which are analysed in the time domain quencies of the cyclic impacts vary in a random
or in the frequency domain, or in both. Prior to sam- manner, due to slip, varying speeds, and varying load
pling, the vibration signals are anti-alias filtered for angle [56, 58, 59]. Consequently, the inconsistency
preventing frequencies higher than half the sampling in the cyclic impacts may complicate the detection of
frequency so as to appear at the lower frequencies. the nominal bearing defect frequencies from the raw
Analyses in the frequency domain are based on vibration spectrum. However, the envelope spectrum
spectrum analyses, which present the vibration data or its square can be successfully used for indicating the
as a function of discrete frequency components. The defect frequencies even with random slip [59, 60].
spectrum is calculated by using fast Fourier transfor- The methods for the determination of vibration
mation. In the simplest form of vibration analyses, parameters representing the rolling surface defects in
the RMS values are monitored. In practice, further bearings have obtained particular attention in recent
processing, such as signal averaging, filtering, enve- years.
lope calculation, calculation of amplitudes of selected Kiral and Karagülle [61, 62] used artificial dynamic
frequency components, cepstrum calculations, syn- loading models for healthy and defected bearings, for
chronized rotation vector and profile monitoring, etc., analysing vibration responses of the rolling bearing
is needed. After these actions, the characteristic fea- structure by a finite-element package. Band-pass fil-
tures (i.e. the vibration parameters are calculated). The tering of the vibration acceleration increased the ratio
number of different parameters, methods, practices, of the RMS value, the crest factor, and the Kurtosis
and their combinations in vibration analysis is unlim- value.
ited. In the following, the use of the most common Heng and Nor [63] have demonstrated that statisti-
methods and parameters are shortly reviewed in rela- cal parameters are sensitive to the rotational velocity
tion to certain factors interfering with the operation of of the bearing. The authors suggested that plots show-
rolling bearings. ing the crest factor against the Kurtosis value could
be used for classifying defects at the rolling surfaces of
6.1.1 Defects in rolling surfaces rolling elements, outer races, and inner races.
According to the results of experiments with faulty
When a rolling element passes a local fatigue crack, outer races of rolling element bearings, by Abdullah
pit or spalled area, or some other irregularities on the and co-workers [64], the statistical RMS parameter
race surface, an impulse occurs and this excites vibra- and the maximum values of the vibration accelera-
tions with frequencies governed by the resonances tion and the AE measurements increase with the size
of the structure. A defect in any of the active rolling of the defect, but not linearly with the bearing speed
surfaces periodically produces periods of vibrations and load. By a similar study, Li and co-workers [65]
due to the impulses from the over-rolling of the sur- showed that a suitable frequency band for the vibra-
face defects. The periods of vibrations are repeated at tion acceleration RMS was 3–5 kHz.

JET656 Proc. IMechE Vol. 224 Part J: J. Engineering Tribology

386 J Halme and P Andersson

Fig. 8 Rolling bearing defect vibration and envelope curves for a radial rolling bearing under a
fixed radial load, a constant rotational velocity of the inner race, and a static outer race.
The abbreviations BPFO (ball pass frequency outer race) and BPFI (ball pass frequency
inner race) refer to the peak frequencies of the envelope curve caused by inner and outer
race defects, respectively. The abbreviations BSF (ball spin frequency) and FTF (fundamen-
tal train frequency) refer to the peak frequency of the envelope curve caused by a rolling
element defect or a cage defect, respectively. RPM stands for the rotational velocity

In a study by Orhan et al. [57], a spectral analysis frequency pattern with an increasing degree of bearing
of the vibration acceleration was successfully used for damage, the Poincare map covering more than 2000
detecting ball bearing looseness and defects in ball shaft revolutions provided an accurate quantification
bearings and cylindrical roller bearings. The authors of the damage level and a reliable indication of the
observed that when a rolling surface defect reached location of the damage in the bearing.
an advanced stage, amplitudes of high frequency com- By performing test rig studies on rolling element
ponents disappeared due to self-hammering of sharp bearings, McFadden and Toozhy [70] showed that syn-
surface flaws originating from the defects. chronous averaging of vibration acceleration could be
Khemili and Chouchane [66] used an adaptive noise used for determining the location of the inner race
cancellation technique to increase the signal-to-noise fault distribution as well as the variations within the
ratio of a vibration acceleration signal from rolling ele- set of rolling elements.
ment bearings with artificially caused defects at the Ericsson et al. [71] analysed vibration acceleration
outer and inner races. Li and Li [67] showed that both signals from 103 measurements carried out at labora-
the short-time energy strength and zero-crossing rates tories and industrial companies, on bearings of which
of the AE signals increases upon an evolution from the true condition was known. As a result of the ana-
healthy rolling bearings to ones with damages on the lysis, the wavelet, the envelope, and the synchronous
outer races and rollers. Using a rolling bearing test rig, time domain averaging, each having an error rate of
Al-Dossary et al. [68] have shown that the transient AE about 10 per cent, were considered the best meth-
burst duration correlated with the size of a defect at ods for vibration data analysis. For extremely slowly
the outer raceway. rotating rolling bearings with an outer race defect,
Choy et al. [69] studied damage identification on the synchronous time domain averaging was found to
artificially caused damages on the inner race and a ball be the best method, while the other methods studied
of a ball bearing in a constantly loaded bearing test rig. were found unreliable.
The change from a healthy bearing into a faulty one At low speeds, the band energy of the envelope
was clearly detected with both the traditional vibra- spectrum has shown to be a powerful tool for outer
tion frequency analysis methods and a Poincare map. and inner race defects but not for rolling element
While no significant change was recognized in the defects [61, 62]. Jamaludin and Mba [72, 73] analysed

Proc. IMechE Vol. 224 Part J: J. Engineering Tribology JET656

 Rolling contact fatigue and wear fundamentals for rolling bearing diagnostics 387

the AE signals from tests with extremely slowly rotating Applying statistical, spectrum, and wavelet analy-
rolling bearings with and without artificially caused sis methods, Prabhakar et al. [80] have investigated
defects on the outer race, on the inner race, and on ball bearings with artificial inner race and outer race
a roller, and concluded that the faulty bearing sur- defects. The RMS and Kurtosis results increased when
faces were identified by signatures of modelled auto- the bearing was changed from a healthy one to one
regressive coefficients of the AE. with defects. The vibration acceleration spectrum
Shiroishi et al. [74] compared healthy bearings with showed an outer race defect, but an inner race defect
ones having artificially produced raceway defects, remained invisible due to some unknown change in
by using the envelope method and an adaptive line the measurement practice. By wavelet analysis, it was
enhancer method for reducing the wide-band noise of possible to detect all the fault types and the wavelet
the envelope signal. It was shown that the wide-band analysis method was effective in detecting both single
noise reduction made the defect amplitude more visi- and multiple defects in a bearing.
ble. Vibration acceleration measurements were found As indicated by the above examples, the number of
to be more sensitive to the fault types studied than applications and applied methods, results, and result-
were the AE measurements, especially with an inner ing analyses is large. So far, no method of vibration
race fault, in which case there are more interfaces analysis has been identified as universally superior to
effectively attenuating the high-frequency vibrations. all the others, for which reason each new application
Williams et al. [75] obtained similar results when will require an evaluation of pre-selected candidate
running new rolling bearings to failure. methods prior to the selection of the most appropri-
In a review study by Tandon and Choudhury [56] ate one.
about rolling bearing fault detection, it is suggested
that out of the time domain statistical parameters, 6.1.2 Starvation
the Kurtosis value is the most effective one for defect
identification. At the initial stage of bearing failure, At a high product of rolling velocity and viscosity, the
high-frequency techniques such as the vibration oil film thickness shows a tendency to level off and
acceleration envelope analysis, the AE measurements, even decrease with increasing velocity [33]. The oil
and the shock pulse method (SPM) have been success- film thinning due to the starvation generates quick
fully applied. pressure pulses, which excite nominal frequencies of
By applying vibration analysis methods and a test the bearing structure. Using a bearing test installa-
rig, Tandon and Kumar [76] have studied bearings tion, Mba [81] has demonstrated that high-frequency
with outer races having artificially caused defects. At vibration, or the AE energy, correlates to the calculated
high rotational speed, the SPM gave a better sensitiv- film thickness in a rolling element bearing. Accord-
ity in comparison with the RMS value of the vibration ing to Miettinen et al. [82], a high base oil viscosity
velocity. in a lubricating grease leads to increased starvation, as
Tandon et al. [77] found that, with different load indicated by increased AE pulse count rates.
and defect sizes, the peak amplitude of the AE showed
the largest increase when the bearing was changed 6.1.3 Contaminated lubricants
from a healthy one to a faulty one. The SPM gave the
second-best response, while the RMS of the vibration The condition monitoring of rolling bearings with
velocity showed the lowest response in this compari- lubricants containing quartz (SiO2 ) dust has been
son. Zhen et al. [78] studied the condition monitoring studied by Maru et al. [83]. Tests with clean oil
of healthy and faulty rolling bearings, and concluded revealed that the RMS value at the high-frequency
that the SPM method is effective in estimating the (600–10 000 Hz) band was less sensitive to the load and
rolling bearing condition. speed than at the other frequencies. The RMS values
Rubini and Meneghetti [79] compared the vibration increased with the contamination particle concentra-
acceleration spectrum, the envelope and the wavelet tion up to a stabilization limit. With increasing size of
analysis methods on a double-row ball bearing with the contamination particles in the tests, the vibration
artificial raceway defects, and found that an inner level first increased until an upper particle size limit
race defect could be detected with the spectrum anal- was reached, after which the vibration level decreased,
ysis method, and that an outer race defect could probably due to a restricted entry of larger particles
be detected by the same method if the loading was into a contact with certain oil film thickness, or due to
increased. Both the envelope analysis method and the particle settling by gravity.
wavelet analysis method were successfully used for the Furthermore, Maru et al. [84] compared high-
determination of all the fault types. While the edges frequency RMS values of the vibrations from rolling
of the cracks were smoothened during repeated over- bearings lubricated with contaminated oil. With clean
rollings by rolling elements, the amplitude of not only oil, the high-frequency RMS values were higher for the
the envelope analysis but also the wavelet analysis roller bearings that were studied than for the ball bear-
results decreased. ings studied, due to a lower lambda value for the roller

JET656 Proc. IMechE Vol. 224 Part J: J. Engineering Tribology

388 J Halme and P Andersson

bearings. However, the increase ratio for the high- than did particles with a larger size. The cleaning and
frequency RMS values, when changing from clean oil re-greasing of a bearing that had been tested with
to contaminated oil, were higher for the ball bearings contaminated grease gave a reduction in the AE level,
than for the roller bearings. although the AE level remained higher than it was
Akagaki et al. [85] analysed vibration acceleration in when the bearing was new and lubricated with clean
a bearing test rig with contaminated oils, and observed grease.
that the shape of the probability density function
changed, depending on the increased particle size of
the white fused alumina contaminant. When increas- 6.2 Wear particle monitoring
ing the contaminant particle size, the wear rate of the Several methods for the detection and analysis of wear
bearing increased to a certain limit. particles in rolling bearings are available. The methods
can be divided into off-line laboratory methods and
6.1.4 Grease lubrication continuous on-line or in-line methods. Hunt [52], and
Roylance and Hunt [91], have listed most of the wear
In comparison with oil lubrication, the use of lubricat- particle detection techniques for in-line or on-line
ing greases may increase or reduce the noise in mea- analyses. In the following, a selected set of methods for
surement signals for rolling bearings. The noise level wear particle monitoring are presented and discussed.
of grease-lubricated deep-groove ball bearings rises
with the bearing diameter [86]. Work by Howard [87]
6.2.1 Composition of particles
showed that, in comparison with a light mineral oil,
all lubricating greases considered as clean gave higher Inductively coupled plasma optical emission spec-
RMS values of the velocity of the bearing-structure- trometry can be used for elementary analysis and
born vibration at a high-frequency band, from 1.8 to detection of different wear metals and their con-
10 kHz, while at a lower frequency band (50–300 Hz), centration in wear particles [92]. Atomic emission
the results were quite similar for greases and for oil. spectroscopy is suitable for an elementary analysis of
Rolling bearing tests by Wunsch [86] on the noise a large number of elements. X-ray fluorescence is used
characteristics of lubricating greases showed that the for classifying a large number of elements from a sam-
vibration peak level decreases with an increasing base ple, which is excited by bombarding with high-energy
oil viscosity. Solid lubricants, such as graphite, cal- X-rays. The energy dispersive X-ray spectrometry fea-
cium carbonate, etc., increased as well the vibration ture of a scanning electron microscope can be used for
peak level of the original bearing. Similarly, bentonite- identifying wear elements as well as for wear particle
clay-thickened and polyurea-thickened greases are sizing and characterization. For particles of more than
expected to be very noisy [88]. 1 μm in diameter, an optical light microscope can be
Work by Komiya [89] on the effect of solid contami- used for analysing the visible features of wear particles.
nation on the noise of grease-lubricated ball bearings In ferrography, ferrous particles are identified on the
showed that the vibration acceleration noise increased basis of their sedimentation on a glass sample plate by
with the increase of aluminium oxide particle size the combined effect of magnetic attraction and gravity.
with a constant concentration of 1000 ppm. The noise
increase tends to occur in the early stage of rota- 6.2.2 Size, shape, and concentration
tion and tends to decrease when the rotation time
is prolonged, which seems to be in good correlation In the traditional filter blockage technique, a flow of oil
with the corresponding oil-lubricated cases. The par- with particles passes through a fine mesh of a known
ticles of the largest size produced a lower vibration pore size, and the change in the flow characteristics
value, which is probably due to the smaller number indicates how many pores are blocked by particles
of particles for the same mass percentage. A 10 ppm larger than the pore size [52, 91]. In inductance-
contaminant particle concentration, regardless of the based detection, a coil is placed around a pipe with
particle size, had no effect on the RMS values of a flow of oil with particles, and metal particles are
the vibration acceleration, but increased the num- detected through an inductance change in the coil,
ber of vibration acceleration pulses exceeding the which is different for ferrous and non-ferrous parti-
threshold value. In general, the lower the base oil vis- cles [52, 91, 93]. Magnetic attraction of wear particles
cosity, the greater the pulse count of the vibration in oil can be used for sensing a particle flow past a
acceleration. magnetic sensor [52, 91, 93].
Miettinen and Andersson [90] investigated the influ- In ferrography analysis, oil-borne particles are sedi-
ence of solid contaminant particles as the cause for mented along a glass slide by the interaction of gravity
vibrations in grease-lubricated, statically loaded deep- and magnetic attraction. Based on the location and
groove ball bearings. The measured AE levels increased orientation, form, size, texture, and colour, wear parti-
non-linearly with the particle concentration. Small- cles and other particles can be identified under a light
sized particles generated a higher AE pulse count microscope [52, 91, 93].

Proc. IMechE Vol. 224 Part J: J. Engineering Tribology JET656

 Rolling contact fatigue and wear fundamentals for rolling bearing diagnostics 389

For the optical determination of the size distribution the probabilistic life estimations performed by cal-
of particles in oil, one option is the Fraunhofer light culations and the recorded and interpreted condition
diffraction (also known as forward scatter) technique, monitoring data, the machine operation statistics, and
by which an array of detectors ahead of a light beam, the maintenance history should be established [97].
passed through a mixture of oil and particles, detects For example, Wang and Zhang [98] have reported the
the proportion of light diffraction caused by small and results of combining oil monitoring information, the
large particles [52, 91]. time of operation, and the failure statistics of 30 air-
With the optical obscuration method, each particle craft engines for predicting the residual life of these
casts a shadow when passing through a light beam, engines.
and a photo detector measures the drop in intensity at In terms of the reliability theory, the most widely
a reference surface [52, 91, 94]. According to Roylance used distribution type calculated for the lifetime of
and Hunt [91], time-of-transition is another optical machine components is theWeibull distribution [4, 98,
technique. As a rotating, focused light beam passes 99]. On the other hand, according to the least sophisti-
some particles, the time-of-transition is measured and cated approach, the operational time is only compared
it is proportional to the particle size. to the recorded lifetime of similar components in sim-
When a flow of oil and particles is directed along a ilar operations as a mean residual life estimate, which
surface coated with a thin conducting layer, the elec- is not very reliable for an individual rolling bearing in
trical resistance across the thin film is increased as the practice [100].
coating becomes thinner due to erosion by the parti- A supervised back-propagation method has been
cles. An ultrasonic beam, which is focused into the oil, successfully tested for the diagnostics of rolling bear-
is reflected from particles in the oil, and the reflected ing defects by using vibration signatures as input
signal, which depends on the size of the discontinuity, values to a neural network [101–103]. In addition, a
is recorded. successful investigation with an unsupervised adap-
The use of wear particle examination as a diagnostic tive resonance method was reported [102]. Rolling
tool requires an estimate of the likely wear mecha- bearing degradation tests were conducted by using
nisms and the regimes in which each of the contacts vibration envelope features as input vectors for unsu-
of interest operates [41]. Condition monitoring of an pervised self-organizing maps (SOM) [104]. It was
oil-lubricated ball bearing in an accelerated bearing shown that by comparing the minimum quantization
life test was studied by Halme [95], who concluded error of new data to the SOM trained with healthy data,
that during a step from the steady state to the final the results were more sensitive for the identification
phase of the life of a ball bearing, the most remarkable of incipient bearing failure than the statistical vibra-
acceleration in a wear process was detected through tion acceleration RMS parameter would have been.
the measured vibration acceleration responses, as well A supervised back-propagation neural network has
as through the amount of relatively large particles in been evaluated for the determination of the degrada-
the oil. tion time and finally the weight application technique
Using a bearing test rig, Harvey et al. [96] stud- for residual life prediction for ball bearings [105]. The
ied the wear monitoring of a lubricated taper roller results reported showed a better prediction in com-
bearing with an artificial fault on the inner race. An parison with the traditional L10 calculation technique.
accelerated test was run until a total damage occurred, However, despite the reported promising results of
and the wear particle responses were measured with rolling bearing diagnostics obtained with different
electrostatic sensors and the vibrations were mea- kinds of neural networks, it should be noticed that in
sured with acceleration sensors. When about 10 per most cases, the studies have been carried out in a lab-
cent of the calculated lifetime of the bearing was left, oratory environment where the data are sufficient in
a first steady increase in the measured electrostatic terms of the amount and one-dimensional in terms of
responses was seen. When about 6 per cent of the life- the application. In many real industrial applications,
time was left, both the electrostatic response and the the situation is totally different and the generalization
vibration amplitude of the inner race fault frequency of the results and neural network models may not be
increased. When about 99 per cent of the calculated justified. Furthermore, this limitation in the applica-
lifetime of the bearing had elapsed, a sharp increase bility applies somewhat to rolling bearing diagnostics
occurred in all the measured parameters. based on fuzzy logics. As an example, it is reported
that, by using vibration features in a test rig, the fuzzy
reasoning gave 100 per cent accuracy in distinguishing
6.3 Combining calculated rating life and between healthy rolling bearings and faulty ones [106].
diagnostics information for refined life For an initial prediction of the rating life for a rolling
expectancy estimates bearing, the statistical methods like the ISO 281 calcu-
lation procedure are the most reliable ones. During
Because of the uncertainty in the determination of the operation of the bearing, the prognosis for the
the rating life of rolling bearings, a link between remaining operational lifetime can be updated and

JET656 Proc. IMechE Vol. 224 Part J: J. Engineering Tribology

390 J Halme and P Andersson

ensured by employing a combination of the calculated condition monitoring measurements can be com-
rating life estimate and condition monitoring mod- bined with a theoretical rating life estimate, in order
els that utilize the progress or trend in the recorded to give an updated and more reliable prognosis for the
and interpreted condition monitoring parameters. remaining operational lifetime of the bearing.
The process for updating the lifetime prognosis and
the order in which the condition monitoring meth-
ACKNOWLEDGEMENTS
ods are applied is important; in vibration analyses of
rolling bearings, the initiation of sub-surface cracks
The authors are grateful for the financial support
and minor metal-to-metal contacts usually generate
from the European Commission, the Sixth Framework
high-frequency vibrations, which can be monitored
programme for Research and Technological Devel-
with AE sensors. When the first sub-surface cracks
opment, under which the present work was carried
reach the surfaces, or when surface cracks are ini-
out as part of Work Package FP6 of the Integrated
tiated, they have sharp edges which cause impacts
Project IP017498 DYNAMITE ‘Dynamic Decisions in
between the rolling surfaces of the bearings and excite
Maintenance’. Research Professor Kenneth Holmberg
the nominal frequencies of the bearing, or vibra-
of VTT is greatly acknowledged for proposing the
tions of typically 500–3000 Hz frequency. The nominal
present study to be carried out and for encouraging
frequencies can be detected by using the vibration
discussions during the progress of the work.
acceleration envelope analysis described in section
6.1. Finally, when the sharp edges are flattened and
© Authors 2010
wider cracks and pits are generated, the nominal bear-
ing defect frequencies such as BPFO, BPFI, BSF, or FTF
(see section 6.1) become detectable from the vibration REFERENCES
velocity spectrum. The high overall vibrations can be
measured when the bearing is approaching its final 1 Harris, T. A. and Barnsby, R. M. Life ratings of roller
failure. bearings. Proc. IMechE, Part J: J. Engineering Tribology,
2001, 215, 577–595. DOI: 10.1243/1350650011543817.
2 Jacobson, B. The Stribeck memorial lecture. Tribol. Int.,
7 CONCLUSIONS 2003, 36, 781–789.
3 Zaretsky, E. V. A. Palmgren revisited – a basis for bearing
life prediction. Lubr. Eng., 1998, 54, 18–23.
Friction, wear, and lubrication mechanisms acting at
4 Weibull, W. A statistical theory of the strength of mat-
the rolling and sliding contacts are central issues for
erials. In Proceedings of the Royal Swedish Academy of
the operation of rolling contact bearings, and consid- Engineering Sciences, Stockholm, Sweden, 1939, 45 pp.,
ered in rating life calculations for bearings at specific vol. 151.
operational conditions. 5 Lundberg, G. and Palmgren, A. Dynamic capacity of
During bearing operation, rolling surface rough- rolling bearings. In Proceedings of the Royal Swedish
ening and wear particle formation may eventually Academy of Engineering Sciences, Stockholm, Sweden,
occur as a consequence of rolling contact wear and 1947, 50 pp., vol. 196.
rolling contact fatigue, unless the operational condi- 6 Lundberg, G. and Palmgren, A. Dynamic capacity of
tions are mild enough to limit the surface alterations roller bearings. In Proceedings of the Royal Swedish
to polishing. Academy of Engineering Sciences, Stockholm, Sweden,
1952, 32 pp., vol. 210.
Rolling contact fatigue, in particular, causes changes
7 Dowson, D. History of tribology, 1998 (Professional
in the bearing kinematics by rolling surface rough- Engineering Publishing Limited, London, UK).
ening and wear particle formation, which in turn 8 Shimizu, S. Fatigue limit concept and life prediction
increase the bearing vibrations. The vibrations and the model for rolling contact machine elements. Tribol.
wear particle formation can be monitored during the Lubr. Technol., 2005, 61, 32–41.
operation of the bearing. 9 International standard ISO 281:2007. Rolling bearings –
Useful methods for vibration-based condition mon- dynamic load ratings and rating life. International Orga-
itoring of rolling bearings are the vibration accelera- nization for Standardization ISO, Geneve, Switzerland,
tion high-frequency techniques, the SPM, the enve- 2007.
lope analysis, and the AE analysis for obtaining an 10 Johnson, K. Contact mechanics, 1985 (Cambridge Uni-
early warning. For the detection of a severe bearing versity Press, Cambridge, UK).
11 Williams, J. and Dwyer-Joyce, R. S. Contact between
defect, the most useful methods are the spectrum
solid surfaces. In Modern tribology handbook (Ed. B.
analysis of the vibration velocity amplitudes at nomi- Bhusan), 2001 (CRC Press LLC, Boca Raton, Florida,
nal bearing frequencies, the determination of the RMS USA).
value of the vibration velocity, and the analysis of 12 Harris, T. A. and Kotzalas, M. N. Essential concepts of
the statistical parameters of the vibration accelera- bearing technology. In Rolling bearing analysis, 5th edi-
tion such as the Kurtosis and RMS values. Measured tion, 2007 (CRC Press, Taylor & Francis Group, Boca
and interpreted vibration and wear particle data from Raton, Florida, USA).

Proc. IMechE Vol. 224 Part J: J. Engineering Tribology JET656

 Rolling contact fatigue and wear fundamentals for rolling bearing diagnostics 391

13 Shigley, J. E. and Mischke, C. R. Mechanical engineering 32 Dowson, D. and Higginson, G. Elasto-hydrodynamic

design, 5th edition, 1989 (McGraw-Hill, Inc., New York, lubrication, edition SI, 1977 (Pergamon press, UK).
USA). 33 Chiu, Y. An analysis and prediction of lubricant film
14 Bowden, F. P. and Tabor, D. The friction and lubrication starvation in rolling contact systems. ASLE Trans., 1974,
of solids. Part I, 1950/1971 (Clarendon Press, Oxford, 17, 22–35.
UK). 34 Cann, P. M. E., Damiens, B., and Lubrecht, A. A. The
15 Egle, D. Wave propagation. In Nondestructive testing transition between fully flooded and starved regimes in
handbook (Ed. R. Miller), 2nd edition, 1987 (American EHL. Tribol. Int., 2004, 37, 859–864.
Society for Nondestructive Testing, USA). 35 Nikas, G., Sayles, R., and Ionnodes, E. Effects of debris
16 Hopwood, T. Acoustic emission applications in civil in sliding/rolling elastohydrodynamic contacts. Proc.
engineering. In Nondestructive testing handbook (Ed. IMechE, Part J: J. Engineering Tribology, 1998, 212,
R. Miller), 2nd edition, 1987 (American Society for 333–343. DOI: 10.1243/135065098154216.
Nondestructive Testing, USA). 36 Cann, P. M. Starved grease lubrication of rolling con-
17 International standard ISO 7626-5:1994. Vibration and tacts. Tribol. Trans., 1999, 42, 867–873.
shock – experimental determination of mechanical 37 Khonsari, M. M., Pascovici, M. D., and Kucinschi, B. V.
mobility. International Organization for Standardiza- On the scuffing failure of hydrodynamic bearings in the
tion ISO, Geneve, Switzerland, 1995. presence of an abrasive contaminant. Trans. ASME, J.
18 Åström, K. J. and Wittenmark, B. Computer controlled Tribol., 1999, 121, 90–96.
systems: theory and design, 3rd edition, 1997 (Prentice 38 Dwyer-Joyce, R. S. Predicting the abrasive wear of ball
Hall Information and System Sciences Series, Upper bearings by lubricant debris. Wear, 1999, 233–235, 692–
Saddle River, New Jersey, USA). 701.
19 Åström, H., Isaksson, O., and Höglund, E.Video record- 39 Sayles, R. and Ionnides, E. Debris damage in rolling
ings of an EHD point contact lubricated with grease. bearings and its effects on fatigue life. J. Tribol., 1988,
Tribol. Int., 1991, 24, 179–184. 110, 26–31.
20 Dowson, D. Elastrohydrodynamic and micro-elasto- 40 Nilsson, R., Svahn, F., and Olofsson, U. Relating con-
hydrodynamic lubrication. Wear, 1995, 190, 125–138. tact conditions to abrasive wear. Wear, 2006, 261,
21 Cann, P. M., Spikes, H. A., and Hutchinson, J. The devel- 74–78.
opment of a spacer layer imaging method (SLIM) for 41 Williams, J. Wear and wear particles – some fundamen-
mapping elastohydrodynamic contacts. Tribol. Trans., tals. Tribol. Int., 2005, 30, 863–870.
1996, 39, 915–921. 42 Kato, K. and Adachi, K. Wear mechanisms. In Modern
22 Spikes, H. A. and Olver, A.V. Basics of mixed lubrication. tribology handbook (Ed. B. Bhusan), 2001 (CRC Press
Lubr. Sci., 2003, 16(1), 1–28. LLC, Boca Raton, Florida, USA).
23 Hamrock, B. and Dawson, D. Ball bearing lubrication, 43 Olver, A. The mechanism of rolling contact fatigue: an
1981 (John Wiley, New York, USA). update. Proc. IMechE, Part J: J. Engineering Tribology,
24 Hamrock, B. Fundamentals of fluid film lubrication, 2005, 219, 313–330. DOI: 10.1243/135065005X9808.
1994 (McGraw-Hill, Singapore). 44 Prashad, H. Appearance of craters on track surface of
25 Jacobson, B. Thin film lubrication of real surfaces. rolling element bearings by spark erosion. Tribol. Int.,
Tribol. Int., 2000, 33, 205–210. 2001, 34, 39–47.
26 Hsu, S. M. and Gates, R. S. Boundary lubrication and 45 Jagenbrein, A., Buschbeck, F., Gröschl, M., and
boundary lubrication films. In Modern tribology hand- Preisinger, G. Investigation of the physical mecha-
book (Ed. B. Bhusan), 2001 (CRC Press LLC, Boca Raton, nisms in rolling bearings during the passage of electric
Florida, USA). current. Tribotest, 2005, 11, 295–306.
27 Hsu, S. M., Munro, R., and Shen, M. C. Wear in 46 Imran, T., Jacobson, B., and Shariff, A. Quantifying
boundary lubrication. Proc. IMechE, Part J: J. Engi- diffused hydrogen in AISI 52100 bearing steel and in
neering Tribology, 2002, 216, 427–441. DOI: 10.1243/ silver steel under tribo-mechanical action: pure rotat-
135065002762355343. ing bending, sliding-rotating bending, rolling–rotating
28 Jacobson, B. Mixed lubrication. Wear, 1990, 136, bending and uni-axial tensile loading. Wear, 2006, 261,
99–116. 86–95.
29 Lugt, P. M., Severt, R. W. M., Fogelström, J., and Tripp, 47 Zaretsky, E. V., Poplawski, J. V., and Miller, C. R. Rolling
J. H. Influence of surface topography on friction, film bearing life prediction – past, present and future.
breakdown and running-in in the mixed lubrication Hanover, Maryland, USA, 2000, National Aeronautics
regime. Proc. IMechE, Part J: J. Engineering Tribology, and Space Administration, NASA/TM-2000-210529.
2001, 215, 519–533. DOI: 10.1243/1350650011543772. 48 Ioannides, E. Contamination and the SKF new life the-
30 Lord, J. and Larsson, R. Effects of slide-roll ratio and ory for rolling bearings. NLGI Spokesman, 1990, 54,
lubricant properties on elastohydrodynamic lubrica- 14–19.
tion film thickness and traction. Proc. IMechE, Part 49 Hamer, J., Sayles, R., and Ioannides, E. Particle defor-
J: J. Engineering Tribology, 2001, 215, 301–308. DOI: mation and counterface damage when relatively soft
10.1243/1350650011543556. particles are squashed between hard anvils. Tribol.
31 Szeri, A. Z. Hydrodynamic and elastohydrodynamic Trans., 1989, 32, 281–288.
lubrication. In Modern tribology handbook (Ed. B. 50 Tasan, Y., Rooji, M., and Schipper, D. Changes in the
Bhusan), 2001 (CRC Press LLC, Boca Raton, Florida, micro-geometry of a rolling contact. Tribol. Int., 2007,
USA). 40, 672–679.

JET656 Proc. IMechE Vol. 224 Part J: J. Engineering Tribology

392 J Halme and P Andersson

51 Olofsson, U., Andersson, S., and Björklund, S. Simu- 68 Al-Dossary, S., Hamzah, R., and Mba, D. Changes in
lation of mild wear in boundary lubricated spherical acoustic emission waveform for varying seeded defect
roller thrust bearings. Wear, 2000, 241, 180–185. sizes. In Proceedings of the International Congress
52 Hunt,T. M. Handbook of wear debris analysis and parti- on Condition Monitoring and Diagnostic Engineer-
cle detection in liquids, 1993 (Elsevier Applied Science, ing Management – Comadem 2006 (Eds U. Kumar, A.
London, UK). Parida, and R. B. Rao), Luleå University of Technology,
53 Choi,Y. and Liu, C. Spall progression life for rolling con- Luleå, Sweden, 2006.
tact verified by finish hard machined surfaces. Wear, 69 Choy, F., Zhou, J., Braun, M., and Wang, L. Vibra-
2007, 262, 24–35. tion monitoring and damage of faulty ball bearings. J.
54 Day, M. J. Condition monitoring of hydraulic systems. Tribol., 2005, 127, 776–783.
In Handbook of condition monitoring (Ed. B. K. N. Rao), 70 McFadden, P. D. and Toozhy, M. M. Application of syn-
1996 (Elsevier Advanced Technology, Oxford, UK). chronous averaging to vibration monitoring of rolling
55 Ehrich, F. E. Handbook of rotordynamics, 1992 element bearings. Mech. Syst. Signal Process., 2000, 14,
(McGraw-Hill, New York, USA). 891–906.
56 Tandon, N. and Choudhury, A. A review of vibration 71 Ericsson, S., Gripa, N., Johansson, E., Persson, L.,
and acoustic measurement methods for the detection Sjöberg, R., and Strömberg, J. Towards automatic
of defects in rolling element bearings. Tribol. Int., 1999, detection of local bearing defects in rotating machines.
32, 469–480. Mech. Syst. Signal Process., 2005, 19, 509–535.
57 Orhan, S., Aktürk, N., and Çelik,V.Vibration monitoring 72 Jamaludin, N. and Mba, D. Monitoring extremely slow
for detect diagnosis of rolling element bearings as a pre- rolling element bearings: part I. NDT&E Int., 2002, 35,
dictive maintenance tool: comprehensive case studies. 349–358.
NDT&E Int., 2006, 39, 293–298. 73 Jamaludin, N. and Mba, D. Monitoring extremely slow
58 Randall, B., Gao, Y., and Ho, D. Enhanced envelope rolling element bearings: part II. NDT&E Int., 2002, 35,
analysis of bearing signal using digital techniques. In 359–366.
Proceedings of the Comadem ’97 – Tenth Interna- 74 Shiroishi, J., Li,Y., Liang, S., Kurfess,T., and Danyluk, S.
tional Congress and Exhibition on Condition moni- Bearing condition diagnostics via vibration and acous-
toring and diagnostic engineering management (Eds tic emission measurements. Mech. Syst. Signal Process.,
E. Jantunen, K. Holmberg, and B. K. N. Rao), 1997 1997, 11, 693–705.
(Technical Research Centre of Finland (VTT), Espoo, 75 Williams, T., Ribadeneira, X., Billington, S., and Kur-
Finland). fess, T. Rolling element bearing diagnostics in run-to-
59 Randall, B. Detection and diagnosis of incipent bearing failure lifetime testing. Mech. Syst Signal Process., 2001,
failure in helicopter gearboxes. Eng. Failure Anal., 2004, 15, 979–993.
11, 177–190. 76 Tandon, N. and Kumar, K. S. Detection of defects at dif-
60 Ho, D. and Randall, B. Optimisation of bearing diag- ferent locations in ball bearings by vibration and shock
nostic techniques using simulated and actual bear- pulse monitoring. Vibr. Shock Pulse Monit., 2003, 9–16.
ing fault signals. Mech. Syst. Signal Process., 2000, 14, 77 Tandon, N., Yadava, G. S., and Ramakrishna, K. M.
763–788. A comparison of some condition monitoring tech-
61 Kiral, Z. and Karagülle, H. Simulation and analysis of niques for the detection of defect in induction motor
vibration signals generated by rolling element bearings ball bearings. Mech. Syst. Signal Process., 2007, 21,
with defects. Tribol. Int., 2003, 36, 667–678. 244–256.
62 Kiral, Z. and Karagülle, H. Vibration analysis of rolling 78 Zhen, L., Zhengjia, H., Yanyang, Z., and Xuefeng, C.
element bearings with various defects under the action Bearing condition monitoring based on shock pulse
of an unbalanced force. Mech. Syst. Signal Process., 2006, method and improved redundant lifting scheme. Math.
20, 1967–1991. Comput. Simul., 2008, 79, 318–338.
63 Heng, R. B. W. and Nor, M. J. M. Statistical analysis 79 Rubini, R. and Meneghetti, U. Application of the enve-
of sound and vibration signals for monitoring rolling lope and wavelet transform analyses for the diagnosis
element bearing condition. Appl. Acoust., 1998, 53, of incipent faults in ball bearings. Mech. Syst. Signal
211–226. Process., 2001, 15, 287–302.
64 Abdullah, M. and Mba, D. A comparative experimen- 80 Prabhakar, S., Monathy, A. R., and Sekhar, A. S.
tal study on the use of acoustic emission and vibration Application of discrete wavelet transform for detec-
analysis for bearing defect identification and estima- tion of ball bearing race faults. Tribol. Int., 2002, 35,
tion of defect size. Mech. Syst. Signal Process., 2006, 20, 793–800.
1537–1571. 81 Mba, D. Acoustic emission generated from bearing
65 Li, Y., Billington, S., Zhang, C., Kurfess, T., Danuluk, S., operational parameters. In Proceedings of the Interna-
and Liang, S. Dynamic prognostics prediction of defect tional Congress on Condition Monitoring and Diagnos-
propagation on rolling element bearings. Tribol. Trans., tic Engineering Management – Comadem 2006 (Eds U.
1999, 42, 385–392. Kumar, A. Parida, and R. B. Rao), Luleå University of
66 Khemili, I. and Chouchane, M. Detection of rolling ele- Technology, Luleå, Sweden, 2006.
ment bearing defects by adaptive filtering. Eur. J. Mech. 82 Miettinen, J., Andersson, P., and Wikström, V. Analy-
A, Solids, 2005, 24, 293–303. sis of grease lubrication of a ball bearing using acoustic
67 Li, C. J. and Li, S. Y. Acoustic emission analysis emission measurement. Proc. IMechE, Part J: J. Engi-
for bearing condition monitoring. Wear, 1995, 185, neering Tribology, 2001, 215, 535–544. DOI: 10.1243/
67–74. 1350650011543781.

Proc. IMechE Vol. 224 Part J: J. Engineering Tribology JET656

 Rolling contact fatigue and wear fundamentals for rolling bearing diagnostics 393

83 Maru, M., Castillo, R., and Padovese, L. Study of solid of Western Australia, 2–5 December 2002, vol. II, pp.
contamination in ball bearings through vibration and 549–553.
wear analyses. Tribol. Int., 2007, 40, 433–440. 96 Harvey, T. J., Wood, R. J. L., and Powrie, H. E. G. Elec-
84 Maru, M., Castillo, R., and Padovese, L. Influence of oil trostatic wear monitoring of rolling element bearings.
contamination on vibration and wear in ball and roller Wear, 2007, 263, 1429–1501.
bearings. Ind. Lubr. Tribol., 2007, 59, 137–142. 97 Vlok, P.,Wnek, M., and Zygmunt, M. Utilising statistical
85 Akagaki, T., Nakamura, M., Monzen, T., and Kawabata, residual life estimates of bearings to quantify the influ-
M. Analysis of the behaviour of rolling bearings in con- ence of preventive maintenance actions. Mech. Syst.
taminated oil using some condition monitoring tech- Signal Process., 2004, 18, 833–847.
niques. Proc. IMechE, Part J: J. Engineering Tribology, 98 Wang, W. and Zhang, W. A model to predict the resid-
2006, 220, 447–453. DOI: 10.1243/13506501J00605. ual life of aircraft engines based upon oil analysis data.
86 Wunsch, F. Noise characteristic of lubricating greases Naval Res. Log., 2005, 52, 276–284.
used for anti-friction bearings. NLGI Spokesman, 1992, 99 Råde, L. and Westergren, B. Beta β mathematics hand-
56, 16–21. book, 1990 (Studentlitteratur, Lund, Sweden).
87 Howard, P. Noise testing of rolling element bearings. 100 Vlok, P. Dynamic residual life estimation of industrial
NLGI Spokesman, 1975, 39, 54–60. equipment based on failure intensity proportions. Doc-
88 Ward, C. Practical aspects of grease noise testing. NLGI toral Thesis, Department of Industrial and Systems
Spokesman, 1994, 58, 8–12. Engineering, University of Pretoria, Pretoria, South
89 Komiya, H. Effect of contaminant in lubricant on Africa, 2001.
noise of ball bearings. NLGI Spokesman, 1992, 56, 101 Alguindigue, I. E., Loskiewicz-Buczak, A., and Uhrig,
9–16. R. E. Monitoring and diagnosis of rolling element bear-
90 Miettinen, J. and Andersson, P. Acoustic emission of ings using artificial neural networks. IEEE Trans. Ind.
rolling bearings lubricated with contaminated grease. Electron., 1993, 40, 209–217.
Tribol. Int., 2000, 33, 777–787. 102 Subrahmanyam, M. and Sujatha, C. Using neural net-
91 Roylance, B. J. and Hunt, T. M. The wear debris ana- works for the diagnosis of localized defects in ball
lysis handbook, 1999 (Coxmoor Publishing Company, bearing. Tribol. Int., 1997, 30, 739–752.
Oxford, UK). 103 Samantha, B. and Al-Balushi, K. R. Artificial neural net-
92 Evans, J. and Hunt, T. Oil analysis, 2004 (Coxmoor work based fault diagnostics of rolling element bearings
Publishing Company, Oxford, UK). using time-domain features. Mech. Syst. Signal Process.,
93 Michael Neale and Associates Ltd A guide to the con- 2003, 17, 317–328.
dition monitoring of machinery, 1979 (Department of 104 Qiu, H., Lee, J., Lin, J., and Yu, G. Robust performance
Industry, Committee for Terotechnology, London, UK). degradation assessment methods for enhanced rolling
94 Toms, L. and Toms, A. Lubricant properties and test element bearing prognostics. Adv. Eng. Inform., 2003,
methods. In Handbook of lubrication and tribology (Ed. 17, 127–140.
G. Totten), vol. 1: application and maintenance, 2006, 105 Huang, R., Xi, L., Li, X., Liu, R., Qiu, H., and Lee, J.
pp. 30.1–30.33 (Taylor & Francis Group, Boca Raton, Residual life predictions for ball bearings based on self-
Florida, USA). organizing map and back propagation neural network
95 Halme, J. Condition monitoring of oil lubricated ball methods. Mech. Syst. Signal Process., 2007, 21, 193–207.
bearing using wear debris and vibration analysis. In 106 Liu, T. I., Singonahalli, J. H., and Iyer, N. R. Detection
Proceedings of the International Tribology Conference of roller bearing detects using expert system and fuzzy
(AUSTRIB’02), Frontiers in tribology, Perth, University logic. Mech. Syst. Signal Process., 1996, 10, 595–614.

JET656 Proc. IMechE Vol. 224 Part J: J. Engineering Tribology

View publication stats