International Journal of Advances in Engineering Sciences Vol.
1, Issue 2, April, 2011
Fatigue Failure Analysis of
Hollow Cylindrical Roller Bearing
Prof. P. H Darji Dr. D. P. Vakharia
Professor, Dept. of Mech. Engg., Professor, Dept. of Mech. Engg.
C. U. Shah College of Engg. & Tech., S. V. National Institute of Technology,
Surendranagar, India Surat, India
pranav_darji@rediffmail.com dpvakharia@yahoo.com
AbstractCylindrical roller bearings with hollow rollers are DN values as high as 3-4 million may be required in the near
advantageous in applications where weight, lubrication and future. At high speeds, balls/rollers in a bearing rotate rapidly.
speed are major considerations in the operations of the bearings. High centrifugal loads increase the Hertzian stresses above the
An improvement in load distribution and thus load capacity may level from applied loads and can reduce significantly bearing
be realized, as well as contact stress is also reduced considerably
fatigue life [1].
by using a bearing with hollow rollers. Since from the
implementation, for hollow roller bearing no method is available
for the calculation of hollowness and contact stresses and II CONTACT MECHANICS OF HOLLOW CYLINDRICAL ROLLER
ultimately premature failure will occur without performing its
optimum life. The contact stresses in hollow members are often calculated
by using the same equations and procedures as for solid
In the present work failure analysis of Hollow Cylindrical Roller specimens. This approach seems to be incorrect. The Hertzian
Bearing (HCRB) made from AISI 52100 is carried out to theory of contact is based on several assumptions. One is that
understand the type and mode of failure. This bearing is basically the profiles of the two bodies are continuous and can be
used in earthmoving equipments where heavy load is applied on
bearing. The objective of the work is to identify the type of
represented to good approximation by a polynomial of second
failure by using material oriented investigations. It was found degree. Hertzian theory does not take into account a situation
that flexural fatigue is the mode of failure occurs in this bearing where the cross sections of either or both of the bodies in
the reason for which is wrong selection of hollowness. contact are multiply connected. Fig. 1 shows solid and hollow
cylinders of the same outside diameter compressed between
two platens and subjected to a load F. The diametral
Keywords-Hollow cylindrical roller bearing, Rolling Contact compression for the hollow cylinder will be greater than that
Fatigue, Failure analysis, Flexural fatigue. for the solid cylinder. The cylinder surface in contact with the
platens will be a rectangle and its width for the hollow
I. INTRODUCTION cylinder will be greater than that for the solid cylinder. This is
With increasing demands for higher speed and load of shown schematically in Fig. 1. It can be predicted a priori that
bearings, Hollow Cylindrical Roller Bearing (HCRB) find the maximum contact stress is less for a hollow cylinder than
wider applications. Hollow cylindrical roller bearing performs for a solid cylinder. The setting-up of mathematics expressions
better than solid bearing. The survey of existing work and derivation of a closed-form generalized solution for the
associated with hollow cylindrical roller bearing is carried out reduced contact stress in a hollow cylinder is a formidable
in this paper and reviewed by two main aspects : analytical problem.
and experimental. The recent advancements in HCRB are also In addition to the contact stresses at the outer contact zone, the
briefly introduced. This paper is a detailed survey of the hollow specimens are subjected to tangential stresses (bending
current studies of HCRB, most of which were published in the stress) whose values are at a maximum on the inner fiber.
last fifteen years. The authors attempt to review the analytical Failure of the roller could be due to either pitting fatigue or
and experimental work regarding contact problems of HCRB. flexural fatigue. Flexural fatigue failures occurred with the
Technological progress creates increasingly arduous AISI 52100 hollow roller when the maximum tangential
conditions for rolling mechanisms. Advances in many fields tensile stress at the bore is 684 N/mm2 [2]. The fatigue cracks
including aeronautics, space and atomic power, involve always began at the bore of the hollow roller. Those that
extreme operating speeds, temperatures and environments, and propagated to the roller surface resulted in surface crack and
demand bearings of improved properties with better strength spalling. The critical stress for failure due to flexural fatigue is
to weight ratios. The trend in gas turbine engines is for larger the tangential stress beneath the point of load application (Fig.
shaft diameters and higher main shaft and auxiliary speeds. 1, point a). This stress is tensile in nature and changes sign,
Current aircraft turbine engines operate in the range of 1.5-2 becoming compressive at point b. Fig. 2 shows the comparison
million DN (bearing bore in mm times shaft speed in rpm). of fatigue strength with ultimate and yield strength [3].
22
International Journal of Advances in Engineering Sciences Vol.1, Issue 2, April, 2011
Improved surface finishes of the inner bore may allow an A starting point in understanding the mechanics of bearing
increase in the limit of the bending stress. However, even with fatigue is the concept of sub-surface stress due to contact of
optimized surface finishes, flexural failures are still expected curved surfaces. Although several potential failure
to occur in thin-wall hollow components when they are mechanisms such as denting, false brinnelling or a host of
subjected to high loads [4]. others can occur in a bearing application, classical material
fatigue cracking typically develops at or beneath the surface of
the load bearing and their contacting raceways as shown in Fig
3 [5]. Rolling contact fatigue is manifested as a flaking off of
metallic particles from the surface of the raceways and/or
rolling elements. For well lubricated, properly manufactured
bearings, this flaking usually commences as a crack below the
surface and is propagated to the surface, eventually forming a
pit or spall in the load carrying surface. Lundberg et al.
postulated that it is the maximum orthogonal shear stress 0
that initiates the crack and that this shear stress occurs at depth
z0 below the surface [6, 7].
Lundberg and Palmgren further postulated that it is the range
of the maximum orthogonal shear stress 0, that is, 20 that
Figure 1. (a) Solid (b) Hollow cylinder under compression with contact area initiates the crack. According to Fig. 3, the depth below the
raceway surface at which 0 occurs is approximately 0.786b
[8] for typical roller bearings. As indicated above, the
microstructural alterations associated with the butterflies and
fatigue cracking tends to occur at a depth 50% greater than 0.
Nevertheless, the LundbergPalmgren theory and subsequent
development of the standard load and life rating formulas are
based on the maximum orthogonal shear stress 0 and the
depth at which it occurs.
Figure 2. Comparison of fatigue strength with Ultimate and Yield strength [3]
III ROLLING CONTACT FATIGUE
In making fatigue endurance considerations with high strength
rolling bearing steel for use as a potential material for
structural applications, it is worth examining the wealth of
experience found from the theory and practice of using this
steels in their native application. Thus, a short background Figure 3. Subsurface fatigue cracking [5]
regarding the mechanics of bearing fatigue, the materials used After nital acid etching of the surface of the sectioned
and some common key effects on the fatigue life of roller component, the butterfly wings appear white in contrast to the
bearings are provided [5]. surrounding matrix of martensite as shown in Fig 4 [3].
It has been considered that if a rolling bearing in service is
properly lubricated, properly aligned, kept free of abrasives,
moisture and corrosive reagents and properly loaded, then all
causes of damage are eliminated save one, material fatigue.
Historically rolling bearing theory postulated that no rotating
bearing can give unlimited service, because of the probability
of fatigue of the surfaces in rolling contact. As discussed in
section II failure of HCRB is also because of endurance limit
of the material. This endurance limit is a level of cyclically
applied, reversing stress, which if not exceeded, the structure
will accommodate without fatigue failure. The endurance limit
for structural fatigue has been established by rotating beam
and/or torsional testing of simple bars for various materials,
which is given in [2].
Figure 4. Fracture study showing planes of weakness along white areas
23
International Journal of Advances in Engineering Sciences Vol.1, Issue 2, April, 2011
Lundberg et al. postulated that fatigue cracking commences at foundation that serves as a basis for continuing analytical and
weak points below the surface of the material. Hence changing experimental investigations into the fundamental mechanisms
the chemical composition, metallurgical structure and involved in bearing failure.
homogeneity of the steel can significantly affect the fatigue Materials oriented investigations have been primarily
characteristics of a bearing, all other factors remaining the phenomenological in nature that is attempts have been made to
same. In referring to weak points, one does not include carefully control test conditions and to correlate material
macroscopic slag inclusions, which cause imperfect steel for response characteristics with test conditions and bearing
bearing fabrication and hence premature failure. Fig. 4 shows performance. Subsequent analysis of the metallurgical
a fracture failure at weak points developed during rolling. This mechanisms involved, are prerequisites for establishing
type of experimental study tends to confirm the Lundberg- reliable analytical models for predicting performance.
Palmgren theory insofar as those at a specified depth below
the rolling contact surface. The basis of the high load carrying capability and reliable
function of rolling bearings is the steel from which they are
IV FAILURE ANALYSIS OF HCRB produced. From a mechanical view point, bearing grade steel
are mainly characterized by high hardness (HRC 58-65) and
In the present work failure analysis of Hollow Cylindrical very high tensile strength. Excellent wear properties and
Roller Bearing (HCRB) made from material AISI 52100 is rolling contact fatigue endurance with reasonable impact
carried out to understand the type and mode of failure. This resistance can also be expected. It is the final manufactured
bearing is basically used in earthmoving equipments where microstructure of the steel that ultimately provides the
heavy load is applied on bearing. The photograph of this necessary mechanical properties desired or rather required. In
bearing is shown in Fig. 5 with hollow roller in breakage the present work we will mainly consider high carbon through
condition. The objective of this work is to identify the type of hardening alloyed steel AISI 52100 or its equivalents, as it is
failure by using material oriented investigations. Also Fig. 6 still the most widely used bearing grade steel across the
shows the cross section of hollow roller with its dimensions. industry.
The hollowness of this roller is 42%. The bearing quality steels are characterized by high hardness,
very high static strength, fine grained microstructures and low
non-metallic inclusion content. Now considering the
microstructure from the metallurgic point of view, bearing
grade steel microstructures in an optimal condition as
characterized by
1. Tempered martensite microstructure
2. Fine and uniform distribution of carbides
3. High degree of cleanliness, i.e. tightly controlled
non-metallic inclusion content.
Above three points are discussed in details for the present case
of hollow roller in the next section. Fig. 7 shows the
photograph of half part of hollow specimen which has been
polished and etched from the failure portion. Optical
microscopy studies were carried out to investigate the nature
Figure 5. Hollow Cylindrical Roller Bearing of failure.
Figure 6. Dimensions of cross section of hollow roller
Figure 7. Specimen of hollow roller made from AISI 52100
A. Metallurgical Analysis and Its Effect on Fatigue Failure
(1) Tempered Martensite Microstructure
Much work has been devoted to characterizing alterations of The microstructure of tempered martensite of this material is
the micro-structure, hardness and residual stress distribution shown in Fig. 8 This microstructure is developed by using 4%
occurring in rolling bearing components during bearing Nital as an etchant and Micrograph is obtained by 1000X
operations. The results of this work provide a technological magnification in optical microscope (As per ASTM Vol 7).
24
International Journal of Advances in Engineering Sciences Vol.1, Issue 2, April, 2011
Micrograph shows the uniform distribution of carbides. subsurface of bearing steels. The microstructural changes
Tampered hardness of this material is 61 HRC measured on could result in fatigue cracks nucleation even when the applied
Rockwell Hardness Testing machine. So the desired properties contact loading is well below to the limits causing
of AISI 52100 material are maintained by the material which deterioration of material strength [9].
is used for hollow roller.
As the elastic continuum beneath the surface of a bearing
material is stressed by the contact pressure any discontinuity
in the otherwise homogenous structure will result in even
higher stress or a stress riser around that discontinuity. Local
plastic deformation is another likely result around such a stress
riser. A non metallic inclusion in an otherwise highly clean
microstucture is one such discontinuity that naturally exists in
bearing steels, even after the manufacturing efforts mentioned.
Fatigue failure may be initiated by an exogenous inclusion in
the bearing steel. Fig. 10 shows the Micrograph of nonmetallic
inclusion. As per ASTM E-45 and 100X magnification the
inclusions are found within the limit i.e. very small in number
and size.
4% nital 1000X
Figure 8. Micrograph of fine tempered martensite microstructure
(2) Fine and Uniform Distribution of Carbides
With a fine grain size, it is expected that the 52100 should
exhibit high static strength. Also finer grains i.e. ASTM 8 or
finer, in ferrous metals produce higher static and fatigue
strengths. Present material of hollow roller shows ASTM 8
grain size. Fig 9 shows the Carbide banding as per IS 4398:94
or Carbide Segregation as per SEP 1520 by 100X
magnification in optical microscope. Carbide distribution is
uniform in the microstructure which is also an optimal
condition. As polished (not etched) 100X
Figure 10 Optical Micrograph of nonmetallic inclusion
(4) Optical Microscopy Analysis
From the above discussion and micrograph it is very clear that
there is no any kind of metallurgical defect in the material of
hollow roller. Thus the material is defect free from
metallurgical view point. Now there are only two possibilities
for the failure of roller, either pitting fatigue or flexural fatigue
as discussed in section II. Which type of failure has occurred,
it can be analyzed again by optical microscopy study. Setup
for this study is shown in Fig. 11 with CCTV camera.
4% natal 100X
Figure 9. Micrograph for Microstructure of fine and uniform carbides
(3) Inclusion Effect on Fatigue Life
Nonmetallic inclusions in the subsurface region of rolling
elements in contact could raise internal stresses to initiate
fatigue cracks and reduce the service life of rolling elements.
The stress field associated with the inclusions provides an
impetus for material deterioration and fatigue crack nucleation
in rolling elements. Under cyclic contact stresses, the
heterogeneous decay of martensite is developed at certain type
of inclusions such as oxides, carbides and silicates particles in Figure 11. Setup for Microscopy study
25
International Journal of Advances in Engineering Sciences Vol.1, Issue 2, April, 2011
Crack Initiation and Propagation to form risk volumes on the inner surfaces of the
Fig. 12 the area shown is in a zone which has not failed by hollow rollers. In such condition failure will be due
pitting. From the inner fiber crack has been developed and to flexural fatigue. Also bearing will fail before its
propagated which results in the failure of hollow roller. The pre calculated fatigue life as in present case study.
crack is observed at higher magnification of 200X in optical So development of hollow roller at its optimum hollowness is
microscope. Thus mode of failure of hollow roller is flexural a very important parameter to increase the fatigue life in case
fatigue. of Hollow Cylindrical Roller Bearing.
REFERENCES
[1] Scott, D., 1976, Hollow rolling elements, Tribology International, pp.
261-264.
[2] Harris, T. A. and Kotzalas, M. N., 2007, Rolling Bearing Analysis -
Advanced Concepts of Bearing Technology, 5th ed., Taylor & Francis,
Boca Raton.
[3] Harris, T. A. and Kotzalas, M. N., 2007, Rolling Bearing Analysis -
Essential Concepts of Bearing Technology, 5th ed., Taylor & Francis,
Boca Raton.
[4] Bamberger, E. N., Parker, R. J. and Dietrich, M. W., 1976, Flexural
Fatigue of Hollow Rolling Elements, NASA TN D 8313,
Washington.
[5] Gabriel F. D. (2006) Fatigue Considerations of High Strength Rolling
Bearing Steels. Schaeffler Group USA, Inc.
[6] Erwin V. Zaretsky (1997) A. Palmgren Revisited A Basis for Bearing
Life Prediction. NASA Technical Memorandum 107440, Prepared for
the STLE Annual Meeting sponsored by the Society of Tribologists and
Lubrication Engineers, Kansas City.
[7] Shigley, J. E. Mechanical Engineering Design. 1 st ed. McGraw-Hall,
Singapore, 1983.
[8] Nortron, R. L. Machine Design An Integrated Approach. 2nd ed. Pearson
Education, Singapore, 2001.
[9] Kuo C. H. (2008) Contact Stress Analysis of an Elastic Half-Plane
Containing Multiple Inclusions. Int. J. Solids Struct., vol. 45, 4562
4573.
Figure 12. Flexural fatigue crack
(a) Path of crack in cross section of hollow roller (b) Magnified view of crack.
V CONCLUSIONS
Metallurgical analysis has been carried out to check the
microstructure of the material AISI 52100 of hollow roller. All
three major microstructures consists uniformly distribution of
carbides. Percentage of inclusions is also very small in number
and size. So failure of hollow roller is not caused by
metallurgical defect. Therefore the crack initiation i.e. fatigue
failure is not according to metallurgical defect but it is due to
flexure fatigue because crack is developed form the inner
surface of the hollow roller. The value of hollowness for this
roller is 42%, which is not optimized and higher than
requirement and finally premature failure has occurred.
For a percentage of hollowness at or less than the
optimum it was found that the risk volume is under
the loading region, not in the bending stress region.
So the failure of the hollow rollers is due to contact
stresses not due to bending stresses. The effect of
bending stress is not significant for hollowness
percentages less than the optimum. Thus pitting is the
type of failure in such case. If the percentage of
hollowness is less than the optimum we can not get
the maximum advantages of using HCRB.
For percentages of hollowness more than the
optimum, bending stresses on the inner surface start
26
