Available online at www.sciencedirect.

com

Procedia Engineering 10 (2011) 1196–1201

ICM11

Load capacity of a three-row roller slewing bearing raceway

P. Göncza*, R. Potoþnikb and S. Glodežc
a,c
University of Maribor, Faculty of Natural Sciences and Mathematics, Koroška cesta 160, 2000 Maribor, Slovenia
b
Rotis d. o. o., Brodišþe 5, 1236 Trzin, Slovenia

Abstract

A calculation model for fatigue lifetime determination of a three-row roller slewing bearing raceway is presented.
First, the contact force distribution between the roller and the raceway is analytically determined from the basic
bearing geometry and the external loads acting on the bearing assembly. Then a numerical model is employed to
calculate subsurface stress distribution in the raceway as a result of the contact force between the roller and the
raceway. Numerically calculated subsurface stress field serves as an input for the fatigue calculation of the raceway,
which is done according to the stress-life approach. Two different types of rollers are investigated: plain cylindrical
roller and a profiled roller. Finally, the influence of the roller geometry on the fatigue life is pointed out.
© 2011 Published by Elsevier Ltd. Open access under CC BY-NC-ND license.
Selection and peer-review under responsibility of ICM11
Keywords: roller slewing bearings; stress-life approach; multiaxial fatigue; numerical calculation

1. Introduction

Slewing bearings are bearings of large dimensions (> 400 mm) used in different applications, such as
cranes, turning tables, excavators, wind turbines etc. [1]. Their main function is to connect structural
parts, allow relative rotation between them and transmission of loads. Typical three-row roller slewing
bearing is schematically shown in Fig. 1.
Roller bearings in general are widely used and highly standardized machine elements. Thus, there are
some well-established industry standards for the calculation of static [2] and dynamic [3] load ratings of
bearings. However, as pointed out in [4] many assumptions used in these standards are not valid for large

* Corresponding author. Tel.: +386-2-229-3785

E-mail address: peter.goncz@uni-mb.si

1877-7058 © 2011 Published by Elsevier Ltd. Open access under CC BY-NC-ND license.
Selection and peer-review under responsibility of ICM11 doi:10.1016/j.proeng.2011.04.199
 P. Göncz et al. / Procedia Engineering 10 (2011) 1196–1201 1197

slewing bearings: only parallel relative displacement of the bearing rings/raceways, no clearances, no ring
deformations, no changing mechanical properties of the raceway, etc.

Fig. 1. Schematic structure of a typical three-row roller slewing bearing assembly

In this paper a calculation model for determination of load capacity of a three-row roller slewing
bearing’s raceway in presented. The influence of two possible roller shapes on the fatigue life of the
raceway is investigated. Stress-life (S-N) approach [5], together with a multiaxial fatigue criteria
(maximum octahedral shear stress [6]) and experimentally determined material properties is used for that
purpose.

Nomenclature

b [/] fatigue strength exponent

DL [m] raceway diameter
Dwe [mm] roller diameter
E [MPa] Young's modulus
Fa [kN] axial force
Fr [kN] radial force
Lwe [mm] roller length
MT [kNm] overturning/tilting moment
n [/] number of rollers
P(x) [mm] profile function
Q [kN] contact force
Ȟ [/] Poisson's ratio
ıƍf [MPa] fatigue strength coefficient
ıi [MPa] principal stresses (i = 1, 2 ,3)
ıu [MPa] ultimate tensile strength
ıy [MPa] yield strength
1198 P. Göncz et al. / Procedia Engineering 10 (2011) 1196–1201

2. Computational model

2.1. Maximum contact load

As mentioned in the introduction, the standard calculation methods for bearing evaluation are based on
some assumptions which take into account only axial Fa and radial Fr loading of the bearing. On the other
hand, most of large slewing bearings are also loaded with an overturning/tilting moment MT (Fig. 1).
Some simple expressions for the determination of contact force distribution between the rolling elements
and the raceway, based on this kind of external loading, can be found in technical literature. According to
an expression from [7] the maximum contact force between the axial roller and the raceway can be
determined as:

4 ⋅ M T Fa
Qmax = ± (1)
n ⋅ DL n

This expression considers that the vertical contact load linearly varies across the diameter of the
raceway (i.e. sinusoidal distribution). However, if slewing bearing rings undergoes considerable
deformations during operation this method does not give enough accurate results. Numerical simulations
[8] or specialized calculation approaches [4] for the determination of contact load distribution should be
used in such cases.

2.2. Material properties of the rollers and the raceway

Rolling elements of large slewing bearing are usually made of steel 100Cr6, while the rings are usually
made of steel 42CrMo4 [4]. To achieve higher loading capacity and wearing resistance, the raceways are
induction quenched. As a result, the mechanical properties of the raceway vary along the depth (in y-
direction). Some of the previously experimentally determined mechanical properties of different layers of
the slewing bearing raceway, used in the presented calculations, are given in Table 1.

Table 1: Mechanical properties of the slewing bearing’s raceway – 42CrMo4 [4]

layer case trans. 1 trans. 2 trans. 3 trans. 4 trans. 5 core

start – yA [mm] 0 -2.6 -2.7 -2.8 -2.9 -2.0 -3.1
end – yB [mm] -2.6 -2.7 -2.8 -2.9 -3.0 -3.1 <
Hardness [HRC] 59.0 55.9 49.7 43.5 37.3 31.1 28.0
E [MPa] 201136
Ȟ [/] 0.3
ıy [MPa] 1947 1808 1529 1250 971 816 816
ıu [MPa] 2352 2161 1812 1507 1246 1028 936
ıƍf [MPa] 2584 2448 2177 1906 1635 1364 1229
b [/] -0.073

2.3. Rolling element – raceway contact model

The numerical calculations were performed with the computer program Abaqus/Standard [9] as a static
simulation with elastic material properties taken into account. Due to the symmetry of the problem, a 1/8
model was employed (Fig. 2a). The numerical model consisted of approx. 100,000 8-node linear brick
elements. The contact between the roller and the raceway was defined as “hard” contact (Fig. 2b). The
 P. Göncz et al. / Procedia Engineering 10 (2011) 1196–1201 1199

contact loading was prescribed as a vertical displacement (in direction of y) of the roller segment. The
resulting contact force Q was then determined for every vertical displacement ǻyi from the results of the
numerical simulation.

Fig. 2. (a) 1/8 symmetry model of the roller – raceway contact, (b) FE mesh of the numerical model

Two types of rollers were simulated in numerical contact model. The first roller was purely cylindrical
(φ25x25 mm) while the other had a slightly modified side profile (Eq. (2)). This special “logarithmic”
profile [10] was developed to reduce the edge contact stresses which occur in case of purely cylindrical
roller.

ª 1 º
P ( x ) = 0.00035 Dwe ln « 2»
(2)
«¬1 − (2 x / Lwe ) »¼

Fig. 3 shows the actual difference between the contact pressure distribution p(x) along the axis x of
plain cylindrical roller and a profiled roller which was calculated with the presented numerical model.
The difference in the contact stress distribution p(x) between the two types of rollers is evident: although
the plain cylinder has a larger contact area (due to the slightly convex shape of the profiled roller), the
contact stress peaks at the roller ends still have a negative influence.

4500 3000 0.05

Contact pressure - p (x) [N/mm2]

Contact pressure - p (x) [N/mm2]

Profile function - P(x) [mm]

4000 ǻy = 0.02 mm
ǻy = 0.04 mm 2500
3500 0.04
ǻy = 0.06 mm
3000 2000
0.03
2500
1500
2000
ǻy = 0.02 mm 0.02
1500 1000
ǻy = 0.04 mm
1000 ǻy = 0.06 mm 0.01
500 profile function
500
0 0 0
4 0 6 2 8 10 12 0 2 4 6 8 10 12
Roller length - x [mm] Roller length - x [mm]
a) b)
Fig. 3. Contact pressure distribution for the (a) unprofiled (cylindrical) roller, (b) contact pressure distribution for the profiled roller
1200 P. Göncz et al. / Procedia Engineering 10 (2011) 1196–1201

The subsurface stress distribution at a given contact load is presented in Fig. 4. As it is shown, the
maximal von Mises stress occurs slightly under the surface and it is considerably higher at the end of the
cylindrical roller (x = 0.5Lwe) than in the center (x = 0). As for the profiled roller, the highest subsurface
von Mises stress occurs in the center of the contact (x = 0). This von Mises stress is 225 MPa lower than
the highest von Mises stress in case of unprofiled roller.

Stress - ı [N/mm2] Stress - ı [N/mm2]

-2500 -1500 -500 500 1500 -2500 -1500 -500 500 1500
0 0
-0.5 -0.5

1 1

Depth - y [mm]
Depth - y [mm]

ı1 - cen. -1.5
-1.5 ı2 - cen.
ı3 - cen. 2
2 ıvM - cen. ı1 - cen.
ı1 - end -2.5 ı2 - cen.
-2.5 ı2 - end
ı3 - end ı3 - cen.
3 3 b)
a)

Fig. 4. Subsurface stress distribution in the raceway at ǻy = -0.03 for the (a) unprofiled (cylindrical) roller, (b) subsurface
distribution for the profiled roller

2.4. Fatigue lifetime calculation

As mentioned in the introduction, stress-life approach (S-N) was used to calculate fatigue lifetime of
the raceway. This approach is usually applicable to dynamically loaded machine elements, which undergo
between 103 and 107 load cycles [5]. The Basquin [5] equation is valid in this region:
1/ b
§ σ eq ·
N f = ¨¨ b ¸
¸ (3)
© 2 ⋅ σ' f ¹

During the operation of the slewing bearing the contact force varies between Qmax and Qmin = 0. Thus,
the mean stress effect is considered with the Goodman equation [5]:
σ eqa
σ eq = (4)
1 − σ eqm / σ u

Both, equivalent mean stress (Eq. (5)) and equivalent amplitude stress (Eq. (6)) are determined
according to the criterion of maximum octahedral shear stress [6]:

1
σ eqm = (σ m1 − σ m2 )2 + (σ m2 − σ m3 )2 + (σ m3 − σ m1 )2 (5)
2

1
σeqa = (σa1 − σa2 )2 + (σa2 − σa3 )2 + (σa3 − σa1 )2 (6)
2
 P. Göncz et al. / Procedia Engineering 10 (2011) 1196–1201 1201

3. Computational results

Fig. 5a shows the relation between the max. contact force and the equivalent stress, while Fig. 5b
shows the influence of the roller type and contact force magnitude on the calculated fatigue life. It can be
seen again, that in case of unprofiled roller the most critical region is at the end, where the contact stress
peaks occur. Due to these peeks the equivalent subsurface stresses are also higher, which consecutively
shortens the calculated fatigue lifetime. At Qmax = 15 kN the calculated fatigue lifetime of the bearing
raceway in contact with a profiled roller is 2.3x longer than in contact with an uprofiled raceway. At
Qmax = 20 kN this ratio is already 6.3.

1600 45

Max. contact force - Qmax [kN]

40
1400
35
1200 30
ıeq [MPa]

25
1000
20
800 15
profiled profiled
unprofiled (end) 10 unprofiled (end)
600
unprofiled (center) 5 unprofiled (center)
400 0
103 104 105 106 107
0 10 20 30 40
Max. contact force - Qmax [kN] Number of load cycles - Nf
a) b)
Fig. 5. (a) Relation between the max. contact force Qmax and the equivalent stress ıeq, (b) relation between the max. contact force
Qmax, roller type and calculated fatigue lifetime Nf

4. Conclusion

In this paper a calculation model for fatigue lifetime determination of a three-row roller slewing
bearing raceway was presented. The subsurface stress distribution in the roller slewing raceway was
numerically determined. These results were used in the fatigue lifetime determination according to the
stress-life approach. The positive effect of the profiled rollers on the fatigue lifetime was shown.

References

[1] Slewing Bearings [Product Catalog]. 2007, Rothe Erde.

[2] ISO 76: Rolling bearings - Static load ratings. 2006, International Organization for Standardization: Switzerland.
[3] ISO 281: Rolling bearings - dynamic load ratings and rating life. 2007, International Organization for Standardization:
Switzerland.
[4] Potoþnik, R., Static capacity and lifetime of large double-row slewing ball bearings [PhD Thesis], in Faculty of Mechanical
Engineering. 2010, University of Maribor: Maribor.
[5] R. I. Stephens, A.Fatemi, R. R. Stephens, H. O. Fuchs, Metal Fatigue in Engineering, 2nd Ed. 2001: John Wiley & Sons, Inc.
[6] Karolczuk, A. and E. Macha, A Review of Critical Plane Orientations in Multiaxial Fatigue Failure Criteria of Metallic
Materials. International Journal of Fracture, 2005. 134(3-4): p. 267-304.
[7] DNV-2.22 - Lifting appliances. 2008, DET NORSKE VERITAS.
[8] Smolnicki, T., D. Derlukiewicz, and M. Stanco, Evaluation of load distribution in the superstructure rotation joint of single-
bucket caterpillar excavators. Automation in Construction, 2008. 17(3): p. 218-223.
[9] Abaqus Simulia 6.10-1 Documentation. 2010, Dassault Systémes.
[10] ISO/TS 16281: Rolling bearings — Methods for calculating the modified reference rating life for universally loaded bearings.
2008, International Organization for Standardization: Switzerland.