Javascript is required
[1] Harris, T.A. & Kotazlas, M.N., Rolling Bearing Analysis, CRC/Taylor & Francis: Boca Raton, 1991.
[2] Cretu, S., Bercea, I. & Mitu, N., A dynamic analysis of tapered roller bearing under fully flooded conditions. Wear, 188, pp. 1–18, 1995. [Crossref]
[3] Bercea, I., Cretu, S. & Nelias, D., Analysis of double-row tapered roller bearings. Tribology Transactions, 46(2), pp. 228–239, 2003. [Crossref]
[4] Tong, V.C. & Hong, S.W., Characteristics of tapered roller bearing subjected to combined radial and moment loads. International Journal of Precision Engineering and Manufacturing–Green Technology, 1(4), pp. 323–328, 2014. [Crossref]
[5] Warda, B. & Chudzik, A., Fatigue life prediction of the radial roller bearing with the correction of roller generators. International Journal of Mechanical Sciences, 89, pp. 299–310, 2014. [Crossref]
[6] Ferreira, J.L.A., Balthazar, J.C. & Araujo, A.P.N., An investigation of rail bearing reliability under real conditions of use. Engineering Failure Analysis, 10, pp. 745–758, 2003. [Crossref]
[7] Bearing Calculation (Chapter 5). Topics in SKF Railway Technical Handbook, SKF Group, pp. 106–121, 2012.
[8] Lai, K., Lim, S. & Teh, P., Optimization of electrostatic separation process for maximizing biowaste recovery using Taguchi method and ANOVA. Polish Journal of Environmental Studies, 24(3), pp. 1125–1131, 2015. [Crossref]
Search

Acadlore takes over the publication of IJCMEM from 2025 Vol. 13, No. 3. The preceding volumes were published under a CC BY 4.0 license by the previous owner, and displayed here as agreed between Acadlore and the previous owner. ✯ : This issue/volume is not published by Acadlore.

Open Access
Research article

Fatigue Life Analysis of a Railway Bearing Using Taguchi Method

t.k. ahn1,
j.y. heo1,
s.h. lee2,
h.j. jung3,
j.w. kim4,
h.b. yi4
1
Department of Automotive Engineering, Hoseo University, Korea
2
Department of Automotive Engineering, Doowon University, Korea
3
Bearing Art, Korea
4
Korea Railroad Corp., Korea
International Journal of Computational Methods and Experimental Measurements
|
Volume 5, Issue 5, 2017
|
Pages 733-740
Received: N/A,
Revised: N/A,
Accepted: N/A,
Available online: 08-31-2017
View Full Article|Download PDF

Abstract:

An axle bearing is one of the most important components to guarantee the service life of a rail car. In order to ensure the stable and reliable bearing life, it is essential to estimate the fatigue life of an axle bearing under the loading conditions. The fatigue life of a bearing is affected by many parameters such as material properties, heat treatment, lubrication conditions, operating temperature, loading conditions, bearing geometry, the internal clearance of bearing, and so on. Because these factors are so complicatedly related to each other, it is very important to investigate the effects of these factors on the axle bearing life. This paper presents the process of estimating the fatigue life of a railroad roller bearing, which takes into account geometric parameters of the bearing in the life calculation. The load distributions of the bearing were determined by solving numerically force and moment equilibrium equations with Lundberg’s approximate model. This paper focuses on analyzing the effects of bearing geometric parameters on the fatigue life using Taguchi method.

Keywords: Fatigue life, Internal geometric parameter, Railway axlebox, Taguchi method, Tapered roller bearing unit

1. Introduction

An axlebox is the linking design element between the rotating wheelset and the railway vehicle. An axlebox bearing has always been a critical component in the reliability of railway vehicles. In order to ensure the stability and reliability of the bearings, it is essential to estimate the fatigue life of an axle bearing under the loading conditions. Tapered roller bearing units are widely used in railway axleboxes due to the high-load capacity against axial loads as well as radial loads.

The fundamental theories of rolling element bearings were early established by Lundberg, Palmgren, and Harris [1]. The application of these theories was the origin of analytical methods to calculate the internal load distribution in a rolling bearing and subsequently to predict its stiffness and fatigue life. With the rapid development of computers, various investigations [2–5] to estimate the load distribution of a bearing have been carried out considering the effect of complex loading conditions as well as the internal bearing geometry. However, most studies are too complicated to use in the industrial field and have focused on the investigation of bearing characteristics under the assumption that the values of external loads acting on the bearing are known. In practice, it is hard to know accurately the bearing load values in many cases.

In this paper, the relationships between external forces, bearing loads, displacements, and load distributions of a bearing unit were investigated. The fatigue life of a tapered roller bearing unit was calculated based on Lundberg’s approximate model which considers the influence of the internal geometric parameters of a bearing unit. Then, the effects of internal geometric parameters on the fatigue life of the bearing unit are investigated by using Taguchi method.

2. Fatigue Life of a Bearing Unit

2.1 Load-Deflection Relationships of a Rolling Bearing

Generally, when a tapered roller bearing with contact angle $\alpha$ is subjected simultaneously to a radial and an axial load, Fig. 1 shows the relative radial displacement $\delta_r$ and the axial displacement $\delta_a$ between bearing rings [1]. At any angular position $\psi$ measured from the most heavily loaded rolling element, the approach of the rings can be expressed as

$\delta_\psi=\delta_a \sin \alpha+\delta_r \cos \alpha \cos \psi$
(1)

The maximum relative deflection at $\psi=0$ is given by

$\delta_{\max }=\delta_a \sin \alpha+\delta_r \cos \alpha=\frac{2 \varepsilon}{2 \varepsilon-1} \delta_a \sin \alpha$
(2)

where $\varepsilon$ is the load distribution factor

$\varepsilon=\frac{1}{2}\left(1+\frac{\delta_a \tan \alpha}{\delta_r}\right)$
(3)

In condition of pure axial deflection for standard tapered roller bearings made of the bearing steel [1], the axial deflection can be approximated as

$\delta_{a 0}=\frac{0.000077}{\sin \alpha} \frac{Q_{\max }^{0.9}}{l_e^{0.8}}=\frac{\delta_{\max }}{\sin \alpha}$
(4)

where $Q_{\max }$ is the maximum roller normal load $(\mathrm{N})$, and $l_e$ is the effective roller length (mm).

Figure 1. Rolling bearing deflection due to combined radial and axial loading

From eqns (2) and (4), the axial deflection $\delta_a$ can be derived as

$\delta_a=\frac{0.000077}{\sin \alpha} \frac{Q_{\max }^{0.9}}{l_e^{0.8}} \frac{2 \varepsilon-1}{2 \varepsilon}.$
(5)

Using load integrals, applied axial bearing load can be expressed as

$F_a=Z \sin \alpha Q_{\max } J_a(\varepsilon)$
(6)

where Z is the number of rollers. J is defined as

$J_a(\varepsilon)=\frac{1}{2 \pi} \int_{-\psi_l}^{\psi_l}\left[1-\frac{1}{2 \varepsilon}(1-\cos \psi)\right]^{1.11} d \psi.$
(7)

Applied radial bearing load can be expressed as

$F_r=Z \cos \alpha Q_{\max } J_r(\varepsilon)$
(8)

where

$J_r(\varepsilon)=\frac{1}{2 \pi} \int_{-\psi_l}^{\psi_l}\left[1-\frac{1}{2 \varepsilon}(1-\cos \psi)\right]^{1.11} \cos \psi d \psi$
(9)
$\psi_l=\cos ^{-1}(1-2 \varepsilon).$
(10)

Consequently, if bearing loads $F_a$ and $F_r$ are given, $Q_{\max }$ and $\varepsilon$ can be calculated numerically from eqns (6) and (8). Using them, $\delta_r$ and $\delta_a$ can be acquired from eqns (3) and (5).

2.2 Equilibrium Conditions of a Tapered Roller Bearing Unit

Figure 2 shows a typical tapered roller bearing unit used in railway vehicles. From the SKF guide principles, which are widely used in railway industry [6, 7], the equivalent radial axlebox load can be calculated as

$K_r=f_0 f_{r d} f_{t r} G$
(11)

where $f_0$ is the payload factor, $f_{r d}$ is the dynamic radial factor, $f_{t r}$ is the dynamic traction factor, and $G$ is the axlebox load.

The equivalent axial axlebox load can be calculated as

$K_a=f_0 f_{a d} G$
(12)

where $f_{a d}$ is the dynamic axial factor.

Considering the loading conditions, the following forces and moment equilibrium equations can be derived.

$K_r-F_{r o}-F_{r i}=0$
(13)
$K_a+F_{a o}-F_{a i}=0$
(14)
$l_d K_a+l_c F_{r o}-\frac{l_c}{2} K_r=0$
(15)

where $l_d$ is the axial axlebox load distance (mm) (Fig. 2), and $l_c$ is the distance between bearing load centres (mm).

Figure 2. Axle bearing unit with applied loads

In order to obtain the four unknown bearing loads, one more equation can be used from the geometric constraint which is assumed that total axial clearance is unchanged after loading.

$\delta_{a o}+\delta_{a i}=\delta_0$
(16)

where $\delta_0, \delta_{a o}$, and $\delta_{a i}$ are the initial axial clearance, the outer bearing axial deflection, and the inner bearing axial deflection, respectively.

All bearing loads can be obtained by solving eqns (13)–(16).

2.3 Fatigue Life of a Tapered Bearing Unit

The fatigue life of a rolling bearing is calculated from the following formula:

$L_{10}=\left(\frac{C}{P}\right)^{10 / 3} \times 10^6(\mathrm{rev.)}.$
(17)

where C is the basic dynamic capacity of a rolling bearing (N).

The equivalent dynamic bearing load P is

$P=\frac{J_r(0.5) J_1(\varepsilon)}{J_1(0.5) J_r(\varepsilon)} F_r$
(18)

where

$J_1(\varepsilon)=\left\{\frac{1}{2 \pi} \int_{-\psi_l}^{\psi_l}\left[1-\frac{1}{2 \varepsilon}(1-\cos \psi)\right]^{4.4} d \psi\right\}^{1 / 4}.$
(19)

Because a tapered roller bearing unit has two bearings, the total fatigue life of a bearing unit is given by

$L_{10_{\_} \text {tot }}=\left(L_{10_{\_} \text {in }}^{-1.125}+L_{10_{\_} \text {out }}^{-1.125}\right)^{-8 / 9}.$
(20)

3. Effects of Internal Geometric Parameters on the Fatigue Life

The fatigue life of a rolling bearing with the same external dimensions can be varied with internal geometric parameters. The effects of four geometric parameters; effective roller length, roller diameter, pitch diameter of the bearing, and number of rollers, on the fatigue life of a tapered roller bearing unit were investigated using Taguchi method. Taguchi method is based on an orthogonal array arrangement which reduces the number of experiments and the cost for the necessary information [8]. A signal-to-noise ratio (S/N) is used in the objective function to analyze the variation of the parameters. For the fatigue life of the bearing unit, the-larger-the-better type of the objective function is taken and is defined as

$S / N=-10 \log \left(\frac{1}{n} \sum_{i=1}^n \frac{1}{y_i^2}\right)$
(21)

where $y_i$ is the observed data, and $n$ is the number of experiments in a trial. A greater $S / N$ ratio shows the best performance. Relative influences of the parameters on the response can be analyzed with the sum of squares $(S S)$ which is defined as:

$S S=\sum(S / N-\overline{S / N})^2$
(22)

where

$\overline{S / N}=\frac{1}{n} \sum S / N.$
(23)

Three levels of a parameter were chosen to investigate a trend of the response to the parameter variation. Consequently, L9 was employed as an orthogonal array because it is suitable for four parameters with three levels. The specifications of the axlebox and the tapered roller bearing unit used in this study are shown in Table 1 and 2 shows four parameters with three levels which were selected considering the implementable range of each parameter.

Table 1. Specifications of the axlebox and the bearing unit

Axlebox

Axlebox load, G (kN)

98.7

Wheel diameter (mm)

860

Payload factor

1.0

Dynamic radial factor

1.3

Dynamic axial factor

0.12

Dynamic traction factor

1.05

Axial axlebox load distance (mm)

16.5

Tapered roller bearing unit

Outside diameter (mm)

207

Bore diameter (mm)

131.75

Width (mm)

152.4

Contact angle (deg)

10

Distance between load centres (mm)

114.8

Basic dynamic capacity (kN)

635

The fatigue life of a bearing unit was calculated according to the orthogonal array and is listed in Table 3. The initial axial clearance was assumed 0. Table 4 shows the analysis of each S/N and SS of parameters. The effects of four parameters on the fatigue life of the bearing unit are shown in Fig. 3.

Table 2. Parameters and their levels

Parameter

Level 1

Level 2

Level 3

Effective roller length (mm)

le

40.7

41.2

41.7

Mean roller diameter (mm)

dr

17.12

18.4

19.35

Effective pitch diameter (mm)

D

164.4

166.4

168.4

Number of rollers

Z

22

23

24

Table 3. Values of the fatigue life with L9 array

No.

le

dr

Dp

Z

Unit life (×10 km)

S/N (dB)

1

40.7

17.12

164.4

22

24.4

107.7

2

40.7

18.4

166.4

23

36.2

111.2

3

40.7

19.35

168.4

24

49.1

113.8

4

41.2

17.12

166.4

22

31.1

109.9

5

41.2

18.4

168.4

23

33.3

110.4

6

41.2

19.35

164.4

24

46.1

113.3

7

41.7

17.12

168.4

22

28.7

109.2

8

41.7

18.4

164.4

23

43.2

112.7

9

41.7

19.35

166.4

24

42.3

112.5

Table 4. S/N and SS of parameters

Parameters

Level

S/N

Mean

SS

le

1

110.9

111.2

0.151

2

111.2

111.2

3

111.5

111.2

dr

1

108.9

111.2

9.347

2

111.4

111.2

3

113.2

111.2

Dp

1

111.2

111.2

0.005

2

111.2

111.2

3

111.1

111.2

Z

1

110.2

111.2

1.786

2

111.2

111.2

3

112.1

111.2

From Fig. 3, the mean roller diameter is the most significant parameter affecting to the fatigue life of a bearing unit. The next parameter is the number of rollers. Other parameters have a little effects on the fatigue life. The fatigue life can be decreased by increasing the effective pitch diameter. However, the fatigue life can be increased by increasing other parameters.

(a)
(b)
(c)
(d)
Figure 3. Effects of four parameters on the fatigue life of the bearing unit

4. Conclusions

An axlebox bearing unit is the critical component for the reliability of railway vehicles. In order to investigate the fatigue life of a bearing unit, equilibrium conditions were formulated on the basis of Lundberg’s approximate model. The fatigue life of bearing units can be varied with internal geometric parameters even if bearing units have all the same external dimensions. Taguchi method has been applied to determine the effects of the internal geometric parameters on the fatigue life. The effects of four internal geometric parameters on the fatigue life are examined by L9 orthogonal array. From the results, the mean roller diameter is the most significant parameter and the number of rollers is the next. The effective roller length and the effective pitch diameter of the bearing unit effect a little on the fatigue life. The fatigue life of the bearing unit can be increased by increasing all parameters except the effective pitch diameter.

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request.

Acknowledgments

In this study, research funding support of the Ministry of Land, Infrastructure and Transport, railway technology research business (15RTRP-B085684-02-000000) was done by.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

References
[1] Harris, T.A. & Kotazlas, M.N., Rolling Bearing Analysis, CRC/Taylor & Francis: Boca Raton, 1991.
[2] Cretu, S., Bercea, I. & Mitu, N., A dynamic analysis of tapered roller bearing under fully flooded conditions. Wear, 188, pp. 1–18, 1995. [Crossref]
[3] Bercea, I., Cretu, S. & Nelias, D., Analysis of double-row tapered roller bearings. Tribology Transactions, 46(2), pp. 228–239, 2003. [Crossref]
[4] Tong, V.C. & Hong, S.W., Characteristics of tapered roller bearing subjected to combined radial and moment loads. International Journal of Precision Engineering and Manufacturing–Green Technology, 1(4), pp. 323–328, 2014. [Crossref]
[5] Warda, B. & Chudzik, A., Fatigue life prediction of the radial roller bearing with the correction of roller generators. International Journal of Mechanical Sciences, 89, pp. 299–310, 2014. [Crossref]
[6] Ferreira, J.L.A., Balthazar, J.C. & Araujo, A.P.N., An investigation of rail bearing reliability under real conditions of use. Engineering Failure Analysis, 10, pp. 745–758, 2003. [Crossref]
[7] Bearing Calculation (Chapter 5). Topics in SKF Railway Technical Handbook, SKF Group, pp. 106–121, 2012.
[8] Lai, K., Lim, S. & Teh, P., Optimization of electrostatic separation process for maximizing biowaste recovery using Taguchi method and ANOVA. Polish Journal of Environmental Studies, 24(3), pp. 1125–1131, 2015. [Crossref]

Cite this:
APA Style
IEEE Style
BibTex Style
MLA Style
Chicago Style
GB-T-7714-2015
Ahn, T., Heo, J., Lee, S., Jung, H., Kim, J., & Yi, H. (2017). Fatigue Life Analysis of a Railway Bearing Using Taguchi Method. Int. J. Comput. Methods Exp. Meas., 5(5), 733-740. https://doi.org/10.2495/CMEM-V5-N5-733-740
T. Ahn, J. Heo, S. Lee, H. Jung, J. Kim, and H. Yi, "Fatigue Life Analysis of a Railway Bearing Using Taguchi Method," Int. J. Comput. Methods Exp. Meas., vol. 5, no. 5, pp. 733-740, 2017. https://doi.org/10.2495/CMEM-V5-N5-733-740
@research-article{Ahn2017FatigueLA,
title={Fatigue Life Analysis of a Railway Bearing Using Taguchi Method},
author={T.K. Ahn and J.Y. Heo and S.H. Lee and H.J. Jung and J.W. Kim and H.B. Yi},
journal={International Journal of Computational Methods and Experimental Measurements},
year={2017},
page={733-740},
doi={https://doi.org/10.2495/CMEM-V5-N5-733-740}
}
T.K. Ahn, et al. "Fatigue Life Analysis of a Railway Bearing Using Taguchi Method." International Journal of Computational Methods and Experimental Measurements, v 5, pp 733-740. doi: https://doi.org/10.2495/CMEM-V5-N5-733-740
T.K. Ahn, J.Y. Heo, S.H. Lee, H.J. Jung, J.W. Kim and H.B. Yi. "Fatigue Life Analysis of a Railway Bearing Using Taguchi Method." International Journal of Computational Methods and Experimental Measurements, 5, (2017): 733-740. doi: https://doi.org/10.2495/CMEM-V5-N5-733-740
AHN T K, HEO J Y, LEE S H, et al. Fatigue Life Analysis of a Railway Bearing Using Taguchi Method[J]. International Journal of Computational Methods and Experimental Measurements, 2017, 5(5): 733-740. https://doi.org/10.2495/CMEM-V5-N5-733-740