The chemical composition of the bearing sleeve material was determined by spectroscopy chemical

analysis method. The microstructure of the bearing sleeve material was observed by scanning electron

* Corresponding author. Tel.: +86 0411 84729613; fax: +86 0411 84728670.

E-mail address: xxiaolei@dlmu.edu.cn (X. Xu).

Engineering Failure Analysis 13 (2006) 8578651350-6307/$ - see front matter 2005 Elsevier Ltd. All rights reserved.1. Introduction

A bearing sleeve made of GCr15SiMn fractured, when being interference tted. The heat treatment of

the bearing sleeve is as follows: quenching (820 C, 30 min)! tempering (200 C, 3.54.0 h). This paperdescribes a detailed metallurgical investigation and fractographic analysis on the failed bearing sleeve.

The possible failure causes were assessed.

2. Investigation methodsAbstract

This paper presents a failure analysis of a bearing sleeve made from GCr15 steel. The bearing sleeve fractured, when

being interference tted. The fracture surfaces were observed by visual and SEM. The crack initiated from the top end

of the bearing sleeve and propagated towards the bottom. Intergranular brittle fracture is the main failure mechanism.

Detailed investigation of the microstructure indicates needle-like bainite, with lamellar troostite occuring at grain

boundaries as a discontinuous network. A banded structure was observed on the bearing sleeve material, but Charpy

impact tests of samples taken from dierent orientations of the failed sleeve show that the failure of the bearing sleeve

has nothing to do with the banded structure. The discontinuous network structure lowers the strength of the grain

boundary. Insucient quenching rate and low quenching temperature are responsible for the formation of the discon-

tinuous grain boundary network.

2005 Elsevier Ltd. All rights reserved.

Keywords: Bearing failure; Intergranular fracture; Alloy steel; Microstructures; FractographyFailure analysis of GCr15SiMn steel bearing sleeve

Xiaolei Xu *, Zhiwei Yu

Institute of Metal and Technology, Dalian Maritime University, Dalian 116026, PR China

Received 14 January 2005; accepted 20 February 2005

Available online 17 May 2005

www.elsevier.com/locate/engfailanaldoi:10.1016/j.engfailanal.2005.02.020

3. Observation results

propagated toward the bottom along the axial direction. No plastic deformation can be observed on the

3.3. Metallurgical examination

The microstructure of the longitudinal and transverse section samples close to and far from the crackorigin zone was observed by SEM. The longitudinal section microstructure shows an obvious banded struc-

ture at low magnication, is composed of bright and dark zones (Fig. 6). There are more carbides in the

bright zone than the dark zone and the carbides should be M3C.

Many observations show that the microstructure of the bearing sleeve material is mainly composed

of martensite and small particulate carbides. XRD shows that the material contains a small amount ofretained austenite (Fig. 7), about 11.0% by quantitative analysis. However, at high magnication, in the

longitudinal and transverse sections, needle-like bainite and lamellar troostite can be observed. It is

important that the bainite and troostites is distributed at the grain boundary in a discontinuous net-

work (Fig. 8). The length of the needle-like bainite or plate-like troostite is about 10 lm, correspondingfracture. It is suggested that embrittlement fracture occurred in the failed bearing sleeve.

3.2. Microscopic features

A radiating crack propagation path can be observed on the fracture surface of the upper end (Fig. 3),

which indicates that the crack initiated at the upper end of the bearing sleeve. The fracture surface in

the crack origin and bottom zones exhibits a mainly intergranular morphology and the intergranular planesare parallel to the fracture surface (Fig. 4). The grain size of the matrix is about 10 lm. In the crack prop-agation zone in the middle of the bearing sleeve, both intergranular and quasi-cleavage fracture can be ob-

served (Fig. 5). All in all, intergranular fracture should be the main failure mode.3.1. Macroscopic features

Two remains of the failed bearing sleeve are shown in Fig. 1. It can be seen that a longitudinal crack

appears in the bearing sleeve. The crack which initiated from the upper end of the bearing sleeve propa-

gated as a plane, but then deected in the middle of the sleeve. A fully planar crack appears at the top

end, and at the bottom of the sleeve the crack appears obviously bent. The macro-appearance is silver gray

and there is no rust on the fracture surface (Fig. 2), no inclusions can be observed at the crack origin region.

From the crack deection, it is suggested that the crack initiated at the top end of the bearing sleeve andmicroscopy (SEM) on a Philips XL-30 scanning electron microscope. Hardness (HRC) of the material of

the failed sleeve was measured. The fracture surfaces were analysed by visual and SEM observation to

study the failure mechanism. The amount of retained austenite was determined by XRD on a Rigaku dif-

fractometer with Coka radiation as per GB836287 [1]. Charpy impact tests were carried out using aninstrumented pendulum type impact testing machine with a 150-J hammer as per GB/T 229-1994 [2]. Im-pact samples were prepared from the circumferential and axial orientations of the bearing sleeve, with a U

notch of 2 mm depth in the circumferential, axial and radial orientations. Two samples were tested for each

condition and an average value was taken. Tempering testing at dierent temperatures was conducted to

study the inuence of the tempering temperature on the microstructure.

858 X. Xu, Z. Yu / Engineering Failure Analysis 13 (2006) 857865to the size of grains from the intergranular fracture. It is suggested that the intergranular fracture of

X. Xu, Z. Yu / Engineering Failure Analysis 13 (2006) 857865 859the bearing sleeve should be related to the presence of the discontinuous network grain boundary bai-

nite and troostite.

3.4. Chemical composition of the bearing sleeve material

The chemical composition of the bearing sleeve materials was determined by spectroscopy chemical anal-ysis. The results are shown in Table 1. It can be seen that the composition of the materials corresponds to

the specied composition range.

Fig. 1. Failed bearing sleeve.

860 X. Xu, Z. Yu / Engineering Failure Analysis 13 (2006) 8578653.5. Hardness and impact toughness

The hardness (HRC) of the failed bearing sleeve at dierent orientations was conducted. The results

show that the hardness values of the failed bearing sleeve material are within HRC5860, which corre-

sponds to the specied range (HRC5763). In order to analyse whether the intergranular fracture is related

to the banded structure, Charpy tests for samples with dierent orientations were conducted. The impact

toughness values are shown in Table 2. It can be seen that the impact toughness of the three kinds of

samples is very low, but not sensitive to the orientation of the sample or the notch. It is suggested that

Fig. 2. Macros of the fracture surfaces.

Fig. 3. Fracture surface of the crack origin zone showing radiating pattern.

Fig. 4. Fracture surface of the crack origin zone showing intergranular morphology.

Fig. 5. Fracture surface of the crack propagation zone showing intergranular fracture and quasi-cleavage.

Fig. 6. Longitudinal section showing banded microstructure.

X. Xu, Z. Yu / Engineering Failure Analysis 13 (2006) 857865 861

Fig. 7. XRD pattern of the sample taken from the failed bearing sleeve.

Fig. 8. Microstructure of the bearing sleeve material.

Table 1

Chemical composition of the bearing sleeve material (wt%)

0.99 0.48 0.95 0.021 0.015 1.47

Specied 0.951.05 0.400.65 0.951.20 60.027 60.02 1.301.65

Table 2

Impact toughness and hardness values

Specimen orientation Notch orientation Impact toughness (J) HRC

Circumferential Axial 3.1 56.9

Circumferential Radial 3.5 56.7

Axial Circumferential 3.0 57.6

862 X. Xu, Z. Yu / Engineering Failure Analysis 13 (2006) 857865

X. Xu, Z. Yu / Engineering Failure Analysis 13 (2006) 857865 863Fig. 9. Macro-morphology of the impact fracture surfaces.the failure of the bearing sleeve has nothing to do with the banded structure. The macro-fracture of the

impact samples exhibits similar brittle fracture characteristics in the fracture surface is smooth, the radiat-

ing origin appears in the middle of the notch and there is no shear lip on the full fracture (Fig. 9). Micro-

fracture of samples shows mainly intergranular fracture morphology. The normal micro-fracture should be

quasi-cleavage morphology for the bearing steel as quenched and low-temperature tempered such as GCr15

or GCr15SiMn [35]. The intergranular fracture is indicative of the very low toughness of the material and

abnormal microstructure resulting from unsuitable heat treatment.

Fig. 10. Microstructure of the tempered sample.

for 2 h. A change in hardness does not occur when tempering at 200 and 270 C. The discontinuous net-

strength of the grain boundaries. When interference tting, longitudinal crack initiated from one end of the

nite and lamellar troostite, which are distributed at the grain boundaries in discontinuous network form.

The discontinuous network decreases the strength of the grain boundaries leading to brittle intergranular

fracture under the circumferential interference t stress. The discontinuous network grain boundary bai-

nite and troostite resulted from the insucient quenching rate and low quenching temperature.

4. An axial banded structure appears in the material, but the failure of the bearing sleeve has nothing to do

with to this according the impact test results of samples of dierent orientations.

References

[1] Retained austenite in steel-quantitative determination method of X-ray diractometer. China Standard, GB8362-87; 1989.bearing sleeves and propagated along the grain boundaries from top to the bottom under the circumferen-

tial stress.The needle-like bainite or lamellar troostite would precipitate when the quenching is less. Additionally, a

lower quenching temperature may make the grain boundaries insuciently alloyed leading to poor harde-

nability in this zone, so that bainite and troostite precipitate more easily than martensite. The discontinuous

network grain boundary bainite and troostite cannot be eliminated by following a low-temperature temper-

ing process. The presence of more retained austenite and appearance of a banded structure in the bearing

sleeve indicates indirectly the insucient quenching rate and low quenching temperature.

5. Conclusions

1. The crack initiated from the top end of the bearing sleeve and propagated toward the bottom along a

longitudinal direction. Intergranular fracture is the main failure mechanism of the bearing sleeve.

2. The chemical composition and hardness of the failed bearing sleeve material correspond to the specied

range.

3. The microstructure is mainly composed of the martensite, small particulate carbides, and needle-like bai-work grain boundary troostite and bainite structure still remains. However, the network structure nearlydisappears when tempering at 360 C (Fig. 10). This indicates that the network structure has low temper-ature tempering resistance. Although the needle-like bainite and lamellar troostite structure can be elimi-

nated by tempering at higher temperature, the hardness value of the material would decrease (HRC53.7)

which would not satisfy the technical specication for hardness.

4. Analysis of the failure causes

From the observations and examination in Section 3, it is inferred that the chemical composition and the

hardness correspond to the specied range. The microstructure is mainly composed of martensite, small

particulate carbides, and discontinuous network grain boundary bainite and troostite. Although martensite

and small particulate carbides are the typical microstructure of a quenching and low-temperature tempered

bearing steel, the appearance of discontinuous network grain boundary bainite and troostite decreases the3.6. Tempering test

Tempering tests were conducted on three samples from the failed bearing sleeve at 200, 270, and 360 C

864 X. Xu, Z. Yu / Engineering Failure Analysis 13 (2006) 857865[2] Metal materials-Charpy notch impact test. China Standard, GB/T 229-1994; 1995.

[3] Vander Voort GV. Embrittlement of steels. Metals handbook. Properties and section, irons, steels and high-performance alloys,

vol. 1. 10th ed. Metals Park (OH): ASM International; 1990.

[4] Engel L, Klingele H. An atlas of metal damage. Munich, Germany: Carl HanserVerlag; 1981.

[5] Metal handbook. Fractography and atlas of fractographs, vol. 9. 8th ed. Metals Park (OH): American Society for Metals; 1974.

X. Xu, Z. Yu / Engineering Failure Analysis 13 (2006) 857865 865

Failure analysis of GCr15SiMn steel bearing sleeveIntroductionInvestigation methodsObservation resultsMacroscopic featuresMicroscopic featuresMetallurgical examinationChemical composition of the bearing sleeve materialHardness and impact toughnessTempering test

Analysis of the failure causesConclusionsReferences