Engineering Failure Analysis 82 (2017) 26–38

Contents lists available at ScienceDirect

Engineering Failure Analysis

journal homepage: www.elsevier.com/locate/engfailanal

Failure analysis on abnormal wear of roller bearings in gearbox for MARK

wind turbine
Yi Gonga, Jing-Lu Feia, Jie Tanga, Zhen-Guo Yanga,⁎, Yong-Ming Hanb, Xiang Lib
a
Department of Materials Science, Fudan University, Shanghai 200433, PR China
b
The Inner Mongolia Autonomous Region Institute of Product Quality Inspection, Huhhot 010070, PR China

AR TI CLE I NF O AB S T R A CT

Keywords: The booming wind power industry as one of the predominant approaches of renewable energy
Wind turbine gearbox resources is not independent of components reliability within the wind turbines, especially the
Roller bearing rotating ones including blades, motors, gearboxes etc. Hence, case studies on failed parts of such
Failure analysis components play an important role for failure prevention and deserves to be publicly reported for
Three-body abrasive wear
experience sharing. In this paper, a failure case concerning severe abnormal wear of the roller
bearing's inner ring in the gearbox of one 1.5 MW wind turbine in China was reported. Since such
roller bearings were imported from a foreign company and operated by a Chinese company, not
only economic losses but also international commercial dispute would be induced if this case
would not be immediately and properly solved. To this end, by means of comprehensive and
systematic investigation into the base materials, process media, surface morphologies, micro-area
compositions and even service environments, root causes of this failure were determined, de-
tailed mechanisms were discussed, and pertinent countermeasures were proposed. Achievement
of this paper would provide the solid evidence to distinguish the responsibilities for the failure,
and would also help to prevent such failures of roller bearings with similar design in wind tur-
bines.

1. Introduction

The installed wind power capacity of China has reached nearly 170 GW till the end of 2016 [1], which still remains the top
position in the world since 2010 [2] and accounts for more than 10% of the total installed capacity in the country. In China, the Inner
Mongolia Autonomous Region holds the largest quantity of wind power generators (commonly known as wind turbines) thanks to its
superior climate and geographical conditions, and over half of them are in the scale of 1.5 MW. That is to say, safe and economic
operation of these 1.5 MW wind turbines ensured by effective reliability assessment and failure analysis of their components, par-
ticularly the rotating ones including blades, motors, gearboxes etc., would play a crucial role in continuous development of this
renewable energy resource domestically, nationally and even globally under the increasing pressure from environmental protection
[3–5].
As for the average failure rates of all the components within a wind turbine, gearbox ranks in the top ones [6–9], only always falls
behind the blades whose failure incidents could be easily found in the literature [10–18]. However, study of the gearboxes primarily
focused on fault diagnosis through vibration monitoring [19–24], while similar failure cases of them have been rarely reported. In
detail, a gearbox is used to increase the rotational speed from the low-speed rotor through the shafts to the high-speed electrical

⁎
Corresponding author.
E-mail address: [email protected] (Z.-G. Yang).

http://dx.doi.org/10.1016/j.engfailanal.2017.08.015
Received 18 April 2017; Received in revised form 13 July 2017; Accepted 28 August 2017
Available online 30 August 2017
1350-6307/ © 2017 Elsevier Ltd. All rights reserved.
Y. Gong et al. Engineering Failure Analysis 82 (2017) 26–38

generator for electricity generation, and its bearings are regarded as one of the most vulnerable subcomponents to failure according
to the statistics [25,26]. Although plentiful investigation has been conducted into the failure mechanisms of the roller bearings within
wind turbine gearboxes, such as the peculiar morphology of the butterfly cracks and the accompanied white etching areas that are
induced by rolling contact fatigue [27–31], they were mostly from the laboratories rather than the engineering practices. Also, the
existing handful of failure analysis incidents on gearboxes mainly dealt with broken of the gear teeth in the shafts [32,33], and fatigue
fracture of the rotor shafts [34–36] and the shear pins that connect the gearbox and the generator [37], while only a little emphasis
was laid on the bearings, e.g. Gould et al. recently elucidated the causes of premature cracking on a failed bearing by utilizing X-ray
tomography [38]. Thus, it can be now recognized that failure analysis on the bearings of wind turbine gearboxes deserves special
attention, and should be immediately reported in public for experience sharing.
In this paper, failure analysis was addressed on abnormal wear of the roller bearings within the gearbox of one 1.5 MW three-
stage wind turbine installed in the Baotou city in the Inner Mongolia Autonomous Region in China. Since this batch of roller bearings
was manufactured by and imported from a foreign company (the supplier), while installed and operated by a Chinese company (the
owner), which one should be blamed for the economic losses due to outage for this failure was urgently required to be identified. To
this end, systematic failure analysis was carried out by means of various characterization instruments and methods based on our
previous experiences on industrial equipment in the past several consecutive years [39–49], including photoelectric direct reading
spectrometer, metallographic microscope (MM) and Vickers micro hardness tester for inspection of the roller bearing base materials;
optical microscope (OM), three-dimensional stereomicroscope (3D SM), scanning electron microscope (SEM) and energy disperse
spectroscope (EDS) for observation of the wear-out surfaces morphologies; and gas chromatography-mass spectrometer (GC–MS),
Fourier transform infrared spectroscopy (FT-IR), and thermogravimetric analysis (TGA) for detection of the environmental medium
lubricant. Then, relevant mechanisms were discussed and pertinent countermeasures were proposed. Considering the fact that under-
performance is a common problem [50] and potential risk is high [51] for most Chinese wind turbines, achievement of this paper
would not simply settle the international economic dispute between these two companies, but help to improve the reliability of wind
turbine gearboxes that are domestically designed, manufactured, tested, operated and maintained in China.

2. Experimental and results

2.1. Visual observation

Fig. 1(a) and (b) respectively displayed the external appearances of the gearbox and the layout of its three planetary gears around
the main shaft, and the roller bearings of concern were located inside these gears, seen in Fig. 1(c). In more detail, it was acquired
from the owner that this cylindrical type of roller bearings consisted of one inner ring with inside diameter of 220 mm, one outer ring
with outside diameter of 340 mm, and two parallel raceways with total width of 160 mm, each containing 24 cylindrical rollers in the
grooves, seen in Fig. 2(a) and (b). After disassembly of some failed bearings, it was found that the inner rings had suffered severe
localized wear on their raceways (Fig. 3(a)), and even had fractured to several parts (Fig. 3(b)).
In order to identify the root causes of such failure, one of the severely worn roller bearing's inner rings and eight of its cylindrical
rollers was provided from the site. As displayed in Fig. 4(a), there were totally three wear-out areas in sizes ranging from several to a
dozen centimeters existing on the raceways of this inner ring. For the purpose of further investigation, five samples were then wire-
cut from the inner ring, and their locations were marked in Fig. 4(b) and (c). As for the rollers, since they were nearly the same and
free of distinct wear traces, only one of them was taken as sample, seen in Fig. 4(d).

2.2. Material examination

Firstly, chemical compositions of the base materials of the roller bearing's inner ring and cylindrical rollers were detected by
photoelectric direct reading spectrometer. Based on the results listed in Table 1, it could be determined that they conformed to the
requirements of 100Cr6 and 100CrMnSi6-4 bearing steels specifications in the ISO 683-17:2014 standard [53] as designed, which are
respectively equivalent to GCr15 and GCr15SiMn in the GB/T 18254-2016 standard [54] of China.
Metallographic structures including both polished states and etched states of the roller bearing's inner ring and cylindrical roller
were then inspected under the metallographic microscope. As presented in Fig. 5(a) and (c) respectively, no inclusions existed in the
inner ring, while Group D (globular oxide type) inclusions with a rating index number between 0.5–1.0 according to the ISO
4967:2013 standard [55] could be obviously observed in the cylindrical roller. After being etched, the inner ring and the cylindrical
roller both exhibited a mixture microstructure of fine martensite, carbides, and retained austenite, seen in Fig. 5(b) and (d).
After that, Vickers hardness (HV) of the surfaces of the inner ring raceway and the cylindrical roller was measured by micro
hardness tester with load of 9.8 N and holding time of 20 s. As listed in Table 2, the average hardness from four sites of the inner ring
raceway was nearly 25% lower than that of the cylindrical roller, which might be responsible for the fact that the former was severely
worn while the latter was almost intact, as displayed in Fig. 4 above.

2.3. 3D stereomicroscopy

Because of similarity of the five samples and limitation of the paper length, only the sample from area 3 (Fig. 4(c) and Fig. 6(a))
was taken as the representative for subsequent investigation. As shown in Fig. 6(b) and (c), indentations in sizes of several millimeters
and in shapes of dots, eggs and rabbit ears etc., and cracks in lengths of several centimeters were found existing on the original

27
Y. Gong et al. Engineering Failure Analysis 82 (2017) 26–38

Fig. 1. External appearances of the gearbox (a) total appearance (b) layout of the planetary gears (c) schematic diagram of the gearbox and the location of the failed
roller bearings (adopted from reference [52]).

Fig. 2. External appearances of the roller bearing (a) bearing inside the planetary gear (b) one unused bearing.

surfaces of the inner ring raceway, which should be the early-stage morphologies of wear in this case. Comparatively, Fig. 6(d) and
(e) exhibited the wear-out areas on the inner ring surface, in which ripples-shaped wear traces from the left to the right in Fig. 6(d),
and plow grooves from different directions in Fig. 6(e) were discovered, demonstrating the typical morphologies of abrasive wear
[56]. By utilizing the surface profiling function of the Hirox KH-7700 3D stereomicroscope for the boundary in Fig. 6(e), it could be
learned that the wear-out surface was pretty rough and its average depth was approximately 400 μm, seen in Fig. 6(f).
On the contrary, only indentations with smaller amount and smaller average size were detected on the surface of the cylindrical
roller, and in general no areas had been worn out, as exhibited in Fig. 7(a). However, a great number of wear traces could also be
noticed within these indentations (Fig. 7(b)), indicating that the rollers had suffered wear as well, but with a much lighter extent than
that of the inner ring raceway.

28
Y. Gong et al. Engineering Failure Analysis 82 (2017) 26–38

Fig. 3. Failed bearings in this 1.5 MW wind turbine (a) wear-out inner ring (b) fractured inner ring.

Fig. 4. External appearances of the failed roller bearing provided from the site.
(a) three wear-out areas on raceway of the inner ring (b) sample area 1 (c) sample areas 2 to 5 (d) eight cylindrical rollers.

Table 1
Chemical compositions of the roller bearing's inner ring and cylindrical roller (wt%).

Element C Si Mn P S Cr Mo Al Cu

100Cr6 0.93–1.05 0.15–0.35 0.25–0.45 ≤0.025 ≤ 0.015 1.35–1.60 ≤ 0.10 ≤ 0.050 ≤ 0.30
Inner ring 0.99 0.32 0.43 0.019 0.004 1.55 0.008 0.025 0.025
100CrMnSi6-4 0.93–1.05 0.45–0.75 1.00–1.20 ≤0.025 ≤ 0.015 1.40–1.65 ≤ 0.10 ≤ 0.050 ≤ 0.30
cylindrical roller 0.97 0.58 1.05 0.015 0.002 1.45 0.018 0.018 0.060

2.4. SEM and EDS

Fig. 8(a) and (b) respectively displayed the microscopic morphologies under SEM of the rabbit-ears-shaped indentation in
Fig. 6(b) and the ripples-shaped wear traces in Fig. 6(d). After magnification, it could be learned that such macroscopic ‘ripples’ were

29
Y. Gong et al. Engineering Failure Analysis 82 (2017) 26–38

Fig. 5. Metallographic structures of the roller bearing's inner ring and cylindrical roller.
(a) polished state 50 ×, and (b) etched state 100 ×, of the inner ring.
(c) polished state 50 ×, and (d) etched state 100 ×, of the cylindrical roller.

Table 2
Vickers hardness (HV) of the roller bearing's inner ring raceway and cylindrical roller.

Subcomponent Site A Site B Site C Site D Average

Inner ring raceway 534 526 538 543 535.25

cylindrical roller 700 684 702 708 698.50

essentially in the form of ‘needles’ from the microscopic point of view, and were also accompanied with impact dents and abundant
particles, seen in Fig. 8(c). Meanwhile, the surface was found peeled off in some areas (Fig. 8(d)), i.e. exhibiting the typical mor-
phology of abrasive wear again. Further magnified (Fig. 8(e)), these particles were actually in sizes not exceeding 10 μm, and
predominantly accumulated in the grooves between the ‘needles’ which had even cracked to some extent as well. With the assistance
of EDS, it was determined that the chemical compositions of these particles like the marked one in Fig. 8(f) were mainly composed of
carbon (C, 9.57 wt%), oxygen (O, 3.37 wt%), chromium (Cr, 1.40 wt%) and iron (Fe, 85.67 wt%), excluding the possibility of ag-
glomerated lubricant due to deterioration. However, considering the fact that the sample had been thoroughly cleaned with ethanol
in the ultrasonic cleaner before SEM, how were such particles introduced and/or generated must be figured out.
In terms of the cylindrical roller, morphology of the indentation on its surface under SEM (Fig. 9(a)) was not as distinct as that
under 3D SM (Fig. 7(b)). However, besides the parallel wear scratches, the plows-shaped wear traces with a lighter extent than that
on the inner ring raceway were also detected, seen in Fig. 9(b), which should be credited to the higher hardness of the cylindrical
roller surface based on the results in Table 2 above.

2.5. Lubricant inspection

Based on the macroscopic and microscopic morphologies observation, it could be basically concluded that abrasive wear, which
should not have occurred under normal service conditions, was the main cause of severe wear on the surface of the roller bearing's
inner ring raceway. Consequently, focus was then naturally turned to the quality of the lubricant that was used for lubrication
between the rings and the cylindrical rollers.
Fig. 10 displayed the gas chromatography-mass spectrometer results of the lubricant sampled by the owner on the site, from

30
Y. Gong et al. Engineering Failure Analysis 82 (2017) 26–38

Fig. 6. Macroscopic morphologies of the surface of the roller bearing's inner ring raceway under 3D SM.
(a) total morphology of sample area 3 (b) indentations in different shapes (c) multiple cracks.
(d) ripples-shaped wear traces (e) plow grooves (f) 3D profile of the wear-out surface.

which it could be learned that the lubricant predominantly consisted of long-chain alkanes with different numbers of carbon atoms
and different kinds of substituent groups. For instance, the highest peaks at retention time of 9.34, 10.19, 10.89, 11.20, 12.15, and
12.64 min corresponded to phenol (C22H18N2O2), 9-octyleicosane (C28H58), 7-hexyldocosane (C28H58), heptacosane (C27H56), tet-
ratetracontane (C44H90), and 11-decyltetracosane (C34H70) respectively. Since not a great deal of organic compounds that contained
carbonyl C]O, carboxylate eCOO– and ether C–O–C were detected, it could be preliminarily deduced that the lubricant had not
deteriorated seriously after use.
By means of Fourier transform infrared spectroscopy, the detailed functional groups of the lubricant were then determined. As
presented in Fig. 11, the absorption peaks in wavenumbers of 2955, 2921 and 2852 cm− 1 stood for the asymmetric and symmetric
stretching vibration of CeH bond in methyl eCH3 and methylene eCH2e for aliphatics, while that of 1377 and 1462 cm− 1 re-
spectively represented the symmetric and the asymmetric bending vibration of methyl eCH3 for aliphatics. Also, the peak in
721 cm− 1 implied that the number of methylene in e(CH2)ne was more than 4 [57,58]. These results accorded with the conclusion
from GC–MS that the lubricant was mainly composed of long-chain alkanes, but the existence of absorption peak in wavenumber
1740 cm− 1 (stretching vibration of carbonyl C]O) indicated that the lubricant had deteriorated to some extent due to oxidation.

31
Y. Gong et al. Engineering Failure Analysis 82 (2017) 26–38

Fig. 7. Macroscopic morphologies of the surface of the cylindrical roller under 3D SM.
(a) total morphology (b) one indentation with wear traces.

Finally, thermogravimetric as well as its derivative analysis (DTG, derivative thermogravimetric analysis) was conducted on the
lubricant. As exhibited in Fig. 12, evident reduction in weight of the lubricant started from nearly 250 °C, with maximum weight loss
rates at around 320 °C and 460 °C, which were all much higher than the actual service temperatures of the roller bearings. Appar-
ently, these phenomena further confirmed that the lubricant had not deteriorated severely during the use, in other words, un-
qualification and/or deterioration of the lubricant could be excluded for causing the abnormal wear of the roller bearings inner ring.

3. Failure analysis

Till now, it's pretty clear that the abnormal wear on the roller bearing's inner ring raceway should primarily ascribe to abrasive
wear. Furthermore, considering the existence of indentations on the original surfaces and a great number of particles on the wear-out
surfaces on the raceway, three-body abrasive wear could then be concluded as the detailed mechanism. Based on the macroscopic and
microscopic morphologies observation presented above, such three-body abrasive wear in our case can be basically divided into three
main stages.

3.1. Stage 1: formation of indentations and macro cracks

In this stage, indentations in sizes of several millimeters (Fig. 6(b) and Fig. 7(b)) and macro cracks in lengths of several cen-
timeters (Fig. 6(c)) were generated due to the compressive stresses from the third body that located between the other two bodies, i.e.
the inner ring and the cylindrical rollers of the roller bearing. Since the surface hardness of the former (the inner ring raceway) was
relatively lower, the amount and the average size and depth of the indentations was much greater than that of the latter (the
cylindrical roller), and macro cracks were also formed. This was the early stage of the three-body abrasive wear, just occurring after
the third body was introduced, and its mechanism seemed like impact on the surfaces of the other two bodies. However, it's not
difficult to infer that the average size of the third body should be in the range of several millimeters according to the sizes of the
indentations, much larger than the particle sizes (less than 10 μm) in the wear-out surfaces (Fig. 8). That is to say, the third body was
actually other foreign substance whose origin needed to be further identified, rather than the wear debris produced from the abrasive
wear.

3.2. Stage 2: localized plastic deformation

Once indentations were formed, such areas were then vulnerable to plastic deformation under the effect of subsequent three-body
abrasive wear, and simultaneously numerous wear debris in smaller sizes (101–103 μm) were produced from the crushed third body
and the base materials of the other two bodies, just like those presented in Fig. 8. Apparently, these debris would also serve as the
third body and further complicate the three-body abrasive wear in a localized level afterwards. From the theoretical point of view,
this procedure commonly involves wedge forming, cutting and plowing, and in our case, relevant morphologies were all observed in
Fig. 8(b), Fig. 8(d) and Fig. 8(e) respectively, and were also accompanied with those particles within the plowing grooves and the
cracks. In essence, this was the key stage of three-body abrasive wear and would develop in a self-acceleration manner, seeming like
the vicious circle illustrated in Fig. 13.

3.3. Stage 3: wear out of the surface

Under the continuous effect of above ‘vicious circle’, the affected areas on the other two bodies gradually grew wider and deeper
with removal of the base materials, and consequently brought about the wear-out surfaces in the sizes of several to a dozen cen-
timeters until being detected or fractured like in Fig. 3. But it is well known that the extent of wear is in a reverse relationship with
the hardness of materials, seen in Eq. 1, in which V is the wear volume, Ks stands for the dimensionless abrasive wear coefficient and
is commonly in the range of 10− 2–10− 3 for three-body abrasive wear, W and L respectively denotes the normal load and the sliding

32
Y. Gong et al. Engineering Failure Analysis 82 (2017) 26–38

Fig. 8. SEM micrographs of the surface of the roller bearing's inner ring raceway.
(a) The rabbit-ears-shaped indentation (b) the ripples-shaped wear traces.
(c) The needles-shaped wear traces and the impact dents after magnification (d) the peeled off areas.
(e) Microscopic cracks and particles (f) one marked particle for its EDS results.

distance of the third body, and H represents the hardness of the affected material surface. Therefore, this could be the reason why the
roller bearing's inner ring raceway with a lower hardness (535.25 HV) was severely worn while the cylindrical rollers with a higher
hardness (698.50 HV) were only embedded with light indentations.
WL
V = Ks
H (1)

So far, the three-body abrasive wear mechanism of the abnormal wear on the roller bearing's inner ring raceway had been
discussed in detail, and the current focus must be shifted to the origin of the third body. Since only one disassembled roller bearing
with failure was provided from the owner, our team particularly took a trip to the site for field investigation. As expected, metallic
debris in sizes of several centimeters were indeed discovered depositing on the cylindrical rollers of such roller bearings, seen in
Fig. 14, some of which with proper shapes and sizes would probably drop into the clearance between the inner ring and the cy-
lindrical rollers during service, and subsequently induce the three-body abrasive wear. Meanwhile, it was also found on the site that
the teeth of the planetary gears outside the roller bearings were severely broken, as revealed in Fig. 15. Now it had eventually come

33
Y. Gong et al. Engineering Failure Analysis 82 (2017) 26–38

Fig. 9. SEM micrographs of the surface of the cylindrical roller (a) one indentation (b) plows-shaped wear traces.

Fig. 10. GC–MS results of the lubricant used for the roller bearing.

Fig. 11. FT-IR results of the lubricant used for the roller bearing.

34
Y. Gong et al. Engineering Failure Analysis 82 (2017) 26–38

Fig. 12. TGA and DTG curves of the lubricant used for the roller bearing.

Fig. 13. Illustration of the development procedure of the key stage of three-body abrasive wear.

Fig. 14. Metallic debris on the cylindrical rollers of one roller bearing.

out that the third body in the three-body abrasive wear that occurred on the roller bearings inner ring raceway was actually the debris
of the broken gear teeth outside the roller bearings. For the purpose of verification, chemical compositions of such metallic debris
were then inspected, and proved to be in accordance with the requirements of 18CrNiMo7-6 low-carbon alloy steel specification in
the BS EN 10084:2008 standard [59] for gears of wind turbine gearbox [60], seen in Table 3. Also, the hardness of the metallic debris

35
Y. Gong et al. Engineering Failure Analysis 82 (2017) 26–38

Fig. 15. External appearance of the broken teeth of one gear.

Table 3
Chemical compositions of the metallic debris on the roller bearing's cylindrical roller (wt%).

Element C Si Mn P S Cr Mo Ni

18CrNiMo7-6 0.15–0.21 ≤0.40 0.50–0.90 ≤ 0.025 ≤ 0.035 1.50–1.80 0.25–0.35 1.40–1.70

Metallic debris 0.20 0.25 0.63 0.006 0.002 1.60 0.28 1.65

was measured (638.25 HV), and found to be coincidently between the inner ring raceway (535.25 HV) and the cylindrical roller
(698.50 HV), which could perfectly explain the reason why the former was severely worn while the latter was only embedded with
lighter and less indentations. Now the root cause of the severely worn inner ring raceway of the failed roller bearing due to three-
body abrasive wear was identified, i.e. it was essentially the consequence of broken of the planetary gear teeth outside the roller
bearing. In other words, the gear teeth that failed in advance possibly because of unqualified manufacture, improper installation, and
irregular operation etc. should be primarily blamed for, which would certainly be the contents of another failure analysis case study.

4. Conclusions and recommendations

1. Properties of the base materials of the failed roller bearing's inner ring and cylindrical rollers were qualified, and the lubricant had
not basically deteriorated as well.
2. Mechanism of the abnormal severe wear on the roller bearing's inner ring raceway was three-body abrasive wear, and it would
develop in a self-acceleration manner.
3. Introduction of the metallic debris from the broken gear teeth outside the roller bearing was the origin of the third body, which
should be the root cause of this failure.
4. The relatively lower hardness of the roller bearing's inner ring was the other important cause of severe wear on its raceway surface
under three-body abrasive wear.

Based on the failure analysis results presented above, several countermeasures were then proposed as follows.

1. Quality of all the components in the wind turbine must be fully controlled during manufacture, installation, operation and
maintenance to avoid broken parts and debris being introduced into the roller bearings.
2. The roller bearing's inner ring raceway surfaces are suggested to be properly finished for hardness increase, which will effectively
improve its resistance to wear in case that three-body abrasive wear is induced again.
3. The conditions of the lubricant including temperature, color, etc. should be strictly monitored to ensure a good lubrication

36
Y. Gong et al. Engineering Failure Analysis 82 (2017) 26–38

between the inner/outer rings and the cylindrical rollers of the roller bearings.

References

[1] Wind Power by Country, Wikipedia, https://en.wikipedia.org/wiki/Wind_power_by_country, (2017).

[2] S. Zhang, P. Andrews-Speed, X. Zhao, Political and institutional analysis of the successes and failures of China's wind power policy, Energy Policy 56 (2013)
331–340.
[3] Y. Gong, Z.G. Yang, F.Y. Yang, Heat strength evaluation and microstructures observation of the welded joints of one China-made T91 steel, J. Mater. Eng.
Perform. 21 (2012) 1313–1319.
[4] Y. Gong, Z.G. Yang, Corrosion evaluation of one dry desulfurization equipment - circulating fluidized bed boiler, Mater. Des. 32 (2011) 671–681.
[5] Y. Gong, J. Cao, L.N. Ji, C. Yang, C. Yao, Z.G. Yang, et al., Assessment of creep rupture properties for dissimilar steels welded joints between T92 and HR3C,
Fatigue Fract. Eng. Mater. Struct. 34 (2011) 83–96.
[6] H. Arabian-Hoseynabadi, H. Oraee, P.J. Tavner, Failure modes and effects analysis (FMEA) for wind turbines, Int. J. Electr. Power Energy Syst. 32 (2010)
817–824.
[7] J.M.P. Perez, F.P.G. Marquez, A. Tobias, M. Papaelias, Wind turbine reliability analysis, Renew. Sustain. Energy Rev. 23 (2013) 463–472.
[8] Z.Q. Li, S.J. Chen, H. Ma, T. Feng, Design defect of wind turbine operating in typhoon activity zone, Eng. Fail. Anal. 27 (2013) 165–172.
[9] C. Su, Y. Yang, X.L. Wang, Z.Y. Hu, Failures analysis of wind turbines: Case study of a Chinese wind farm, ‘Prognostics and System Health Management
Conference’, Chengdu, Sichuan Province, China, 2016.
[10] J.C. Marín, A. Barroso, F. París, J. Cañas, Study of fatigue damage in wind turbine blades, Eng. Fail. Anal. 16 (2009) 656–668.
[11] J.S. Chou, W.T. Tu, Failure analysis and risk management of a collapsed large wind turbine tower, Eng. Fail. Anal. 18 (2011) 295–313.
[12] J.S. Chou, C.K. Chiu, I.K. Huang, K.N. Chi, Failure analysis of wind turbine blade under critical wind loads, Eng. Fail. Anal. 27 (2013) 99–118.
[13] S. Ataya, M.M.Z. Ahmed, Damages of wind turbine blade trailing edge: forms, location, and root causes, Eng. Fail. Anal. 35 (2013) 480–488.
[14] X. Chen, W. Zhao, X.L. Zhao, J.Z. Xu, Preliminary failure investigation of a 52.3 m glass/epoxy composite wind turbine blade, Eng. Fail. Anal. 44 (2014)
345–350.
[15] J.A. Vargas, J.E. Wilches, H.A. Gómez, J.A. Pacheco, R.J. Hernandez, Analysis of catastrophic failure of axial fan blades exposed to high relative humidity and
saline environment, Eng. Fail. Anal. 54 (2015) 74–89.
[16] X. Chen, J.Z. Xu, Structural failure analysis of wind turbines impacted by super typhoon Usagi, Eng. Fail. Anal. 60 (2016) 391–404.
[17] B. Amirzadeh, A. Louhghalam, M. Raessi, M. Tootkaboni, A computational framework for the analysis of rain-induced erosion in wind turbine blades, part I:
stochastic rain texture model and drop impact simulations, J. Wind Eng. Ind. Aerodyn. 163 (2017) 33–43.
[18] B. Amirzadeh, A. Louhghalam, M. Raessi, M. Tootkaboni, A computational framework for the analysis of rain-induced erosion in wind turbine blades, part II:
drop impact-induced stresses and blade coating fatigue life, J. Wind Eng. Ind. Aerodyn. 163 (2017) 44–54.
[19] W.X. Yang, P.J. Tavner, C.J. Crabtree, Y. Feng, Y. Qiu, Wind turbine condition monitoring: technical and commercial challenges, Wind Energy 17 (2014)
673–693.
[20] A.R. Nejad, P.F. Odgaard, Z. Gao, T. Moan, A prognostic method for fault detection in wind turbine drivetrains, Eng. Fail. Anal. 42 (2014) 324–336.
[21] Y.N. Qiu, Y.H. Feng, J. Sun, W.X. Zhang, D. Infield, Applying thermophysics for wind turbine drivetrain fault diagnosis using SCADA data, IET Renew. Power
Gener. 10 (2016) 661–668.
[22] Z. Li, X. Yan, X. Wang, Z. Peng, Detection of gear cracks in a complex gearbox of wind turbines using supervised bounded component analysis of vibration signals
collected from multi-channel sensors, J. Sound Vib. 371 (2016) 406–433.
[23] J.M. Ha, H. Oh, J. Park, B.D. Youn, Classification of operating conditions of wind turbines for a class-wise condition monitoring strategy, Renew. Energy 103
(2017) 594–605.
[24] W. Teng, X. Ding, Y. Zhang, Y. Liu, Z. Ma, A. Kusiak, Application of cyclic coherence function to bearing fault detection in a wind turbine generator under
electromagnetic vibration, Mech. Syst. Signal Process. 87 (Part A) (2017) 279–293.
[25] J. Ribrant, L.M. Bertling, Survey of failures in wind power systems with focus on Swedish wind power plants during 1997–2005, IEEE Trans. Energy Convers. 22
(2007) 167–173.
[26] Y. Lin, L. Tu, H. Liu, W. Li, Fault analysis of wind turbines in China, Renew. Sust. Energ. Rev. 55 (2016) 482–490.
[27] M.H. Evans, J.C. Walker, C. Ma, L. Wang, R.J.K. Wood, A FIB/TEM study of butterfly crack formation and white etching area (WEA) microstructural changes
under rolling contact fatigue in 100Cr6 bearing steel, Mater. Sci. Eng. A 570 (2013) 127–134.
[28] S. Janakiraman, O. West, P. Klit, N.S. Jensen, Observations of the effect of varying hoop stress on fatigue failure and the formation of white etching areas in
hydrogen infused 100Cr6 steel rings, Int. J. Fatigue 77 (2015) 128–140.
[29] B. Gould, A. Greco, Investigating the process of white etching crack initiation in bearing steel, Tribol. Lett. 62 (2016) 26.
[30] Y.J. Li, M. Herbig, S. Goto, D. Raabe, Atomic scale characterization of white etching area and its adjacent matrix in a martensitic 100Cr6 bearing steel, Mater.
Charact. 123 (2017) 349–353.
[31] V. Šmeļova, A. Schwedt, L. Wang, W. Holweger, J. Mayer, Microstructural changes in white etching cracks (WECs) and their relationship with those in dark
etching region (DER) and white etching bands (WEBs) due to rolling contact fatigue (RCF), Int. J. Fatigue 100 (2017) 148–158.
[32] W. Teng, X. Ding, X. Zhang, Y. Liu, Z. Ma, Multi-fault detection and failure analysis of wind turbine gearbox using complex wavelet transform, Renew. Energy 93
(2016) 591–598.
[33] Q. Wang, Y. Zhu, Z. Zhang, C. Fu, C. Dong, H. Su, Partial load: A key factor resulting in the failure of gear in the wind turbine gearbox, J. Fail. Anal. Prev. 16
(2016) 109–122.
[34] Z. Zhang, Z. Yin, T. Han, A.C.C. Tan, Fracture analysis of wind turbine main shaft, Eng. Fail. Anal. 34 (2013) 129–139.
[35] J. Herrmann, T. Rauert, P. Dalhoff, M. Sander, Fatigue and fracture mechanical behaviour of a wind turbine rotor shaft made of cast iron and forged steel,
Procedia Structural Integrity 2 (2016) 2951–2958.
[36] T. Rauert, J. Herrmann, P. Dalhoff, M. Sander, Fretting fatigue induced surface cracks under shrink fitted main bearings in wind turbine rotor shafts, Procedia
Structural Integrity 2 (2016) 3601–3609.
[37] S. Sankar, M. Nataraj, V.P. Raja, Failure analysis of shear pins in wind turbine generator, Eng. Fail. Anal. 18 (2011) 325–339.
[38] B. Gould, A. Greco, K. Stadler, X. Xiao, An analysis of premature cracking associated with microstructural alterations in an AISI 52100 failed wind turbine
bearing using X-ray tomography, Mater. Des. 117 (2017) 417–429.
[39] Y. Gong, J. Cao, X.H. Meng, Z.G. Yang, Pitting corrosion on 316L pipes in terephthalic acid (TA) dryer, Mater. Corros. 60 (2009) 899–908.
[40] Y. Gong, J. Zhong, Z.G. Yang, Failure analysis of bursting on the inner pipe of a jacketed pipe in a tubular heat exchanger, Mater. Des. 31 (2010) 4258–4268.
[41] Y. Gong, C. Yang, C. Yao, Z.G. Yang, Acidic/caustic alternating corrosion on carbon steel pipes in heat exchanger of ethylene plant, Mater. Corros. 62 (2011)
967–978.
[42] Z.G. Yang, Y. Gong, J.Z. Yuan, Failure analysis of leakage on titanium tubes within heat exchangers in a nuclear power plant. Part I: electrochemical corrosion,
Mater. Corros. 63 (2012) 7–17.
[43] Y. Gong, Z.G. Yang, J.Z. Yuan, Failure analysis of leakage on titanium tubes within heat exchangers in a nuclear power plant. Part II: mechanical degradation,
Mater. Corros. 63 (2012) 18–28.
[44] Y. Gong, Z.G. Yang, X.H. Meng, Failure analysis of one peculiar 'Yin-Yang' corrosion morphology on heat exchanger tubes in purified terephthalic acid (PTA)
dryer, Eng. Fail. Anal. 31 (2013) 203–210.
[45] F.J. Chen, C. Yao, Z.G. Yang, Failure analysis on abnormal wall thinning of heat-transfer titanium tubes of condensers in nuclear power plant part I: corrosion and
wear, Eng. Fail. Anal. 37 (2014) 29–41.

37
Y. Gong et al. Engineering Failure Analysis 82 (2017) 26–38

[46] F.J. Chen, C. Yao, Z.G. Yang, Failure analysis on abnormal wall thinning of heat-transfer titanium tubes of condensers in nuclear power plant part II: erosion and
cavitation corrosion, Eng. Fail. Anal. 37 (2014) 42–52.
[47] Y.Y. Ma, S. Yan, Z.G. Yang, G.S. Qi, X.Y. He, Failure analysis on circulating water pump of duplex stainless steel in 1000 MW ultra-supercritical thermal power
unit, Eng. Fail. Anal. 47 (Part A) (2015) 162–177.
[48] Z.G. Yang, Y. Gong, Chapter 16 - failure analysis of heat exchanger tubes in petrochemical industry: microscopic analysis approach, Handbook of Materials
Failure Analysis with Case Studies from the Oil and Gas Industry, Butterworth-Heinemann, 2016, pp. 329–352.
[49] Q. Ding, X.F. Tang, Z.G. Yang, Failure analysis on abnormal corrosion of economizer tubes in a waste heat boiler, Eng. Fail. Anal. 73 (2017) 129–138.
[50] J. Yuan, C. Na, Y. Xu, C. Zhao, Wind turbine manufacturing in China: a review, Renew. Sust. Energ. Rev. 51 (2015) 1235–1244.
[51] X. Chen, C. Li, J. Xu, Failure investigation on a coastal wind farm damaged by super typhoon: A forensic engineering study, J. Wind Eng. Ind. Aerodyn. 147
(2015) 132–142.
[52] Y.H. Feng, Y.N. Qiu, C.J. Crabtree, H. Long, P.J. Tavner, Monitoring wind turbine gearboxes, Wind Energy 16 (2013) 728–740.
[53] Heat-Treated Steels, Alloy Steels And Free-Cutting Steels Part 17: Ball and Roller Bearing Steels, International Organization for Standardization, Geneva,
Switzerland, 2014.
[54] GB/T 18254-2016, High-Carbon Chromium Bearing Steel, China Standards Publisher, Beijing, China, 2016.
[55] Steel - Determination of Content of Non-Metallic Inclusions - Micrographic Method Using Standard Diagrams, International Organization for Standardization,
Geneva, Switzerland, 2013.
[56] L.H. Sun, Z.G. Yang, X.H. Li, Mechanical and tribological properties of polyoxymethylene modified with nanoparticles and solid lubricants, Polym. Eng. Sci. 48
(2008) 1824–1832.
[57] Y. Gong, S.M. Hu, X.L. Yang, J.L. Fei, Z.G. Yang, et al., Comparative study on degradation of ethylene-propylene rubber for nuclear cables from gamma and beta
irradiation, Polym. Test. 60 (2017) 102–109.
[58] Y. Gong, Z.G. Yang, Fracture failure analysis of automotive accelerator pedal arms with polymer matrix composite material, Compos. Part B 53 (2013) 103–111.
[59] BS EN 10084, Case Hardening Steels - Technical Delivery Conditions, European Committee for Standardization, Brussels, Belgium, 2008.
[60] GB/T 19073-2008, Gearbox of Wind Turbine Generator System, China Standards Publisher, Beijing, China, 2008.

38