Advertisement

Journal of Iron and Steel Research International

, Volume 25, Issue 1, pp 108–119 | Cite as

Effect of laser shock peening on combined low- and high-cycle fatigue life of casting and forging turbine blades

  • Cao Chen
  • Xiao-yong Zhang
  • Xiao-jun Yan
  • Jun Ren
  • Da-wei Huang
  • Ming-jing Qi
Original Paper
  • 43 Downloads

Abstract

Laser shock peening (LSP) is a novel effective surface treatment method to improve the fatigue performance of turbine blades. To study the effect of LSP on combined low- and high-cycle fatigue (CCF) life of turbine blades, the CCF tests were conducted at elevated temperatures on two types of full-scale turbine blades, which were made of K403 by casting and GH4133B by forging. Probabilistic analysis was conducted to find out the effect of LSP on fatigue life of those two kinds of blades. The results indicated that LSP extended the CCF life of both casting blades and forging blades obviously, and the effect of LSP on casting blades was more evident; besides, a threshold vibration stress existed for both casting blades and forging blades, and the CCF life tended to be extended by LSP only when the vibration stress was below the threshold vibration stress. Further study of fractography was also conducted, indicating that due to the presence of compressive residual stress and refined grains induced by LSP, the crack initiation sources in LSP blades were obviously less, and the life of LSP blades was also longer; since the compressive residual stress was released by plastic deformation, LSP had no effect or adverse effect on CCF life of blade when the vibration stress of blade was above the threshold vibration stress.

Keywords

Laser shock peening Combined low- and high-cycle fatigue life (CCF) Full-scale turbine blade SN curve Threshold vibration stress 

1 Introduction

Gas turbine blades are an indispensable component of aero-engine, which have to endure various loads (such as thermal load, centrifugal load, aerodynamic load and vibration load) during the operation of aero-engine. The harsh service environment gives rise to failures of turbine blades. Therefore, how to extend the life of turbine blades has aroused general interest in the field of aero-engine.

Laser shock peening (LSP) is a novel effective surface treatment method for improving the surface performance of metallic materials. Laser-induced high-energy shock wave is used to induce compressive residual stresses in the impacted area of metal in LSP treatment, which effectively improves the fatigue life of treated zone [1, 2]. At present, LSP has been successfully adopted in the manufacture of fans, compressor blades, and blisks and effectively extends the fatigue life of those components [3]. However, the effect of LSP on fatigue performances of turbine blades has seldom been reported at present.

Current research about the effect of LSP on fatigue performances mainly focuses on the single fatigue failure mode (pure high- or low-cycle fatigue) [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15]. Yang et al. [4] investigated the fatigue crack growth rates of 2024-T3 aluminum alloy treated by LSP and found that LSP was an effective surface treatment method for improving the fatigue performance of 2024-T3 aluminum alloy. Hong [8] studied the effect of laser energy and intensification time on the low-cycle fatigue of Ti17 titanium alloy. Their results showed that the roughness and residual stress were not adversely affected by laser energy and intensification time; the fatigue life of Ti17 increased with the rise of laser energy and intensification time. Altenberger et al. [12] treated the Ti–6Al–4V titanium alloy specimens using LSP and conducted fatigue tests on those specimens. They found that the nano-crystalline structures and high-density dislocation entanglements in the surface hardened layer still existed in the high-temperature environment, which ensured that the LSP strengthening effect could still play a role in the high-temperature environment.

Since the turbine blades endure both low-cycle loads (centrifugal load, bending load) and high-cycle loads (aerodynamic load, vibration load) simultaneously [16, 17, 18, 19, 20], it is necessary to take both low-cycle loads and high-cycle loads, as well as their coupling effect into consideration in the fatigue life evaluation of turbine blades. Consequently, it is necessary to study the effect of LSP on the combined low- and high-cycle fatigue (CCF) life of turbine blades.

Currently, the manufacturing methods of turbine blades are mainly casting and forging. The microstructures of casting blades and forging blades are different due to the entirely different manufacturing processes. The reason that LSP can improve the fatigue performance is the compressive residual stresses induced by laser impact. However, different microstructures are likely to affect the formation of compressive residual stresses and further affect the improvement in fatigue performance. Therefore, it is necessary to find out the effect of LSP on the fatigue performance of turbine blades manufactured by casting and forging.

Accordingly, it is the aim of this work to study the effect of LSP on CCF life of full-scale turbine blades manufactured by casting and forging.

2 Experimental procedures

2.1 Specimens and critical location selection

The turbine blades used in this study are taken from turbines of an in-service aero-engine. Blades from the second stage (referred to as “casting blade”) are cast with nickel-base superalloy K403. (Chemical compositions are shown in Table 1.) The heat treatments of those blades are carried out at 1210 °C in a pit-type gas carburizing furnace for 4 h. Subsequently, the blades are cooled in air. After above-mentioned treatments, the microstructures of casting blades are composed of γ phase, γ′ phase, (γ + γ′) eutectic and MC carbide. The γ′ phase is the main strengthening phase of casting blades, of which the mass fraction is 57%. The volume fraction of (γ + γ′) eutectic is 2%.
Table 1

Chemical compositions of GH4133B and K403 (wt%)

Material

Chemical composition

GH4133B

C

Cr

Al

Ti

Fe

Nb

Mg

Zr

B

Ce

Mn

Si

0.06

20.50

1.45

2.75

1.50

1.50

0.005

0.05

0.01

0.01

0.35

0.65

P

S

Cu

Bi

Sn

Sb

Pb

As

Ni

0.015

0.007

0.07

0.0001

0.0012

0.0025

0.001

0.0025

Balance

K403

C

Cr

Co

W

Mo

Al

Ti

Fe

B

Zr

Ce

Mn

0.14

11.00

5.2

5.15

4.15

5.60

2.60

2.0

0.017

0.05

0.01

0.50

Si

P

S

Pb

Sb

Bi

Sn

As

Ni

0.50

0.02

0.01

0.001

0.001

0.0001

0.002

0.005

Balance

The third stage turbine blades (referred to as “forging blade”) are made of nickel-base superalloy GH4133B. (Chemical compositions are shown in Table 1.) The heat treatments of forging blades are carried out at 1080 °C for 8 h followed by air-cooling. Subsequently, the blades are heated at 750 °C for 16 h followed by air-cooling. The microstructures of forging blades are composed of γ phase, γ′ phase, MC and M23C6 carbides. The γ′ phase is also the main strengthening phase of forging blades, of which the mass fraction is 14%.

Both casting blades and forging blades are of parallelogram tip shroud with the labyrinth seals on it. The root of the blade is unilateral serration, and all the blades are installed in pairs into the fir-tree grooves on disk (Fig. 1).
Fig. 1

Blades installed in pairs into fir-tree grooves on disk

Unlike the standard specimens (e.g., dog-bone specimens) which have a pre-designed gauge section (or critical location) and failure can occur in this section, the real turbine blade specimens have a rather complicated geometric profile and thus the critical location of blades needs to be known. In general, the critical location can be determined by both failure statistics in the outfield and numerical simulation. Both casting blades and forging blades were found fractured in the outfield. The fracture occurred in the location of first serration (Fig. 2). Moreover, the finite element analysis results indicated that the maximum stress of these two types of blades under real flight condition also occurred in the first serration (Fig. 3). Consequently, the first serration was selected to be the critical location of those two types of blades.
Fig. 2

Blade ruptured in the first serration in the outfield

Fig. 3

Stress distribution of casting blade and forging blade under real flight condition. a Casting blade; b forging blade

2.2 LSP treatment parameters

To extend the fatigue life of those two types of blades, the fatigue performance of the weakest area (the serrations) should be enhanced. LSP was used to strengthen the serrations of those blades (Fig. 4), and detailed LSP treatment parameters are listed in Table 2. The blades treated by LSP are referred to as “LSP blades” in the following text. Correspondingly, the untreated blades are referred to as “untreated blades.”
Fig. 4

Serrations of blades treated by LSP

Table 2

Parameters of LSP treatment for blades

Parameter

Value

Laser power density/(GW cm−2)

5.8

Laser wavelength/nm

1064

Diameter of spot/mm

4

Pulse width/ns

20

Times of shock

2

Overlapping rate/%

60

2.3 Test conditions and experimental setup

According to the typical working conditions of blades, the test conditions in this study are shown in Table 3. Since it is difficult to determine the high-cycle loads, several vibration amplitudes are used in this study based on the inverse method [21]. The schematic diagram of the inverse method is shown in Fig. 5 (in which σv represents the vibration stress and N is the high-cycle fatigue cycles). The actual range of vibration stress can be estimated by comparing the flight failure data of blades and σvN curve in Fig. 5. For most vibration amplitudes, only 1 specimen is tested to get the σvN curve. However, for some specific amplitudes, 4 or more specimens are tested to take into consideration of scatter.
Table 3

Test conditions and corresponding specimen number used in this study

Type

Low-cycle load/kN

Ratio of minor cycle per major cycle

Temperature/°C

LSP

Vibration amplitude/mm

Number

Casting blades

43.6

5000

510

Yes

1.4

1

1.5

6

1.6

1

1.9

1

2.0

5

No

0.8

1

1.0

1

1.4

1

1.5

7

2.0

6

2.5

1

Forging blades

46.2

5000

530

Yes

1.4

1

1.6

1

1.7

5

1.9

1

2.1

4

No

1.2

1

1.6

1

1.7

5

1.9

1

2.0

1

2.1

4

2.2

1

Fig. 5

Schematic diagram of inverse method

A CCF test system shown in Figs. 6 and 7 was established. As shown in Fig. 6, the fixture used in this study can add both low-cycle load and high-cycle load on the blade simultaneously. The test system is mainly composed of a hydraulic testing machine, an outer clamp, an inner clamp and an electromagnetic exciter. The low-cycle load is not added on the blade directly, and it is added on the top of the outer clamp (the n-shaped structure in Fig. 6) and transmits to blade through the inner clamp following a path marked by red arrows in Fig. 6. The electromagnetic exciter, which is linked to the inner clamp, is used to generate the high-cycle vibration load. To ensure that the inner clamp can move freely during vibration, two bearings are used to reduce the friction between the outer clamp and inner clamp.
Fig. 6

Fixture used in CCF tests to apply low-cycle load and high-cycle vibration load on blades

Fig. 7

CCF test system composed of a fixture, a hydraulic test machine, an electromagnetic exciter and an electron microscope

According to the test system shown in Fig. 6, the level of high-cycle load can be controlled by limiting the amplitude of electromagnetic exciter. However, the relation between the vibration amplitude and vibration stress on the critical location is unknown. Finite element method was used to analyze the stress distribution of blade under different vibration amplitudes [22], and the relation between vibration stress and vibration amplitude on the critical location was obtained (Fig. 8).
Fig. 8

Relation between vibration stress and vibration amplitude on critical location

3 Results and discussion

3.1 Test results

For the casting blades, 31 valid data were obtained, 14 of which were obtained from LSP blades. (Some of the ruptured blades are shown in Fig. 9a.) All the blades are ruptured from the first serration, which coincides with the critical location determined above. The full test results are listed in Table 4.
Fig. 9

Some of ruptured blades in CCF tests. a Casting blades; b forging blades

Table 4

Experimental data of casting blades under various vibration amplitudes

No.

LSP blades

Amplitude/mm

Vibration stress/MPa

LCF life

HCF life

A1

1.4

114.3

133

665,000

A2

1.5

122.5

41

205,000

A3

49

245,000

A4

49

245,000

A5

129

645,000

A6

230

1,150,000

A7

261

1,305,000

A8

1.6

130.6

656

3,280,000

A9

1.9

155.1

22

110,000

A10

2.0

163.3

30

150,000

A11

31

155,000

A12

33

165,000

A13

33

165,000

A14

39

195,000

No.

Untreated blades

Amplitude/mm

Vibration stress/MPa

LCF life

HCF life

B1

0.8

65.3

2687

13,435,000

B2

1.0

81.6

443

2,215,000

B3

1.4

114.3

412

2,060,000

B4

1.5

122.5

25

125,000

B5

30

150,000

B6

31

155,000

B7

68

340,000

B8

144

720,000

B9

173

865,000

B10

356

1,780,000

B11

2.0

163.3

28

140,000

B12

31

155,000

B13

63

315,000

B14

72

360,000

B15

80

400,000

B16

211

1,055,000

B17

2.5

204.1

11

55,000

For the forging blades, 26 valid data were obtained and 12 among them were LSP blades. (Some of the ruptured blades are shown in Fig. 9b.) Similarly, all the blades are ruptured from the critical location. Full results can be found in Table 5.
Table 5

Experimental data of forging blades under various vibration amplitudes

No.

LSP blades

Amplitude/mm

Vibration stress/MPa

LCF life

HCF life

C1

1.4

138.7

296

1,480,000

C2

1.6

157.0

275

1,315,000

C3

1.7

168.4

53

265,000

C4

63

315,000

C5

96

480,000

C6

157

785,000

C7

217

1,085,000

C8

1.9

188.2

47

235,000

C9

2.1

208.1

24

120,000

C10

29

145,000

C11

41

205,000

C12

81

405,000

No.

Untreated blades

Amplitude/mm

Vibration stress/MPa

LCF life

HCF life

D1

1.2

118.9

515

2,575,000

D2

1.6

157.0

170

850,000

D3

1.7

168.4

37

185,000

D4

56

280,000

D5

63

315,000

D6

112

560,000

D7

251

1,255,000

D8

1.9

188.2

61

305,000

D9

2.0

198.2

79

395,000

D10

2.1

208.1

18

90,000

D11

31

155,000

D12

50

250,000

D13

115

575,000

D14

2.2

218.0

59

295,000

3.2 Active effect of LSP on fatigue life and scatter

The test results of casting blades under 1.5 mm amplitude and 2.0 mm amplitude are plotted on logarithmic normal probability diagram, respectively (as shown in Figs. 10, 11). The lines in the figures represent the results obtained from linear regression estimation using a statistical software (lognormal, SMITH (TM) V 5.15A). The estimation distribution parameters and coefficients are also reported in these figures. As can be seen, the correlation coefficients r2 are relatively large, indicating that the test results can be well described by the lognormal distribution model.
Fig. 10

Distribution of fatigue life of casting blades under 1.5 mm amplitude

Fig. 11

Distribution of fatigue life of casting blades under 2.0 mm amplitude

The safe lives (which were calculated under the reliability of 99.87%, confidence level of 95%) of casting blades under 1.5 mm amplitude and 2.0 mm amplitude are shown in Table 6. The results show that the safe life is significantly increased by LSP. The safe life of LSP blades is 1.63 times larger than that of untreated blades under 1.5 mm amplitude, and the multiple is 1.56 under 2.0 mm amplitude. The ratio of maximum cycles and minimum cycles, which is used to represent the scatter, is reduced by 55% under 1.5 mm amplitude and 83% under 2.0 mm amplitude.
Table 6

Safe life and scatter of casting blades under 1.5 mm amplitude and 2.0 mm amplitude

Amplitude/mm

Life and scatter

LSP blades

Untreated blades

Ratio of LSP to untreated

1.5

Safe life

189,150

116,153

1.63

Maximum/minimum

6.36

14.24

0.45

2.0

Safe life

144,828

92,601

1.56

Maximum/minimum

1.30

7.54

0.17

In addition, by comparing the test data of casting blades under 1.5 mm amplitude and 2.0 mm amplitude, it can be seen that the scatter of casting blades decreases significantly and the safe life is increased after treated by LSP, which means that LSP has a strengthening effect on the CCF life of casting blades. Meanwhile, the strengthening effect is more obvious when the vibration stress is lower.

For the forging blades, the data of 1.7 mm amplitude and 2.1 mm amplitude are processed by the same method (as shown in Figs. 12, 13). The safe lives and scatters are shown in Table 7. The results indicate that the safe life is increased evidently by LSP both under 1.7 mm amplitude and 2.1 mm amplitude. The safe life increases by a factor of 1.32 under 1.7 mm amplitude and by a factor of 1.47 under 2.1 mm amplitude. Meanwhile, the scatter of forging blades is also decreased by LSP, but it is not as obvious as the casting blades.
Fig. 12

Distribution of fatigue life of forging blades under 1.7 mm amplitude

Fig. 13

Distribution of fatigue life of forging blades under 2.1 mm amplitude

Table 7

Safe life and scatter of forging blades under 1.7 mm amplitude and 2.1 mm amplitude

Amplitude/mm

Life and scatter

LSP treated blades

Untreated blades

Ratio of LSP to untreated

1.7

Safe life

189,806

143,598

1.32

Maximum/minimum

4.09

6.78

0.60

2.1

Safe life

88,636

60,174

1.47

Maximum/minimum

3.31

6.41

0.52

3.3 Effect of LSP on SN curve

The median SN curve, which shows the relation between vibration stress and fatigue life under a 50% survival probability, is a useful method to evaluate the fatigue performances of materials [23, 24, 25]. Based on the CCF test results of casting blades, the fitting curves obtained in double logarithmic coordinates by the least-square method are shown in Fig. 14. As can be seen evidently in the figure, the two curves intersect at a vibration stress of 131 MPa (mentioned as “threshold vibration stress” in this paper), below which the LSP shows an active effect on the CCF life of blades, while above which the LSP shows no effect or even a negative effect on the CCF life.
Fig. 14

Median SN curves of casting blades treated and untreated by LSP

Similarly, for the forging blades, the median SN curves of LSP blades and untreated blades are shown in Fig. 15. Obviously, a threshold vibration stress of 170 MPa can also be found. When the vibration stress is below the threshold stress, the CCF life is extended by LSP treatment. However, the CCF life decreases when the vibration stress is above the threshold stress.
Fig. 15

Median SN curve of forging blades treated and untreated by LSP

To find out the reason why active or negative effect can be found below or above the threshold vibration stress after LSP treatment, the relation between threshold vibration stress and yield stress was studied. Stresses that may affect the fatigue life are listed in Table 8, and it shows that the total stress (the sum of the threshold vibration stress and static stress) is very close to the yield stress, which means that the LSP treatment will have no effect or adverse effect on the CCF life of blades when the total stress is close to or surpasses the yield stress.
Table 8

Relation between threshold vibration stress and yield stress of experimental materials

Item

Casting blades

Forging blades

Material

K403

GH4133B

Static stressa/MPa

701

628

Vibration stress/MPa

131

170

Total stress/MPa

832

798

σ 0.2 b /MPa

819

731

aThe yield stresses at 0.2% offset were got at 510 °C, which is the temperature used in CCF tests

bStatic stress is the stress generated by the low-cycle load reported in Table 3

3.4 Fractography

A systematic fractographic analysis of the fractured blades was carried out. Figure 16 shows the macrofracture of both casting blades and forging blades. The specimen numbers (coinciding with the numbers in Tables 4, 5) and vibration amplitude are listed on the right of the figure. The red arrows indicate the crack initiation sources. As can be seen from the figure, LSP blades have less crack initiation sources than untreated blades under the same vibration amplitude. According to the research of Yi et al. [26], the fatigue crack always initiates from the defects in the material. Due to the shocks of LSP treatment, the defects on the surface of the blade are inhibited by the compressive stress. Therefore, the crack initiation sources in LSP blades are obviously less.
Fig. 16

Macrofracture of casting blades and forging blades. The specimen number and vibration amplitude are listed in yellow. Red arrows indicate the crack initiation sources

In addition, the refined grains induced by LSP can also effectively inhibit the propagation of the crack. Figure 17 shows the crack propagation region of LSP blade (A3) and untreated blade (B10). The observed areas of LSP blade and untreated blade are 0.8 mm away from the crack initiation source. Based on the measurement results of fatigue striation marked in Fig. 17, the fatigue striation width of LSP blade is smaller than that of the untreated blade. Therefore, due to the presence of compressive residual stress and refined grains, the LSP has an active effect on the extension of fatigue life.
Fig. 17

Crack propagation region of LSP blade (A3) and untreated blade (B10). The measurement results of fatigue striation are reported in the figure

In order to study the mechanism of threshold vibration stress, the crack initiation regions of A3, A14, B10 and B16 were observed by using a scanning electron microscope (Fig. 18). The cracks of untreated blades (B10 and B16) initiate from the surface of blades. Due to the presence of compressive residual stress, the maximum stress occurs in the subsurface of the blade. Therefore, the cracks of LSP blade initiate from the subsurface when vibration stress is below the threshold value (A3). Similar results are also obtained in the researches of Liu et al. [27] and Jiang et al. [28]. However, the abnormal phenomenon arises in A14 (LSP blade), of which the cracks initiate from the surface. Since the vibration stress of A14 is above the threshold vibration stress, the total stress of A14 is above the yield stress. According to the research of Benedetti et al. [29], the residual stress induced by shot peening is released when the material plastic flow stress was achieved. Therefore, the abnormal phenomenon of A14 is considered to be caused by the release of compressive residual stress. Since the compressive residual stress of A14 is released by plastic deformation, the maximum stress reappears on the surface of the blade and the cracks also initiate from the surface of the blade. Meanwhile, due to the release of compressive residual stress, the LSP will have no effect or adverse effect on CCF life of blades.
Fig. 18

Crack initiation region of LSP blade (A3, A14) and untreated blade (B10, B16). The vibration stress of A3 and B10 is below the threshold vibration stress, and the vibration stress of A14 and B16 is above the threshold vibration stress. The red arrows indicate the location of crack initiation

4 Conclusions

  1. 1.

    LSP evidently extended the CCF life of both casting blades and forging blades. For the casting blades, the safe life was increased by 1.63 times and 1.56 times under the vibration amplitudes of 1.5 and 2.0 mm, respectively. For the forging blades, the safe life was also increased by 1.32 times and 1.47 times under the vibration amplitude of 1.7 and 2.1 mm. The effect of LSP on CCF life and scatter was more evident in casting blades.

     
  2. 2.

    Threshold vibration stress for both casting and forging blades was found. The CCF life tended to be extended by LSP only when the vibration stress was below the threshold vibration stress.

     
  3. 3.

    Due to the presence of compressive residual stress and refined grains induced by LSP, the crack initiation sources in LSP blades were obviously less, and the life of LSP blades was also longer.

     
  4. 4.

    Since the compressive residual stress was released by plastic deformation, LSP had no effect or adverse effect on CCF life of blades when vibration stress was above the threshold vibration stress.

     

Notes

Acknowledgements

This work was supported by National Natural Science Foundation of China (Grant Nos. 11602010 and 51505018).

References

  1. [1]
    X.C. Zhang, Y.K. Zhang, J.Z. Lu, F.Z. Xuan, Z.D. Wang, S.T. Tu, Mater. Sci. Eng. A 527 (2010) 3411–3415.CrossRefGoogle Scholar
  2. [2]
    G. Gomez-Rosas, G.C. Rubio, J.L. Ocaña, C. Molpeceres, J.A. Porro, W. Chi-Moreno, M. Morales, Appl. Surf. Sci. 252 (2005) 883–887.CrossRefGoogle Scholar
  3. [3]
    W. Jia, Q. Hong, H. Zhao, Mater. Sci. Eng. A 606 (2014) 354–359.CrossRefGoogle Scholar
  4. [4]
    J.M. Yang, Y.C. Her, N.L. Han, Mater. Sci. Eng. A 298 (2001) 296–299.CrossRefGoogle Scholar
  5. [5]
    C. Rubio-González, G. Gomez-Rosas, J.L. Ocaña, Appl. Surf. Sci. 252 (2006) 6201–6205.CrossRefGoogle Scholar
  6. [6]
    C. Rubio-González, C. Felix-Martinez, G. Gomez-Rosas, Mater. Sci. Eng. A 528 (2011) 914–919.CrossRefGoogle Scholar
  7. [7]
    J. Sheng, S. Huang, J.Z. Zhou, Opt. Laser Technol. 77 (2016) 169–176.CrossRefGoogle Scholar
  8. [8]
    C.Q. Hong, Appl. Surf. Sci. 351 (2015) 524–530.CrossRefGoogle Scholar
  9. [9]
    B. Lin, C. Lupton, S. Spanrad, Int. J. Fatigue 59 (2014) 23–33.CrossRefGoogle Scholar
  10. [10]
    A. Chahardehi, F.P. Brennan, A. Steuwer, Eng. Fract. Mech. 77 (2010) 2033–2039.CrossRefGoogle Scholar
  11. [11]
    X. Wu, C. Huang, X. Wang, Int. J. Impact Eng. 38 (2011) 322–329.CrossRefGoogle Scholar
  12. [12]
    I. Altenberger, R.K. Nalla, Y. Sano, Int. J. Fatigue 44 (2012) 292–302.CrossRefGoogle Scholar
  13. [13]
    N.F. Ren, H.M. Yang, S.Q. Yuan, Mater. Des. 53 (2014) 452–456.CrossRefGoogle Scholar
  14. [14]
    R.K. Nalla, I. Altenberger, U. Noster, Mater. Sci. Eng. A 355 (2003) 216–230.CrossRefGoogle Scholar
  15. [15]
    T.T. Xiong, F. Zhang, J. Iron Steel Res. Int. 22 (2015) 55–59.CrossRefGoogle Scholar
  16. [16]
    H. Wang, H.F. Zuo, H.E. Xun, Fail. Anal. Prev. 2 (2007) 24–28.Google Scholar
  17. [17]
    C. Schweizer, T. Seifert, B. Nieweg, Int. J. Fatigue 33 (2011) 194–202.CrossRefGoogle Scholar
  18. [18]
    X.J. Yan, R.J. Sun, D. Ying, J. Aerospace Power 26 (2011) 1824–1829.Google Scholar
  19. [19]
    Y. Tao, J. Jia, L.J. Tao, J. Iron Steel Res. Int. 17 (2010) No.9, 73–78.CrossRefGoogle Scholar
  20. [20]
    H.Y. Qin, G. Chen, Q. Zhu, J. Iron Steel Res. Int. 22 (2015) 551–556.CrossRefGoogle Scholar
  21. [21]
    R.J. Sun, X.J. Yan, J.X. Nie, J. Aerospace Power 27 (2012) 289–294.Google Scholar
  22. [22]
    N. Kang, X.J. Yan, L.I. Xiao, Gas Turb. Exp. Res. 23 (2010) 22–25.Google Scholar
  23. [23]
    X.J. Yan, X. Chen, R.J. Sun, J. Eng. Gas Turb. Power 136 (2014) 102502.CrossRefGoogle Scholar
  24. [24]
    T. Shimokawa, Y. Hamaguchi, J. Mater. Sci. Technol. 107 (1985) 214–220.Google Scholar
  25. [25]
    P.H. Kvam, E.P. Kvam, Technometrics 41 (1999) 277–290.CrossRefGoogle Scholar
  26. [26]
    J.Z. Yi, Y.X. Gao, P.D. Lee, Metall. Mater. Trans. A 34 (2003) 1879–1890.CrossRefGoogle Scholar
  27. [27]
    W. Liu, J. Dong, P. Zhang, Mater. Trans. 50 (2009) 791–798.CrossRefGoogle Scholar
  28. [28]
    X.P. Jiang, C.S. Man, M.J. Shepard, Mater. Sci. Eng. A 468-470 (2007) 137–143.CrossRefGoogle Scholar
  29. [29]
    M. Benedetti, V. Fontanari, P. Scardi, Int. J. Fatigue 31 (2009) 1225–1236.CrossRefGoogle Scholar

Copyright information

© China Iron and Steel Research Institute Group 2018

Authors and Affiliations

  1. 1.School of Energy and Power EngineeringBeihang UniversityBeijingChina
  2. 2.Collaborative Innovation Center of Advanced Aero-EngineBeijingChina
  3. 3.National Key Laboratory of Science and Technology on Aero-Engine Aero-ThermodynamicsBeijingChina
  4. 4.Beijing Key Laboratory of Aero-Engine Structure and StrengthBeijingChina

Personalised recommendations