Effects of intermittent loading time and stress ratio on dwell fatigue behavior of titanium alloy Ti-6Al-4V ELI used in deep-sea submersibles

doi:10.1016/j.jmst.2020.10.063

Abstract

Abstract:

Different components of deep-sea submersibles, such as the pressure hull, are usually subjected to intermittent loading, dwell loading, and unloading during service. Therefore, for the design and reliability assessment of structural parts under dwell fatigue loading, understanding the effects of intermittent loading time on dwell fatigue behavior of the alloys is essential. In this study, the effects of the intermittent loading time and stress ratio on dwell fatigue behavior of the titanium alloy Ti-6Al-4V ELI were investigated. Results suggest that the dwell fatigue failure modes of Ti-6Al-4V ELI can be classified into three types, i.e., fatigue failure mode, ductile failure mode, and mixed failure mode. The intermittent loading time does not affect the dwell fatigue behavior, whereas the stress ratio significantly affects the dwell fatigue life and dwell fatigue mechanism. The dwell fatigue life increases with an increase in the stress ratio for the same maximum stress, and specimens with a negative stress ratio tend to undergo ductile failure. The mechanism of dwell fatigue of titanium alloys is attribute to an increase in the plastic strain caused by the part of the dwell loading, thereby resulting in an increase in the actual stress of the specimens during the subsequent loading cycles and aiding the growth of the formed crack or damage, along with the local plastic strain or damage induced by the part of the fatigue load promoting the cumulative plastic strain during the dwell fatigue process. The interaction between dwell loading and fatigue loading accelerates specimen failure, in contrast to the case for individual creep or fatigue loading alone. The dwell fatigue life and cumulative maximum strain during the first loading cycle could be correlated by a linear relationship on the log-log scale. This relationship can be used to evaluate the dwell fatigue life of Ti alloys with the maximum stress dwell.

Key words: Ti-6Al-4V ELI, Dwell fatigue, Intermittent loading time, Stress ratio, Creep fatigue interaction

Chengqi Sun, Yanqing Li, Kuilong Xu, Baotong Xu. Effects of intermittent loading time and stress ratio on dwell fatigue behavior of titanium alloy Ti-6Al-4V ELI used in deep-sea submersibles[J]. J. Mater. Sci. Technol., 2021, 77: 223-236.

Figures/Tables 22

Fig. 1. Microstructures in the transverse section of specimens. (a) EBSD grain map and (b) EBSD phase map corresponding to (a); α- and β-phase are denoted in green and red, respectively.

Fig. 2. Shapes and dimensions of specimens (in mm) for (a) tensile test, (b) creep test, and (c) conventional fatigue test and dwell fatigue test at stress ratio R = 0.

Fig. 3. Schematics of loading waveforms. (a) Conventional fatigue test performed using continuous triangular wave (dt,min = 0) or intermittent triangular wave (dt,min > 0); (b) Dwell fatigue test performed using continuous trapezoidal wave ( dt,min = 0) or intermittent trapezoidal wave (dt,min> 0).

Fig. 4. Variations in dwell fatigue life and conventional fatigue life with intermittent loading time.

Table 1. Loading information and associated fatigue life and failure mode data for specimens in Fig. 4.

Maximum stress (MPa)	Dwell timed_t_,max(s)	Intermittent loading timed_t_,min(s)	Fatigue life (cyc)	Failure mode
815	120	0	2373	Fatigue
815	120	0	2384	Mixed
815	120	0	1028	Mixed
815	120	2	3669	Ductile
815	120	2	3391	Fatigue
815	120	2	2099	Mixed
815	120	12	5096	Ductile
815	120	12	2198	Mixed
815	120	12	2456	Mixed
815	120	60	1844	Mixed
815	120	60	1714	Mixed
815	0	0	18,317	Fatigue
815	0	0	19,585	Fatigue
815	0	12	15,920	Fatigue
815	0	12	17,098	Fatigue

Fig. 5. Photographs and fracture surface morphologies of failed specimens subjected to dwell fatigue tests at σmax=815 MPa, dt,max = 120 s, σmin=0 MPa, and dt,min = 0 s. A-1 to A-4: Nf = 2373 cycles; A-1 is photograph of fractured specimen, A-2 is SEM image of fracture surface, and A-3 and A-4 are magnified images of regions corresponding to up and down arrows in A-2, respectively. B-1 to B-4: Nf = 2384 cycles, B-1 is photograph of fractured specimen, B-2 is SEM image of fracture surface, and B-3 and B-4 are magnified images of regions corresponding to up and right arrows in B-2, respectively. C-1 to C-4: Nf = 1028 cycles; C-1 is photograph of fractured specimen, C-2 is SEM image of fracture surface, and C-3 and C-4 are magnified images of regions corresponding to lower left and down arrows in C-2, respectively.

Fig. 6. Photographs and fracture surface morphologies of failed specimens subjected to dwell fatigue tests at σmax=815 MPa, dt,max = 120 s, σmin=0 MPa, and dt,min = 2 s. A-1 to A-4: Nf = 3391 cycles, A-1 is photograph of fractured specimen, A-2 is SEM image of fracture surface, and A-3 and A-4 are magnified images of regions corresponding to arrows in A-2 and A-3, respectively. B-1 to B-4: Nf = 3669 cycles, B-1 is photograph of fractured specimen, B-2 is SEM image of fracture surface, and B-3 and B-4 are magnified images of regions corresponding to up and right arrows in B-2, respectively. C-1 to C-4: Nf = 2099 cycles; C-1 is photograph of fractured specimen, C-2 is SEM image of fracture surface, and C-3 and C-4 are magnified images of regions corresponding to left and down arrows in C-2, respectively.

Fig. 7. Photographs and fracture surface morphologies of failed specimens subjected to dwell fatigue test and tensile test. A-1 to A-4, B-1 to B-4, and C-1 to C-4: specimens subjected to dwell fatigue test, σmax=815 MPa, dt,max = 120 s, σmin=0 MPa, and dt,min = 12 s. A-1 to A-4: Nf = 2198 cycles; A-1 is photograph of fractured specimen, A-2 is SEM image of fracture surface, and A-3 and A-4 are magnified images of regions corresponding to arrows in A-2 and A-3, respectively; B-1 to B-4: Nf = 5096 cycles; B-1 is photograph of fractured specimen, B-2 is SEM picture of fracture surface, and B-3 and B-4 are magnified images of regions corresponding to up and down arrows in B-2, respectively. C-1 to C-4: Nf = 2456 cycles, C-1 is photograph of fractured specimen, C-2 is SEM image of fracture surface, and C-3 and C-4 are magnified images of regions corresponding to left and down arrows in C-2, respectively. d-1 to d-4: tensile specimen, d-1 is SEM image of fracture surface, d-2 is SEM image of fracture surface, and d-3 and d-4 are magnified images of regions corresponding to arrows in d-2 and d-3, respectively.

Fig. 8. Photographs and fracture surface morphologies of failed specimens subjected to dwell fatigue tests at σmax=815 MPa, dt,max = 120 s, σmin=0 MPa, and dt,min = 60 s. A-1 to A-4: Nf = 1844 cycles; A-1 is photograph of fracture specimen, A-2 is SEM image of fracture surface, and A-3 and A-4 are magnified images of regions corresponding to arrows in A-2 and A-3, respectively. B-1 to B-4: Nf = 1714 cycles; B-1 is photograph of fractured specimen, B-2 is SEM image of fracture surface, and B-3 and B-4 are magnified images of regions corresponding to left and down arrows in B-2, respectively.

Fig. 9. Fracture surface morphologies of failed specimens subjected to conventional fatigue test and intermittent conventional fatigue test. A-1 to A-3 and B-1 to B-3: specimens subjected to conventional fatigue test, σmax=815 MPa and σmin=0 MPa. A-1 to A-3: Nf = 18,317 cycles; A-1 is SEM image of fracture surface, and A-2 and A-3 are magnified images of regions corresponding to up and down arrows in A-1, respectively; B-1 to B-3: Nf = 19,585 cycles; B-1 is SEM image of fracture surface, and B-2 and B-3 are magnified images of regions corresponding to arrows in B-1 and B-2, respectively. C-1 to C-3 and d-1 to d-3: Specimens subjected to intermittent conventional fatigue test, σmax=815 MPa, σmin=0 MPa, and dt,min = 12 s; C-1 to C-3: Nf = 15,920 cycles; C-1 is SEM image of fracture surface, and C-2 and C-3 are magnified images of regions corresponding to right and lower-right arrows in C-1, respectively; d-1 to d-3: Nf = 17,098 cycles; d-1 is SEM image of fracture surface, and d-2 and d-3 are magnified images of regions corresponding to left and right arrows in d-1, respectively.

Fig. 10. Variations in cumulative maximum strain (εmax) and minimum strain (εmin) with number of loading cycles for specimens fractured within extensometer during dwell fatigue tests performed at σmax=815 MPa, σmin=0 MPa, and dt,max = 120 s.

Fig. 11. Dwell fatigue life at different stress ratios R. Lines denote linear regression results. (a) Fatigue life versus R and (b) fatigue life versus (1-R)/2.

Table 2. Loading information and associated fatigue life data for negative stress ratios shown in Fig. 11.

Maximum stress (MPa)	Dwell timed_t_,max(s)	Minimum stress (MPa)	Stress ratioR	Fatigue life (cyc)
815	120	-815	-1	207
815	120	-815	-1	269
815	120	-407.5	-0.5	761

Fig. 12. Photographs and fracture surface morphologies of failed specimens subjected to dwell fatigue tests at negative stress ratios. A-1 to A-4: σmax=815 MPa, dt,max = 120 s, σmin=-815 MPa, dt,min = 0 s, and Nf = 207 cycles; A-1 is photograph of fractured specimen, A-2 is SEM image of fracture surface, and A-3 and A-4 are magnified images of regions corresponding to down and up arrows in A-2, respectively. B-1 to B-4: σmax=815 MPa, dt,max = 120 s, σmin=-815 MPa, dt,min = 0 s, and Nf = 269 cycles; B-1 is photograph of fractured specimen, B-2 is SEM image of fracture surface, and B-3 and B-4 are magnified images of regions corresponding to down and up arrows in B-2, respectively; C-1 to C-4: σmax=815 MPa, dt,max = 120 s, σmin=-407.5 MPa, dt,min = 0 s, and Nf = 761 cycles, C-1 is photograph of fractured specimen, C-2 is SEM image of fracture surface, and C-3 and C-4 are magnified images of regions corresponding to down and up arrows in C-2, respectively.

Fig. 13. Variations in cumulative maximum strain (εmax) and minimum strain (εmin) with number of loading cycles for dwell-fatigued specimens fractured within extensometer at σmax=815 MPa, dt,max = 120 s, and dt,min = 0 and different stress ratios.

Fig. 14. Effects of prior creep and fatigue loading on tensile behavior. Prior-creep stress is 815 MPa and fatigue loading is σmax=815 MPa, dt,max = 0 s, σmin = 0, and dt,min = 0.

Table 3. Tensile strength and yield strength for specimens after being subjected to prior creep or fatigue loading.

Loading type	Tensile strength (MPa)	Yield strength (MPa)
Tensile	930	855
Tensile	929	846
Tensile	940	857
Creep for 100 h then tensile	950	945
Creep for 200 h then tensile	959	956
500 cycles then tensile	917	860

Fig. 15. Effect of prior creep on conventional fatigue life, diameter of test section of specimens, and cumulative strain as function of number of loading cycles. (a) Fatigue life versus prior creep time; prior-creep stress is 815 MPa as calculated based on initial diameter of test section of specimens, subsequent fatigue loading is σmax=815 MPa, dt,max = 0 s, σmin = 0, dt,min = 0, and maximum stress for fatigue loading was calculated based on initial diameter of test section of specimens. (b) Ratio of diameter of test section of specimens after prior creep to that before prior creep in (a). (c) Maximum stress versus fatigue life. Here, prior-creep stress is 815 MPa as calculated based on initial diameter of test section of specimens, and maximum stress during subsequent fatigue loading was calculated based on diameter of test section of specimens after being subjected to prior-creep. (d) Cumulative maximum strain (εmax) and minimum strain (εmin) versus number of loading cycles for prior-creep specimens fractured within extensometer. Herein, prior-creep stress is 815 MPa as calculated based on initial diameter of test section of specimens, and maximum stress during subsequent fatigue loading is 815 MPa as calculated based on initial diameter of test section of specimens.

Table 4. Cumulative strains after prior creep for 10 h and immediately after complete unloading after prior creep for 10 h for specimens subjected to prior creep for 10 h.

Specimen No.	After prior creep for 10 h	Immediately after complete unloading after prior creep for 10 h
1	3.45%	2.68 %
2	3.69%	2.96 %
3	3.83%	3.08 %

Fig. 16. Variations in cumulative maximum strain with loading time for specimens fractured within extensometer during dwell fatigue test with σmax=815 MPa and σmin=0 MPa. Results for creep tests are shown for comparison. The stress for creep tests is calculated based on initial diameter of test section of specimens.

Fig. 17. Effects of dwell loading and prior creep on cumulative maximum strain of specimens fractured within extensometer. (a) Cumulative maximum strain versus number of loading cycles, σmax=815 MPa and σmin=0 MPa. Specimens subjected to prior creep for 200 h at σ =815 MPa exhibited plastic strain because of creep and then fatigue. The stress for prior creep is calculated based on initial diameter of test section of the specimen. (b) Cumulative maximum strain versus loading time, σmax=815 MPa and σmin=0 MPa.

Fig. 18. Fatigue life versus cumulative maximum strain (εmax) during first loading cycle of dwell fatigue test. Line denotes linear regression result for dwell fatigue life (Nf) on logarithmic scale and strain (εmax) on logarithmic scale.

References 30

[1]	Z. Song, D.W. Hoeppner, Int. J. Fatigue 10 (1988) 211-218. DOI URL
[2]	M.R. Bache, Int. J. Fatigue 25 (2003) 1079-1087. DOI URL
[3]	P. Lefranc, V. Doquet, M. Gerland, C. Sarrazin-Baudoux, Acta Mater. 56 (2008) 4450-4457. DOI URL
[4]	J. Qiu, Y. Ma, J. Lei, Y. Liu, A. Huang, D. Rugg, R. Yang, Metall. Mater. Trans. A 45 (2014) 6075-6087. DOI URL
[5]	S. Waheed, Z. Zheng, D.S. Balint, F.P.E. Dunne, Acta Mater. 162 (2019) 136-148. DOI URL
[6]	Q. Song, Y. Li, L. Wang, R. Huang, C. Sun, Metals 9 (2019) 914. DOI URL
[7]	V. Sinha, M.J. Mills, J.C. Williams, Metall. Mater. Trans. A 35 (2004) 3141-3148. DOI URL
[8]	C. Lavogiez, S. Hémery, P. Villechaise, Int. J. Fatigue 131 (2020), 105341. DOI URL
[9]	W.J. Evans, C.R. Gostelow, Metall. Trans. A 10 (1979) 1837-1846. DOI URL
[10]	M.R. Bache, M. Cope, H.M. Davies, W.J. Evans, G. Harrison, Int. J. Fatigue 19 (1997) S83-S88.
[11]	F. Wang, W. Cui, Mater. Sci. Eng. A 642 (2015) 136-141. DOI URL
[12]	P.J. Bania, D. Eylon, Metall. Trans. A 9 (1978) 847-855. DOI URL
[13]	D. Munz, V. Bachmann, Materialwiss. Werkst. 11 (1980) 168-172. DOI URL
[14]	W. Shen, W.O. Soboyejo, A.B.O. Soboyejo, Mech. Mater. 36 (2004) 117-140. DOI URL
[15]	Y. Li, C. Bian, K. Wang, X. Sun, C. Qin, J. Ship Mech. 22 (2018) 1124-1135.
[16]	M.T. Whittaker, W.J. Evans, W. Harrison, in: Icf12 Ottawa (2013)
[17]	W.J. Evans, Fatigue Fract. Eng. Mater. Struct. 27 (2004) 543-557. DOI URL
[18]	W.J. Evans, M.R. Bache, Int. J. Fatigue 16 (1994) 443-452. DOI URL
[19]	L. Yang, J. Liu, J. Tan, Z. Chen, Q. Wang, R. Yang, J. Mater. Sci. Technol. 30 (2014) 706-709. DOI URL
[20]	M.E. Kassner, Y. Kosaka, J.S. Hall, Metall. Mater. Trans. A 30 (1999) 2383-2389. DOI URL
[21]	C.A. Stubbington, S. Pearson, Eng. Fract. Mech. 10 (1978) 723-756. DOI URL
[22]	W.J. Evans, J. Mater. Sci. Lett. 6 (1987) 571-574. DOI URL
[23]	W.J. Evans, M.R. Bache, Scr. Metall. Mater. 32 (1995) 1019-1024. DOI URL
[24]	J. Liu, L. Yang, Q. Wang, R. Yang, in: V. Venkatesh, A.L. Pilchak, J.E. Allison, S. Ankem, R. Boyer, J. Christodoulou, H.L. Fraser, M.A. Imam, Y. Kosaka, H.J. Rack, A. Chatterjee, A. Woodfield (Eds.), Proceedings of the 13thWorld Conference on Titanium, The Minerals, Metals & Materials Society, California, 2016, pp.955-958.
[25]	Z. Wu, H. Kou, N. Chen, M. Zhang, K. Hua, J. Fan, B. Tang, J. Li, J. Mater. Sci. Technol. 50 (2020) 204-214. DOI URL
[26]	Y. Li, Q. Song, S. Feng, C. Sun, Materials 11 (2018) 1628. DOI URL
[27]	Z. Zhang, L. Li, Z. Zhang, P. Zhang, J. Mater. Sci. Technol. 33 (2017) 603-606. DOI URL
[28]	Z. Zheng, D.S. Balint, F.P.E. Dunne, J. Mech. Phys. Solids 96 (2016) 411-427. DOI URL
[29]	R.J. Morrissey, D.L. McDowell, Int. J. Fatigue 21 (1999) 679-685. DOI URL
[30]	J.R. Taylor, An Introduction to Error Analysis, second ed., University Science Books, California, 1997.