Issue 78

A. P. Soldatenkov et alii, Fracture and Structural Integrity, 78 (2026) 248-259; DOI: 10.3221/IGF-ESIS.78.15

Fig. 12 shows the dependences of the fatigue striation spacing and the crack propagation duration vs the crack lengths in the thickness direction of the specimens cut out from the sample with the heat treatment 2. The maximum values of fatigue striation spacing were ~350 nm, with typical values of several microns for titanium alloys [28]. Therefore, the crack propagation duration was shortened greatly. This phenomenon could be caused by the exhaustion of ductility of the Ti 5Al-5V-5Mo-1Cr-1Fe alloy during the applied treatment and its inability to achieve the typical values greater than 1 µm, on the one hand. On the other hand, the sizes of the tested specimens did not allow achieving high SIF values when propagating the semi-elliptical surface cracks. It is the high SIF value that determines the high value of fatigue striation spacing.

Figure 12: The dependences of the fatigue striation spacing δ and the crack propagation duration N p on the crack lengths a in the thickness direction of the specimens with the heat treatment 2, tested at 1200 (a) and 900 (b) MPa. Nevertheless, it should be highlighted that the fatigue durability of the specimens cut out from the sample with the heat treatment 2 were much greater than typical values for the Ti-5Al-5V-5Mo-1Cr-1Fe alloy at similar cyclic stresses. So, the reduced fatigue striation spacing should be related to the specimen sizes, but not to the treatment conditions. According to [29], the fatigue limits at the number of cycles of 10 7 for six different semi-finished products of the Ti-5Al 5V-5Mo-1Cr-1Fe alloy spanned from 490 MPa to 590 MPa. The maximum value of fatigue limit was obtained for extrusions in annealed state, the fatigue curve of which was compared with the data of fatigue tests for specimens with the heat treatment 2. From the fatigue curve for an extruded and annealed section, the fatigue durability was observed to be 5·10 4 – 10 6 cycles at the maximum stress of ~690 MPa. Thus, the fatigue strength of 800 MPa at 2·10 6 cycles obtained in this study for the specimens with the heat treatment 2 can be compared with fatigue strengths at 10 6 cycles for the alloy with six different semi-finished products. The fatigue strength at 10 6 cycles for the alloy with the heat treatment 2 was increased by more than 15 % as compared to the above-mentioned reference data [29]. However, the reference data correspond to specimens subjected to symmetrical cyclic loading and, most likely, by rotary bending, while in this study, the fatigue tests were conducted under positive R -ratio. The variations in both cycle stress ratios and types of loading could give rise to change in the value of fatigue strength over the discussed number of cycles, so the obtained results should be considered as preliminary. For a more accurate comparison of these data, similar specimens should be tested under identical conditions (loading type and symmetrical stress cycle). By analyzing the dependences of the fatigue striation spacing on the crack length, it was possible to estimate the relative duration of crack propagation period N p / N f in the fatigue durability N f of the specimens. It reduced with a decrease in the maximum cycle stress (or with an increase in the fatigue durability N f achieved in the test) and could be approximated by a monotonic dependence (Fig. 13a). By comparing with the results reported for the heat treatment 1 [15], no significant difference in the relative duration of crack propagation period N p / N f was found at similar fatigue durability N f (Fig. 13a). In addition, a greater scatter of this characteristic depending on the fatigue durability N f was observed for the specimens cut out from the sample with the heat treatment 1. Since the fatigue durability N f was greater for the heat treatment 2 at the same stress level, its crack initiation period was longer as well. For alloy samples with the heat treatment 2, the transition to the final fracture stage was observed at the values of K I,max from 21.8 up to 31.1 MPa·m 1/2 , while they were from 15.6 up to 25.4 MPa·m 1/2 for the heat treatment 1 (Fig. 13b). Its high levels of about 28 MPa·m 1/2 were also reported in [30] for the coarse-grained Ti-5Al-5V-5Mo-1Cr-1Fe alloy at the numbers of cycles to failure within 1.5  10 4 . Obviously, a correlation between the SIF values K I,max and the fatigue durability N f was weak in the studied cases. Dependences of the fatigue striation spacing δ at transition to the final fracture stage on the corresponding SIF values K I,max enabled to reveal a weakly expressed monotonous increasing relationship between them (Fig. 13c). For the heat treatment 2, the transition to the final fracture stage occurred at higher values of fatigue striation spacing, testifying for higher ductility

256