**4. Results**

Fatigue tests were performed with specimens from AISI 347 batch A taken from an original surge line pipe. Specimens from rolled bars from AISI 347 batch B represented the more metastable material and were used for selected fatigue tests to show the influence of metastability on the deformation and phase transformation behavior in the LCF regime as well as for characterizing HCF behavior at ambient temperature.

## *4.1. LCF Behavior*

#### 4.1.1. Ambient Temperature

Single step (constant amplitude) tests under total-strain control were performed in the LCF regime with total strain amplitudes 0.6% ≤ εa,t ≤ 1.6%, load frequency f = 0.01 Hz and triangular waveform. Figure 6 presents the development of stress amplitude (σa), the change in specimen temperature (ΔT) and α'-martensite formation (ξ) for batch A and, additionally, the results from the test with total strain amplitude εa,t = 1.0% for batch B. The cyclic deformation behavior of the investigated steel at ambient temperature was fundamentally determined by deformation-induced austenite-α´-martensite transformation. After a load-dependent number of cycles N, the formation of α´-martensite started and increased continuously with increasing cycle number until specimen failure (Figure 6c). The σa,N-curves (Figure 6a) illustrate the associated cyclic hardening processes, which led to a maximum stress amplitude for εa,t ≥ 1.2% in the range of the tensile strength σ<sup>f</sup> = 569 MPa (batch A) of the solution-annealed material. Cyclic hardening was detected by temperature measurement, temperature increases in total-strain controlled fatigue tests being indicative of increases of plastic strain energy. Because of very low load frequency, the maximal change in the specimen temperature was below 3 K, which corresponds with an absolute temperature of about 28 ◦C and consequently bears no significant influence on the formation of α´-martensite. Figure 6c shows the development of α´-martensite content during the fatigue tests discussed above. After an incubation period, which depended on the total strain amplitude, the α´-martensite volume fraction increased continuously with the number of cycles. With increasing εa,t, the onset of α´-martensite formation was shifted to lower N, and a higher volume fraction of α´-martensite was measured at specimen failure. To characterize the influence of metastability of nominally the same type of austenitic stainless AISI 347 steel, specimens from batches B were cyclically loaded with total strain amplitude εa,t = 1%. Both batches A and B showed a continuous cyclic hardening, however, with different gradients after the first ten cycles. In the cycle range 10 < N < 100, batch A showed only a slight increase of stress amplitude, while batch B underwent a larger increase of σa, which correlated directly with a significant development of α'-martensite (Figure 6c) from 1 FE% to 41 FE%. As mentioned in the methods section above, ferromagnetic α'-martensite was measured in situ during fatigue testing using a FeritscopeTM sensor, for which readings above 60 FE% could be inaccurate due to lack in instrument linearity and calibration difficulties at high α'-martensite contents. For this reason, measured data in the range above 60 FE% were plotted as dashed lines in Figure 6c. After a high rate of σ<sup>a</sup> increase between N = 10 and N = 100, batch B showed a decrease in the stress amplitude rate dσa/dN. However, further cyclic hardening took place which led to specimen failure at σ<sup>a</sup> = 760 MPa, which is 120 MPa higher than the tensile strength of this material in its as-received state (Table 2). Batch A also showed evidence of a correlation between the development of α'-martensite and an increase in stress amplitude. However, the start of α'-martensite formation occurred at a higher number of cycles in comparison to batch B and the α'-martensite content was found to be, at the same number of cycles, significantly lower with the maximum value at specimen failure ξ = 30 FE%. The results of these tests clearly depicted how substantial the differences can be between the formation of deformation-induced α'-martensite in the same type of austenitic stainless steel due to differences in chemical composition and grain size and consequently the material's metastability. It should be noted that both batches had a chemical composition within the range given in international standards for the investigated steel type.

**Figure 6.** Development of (**a**) stress amplitude, (**b**) change in temperature and (**c**) α'-martensite versus N during total-strain controlled LCF test of batch A and B.
