PSI - Issue 42

Matěj Mžourek et al. / Procedia Structural Integrity 42 (2022) 457 – 464 Matěj Mžourek / Structural Integrity Procedia 00 (2019) 000 – 000

460

4

(for notched specimens) Hz was chosen, as there is little impact on the relative value of obtained fatigue life. The resonating frequency of the machine with the A01-A02-A03-A04-A51-A52 specimens was 125-100-160-130-184 184 Hz, respectively. Thermographic measurements of the surface via a FLIR A315 camera were taken during testing to be utilized in a parallel research project on fatigue thermodynamics. No external cooling of the specimens was thus induced. 3. Experimental Results 3.1. S-N curves Raw obtained fatigue data points are available in Fig. 3. a), where the vertical axis corresponds to axial nominal stress σ y,nom (calculated from the net cross section for notched specimens). A Basquin-type curve was used to approximate the near-linear region of each curve. Afterwards, modifications based on roughness and hardness were applied, as explained in the following chapter. Finally the curves were transformed into the σ y,max – N space via multiplying the nominal stresses by K t obtained through FEM – this was done for unnotched specimens as well. This ensures the usability of the model given by Eq. (1). The modified curves used for the analyses are plotted in Fig. 3. b), with their parameters available in Table 2.

Table 2. Basquin-type parameters of the modified S-N curves Specimen type A01 A02 A03 A04

A51 5.23

A52 5.29

W [-] C [-]

12.31

19.62

16.98

20.88

1.3E+39 1.5E+59 9.2E+51 1.1E+62 1.9E+20 7.1E+19

150 250 350 450 550 650 750 850 1E+2 1E+3 1E+4 1E+5 1E+6 1E+7 1E+8 σ y,nom [MPa] Number of cycles [-] A01 A02 A03 A04 A51 A52

1050 1250 1450 1650

A01 A03 A51

A02 A04 A52

σ y,max [MPa]

450 650 850

1000

10000 100000 1000000

Number of cycles [-]

Fig. 3. a) Obtained fatigue data points, b) S-N curves with modifications applied according to Secs. 3.3 and 3.4.

3.2. Thermal response Pronounced self-heating behavior has been observed during testing via the FLIR A315 camera. Temperatures notably higher than in the rest of the tested specimens were observable in the elongated smooth specimens A02 and A04. Maximal temperature reached during a test generally rises with the set loading amplitude, with the extremes reaching up to 200 °C. A change in the stiffness of the specimens due to elevated temperatures is believed to be the cause of some of the tests ending prematurely (due to a drop of the resonant frequency), where no crack was detected via dye-penetrant test after the fatigue experiment (4 specimens in A02, 8 specimens in A04). A CT scan was performed for some specimens where the dye penetrant test failed to show any crack - no internal cracks were revealed. The elevated temperatures should not be able to induce phase transformation or residual stress relief.