PSI - Issue 42

Kamila Kozáková et al. / Procedia Structural Integrity 42 (2022) 270–275 K. Koza´kova´ et al. / Structural Integrity Procedia 00 (2019) 000–000

272

3

3. Experimental procedure

Fatigue tests were performed using an ultrasonic fatigue testing machine which works at the frequency of 20 kHz. These experiments require a special design of specimens. All specimens must be designed so that their intrinsic frequencies of longitudinal oscillations are equal or close to 20 kHz. The design of smooth and notched specimens is shown in Fig. 5. The lengths of notched specimens are listed in Tab. 2. During fatigue tests, closed circuit of deionized water is used for specimen cooling to prevent overheating and subsequent changes in the microstructure. The direction of loading is equal to the direction of extrusion. The middle part of machined smooth specimens was ground and polished (SiC papers, diamond paste). It was not possible to perform mechanical polishing of notched specimens due to the geometry of notches, so the notched specimens remained in the post-machined state.

Fig. 5: Parameters of smooth and notched specimens

Table 2: The lengths of notched specimens

notch radius r [mm]

0.1

0.2

0.4

0.8

1.5

3

length of specimen L [mm]

78.1

78

77.9

77.6

77.3

76.7

4. Results

All the tests were performed at the ultrasonic fatigue testing device. Number of cycles to fracture is between 10 6 and 10 10 . Fig. 6 shows experimental results. All specimens were broken. These results were approximated by S - N curves in the form of an exponential relationship between applied stress σ a and number of cycles to failure N f : σ a = A · N B f . The coe ffi cients of S - N curves are listed in Tab. 3.

Table 3: Coe ffi cients of S - N curves

r [mm]

0.1

0.2

0.4

0.8

1.5

3

smooth 318.39 -0.0304

A B

310.39 -0.0795

297.77 -0.0727

269.77 -0.0712

419.85 -0.0833

602.21 -0.0969

297.03 -0.0472

5. Observation and discussion

It was expected to obtain the set of S - N curves with a clear influence of the notch radius. The smaller notch radius should lead to shorter fatigue life for the same stress level. Also the slope of the curves should be similar. However, there have been variances from these expectations. The life curve of the specimens with a notch radius of 1.5 mm descends more steeply than the others. Further, it should be noted that the life curve of specimens with a notch radius