PSI - Issue 43

Jan Klusák et al. / Procedia Structural Integrity 43 (2023) 142–147 Author name / Structural Integrity Procedia 00 (2022) 000 – 000

143

2

2021, Trávníček et al. 2021, Seitl et al. 2022) and a study of construction steels S355 (Klusák et al . 2021, Seitl et al. 2018, Seitl et al. 2018a) Ultrasonic testing device was used to generate high frequency loading at 20 kHz. Thus, we could reach high number of cycles in real time. Fully reversed push-pull loading cycle with the stress ratio R = -1 was applied.

Nomenclature A

Coefficient of Basquin’s law Coefficient of Basquin’s law

B

Dynamic modulus of elasticity [GPa] Intrinsic frequency of the specimen [Hz]

E d

f n R S f

Stress ratio

Stress factor [MPa/  m] Poisson’s ratio [-]

 

Density [kg/m 3 ] Fatigue limit

σ f

2. Experimental testing Fatigue lifetime of the studied steels is described by S-N curves, that were obtained from tests of cyclic loading at ultrasonic frequency. The specimens were designed to exhibit their intrinsic frequency close to the frequency of the loading device. Firstly, basic material properties had to be determined. Dynamic modulus of elasticity was ascertained by means of an impulse excitation technique , and together with the Poisson’s ratio and density, the properties are in Table 1. Dynamic modulus, Poisson’s rati o and density of materials enter to the modal analysis performed in the finite element software ANSYS. Thus, the dimensions and the intrinsic frequency of the specimens were adjusted and calculated, see Table 1 and Fig. 1. Finally, the stress factor S f of the specimens were calculated by the harmonic analysis, and they are also stated in Table 1. To observe the materials microstructure, the fatigue samples were cut longitudinally. Specimens for the microstructure characterization were ground using sandpaper followed by electrolytic polishing in solution of 600 ml methanol, 360 ml ethylene glycol monobuthyl ether and 60 ml perchloric acid. Using Tescan Lyra 3 XMU scanning electron microscope, the gauge lengths and the specimens' heads were observed. The deformation structures were observed using the electron channeling contrast imaging technique (ECCI), while the microstructures in as-delivered conditions were characterized using electron back-scattered diffraction (EBSD). Microstructures of both experimental materials are shown in Fig. 2. Both materials exhibited similar microstructure with equiaxed austenitic grains, containing many annealing twins and delta ferrite stringers along the rolling direction. Area- weighted mean grain size was 59.78 μm for 1.4306 steel and 57.92 μm for 1.4307 steel. No preferential crystallographic orientation was recorded.

Table 1. Properties of the materials and the stress factors of the test specimens.

Material

Dynamic modulus E d [GPa]

Poisson’s ratio  [-]

Density  [kg/m 3 ]

Intrinsic frequency f n [Hz]

Stress factor S f [MPa/  m]

304L/4306 304L/4307

199 198

0.3 0.3

7899 7884

20005 19974

26.3 26.2

After machining, the test samples were polished electrolytically to get fine surface without defects and traces of machining. High frequency measurements are performed at resonant frequency of 20 kHz. Cyclic loading of austenitic steels leads to high heat generation. For this reason, the specimens must be efficiently cooled. In our case, cooling by water in a closed circuit was used, see Fig. 3.