Safe choice of structural steels in a region of ultra-high number of load cycles

Abstract In this paper the authors introduce their own selected experimental results in the field of the investigation of fatigue resistance of structural steels. The experiments were carried out on the nine structural steels including high strength steels, DOMEX 700MC, HARDOX 400, HARDOX 450, 100Cr6 (UTS from 446 MPa to 2462 MPa) at high-frequency cyclic loading (f = 20 kHz, T = 20 ± 5 °C, R = -1) in the region of number cycles ranged from N ≈ 2×106 to N ≈ 2×109 cycles of loading. The continuous decrease of fatigue strength in dependence on the number of loading cycles was observed with the average value of ratio σa2×109/σa2×106 = 0.69.


Introduction
Fatigue resistance of structural materials began to be experimentally studied at the beginning of the 19 th century. The strain or stress vs. number of cycles (S-N, σa = f(N)) dependence including conventional fatigue limit usually referred to as N = 2×10 6 ÷ 10 7 cycles (steels and cast irons) are the main parameters used at evaluation of fatigue resistance of structural materials. However, targeted research the fatigue fractures occur not only at low, high but also at ultra-high cycle region of loading.
Modern industry is focused on the increase of the performance of industrial equipment. This fact is connected with prolongation of fatigue life of individual parts of the equipment where, on the other hand, safety is the most important factor. With the aim to prolong the service life of parts was the fatigue resistance of structural steels in the region from N ≈ 1×10 6 cycles to N ≈ 1×10 10 cycles intensively studied during last years. In this region of ultra-high cycle loading are open questions about the influences of cyclic plasticity processes at very low loading amplitudes, surface and sub-surface fatigue crack initiation, micropurity, microstructural heterogeneity, inclusions, holes, shrinkages, pores, grain size, long grain boundaries and their effect on fatigue behaviour. Moreover, the fatigue limit existence and physical nature, shape of S-N, σa = f(N) curve, type of loading and frequency influence are discussed intensively. Research in this field has shown that in structural materials (steels and cast irons) fatigue strength decrease continuously with increasing number of loading cycles even behind, in past sufficient, N = 2×10 6 -2×10 7 cycles. For this reason, is necessary to know values of fatigue strength in region of ultra-high number of cycles loading, if designing structural parts are intended to be cyclic loaded in this region (Bathias et al., 2004;Bathias et al., 1999, Bokůvka et al., 2002Bokůvka et al., 2014;Nový et al., 2012;Chapetti, 2010).
The study of fatigue resistance of structural materials with the use of low-frequency loading (f = 10 -200 Hz) in the region of ultra-high number of loading cycles is time consuming. Using of high-frequency, ultrasonic fatigue testing devices is therefore necessity. Time and economical effectiveness of the determination of the fatigue characteristics of structural materials is evident (Bokůvka et al., 2002;Bokůvka et al., 2014;Kazymyrovych et al., 2009;Stanzl-Tschegg, 1999;Trško et al., 2018, Höppel et al., 2010.
In this paper the authors introduce their own experimental results from the field of fatigue resistance of structural steels obtained in the region of ultra-high number of loading cycles.

Experimental
Experimental works, the quantitative chemical analysis, tensile tests, fatigue test were carried out on the nine structural steels incl. high strength steels (DOMEX 700MC, HARDOX 400, HARDOX 450, 100Cr6). Chemical analysis was performed with the help of emission spectrometry on an ICP (JY 385) emission spectrometer using a fast recording system Image. Tensile tests were carried out on a ZWICK Z050 testing machine at ambient temperature of T = 20 ± 5 °C, with the loading range in interval F = 0 -20 kN and the strain velocity of έm = 10 -3 s -1 . Round cross-section specimens were used; the shape and dimensions of the test specimens met the requirements of EN 10002-1 standard (3 specimens were used). Fatigue tests were carried out at high-frequency sinusoidal cyclic tension-compression loading (f = 20 kHz, T = = 20 ± 5 °C, R = -1, cooled by distilled water with anticorrosive inhibitor) and with the use of ultrasonic fatigue testing device, see Fig.  1. Smooth round bar specimens (min. 10 pieces) shown in Fig.  2, with diameter of 4 mm, ground and polished by metallographic procedures were used for the fatigue tests (Trško et al., 2018). The investigated region of number of cycles ranged from N ≈ 2×10 6 to N ≈ 2×10 9 cycles of loading.
where ∆σa is given from ∆σa = 20 MPa to ∆σa = 200 MPa. From the Tab. 3 is visible the influence of tested structural steel ultimate tensile strength UTS on the fatigue properties. The structural steel with high ultimate tensile strength UTS level exhibit the high values of σa2×10 6 and σa2×10 9 ; more precisely thresholds for fatigue crack initiation. In contrary the thresholds for cracks growth (defect tolerant approach) are decreased. The fatigue ratio is decreasing from 0.63 to 0.35 (σa2×10 6 ), from 0.45 to 0.25 (σa2×10 9 ), decreasing σa2×10 6 vs. σa2×10 9 with number of cycles increasing (N = 2×10 6 vs. N = 2×10 9 cycles of loading). The average value of ratio σa2×10 9 /σa2×10 6 is about 0.69. Microstructural factors which improve resistance to fatigue crack growth do not necessarily guarantee improved resistance to crack initiation and philosophy of short cracks must be considered (Ritchie, 1981).

Conclusion
Based on the results of the fatigue tests it can be stated: • in tested nine structural steels a continuous decrease of fatigue strength in dependence on the number of loading cycles was observable in the ultra-high region of loading cycles (2×10 6 < N < 2×10 9 ); • the average value of ratio σa2×10 9 /σa2×10 6 was 0.69; • the values of conventional fatigue limit (usually referred to N = 2×10 6 -10 7 cycles) are overestimated and, therefore, do not meet demands of safety of structural parts; • these facts must be taken into consideration with reference to safety when designing structural parts in the ultra-high region of loading cycles.