PSI - Issue 19

Keiji Yanase et al. / Procedia Structural Integrity 19 (2019) 504–512 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

508

5

2.2. Experimental method

In practice, it is difficult to perform a stable crack growth test for a shear-mode crack in hard steels in the high-cycle fatigue regime because of mode I crack branching (Fig. 7, cf. Schönbauer et al. (2017), Nishimura et al. (2018) and Endo and Yanase (2019)). Further, to achieve the present research objective, a relatively long shear-mode fatigue crack (cf. Figs. 2 and 3) must be reproduced in a laboratory. However, it is not a straightforward task to study the shear-mode fatigue crack growth behavior in a laboratory. In practice, the scarcity of experimental data associated with the lack of an efficient experimental setup to study the shear-mode crack behavior hinders the development of reliable quantitative and qualitative analyses (Endo et al. (2016)). Recently, a new testing method was proposed and implemented to obtain stable shear-mode fatigue crack growth on the surface of a specimen by applying cyclic torsion with superimposed static compression, as illustrated in Fig. 8. This method makes it possible to directly observe the behavior of a small shear mode crack on the surface of a specimen (Matsunaga et al. (2011), Okazaki et al. (2014, 2017), Endo et al. (2016), Akaki et al. (2017)). Since the possibility to observe a shear-mode crack on the surface of a roll steel specimen was unknown, the effectiveness of testing technique was firstly examined with Material A in this research. Torsional fatigue tests were carried out in air at room temperature by a servo-hydraulic combined axial/torsional loading fatigue testing machine (MTS model 809). The static compressive stress of  static = – 1000 MPa was simultane ously applied in the direction of the specimen axis to obtain a stable shear-mode fatigue crack growth. The test frequency was f  20Hz, and the stress ratio was R = – 1. Surface cracks were successively observed using the replica method at the appointed number of cycles, N . If the amplitude of twist angle,  a , was increased up to   a = 1 o from the onset of fatigue test, the specimen was defined to be failed. When the fatigue test continued until N = 10 7 cycles without failure, the specimen was defined to be run-out. Given the actual inspection and maintenance period of N  10 6 cycles for the rolls, N = 10 7 cycles is enough for the fatigue testing of the roll materials.

Specimen axis

Specimen axis

200 m m

200 m m

Fig. 7. Torsional fatigue failure with mode I crack branching in 17-4PH stainless steel (Schönbauer et al. (2017)).

Static compressive stress,  static

Static compression

Shear-mode crack initiation & propagation

Cyclic torsion

Cyclic shear stress,  a

Fig. 8. Torsional fatigue test with superimposed static compression.