PSI - Issue 18

V.N. Shlyannikov et al. / Procedia Structural Integrity 18 (2019) 322–329 Author name / Structural Integrity Procedia 00 (2019) 000–000

328

7

A qualitative comparison between the experimental and full scale crack path shows a satisfactory agreement for biaxial loading conditions at imitation model II (Fig. 1c and Fig. 7a). Thus, the biaxially loaded imitation model II most accurately reproduces the state of the critical zone of the compressor disk at operation.

a)

b)

Fig. 6. Crack path (a) and fracture surface (b) for imitation model I.

Cut surface to open crack Crack length, a 1

Crack length, a 2 Ruptured zone

a)

b)

Fig. 7. Crack path (a) and fracture surface (b) for imitation model II.

The crack front positions of the imitation model II fracture surface are shown in Fig. 7b. In this figure, the depth of quarter-elliptical crack is denoted by c , crack length a 1 is the current crack length monitored by using the optical microscope, and crack length a 2 is the crack length on the second free surface of the imitation model II, which is measured by benchmarks. The crack propagation process in imitation model II can be divided into two stages. In the first stage the quarter-elliptical crack is described by a part-through thickness crack. In the second stage quarter elliptical crack completely crosses the imitation model II thickness t and becomes a through-thickness crack. In particular, using the optical microscope, it was possible to obtain the curve of crack length propagation versus cycle numbers for both imitation models (Fig. 8a) and the curves of through-thickness crack propagation (for both free surfaces) versus cycle numbers for imitation model II (Fig. 8b). The crack front shape and the parameters a and c for quarter-elliptical part-through thickness cracks were obtained by detailed fracture surface analysis. The relation between the aspect ratio a/c and cycle numbers is