Fatigue Crack Paths 2003

encountered with the sequence composed of 5 104 cycles blocks. These evolutions of

the crack path under sequential loads tend to prove that the expected crack orientations

are observed and hence the stage I (mode II) and stage II (mode I) are relevant for a

damage description. However, it is also clear that tension and torsion are not only

different from the crack orientation point of view, the corresponding damage evolutions

being not the same for these two types of loading (Fig. 1).

Specimen axis

9

8

7

1 →6

Figure 2. Surface crack path for a 105 cycles block (tension, torsion) test (Nf=8 105)

applied to a mild steel C36 and corresponding block numbers.

Fatigue tests under blocks applied to C18 [9]

Lasserre and Froustey [9] tested a mild steel C18 whose fatigue limit in torsion is

t-1=179 M P aand in bending f-1=310 MPa. In the finite life regime, the bending level

Σa=350 M P aleads to the mean lifetime (7 specimens) Nf=1.65 105 whereas in torsion

the amplitude Τa=200 M P aleads to Nf=1.73 105(5 specimens).

sequence

The authors showed that a

applied until initiation and composed by 104

cycles of bending followed by 10 4 c y c l e s of torsion is as damaging (that is leads to the

same number of cycles to initiation, i.e. 1.7 105) as a test carried out with either bending

or torsion loads. This observation was made by applying (on 5 specimens) in the block

tests the amplitudes of bending (350 MPa) and torsion (200 MPa) that lead to 1.7 105

during the simple tests.

They also study the influence of a block sequence composed by 104 cycles of

bending Σa followed by 104 cycles of combined in phase bending-torsion (Τa+Σa)

followed by 104 cycles of torsion Τa where Σa=350 MPa, (Τa+Σa) =(350 MPa,144 MPa)

and Τa=200 MPa. This sequence was repeated until initiation of a crack. The mean

lifetime (3 specimens) was found to be Nf=1.73 105.