FIGURE 6.29 Crack instability curve for 32% SCS-6/Ti-15-3. Value of o'ym = 556 MPa was used.

These complications urged engineers to suggest that accurate modeling using the models presented in Section 6.4.2, is possible only within specific limits that could secure the steady behavior of the toughening mechanisms [69, 96]. Such limitation reflects the incapacity of current fatigue modeling methodologies (LEFM, EPFM) to cope with: (a) fiber failure ahead of the crack tip (degradation of FB), (b) degradation of FCE, and (c) degradation of flow resistance.

In 1996, de los Rios et al. [96] published a work where the TZMM was quoted to operate extremely accurately until a specific crack length. The so-called onset of unsteady crack growth has been previously observed from fatigue experiments conducted in plane and notched SCS-6/Ti-15-3 uMMC specimens [51, 69, 92, 114]. Those experiments showed a distinct change in the slope of the fatigue crack growth rate (Fig. 6.30). The fact that this change takes place well before the final failure of the specimen urged the researchers to assume that such behavior is dominated by fiber failure in the crack wake and not ahead of the crack tip. Ibbotson et al. [51] concluded that this change may mark the initiation of an unstable crack growth. Similar, hypothesis has been upheld by de los Rios et al. [96].

In [11], it was suggested that negligible closure stress and substantial crack length could signify a hypothetical lower bound for fatigue failure (or a typical bound of the operational life of the material). In [10] it was reported that the onset of FB degradation can be predicted by assuming that the COD close to crack tip is equal to the CODcr (Fig. 6.31) and employing a modified version of

FIGURE 6.30 Crack length vs. loading cycles of 32% SCS-6/Ti-15-3 loaded at amax = 800 MPa (R = 0.01) with SEN of 650 ^m.
0 0

Post a comment