Contribution on “slow fatigue crack growth and threshold behaviour of a medium carbon steel in air and vacuum” by R. J. Cooke, P. E. Irving, G. S. Booth and C. J. Beevers

✍ Scribed by R.O. Ritchie

• Publisher: Elsevier Science
• Year: 1975
• Tongue: English
• Weight: 217 KB
• Volume: 7
• Category: Article
• ISSN: 0013-7944
• DOI: 10.1016/0013-7944(75)90075-2

✦ Synopsis

I SHOULD like to compliment Dr Beevers and his co-workers on their paper. It is pleasing to see at last some mechanistic basis to the problem of fatigue crack propagation in steels at stresses approaching the threshold stress intensity for crack growth. Such data are still extremely scarce in the literature. Further, they are to be complimented on their attempt at understanding the marked influence of mean stress (or K,, the maximum stress intensity) on the growth rate at these low stress intensities. This effect is known to disappear at higher values of AK (the alternating stress intensity), until at very high propagation rates, as K, approaches Kc (the stress intensity at failure), growth becomes once more mean stress dependent [l-3]. In the latter case it has been relatively straightforward to attribute this mean stress sensitivity to fracture mechanisms such as cleavage, intergranular cracking and void coalescence ("static" modes), which occur during striation formation [ l-31. Although a nominally static fracture mode, namely intergranular separation, is also observed at very low growth rates in steels (see, for exampk, Refs. l-5), I feel that the authors are correct in asserting that this, in itself, is not the root cause of the marked effect of mean stress near the threshold. Their interpretation of the mean stress dependence at low growth rates being due primarily to the effect of the environment is far more appealing This has been borne out particularly well by the results of CookelS], who showed that mean stress sensitive growth at low values of AK was absent in several steels tested under vacuum, whereas the same steels tested in laboratory air showed a marked mean stress dependence. Similar observations have been made in Ti4AMV alloysl6.7).

However, to explain further this dependence of growth rate on mean stress due to environmental action, Beevers and his co-workers introduce a two-component mechanism, and it is here that I find myself at variance with some of their conclusions. Briefly, this two-fold mechanism for steels involves (1) creation of damage ahead of the crack tip in the form of intergranular facets, which is considered to be AK controlled, and (2) the linkage of these facets to the main crack front by a tearing mechanism, regarded as K, controlled. As the authors attribute the marked mean stress (or K,) influence on growth rate to the environment, one must conclude that the environment is affecting the tearing component (K, controlled) rather than the creation of intergranular damage. This I find surprising. Without claiming that the environment will not affect a tensile tearing process (see, for example, the model due to Krafft and Mulherin for "wet" enviromnents[8]), it seems far more likely that the environment would primarily aid the intergranular separation. This could occur by the diffusion of hydrogen from the water vapour in air to grain boundaries ahead of the crack tip, thereby weakening them relative to the grain interiors. Support for this can be found from the authors' own results, which show clearly little evidence of intergranular fracture (and no effect of mean stress) for fatigue crack tPubkthcd in ikgng FmctwchiecA,7,6947(19l5).