Page 391 - Biaxial Multiaxial Fatigue and Fracture
P. 391
The Environment EfJect on Fatigue Crack Growth Rates in 7049 Aluminium Alloy at ... 315
Sadananda [40,48]. The UA alloy shows an L-shape at positive R-ratios too, whereas a loss of
fatigue crack propagation resistance may be recognised at negative R-ratios again. AK reaches
a plateau for R<O and Kmx therefore decreases with decreasing R to comply with constant AK
required for crack growth. This means that the controlling mechanism switches from a K,,
controlled behaviour for positive R-ratios to a AK controlled for negative R. This implies that
reversed cyclic plasticity may become the governing factor for fatigue crack growth.
Observing the experimental results, the role of environment and microstructure on the
threshold Mth-KmX curve for the same 7049 alloy and summarising the contributions of
several investigators [3,19,20,22,25,46,47], it is clear that the introduction of moist air
environment reduces strongly the hK*th values of the UA and OA alloys. The wavy slip mode
in the OA alloy probably is the reason for the reduced Math in comparison to the UA alloy in
moist air and in vacuum. The two microstructures likewise show different fatigue crack
growth behaviour in moist air, with higher thresholds of the UA structure than those of the
OA alloy.
Figures 8 (a) shows the better fatigue crack growth properties of both alloys in vacuum
than in humid air which has been found in a similar extend by Kirby and Beevers [25,36]. The
data of Kirby and Beevers, resulting in a AKth-Kmax curve with a slope of 1, point to the
prevailing influence of fatigue loading and microstructure. The different results of the present
study probably are caused by two facts. First, the studied alloy (AI 7049 alloy instead of
7075) is more susceptible to corrosive influences, and second, the vacuum was not a high
vacuum (-2.6~10” Pa).
Figures 9 (a) to (0 show the fracture surfaces typical for fatigue loading of the UA-OA
alloy in air and vacuum at R=-1. The UA alloy shows planar slip in air in contrast to the rather
ductile fracture surface in vacuum (Figs 9 (e) and (d)). In ambient air both the UA and OA
alloy (Figs 9 (a) and (c)) shows a brittle crystallographic fracture mode. As an additional
example for the influence of the environment, Fig. 9 (d) shows that fatigue loading of the UA
alloy in vacuum leads to a rather ductile fracture surface, whereas humid air causes some
embritteling, as visible in Fig. 9 (c) and (e): The main influence seems to come from the load
ratio, showing extensive crystallographic brittle fracture features at R = -1 in air. The UA
alloy shows in addition crack branching, and the crack advance profile is zig-zag like. The
main differences in FCGR behaviour of the OA and UA microstructure indeed arise from
different slip deformation behaviour: homogeneous and wavy slip in the OA alloy (more
brittle in ambient air than in vacuum, probably induced by hydrogen) and localised planar slip
in the UA microstructure.
The present results for 7049 aluminium alloy tested in ambient air show a distinct trend of
lower threshold hKth values and higher near threshold growth rates with increasing aging
treatment. These features can be rationalised in terms of several competing mechanistic
processes: intrinsic and microstructural effects and microstructure environment interactions.
In the absence of any environment effect, in vacuum, the crack propagation mechanism is
governed only by microstructural factors whose action in turn is governed by the loading
conditions [54,55]. Crack propagation is intergranular, controlled by slip in one or many
active planes. In the crack growth range where the Paris law is valid, Le., in stage 11, the crack
tip loading conditions permit at least two slip systems to be active which in turn leads to a
plane crack growth path affected only by the presence of large inter-metallic precipitates [14].
The chemisorption phenomenon describes the formation of hydrogen by the dissociation
of the absorbed water molecules. In such a case a hydrogen embrittlement mechanism can be
brought into action [26,50,53]. Accordingly the thresholds uti, for both aging conditions are
higher in vacuum than in humid air at all load ratios. Moreover, the differences between the