utilized Al–Cu specimens with GP zones aligned in a
common direction. This allowed for the effect of the angle
between the Burgers vectors of active dislocations
and the GP zone planes to be examined collectively.
The experiments revealed that at room temperature
there is no noticeable effect of the Burgers vector–GP
zone orientation on strength. However, at 77 K the
strength of specimens with active dislocations having
Burgers vectors parallel to the plane of GP zones
showed a significantly enhanced strength. To connect
the results of the simulations performed here with the
experiments of Eto et al. and Muraishi et al., we utilize
data from our previous work on edge dislocation–GP
zone interactions [19] (Table 1). While the effective resistance
to dislocation glide in experiment is likely to be a
complex function of sc for both screw and edge dislocation
characters, reasonable bounds can be assumed. A
lower bound of effective glide resistance would consist
of the arithmetic mean of sc from both screw and edge
character dislocation–GP zone interactions. An upper
bound would consist of only sc from the stronger dislocation
character interaction. Regardless of which bound
is considered, the atomistic simulation results are consistent
with experiments in two important ways: (i) GP
zones act as stronger obstacles to dislocation glide at
0 K when the Burgers vectors of active dislocations are
parallel to the planes of GP zones due to the large
strength increase associated with screw dislocations
cross-slipping onto impeded planes. It should be noted
that if only edge dislocations are considered the opposite
conclusion contrasting experiment is reached and (ii) the
strengthening effect from screw dislocations, with Burgers
vectors parallel to GP zone planes, cross-slipping
onto impeded planes varies inversely with temperature
due to thermal activation aiding the second cross-slip
event, while the other orientations and mechanisms
show a much milder temperature dependence at molecular
dynamics strain rates.
In summary, the atomistic simulations performed
here have highlighted the importance of dislocation
cross-slip in precipitation hardening. It is found that
cross-slip can either decrease or increase the hardening
effect of precipitates, depending upon whether the
cross-slip plane is impeded by the precipitate. Cross-slip
is observed to occur at both 0 and 300 K, suggesting that
it can occur in the absence of thermal activation. The
importance of cross-slip in precipitation hardening is
highlighted when considering the simulation results in
the context of experimental data which cannot be explained
solely by edge dislocation precipitate interactions.
When cross-slip does not occur or does not
provide an unimpeded plane for the dislocation to overcome
the precipitate, Orowan looping is observed in 0 K
simulations. Considering that the precipitates are GP
zones, this is in contrast to the long-held continuum
view that dislocations overcome small precipitates via
cutting. This analysis, in combination with our previous
report [19], forms the foundation for atomistic-based
modeling of age hardening, the focus of our forthcoming
work.