Deformation Mechanisms: Granular Flow - Materials Science for Structural Geology - page 227

It is widely observed that pronounced grain growth occurs during deformation
in the superplastic regime unless there are factors such as the presence of second
phases, to inhibit it (Clark and Alden 1973 ; Suery and Baudelet 1973 , 1980 ; Watts
and Stowell 1971 ; Wilkinson and Cáceres 1984 ). It seems therefore that within a
given phase there is some sort of coupling between the sliding on grain boundaries
and their migration. Possibly, during the relative movement of the contacting
grains, there are increased opportunities for the boundary atoms belonging to one
grain to enter low energy sites belonging to the other grain, thereby promoting the
migration of the boundary. On such a view, there might exist independently a
driving force for the migration, deriving for example from interfacial energy
decrease, and the grain boundary sliding would simply serve to accelerate the
kinetics of the migration. However, it is possible that sometimes a driving force for
migration may derive from the deformation, if there is an increase in the possibility
of grain boundary sliding through bringing adjacent boundaries into coplanar
configuration.
As far as granular flow itself is concerned, the only kinematic quantity of
importance is the instantaneous relative movement of the main part of one grain
relative to the main part of its neighbour, as reflected in the relative motion of the
centres of mass existing at the instant. However, because of the grain boundary
migration, what constitutes the grain has to be redefined from instant to instant.
This circumstance should not present a conceptual problem if the deformation is
viewed as a summation of successive increments, each taking place at one such
instant and involving the grains as they exist at that instant.
The actual movement pattern of the grains in superplastic flow is not well
known although observations have confirmed neighbour exchange (Ashby and
Verrall 1973 ; Duclos 2004 ; Rai and Grant 1983 ). By analogy with pure particulate
flow ( Sect. 7.2.2 ), it could be expected that there would be both substantial short-
range coordination of relative grain displacements and considerable irregularity.
Associated with this pattern, the stress distribution could be expected to be highly
irregular on the grain scale. Some groups of grains would be subjected to higher
local stress than others and there would be pronounced stress concentrations or
diminutions where relative grain movements tend to produce interference or create
voids. These stress heterogeneities would provide the driving forces for the
accommodation processes, whether diffusional or crystal-plastic.
If the accommodation is by diffusional transfer of material, the principles of this
have been set out in Chap. 5 . If it is by crystal plasticity, the processes are
discussed in Chap. 6 . The principal problem in either case is to elucidate and
quantify the actual grain-scale driving forces and the detailed geometries of the
processes.
In polyphase materials there are further constraints. Thus, grain growth may be
limited for various reasons: for example, for geometric or topological reasons,
depending on the proportions of the phases; or for surface energy reasons if a
finely dispersed phase is distributed along the grain boundaries of another phase.
Where accommodation by diffusive mass transfer is involved, the chemical
identity of the phases must be retained by appropriate constraints on the fluxes of