Geoscience Reference
In-Depth Information
promotes further deformation and the generation
of small grains, leading to the runaway instabil-
ity. The necessary condition for this localization
mechanism to operate is that the size of dynam-
ically recrystallized grain is smaller than that of
a critical size for the transition between diffusion
and power-law dislocation creep, and is given by
bulk diffusion of various diffusing species makes
the application of diffusion data complicated (e.g.,
Gordon, 1973, 1985). In fact, there have been
no theoretical studies on the creep strength of
any geological materials that predicted the plas-
tic properties including the creep strength and
the dominant slip systems correctly for miner-
als. Consequently, experimental studies play the
most important role in our understanding of the
rheological properties of the Earth and planetary
interiors.
Plastic properties depend on a number of param-
eters and many of which change plastic properties
(e.g., strain-rate for a given stress) by several or-
ders of magnitude. Therefore a careful control
and characterization of chemical environment
and microstructures, and a choice of appropri-
ate method of experiments (or the development
of new methods when needed) are critical in the
study of plastic properties. Important factors that
need to be characterized or controlled include the
water content (water fugacity) and the grain-size.
The importance of controlling or characterizing
water content must be emphasized because the
water effects are large and water may be lost or
added to the sample during an experiment. Water
content of a sample must be measured both before
and after each experiment. Similarly grain-size
must be measured both before and after each ex-
periment when a polycrystalline sample is used.
Although effects are relatively small, oxygen fu-
gacity and oxide activity must also be controlled.
The choice of an apparatus must be made based
on the proper understanding of theory of plas-
tic deformation and of thermodynamic properties
of materials under high-pressure and temperature
conditions. If a necessary apparatus does not exist,
then one should design a new one to achieve the
scientific goal. Key aspects in choosing or devel-
oping deformation apparatus or methods are (i) the
range of pressure and temperature in which the
machine can be operated and (ii) the resolution of
mechanical measurements. A common trade-off
is that a high-resolution testing machine such as a
gas-medium deformation apparatus has a limited
pressure range of operation (P
<
0
.
5GPa), whereas
apparatus that can be operated to higher pressures
A
r
A
pl
A
diff
a
−
n
−
m
exp
H
pl
−
H
∗
D
mRT
.
m
<
σ
μ
(4.16)
H
∗
D
>
0formost
materials, this condition means that shear lo-
calization likely occurs at high stresses and low
temperatures.
This is the necessary condition for shear lo-
calization. In order for substantial deformation
to occur by this mechanism, the growth rate of
newly formed grains must be small enough. Low
temperature favors slow growth rate, but for pure
material such as pure olivine aggregates, growth
rate is still so fast (Karato, 1989b) that substantial
shear localization does not occur. The presence of
a secondary phase that retards grain-growth rate
is needed to achieve substantial shear localization
(Karato, 2008). I will come back to this issue later
in this chapter when I discuss the deformation of
the lithosphere.
Localized deformation due to stress-induced
melt migration was studied by Holtzman
et al
.
(2003a,b). Such a process is likely important near
mid-ocean ridges where extensive partial melting
occurs. However, importance of these processes
in the asthenosphere far from ridges is unclear.
n
−
m
>
0and
H
pl
−
1
Because
a
−
4.3 Experimental Methods in Deformation
Studies
Because plastic deformation involves a number
of processes as discussed above and also because
defects involved in plastic deformation have low
symmetry, theoretical modeling is difficult and
it is essential to obtain experimental results (see
Section 4.6, Theoretical Studies, below). Even in
the case of diffusion creep where theory is well
established, the interplay of grain-boundary and
