The viscosity of the outer core is a critical parameter in geodynamo
modelling. It has been assumed from theoretical arguments that the liquid
Fe-alloy in the outer core is inviscid. This has never been verified experimentally
and the many estimates of outer core viscosity, from geomagnetism, seismology,
geodesy, and theory, span 14 orders of magnitude. Viscosity experiments
on Fe-S liquids are being carried out in a multianvil apparatus at pressures
and temperatures that are low compared with core conditions but with the
aim of determining pressure and temperature dependences (activation volume
and energy, respectively) of viscosity in this important geophysical system.
The typically used high pressure Stokes viscometry 'quench and probe' technique
is not suitable for this system because of the low viscosities and the
lack of requisite control of the rise/fall time of a sphere. Instead, we
employ the 'electro-detection method' which was developed for in-situ
sphere speedometry measurement. The method is based on the detection of
sphere position from an electrical resistance anomaly produced by an insulating
sphere passing through a pair of electrodes located in the conducting liquid
sample (Fig. 3.6-14). Spheres of ruby or sapphire,
|
0.5-0.7 mm diameter, are used in boron nitride contained samples of
2.5 mm diameter and 2.5-3.0 mm height. Ruby is inert in Fe-S liquids and
provides the requisite electrical conductivity contrast (~107)
with the sample. Using Fe electrodes, the measured sphere velocity incorporates
the electrical size of the sphere, determined from previous potential field
modelling experiments, which is ~2.6 times the physical size. The experiments
are carried out with heating rates as high as 2750oC/min using
LaCrO3 furnaces and W/Re thermocouples. Viscosity results for
Fe-20.7wt% at 7 GPa show an Arrhenian T-dependence (Fig. 3.6-15). At pressures
of 6 and 8 GPa and for temperatures in the range 1100-1350oC,
the viscosities of Fe-27wt%S are in the range 1-6 Pa-s, or ~3 orders of
magnitude higher than pure liquid Fe at 1 atm. These values are in agreement
with extrapolations of previous viscosity measurements on the same composition
up to 5 GPa and 1300oC in a cubic anvil press. Using the same
configuration and sample volume, we plan to extend the pressure range of
viscosity measurement to 15 GPa in a 5000 ton press.
|