Laboratory studies of Dislocation Damping

Robert Farla¹, Harri Kokkonen¹, John Fitz Gerald¹, Auke Barnhoorn^1,2, Ulrich Faul³ and Ian Jackson¹

1 Research School of Earth Sciences, Australian National University, Canberra, ACT 0200, Australia
2 Now at Department of Earth Sciences, Utrecht University, 3508 TA Utrecht, The Netherlands.
3 Department of Earth Sciences, Boston University, Boston, MA 02215, USA

Last year we presented some preliminary results on dislocation annealing in fine-grained synthetic olivine [1]. Since analyses of these results showed that dislocations can be preserved over laboratory timescales, the first experiments to search for dislocation damping in polycrystals have now been conducted.

Two similar deformed synthetic olivine specimens have been investigated so far. The first specimen, 6618, has been deformed in compression up to 2.3% strain, the second, 6646, to 22%. The resulting dislocation densities are 2.3 and 3.6 µm-2 respectively. The specimens were exposed to a maximum temperature of 1100°C, and to oscillating torques of 1 - 1000s periods at each temperature decrement, generating maximum shear strains of around 10-5. The strains are sensed from displacements measured with parallel-plate capacitance transducers.

The first attenuation experiment in torsion on specimen 6616 yielded a surprise null result. This led to some calculations for resolved shear stresses in relation to deformation in compression or torsion and assuming that, at high temperatures, the dominant slip system for olivine involves slip in the [100] direction on the (010) glide plane. The main results are shown in figure 1 for [100](010) slip in a single crystal olivine. The two panels (a) and (b) represent deformation in compression and torsion respectively. The parameters α and β (and γ) are the direction cosines relative to the crystal axes [α, β, γ] and some major crystal directions are labelled. In uniaxial compression, the resolved shear stress as indicated by the contours in (a) has its maximum value of 0.5 for compression parallel to [110]c. The contours in panel (b) describe the azimuth-dependent  resolved shear stress for different torsion axes. Note that torsional deformation around the [110]c is the least favoured orientation for stressing [100](010). In contrast, a cylinder axis parallel to [111]c (ie making equal angles with all 3 principal unit-cell axes) is a good trade-off if a single crystal were to be deformed in compression then transferred to attenuation measurements in torsion. For the case of polycrystalline olivine, preliminary calculations show that prior torsional deformation should increase the anelastic strain by 6-fold relative to prior compressional deformation.

Figure 2 shows measurements of Q-1 for both deformed specimens with different dislocation densities in comparison to a similar synthetic but undeformed olivine specimen. Only specimen 6646 with the higher dislocation density shows enhanced attenuation and only at the highest temperatures (> 1000°C). In nature, dislocation density should be in equilibrium with the prevailing tectonic stresses in the upper mantle. Dislocation damping will become progressively more important as the grain-boundary dissipation decreases with increasing grain size. Accordingly, dislocation damping could dominate in melt-free material under upper-mantle conditions.

Since the resolved shear stress calculations indicated the need, prior torsional deformation experiments will be done through a collaboration with the University of Minnesota in 2009. Also for single crystals we plan to investigate the possibility of deforming along [111]c in compression. For both materials, subsequent measurement of the forced torsional behaviour will proceed at ANU.

 

 

---

[1] Farla RJM , Kokkonen H, Fitz Gerald JD, Barnhoorn A, Faul UH and Jackson I (2008) Dislocation recovery in fine-grained synthetic olivine. In preparation for submission to Physics and Chemistry of Minerals.