Grain-size-sensitive creep and its relationship to grain-size-insensitive attenuation in ice-I

Abstract:

Tidal dissipation in the ice shell of, e.g., Europa, occurs in the context of a periodic strain amplitude ε_a ~10^–5 imposed upon a material with a deformation-effected microstructure related to tectonic activity. Tidal flexing “samples” this microstructure; the microstructure's anelastic (attenuation, Q^–1) response effects dissipation. Experiments combining steady-state creep of polycrystalline ice-I with superposed sinusoidal loading demonstrated an attenuation response that is (a) an order of magnitude more attenuating than predicted by the Maxwell Solid model, (b) similar in form to the Andrade Solid model, which at high-temperature and/or low-frequency conditions is approximated by a power law, (c) modestly non-linear (i.e., Q^–1 is a function of the periodic strain amplitude) and (d) insensitive to grain size [1]. The grain-size insensitivity is profound, as steady-state creep of the specimens occurred by the geologically relevant mechanism of grain boundary sliding accommodated by basal dislocation glide (GBS)—a grain-size-sensitive rheology [2]. Here GBS involves emission, motion and interaction of lattice dislocations, and dislocation structures (e.g., subgrain boundaries and secondary grain boundary dislocations) are part of the microstructure sampled by the periodic stress.

 Transient creep responses sample the aspects of creep microstructure that effect attenuation. In our experiments, polycrystalline ice-I specimens are crept to steady state in the GBS regime (σ = 0.5-5.0 MPa, T = 243K, d = 30-150µm) and subjected to instantaneous drops in differential stress: initial strain-rate recovery is related to the creep microstructure of the previous, higher stress. Our results map-out in stress-strain rate space a “hardness” curve consistent with dislocation microstructure self-similarity - a requirement for grain size-insensitive attenuation. Cryogenic electron backscatter diffraction (EBSD) characterizes the associated microstructure. We have not observed significant subgrain development in our samples; the result suggests a first-order role for the grain boundary dislocation structure proper [cf. 3].

[1] C. McCarthy, Ph.D. Dissertation, Brown Univ. (2009)

[2] D.L. Goldsby & D.L. Kohlstedt, JGR 107 (2001)

[3] R.C. Gifkins, Metall. Trans. A 7A (1976)