Dynamics of ice mass deformation: Linking processes to rheology, texture, and microstructure

Abstract

[1] Prediction of glacier and polar ice sheet dynamics is a major challenge, especially in view of changing climate. The flow behavior of an ice mass is fundamentally linked to processes at the grain and subgrain scale. However, our understanding of ice rheology and microstructure evolution based on conventional deformation experiments, where samples are analyzed before and after deformation, remains incomplete. To close this gap, we combine deformation experiments with in situ neutron diffraction textural and grain analysis that allows continuous monitoring of the evolution of rheology, texture, and microstructure. We prepared ice samples from deuterium water, as hydrogen in water ice has a high incoherent neutron scattering rendering it unsuitable for neutron diffraction analysis. We report experimental results from deformation of initially randomly oriented polycrystalline ice at three different constant strain rates. Results show a dynamic system where steady-state rheology is not necessarily coupled to microstructural and textural stability. Textures change from a weak single central c axis maxima to a strong girdle distribution at 35 to the compression axis attributed to dominance of basal slip followed by basal combined with pyramidal slip. Dislocation-related hardening accompanies this switch and is followed by weakening due to new grain nucleation and grain boundary migration. With decreasing strain rate, grain boundary migration becomes increasingly dominant and texture more pronounced. Our observations highlight the link between the dynamics of processes competition and rheological and textural behavior. This link needs to be taken into account to improve ice mass deformation modeling critical for climate change predictions.

Dynamic recrystallization during deformation of polycrystalline ice: insights from numerical simulations

“…When DRX is active this single maximum tends to shift towards being obliquely oriented with respect to σ 1 (or X axis) in the YZ plane (figure 5). Pure shear and uniaxial compression experiments [59,60] of polycrystalline ice result in a girdle fabric of c-axes around the direction of σ 1 (compression direction). The lack of c-axes oriented parallel to σ 1 in our simulations is interpreted as due to the activation of recrystallization processes, coherently with temperatures near to the melting point in the experiments (between −2°C and −7°C).…”

Rheology, microstructure and crystallographic preferred orientation of matrix containing a dispersed second phase: Insight from experimentally deformed ice

“…Even at high homologous temperatures (T h ), there is no chemical reaction between ice and a second phase, allowing evaluation of the physical processes alone. We use deuteric (D 2 O, "heavy water") ice, instead of H 2 O ice, as D 2 O ice has significantly better signal to noise ratio for neutron scattering, with little absorption (Piazolo et al, 2013). Importantly, D 2 O ice has similar structural characteristics (Peterson and Levy, 1957) and rheology (Wilson et al, 2014) to H 2 O ice.…”

“…At the same time, stress and strain heterogeneities are accommodated by strain-induced grain boundary migration (GBM) recystallization and heterogeneous nucleation in areas of high dislocation density (Gottstein and Mecking, 1985). Those two competing processes control the grain size of deforming ice (Wilson and Peternell, 2012;Piazolo et al, 2013). Surface energy-induced GBM may also contribute to ice recrystallization, however, its influence on grain boundary network is negligible as grain boundary energy induced by strain is two orders of magnitude higher than that induced by surface energy (Gottstein and Shvindlerman, 1999).…”

“…During the experiment ice crystal orientation data acquisition is limited to an orientation range of ±40 • from σ 1 , due to the geometry of the measurement (Piazolo et al, 2013 supporting information). Using the intensity of three ice diffraction peaks (10-10), (0001), and (10-11), we are able to calculate the strength of the CPO during the experiment (for details see Piazolo et al, 2013). In addition, we analyse the (001) orientations of graphite.…”

“…We calculated n for ice mixtures using the linear regression method. We used σ ss values at both experimental strain rates (Table 2) and σ ss values (reduced by f ) previously published for pure D 2 O ice (Piazolo et al, 2013). As graphite-bearing samples do not reach the near steady-state, we used the stress and strain rate Fig.…”