An Experimental Investigation of the Effect of Grain Size on "Dislocation Creep" of Ice. Issue 9 (1st September 2021)

• Record Type: Journal Article
• Authors: Qi, Chao; Goldsby, David L.
• Is Part Of: Journal of geophysical research. Volume 126:Issue 9(2021)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-09-01
• Subjects: ice -- rheology -- dislocation creep -- grain boundary sliding -- creep; Geomagnetism -- Periodicals; Geochemistry -- Periodicals; Geophysics -- Periodicals; Earth sciences -- Periodicals; 551.1
• DOI: 10.1029/2021JB021824
• Languages: English
• ISSNs: 2169-9313

Abstract

Abstract: The creep behavior of ice is believed to be dominated by dislocation creep at high stresses ( > 1 MPa) and high homologous temperatures ( > 0.9). Dislocation creep of ice is often described by the Glen law, ε ̇ = B σ n, with a canonical value of n = 3, and is independent of grain size. Laboratory studies of ice deformation at elevated pressures, however, suggest that dislocation creep of ice is characterized by a value of n ≈ 4. Here, we deformed ice samples with initial grain sizes of 0.23 and 0.63 mm, at 263 K and confining pressures of 10 or 20 MPa. The experiments yield a value of n = 3.6 using the peak stresses obtained at a strain ε ≈ 0.02, and reveal a marked dependence of the stress on initial grain size. The nominally constant flow stresses at larger strains, up to 0.2, yield a value of n = 3.9, and no influence of the initial grain size on the stress is observed. The lower value of n at peak stresses and the decrease in peak stress with decreasing initial grain size suggest a contribution to the strain rate from a grain‐size‐sensitive process, such as grain boundary sliding. Our data yield two flow laws, one for isotropic ice and one for highly anisotropic ice deformed to large strains, the latter of which may be used to model flow in high‐strain natural environments. Plain Language Summary: The flow of ice in glaciers and ice sheets is believed to be dominated by the dislocation creep mechanism, in which the strain rate depends on stress raised to a power n and is insensitive to grain size. A canonical value of n = 3 is widely used; however, laboratory experiments often reveal a value of n = 4. To reconcile this discrepancy, we carried out laboratory deformation experiments on ice and analyzed mechanical data at different strains. Data obtained at small strains, representing a homogeneous, isotropic microstructure, reveal n = 3.6. Data obtained at larger strains, representing an anisotropic microstructure, reveal n = 3.9. More importantly, a grain‐size dependence was found in the putative "dislocation creep" regime, suggesting that grain boundary sliding contributes to the deformation. This discovery implies that the rheological behaviors of ice need to be reinvestigated. Key Points: Peak‐stress data from constant‐strain‐rate tests reveal a stress exponent of 3.6, for ice with a homogeneous, isotropic microstructure Flow‐stress data from constant‐strain‐rate tests reveal a stress exponent of 3.9, for ice with an anisotropic microstructure A grain‐size dependence is observed in the peak‐stress data, suggesting a contribution to the strain rate from grain boundary sliding