RHEOLOGY OF ICE II AND ICE IIl FROM HIGH-PRESSURE EXTRUSION

Rheological parameters for ice II Experimental Method and ice III, needed in tectonic models of the icy satellites of the major planets, are obtained from Extrusion experiments were carried out as extrusion experiments and compared with the rheolfollows. Approximately 2 cm $ of distilled water ogy of ice I at pressures ~2 kbar and temperatures was introduced into the 9.5-ram bore of a piston~240K, .Ice II has a higher effective viscosity cylinder apparatus, the upper piston of whicb was . (by a factor ~10) than ice I at similar stress loaded by a hydraulic ram to pressurize the sam, levels, whereas ice III has a lower viscosity (by ple. Extrusion of the sample took pla'c'e through a a factor ~0.01). The rheological contrasts a•ong narrow cylindrical orifice along the axis of the the ice phases are related to differences in the lower piston, which was supported in fixed posidielectric relaxation behavior and state of proton tion in such a way that the extruded ice could order/disorder in the structures in a way that escape at atmospheric pressure from the bottom. sheds light on the nature of dislocation motion in Orifices of diameter 1 to 2 ram, and of length ice. A striking transformation plasticity accom12.5 mm, were used in different experiments. The panies the ice I-III transition and could have bomb was immersed in a cooling bath for large tectonic effects. temperature control to ñIøC. In each experiment, sample temperature and pressure were brought to Introduction desired values (Fig. 1), and the rate of extrusion ß was measured by monitoring the advance of the Because the interiors of several of the moons upper piston,. detected with a displacement transof the major planets consist to a large extent of ducer and recorded on a strip-chart recorder along ice in its high-pressure forms [Lupo and Lewis, with the hydraulic-ram pressure. 1979], the thermal and structural evolution of these moons (e.g., subsolidus convection) and its Results and Analysis of Data reflection in the morphology of their surfaces (e.g., impact crater relaxation) is dependent on At sample pressures below the phase transithe rheological properties of the dense ice phases tion to ice II or III the extrusion rate is gov[Poirier, 1982]. Additional interest in the erned by the rheology of ice I, averaged over the rheology of these phases arises at a fundamental pressure range from sample pressure to atmospheric level with regard to general relationships between and over the shear stress range across the orifice. If ice I obeys the non-linear flow law ß n 1 rheology and structure among polymorphs Because • = Ai• where ß is shear stress, • is shear proton disorder and the associated mechanism of ' proton rearrangement i the hydrogen-bonded strucstrain rate, and A 1 and n 1 are rheological constants, whose variation with pressure is small tule of ice I are thought to have an important effect on the creep process [Glen, 1968], and because the other ice phases have different structures and proton order/disorder properties [Kamb, 1973], a comparison of their rheologies offers the possibility of obtaining new insight into the role of these structural features in creep. Preliminary results of an ongoing experimental study of the creep properties of ice II and III by constant-strain-rate tests at high pressure have been reported by Kirby, Durham, and Heard [1985]. Because of the importance of the subject, we believe it of value to present briefly the results of an independent set of experiments we have carried out by a quite different method, namely, extrusion from high pressure. Our results agree in a general way with those of Kirby et al. [1985], and in addition reveal a flow runaway phenomenon that is a dramatic manifestation of transformation plasticity. Copyright 1986 by the American Geophysical Union. Paper number 6L6981. 0094-827 6 / 86 / 006L-6981 $ 03. O0 enough to be neglected here [Weertman, 1983], and if there is no slip at the walls of the orifice in the extrusion, then the extrusion rate, expressed in terms of .the speed of piston advance U, will be U -Ai(P/L) nl anl+3/[b2(nl+3)2 nl] (1)