Parameters defining a single constitutive relation are often ˚tuned' on an ad hoc basis in individual studies, with the result that an acceptable description of field observations may be obtained. The most obvious case for such a development involves the inclusion of a separate rheology for the debris-rich basal ice layer that lies at the base of most ice masses. Incorporation of variability in ice character, and therefore rheology, into models of ice-mass motion has proven difficult due to lack of evidence relating to the systematic spatial variability of these properties. Controls over rheology in ice masses are multivariate, and isolating the roles of individual factors has proved difficult in both field- and laboratory-based studies. Relationships between ice character and rheology are difficult to investigate.

These rheological variations have not yet been incorporated into any glacier flow models, and there are at least three major reasons for this: Experimental and field observations have supported this ( Reference SwinzowSwinzow, 1962 Reference Kamb and LaChapelleKamb and LaChapelle, 1964 Reference Kamb and ShreveKamb and Shreve, 1966 Reference Goughnour and AnderslandGoughnour and Andersland, 1968 Reference Hooke, Dahlin and KauperHooke and others, 1972 Reference Echelmeyer and ZhongxiangEchelmeyer and Wang, 1987 Reference PatersonPaterson, 1994). Intuitively, rheological variability must occur as a result of the differences in the physical character of ice found throughout ice masses but most profoundly within the lower basal ice layers. Field sociologists have known for a long time that ice character varies markedly in real ice masses, especially within the lower ˚basal ice' layers of glaciers ( Reference Kamb and LaChapelleKamb and LaChapelle, 1964 Reference LawsonLawson, 1979 Reference KnightKnight, 1997) and within the deforming zones of alpine rock glaciers ( Reference Vonder Muhll and HolubVonder Mühll and Holub, 1992). However, the importance of the transient creep effects has been demonstrated by Reference Azizi, Whalley, Chung, Sayed and GresnigtAzizi and Whalley (1995), although parameterization of such effects remains problematic. Reconstructions of the dynamics of small ice bodies such as alpine rock glaciers ( Reference KääbKääb, 1996) have used steady-state rheologies for simplicity. Such transient effects dominate the behaviour of small ice bodies in their initial stages. Where A and m are creep parameters, A being the same as that used in Equation (1).Ĭlearly, where m = 0, steady-state conditions apply, and where m = 1, accelerating (tertiary) creep is assumed. Values for A and n in the flow law derived from the experiments are reasonable and indicate the potential of this method for ice-deformation studies. Samples are instrumented with displacement markers and thermocouples. Deformation can therefore be treated as simple shear. Models are unconfined on their downslope ends, rendering longitudinal stresses negligible. Models are confined laterally by the Perspex strongbox walls, preventing lateral extension within the sample during testing. Thus we were able to model the deformation of the lower 14 m of an ice mass approximately 50 m thick.

Steel plate was placed on top of the ice model, replicating an overburden of approximately 36 m of ice (at 80 g). In the following paper we describe the physical modelling of two (laboratory-prepared) isotropic, polycrystalline ice models (0.75 × 0.25 × 0.18 m) at enhanced gravity levels (80 g) in a geotechnical beam centrifuge. Obviously the best method for investigating ice in the laboratory would be if stress gradients similar to those found in Nature could be replicated. Detailed experimental analyses of ice samples in the laboratory have until now only been undertaken using uniform stress fields in uniaxial or triaxial tests. ice thickness, gravity and ice density) which produces a stress gradient within the ice column. In Nature, ice deformation is the response of the glacier/ice sheet to the applied self-weight stress of the ice mass (i.e. Glen’s flow law is a well-established general law for steady-state glacier ice deformation, and many laboratory tests and field measurements have been undertaken which have shown the generality of the law to be correct.

Strongbox thermal expansion