Herbert T. Ueda

Structure and temperature dependence of the flexural properties of laboratory freshwater ice sheets

Abstract In this report we present results of small-beam testing conducted in a test tank on ice corresponding in structure to the two major ice types, S-1 and S-2 encountered in lake ice-sheets. Tests of 730 beams in the temperature range −1 to −19°C showed that macrocrystalline (S-1) and columnar (S-2) ice differ appreciably in their flexural characteristics and that these differences are attributable to variations in the size and orientation of the crystals in the ice and the thermal condition of the beams. Parallel testing of cantilever and simply-supported beams indicated a virtual non-dependence of flexural strength on the temperature of the fiber in tension. It was also determined that the sharply terminated corners of conventional cantilever beams are a source of appreciable stress concentration that can reduce the intrinsic flexural strength by as much as one-half but which, in most cases, can be substantially relieved by drilling holes at the beam roots. Overall, flexural strengths did not exceed 1200 kPa for cantilever beams or 1650 kPa for simply-supported beams tested in parallel. The highest flexural strengths were measured on isothermal simply-supported beams of S-2 ice tested with the top surface in tension, with average strengths for such ice increasing from 1650 kPa at −1°C to nearly 2600 kPa at −19°C. Beams made to fail with bottom in tension, tested about 35% weaker due to the greatly increased size of crystals in the bottom of S2 ice sheets. Beams of S1 ice yielded flexural strengths mid-way between those measured on S2 ice. This behavior, which occurs despite the fact that crystal size in S1 ice is always very much larger than in the coarsest-grained S2 ice, is attributed to the dominant vertical c-axis structure of S1 ice that forces tensile failure to propagate in the hard-fail plane of the crystals. It was also determined that temperature gradients in simply-supported beams can decrease flexural strengths by as much as 50% compared to isothermal beams tested at the same ambient air temperatures. Strain-modulus measurements showed little dependence on either the temperature of the beam or its flexural strength, with values ranging from 4–6 GPa for cantilever and parallel simply-supported beams, and from 6–8 GPa for isothermal simply-supported beams.