The dynamic tensile strength of ice and icesilicate mixtures

We determined the dynamic tensile of fracturing and fragmentation properties of icy strength of ice and ice silicate mixtures at substances. Our goal in previous studies has been strain rates of •104 s -1. At these strain rates, to establish some data and scaling laws related to ice has a tensile strength of ~17 MPa, and impact crater formation and fragmentation of ice ice-silicate mixtures with 5 and 30 wt % sand and ice-silicate mixtures [Lange and Ahrens, 1981, content have strengths of ~20 and 22 MPa, 1982a, b]. A central observation, pertinent to respectively. These values lie significantly most of these experiments, was the occurrence of above tensile strengths of ~1.6 MPa for ice and of tensile failure as the principal mechanism in the ~5-6 MPa for frozen silt, measured at strain rates fragmentation of impacted icy targets. of ~10 -2 to 10 ø s -1 but markedly below values The goal of the present study is the found for a variety of rocks at comparable strain determination of the dynamic tensile strength of rates. Results of the present experiments are icy media at strain rates of ~104 s -1. These data used to derive parameters for continuum fracturing will help to understand better the macroscopic models in icy media, which are used to determine phenomena observed in our impact experiments and relations between tensile strength and strain will be used in the derivation of scaling laws for rate, and to predict stress and damage histories fragmentation and crater formation in icy as well as size frequency distributions for ice substances. and ice-silicate fragments. It is found that The dynamic tensile strength of ice and tensi.l• streng• o M is related to strain rate by ice-silicate samples with 5 and 30 wt % sand o M • •[o 0'2õ-0' similar to results obtained for content at temperatures between ~230 and 250 K was other geological materials. The increase of small obtained by carrying out experiments in which fragments relative to larger fragments with plexiglas plates impacted target pellets increasing strain rate, as predicted by the consisting of these materials. Upon wave continuum model, is a result which parallels reflection, tensile stress pulses of •0.75 •s findings in laboratory impact experiments. duration and strain rates of ~104 s -1 result. The Introduction shock-loaded samples were recovered and analyzed subsequent to each experiment, and the stress at The strength of crystalline solids is dependent which spallation or breakage of the samples on the mode and rate of stress loading. occurred was defined as the dynamic tensile Compressire and tensile strengths of many strength of the substance. geological materials vary by as much as 1 order of We describe in the next sections the basic magnitude when subjected to either static (strain experimental techniques and results and attempt to rates of •10 -3 s -1) or dynamic (strain rates of use these results to model the fracturing process ~>102 s -1) tests [Grady and Hollenbach, 1979]. of icy media in terms of a continuum model. •nowledge of the dynamic tensile strengths of rocks is essential for an understanding of Experimental Techniques fracturing and fragmentation processes. Applications include many industrial processes Sample Preparation and Assembly involvi• •'•-•' ............ • .... • mining operations [Carter, 1978], impact or All samples were prepared by compressing finely explosive crater formation [O'Keefe and Ahrens, crushed ice (mean grain size ~0.1 to 0.5 mm) or 1976], and processes related to the accretion of ice-silicate mixtures containing specified amounts planetary bodies [Matsui and Mizutani, 1977]. of silica sand (mean grain sizes of 0.1 to 0.5 mm) Although water ice is a common geological into sample pellets of 20 mm diameter and 6 mm material, relatively little has been done to thickness. The grain size of crushed ice was determine its mechanical properties over a wide reduced using a chilled food blender. The ice range of stress and strain rates. Voyager powder or a homogeneous mixture of ice and silica discoveries of impact cratered surfaces on the sand was then filled in a mold and compacted by satellites of Jupiter and Saturn [e.g., Smith et use of a hand-operated hydraulic press. During al., 1979, 1981] have refocused attention on the compression, the sample was evacuated, thus study of physical properties of ice and avoiding extensive trapping of air bubbles. This ice-silicate mixtures. Many of the processes procedure .gave mostly transparent or related to the origin and evolution of the icy semitransparent samples with no observable void moons of Jupiter and Saturn require the knowledge spaces. Extensive recrystallization along grain boundaries of single ice grains took place during *Present address: Alfred-Wegener-Institut f•r sample compression. The sample pellets were Polarforschung, Columbus-Center, D-2850 pressed into annular stainless steel target plates Bremerhaven, Germany. which were cooled via the cooling coil circuit surrounding the sample (Figure 1). Temperatures Copyright 1983 by the American Geophysical Union. were monitored with a thermocouple attached to the target plate, close to the sample. Since the Paper number 2B1806. target plate was thinner than the sample (5 mm 0148-0227/83/002B-1806505.00 versus 6mm, respectively) the plexiglas flyer