Acoustic Double Refraction in Low-porosity Rocks by Terry Todd, Gene Simmons, and W. Scott Baldridge

Anisotropy in physical properties of rocks can arise from preferred mineral orientation, mineral layering, nonhydrostatic stress, and anisotropic crack distribution. For instance, all of the following cause acoustic double refraction: preferential orientation of olivine grains in dunites, alternating layers in laboratory-sizedsamples of such mineral pairs as olivine-feldspar, wollastonite-diopside, and garnetpyroxene, alternating layers of basalt flows and lunar breccias, anisotropy in crack distribution of most granites, and anistropy in crack distribution induced by uniaxial stress. We discuss, both experimentally and theoretically, shear-wave propagation in these rock types and indicate how the laboratory data may be applied to the interpretation of the anisotropy observed in the Earth's crust and upper mantle. We discuss the possibility of elastic anisotropy in the Moon. I N T R O D U C T I O N Values of physical properties of rocks at effective pressures below a few kilobars are determined primarily by mineral content and the presence or absence of microcracks. Preferred orientation ofmicrocracks or individual minerals causes anisotropy in physical properties. Mineral anistropy can be caused by alignment of individual grains or layering of one or more minerals. Crack anisotropy can be naturally occurring or artificially induced. The phenomenon of anistropy can be used to improve our understanding of the Earth's crust and upper mantle through a study of physical properties which are sensitive to mineral orientation and crack distribution. Elastic properties, i.e., shear and compressional velocities, which can readily be measured in situ, are such physical properties. Thill et al. (1969) have shown that compressional-wave velocity is a particularly sensitive indicator of mineral anistropy and crack distribution in dry low-porosity rocks. They found that velocities in the Yule marble were slowest parallel to a direction of distinct preferred orientation of optic [0001] axes of calcite, the slow velocity direction in single crystal calcite, and that velocities in the Salisbury granite were lowest for a direction normal to a plane containing a dense concentration of cracks in quartz grains. However, there are certain inherent problems in the use of their data to interpret compressional velocities of rocks in situ. For instance, the compressional velocity of lowporosity rocks when saturated is typically 20 to 30 per cent greater than the same rocks when completely dry (Nur and Simmons, 1969b). The degree of saturation in situ is generally not known but is probably near 100 per cent. If saturation is complete, pore pressure effects (Todd and Simmons, 1972) must be considered. In addition, in order to determine anisotropy with compressional waves, two perpendicular measurements of compressional velocity are required. The two waves travel different paths, and problems arise involving compositional variations between the two propagation paths. It would be advantageous to measure some physical property which would provide directional information without having to cope with local values of, and local variations in, the percentage of saturation, pore pressure, and composition. For a particular direction of propagation in elastically anisotropic media, there can exist a quasi-longitudinal wave and two quasi-shear waves. Particle motions of the three