The samples used in this study were taken from Sites 897, 899, and 900 (Table 1). Petrographic descriptions given in the discussion below are taken from the shipboard core descriptions (Sawyer, Whitmarsh, Klaus, et al., 1994) and visual inspection of the samples. Site 897 was the most distal site, and Site 900 the most landward site. All samples were taken from those collected for routine shipboard velocity analyses. Minicores 25 mm in diameter, oriented perpendicular to the core (horizontal to the drill hole), were taken from the selected pieces, and the ends were trimmed to form parallel faces perpendicular to the minicore axis. The cores ranged in length from 19.5 to 25.6 mm. The direction of wave propagation was along the axis of the core, and so represents horizontal wave propagation in the Earth. All velocity measurements were done on dry samples with no confining pressure.
The experimental apparatus is shown in Figure 1. The technique is similar to that developed by Sondergeld et al. (1990), except that we examined the arrival time of a constant phase in the waveform (i.e., the first trough or peak) rather than picking a first break. The samples were placed in a rotating clamp, with shear-wave transducers on either end of the minicore. A 400-kHz electronic pulse similar to that described by Harry and Batzle (this volume) was sent through the transducer at one end of the sample, generating an acoustic signal that was detected by the transducer at the other end of the sample. The orientation of the transducers did not vary during the experiment, so that particle displacement was always vertical. Traveltimes were recorded at 10° intervals as the samples were rotated 360°. The shear-wave velocity at each azimuth was calculated by dividing the measured traveltime by the sample length (after correcting for the time required for the pulse to travel through the transducer assembly). Thus, minimum traveltimes were recorded at azimuths corresponding to the greatest shear-wave velocity.
Acoustic anisotropy (in percent) is given by
where Vmax and Vmin are the maximum and minimum velocities and Vmean is the mean velocity (Kern, 1993). For simple transverse anisotropy (such as might arise from a single family of preferred mineral grain orientation, foliation, or veining) the samples should display fast and slow directions oriented 90° from each other (e.g., Kern and Tubia, 1993). This appears as two cycles of a periodic wave in the velocity vs. azimuth plot. Sample inhomogeneities or inconsistent transducer coupling introduces irregularities in the appearance of the simple harmonic wave. Microcracks opening after sample recovery may also contribute to anisotropy (Christensen and Wepfer, 1989). The samples studied show no obvious signs of post-recovery macrofractures, but measurements at elevated confining pressure would be required to test for the presence and effect of microcracks. The closure of microcracks is believed to cause the increase in velocity and decrease in anisotropy with increasing pressure that are typically observed in laboratory measurements of crystalline rocks (Christensen and Wepfer, 1989).