The results for Hole 897D are shown in Figure 2. Sample 149-897D-23R-6 (Piece 3B) was taken from an interval of the core with sparse serpentinite-filled fractures oriented approximately 30° to vertical. The sample was taken from between veins, and it has only two thin veins visible in it. The 0° azimuth is oriented parallel to the veins. The core description reports that the piece from which this sample was taken has been replaced by 98% mesh serpentinite (Sawyer, Whitmarsh, Klaus, et al., 1994). The shear velocity varies between 1535 and 1620 m/s, with a maximum anisotropy of 5%. Slow directions are oriented at azimuths of 90° and 280°, which corresponds to particle displacement perpendicular to the orientation of the veins. Velocities vary smoothly between the fast and slow directions. The direction of slow wave propagation is consistent with the lattice-preferred orientation of vein-filling serpentine in the fractures, because serpentine is anisotropic with a slow direction of wave propagation corresponding to particle displacement across the fibers (Kern, 1993). However, mesh serpentinite is also anisotropic and the mesh texture may be at least partially inherited from the primary mineralogy (Deer et al., 1978). Thus, it is not possible to determine with certainty whether the anisotropy arises from the vein orientation or from primary preferred mineral orientation. The 5% anisotropy observed in this sample can be reasonably attributed to preferred mineral orientation, which may develop anisotropy up to 15% in olivine group minerals (Christensen and Salisbury, 1979) and has been observed as high as 35% in serpentinites at high pressure (600 MPa; Kern, 1993).
Sample 149-897D-24R-1 (Piece 3C) was taken from a moderately veined interval, and it contains less primary olivine and more primary pyroxene than Sample 149-897D-23R-6 (Piece 3B). Vein density is sufficiently high that the sample contains several visible veins, oriented approximately 30° to vertical. The 0° azimuth is parallel to the veins. Shear-wave velocities in this sample range from 1478 to 1608 m/s, for an anisotropy of about 8%. Except for a narrow range of azimuths showing slow wave propagation near 100° and a narrow range of fast propagation near 260°, the velocities are fairly uniform at ~1570 m/s. The slow direction of propagation corresponds to particle displacement nearly perpendicular to the vein orientation. If the sample was transversely isotropic in the plane of wave propagation, then two slow and two fast azimuths are expected. For example, the fast azimuth should have an equally fast azimuth oriented 180° away, because this would result in identical senses of particle displacement in the sample. The lack of two fast and slow azimuths in Sample 149-897D-24R-1 (Piece 3C) suggests that the sample is not simply homogeneous and transversely isotropic, and, in fact, is typical of samples with veins cutting at oblique angles through the sample. When a shear wave impinges on a vein at an oblique angle, it is split into two waves with orthogonal particle-displacement directions (e.g., Kern, 1993), The wave with particle displacement oriented parallel to the vein is faster than the wave with particle displacement oriented perpendicular to the vein. In this sample, an obliquely oriented vein much larger than the other veins cuts the sample at an angle of about 30° to the minicore axis. We interpret the velocity data to indicate shear-wave splitting of waves striking at high angles to the orientation of this vein. The steady trend of slightly increasing velocity at azimuths between 130° and 260° indicates that the fast wave is being picked from the traveltime data, producing the narrow range of azimuths with velocities significantly different than 1570 m/s. There is no indication for significant anisotropy in the sample matrix.
Sample 149-897D-25R-1 (Piece 3) was taken from an interval showing fairly dense vertically oriented veining. The sample contains numerous veins of similar size and orientation. Like Sample 149-897D-23R-6 (Piece 3B), the primary mineralogy of this sample is mostly olivine (85%), with 15% pyroxene. The 0° azimuth is oriented parallel to the veins. This sample shows remarkably large variations in shear-wave velocity, from 1307 to 1677 m/s. Anisotropy is calculated to be 25%, with fast directions of wave propagation at 0° and 180°. This corresponds to particle displacement directions parallel to the vein orientation and is consistent with the anisotropy characteristics of fibrous vein-filling serpentinite. The shear velocity varies smoothly between the fast and slow azimuths, as is expected if the vein orientation controls the anisotropy and no single vein dominates. The anisotropy in this sample is much larger than that observed in any of the other samples studies, but is reasonable in view of the 35% anisotropy observed in serpentinite under high pressure by Kern (1993).
The results for Hole 899B are shown in Figure 3. Sample 149-899B-21R-1 (Piece 1B) was taken from a pyroxene-rich clast in the Lower Breccia Unit at this site. The sample contains no significant veining. The shear-wave velocity for this sample shows little variation with azimuth, remaining close to 1590 m/s, and calculated anisotropy is less than 1%. This indicates that preferred grain-boundary orientation is absent or is insufficient to produce measurable anisotropy.
Sample 149-899B-31R-1 (Piece 12C) was taken from a nonbrecciated pyroxene-rich peridotite that has minor serpentinite veining. The 0° azimuth is oriented parallel to the veins. The sample shows a smooth variation in shear-wave velocity from 1675 to 1796 m/s, resulting in 7% anisotropy. The slowest propagation speeds are at azimuths of 90° and 280°, where particle displacement is approximately perpendicular to the dominant vein orientation.
Sample 149-899B-34R-1 (Piece 5) was taken from a piece of mylonitized peridotite, with the 0° azimuth oriented parallel to the mylonitization. The sample has unusually high shear-wave velocities at azimuths less than about 80° and greater than 320°. Arrival times at these azimuths are interpreted to come from a converted wave that is reflected off a break in the sample, which occurred during sample preparation. The break is oriented approximately 30° to the axis of the minicore and its strike is perpendicular to the 50° azimuth. The break cuts one end of the minicore approximately 3 mm from the edge. The orientation of the break is favorable for conversion of the shear wave to a compressional wave at azimuths close to 50°, producing the anomalously high velocities. Inconsistent coupling of the transducer to the sample may also be partly responsible for mode conversion, although no visual indication of coupling problems was apparent during measurement. Velocities between 80° and 320° are inferred to represent the direct shear wave. The sample shows a smooth variation in shear-wave velocity in this interval, with a maximum propagation speed at of 1690 m/s at 190° and slow propagation speeds of about 1590 m/s at 110° and 270°. This results in about 6% anisotropy, with the slow propagation direction corresponding to particle displacement perpendicular to the orientation of the mylonite fabric.
The results for Hole 900A are shown in Figure 4. Sample 149-900A-81R-3 (Piece 1) was taken from a foliated breccia clast, interpreted to have originally been of a mafic protolith (Sawyer, Whitmarsh, Klaus, et al., 1994). The 0° azimuth is parallel to the foliation orientation. Shipboard fabric analyses indicated that brecciation postdated the foliation, or at least continued after foliation. In spite of the well-developed foliation, this sample shows little azimuthal variation in shear-wave velocity. Anisotropy is calculated to be less than 2%, with an average shear-wave velocity of about 1900 m/s.
Sample 149-900A-85R-4 (Piece 1D) is a strongly foliated piece with a later generation of calcite veins oriented parallel to the 150° azimuth. The 0° azimuth is oriented parallel to the foliation direction. Velocity varies from 1870 to 1990 m/s, with the maximum velocity occurring at 0° azimuth, where particle displacement is parallel to the foliation. The minimum velocity occurs at 240°, where particle displacement is perpendicular to the vein orientation. An intermediate maximum velocity is measured at an azimuth of 180°, corresponding to particle displacement direction parallel to the foliation direction. The maximum anisotropy is calculated to be 6%. We infer that anisotropy arising from the foliation may be the dominant control on shear-wave propagation speed, except at azimuths where the particle displacement is nearly perpendicular to the vein orientation. The highest rate of wave propagation occurs when particle displacement is parallel to foliation. However, where the particle displacement is close to perpendicular to the vein orientation, the vein appears to dominate. The slowest propagation speeds thus occur at an azimuth perpendicular to the vein orientation.
Sample 149-900A-85R-6 (Piece 8) is a mafic rock showing some foliation and at least two generations of veining. The 0° azimuth is parallel to the foliation direction. The dark-colored, older family of veins (epidote?) reported in the core description (Sawyer, Whitmarsh, Klaus, et al., 1994) are oriented parallel to the 310° azimuth and cut the sample obliquely at an angle of about 20° to the minicore axis. A well-developed later family of epidote veins is oriented parallel to the 340° azimuth. Velocities range from 1680 to 1890 m/s, for a calculated anisotropy of 12%. As with Sample 149-900A-85R-4 (Piece 1D), the multiple vein and foliation directions result in a complicated variation of shear velocity with azimuth. In this sample, the vein orientation appears to dominate the velocity variation, because neither major nor minor maxima and minima correlate well with the foliation orientation (Fig. 4). The highest velocities are at azimuths close to the orientation of the two sets of veins (near 150° and between 300° and 340°), where particle displacement is approximately perpendicular to the veins. The second set of veins (at an azimuth of 150°) appears to best match the fastest directions of propagation.