Physical properties measured on whole cores at Site 897 included magnetic susceptibility, Gamma-Ray Attenuation Porosity Evaluator (GRAPE) bulk density, and thermal conductivity. Measurements of split cores included Digital Sound Velocimeter (DSV) velocity in unlithified sediments and resistivity. Measurements in discrete samples included wet and dry masses and wet and dry volumes for determining index properties and acoustic velocities using the Hamilton Frame Velocimeter.
Despite slight differences in the depths of the lithostratigraphic units among Holes 897A, 897C, and 897D (see "Lithostratigraphy" section, this chapter), the physical properties from all holes have been plotted together and are discussed as a single section representative of Site 897. These data show several systematic trends that correlate with downhole lithostratigraphic and chemical changes and may point to important variations in hydrologic and mechanical conditions.
Bulk densities, porosities, and grain densities were determined in discrete samples recovered from Holes 897A, 897C, and 897D using gravimetric methods (Table 15; Fig. 61). Uncertainties when determining these quantities can be estimated from the precision of the mass and volume measurements (see "Explanatory Notes" chapter, this volume) and have been estimated as ±0.002 g/cm3 for densities, and ±2% for porosities. Probably less reliable data points near the top of Hole 897C are attributed to instrument difficulties early in the measurement program. The suspect data points are circled in Figure 61. The overall plots of bulk density and porosity vs. depth indicate a general downhole decrease in the water content. The trends are marked by several distinct offsets and changes in slope, some of which coincide with lithostratigraphic unit boundaries.
Bulk densities generally increase from about 1.6 g/cm3 near the seafloor to about 2.45 g/cm3 within basement at a depth of 830 mbsf in Hole 897D. Porosities decrease from about 60% near the seafloor to 10%-20% in basement rock.
Between 600 and about 650 mbsf, bulk density decreases from about 2.3 to 1.9 g/cm3. Porosity increases from 40% to 60% across the same interval. A similar downhole inversion also is visible in the GRAPE bulk density and thermal conductivity data (discussed below). The base of this zone corresponds to the contact between lithostratigraphic Units III and IV (debris flow). Overlying this contact is a 23-m-thick section of very fissile, relatively uniform reddish-brown claystone (lithostratigraphic Subunit IIIA). The reversals in trends of downhole density and porosity originate above this section.
At 650 mbsf, one observes a sharp offset in bulk density and porosity. Bulk density increases from about 1.9 to 2.4 g/cm3, while porosity decreases from 60% to less than 40%. This step coincides with the contact between lithostratigraphic Units III and IV, at the top of the debris flow overlying the basement. Index properties show significant scatter within Unit IV, which reflects the variable lithologies sampled in the debris flow.
Crystalline basement rock composed of serpentinized peridotite (see "Igneous and Metamorphic Petrology and Geochemistry" section, this chapter) was encountered at 677.5 mbsf in Hole 897C, and at 693.8 mbsf in Hole 897D. Bulk densities within this material generally are higher than in the sedimentary section and range from 2.3 to 2.5 g/cm3.Porosities cluster between 15% and 20%. Grain densities within the basement range from 2.5 to 2.75 g/cm3, with a slight decrease downward. These grain densities are lower than those in the sedimentary sequence, which maintain a nearly constant value of about 2.8 g/cm3. The downward decrease in grain density in the basement was interpreted as a consequence of decreasing calcite content in the serpentinized peridotite with depth (see "Igneous and Metamorphic Petrology and Geochemistry" section, this chapter). The decrease in grain density with depth appears to stabilize near 2.6 g/cm3 in the deepest rocks, which is consistent with the density of serpentine (chrysotile) of about 2.55 g/cm3 (Deer et al., 1966).
Several measurements were performed to test the grain density calculations for representative lithologies using a flask pycnometer. Results generally corroborated the grain densities determined using the helium pycnometer, producing values within 0.05 g/cm3 of the measured value.
Bulk densities also were estimated using whole-core GRAPE measurements performed on most sections recovered from Holes 897A, 897C, and 897D. Densities were corrected for the composition of the pore fluids (Boyce, 1976) and for shorter gamma-ray paths in the incompletely filled RCB core liners (Evans and Cotterell, 1970; see "Explanatory Notes" chapter, this volume). Following the methods of Boyce (1973) and Gealy (1971), the maximum values for GRAPE densities were assumed to provide the best estimates of bulk density. The curve describing the maximum GRAPE density was estimated visually and is shown in Figure 61.
The gravimetrically and GRAPE-derived determinations of bulk density show a close correlation throughout the sedimentary section (Fig. 61). The GRAPE-determined bulk densities increase slightly downhole in the sedimentary section, from about 1.8 g/cm3 at 45 mbsf to 2.1 g/cm3 at 600 mbsf. Between 600 and 640 mbsf, GRAPE bulk density decreases from 2.1 to 1.9 g/cm3. This decrease is corroborated by the estimates of gravimetric bulk density and correlates with a sudden increase in porosity near the interface between lithostratigraphic Units II and III. Bulk densities within the basement serpentinized peridotites exhibit a mean value of 2.7 g/cm3, which varies with the visually observed degree of alteration. This value is about 0.15 g/cm3 higher than the bulk density that was obtained gravimetrically. One cause of this discrepancy is the Fe/Mg-rich minerals in the Serpentinized peridotite. These have a higher gamma-ray attenuation coefficient than quartz, which was the standard assumed during data processing (Peterson, Edgar, et al., 1970). Consequently, an artificially high density was computed for the serpentinized peridotite during data reduction.
Electrical resistivities were measured at intervals of 0.5 to 0.75 m in split cores taken from between 110 and 690 mbsf. Resistivity was used to calculate the formation factor, which is the ratio of the electrical resistivity of the sediment to that of the interstitial water (see "Explanatory Notes" chapter, this volume). The formation factor commonly correlates with porosity, but the relationships must be determined empirically for individual lithologies (Boyce, 1980). Formation factor was calculated using the resistance of seawater at room temperature.
The plotted values indicate a general increase in formation factor with depth (Fig. 61). Because resistivity measurements are obtained within the uppermost few millimeters of the surface of the split core and are sensitive to changes in near-surface water content, the computed formation factors must be treated cautiously.
Acoustic velocity was measured in each core recovered in Holes 897C and 897D (Table 16). The DSV was used to measure compressional-wave velocity in split cores containing unconsolidated sediment from Cores 149-897C-1R through -29R. Below Core 149-897C-29R, sediments were sufficiently lithified to allow us to cut dis crete samples from the split core, and velocity was measured using the Hamilton Frame Velocimeter. The quality of the RCB core was inadequate for measurement of acoustic velocity using the MST (see "Explanatory Notes" chapter, this volume).
Discrete samples taken from the sedimentary units were trimmed to a roughly cubic shape, and velocity was measured in three mutually orthogonal directions, one of which was parallel to the long axis of the core. Below Cores 149-897C-63R and 149-897D-11R, 25-mm-diameter minicores were cut from basement units. The minicores were oriented perpendicular to the split face of the core, and velocity was measured along the axis of the minicore. Sample spacing within the basement cores was varied to sample representative lithologies. The accuracy of the velocity measurements taken with the DSV and the Hamilton Frame Velocimeter was estimated as 2% to 3% on the basis of repeated measurements of numerous samples and calibration standards.
Measured vertical Hamilton Frame Velocimeter velocities are remarkably uniform throughout the sedimentary section and increase only slightly from about 1700 m/s at 324 mbsf to about 2000 m/s at 689 mbsf (Fig. 62). Acoustic anisotropy throughout most of the sedimentary section is less than 5% (Fig. 62; see "Explanatory Notes" chapter, this volume). On the basis of the accuracy of the velocity measurements, anisotropy was estimated to be accurate to within 2% to 3%.
In contrast to the cores from the Cenozoic sedimentary section, those from basement and lithostratigraphic Subunit IIIB and Unit IV (see "Lithostratigraphy" section, this chapter) have highly variable velocities that range from about 1950 to 7100 m/s (Fig. 62). The slowest velocities in these cores were measured in Cretaceous(?) clays within lithostratigraphic Unit IV in Hole 897C. Velocities greater than 2700 m/s are from serpentinized peridotite (see "Igneous and Metamorphic Petrology and Geochemistry" section, this chapter). Velocities within the ultramafic rocks segregate into distinct groups that strongly correlate with the degree of alteration estimated from visual inspection (Fig. 63). The least-altered samples (well-indurated, with dark green to black color and lacking common veining) have velocities greater than 6000 m/s. Well-indurated samples containing common serpentine and calcite veins have velocities ranging from 4200 to 5300 m/s. The most highly altered ultramafic rocks have velocities that range from 2800 to 3800 m/s. These rocks are poorly indurated, with common calcite and serpentine veins and clayey texture.
Magnetic susceptibility was measured at 3- to 5-cm intervals in all cores collected at Site 897. Results are discussed within the "Paleomagnetism" section (this chapter).
Thermal conductivity was measured on all cores recovered from Site 897 at a frequency of two measurements per section in the sedimentary sequence and one measurement per section in the basement cores. The mean error associated with these measurements was estimated as 0.2 W/(m·K). Above about 650 mbsf, values of thermal conductivity range between 1.1 and 1.6 W/(m·K) (Fig. 64; Table 17). The data suggest a slight increase with depth, which most likely results from a decrease in porosity (Fig. 61). Below 650 mbsf, within lithostratigraphic Unit IV and the crystalline basement, consistently higher values were obtained for thermal conductivity that range between 1.7 and 2.3 W/(m·K). The high scatter in thermal conductivity values in the deeper rocks may be related to varying concentrations of calcite and serpentine veins (see "Igneous and Metamorphic Petrology and Geochemistry" section, this chapter).