Previous   |   Next

doi:10.2204/iodp.proc.301.106.2005

Physical properties

Physical property measurements provide indications of lithology, texture, and degree of alteration. Shipboard measurements of physical properties are used to help correlate data from downhole logging with those from cored sections to provide additional insight on the physical state of the section. Data also provide context that aids in the interpretation of other information such as microbiological and biogeochemical results. The sediment section at Site U1301 was cored using the APC in Holes U1301C and U1301D. The basement section was cored using the RCB in Hole U1301B. The following sections describe the results of the shipboard analyses for both the sediment and basement sections.

Hole U1301C

Nineteen APC sediment cores were obtained over a depth interval of 0–265.3 mbsf in Hole U1301C, with the majority (16) taken in lithologic Unit I (0–13.1 mbsf) and the remainder taken from Unit II (Fig. F58A). The section was not continuously cored because of time and operational constraints and because an essentially identical section was cored at Site 1026 during Leg 168. At the end of the expedition, we returned to Site U1301 to core the interval 120–177 mbsf in Hole U1301D. However, no physical property measurements were made from this last set of cores because of the short transit time at the end of the expedition. Physical property measurements were made on whole cores (multisensor track and thermal conductivity), split cores (P-wave velocity and shear strength), and discrete samples (moisture and density [MAD] properties); data are presented in Table T18.

In an effort to assess potential sampling bias, each sample was classified as clay, sand, or mixed lithology. This nomenclature may not correspond precisely to the lithologic description of the section (see "Lithostratigraphy") but is correct at the scale of physical property samples. Table T19 shows the percentage of sand in the total thickness of the described section (excluding whole-round samples collected from the catwalk) and also shows the percentage of measurements that were made in sand layers. This demonstrates that measurements of thermal conductivity, P-wave velocity, MAD properties (bulk density, grain density, and porosity), and shear strength are all biased toward an oversampling of clay layers, suggesting that a weighted averaging scheme should be employed when characterizing overall conditions throughout the cored interval. However, it should also be noted that recovery was incomplete in many cores, cores containing sand are often highly disturbed, we did not core the sections continuously, and many whole-round sections were removed for microbiology. Thus, even if we had made physical property measurements in each recovered lithology according to its fractional representation in the cores, there would still be an unquantifiable bias in the data set.

Multisensor track

Magnetic susceptibility

Magnetic susceptibility ranged from 44 × 10^–6 to 3460 × 10^–6 SI units (Fig. F58B), with large spikes in the data set corresponding to sand layers of turbidite sequences and lower values corresponding to silt and clay layers. The highest values were measured in a series of interbedded turbidites recovered in Cores 301-U1301C-6H and 7H over an interval of 12 m beginning at 45 mbsf and likely result from the presence of magnetite grains (see "Lithostratigraphy"). Figure F59 displays the variability within a typical 1.5 m section of interbedded sand and clay recovered from Subunit IB (Section 301-U1301C-3H-5), along with colocated magnetic susceptibility and bulk density data. Magnetic susceptibility is initially low in the uppermost clay layer and increases with depth through a ~50 cm thick turbidite layer. Oscillations in the record correspond to thin interbedded layers of sand and clay. The variations in the magnetic susceptibility record are tracked nicely by those in the bulk density record, where low- and high-density layers correspond to clay and sand, respectively. Similar profiles of magnetic susceptibility and bulk density are seen deeper in Subunit IB (Fig. F60) over a longer, 7 m thick portion of Cores 301-U1301C-8H and 9H. Magnetic susceptibility is initially near 1000 × 10^–6 SI in the sand layer near the top of the core, before dropping and oscillating through a 1.5 m thick interbedded sequence, and attaining a constant low value of ~250 × 10^–6 SI throughout the underlying clay layer. As with Section 301-U1301C-3H-5, the bulk density record displays a similar pattern. Magnetic susceptibility decreases both in magnitude and variability across the transition from Unit I to Unit II (dropping from 560 × 10^–6 ± 318 × 10^–6 SI to 280 × 10^–6 ± 95 × 10^–6 SI), along with the shift from sand (containing magnetite) to hemipelagic clay lithology.

Natural gamma radiation

Although pore volume may have an effect on the natural gamma radiation (NGR) signal where emissions from NGR-producing elements are low, the porosity in Hole U1301C varied by less than a factor of two (~50%–80%), whereas the NGR data varied by one order of magnitude (5–50 cps) (Fig. F58E). In general, NGR counts were higher in clay than in sand, resulting from the higher content of radioactive isotopes in the clay. The highest values were obtained from a 5 m thick clay layer in Cores 301-U1301C-9H and 10H (Figs. F58E, F60). The resolution of the NGR data is lower than that of the other MST instruments (measurements every 15 cm, as opposed to 2.5 cm), so the pattern does not show the same degree of variability in the thinly interbedded sand and clay zones (Fig. F59).

Gamma ray attenuation density

The bulk density of the core is estimated by gamma ray attenuation (GRA). Results vary by as much as 15% between GRA and MAD property samples within the highly porous clay layers at depths <100 mbsf, whereas the difference is smaller for less porous layers deeper within the section (Fig. F58C). In general, the data show a systematic increase in bulk density with depth from ~1.5 to 2 g/cm³ in the upper 75 m of the section, below which they scatter around a mean value of ~2 g/cm³ (Fig. F58C).

P-wave logger

The P-wave logger (PWL) showed anomalous results and ceased to operate beginning with Core 301-U1301C-4H and was not repaired in time for use on any further sections. The limited data obtained over the interval of 0–24 mbsf are loosely consistent with velocity data obtained using the P-wave sensor (PWS) system on split cores (Fig. F58F).

Moisture and density properties

A total of 78 discrete 10 cm³ samples were collected immediately after splitting of cores for the determination of MAD properties and were classified as being predominantly either clay, sand, or mixed lithology.

Bulk density of the clay layers increases systematically from 1.4 g/cm³ at the seafloor to ~2 g/cm³ at a depth of 100 mbsf and correlates with a ~30% decrease in porosity over the same depth interval (Fig. F58C, F58I). The porosity of sand layers remains relatively constant at ~40% to a depth of 115 mbsf. Analysis of trends at greater depth is precluded by the discontinuous core. Bulk density values from clay lithologies recovered in Cores 301-U1301C-15H through 19H vary slightly about a mean of 1.9 g/cm³. The bulk density of sand layers is relatively consistent at 2.0 ± 0.1 g/cm³, with the exception of a highly variable zone located within Subunit IB at ~79–82 mbsf. Grain density is remarkably consistent at 2.8 ± 0.1 g/cm³ regardless of depth or lithology. The higher than expected grain density could be attributable to pyrite, which has a grain density of ~5 g/cm³.

Thermal conductivity

Thermal conductivity data were collected at a frequency of at least one measurement per section, with additional data acquired in some turbidite layers and in sections bracketing regions where downhole temperature tools (APCT and DVTP) were used. A total of 238 discrete measurements were made at 88 different locations on whole cores using the full-space needle probe configuration. Each sample location was classified as being either clay, sand, or mixed. In order to estimate the degree of uncertainty in any single value, multiple measurements were made at each sampling location. Thermal conductivity was strongly controlled by lithology, with values for clay being significantly lower than values for sand, averaging 1.12 ± 0.12 and 1.53 ± 0.19 W/m·K, respectively. Values of mixed lithology span the range in between (Fig. F61A; Table T18). A systematic increase of thermal conductivity is apparent in the upper 100 mbsf, where thermal conductivity increases from ~0.9 to ~1.2 W/m·K, whereas the porosity of clay layers decreases concurrently from >80% to ~50%. The observed relationship between thermal conductivity and porosity suggests that thermal conductivity varies consistently with porosity only within the clay layers (Fig. F61B).

We calculate a range of potential matrix thermal conductivities by employing a geometric mean mixing model (e.g., Woodside and Messmer, 1961) and estimate matrix thermal conductivities ranging from 1.3 to 6.0 W/m·K. A matrix thermal conductivity of ~2.5 W/m·K was calculated for clay lithologies, which is ~1 W/m·K less than those estimated by Shipboard Scientific Party (1997) for Hole 1026A located <2 km north northeast along the same buried basement ridge as Hole U1301C.

Thermal conductivity drops at ~100 mbsf from ~1.25 to 1.05 W/m·K (Fig. F58G), concurrent with a lithologic change from medium–coarse massive sand to gravel, consistent with a similar decrease seen at Sites 1026 and 1027. Below 200 mbsf, thermal conductivity varies little with depth within the hemipelagic clay recovered in Cores 301-U1301C-17H, 18H, and 19H, having a mean value of 1.16 W/m·K.

P-wave velocity

A total of 107 measurements of P-wave velocity were made on split sections using the Hamilton Frame PWS3 contact probe system. Cracking of lithified sediments precluded the acquisition of reliable velocities by the z- and y-axis transducers in sediments from deeper than ~35 mbsf, but x-axis velocities (perpendicular to the cut face) were successfully measured using a noninvasive transducer pair. The digitally recorded distance between the PWS3 transducers was found to be in error, and postprocessing corrections were applied to all x-axis velocity data by simply correcting for the actual sample thickness and recalculating the correct velocity as

where

v = the correct velocity;

t_meas = the measured traveltime;

d_meas = the original recorded transducer separation,

m = the slope of the regression line defining the correct transducer separation;

b = the intercept of the regression line defining the correct transducer separation;

x_liner = the thickness of the core liner;

v_liner = the seismic velocity of the core liner; and

t_delay = the instrument electronic delay (Fig. F62A; see caption for variable values and equation reference).

Data were classified by lithology (clay, sand, or mixed), and are shown in Figure F58F.

Velocity values range from ~1480 to 1780 m/s over the 265 m drilled interval, with an increase of ~10% occurring within the uppermost 50 mbsf (Fig. F62B). Velocities measured in clay lithologies consistently display a lesser degree of variability than those measured in sand and also display a systematic increase with depth. Mobilization and redeposition of sand layers in turbidite sequences during coring precludes determining a statistically significant trend with depth for sandy units. We evaluate the degree of anisotropy by comparing horizontal and vertical velocities where the vertical spacing between measurements is <5 m and find no significant directional preference (Fig. F62C).

Undrained shear strength

A total of 46 shear strength measurements were made on split cores immediately after splitting using either a handheld Torvane device or a pocket penetrometer (Fig. F58H). The strength of the sediment determined which instrument was used. Measured ranges were 0–184 kPa for the Torvane and 86–221 kPa for the penetrometer. Three different adapters are available for the Torvane, and the strength of the sediment determined which adapter was used. Measurements needed to be converted to account for the size of the adapter and to change the units from kilograms per square centimeter into kilopascals (1.96 for the large adapter, 9.81 for the medium adapter, and 24.52 for the small adapter). In addition to the unit conversion, penetrometer readings were divided by 2 because the penetrometer is calibrated as an unconfined compression test equal to twice the undrained shear strength.

Most measurements were carried out on clayey material, and only a few measurements were made on sand. Values measured on sand are consistently lower than values measured on clayey material, probably because they are less consolidated and sand has been resuspended. Values of shear strength range from 1 to 221 kPa. In the upper part of the section (above 120 mbsf), shear strength increases with depth, most likely because of consolidation of the sediment caused by the increase of the effective stress. Consolidation in the upper part of the hole is confirmed by MAD measurements, which show an increase in bulk density and decrease in porosity with depth.

Hole U1301B

Physical property measurements in the basement section (Hole U1301B) included magnetic susceptibility measured on every section using the MST and discrete measurements of thermal conductivity, P-wave velocity, and MAD properties (bulk density, dry density, grain density, wet density, wet weight, dry weight, porosity, and void ratio) (Fig. F63; Table T20). Sampling frequency for the discrete samples was generally one per section, with extra samples collected where there were visibly significant changes in lithology, alteration, or texture of the core.

Magnetic susceptibility

Whole sections were run through the MST after the core liner was split, with sampling resolution set at 1 cm regardless of section continuity. Magnetic susceptibility was as great as ~4000 × 10^–6 SI, with the highest values corresponding to massive lava flows recovered in Cores 301-U1301B-13R, 15R, and 18R, which were located at ~430, 445, and 472 mbsf, respectively (Fig. F63A, F63B). Other lithologies (pillow lava and hyaloclastite) generally yield values <2000 × 10^–6 SI, partly due to highly fractured layers and poor core recovery leading to large discontinuous sections, both of which act to reduce the magnetic susceptibility. No effort was made to correct for incomplete filling of the core liner or for discontinuous rock that is characteristic of this formation or to correct for voids due to the removal of whole-round sections for microbiological analysis (Fig. F64).

Thermal conductivity

A total of 365 individual measurements were made on 68 basalt samples over the interval of 351.2 to 576.3 mbsf (Fig. F63C; Table T20), with samples being chosen based on size, continuity, and absence of fractures or veins. A minimum length of 6 cm was required to achieve full contact between the sample and the thermal conductivity probe. Thermal conductivity measured on samples with lengths ≤6 cm tended to be slightly lower than the overall average, suggesting a bias due to small sample size. These values are plotted as open squares on Figure F63C. By making three to six individual measurements on each sample, we are able to evaluate the reliability of each point measurement by including error bars of 1σ.

Thermal conductivity values ranged from 1.17 to 1.84 W/m·K (average = 1.70 ± 0.10 W/m·K) over the depth range of 351.2–576.3 mbsf. Measurements were made on samples having varying lithology, porosity, and degrees of alteration; there is no statistically significant change in thermal conductivity with depth. Values >1.75 W/m·K consistently came from large massive samples (>6 cm in length), recovered in either massive flows or pillow basalts. The lowest values of 1.17 and 1.37 W/m·K correspond to the two hyaloclastite samples from Cores 301-U1301B-1R and 35R at depths of 351.2 and 564.7 mbsf, respectively. This suggests that recovery and sampling biases toward unfractured basalt skew the data toward higher values and likely provide an upper bound on the effective thermal conductivity of uppermost basement in this region.

P-wave velocity

A total of 106 discrete samples were initially collected from oriented cores for measurement of P-wave velocity, with roughly half being shared with the paleomagnetism group. Unlike samples collected for MAD property analysis, those collected for paleomagnetic investigation were not carefully cut and polished to ensure that the cut faces were flat and parallel. More than 95% of the samples were cut as 2 cm × 2 cm × 2 cm minicubes, allowing two horizontal velocities to be measured in addition to the vertical velocity. Only where lithology or structure precluded cutting of a minicube were round minicores collected, allowing P-wave measurement in one direction.

Following initial processing, the PWS3 transducer pair was found to be out of calibration. Samples shared with the paleomagnetism group that had not been heated were remeasured, along with those collected initially for P-wave determination only, reducing the overall sample number from 166 to 106 (Janus database run numbers 3512–3803). Compressional velocities were measured on 94 samples in the x-direction (perpendicular to the half-cutting surface), 92 samples in the y-direction (parallel to the half-cutting surface), and 105 samples in the z-direction (vertical). The assignment of x- and y-directions is arbitrary because basalt cores cannot be oriented in azimuth.

Velocity values ranged from a minimum of 3940 m/s to a maximum of 5750 m/s (Fig. F63D) (average = 5130 ± 280 m/s). This average value is more than regional values determined using seismic reflection techniques (Rohr, 1994) and is consistent with shipboard sample values from Leg 168 (Shipboard Scientific Party, 1997). Velocities measured on samples used for paleomagnetism analysis tend to display a slightly higher degree of variability (due to the cutting/polishing differences stated previously), but mean values are generally consistent with those samples prepared for MAD property measurements. The lowest velocity was measured on a highly vesicular sample recovered from Core 301-U1301B-18R, located within a massive flow unit at a depth of ~475 mbsf. Additional samples recovered from the same lithologic unit include velocities as great as ~5500 m/s, demonstrating a high degree of heterogeneity. There is no statistically significant overall velocity trend with depth, although it appears that P-wave velocity may be locally reduced in areas of higher porosity or greater alteration and fracturing (discussed below). A comparison of horizontal velocity with vertical velocity for paleomagnetic and MAD property samples displays no anisotropy (Fig. F65).

MAD properties

A total of 83 discrete samples were collected from Hole U1301B for the determination of MAD properties (Fig. F63E, F63F, F63G). More than 95% of the samples were cut as minicubes and were also used for velocity measurements, with the remainder consisting of nonoriented minicores or rock fragments with a minimum size of ~5 cm³.

Bulk density spanned the range 1.86–3.03 g/cm³ (average = 2.75 ± 0.13 g/cm³) (Fig. F63E). Grain density exhibited a similar distribution of 2.23–3.11 g/cm³ (average = 2.86 ± 0.09 g/cm³) (Fig. F63F). The lowest values of both grain and bulk density were made in a highly brecciated hyaloclastite sample recovered from Core 301-U1301B-1R at 351.2 mbsf, whereas the highest density comes from the top of Core 14R at the boundary between a massive and pillow flow at ~432 mbsf. Low values were also measured in a hyaloclastite flow recovered in Core 301-U1301B-35R from ~562 mbsf. Porosity values span the range 1.9%–30.3% (average = 5.8% ± 3.5%) (Fig. F63G). The strong heterogeneity of fractured crystalline rock is displayed by a single sequence recovered in Core 301-U1301B-18R at ~472 mbsf, where density and porosity are as variable over the 9 m thick massive section as they are over the entire 225 m cored interval from Cores 301-U1301B-2R through 36R.

Porosity values are inversely correlated with grain density, with minimum porosity values generally measured where grain densities were highest (Fig. F66A). Additionally, grain density variability decreases with decreasing porosity, possibly resulting from infilling of vesicles, fractures, and microcracks with precipitates and alteration minerals, as a result of basement alteration. Similar correlations between porosity and grain density were shown by Busch et al. (1992), Broglia and Moos (1988), and Carlson and Herrick (1990).

Similarly, seismic velocity and porosity are inversely correlated (Fig. F66B), which may be due in part to formation porosity but may also be an artifact of sample cracking and other core damage. The effect of decreasing grain density, which may be inferred to be a proxy for increasing basement alteration, on seismic velocity is shown in Figure F66C. Seismic velocity displays a loose positive correlation with grain density, with the majority of velocities decreasing from >5500 to ~4750 m/s while grain density decreases from ~3.0 to ~2.7 g/cm³.

Top of page   |   Previous   |   Next