PHYSICAL PROPERTIES

Introduction and General Objectives

Shipboard physical properties determinations provide a first look at variations in core material characteristics and may be correlated with core lithology, downhole geophysical results, and regional seismic data. The principal objectives of the physical properties measurement program are closely connected to the main scientific and operational goals of Leg 183. They can be grouped together as follows:

  1. Providing comprehensive physical properties datasets, including porosity and acoustic variations.
  2. Integrating core physical properties data with physical parameters derived from downhole logging results. Bulk density, porosity, acoustic velocity, and thermal conductivity data can aid log interpretation.
  3. Providing cross-hole correlation. Magnetic susceptibility was measured on whole-round sections along the length of the recovered core to enable correlation of stratigraphic horizons from adjacent holes.
  4. Constraining the interpretation of seismic reflection and other geophysical data.

Additionally, physical properties data may help determine how the Kerguelen Plateau grew and its tectonic history, thereby contributing to the principal objectives of Leg 183.

All instruments/apparatus used in the shipboard laboratory and principles of methods are described in Blum (1997). Measurements were made on whole sections of cores using the MST. We determined thermal conductivity for sediments and basement rocks using the needle probe and half-space puck methods, respectively. When the sediment was sufficiently soft, we determined compressional-wave (P-wave) velocities (Vp) on the working half of the core. Using the insertion probe system (PWS1 and PWS2) and/or the contact probe system (PWS3). For consolidated or lithified sediments and for hard rocks, we cut samples and we measured velocity using the contact probe system. Whenever possible, we measured index properties in samples where velocity was also measured.

Sampling Strategy

The sampling program for physical properties was designed to fulfill the following two requirements:

  1. Whole-core sections were scanned with the MST before being split. We then selected physical properties samples from the split cores. Where recovery permitted, we chose one or two samples per section to represent the dominant lithology.
  2. Core specimens for analyses were selected in conjunction with sedimentologists and structural geologists, we selected additional samples to represent intervals of unusual lithology or structure.

Whole-Core Measurements

Multisensor Track

The MST included four physical properties sensors (magnetic susceptibility meter, gamma-ray attenuation porosity evaluator [GRAPE], P-wave logger [PWL], and natural-gamma radiation detector [NGR]). Individual, unsplit core sections were placed on the MST, which automatically moved the section through the sensors on a fiberglass track. MST data were sampled at discrete intervals, with the sampling rate chosen to optimize data resolution and the time limitations of running each core section through the device. GRAPE data, compressional wave velocity, and magnetic susceptibility data were all logged at an interval of 4 cm and acquisition times of 5 s. Core sections were run through the MST after they had warmed to at least 16C. GRAPE data are most reliable in undisturbed cores and offer the potential of direct correlation with downhole bulk density logs. For highly fragmented core, GRAPE acquisition was turned off. Many cores of (unconsolidated) soft sediments were disturbed to some degree. Where cores were not filling the lines, disturbed, or fractured, we expect the GRAPE density to have a general lower value.

The PWL operates simultaneously with the GRAPE and transmits a 500-kHz P-wave pulse (2-s wave period; 120 V) through the core. A pair of displacement transducers monitors the separation between the P-wave transducers. Data are collected at 3-cm intervals. The quality of the data was assessed by examining the arrival time and amplitude of the received pulse. Data with anomalously large travel times or low amplitudes were discarded. Because RCB cores do not fill liners, only selected cores from each site were run through the PWL.

Magnetic susceptibility was determined on all sections at 3- to 5-cm intervals using the 1.0 (1 s integration time) range on the Bartington meter (model MS2C), which has an 88-mm coil diameter. Magnetic susceptibility helps detect variations in magnetic properties caused by lithologic changes or alteration. The quality of these results is degraded in RCB sections if the core liner is not completely filled or the core is disturbed. However, general downhole trends may still be used for laboratory to well-log correlation. During Leg 183, magnetic susceptibility was also determined with a point-susceptibility meter on the AMST. We routinely compared data from the two instruments (see "Paleomagnetism" sections in the site chapters for discussion).

NGR emission was routinely recorded for all core sections, both to monitor variations in radioactive counts of sample rocks and to provide a correlation with the geophysical logging. The NGR system records radioactive decay of 40K, 232Th, and 238U, three long half-life isotopes. The total gamma-ray count is a function of the combined influence of these three isotopes. The installation and operating principles of the NGR system used during Leg 183 are discussed by Blum (1997).

The area of influence for the four NGR sensors was ~10 cm from the points of measurements along the core axis. As gamma-ray emission is random, count times have to be sufficiently large to average for short-period variations. This was achieved on the MST system by utilizing the long area of influence on the sensors and using a moving average window to smooth count-rate variations and to achieve a statistically valid sample.

The NGR system was calibrated in port against a thorium source. We generally made measurements every 16 cm and, in some basement rock, every 6 cm. Results were output in CPS (counts per second) units.

Thermal Conductivity

Thermal conductivity is the rate at which heat is transmitted by molecular conduction. Thermal conductivity is an intrinsic material property that depends on the chemical composition, porosity, density, structure, and fabric of the material. Thermal conductivity profiles of sediments and rock sections are mainly used, along with temperature measurements, to calculate heat flow. Heat flow is not only characteristic of the material but also helps to indicate age of ocean crust and fluid circulation processes at a range of depths (Blum, 1997). Whole-round core sections were allowed to equilibrate to room temperature for at least 2 hr in preparation for thermal conductivity measurements. We used the needle-probe method in full-space configuration for soft sediments and in half-space mode for lithified sediment and hard-rock samples. We typically acquired data in every soft-sediment core and every section in basement rocks. Data are reported in units of W/(mK). The mean error associated with these determinations is estimated as 0.2 W/(mK) in sediments. For technical details see Blum (1997).

Soft-Sediment Full-Space Determinations

We inserted a needle probe containing a heater wire and a calibrated thermistor into the sediment through a small hole drilled in the core liner before the sections were split. At the beginning of each determination, we monitored temperatures in the samples without applying current to the heating element to verify that temperature drift was <0.04C/min. We then turned on the heater and recorded the temperature rise in the probes. After heating for ~60 s, the needle probe behaves nearly as a line source with constant heat generation per unit length. We fit the temperatures recorded between 60 and 240 s to the following equation using the least-squares method (von Herzen and Maxwell, 1959):

T(t) = (q/4k) ln(t) + L(t), (2)

where k is the apparent thermal conductivity (W/[mK]), T is temperature (C), t is time in s, and q is the heat input per unit length of wire (W/m). The term L(t) corrects for temperature drift, described by the following equation:

L(t) = At + Te, (3)

where A represents the rate of temperature change, and Te is the equilibrium temperature. L(t) therefore corrects for the background temperature drift, systematic instrumental errors, probe response, and sample geometry. The best fit to the data determines the unknown terms k and A.

Lithified Sediment and Hard-Rock Half-Space Determinations

We made half-space determinations on selected lithified sediments and basaltic rock samples after the cores were split and their faces polished. The needle probe rested between the polished surface and a grooved epoxy block with relatively low conductivity (Sass et al., 1984; Vacquier, 1985). We conducted half-space measurements in a water bath to keep the samples saturated, to improve the thermal contact between the needle and the sample, and to reduce thermal drift. EG&G thermal joint compound was used to improve the thermal contact. Data collection and reduction procedures for half-space tests are similar to those for full-space tests except for a multiplicative constant in equation 3 that accounts for the different experimental geometry.

Discrete Measurements in Split-Core Samples

Index Properties

We extracted samples of ~10 cm3 from the fresh core for determination of index properties. We calculated bulk density, grain density, water content, porosity, and dry density from wet and dry sample and dry volumes. The error associated with mass determinations using two Scitech electronic balances is 0.05%. The balance was equipped with a computer averaging system that corrected for ship accelerations. The sample mass was counterbalanced by a known mass such that the mass differentials generally were <1 g. We determined sample volumes using a Quantachrome Penta-Pycnometer, a helium-displacement pycnometer with a nominal precision of 0.02 cm3, but with apparently a lower experimental precision of 0.04 cm3. We determined sample volumes at least three times, until readings were consistent. A standard reference volume was included with each group of samples during the measurements and rotated among the cells to check for instrument drift and systematic error. This exercise demonstrated that the measured volumes had a precision of ~0.02 cm3. We carefully calculated the sample beakers used for discrete determinations of index properties before the cruise. After the samples were oven dried at 105 5C for 24 hr and allowed to cool in a desiccator, we measured dry weight and volumes. The main problem with this drying temperature is that chemically bound water in clay minerals is largely lost in addition to interstitial water.

Water Content

The determination of water content as a fraction of total mass or as a ratio of water mass to solid mass followed the methods of the American Society for Testing and Materials (ASTM) designation (D) 2216 (ASTM, 1989). We measured total mass (Mt) and dry mass (Md) using the electronic balance and the difference was taken as the uncorrected water mass. Corrections for salt assumed a pore-water salinity (r) of 0.35%, following the discussion by Boyce (1976). The equations for the two water content (W) calculations are

Wd (percent dry mass) = [(Mt - Md)/(Md - rMt)] 100, (4)
Ww (percent wet mass) = {(Mt - Md)/[(1 - r) Mt]} 100, (5)

where Mt and Md are measured in grams.

Bulk Density

Bulk density (bulk) is the density of the total sample including the pore fluid (i.e., bulk = Mt/Vt), where Vt is the total sample volume (in cubic centimeters).

Grain Density

Grain density (g) was determined from the dry mass and dry volume measurements. Both mass and volume must be corrected for salt, leading to the following equation:

g = (Md - Ms)/[Vd - (Ms /s)], (6)

where Md is the dry mass (g) and s is the density of salt (2.257 g/cm3).

Ms = r Mw (g) is the mass of salt in the pore fluid, Mw (g) is the salt-corrected mass of the seawater:

Mw = (Mt - Md)/(1 - r). (7)

Porosity

Porosity () represents the ratio of pore-water volume to total volume. The following relationship was employed:

= [(g - bulk)/(g - w)] 100, (8)

where g represents the calculated grain density, bulk is the bulk density, and w is the density of seawater.

Dry Density

The dry density (d) is the ratio of the dry mass (Md) to the total volume (Vt). We calculated the dry density using the corrected water content (Wd) and porosity () as follows:

d = (/Wd) w. (9)

Velocity Determinations

For discrete velocity measurements in split cores, we used the insertion probe system (PWS1 and PWS2) and the contact probe system (PWS3), using a signal frequency of 500 kHz. In soft sediments, we employed the insertion probe system that determines the traveltime over a fixed interval in the y (across the core) and z (along the core) directions. For consolidated and lithified sediments and for hard rocks, only the PWS3 was used, measuring samples within the liner or in cut samples. When lowering the transducer to the core surface, an indeterminate pressure is applied. For all measurements, we lowered the transducer to the point where it just reached the surface. We then applied an additional force of 20 psi, sufficient for achieving good contact yet nondestructive on soft sediments. Measurements of ultrasonic velocities are known to be relatively more stress sensitive in the low-pressure regime. The system was calibrated with Plexiglas standards, corresponding to velocities being found in sediments. The system was not calibrated with standards representing high velocities, and this may account for velocities >6000 m/s in some hard rocks. If the high velocities are a consequence of the calibration method, it would result in an increasing error as the velocities increase, but the general trend would be valid. At Site 1137, the discrete velocity measurements agree well with sonic log data, and all velocities are <6000 m/s. At later sites, and where we encounter higher velocities, we did not have sonic log data for comparison. Determining velocity over a short distance, as in some cut samples, can also increase the uncertainty of the measurements. We measured compressional wave velocity (500 kHz) once or twice per section and in more than one direction of the core where possible. Distilled water constituted the coupling fluid at the transducer/core interface. Sediments were measured in half liners or in discrete samples taken from the core. For disturbed or biscuited sediments, we employed a needle-probing method to identify harder and preserved biscuits for measurement. We sampled crystalline and basaltic rocks as minicores (2.54 cm in diameter) drilled perpendicular to the axis of the core (x-direction) or sawed as oriented cubes. The ends of the minicores were trimmed parallel with a rock saw, and we measured traveltimes and distances along the axis of the minicore. Velocities for cubes were determined in two or three mutually perpendicular directions, Vz (along the core), Vx (into the split core, perpendicular to core axis), and Vy (across the split core). Velocity anisotropy follows the relationship

Anisotropy = 3(Vmax - Vmin)/(Vx + Vy + Vz), (10)

where Vmax and Vmin are the maximum and minimum velocities (among Vx, Vy, and Vz). During Leg 183, we made measurements for velocity determinations either adjacent to paleomagnetic minicores or directly on the minicores to save core material for shipboard physical properties samples. In intervals where discrete sampling was sparse, and if time allowed, we made additional measurements on split pieces of hard rocks. Velocities from these measurements may have slightly lower values because of a more irregular contact surface. Nevertheless, velocities determined from hard-rock pieces should exhibit trends similar to those from discrete sample measurements.

We determined the dimensions of samples analyzed in the contact probe system with digital calipers. During Leg 183, we noted that the display of the digital caliper needed to be monitored regularly or erroneous readings could occur. It is imperative to keep the caliper's track clean. We estimated traveltime by identifying the first break of the stacked waveforms, and we corrected traveltime for system delays. Using the corrected traveltime and path length, we calculated velocities. Velocity data are reported here in raw form.

NEXT