Previous   |    Next

doi:10.2204/iodp.proc.345.109.2014

Physical properties

Physical properties of gabbroic rock recovered in Hole U1415I were characterized through a series of measurements on whole-core sections, half-core sections, half-core pieces, and discrete samples as described in “Physical properties” in the “Methods” chapter (Gillis et al., 2014e). We measured gamma ray attenuation (GRA) density and magnetic susceptibility on the Whole-Round Multisensor Logger (WRMSL); natural gamma radiation (NGR) on the Natural Gamma Ray Logger (NGRL); point magnetic susceptibility, reflectance spectrophotometry, and colorimetry on the Section Half Multisensor Logger (SHMSL); and thermal conductivity, compressional wave velocity, density, and porosity on discrete samples. The rock names reported in data tables correspond to the primary lithologies assigned by the igneous group. Data are summarized as a function of depth in Figure F31.

Raw GRA density, magnetic susceptibility, reflectance spectrophotometry, and colorimetry data were uploaded to the Laboratory Information Management System database and subsequently filtered following the procedures described in “Physical properties” in the “Methods” chapter (Gillis et al., 2014e) to remove spurious points that correspond to empty intervals in the liner, broken pieces, and pieces that were too small. Both raw and filtered data are provided in PHYSPROP in “Supplementary material.”

Multisensor core logger data

Natural gamma radiation

Only two sections (345-U1415I-2R-1 and 4R-1) were measured on the NGRL; other sections contained pieces too small to provide reliable data with this instrument. NGR is overall very low (0–0.4 cps) in this core, significantly lower than background level (~5 cps).

Gamma ray attenuation density

Two cores (345-U1415I-2R and 4R) were measured on the WRMSL. GRA density measurements are volume dependent and range between ~2.2 and 2.6 g/cm³, which is ~0.4 g/cm³ lower than bulk density measured on discrete samples in the same cores (Fig. F31).

Magnetic susceptibility

Magnetic susceptibility was measured on both the WRMSL and SHMSL. The whole-round core measurements are volume measurements that give an average apparent susceptibility value over an 8 cm long interval, whereas the SHMSL values are given by point measurements. When measured on whole-round cores, magnetic susceptibility is generally underestimated, with values significantly lower than the point magnetic susceptibility (Fig. F31). The mean magnetic susceptibility of rock recovered in Hole U1415I is very low (311 × 10^–5 ± 544 × 10^–5 SI for point magnetic susceptibility; maximum recorded is 3600 × 10^–5 SI), reflecting the absence of magmatic Fe-Ti oxides. The highest measured values correspond to intervals with a higher abundance of magnetite associated with serpentinized olivine (Section 345-U1415I-4R-1; ~27.7 mbsf).

Reflectance spectrophotometry and colorimetry

Reflectance spectrophotometry and colorimetry data, together with point magnetic susceptibility data, were systematically acquired using the SHMSL, with a step size of 2 cm. No significant variation is evident in the recorded reflectance and chromaticity parameters (L*, a*, and b*) along the cores recovered in Hole U1415I. L*, a*, and b* mean values are 44.8 ± 6.6, 1.19 ± 0.75, and –6.73 ± 2.39, respectively.

In an attempt to better detect variations related to mineral composition of the recovered rocks, we conducted a series of measurements with a measurement interval of 1 mm on Pieces 8, 9, and 10 of Section 345-U1415I-4R-1, which is layered gabbronorite with significant variations in olivine content. The values returned by the instrument were substantially different from those measured previously with a 2 cm measurement interval. We then made several series of measurements with varying intervals (1 and 5 mm and 5 cm), the results of which are presented in Figure F32.

The instrument was calibrated immediately before the first run with 1 mm measurement intervals. The a* and b* values drifted away from the first measurements (note the different slope for the first 1 mm interval curves in Fig. F32A, F32B), and the results obtained for each measurement series were different (note the offsets between the first 1 and 5 mm and 1 cm curves in Fig. F32A, F32B). In contrast, the L* values appeared to be stable over the different series of measurements (Fig. F32C). After running a series of tests with color standards, we established that the measured chromaticity values a* and b* oscillated over time in an unpredictable manner (see “Physical properties” in the “Methods” chapter [Gillis et al., 2014e]). This behavior appeared to be related to the sensitivity of the spectrophotometer to temperature changes. After this problem was fixed by the technical staff, we remeasured the same core pieces with measurement intervals of 1 mm and 1 cm. Comparison of these more reliable data with the core (interval 345-U1415I-4R-1, 45–125 cm; Fig. F32) shows that there is no obvious systematic variation of the reflectance parameters with the olivine content of the rock. This may be related to the relatively coarse grains (as large as several millimeters in interval 345-U1415I-4R-1, 45–125 cm) or the sensor (7 mm diameter) measuring individual or a few grains rather than a representative aggregate.

Discrete sample measurements

Moisture and density

Bulk density, grain density, and porosity were calculated from wet masses, dry masses, and volume following the procedure described in “Physical properties” in the “Methods” chapter (Gillis et al., 2014e) measured on three cubic (2 cm × 2 cm × 2 cm) samples taken from the working halves of Sections 345-U1415-2R-1 and 4R-1 (Table T7; Fig. F31). Average bulk density and grain density are 2.88 ± 0.06 and 2.90 ± 0.06 g/cm³, respectively, and are similar to densities measured at Hess Deep (Ocean Drilling Program [ODP] Leg 147 Site 894) (Fig. F33). Porosity is low, ~1% on average.

P-wave velocity

The same three cubic samples used for moisture and density analyses were measured for P-wave velocities (V_P) along the three principal directions (x, y, and z) in the core reference frame (see Fig. F2 in the “Methods” chapter [Gillis et al., 2014e]). Results are listed in Table T7 and plotted in Figure F31. Average V_P is 6.32 ± 0.22 km/s, and the apparent anisotropy varies from 2% to 4.3%. We performed two successive series of measurements (≥10 in each direction) for each of these three samples, and the results varied by as much as 0.22 km/s (~3.4%) in a given direction from one series of measurements to another. The related standard deviation of the measured V_P is as much as 1.74%, and the fastest direction changed from y to x for Sample 345-U1415I-4R-1W, 50–52 cm (Piece 8A). Measured apparent anisotropies are low and should therefore be treated with caution; values on the order of 2% are unlikely to be significant. For the remainder of the expedition, only one series of measurements was performed on each sample. However, the measured velocities are probably not as accurate as suggested by the low standard deviations (typically <0.1%) for each individual measurement series. We consider that the error associated with our measurements is on the order of 2%.

Results are compared in Figure F33 with V_P and grain density measurements made during previous ODP legs and IODP expeditions on gabbroic samples from fast-spreading and slow-spreading oceanic crust. V_P values are consistent with measurements made at Hess Deep (Site 894). Figure F33 shows a large overall dispersion of V_P values. For example, the average value for ODP Leg 118 Hole 735B (Southwest Indian Ridge) is ~1 km/s higher than that for IODP Expedition 304/305 Site U1309 (Mid-Atlantic Ridge). We interpret this dispersion as being primarily caused by differences in the measurement protocols and in the performance of the measuring devices used during these different legs/expeditions. No obvious geologically meaningful explanation can be discerned for such large differences in V_P from one site to another with very similar lithologies, porosities, and degrees of alteration. This illustrates the intrinsic limitation of the quality of V_P measurements made on board, and any interpretation using these data should be treated with caution. As expected, the measured V_P at room pressure depends primarily on porosity, which is a reasonable proxy for sample groundmass petrophysical variability (e.g., background alteration) as we avoided taking discrete samples with metamorphic or alteration veins, rather than on lithology.

Thermal conductivity

Thermal conductivities were measured in three gabbroic rock samples taken at irregularly spaced intervals in Hole U1415I (Table T8; Fig. F31). Measured values range from 2.16 to 2.51 W/(m·K) and are averages of 10–20 measurements for each piece, with a standard deviation of <1.2%. We attempted to measure anisotropy in a foliated orthopyroxene-bearing olivine gabbro (Section 345-U1415I-4R-1A [Piece 9]) by using the shorter probe (see “Physical properties” in the “Methods” chapter (Gillis et al., 2014e), collecting two series of measurements with the probe needle aligned parallel and perpendicular to the foliation. The obtained apparent anisotropy is very weak (0.3%), lower than the measurement error; therefore, the anisotropy results are not meaningful. The small probe tends to return less stable values than the large probe, which makes the exercise of estimating the anisotropy of thermal conductivity difficult in this type of lithology.

Top of page   |    Previous   |    Next