A major objective of the testing program on the core sample from ODP Site 897 was to deduce or at least constrain the in situ stress conditions at the sample depth, including both the effective vertical and horizontal stresses. Because of the very low apparent yield stress obtained from initial testing, which could reflect either a very high pore-fluid pressure or anomalous sediment behavior, a reference consolidation test was run on mechanically disaggregated remnants of the core. Finally, a reconsolidation-consolidation test with multiple unload-reload cycles was run on an undisturbed sample to explore the effects of the initial test state on the yield stress. Compressional seismic velocities were measured during all tests except the consolidation of the disaggregated material.
Most of the consolidation tests were performed in a large triaxial cell driven by a computer-controlled servo-hydraulic load frame. All but one of these tests were run with uniaxial strain using a pseudoderivative algorithm that adjusted the confining pressure and axial load to keep the sample cross section constant. The cross-sectional area was monitored with an array of eight linear variable displacement transducers (LVDTs) and was maintained constant to within 0.003%. Axial strains were determined from two LVDTs fitted to rings fastened so as to measure displacement over the middle half of the sample (Fig. 4).
The vertical stresses during these tests were increased or decreased at a constant rates near 10-3 MPa/min, which generated strain rates on the order of 10-8/s. These low strain rates, together with drainage from both sample ends, ensured dissipation of excess pore pressure and, in comparison to standard geotechnical consolidation rates, included a significant portion of the secondary consolidation.
The consolidation of the disaggregated material was performed in a large (3-in. diameter) consolidation cell, driven by the same computer-controlled load frame (Karig and Hou, 1992). This consolidation was also performed at a constant stress rate except at stresses less than 1 MPa. The sample was prepared as a slurry and loaded in the cell to a depth such that, when consolidated to a porosity of 25%, it would be just thick enough to cover the lateral stress sensor in the cell wall. In addition to making the sample as short as possible to reduce sidewall friction, the cell wall was coated with lubricant (Karig and Hou, 1992).
The sediment used for testing was a burrow-mottled calcareous mudstone from Section 149-897D-3R-2, 73-103 cm, representing a depth from 618.7 to 619 m below seafloor (mbsf). This mudstone lies near the base of stratigraphic Subunit IIC, of probable middle Eocene age, and has an estimated 45% carbonate content (Shipboard Scientific Party, 1994a). Smear slide analysis indicates that most of this carbonate is in the form of nannofossils, primarily coccoliths, and that there is a very low ratio of silt to clay size fraction. Although no detailed clay mineralogy was available, the shipboard analyses and the plastic behavior of the sample during handling indicate a fair fraction of smectitic clay. This type mudstone is a very common deep sea lithology and it is useful to compare its mechanical behavior with that of a nearly noncalcareous silty clay from the Nankai Trough, which was previously tested (Karig, 1993; Feeser et al., 1993).
Bulk densities of these subsamples, measured in our laboratory from sample volumes and wet weight, varied between 2.18 to 2.0 g/cm3, with a strong mode near 2.195 g/cm3. The bulk density of adjacent sections of core from shipboard measurements was 2.21 g/cm3 (Shipboard Scientific Party, 1994a). The corresponding porosity was given as 36.2%, but calculated from the given grain density of 2.78 g/cm3, it would be 32.6%. With that grain density, the porosity of our subsamples is 33.4%.
Because this sediment appeared to be much more plastic than previously studied Nankai clays, the plasticity index (PI) was determined by measuring water contents at the plastic limit (PL) and liquid limit (LL): the PI is the difference in water content between that at the LL and the PL. For this sediment, LL = 42.7%, and PL = 19.2%, resulting in a PI of 23.5%, which indicates moderate plasticity for an inorganic clay (e.g., Holtz and Kovacs, 1981). The water content of the core sample before testing was 18.4%, indicating that the material was naturally quite close to its PL. This relatively high water content and high PI can account for the more plastic behavior during handling of this sample than for the previously tested silty clays.
The core section was subsampled with a water-lubricated rotary coring tube that cut cylinders 20 mm in diameter. For this test series, the cylinders were all cut with vertical axes and trimmed to lengths near 55 mm. For the tests, the cylindrical subsamples were placed in latex jackets and fitted with LVDT arrays.