Previous   |    Next

doi:10.2204/iodp.proc.343343T.201.2015

Methodology

At The Pennsylvania State University (PSU), whole-round core samples are kept in sealed aluminum vacuum bags with wet sponges to minimize desiccation and stored in a refrigerator until subsampling and trimming. Once subsampling was complete, the remaining material was resealed in plastic core liners using plastic caps and electrical tape. During sample preparation, samples of whole-round Sections 343-C0019E-5R-1 and 343-C0019E-20R-1 were trimmed to form cylinders 25.4 mm in diameter, with lengths ranging from 15.7 to 28.3 mm, using fine point carving tools. These samples were aligned parallel to the whole-round core axis and collected from the same interval in each core, 697.18 mbsf for the prism sample (Section 5R-1) and 831.45 mbsf for the underthrust sample (Section 20R-1). All uniaxial experiments were conducted at room temperature conditions (24.0°C) in an oedometer system (Fig. F2). The triaxial experiments were conducted in a triaxial apparatus and maintained at a constant temperature of 29.5°C using a temperature chamber that surrounds the apparatus (Fig. F3). Syringe pumps controlled the pore pressure for all experiments, and deionized water was used to fill the pore fluid lines. A syringe pump also controlled both the confining and axial stress in the triaxial experiments, and silicone oil was used as the confining medium.

Uniaxial constant rate of strain tests

We conducted two uniaxial CRS tests in the Rock and Sediment Mechanics Laboratory at PSU. In the uniaxial CRS tests, the sample was placed within a steel ring and positioned in an oedometer system capable of reaching 50 kN axial load (Fig. F2). Each sample was equilibrated for 24 h at 300 kPa backpressure to ensure fluid saturation and that any air present in the pore pressure lines was completely dissolved. After backpressuring, a progressively increasing axial load was applied to the sample by specifying a constant displacement rate of 3.3 × 10^–5 to 5.0 × 10^–5 mm/s using a computer-controlled load frame. In this configuration, pore pressure at the top of the sample is kept at a constant 300 kPa during the consolidation stage. The base of the sample was undrained, and basal pore pressure was monitored throughout consolidation (ASTM International, 2006). Axial displacement and axial load were measured externally, and the measured displacement was corrected to account for the compliance of the oedometer system. Total axial stress (σ_A), basal pore fluid pressure (P_f), backpressure, and axial displacement were recorded every 10 s throughout the experiment.

Differential fluid pressure across the sample length during continuous loading induced fluid flow through the sample, and thus, the hydraulic conductivity and permeability of the sample could be determined from the known strain rate () and excess basal pore pressure (Δu), which is defined as the difference between the measured basal pore pressure and the backpressure (ASTM International, 2006; Long et al., 2008). Because the increase in basal pore pressure during loading depended on the rate of displacement, sample length, and hydraulic conductivity of the sample, we chose an appropriate constant displacement rate (or strain rate) for each sample such that Δu was sufficiently large to be measurable with minimal error but remained <12% of the effective vertical stress (e.g., ASTM International, 2006). Hydraulic conductivity (K) is given by

K = (HnH0ρwg)/2Δu,

(1)

where

= strain rate,
• ρ_w = density of water,
• g = gravitational acceleration,
• H_n = sample height at the time of interest,
• H₀ = initial sample height, and
• Δu = excess basal pore pressure.

Permeability (k) is then given by

where µ is the fluid viscosity at room temperature (0.001 Pa·s).

To allow for direct comparison of both the uniaxial CRS and triaxial tests, the effective mean stress (σ_m′) is reported for all of our experiments. To estimate the effective mean stress for the uniaxial CRS tests, in which a zero lateral strain boundary condition was imposed but lateral stresses were not measured, the coefficient of the horizontal stress at rest (denoted as K₀) is used; K₀ represents the ratio of horizontal to axial effective stresses. We assume a value of K₀ = 0.6, which is typical for clay-rich sediment (Karig, 1990). The effective mean stress (σ_m′) is then calculated as follows:

where σ₁′, σ₂′, and σ₃′ are the effective principal stresses.

In the CRS tests, the effective principal stresses σ₂′ and σ₃′ were equal (Fig. F2), and Equation 3 becomes

where σ′_A and σ_H′ are the effective axial stress and effective horizontal stress, respectively. Given that K₀ = σ_H′/σ′_A = 0.6,

Using the hydraulic conductivity and the coefficient of volume compressibility (m_v) of the specimen, the coefficient of consolidation (C_v) can be calculated as follows (ASTM International, 2006):

and

where Δε and Δσ′_A are the changes in axial strain and effective axial stress. Specimen void ratio (e) is calculated from the measured displacement (ASTM International, 2006):

where H_n is the specimen height at a given point in the test, and H_s is the equivalent height of the solids, which remains constant throughout the experiment. We determined H_s for each sample from the wet mass and dry mass after oven drying at 105°C for 24 h.

Porosity (ϕ) is related to void ratio by

A salt mass correction for the pore fluid was done following the IODP method used for moisture and density (MAD) measurements that accounts for salt in seawater (Blum, 1997). For our uniaxial tests, we also report the compression index (C_c), which is the slope of the virgin compression line in e-log σ′_A space (Table T3; e.g., Fig. F4):

The preconsolidation stress (P_c′) for each sample represents the maximum past burial stress in a monotonically increasing stress regime (Holtz and Kovacs, 1981). The preconsolidation stress is determined using two commonly used methods: the Casagrande and strain energy density (SED) methods. The Casagrande method defines P_c′ graphically from the intersection of the virgin compression line and the bisector of a tangent and constant void ratio line from the point of maximum curvature in e-log σ′_A space (Fig. F4) (Casagrande 1936). The SED method defines P_c′ using the change in the rate of total work as a function of effective axial stress (Becker et al., 1987). In the case of 1-D consolidation, SED is given by (Becker et al., 1987; Germaine and Germaine, 2009)

(11)

Preconsolidation stress for the SED method was highly sensitive to the portion of the virgin compression line used to define a linear fit, resulting in large uncertainties. Therefore, the Casagrande values are used for determining the degree of consolidation. The overconsolidation ratio (OCR) for each sample is defined as

where σ_vh′ is the expected in situ effective vertical stress for hydrostatic conditions.

Effective vertical stress is calculated using the shipboard bulk density data (Expedition 343/343T Scientists, 2013). An OCR value <1 indicates that the sample is underconsolidated and that in situ pore pressures may exceed hydrostatic conditions, an OCR value of 1 indicates that the sample is normally consolidated, and an OCR value >1 indicates that the sample is overconsolidated.

Isostatic loading and permeability tests

A second set of deformation and permeability experiments were conducted in parallel with the uniaxial CRS tests to produce independent values of porosity and permeability under an isostatic stress path (σ₁′ = σ₂′ = σ₃′ = P_c, P_c = confining pressure), as opposed to the uniaxial stress path achieved in the uniaxial apparatus. For these tests, the sample was placed within a rubber jacket and loaded in a triaxial apparatus (Fig. F3). The sample was equilibrated for 24 h at 400 kPa pore pressure to ensure saturation and that any air in the pore pressure lines was dissolved. After saturation and equilibration, specimens were loaded under drained isostatic stress conditions. The pore pressure was maintained at a constant 400 kPa, and the confining pressure and axial stress (P_c = σ_A) was increased in a series of 8–10 stress steps, from 4 to 55 MPa for the prism sample (Section 343-C0019E-5R-1) and from 4 to 87 MPa for underthrust sample (Section 20R-1) (see TESTDATA in “Supplementary material”). After each stress increment, we allowed the sample to equilibrate for ≥24 h prior to measuring permeability. Confining pressure and total axial stress, pore fluid pressure (P_f) of the upstream and downstream sample ends (P_{p_up} and P_{p_down}), axial displacement, and pore volume changes are recorded every 10 s throughout the experiment. Specimen porosity throughout the experiment is calculated based on final specimen porosity determined from wet and dry mass and from changes in pore volume recorded continuously during the test.

To allow for direct comparison of the triaxial tests with the uniaxial CRS tests, the effective mean stress (σ_m′) is reported for all of our experiments. The effective mean stress (σ_m′) for the triaxial experiments is defined as σ_m′ = P_c – P_f. Permeability was measured at each load step using a steady-state constant head method. In these tests, pore pressure gradients from 100 to 600 kPa were imposed on the samples, and volumetric flow rates (Q) were measured after reaching steady-state, defined using the criterion that the difference between upstream and downstream flow rates was <5%. Permeability was then calculated using Darcy’s law:

where A is the cross-sectional area of the sample and the pore pressure at the upstream, and downstream sample ends are P_{p_up} and P_{p_down}, respectively.

Top of page   |    Previous   |    Next