| IODP Proceedings Volume contents Search | |||||||||||||||||||||||||||||
 | |||||||||||||||||||||||||||||
| Expedition reports Research results Supplementary material Drilling maps Expedition bibliography | |||||||||||||||||||||||||||||
|
doi:10.2204/iodp.proc.322.203.2012 Methods and materialsSampling and sample handlingSampling of whole-round cores during Expedition 322 was conducted after partitioning the recovered core into 1.5 m length sections and temperature equilibration for multisensor core logging. Sample intervals of 10–20 cm of undisturbed silty claystones suitable for geotechnical testing were identified by X-ray computed tomography. The selected sample intervals were cut and sealed in the core liner and packed in aluminum vacuum bags. Samples of sandy sediments were taken from the working halves, with exception of Sample 322-C0012A-45R-4, 79–95 cm (Table T1). Selected specimens were put in a bisected core liner, wrapped in plastic film, and sealed in an aluminum vacuum bag. During storage and transport, all samples were kept refrigerated at ~4°C until used for laboratory testing. Incremental loading consolidation testsIncremental loading consolidation tests were conducted on all silty claystone and sandstone samples with a fixed-ring oedometer in the Center for Marine Environmental Sciences (MARUM) geotechnical laboratory at the University of Bremen. The oedometer ring that confines the sample is fixed to the bottom of an open cell filled with artificial seawater to which the pore water in the sample has free access. The ring has an inner diameter of 25.40 mm (area = 5 cm2) and a height of 14.75 mm. The small ring size was necessary to prepare intact samples extracted from half rounds parallel to the longitudinal axis of the core. All samples were carefully trimmed with a knife to fit exactly into the ring. Because of the presence of microcracks (possibly drilling induced), sample height varied between 10.56 and 14.75 mm, and no intact specimen could be prepared from Sample 322-C0011B-28R-1, 0–17 cm. This sample was remolded for the consolidation test. For the consolidation experiments we used two loading frames with maximum vertical stress capacities of 16 and 20 MPa for an area of 5 cm2. Stress is applied to both frames with weights on a lever arm. At the beginning of each experiment, the specimen was saturated with artificial seawater (35‰ salinity) in the oedometer cell for 24 h with applied vertical stresses (σv) of 0.01–0.06 MPa. For the loading phase of the consolidation test, the applied vertical stress is doubled until the maximum capability of the respective loading frame was reached. Each load step was maintained for 24 h to allow complete pore pressure dissipation such that the applied vertical stress is equal to the vertical effective stress (σ′v). For the unloading phase, we reduced the applied vertical stress by ~75% every 24 h. During the experiment, the time (t), the applied vertical stress (σv), and the relative displacement h (accuracy = ±0.01 mm) were recorded. After the experiment, the final sample height was measured with a caliper (accuracy = ±0.01 mm). The final void ratio eend is calculated by
where Vw is the volume of pore water (in cm3) and Vb is the bulk volume of the compacted specimen after the test (in cm3). Vb is given by the height of the sample and the ring area (5 cm2), and Vw is determined following Blum (1997) by oven drying the sample for 24 h at 105°C. The change of void ratio during the consolidation test was calculated using the displacement data with the final void ratio as reference. Plots of consolidation data from incremental loading tests show void ratio e at the end of loads steps when σv = σ′v against the logarithm of the vertical effective stress (σ′v). The compressibility characteristics of the samples are reported as compression indexes. The compression index is the slope of a linear portion in the plot given by
A preconsolidated sample undergoes a recompression phase, a primary consolidation phase, and an unloading phase. For the different phases of the consolidation test, we determined the recompression index (Cr), the compression index (Cc) and the expansion index (Ce), respectively. Consolidation test data also allow estimation of the maximum past effective stress (P′c) experienced by a clay-rich sediment. In the present study, we used the Casagrande method to estimate the value of P′c (Casagrande, 1936). In this method, the value of P′c is obtained from the intersection between the backward projected line of primary consolidation and the bisecting line between the horizontal and tangent line at the point of maximum curvature of the consolidation curve. Because of low data resolution of incremental loading consolidation tests, we interpolated consolidation data on the basis of a rational equation fit. The point of maximum curvature of the fit was computed using a second-order numerical solution following Dawidowski and Koolen (1994). The consolidation state of the samples is defined by the ratio of P′c to the in situ vertical hydrostatic effective stress σ′vh. This is known as the overconsolidation ratio (OCR) and defined as P′c/σ′vh. The vertical hydrostatic effective stress corresponds to the stress at depth assuming that only hydrostatic pore pressure exists in the sediment and is calculated by integrating the sediment’s bulk density from shipboard measurements upward from the sample depth to the seafloor combining Expedition 322 and 333 data (see the “Expedition 322 summary” chapter [Underwood et al., 2010]; Expedition 333 Scientists, 2011). σ′vh is approximated by
where
If the maximum past effective stress P′c is equal to σ′vh (OCR = 1), the tested sample is considered to be normally consolidated. For OCR < 1, the consolidation state is referred to underconsolidated, and in the inverse case the sample is overconsolidated. Yielding of sandstones does not reflect the maximum past effective stress and may have other causes (e.g., Karig and Hou, 1992). Nonetheless, we applied the same analysis as for the silty claystones and keep also the nomenclature. Thus, we determined compression indexes, employed the Casagrande (1936) method to estimate the yieldpoint at which a rapid increase in compressibility occurs, and related the yieldpoint to the in situ vertical hydrostatic stress to obtain OCR. As part of the consolidation analysis, the hydraulic conductivity K (m/s) was computed by
Cv is the coefficient of consolidation (m2/s) and is determined for each load step using the Taylor method as described in Craig (2004). γw denotes the unit weight of the pore water (kg/[m2 × s2]), and mv is the coefficient of volume compressibility (m2/MN), calculated by
where
The logarithm of the obtained hydraulic conductivity is presented as a function of void ratio for each sample (Fig. F3) and fitted with a relation of
where α is the change of hydraulic conductivity with void ratio and log(K0) is the projected value at zero void ratio. Hydraulic conductivity is a function of fluid properties, which may be different for laboratory and in situ conditions. Therefore, it is convenient to convert K into intrinsic permeability, which is a function of the porous medium alone (Fetter, 2001). The relation between hydraulic conductivity and the intrinsic permeability is
where ν is the dynamic fluid viscosity of water at 20°C (0.001 Pa·s) (Fetter, 2001). We assume that the void ratio at the laboratory-determined maximum past effective stress (P′c) represents the in situ void ratio eP′c (Dugan and Daigle, 2011). To determine the in situ intrinsic permeability, we combine the fitted relation between void ratio and hydraulic conductivity with the conversion from hydraulic conductivity into intrinsic permeability k (m2) such that the in situ intrinsic permeability ki can be estimated by
The in situ intrinsic permeability ki of sandstones was also determined for void ratios that correspond to the maximum past effective stress in incremental loading tests. Constant rate of strain consolidation testsInsufficient resolution of incremental loading test data prevented the calculation of Cv and K for intact sandstone samples. Therefore, we conducted CRS tests on remolded samples in addition to the incremental loading consolidation tests on intact samples to determine the e-log(K) relation. Void ratios from incremental loading tests are put into the best fit equation to calculate the associated K for intact samples. Together with the calculated mv from the incremental loading tests, Cv is computed by
We used a custom-made CRS cell with a diameter of 63 mm in combination with a computer-controlled hydraulically driven load frame (Fig. F4). The larger diameter of the cell made it necessary to remold the samples. Samples were disaggregated by careful crushing of the specimen with a flexible spatula and simultaneous addition of artificial seawater until a void ratio of ~1 was achieved. The height of the sample varied between 21.84 and 9.96 mm depending on available sample material. The sample was confined by the stainless steel cell wall, which ensures only uniaxial strain. The sample was allowed to drain upward, with atmospheric pressure at the top while the base is sealed. Vertical stresses of up to ~70–80 MPa were applied at strain rates up to 22%/h to induce a sufficient basal pore pressure in the specimen. We neglected the holding phase at maximum applied stress because compression behavior was already determined from incremental loading tests. The time (t), applied vertical stress (σv), excess pore water pressure at the base (Δu), and the relative displacement (accuracy ± 0.01 mm) were recorded during deformation. After the experiment the sample height and the final void ratio were determined in the same way as the postsample handling of incremental loading tests. We computed vertical effective stress σ′v and hydraulic conductivity K according to American Society for Testing and Materials (ASTM) Standard D854-06 (ASTM, 2006) with
where
The strain rate is calculated by
The logarithm of hydraulic conductivity is presented as a function of void ratio. Curve fitting was performed similar to incremental loading tests, again using the following linear relation:
|
||||||||||||||||||||||||||||
× hn × h0 × γw)/(2 × Δu),
= strain rate.
= [(h0 – hn)/h0]/(tn – t0).