v' = 0,
h' would be 0.2 ± 0.05
MPa.
The reconsolidation-consolidation tests showed clearly that yield stresses for this section of core are very low considering its depth of burial. Test T-70 displayed the clearest definition of yield stress, which ranged between 0.9 and 1.3 MPa using various mechanical relationships (Fig. 5). A similar low yield stress was observed on or constrained by two other tests. At the estimated yield stress the effective horizontal stress was 0.8 ± 0.1 MPa.
Such a low yield stress might be confused with mechanical effects sometimes encountered near the initiation of a test, such as the dissipation of negative pore pressure (e.g., Bishop et al., 1975), but the stress path beyond this apparent yield condition clearly shows a virgin consolidative behavior. For example, the compressibilities at higher stress decrease continuously (Fig. 5) to effective vertical stresses greater than 10 MPa, which is well above any expected in situ stress. In addition there were no indications in the shape of the test curves for any sediment cementation.
Post-yield stress paths for three uniaxial strain tests defined a stress ratio with a slope of 0.56 ± 0.1 and was quite constant with respect to stress over the stress range probed. Extrapolation of the stress ratio curves did not pass through the origin; at
v' = 0,
h' would be 0.2 ± 0.05
MPa.
The consolidation characteristics of this material were corroborated and extended by a uniaxial strain test with multiple unload-reload cycles, which also provided data on the elastic characteristics of sediment
(Fig. 6). The primary objective of this test was to explore the nature of the reconsolidation stress path as a function of the initial test stress state. Starting at an isotropic stress of 0.69
MPa, the vertical stress was increased, under uniaxial strain conditions, to 3.5
MPa, which is well beyond the yield stress. From this state an unload-reload cycle was run, with the vertical stress during the reload phase increased to 6.5
MPa. This reload path constituted a "normally consolidated" stress path. The sample was again unloaded from the 6.5-MPa stress under uniaxial strain, but was then returned to the same stress state as that at the first unload phase, which required a triaxial strain path. The final phase was another uniaxial reconsolidation-consolidation to
v' = 8.5
MPa, with the sample maintaining a constant but slightly larger cross-sectional area from that during the earlier phases. This last phase provided an "overconsolidated" path (or initial stress state).
Because of the similar elastic and plastic compressibilities at higher stress levels, axial stress-strain relationships did not define the yield stress clearly during this test. Yield stresses were more clearly delineated during both reload phases by changes in other ratios, but were at values of
v' slightly less than the previous consolidation stresses. The stress ratios during consolidation phases were well-defined at 0.55 ± 0.1, and the elastic stress ratios were less well defined at about 0.44. ± 0.3. A phase of constant
v' at 8.5 MPa produced some axial creep, but no significant change in lateral stress (h').
The overconsolidated path during this test did not show any similarities to those of cemented sediments (Fig. 6), which could reflect a number of possible causes. The relatively small differences in elastic and consolidation behavior, plus the relatively small degree of overconsolidation, was probably the greatest cause. In addition, the Ko line progressively shifted upward on the q-p' plane with each cycle, which reduced the need for a "cementation" effect.
The reference consolidation test involved the primary consolidation of disaggregated remnants of the section of core remaining after subsampling. Disaggregation was accomplished by crushing the dried material to millimeter size followed by reduction in a ball mill to a mean diameter of less than 5 µm, as estimated optically. Because there were very few grains in the original sediment larger than 5 µm, the disaggregated material was presumed to closely mimic the physical characteristics of the original sediment. The powdered material was mixed with distilled water into a slurry with a porosity of approximately 70%.
Consolidation of this slurry began with several steps of constant vertical stress, but from a stress of less than 1 MPa, proceeded at a constant rate of stress increase of 0.035 MPa/hr, except for a short stress hold at 1.6 MPa, to a maximum stress of 9.82 MPa, which was maintained for 26 hr. A sidewall correction, based on past experience (Karig and Hou, 1992), would have increased Ko by 0.01 or less and was ignored. Strains and porosities were determined by back-calculating from the final sample dimensions and weight, assuming a grain density of 2.78 g/cm3 (Shipboard Scientific Party, 1994a).
Porosity decreased logarithmically with
v' to 30.4%, with an additional 0.3% porosity decrease during the final stress hold
(Fig. 7). The porosity-log v' curve generated a value for Ac
of 1.3%/decade of log v' The stress ratio (
h'/v') was constant at 0.60 between
v' from 1 to 9.8
MPa, the range over which h' could be reliably measured. These parameters differed significantly from those of the post-yield consolidation on the natural samples
(Fig. 7). The disaggregated equivalent had significantly lower porosities, was more compressible, and had a higher Ko. During the final period of secondary consolidation,
h' increased by only 0.016 MPa and that occurred during the first 1.5 hr of the hold. This increase resulted from dissipation of a very small degree of fluid overpressure during the constant stress-rate test, with the subsequent stability in
h' demonstrating the relative insensitivity of Ko to secondary consolidation.
