It is postulated that the presence of swelling clay minerals in unconsolidated fine‐grained saturated marine sediments leads to deviations from the normally assumed ideal elasticity of the solid phase. In the proposed “effective grain model” (EGM), the elastic grain material is consequently replaced by an effective medium made up of a homogeneous elastic mineral phase that is isotropically interspersed with cylindrical, “penny‐shaped” inclusions of low aspect ratio representing the intracrystalline water layers in the swelling clay minerals. The two‐phase nature of the grain material thus specified results in a wave‐energy consuming squirt‐flow process from the inclusions into the pore space. Introducing the EGM into the Biot‐Stoll model (BSM) via the complex bulk modulus of the dissipative grain material leads to a better fit to literature data on the attenuation of compressional waves than does the original BSM alone. Since swelling clay minerals occur in nearly all clay‐bearing sediments, it is concluded that the attenuation mechanism of the EGM may represent a universal contribution to the overall intrinsic anelasticity of unconsolidated fine‐grained saturated marine sediments in the frequency range from a few kilohertz to about 1 MHz.

The elastic properties of sand strongly depend on the grains' contact stiffness, which can be increased significantly by solid matter and, depending on frequency, viscous fluid acting as contact cement. To calculate seismic velocities in precompacted fluid-cemented sand, we examine how a small amount of viscous fluid at the grain contacts influences their normal and tangential stiffnesses as a function of effective pressure. Using the Hertz-Mindlin approach and considering oscillatory loading in addition to precompaction of a combination of two elastic spheres, we extend the dry-contact elastic theory by a viscoelastic formulation.Here, we describe the radial flow of the fluid cement induced by the oscillations of the grains' surfaces around the direct contact, a process that leads to a complex normal stiffness and stiffness/frequency dispersion. In the resulting combined model, the low-frequency real part of the complex normal stiffness identical is to the original HertzMindlin expression. The magnitude of the dispersion is governed by the amount of viscous cement; magnitude decreases as effective pressure increases. The frequency of the maximum imaginary part of the normal stiffness is determined mainly by cement viscosity and contact geometry. The tangential contact stiffness virtually is not influenced by the viscous fluid.Comparison of predicted results with data from pulse transmission experiments (500 kHz) on glass beads with two different fluids shows an excellent fit in P-wave velocities (V p ), whereas S-wave velocities (V s ) are systematically overestimated by the model. The experimental results confirm, however, the predicted change with effective pressure in the V P /V S ratio for both examined cases as well as reflect the predicted increase in V P and V S , respectively, between the two cases. This implies that our viscoelastic formulation represents a reasonable way to describe the role of viscous cement in sand.

This paper presents a model of acoustic wave propagation in unconsolidated marine sediment, including compaction, using a concept of a simplified sediment structure, modeled as a binary grain-size sphere pack. Compressional- and shear-wave velocities and attenuation follow from a combination of Biot's model, used as the general framework, and two viscoelastic extensions resulting in complex grain and frame moduli, respectively. An effective-grain model accounts for the viscoelasticity arising from local fluid flow in expandable clay minerals in clay-bearing sediments. A viscoelastic-contact model describes local fluid flow at the grain contacts. Porosity, density, and the structural Biot parameters (permeability, pore size, structure factor) as a function of pressure follow from the binary model, so that the remaining input parameters to the acoustic model consist solely of the mass fractions and the known mechanical properties of each constituent (e.g., carbonates, sand, clay, and expandable clay) of the sediment, effective pressure, or depth, and the environmental parameters (water depth, salinity, temperature). Velocity and attenuation as a function of pressure from the model are in good agreement with data on coarse- and fine-grained unconsolidated marine sediments.

No abstract