Michael A. Homel

Mesoscale Validation of Simplifying Assumptions for Modeling the Plastic Deformation of Fluid-Saturated Porous Material

Mesoscale simulation of fluid-saturated porous materials is used to investigate unvalidated assumptions in continuum modeling of geomaterials at high rates. The importance of pressure-dependence of the matrix material on the applicability of the effective-stress approach during plastic deformation is explored. The hydrostatic load-unload response of saturated porous structures having both pressure-dependent and pressure-independent matrix materials is modeled using the material point method. Results are used to validate a semi-empirical strain-to-yield effective-stress model in which the pore pressure evolves with local material deformation, an approach that is applicable to materials with closed porosity or those loaded at sufficiently high rates that fluid transport can be neglected. Mesoscale simulations are used to estimate the strain rates beyond which fluid flow through the matrix can be neglected.

Relaxing the multi-stage nested return algorithm for curved yield surfaces and nonlinear hardening laws

The multi-stage nested return algorithm is an efficient approach for implementing plasticity and damage models that addresses issues of nonuniqueness and nonconvergence, and which was previously shown to perform well in a variety of verification tests. Straightforward modifications to this algorithm are shown to improve the robustness for high-curvature yield surfaces and nonlinear hardening laws. Improved methods for computing the initial “fast return” are presented with application to two-surface geomodels. These modifications reduce the need for subcycling while maintaining the efficiency of the algorithm.

Microscale investigation of dynamic impact of dry and saturated glass powder

The response of particulate materials to impulsive loading includes complex interactions between grains due to fracture and comminution and the presence of interstitial material. The quasi-static strength of saturated powders is related to the concept of “effective stress” in which the fluid stiffens the material response and reduces the shear strength. However, detailed information regarding the effects of saturation under dynamic loading is lacking since static equilibrium between phases cannot be assumed and the interaction becomes more complex. Recent experiments on the IMPULSE (IMPact System for ULtrafast Synchrotron Experiments) capability at the Dynamic Compression Sector (DCS) of the Advanced Photon Source (APS) have captured in-situ X-ray phase-contrast images of shock loaded soda lime glass spheres in dry and saturated conditions. Previous investigations have observed reduction of fragmentation attributed to “cushioning” of an interstitial fluid in impact recovery experiments. The differences between the modes of deformation and compaction are compared with direct numerical simulations showing that the cause of fracture is different. In drained (dry) impact experiments at 300 m/s, the fractures initiate near the contact point between grains. In fully saturated experiments with identical impact conditions, spallation is observed during the incident stress-wave passage in the glass before the H2O has equilibrated.

On mesoscale methods to enhance full-stress continuum modeling of porous compaction

The dynamic compaction of initially porous material is typically treated in continuum dynamics simulations via adjustments to the scalar equation of state (EOS) of the bulk, porous material relative to that of the solid. However, the behavior during compaction is governed by inelastic processes, as the solid material deforms, largely by shearing, to fill the voids. The resulting response depends on the strain path, e.g. isotropic versus uniaxial loading. Adjustments to the EOS are therefore fundamentally unsuited to describing porous compaction, and it is desirable to consider porous effects through the stress and strain tensors. We have performed mesoscale simulations, resolving the microstructure explicitly, to guide the construction of continuum models. These simulations allow us to study the interplay between strength and EOS in the solid, the extent of dissipative flow versus non-dissipative displacement, and the evolution of porosity and micro-morphological features.