Description |
To facilitate material point method (MPM) simulations of hypervelocity shaped-charge jet penetration into fluid-saturated porous geomaterials, a new constitutive model has been developed in which fluid pressure accumulates under pore collapse with a corresponding change in elastic properties and strength. The approach is inspired by an idealized model for the elastic-plastic collapse of a fluid-filled thick spherical shell, for which the solution can be expressed in terms of an effective stress. The continuum implementation evolves an isotropic backstress as a state variable within a classical plasticity framework, so that increasing pore pressure strengthens the material in hydrostatic compression while reducing the shear strength. This approach is shown to produce experimentally observed trends for the depth of penetration of a tungsten jet into drained and undrained sandstone. A new approach for numerical solution of plasticity models is presented that decouples the nonlinear solution of the consistency parameter from the calculation of the projection to the yield surface, allowing for robust convergence for models with highly nonlinear hardening laws. A nongradient algorithm is used for finding the closest-point projection in energy-mapped stress space; this method eliminates many difficulties associated with traditional solution methods and can be easily applied to yield criteria of arbitrary complexity. Enhancements to an existing multistage nested return solution method are also described. Mesoscale simulation is used to investigate unvalidated assumptions in the continuum model development. In silico hydrostatic testing explores the limitations of the effective stress approach for plastic deformation of a limestone microstructure with pressure-dependent matrix properties. Dynamic simulations determine the crush response of fluid-saturated granular structures with implication on the need for coupled multiple velocity fields in high-rate loading. The convected particle domain interpolation (CPDI) method is modified to allow for efficient parallelization and to enable control of numerical fracture. Testing of this method reveals that the smeared damage approach in the material point method leads to significant mesh dependence and advection error. Implications to the simulation of fracture and fragmentation are discussed |