A major goal in earthquake physics is to derive a constitutive framework for
fault slip that captures the dependence of shear strength on fault rheology,
sliding velocity, and pore-fluid pressure. In this study, we present H-MEC
(Hydro-Mechanical Earthquake Cycles), a newly-developed two-phase flow
numerical code - which couples solid rock deformation and pervasive fluid flow
- to simulate how crustal stress and fluid pressure evolve during the
earthquake cycle on a fluid-bearing fault structure. This unified,
continuum-based model, incorporates a staggered finite
difference-marker-in-cell (SFD-MIC) method and accounts for full inertial (wave
mediated) effects and fluid flow in poro-visco-elasto-plastic compressible
medium. Global Picard-iterations and an adaptive time stepping allows the
correct resolution of both long- and short-time scales, ranging from years
during slow tectonic loading to milliseconds during the propagation of dynamic
ruptures. We present a comprehensive in-plane strike-slip setup in which we
test analytical poroelastic benchmarks of pore-fluid pressure diffusion from an
injection point along a finite fault width. We then investigate how pore-fluid
pressure evolution and solid-fluid compressibility control sequences of seismic
and aseismic slip on geologic faults. While the onset of fluid-driven shear
cracks is controlled by localized collapse of pores and dynamic
self-pressurization of fluids inside the undrained fault zone, subsequent
dynamic ruptures are driven by solitary pulse-like fluid pressure waves
propagating at seismic speed. Furthermore, shear strength weakening associated
with rapid self-pressurization of pore-fluid can account for the slip-fracture
energy scaling observed for large earthquakes.