Energy demand is driven by increasing population and quality of life. Fractures localize
mechanical deformations and fluid flow, and they impede heat flow through the rock
matrix. Therefore, fractures present a challenge to both the recovery of underground
energy and long-term waste disposal solutions like carbon geological storage.
Fracture are planar discontinuities that form when brittle rocks. The discrete
element method can model the complex micromechanics of rock failure. In this thesis
we present a digital rocks analogue which is used to explore 1) the rock brittle-to-ductile
transition with increased confining stress 2) the meaning of friction in intact rocks and
what factors control confinement-dependent strength, 3) exhumation damage and its
effect on rocks strength, and 4) multistage loading.
The design, analysis and construction of a large-scale true triaxial load frame
opens the door to geophysical studies on fractured rock masses. The frame can subject
large cubical rock specimens (50cm × 50cm × 50cm) to boundary stresses up to 3 MPa.
Auxiliary systems include active acoustic monitoring, passive acoustic emissions
sensing, and high-pressure fluid injection. The evolution of P-wave velocity under
anisotropic stress demonstrate the device’s capabilities.
The true triaxial load-frame and the high-pressure fluid injection system are used
to study hydraulic fracturing in pre-fractured media. We explore the competing
influences of stress and rock mass fabric. Notably, even under extreme stress
anisotropy, the fluid invades all fracture sets of our pre-fractured specimen. Fracture
intersections act as flow conduits and feed the invading fluid into the adjacent fractures,
and local phenomena such as gouge-displacive fingering are identified.
Thermal contact resistance impedes heat flow between neighboring rock blocks
in fractured rocks. Contact resistance manifests as a discontinuous thermal gradient. It
strongly influences the rock effective thermal conductivity and makes it sensitive to
water saturation, stress, and the presence of gouge material. Finally, we conduct
detailed thermal conductivity measurements on sand-silt gouge mixtures and propose
physics-inspired models that accurately predict the thermal conductivity and mass
density of dry and wet specimens as a function of stress and fines content.
|Date of Award
- Physical Sciences and Engineering
|J. Carlos Santamarina (Supervisor)