The evolution of fault strength, permeability, and acoustic properties in experimental studies from fault initiation through the seismic cycle

Within Earth's crust, fault zones accommodate significant deformation and strain resulting from plate tectonics and other processes. Due to the hazards associated with fault slip, much work has been done to understand the factors controlling deformation style within these zones, which can range from quiescent aseismic slip to devastating earthquakes, such as the 2011 Mw9 Tohoku Oki earthquake. In particular, our understanding of processes like slow earthquakes and healing within fault zones remains unclear. Additionally, as fault zones develop they become highly differentiated from their parent material, as fault materials mix, break, rotate, and develop into fabrics. These changes, which vary with fault composition, chemistry, stress, and strain, can cause significant strength changes and permeability decrease. In particular, fault permeability can dictate regional fluid flow and may allow faults to act as petroleum traps and seals. Despite the importance of such faults, our understanding of their permeability evolution, especially in marinesediment basins, is relatively poor.;In this dissertation, I investigated the evolution of fault zones as they initiate and proceed through the seismic cycle. In particular, I studied the origins of slow earthquake slip, the mechanisms controlling deformation band formation, and the evolution of fault fabric and permeability with fault zone development. This work was predominantly conducted on laboratory fault zones in a biaxial forcing apparatus under conditions appropriate for fault development in Earth's upper crust. In chapter 1, I present the first laboratory observations of repetitive, slow stick-slip in fault zone materials (serpentine) and mechanical evidence for their origin. In particular, we document a transition from unstable to stable frictional behavior above a threshold velocity of ~10 microm/s. Additionally, these events are accompanied by precursory elastic wave speed reduction (2--21%) that begins up to 60 seconds before failure, perhaps suggesting a reliable earthquake predictor. In chapter 2, I investigate fault zone evolution through the seismic cycle and as it initiates, documented via elastic wave speed measurements. These experiments were conducted on halite under conditions where pressure-solution is operative, and they show the interplay of elastic wave speed measurements with porosity and fabric formation. Indeed, these observations point to a new technique for non-invasive fabric observation within laboratory and natural fault zones. In chapter 3 and chapter 4, I also discuss fault zone initiation and development for two specific cases: deformation bands and clay-rich marine-sediment faults.;Chapter 3 highlights deformation band formation through laboratory experiments, and shows that fault strengthening via shear-driven comminution is the likely mechanism limiting strain. I observe significant strengthening at low shear strains (e.g., gamma < 5), and tie these observations directly to particle-size reduction. To accommodate fault-like strain, many deformation bands form within a given region, significantly limiting fluid flow, similar to standard faults. In chapter 4, I discuss the permeability evolution of faults under conditions appropriate for marine-sediment basins, like the Gulf of Mexico. In particular, the role of halite and clay within faults adjacent to salt domes and the possibility for multimechanism behavior, including brittle deformation at high strain rates and ductile deformation and pressure solution at slower rates is unclear. We found that that fault permeability can be reduced by up to 2-4 orders of magnitude with clay content, that small load (sigman < 6 MPa) and small strain (gamma < 5) can cause < 1 and < 2 orders of magnitude permeability reduction, respectively, and that halite is largely interchangeable with quartz for permeability.;This dissertation provides mechanical insight on a variety of fault deformation styles, as well as their implications. I document the first observations of slow earthquakes in the laboratory, evidence for their origins, and evidence for viable premonitory earthquake signals. I characterize fault fabric evolution leading up to and through the seismic cycle, and suggest a new tool for these observations. I provide significant evidence for the mechanism controlling deformation band formation and arrest. And I show how permeability may evolve within marine-sediment basin faults.