Judith S. Chester

Fracture surface energy of the Punchbowl fault, San Andreas system.

Fracture energy is a form of latent heat required to create an earthquake rupture surface and is related to parameters governing rupture propagation and processes of slip weakening. Fracture energy has been estimated from seismological and experimental rock deformation data, yet its magnitude, mechanisms of rupture surface formation and processes leading to slip weakening are not well defined. Here we quantify structural observations of the Punchbowl fault, a large-displacement exhumed fault in the San Andreas fault system, and show that the energy required to create the fracture surface area in the fault is about 300 times greater than seismological estimates would predict for a single large earthquake. If fracture energy is attributed entirely to the production of fracture surfaces, then all of the fracture surface area in the Punchbowl fault could have been produced by earthquake displacements totalling <1 km. But this would only account for a small fraction of the total energy budget, and therefore additional processes probably contributed to slip weakening during earthquake rupture.

Microfracture analysis of fault growth and wear processes, Punchbowl Fault, San Andreas system, California

Abstract Fabric and timing relations of mode I microfractures are used to test current hypotheses for the origin of damage along large-displacement faults by the processes of fault growth and wear. Oriented samples 0.075 m to 1 km from the Punchbowl fault surface (i.e. ultracataclasite layer) document an increase in development of preferred orientation and increase in density of microfractures towards the ultracataclasite layer, defining a zone of fault-related microfracture damage about 100 m thick. A distinct microfracture set that is perpendicular to the slip direction of the fault is present throughout the damage zone. This implies that the average orientation of the maximum principal compressive stress within the damage zone was nearly normal to the fault surface. Two additional microfracture sets are present, one oriented at low angles to the fault within meters of the ultracataclasite layer, and another low-angle set that occurs in the outermost damage zone. The preferred orientations and timing relations are most consistent with local damage accumulation from stress cycling associated with slip on a geometrically irregular, relatively weak fault surface. Low-angle microfractures nearest the ultracataclasite layer also may record wear associated with the passage of earthquake ruptures, and those in the outermost damage zone may be consistent with Andersonian fault formation and subsequent fault weakening.

Stress and deformation along wavy frictional faults

Locally inhomogeneous stress states are expected along faults owing to slip on geometrically irregular fault surfaces. We use an analytical model of elastic deformation along a wavy frictional fault to evaluate the variation in local stress state as a function of surface roughness, elastic modulus, slip, coefficient of friction, and far-field stress. The total stress state along the fault may be described by the sum of a basic stress component resulting from frictional slip on a planar fault surface and a perturbed stress component resulting from the presence of roughness. Roughness produces a variation in normal stress across the fault surface, and assuming roughness and modulus appropriate to crustal faults, the normal stress should be reduced to a near-zero magnitude locally, such that separation of fault walls is likely. The large variation in normal stress along the fault surface resulting from fault roughness may be responsible, in part, for complexity in moment release during large earthquakes and for lateral variation in seismic coupling along faults. The variation in principal stress orientations and magnitudes along a fault increases with a decrease in the coefficient of friction of the fault. The location and size of regions with a high likelihood for brittle failure depend on the orientation of the far-field principal stress and fault friction. The average orientation of the principal stresses in the region of likely failure is not the same as the far-field principal stress orientation. Although inversion of earthquake and fabric data for stress orientation along a fault may be possible, the model results suggest that inversion results are insufficient to determine far-field stress states and fault friction without additional independent data.

Fault-propagation folds above thrusts with constant dip

Abstract Previous models of fault-propagation folds have been extended to include the case where folding initiates above a thrust ramp of constant dip. The model may be useful in analyses of fold-thrust structures in thin- and thick-skinned compressional terrains. It may be most applicable to fault-propagation folds that develop as a fault propagates from a thick, massive, brittle unit into a thinly layered, anisotropic unit.

Mechanisms of compaction of quartz sand at diagenetic conditions

The relative contribution of cracking, grain rearrangement, and pressure solution during experimental compaction of quartz sand at diagenetic conditions was determined through electron and optical microscopy and image analysis. Aggregates of St. Peter sand (255±60 μm diameter grain size and porosity of approximately 34%) were subjected to creep compaction at effective pressures of 15, 34.5, 70, and 105 MPa, temperatures of 22 and 150°C, nominally dry or water-saturated (pore fluid pressure of 12.5 MPa) conditions, and for times up to one year. All aggregates displayed transient, decelerating creep, and volume strain rates as low as 2×10−10 s−1 were achieved. The intensity of fracturing and degree of fragmentation increase with volume strain and have the same dependence on volume strain at all conditions tested, indicating that impingement fracturing and grain rearrangement were the main mechanisms of compaction throughout the creep phase. The increase in fracture density and decrease in acoustic emission rate at long times under wet conditions reflect an increase in the contribution of subcritical cracking. No quantitative evidence of significant pressure solution was found, even for long-term creep at 150°C and water-saturated conditions. Comparison of our findings to previous work suggests that pressure solution could become significant at temperatures or times somewhat greater than investigated here.

Influence of layering and boundary conditions on fault-bend and fault-propagation folding

The influence of heterogeneous layering and boundary conditions on the structural development of fault-bend and fault-propagation folds has been investigated through petrographic study of nonscaled rock models. The models are deformed in a triaxial rock-deformation apparatus at room temperature and a 50-MPa confining pressure. The models consist of a single layer of sandstone containing a saw-cut ramp that is inclined 20° to the layering, and an overlying, intact, thinly layered unit that is composed of limestone interlayered with lead or mica. Analysis of the fold-thrust structures generated in sequentially shortened models with different loading conditions and layer types suggests that the mode of fold-thrust interaction activated upon shortening will depend on fault zone drag, bending and shearing resistance of the hanging wall, shear strength of layer interfaces, and loading conditions. For the models, these parameters may be expressed as a strength ratio describing the resistance to foreland translation relative to the resistance to internal deformation of the thrust system. Low strength ratios favor fault-bend folding. High strength ratios favor internal shortening of the sheet; isotropic and thick (relative to ramp height) units above a propagating thrust tip will shorten primarily by faulting, whereas thinly layered, anisotropic units will shorten by fault-propagation folding. During both modes of fold-thrust interaction, the dips of the fold limbs increase, interlimb angles decrease, and imbricate faults form in the hanging wall or footwall with shortening. In one model suite, the imbrication is associated with a transition from fault-bend folding to fault-propagation folding and produces a highly asymmetric ramp anticline similar to a second-mode fault-bend fold or to a transported fault-propagation fold. The model data suggest that fault-propagation folding in heterogeneously layered rock occurs by the discontinuous formation, growth, and linkage of faults below the growing fold. Amplification of the fault-propagation fold is affected by the amount of slip transferred out of the deforming region, imbrication, and buckling. The changes in the mode of fold-thrust interaction and modifications in the local geometry and strain distribution that occur during shortening result from slip hardening on faults, or from rotation- or strain-induced variations in the strength of the layers.

Laboratory deformation of granular quartz sand: Implications for the burial of clastic rocks

We explore the influence of mechanical deformation in natural sands through experiments on water-saturated samples of quartz sand. Stresses, volumetric strain, and microseismicity (or acoustic emission, AE) rates were monitored throughout each test. Deformation of quartz sand at low stresses is accommodated by granular flow without significant grain breakage, whereas at high stresses, granulation and cataclastic flow are dominant. Sands deformed under isotropic conditions show compactive strains with an inverse power-law dependence of macroscopic crushing strength on mean grain size. Triaxial compression at high effective pressures produces compactive strain and a high AE rate associated with considerable particle-size reduction. Triaxial compression at low effective pressure produces dilatant granular flow accommodated by grain boundary frictional sliding and particle rotation. On the basis of experiment results, we predict the evolution of porosity and macroscopic yield strength as a function of depth for extensional and contractional basins. Sand strength increases linearly with depth for shallow burial, whereas for deep burial, strength decreases nonlinearly with depth. At subyield stresses, porosity evolves as a function of applied mean stress and is independent of distortional stress. Our predictions are in qualitative agreement with observations of microfracture density obtained from laboratory creep-compaction experiments and with natural sandstones of the Gulf of Mexico basin. Mechanical deformation contributes as much as a 30% increase to fluid pressure evolution, which has particular application to sedimentary systems that display zones of fluid overpressure. Furthermore, deformational strains cannot be fully recovered during uplift, erosion, and unloading of a sedimentary basin.

Displacement and dynamic weakening processes in smectite‐rich gouge from the Central Deforming Zone of the San Andreas Fault

The strength of clay-rich gouge from the Central Deforming Zone (CDZ) of the San Andreas Fault (SAF) was measured using a high-speed rotary shear apparatus to evaluate the potential for unstable slip along the creeping segment of the SAF. Wet and dry gouge was sheared at 0.1–1.3 m/s, 0.5–1.5 MPa normal stress, and 1–20 m displacement. CDZ gouge is weaker wet than dry and exhibits displacement strengthening to peak friction followed by weakening to steady state strength that decreases with increasing velocity. A clay foliation (Unit 2) develops from the initial microstructure (Unit 1) during the first 1.5 m of slip coincident with increasing strength. Subsequent weakening occurs during shear within Unit 2, and subsequently with development of a localized foliated slip zone (Unit 4) and fluidized material (Unit 3). Displacement and dynamic weakening result from slip along clay foliation assisted by shear-heating pressurization of pore fluid in wet gouge and additional grain-size reduction and possible clay dehydration in dry gouge. Peak strength is proportional to normal stress, but steady state strength is insensitive to normal stress probably because pore pressure approaches the normal stress. As such, CDZ gouge is weak at coseismic rates relative to interseismic creep strength. The potential for sustaining rupture propagation into the CDZ from an adjacent seismic segment is sensitive to the relationship used to extrapolate the critical weakening displacement from experimental to in situ conditions. Rupture propagation from a microseismic patch within the CDZ is unlikely, but sustained propagation from a large earthquake (e.g., Parkfield event) may be possible.

Mechanical stratigraphy and fault-fold interaction, Absaroka thrust sheet, Salt River Range, Wyoming

Abstract A distinct contrast in deformation style from tightly folded and faulted Cambrian through Ordovician strata to broadly folded and faulted Mississippian and younger strata is displayed in the hanging wall of the Absaroka thrust, Salt River Range, Wyoming. Based on style, the Paleozoic section can be divided into two main structural lithic units. Each unit is characterized by a relatively weak, ductile, anisotropic lower section and a relatively strong, brittle, more isotropic upper section. Where the two units were stacked within the thrust sheet, inverted fault–propagation folds formed in the center of each unit, and the overall transition upward from close- to wide-spaced folds and imbricate faults developed in the multilayer. Where the upper unit was isolated, deformation was dominated by imbricate faulting with little associated folding, and inverted fault–propagation folds did not form. Observations illustrate that the mechanical interaction between the two units and boundary conditions imposed on them were significant to defining the deformation response. It is note-worthy that the boundary between the two structural lithic unts, in part defined by the roof thrust of the Stewart passive roof duplex, occurs within a dolostone rather than in overlying shale of the Darby Formation. I propose that the roof thrust and other detachments in dolostone reflect the relatively low frictional strength of dolostone under shallow crustal conditions. Cross-cutting relations suggest that the Absaroka sheet was cut by large imbricate faults during late-stage movement over a major dip ramp. A high deviatoric and low mean compressive stress state just beyond the upper ramp–flat corner promoted break-back imbrication and allowed tectonic thickening of the thrust wedge necessary to reestablish critical taper and for forward translation.

Stress distribution and failure in anisotropic rock near a bend on a weak fault

Large transform faults, thin-skinned thrust faults, and listric normal faults often contain restraining and releasing bends that can alter the state of stress in the adjacent crust during faulting episodes. Crustal rock commonly has anisotropic mechanical properties due to the presence of sedimentary layering, subsidiary fractures and faults, schistosity, or other foliation. We present an analytical solution for the stress distribution in anisotropic rock produced by sliding on a wavy, frictionless surface. The frictionless surface represents the limiting case of a weak fault. The fault shape treated is either a sinusoid, or a periodic array of isolated restraining and releasing bends. The rheological behavior of the rock is that of an incompressible linear viscous fluid with an orthotropic anisotropy characterized by a greater viscosity for shortening and extension than for shear in the principal directions of anisotropy. Our results illustrate that stress and flow associated with a bend in a fault will extend to much greater distances from the fault when the medium is anisotropic. The magnitude of the stress perturbation increases with degree of anisotropy and decreases with radius of curvature of the fault surface. Principal stress directions tend to align parallel to the principal directions of anisotropy except in the immediate region of the fault bend. Mean stress and maximum shear stress magnitudes vary along the fault in a cellular manner, with multiple maxima near a bend. Stress concentrations emanate from the bend and are elongate in directions parallel to the principal directions of anisotropy. For isotropic rock, locations and orientations of shear failure near restraining bends are distinctly different from those near releasing bends. In contrast, with application of an anisotropic failure criterion, as the magnitude of anisotropy increases, the patterns of shear failure at a restraining and a releasing bend become increasingly similar. The model may help explain field observations from dip-slip and strike-slip regimes that indicate a complex stress history and local stress reorientation adjacent to a bend in a fault.