T13A-1901
Time-dependent Brittle Deformation in Darley Dale Sandstone
The characterization of time-dependent brittle rock deformation is fundamental to understanding the long- term evolution and dynamics of the Earth's upper crust. The chemical influence of water promotes time- dependent deformation through stress corrosion cracking that allows rocks to deform at stresses far below their short-term failure strength. Here we report results from a study of time-dependent brittle creep in water- saturated samples of Darley Dale sandstone (initial porosity of 13%). Conventional creep experiments (or 'static fatigue' tests) show that time to failure decreases dramatically with the imposed deviatoric stress. They also suggest the existence of a critical level of damage beyond which localized failure develops. Sample variability results however in significant scattering in the experimental data and numerous tests are needed to clearly define a relation between the strain rate and the applied stress. We show here that stress-stepping experiments provide a means to overcome this problem and that it is possible this way to obtain the strain rate dependence on applied stress with a single test. This allows to study in details the impact of various thermodynamical conditions on brittle creep. The influence of effective stress was investigated in stress-stepping experiments with effective confining pressures of 10, 30 and 50 MPa (whilst maintaining a constant pore fluid pressure of 20 MPa). In addition to the expected purely mechanical influence of an elevated effective stress our results also demonstrate that stress corrosion appears to be inhibited at higher effective stresses. The influence of doubling the pore fluid pressure however, whilst maintaining a constant effective stress, is shown to have no effect on the rate of stress corrosion. We then discuss the results in light of acoustic emission hypocentre location data and optical microscope analysis and use our experimental data to validate proposed macroscopic creep laws. Finally, using an internal electric furnace, we show the influence of temperature on time-dependent brittle deformation.
T13A-1902
Deformation and Failure Properties of Colli Albani Tuff
The Colli Albani are an explosive Quaternary volcanic district, which may pose a threat to the city of Rome, Italy. Recent seismic swarms and hydrothermal activity suggest that the main faults, affecting the volcanic complex are cyclically active. A 350 m scientific borehole was therefore drilled into this volcanic area to elucidate its inner structure for the first time (Mariucci et al., 2008). The main unit drilled, called Tufo Pisolitico di Trigoria unit, is a fine-grained, matrix-supported pyroclastic deposit with rare lithic lava clasts and sparse pumice. Elastic wave velocities (Vinciguerra et al., 2008) revealed how, within the same lithology, the different degree of lithification and presence of clasts can affect significantly physical property values. In this study we investigate the micromechanics leading to failure in Colli Albani tuff. Samples of 38 mm in length and 18 mm in diameter were cored in the vertical direction from a core at the Colli Albani site. Our block of tuff had a nominal connected porosity (measured by water saturation) of 32%. We performed a series of hydrostatic and conventional triaxial compression experiments on water saturated samples in drained conditions with 10 MPa pore pressure at room temperature. Under hydrostatic condition, the onset of grain crushing and pore collapse P* occurred at an effective pressure of 41 MPa. Intense microcracking was observed in the deformed sample which failed by cataclastic flow. Four samples were deformed at effective pressures between 5 and 30 MPa. Shear-enhanced compaction was observed at all pressure conditions. In the stress space, the resulting yield envelope mapped out an elliptical cap comparable to previous observations on porous sandstones and carbonate rocks. In the deformed samples we observed pore collapse and microcracking. A theoretical model is being developed for the micromechanics of compaction and failure in Colli Albani tuff. Results support the interpretation of the mechanisms of deformation monitored at the field scale.
T13A-1903
Friction of Granite and Gabbro at Subseismic Sliprate
To investigate the transition process of the friction between slow sliprate less than 1μm/s, whose coefficient of friction is between 0.6 and 0.85, and high sliprate of about 1 m/s, whose coefficient of friction is less than 0.4, we conducted friction experiments on two kinds of rock (granite and gabbro) at intermediate sliprate ranging between 1 μm/s and 0.1 m/s (subseismic sliprate) under dry and wet conditions. The experiments were conducted using a new rotary-shear friction apparatus made in 2007 at NIED, whose design is based on that of Prof. T. Shimamoto of Hiroshima University [Shimamoto and Tsutsumi, 1994]. A pair of cylindrical specimens of rock samples was in contact with each other applying specific normal stress (σ) and rotating one of the samples fixing the other. And torque (τ) is measured at the fixed sample. Then the contacting surface between two specimens becomes a simulated fault plane sheared at a given velocity (V) and friction (μ) is estimated using σ and τ. A series of experiments were performed at σ of 2.04 MPa and V ranging from 0.74 to 88 mm/s to investigate the velocity- dependence of the frictional behavior. The experiments were done on granite and gabbro in atmospheric environment (dry condition) and a water-saturated open vessel (wet) to evaluate the effects of rock type and dry/wet condition on the frictional behavior at the subseismic sliprates. For all cases, the coefficients of friction decreased from an initial peak friction, μi, to a steady state friction, μss, with increasing displacement up to 100 m. μss was dependent on slip rate, rock type and wet/dry conditions. Increasing V from 0.74 to 88 mm/s, μss decreased from more than 0.8 to less than 0.4 for dry granite and dry gabbro. μss for dry granite was almost stable and located between μss at low and high sliprates. In contrast, μss for dry gabbro oscillated with displacement between the values of μss at low sliprate and that at high sliprates. The results on wet granite and wet gabbro showed that μss at V equal to 0.74 mm/s was less by 0.3 than that for the dry rocks. This suggests that the presence of water decreases μss for the rocks. muss for wet granite was almost independent on sliprate, while μss for wet gabbro oscillated between 0.4 and 0.8 at the sliprates ranging between 0.74 and 88 m/s, although the oscillation frequency varied with sliprates. In this study, we found that these two rocks exhibited velocity- and water-weakening at the subseismic slip rates. Further work is needed to investigate the weakening mechanisms for the rocks and to understand the cause of the fluctuation of μss for gabbro.
T13A-1904
Experimental Investigation of a Transition Between Stick-Slip and Creep as a Function of Slip Velocity, Temperature and Normal Stress
We investigate frictional properties of crystalline rocks to map the transition between stick-slip (velocity-weakening) and stable creep (velocity-strengthening) behavior as a function of: slip velocity, temperature and normal stress. We performed an initial series of direct shear tests on diabase and novaculite for velocities of 10-8 - 3× 10-5 m/s, temperatures of 25-500°C and normal stresses of 1-15 MPa. Analysis of data reveals four basic types of frictional behavior: stick-slip, episodic slow-slip events, quasi-sinusoidal accelerated creep and stable sliding. Episodic acceleration and peak slip velocities progressively decrease through these phases in the above order. The transition between creep and unstable stick-slip strongly depends on all three parameters (strain rate, temperature, normal stress). As temperature and normal stress increase, and forcing velocity decreases, the sample progresses from stable sliding, to sinusoidal accelerated creep, to slow-slip, and then to stick-slip. For the range of applied normal stresses, the stick-creep transition seems to occur around 150-200°C for dry diabase and dry novaculite. Our observations are broadly consistent with predictions of the rate and state friction theory, indicating that lower slip rates and higher normal stresses result in enhanced interlocking of contacts on a frictional interface. Our data show that elevated temperatures give rise to the same effect, suggesting a thermally-activated nature of the asperity contacts. Surprisingly, we did not observe a high-temperature transition from stick-slip back to stable sliding in novaculite (purely silicic lithology), even at temperatures as high as 500°C in our dry experiments. Such a transition is widely believed to be responsible for the brittle-ductile boundary defining the bottom of the seismogenic layer. At temperatures of 400-500°C we did observe some minor stress relaxation during hold tests, suggesting that the stick-creep transition is also rate dependent. Our observations may highlight the role of both water and strain rate on the brittle-ductile transition at the bottom of the seismogenic layer. In particular, the velocity weakening behavior may extend deeper than typically thought, at least in the dry middle crust given elevated strain rates. We also point out potential similarities between the periodic accelerated creep observed in our experiments at the boundary between the stick-sip and stable sliding regimes, and episodic slow-slip events reported near the velocity-neutral transition in a number of subduction zones. The slip rates observed during slow-slip events in our experiments have the same order of magnitude (10-8 m/s) as the slow-slip events in the Cascadia subduction zone.
T13A-1905
Time-Dependent Strengthening Rates in Simulated Fault Gouge and Implications for Fault Zone Processes.
The frictional behavior of tectonic faults is ultimately a product of asperity contact processes, fault zone fabric, and granular processes in gouge zones. Contact area evolution likely exhibits significant control on overall fault stability by influencing 1) healing during the interseismic period, 2) rate-dependent friction, and 3) restrengthening at short time scales of dynamic rupture. We report on a series of slide-hold-slide tests conducted on simulated fault gouges, to explore the role of gouge mineralogy on the evolution of contact area and associated fault healing at intermediate time scales. We conducted friction experiments in a double-direct shear configuration, under room conditions using four simulated fault gouges: talc (particles < 125μm), kaolinite (1-4μm), illite shale (< 106μm), and feldspar (andesine) (<106μm). Layers of simulated gouge were sheared at load point velocities of 1, 10, and 100 μm/s under constant normal stress (20 MPa) between rough rigid forcing blocks to a shear strain of ~25, and then subjected to slide-hold-slide tests with holds ranging from 1 to 1000 seconds, with 0.5 mm of slip between each hold. Talc shows little to no healing (Δμ < 0.001) over all hold times and all loading rates, whereas kaolinite (Δμ = 0.0025 - 0.025), andesine (Δμ = 0.002 - 0.021) and illite shale (Δμµ = 0 - 0.011) all exhibit healing. Illite shale shows increasing healing in tests that were run at increasing load point velocity during shear before and after the holds. All gouges exhibit log-linear healing rates over all velocities and hold times, with the exception of feldspar, which deviates from this relationship at hold times > 100s. Layer compaction during holds increases with hold time and with the load point velocity. Compaction ranges from < 1 μm at 1 μm/s to 3.5 μm at 100 μm/s. Talc compacts the least and feldspar compacts the most. High compaction at high velocities is attributable to gouge dilation during shear. Low compaction is associated with the weak mineral talc, suggesting that most of the porosity reduction in weak minerals occurs under initial loading and that large porosity changes are unlikely. The low strength of talc likely facilitates saturation of contact area even at low effective stress, as evidenced by extremely low rates of healing and compaction. Because velocity-weakening frictional behavior (a necessary condition for unstable slip) is predicated on the disruption of time-dependent strengthening processes, the presence of talc or similarly weak minerals would promote stable sliding in a fault zone. Ongoing work will consider water saturated conditions and the possible effects of compaction and porosity reduction on permeability and pore pressure evolution.
T13A-1906
Strong Velocity-Weakening of Nanograins at High Slip-Rates
It has been observed that slip localization zones in some experimental and natural faults consist of crystalline or amorphous nanograins of different minerals. Prolonged grinding of silicate rocks (e.g., quartz rock and granite) is known to produce amorphous silica nanograins and mechanical properties of the material (especially under wet condition) have been attributed to a mechanism of fault weakening. Also, recent high- velocity friction tests on carbonate rocks showed that faults can be weakened by thermal decomposition of calcite into nanograins of lime and carbon dioxide and the lubrication effect of the nanograins would be critical for the fault weakening. However, mechanical behavior(s) and friction mechanism(s) of fault slip zones with nanograins, especially at high slip-rates, are still poorly understood, despite their potential importance to the understanding of seismic faulting. In this contribution, we show you our experimental results indicating velocity-weakening of nanograins (probably caused by still unknown mechanical behaviors of nanograins) rather than by temperature-related weakening behavior. In our high-velocity friction tests on Carrara marble at seismic slip-rates, we have tried to "cool" the simulated fault with liquid nitrogen and compressed air during frictional sliding, and found, in the simulated fault coated with nanopowders of lime (CaO) formed by thermal decomposition, no correlation between friction and temperature measured with thermocouples (i.e., no temperature-related weakening behavior), although strong "velocity-weakening" behavior appeared. The observation was confirmed by another experiment: from (1) the calculated "maximum" sliding surface temperature [Carslaw and Jaeger, 1959] using the mechanical data, with an assumption of strong slip localization into a very thin layer, and (2) the measured temperature with thermocouples at a place just below the sliding surface and close to the periphery of the specimen, it was found that, at the "constant" velocity of 1.3 m/s, friction at ~420°C was very low (~0.15 in terms of friction coefficient) and was almost the same as that at 1260°C. On the other hand, interestingly, we could observe friction at the deceleration stage (when sliding velocity decreases gradually from 1.3 m/s to zero) was recovered notably to a higher value with decreasing velocity (thus velocity-weakening), although temperature change was < 100°C. Also, this weakening may not be explained by flash heating weakening, because of extremely small contact sizes of nanograins. From this previously unexpected behavior, now we are open to a possibility that friction mechanism(s) in nanograins may be different from those in ordinary grains (> μm) and we wonder what kind of friction mechanism operates in the nanograins. To get an insight into the questions, additional friction experiments on nanograins of different minerals are being conducted, and the experimental results as well as TEM observation results will be presented at the meeting.
T13A-1907
Analyzing Fault Gouge Under Seismic Conditions in Laboratory Experiments
The slip along a fault zone during an earthquake is associated with dense granular flow of the fault gouge. The gouge is a mixture of powder formed during prior slip as well as the dynamic rupture in the process zone of the propagating earthquake. Experimental tests of dense granular flow are usually conducted under low to moderate shear-rates and under low to vanishing normal stresses, and commonly without confinement. These experimental conditions do not match the conditions of natural earthquakes: slip of a few meters; normal stress of tens to hundreds MPa; slip velocity of ~1 m/s; rise time < 1 sec; and elevated, undrained fluid pressure. We build an apparatus for testing the mechanical behavior of fault gouge (and other dense granular materials) under these natural conditions. The apparatus has the following capabilities: (1) Control of fluid pore pressure in the gouge by leak-proof design for dry, wet, or partially saturated experiments, and control of pore pressure in wet experiments; (2) Continuously variable speed control 0.01- 1.5 m/s: (3) Cumulative slip up to 10 m; (4) Normal stress up to 30 MPa; (5) fast step-loading (short rise- time); and (6) Testing either gouge or solid rocks. We use this instrument to test weakening mechanisms during earthquake slip. Preliminary experimental results will be presented.
T13A-1908
High Shear Strain Behavior of Synthetic Muscovite Fault Gouges Under Hydrothermal Conditions
Phyllosilicates are common constituents of mid-crustal fault zones and are widely believed to exert a strong influence on the strength of fault rocks, in particular around the brittle-ductile transition. The present study aims to experimentally determine the mechanical strength and frictional behaviour of muscovite gouge, in order to establish whether the presence of muscovite might contribute to the long term weakness often inferred to hold for large scale crustal fault zones. Rotary shear experiments in the temperature range 20-700°C, at 100 MPa fluid pressure, have been performed on synthetic muscovite gouges (average grain size 13 μm, < 10% quartz). The effects of sliding velocity, normal stress and shear strain on the frictional behavior of the muscovite gouges have been studied. Microstructural analysis of the experimentally deformed samples has been done using SEM and XED in order to obtain insight in the operating microphysical processes. The mechanical behavior under the conditions applied is independent of sliding velocity in the range 0.03-3.8 μm/s. The results show frictional behavior in samples deformed in velocity stepping experiments under 100 MPa normal stress with coefficients of friction of ~0.34 at 20°C, in agreement with previous data, gradually increasing to values of ~0.55 at 300°C, remaining around this value up to 600°C. During shear, samples compact continuously (15-20% at γ~25) at low temperatures (20-300°C). At higher temperatures, similar compaction is preceded by initial dilatation of up to 10%. The low temperature microstructures (< 300°C) show strong grain size reduction to form a well- foliated muscovite matrix (grain size < 2 μm) enclosing quartz and widespread muscovite porphyroclasts (~10 μm). Numerous shear bands in the Riedel and Y-shear orientations are present. Towards higher temperature, the microstructures are increasingly characterized by dense, elongate lenses (20-50 by 10 μm) of fine, folded and kinked grains (5-10 μm), separated by an anastomosing network of zones of ultra-fine (<< 1 μm) grain size. At 700°C the microstructure appears altered, possibly due to chemical alteration or partial melting. The muscovite gouge is inferred to have been deformed by cataclastic flow at low temperatures, while intra- crystalline plastic deformation and healing by diffusive mass transfer processes become increasingly important towards higher temperatures. The results of the experiments can be extrapolated to natural fault zones by constructing crustal strength profiles. The experimental results obtained at higher temperatures correspond to greater depths, thus indicating that muscovite gouge becomes stronger with depth. Based on this, it is questionable whether the presence of muscovite can account for the weakness inferred from geophysical and geological observations for large scale crustal fault zones, such as the San Andreas fault zone.
T13A-1909
Talc Formation Along Continental Low-Angle Normal Faults: Growth, Deformation, and Mechanical Consequences
Fault zones which slip under low-resolved shear stress, i.e. weak faults, represent a mechanical problem. Low-angle normal faults in particular have been the subject of much debate since they have been widely documented in the extending continental crust, yet no moderate to large earthquakes have been positively identified along these structures. Here we present a study of a regional low-angle normal fault, the Zuccale fault on the Island of Elba (Central Italy), which has a displacement of 6-8 km and carries fault rocks exhumed from between 3-8 km depth. The key feature of the fault zone is a mesoscopically ductile fault core which is sandwiched between a hanging-wall and footwall where the deformation is exclusively brittle. A significant amount of deformation was accommodated within a pervasively foliated basal horizon within the fault core, containing calcite-mylonites and talc-phyllonites. Field, optical microscope, XRD, SEM and TEM studies show that the phyllonitic foliation mostly consists of talc lamellae, 50-100 nanometres thick. Talc was formed by fracturing of dolomite and interaction with silica-bearing hydrothermal fluids, and often crystallizes within strain shadow regions around carbonate porphyroclasts, indicating a syn-tectonic origin. With increasing deformation, the amount of dolomite within these fault rocks decreases from 80-90% to zero and the amount of talc increases from 5-10% to 60-70%. At the same time, talc lamellae rotate into parallelism with the margins of the fault core, and ultimately form an interconnected network. TEM analysis shows that the talc lamellae contain numerous interlayer delaminations, and that sliding, translation and/or rotation along 001 foliation planes is common, strongly suggesting that deformation was accommodated by frictional slip. Kinking and interlayer bending of talc are rare. Talc has one of the lowest friction coefficients of any mineral (μ <0.2), and its occurrence within the Zuccale fault as continuous and interconnected layers may have fundamental mechanical consequences. The derivation of talc from dolostones suggests that this weakening process may be applicable to many faults within the continental crust, provided that silica-bearing fluids are available. Finally, the syn-tectonic production of stable sliding minerals such as talc may provide an explanation for the lack of large earthquakes on low-angle normal faults worldwide.
T13A-1910
What Controls Creep on the San Andreas Fault at the SAFOD Drillhole?
There is general agreement that phyllosilicates contribute to a mechanically weak San Andreas Fault, but there are contrasting opinions as to which minerals are responsible for low frictional coefficients and fault creep behavior along the active section that is sampled by the SAFOD (San Andreas Fault Observatory at Depth) drillhole. One recent hypothesis proposes that talc and serpentine are responsible for the apparent weakness of the fault rock (Moore and Rymer 2007). This hypothesis is based on the occurrence of these minerals both in outcrop and at depth in the SAFOD borehole. However, in our study of thin film mineral phases from mineralized slip planes at ca. 3.0-3.4 km measured depths, we have not been able to identify either talc or serpentine on displacement surfaces. Instead, we find localized precipitation of hydrous mixed- layered clay minerals that grow preferentially on fractured clasts with polished and occasionally slickensided displacement surfaces. This localization is of particular interest due to the large surfaces areas, cation exchange possibilities, related changes in hydration state, and strong orientation produced by substrate- controlled growth. We suggest that the majority of slow creep occurs along these heterogeneously distributed, micron-scale thin films (or nano-coatings), accommodated by a combination of 1) slip along particles surfaces, 2) displacement along hydrated interlayers and 3) intracrystalline deformation of the clay lattice, pos¬sibly associated with repeated nucleation and growth. In our study, we emphasize that it is the nature of minerals occurring along slip surfaces of the San Andreas Fault that contribute most to the creep behavior, rather than matrix minerals that do not otherwise appear to have a genetic link to the faulting process. However, more detailed work at the recently cored foliated gouge from the actively creeping part of the fault will be necessary in the future.
T13A-1911
Helium isotopes in matrix pore fluids from SAFOD drill core samples suggest mantle fluids cannot be responsible for fault weakening
To quantify fluid flow in the San Andreas Fault (SAF) (and since direct fracture fluid sampling of the fault zone was not available), we have adapted a method to extract rare gases from matrix fluids of whole rocks by diffusion. Helium was measured on drill core samples obtained from 3054 m (Pacific Plate) to 3990 m (North American Plate) through the San Andreas Fault Zone (SAFZ) ~3300 m during SAFOD Phases I (2004), II (2005), III (2007). Samples were typically collected as 2.54 cm diameter subcores drilled into the ends of the cores, or from the core catcher and drillcore fragments within <2hr after core recovery. The samples were placed into ultra high vacuum stainless steel containers, flushed with ultra high purity nitrogen and immediately evacuated. Helium isotopes of the extracted matrix pore fluids and the solid matrix were determined by mass spectrometery at LDEO. Matrix porefluid 3He/4He ratios are ~0.4 – 0.5xRa (Ra: atmospheric 3He/4He = 1.384 x 10-6) in the Pacific Plate, increasing toward the SAFZ, while pore fluids in the North American Plate have a 3He/4He range of 0.7-0.9Ra, increasing away from the SAFZ (consistent with results from mud gas samples (Wiersberg and Erzinger, 2007) and direct fluid samples (Kennedy et al., 2007)). Helium isotope ratios of the solid matrix are less than 0.06Ra across the SAF in samples from both the North American and the Pacific plates, thereby excluding the host matrix as source for the enhanced isotopic signature. If the system is assumed to be in steady state, then the flux of mantle helium must be from the North American Plate to the Pacific plate. The steeper gradient in the Pacific Plate relative to the North American plate is consistent with a porosity corrected effective diffusivity. The source for this mantle helium in the North American Plate is likely related to a low crustal conductivity zone identified by magnetotelluric signals (Becken et al., 2008) that provides a channel for transport of mantle helium within brittle crust under high strain rates (Kennedy et al., 2007). The helium isotope gradients suggest that fault weakening by mantle-derived fluid pressure is unlikely. More likely, mantle fluids "bleed" into the North American plate below seismogenic depths and are transported across the fault by nonseismic, diffusive processes.
T13A-1912
Geochemical Characteristics of Fluids in the Chelungpu fault of Taiwan and its Implications
The Chi-Chi earthquake (Mw 7.6) was characterized by high rupturing and slip velocity with large displacement in the North suggested that the fault lubrication may be occurred during co-seismic period. The characteristics of fluid, thus, involved in the Chelungpu fault system are important for realizing the mechanism and behavior of the fault-fluid interaction. We analyzed geochemical characteristics, such as hydrogen and oxygen isotopes, physical properties and ionic concentrations of fluid samples retrieved in various depths along the boreholes the Hole-A and Hole-B of Taiwan Chelungpu fault Drilling Project (TCDP) to trace fluid sources. The results show that the source of fluid in the Hole-B was mainly from tap water, while there were two probable sources in the Hole-A owing to abrupt shift of ionic concentrations at depth of 200-300 m. The shallower fluid with lower ionic concentrations may be from the leakage of groundwater above the depth of 300 m. However, the deeper fluid may be the thermal water from the Kueichulin Formation because of high ionic concentrations, especially HCO3-, and higher oxygen isotope. Two sources of fluid in the Hole-A are representative of the fluid systems in the hanging wall and foot wall, respectively. The geochemical characteristics of fluid in the Hole-A imply that the fault zone served as a barrier in the inter-seismic period, resulting in distinctly different fluid between the hanging wall and foot wall. The frequent occurrences and the distributions of calcite vein provide the evidence for the upwelling of HCO3--rich fluid from the Kueichulin Formation and indicate that the fault served as fluid conduit during seismic faulting and allowed the fluid flow across the fault zone to precipitate calcite veins in fractures of hanging wall. This study, therefore, provides a mechanism to depict the fluid behavior in the Chelungpu fault system during different stages of fault development.
T13A-1913
Pore-pressure variations: A possible explanation for strain-rate-dependent faulting with wet clay
Localized shear zones (i.e., faults) develop in extensional analog models with wet clay only after a significant amount of extension has occurred. The amount of strain needed for localization, as well as the proportion of total extension that is accommodated by shear zones, depend on the strain rate. With lower strain rates, less strain is needed before localized shear zones appear and more of the total extension is accommodated by visible faults. Thin sections from extensional analog models show that: (1) clay minerals in unfaulted domains are aligned horizontally; (2) clay minerals in shear zones have a strong preferred orientation with clay minerals aligned slightly oblique to the shear-zone boundary; and (3) shear zones are extremely narrow, i.e. 30 microns wide. These observations lead us to consider the possible role of pore-pressure changes in the development of shear zones in clay models. The reorientation of clay minerals in the shear zones would produce a period of dilation and related localized decrease in pore-pressure, thereby inhibiting localized deformation. Experiments with lower strain rates would have more time to restore pore pressures in shear zones through permeable flow and, thus, promote strain localization. We use a 2D numerical model to investigate how pore-pressure changes and permeable flow would affect fault growth in extensional models with wet clay. As in previous numerical studies, brittle strain localization is achieved through reduction of cohesion and friction with increasing strain. Initially high pore pressures are reduced with increasing strain and permeable flow leads to a diffusional restoration of pore pressures. With these numerical setup we can reproduce the above-mentioned strain-rate dependency of (1) the onset of faulting and (2) the portion of extension accommodated along discrete faults. We propose that strain-related, dilatational, pore-pressure drops in wet clay might cause the observed rate-dependent faulting.
T13A-1914
Permeability evolution of fault zone in high-velocity frictional tests
Fluid transport property of fault zone is an important parameter to understand dynamic weakening and fault strengthening processes during large earthquakes. It also influences on the coseismic pore pressure response near surface. Permeability is one of the important transport properties, and in laboratory experiments, permeability changes under very low strain and low slip rates were previously reported. However, these parameters may not adaptable for the analysis of coseismic process at a high slip rate. Therefore, we estimated a dynamic permeability change from high-velocity friction tests. Berea Sandstone (initial porosity =23%), Inada Granite (1.5%) and India Sandstone (14%) were chosen as test specimen. Both high velocity friction test and permeability measurement were carried out using cylindrical samples with 25 mm of diameter and 20 to 40 mm length. High velocity friction tests were performed at room temperature and 1 MPa of normal stress. We sheared samples around 5 m at 1.3 m/s of an average slip velocity. Friction coefficients in all tests were suddenly increased at the initiation of sliding (Berea Sandstone: 1.03, Inada Granite: 1.15), then they decreased gradually and achieved steady state level. All samples were shortened continuously during sliding due to formation of gouge and compaction of samples. Intrinsic permeability was measured at high confining pressure condition up to 120 MPa, and distilled water and nitrogen gas were used as a pore fluid. Permeability was evaluated using the steady flow method and the transient pulse method. Permeability of Berea Sandstone, Inada Granite and India Sandstone before friction tests showed 1×10-13 ~ 3×10-14 m2, 2×10-19 ~ 2×10-20 m2, and 4×10-17 ~ 2×10-17 m2 respectively at 5 to 100 MPa of effective pressure. Permeability of Berea Sandstone after friction tests showed 5×10-14 ~ 1×10-14 m2, which are smaller than that before friction test, though permeability of India Granite and India Sandstone after friction increased relatively (Inada: 2×10-18 ~ 4×10-19 m2, India: 8×10-17 ~ 4×10-17 m2). The effective pressure sensitivity of the permeability did not change after experiments. Newly formation of fine grained gouge and micro-cracks were observed at around slip surfaces in all rocks. Micro-XCT observation showed that the crack density in Berea sandstone after friction test was higher than that in Granite. Our results suggest that the evolution behavior of Berea Sandstone and Inada Granite is controlled by the relative relationship among bulk, gouge and crack permeabilities. Permeability reduction of Berea Sandstone could be caused by the evolution of fine grained gouge layer that might be lower permeability than bulk permeability. On the contrary, a micro-crack enhancement might mainly affect the increase of permeability in Inada Granite. Thermal pressurization process is thought to be effective in the fault zone originated from granitic rock, because granitic rock and its correlated fault rocks show low permeability value (Mizoguchi et al., 2007 in JSG). However, our results indicate that thermal pressurization could be ceased by the evolution of permeability which associates with the abrupt recovery of the fault strength due to rapid dissipation of the elevated pore pressure.
T13A-1915
Fault Damage Zone Permeability in Crystalline Rocks from Combined Field and Laboratory Measurements
In nature, permeability is enhanced in the damage zone of faults, where fracturing occurs on a wide range of scales. Here we analyze the contribution of microfracture damage on the permeability of faults that cut through low porosity, crystalline rocks by combining field and laboratory measurements. Microfracture densities surrounding strike-slip faults with well-constrained displacements ranging over 3 orders of magnitude (~0.12 m – 5000 m) have been analyzed. The faults studied are excellently exposed within the Atacama Fault Zone, where exhumation from 6-10 km has occurred. Microfractures in the form of fluid inclusion planes (FIPs) show a log-linear decrease in fracture density with perpendicular distance from the fault core. Damage zone widths defined by the density of FIPs scale with fault displacement, and an empirical relationship for microfracture density distribution throughout the damage zone with displacement is derived. Damage zone rocks will have experienced differential stresses that were less than, but some proportion of, the failure stress. As such, permeability data from progressively loaded, initially intact laboratory samples, in the pre-failure region provide useful insights into fluid flow properties of various parts of the damage zone. The permeability evolution of initially intact crystalline rocks under increasing differential load leading to macroscopic failure was determined at water pore pressures of 50 MPa and effective pressure of 10 MPa. Permeability is seen to increase by up to, and over, two orders of magnitude prior to macroscopic failure. Further experiments were stopped at various points in the loading history in order to correlate microfracture density within the samples with permeability. By combining empirical relationships determined from both quantitative fieldwork and experiments we present a model that allows microfracture permeability distribution throughout the damage zone to be determined as function of increasing fault displacement.
T13A-1916
Fluid Overpressure and Earthquake Triggering in Faulted Evaporitic Sequences
The mainshocks of the 1997 Umbria-Marche seismic sequence (M above 5) nucleated at about 6 km depth, within the Triassic Evaporites (TE), made of interbedded anhydrites and dolostones. Previous studies suggest that the time-space evolution of the aftershock sequence was driven by the coseismic release of high-pressure fluids trapped within the TE. In order to understand whether CO2 fluid overpressure can be maintained up to the coseismic period, and trigger earthquake nucleation, we modelled fluid flow through a mature fault zone within the TE. Our model's parameters are the structure and permeability (k) of the major fault zone, as inferred from the integration of field observations and k laboratory data. The fault zone architecture is given by a distinct fault core, up to 3 meters thick, of very fine-grained fault gouge and cataclasites, surrounded by a geometrically complex damage zone (up to few tens of meters wide). The damage zone is characterized by adjacent zones of heavily- (dolostones) and non-fractured rocks (foliated anhydrites). The permeability of the fault core is inferred to be relatively low (k < 10E-18 m2), due to the presence of fine grained fault rocks. Where the connectivity of the fractured dolostone layers is high, the permeability of the damage zone is high (> 10E-17 m2) and controlled by the development of mesoscale fracture patterns. The permeability of the damage zone can be assumed to be as low as the values measured during our lab experiments on anhydrite rocks (k = 10E-17 - 10E-20 m2), where we observed the foliated Ca-sulphate rocks, with no mesoscale fracture patterns, completely surround the fractured blocks of dolostones. Our model results show that, during the seismic cycle, within portion of the fault zone where the fractured dolostone layers have a low connectivity, the lateral fluid flux across the fault zone is always lower than the vertical parallel fluid flux. We propose that overpressured patches cause dramatic fault weakening and trigger earthquake nucleation.
T13A-1917
The Role of Fluids in Triggering Earthquakes: Observations From Reservoir Induced Earthquakes
We relocate micro-earthquakes induced by the Açu reservoir in Brazil and observe seismicity migration consistent with pore-pressure diffusion. Reservoir induced seismicity provides a natural laboratory in which to investigate the spatio-temporal evolution and triggering of earthquakes caused by fluid diffusion. Our results can be used to quantify and model pore-pressure diffusion, and to investigate the role of fluids in triggering earthquakes in other tectonic settings. Do Nascimento et al. (2004) recorded and located 267 earthquakes (M ≤ 2.1) beneath the Açu reservoir between 1994-1997. The seismicity increased several months following annual water level peaks, implying that fluid pressure diffusion is the principal triggering mechanism. The small station spacing and very low-attenuation, Precambrian basement rock enabled location of the earthquakes with uncertainties of only a few hundred meters. The earthquakes were located in three clusters, and the time delay to activation of each cluster increased with the depth of the cluster. The location uncertainties were too large to resolve any seismicity migration within a single cluster. We relocate these earthquakes using waveform cross-correlation to obtain groups of similar events, and improve picks. We also apply time-corrections for the stations with poor absolute timing. We begin by relocating 173 earthquakes from the largest cluster using only S-P times because of the poor timing. We obtain more tightly clustered locations with uncertainties on the order of 10-50 m (sub-sample pick accuracy in 200 samples/s data). We observe temporal migration of the earthquakes within the cluster, both along strike, and to increasing depth. We observe a seismicity migration rate between 5 and 15 m/day. The rate is highest during the time of peak seismicity rate, and there is some suggestion that the rate decreases with increasing depth. We then include the best constrained time-corrections, and apply our methods to the entire data set.
T13A-1918
Experimental Observations of Permeability Enhancements by Dynamic Stresses
Shaking produced by seismic faulting often triggers distant and nearby earthquakes. Seismic waves are also known to increase stream flow and spring discharge and enhance oil production; in some cases tripling the effective permeability of the natural system. These observations have been attributed to shaking-induced increases in permeability. However, the underlying mechanism is poorly understood. Here we present experimental evidence of permeability enhancement in fractured rock samples subject to dynamic stresses. We use Berea sandstone samples under triaxial stresses with confining pressure of 9 MPa and 20 MPa of normal stress. We flow deionized water through a fracture produced in-situ and find that oscillations in pore pressure, of 20 second period and 120 second duration, induce transient increases in permeability. Permeability increases scale with the amplitude of pore pressure oscillations. The maximum value of the permeability enhancement is 5x10-16 m2 over a background permeability of 1x10-15 m2. After the oscillations, permeability recovers as the inverse square root of time. The recovery indicates a reversible mechanism, such as clogging/unclogging of fractures, as opposed to an irreversible one, like micro-fracturing. Our result has clear consequences for earthquake triggering mediated by permeability enhancement. Moreover, our data point at the feasibility of dynamically controlling permeability of fractured systems with applications to hydrology and oil reservoir engineering.
T13A-1919
Fluid Flow and Rock Deformation: Evolution of Permeability During Experimental Dehydration of Gypsum
The rise of pore fluid pressure in rocks produces a reduction of effective pressure and results in mechanical embrittlement in rocks under tectonic stress. The two main mechanisms by which pore fluid pressure may rise are rapid compaction of the pore space hosting the fluid or the release of fluids by metamorphic reactions when Clapeyron slopes are positive. In both cases, the permeability of the affected and surrounding rocks will determine whether available drainage will be sufficient to prevent fluid overpressures and subsequent fracturing. The role of dehydration reactions triggering brittle events in stressed rocks has been targeted before in the laboratory. However, one of the key elements in laboratory experiments is to constrain the evolution of drainage during dehydration reactions, which is a function of evolving permeability and compaction. During dehydration, water is continuously being evolved from the rock and the two techniques most widely used for measuring permeability, the flow through method and the pulse transient technique, are not ideal methods to measure permeability reliably with ongoing reaction. We use the pore fluid pressure oscillation technique to measure permeability while the dehydration reaction is under progress in specimens of gypsum. Gypsum, when present, is important in the deformation of foreland fold and thrust belts, but we mainly use it as an analog material to study processes that occur at greater depths in subduction zones in relation to the generation of seismicity. As an experimental material, Volterra gypsum has several features of interest. It has low porosity (0.5 %) and very low permeability (10-18 to 10-20 m-2 depending on effective pressure) thus making permeability very sensitive to rise in pore fluid pressure with small degree of reaction. At standard laboratory strain rates (10-3 to 10-9 s-1) it deforms ductilely by cataclastic flow and kinking, but can also deform by diffusion creep, both processes affecting porosity and permeability. The experiments presented here were run at constant confining and fluid pressure of 100/40 MPa and at constant temperature. At this effective pressure (60 MPa) the starting material has an initial permeability of 2 x 10-20 m-2 which increases progressively to 2 x 10-17 m-2 after full reaction. The variation of three orders of magnitude in permeability is simultaneous to partial compaction of the porosity generated by the reaction.
T13A-1920
Weakening Process Due To Phyllosilicate Dehydration And Dehydroxylation Reactions And Its Possibility To Trigger Earthquakes
We experimentally simulated the dewatering reactions from the phylloscilicate minerals to weaken the fault due to the abnormal pore pressure generation, and discussed those possibilities to trigger earthquakes. For the experiment, the gouge samples, Na-montmorillonite, kaolinite and antigorite were subjected to sliding with heating at 9.6° C min-1 to 500° C and 800° C under 80 MPa of confining pressure. Significant frictional-strength reductions were found in the Na-montmorillonite gouge due to dehydration and in the kaolinite and antigorite gouges due to dehydroxylation when we controlled the drainage condition surrounding the gouge zone as to be "undrained". However, frictional strength continued to increase in the fully drained condition that allowed released water to move outside the gouge. Frictional strengths under the two drainage conditions were compared during dehydration and dehydroxylation reactions to estimate the pore pressure generated in the gouge zone under the undrained condition. The 6-8 wt.% of dehydroxylated water generated from the kaolinite caused a rapid increase in pore pressure to 68 MPa at 500° C. The dehydroxylation of antigorite induced weakening gradually from 500° C prior to the rapid reduction appeared at 620° C, corresponding to ca. 70 MPa of pore pressure. In contrast, in the Na-montmorillonite gouge, the pore pressure gradually increased to 72 MPa as the temperature rose to 500° C. Especially on a coseismic stage, these results suggest that dehydration/dehydroxylation could advance the frictional sliding acceleration induced by frictional heating, even if the fault is in a dry condition so long as the fault rock contains the minerals having water in their crystals, such as phyllosilicates.
T13A-1921
Frictional Melt: Fault Lubrication or Brake?
Slip weakening is an essential process for large earthquakes to occur. Frictional melting has been known to be one of the effective slip weakening mechanisms, although there are some suggestions that frictional melt may strengthen fault during seismic slip. Our high-velocity friction experiments on a siltstone imply that frictional melt may act as a brake rather than lubricant. In the high-velocity rotary shear tests on siltstone at slip rates of 1.14-1.18 m/s and normal stresses of 9.8- 20.9 MPa, three cycles of strengthening and weakening occur before the simulated faults exhibit a steady- state friction with an average friction coefficient (μss) of 0.27. After the final weakening, the simulated fault zone consists of a molten layer (0.30-0.75 mm thick) mantled by ¡®damage' layer (0.1-0.2 mm thick). The microstructures of the fault zone at the second peak friction (μp = 0.66) after the second strengthening but before the second weakening) include fragments of melt mixed with gouge. In contrast, the fault zone immediately after the second weakening (before the final strengthening) consists of a continuous molten layer (0.2-0.6 mm thick) with a high fraction of clasts. At the final strengthening to the third peak friction (μp = 0.7-0.8), the molten layer is fragmented and mixed with gouge again. The surface temperature of the fault zone measured by a radiation thermometer is 650-700°C before reaching the final peak friction. In contrast, the measured temperature of the fault zone at the final peak friction is 800-850°C. High-velocity friction of gabbro and tonalite also shows similar behavior to that of siltstone. However, the high- velocity friction behavior of peridotite is somewhat different from that of siltstone, tonalite and gabbro, in that it does not involve a significant strengthening except first, instant strengthening (presumably due to lower viscosity of melt). Our experimental results from siltstone indicate that a continuous molten layer forms at the initial stage and that fault strengthening due to high viscosity of melt can result in the breakage of the molten layer. At least two cycles of the molten layer fragmentation before the final stress drop suggest that frictional melt may arrest further slip unless melt viscosity is much lowered with a significant rise in temperature and decrease in the fraction of clasts.
T13A-1922
A Model for Flash Melting at Asperity Contacts as a mechanism for Fault Weakening at High Slip Rates
The processes that influence the strength of mature fault zones remain a subject of ongoing debate. Thermal pressurization, acoustic fluidization, and elastohydrodynamic lubrication have all been proposed as possible mechanisms for fault weakening. We focus on flash-melting at asperity contacts as a potential source for weakening along mature faults. Motivated by the results of recent high-velocity friction experiments on multi and mono-minerallic sliding systems, we model the evolution of frictional and viscous resistance as melt layers on asperity contacts form, grow, and slide out of existence. For each composition we make use of available thermodynamic and physical properties and account approximately for the effects of phase transitions during heating. Our elementary model makes a priori predictions for the effective friction coefficient μ as high (melting) temperatures and confining pressures are attained at microscopic asperities. The goal is to assess whether melting at these contacts is a viable explanation for experimentally observed changes in strength. Our model predicts that changes in μ are strongly controlled by the ratio of slip rate V to the critical value required for the onset of weakening Vw, and by the Stefan number S, which is the ratio of the latent heat of fusion to the sensible heat required to raise the temperature from ambient levels. The melt viscosity itself is found to not be a significant factor for the range of parameter values that most likely dominates the sliding behavior. This somewhat counter-intuitive result arises because the average melt thickness and shear resistance are both proportional to the viscosity, but the latter is also inversely proportional to the former so viscosity drops out of the problem. Despite its simplicity, our model predicts frictional behavior that is comparable to available experimental results and may be easily implemented within more complex models dedicated to describing the dynamic behavior of mature fault systems.
T13A-1923
Thermo-Chemical Pressurization of Fault Gouges During Coseismic Slip
This work deals with thermo-hydro-mechanical couplings within fault gouges during coseismic slip, including effects of possible thermal dehydration of hydrous minerals. The framework of thermal pressurization of pore fluid is extend to include the dehydration effect as a source term for pore pressure and a sink for temperature. The dehydration kinetics is modelled by a first order reaction rate which is bounded by the ratio between heating rate and the reaction enthalpy variation. We first solve analytically the equations in the case of no fluid and heat transport and with a constant reaction rate. It shows that the dehydration reaction, if the rate constant is of the order of 1~s-1, induces a pore pressure increase that can grow beyond the normal stress apllied on the fault. At the same time, the temperature slightly decreases as the reaction progresses. If the kinetics is calculated from the ratio between frictional heating and enthalpy variation, the temperature is kept constant during the reaction, and the pore pressure increases asymptotically to the normal stress. This corresponds to a transient equilibrium where the pressurization is maintained by the mineral reaction. Then we model the phenomenon more precisely by taking into account transport of fluid and heat and we use an Arrhenius law to calculate the rate constant as function of temperature. The overall behaviour of the system is characterized by a sudden increase of pore pressure and an almost constant temperature as the reaction starts. The parameters values are then discussed, showing that overpressures can occur for low reaction temperatures, low enthalpy change, deep faults and/or thick slipping zones. It shows that dehydration is an effective mechanism for delaying or preventing melting during coseismic slip, if the mass fraction of released water is larger than ~ 1%.
T13A-1924
Fault deformation and movement under wide range of slip rates in boundary-value problem simulations
We implement various frictional constitutive laws for fault deformation and movement in the framework of the extended finite element method. Slip rates considered span from microns per second characteristic of fault creep, to meters per second characteristic of seismic slip speeds. For slow slip rates the frictional law is characterized by the regularized Dieterich-Ruina slowness law; for fast slip rates the frictional shear strength varies with the normal stress raised to the power 1/4, in accordance with melt lubrication theory. The extended finite element method allows the fault to pass through the interior of the finite elements, thus resolving the intense deformation without having to align the fault on the element sides. Casting the frictional constitutive laws in a boundary-value problem allows the investigation of the effect of variable coefficient of friction on fault rupture and propagation under traction, displacement, and mixed traction/displacement boundary conditions. In addition to the variable coefficient of friction models, we also implement a slip- weakening model within the extended finite element framework. We present two-dimensional plane strain simulations showing effects of slip rates on fault propagation and slip distributions under various traction and displacement boundary conditions.
T13A-1925
How to Study Earthquake Triggering and Stress Field Evolution by new FEM
The change in the Coulomb failure stress is widely used to study the aftershock triggering and stress transfer of large earthquakes. The change in the Coulomb failure stress is usually applied to predict the regions where seismic risk increases or decreases. But it is hard to judge whether earthquake is triggered or not, because the occurrence of earthquake is not only related to the change in the Coulomb failure stress, but also relevant with the initial stress field and the strength of earthquake fault zone before the earthquake. In order to study aftershock triggering and the continuous evolution of stress field considering the influence of tectonic stress field in the earthquake sequence, a new finite element method is proposed in the paper. This new FEM can be applied to calculate the continuous evolutions of displacement and stress field in an earthquake sequence. Earthquake triggering factor and seismic risk degree are proposed for determining if the earthquake can be triggered or not and assessing the degree of seismic risk caused by the earthquake, respectively. The new FEM and these two concepts are used to study the aftershock triggering of the 1992 Joshua Tree/Landers/Big Bear earthquake sequence. The results of this investigation indicate that the method is suitable to characterize the earthquake triggering relationship and the evolution of stress field perturbed by earthquakes.
T13A-1926
Micro-seismicity of the submerged section of the North-Anatolian Fault within the Sea of Marmara : results from Ocean Bottom Seismometers
Four Ocean Bottom Seismometers (OBS) were deployed for 2,5 months in the Tekirdag Basin (eastern Sea of Marmara), within the submerged section of the North Anatolian Fault. Seismological data from land stations deployed onshore by the Kandilli Earthquake Observatory (KOERI) were also used to improve the determination of the earthquake characteristics. The OBS array was centered a cold seep site located within the fault zone, which was extensively explored with submersibles, ROV in 2002 and Nautile in 2007. Near the cold seep, one piezometer and 3 flowmeters were also deployed. During the 10-weeks long deployment, the OBSs recorded a total of about 150 events from below the basin and the immediately adjacent submarine areas, while only about 50 events were recorded by the land stations. The data thus confirm previous experiments in the Marmara Sea showing the efficiency of seabottom instruments to lower the detection threshold compared to land stations. The recorded events are distributed in two groups. The first is a cluster of events located below the Western Ridge, below a zone where gas and oil seeps originating from Thrace Basin source rocks where found at the seafloor. Events of the second group are located deep in the crust and aligned along a NNE direction crossing the northern escarpment of the Tekirdag Basin. Visual observations with Nautile show that gas emissions occur through NNE oriented tensile cracks nearthe base of this escarpment. The seismological data thus suggests these cracks are the surface expression of a deep seated active fault.
T13A-1927
Re-examination of the Present Stress State of the Atera Fault, Central Japan, Based on the Calibrated Crustal Stress Data of Hydraulic Fracturing Test by Measuring the Tensile Strength of Rocks
To infer the activity and physical state of intraplate faults in Japan, we re-examined the crustal stress with the hydraulic fracturing test by measuring the tensile strength of rocks. The tensile strength was measured by fracturing hollow cylindrical rock samples (inner and outer radius are 25.0-25.2 mm and 55.1-101.5 mm, respectively, length is 137.0-140.1 mm) which were obtained close to the in situ stress measurement locations by pressurizing the inner hole of the sample. Confining pressure is not applied to the samples in this test. To check the reliability and accuracy of this test, we conducted similar experiments with the standard rock sample (Inada granite) whose physical property is well known. Then, we measured the tensile strength of all available core samples including the Atera fault (at Ueno, Fukuoka, and Hatajiri), the Atotsugawa fault, and the Nojima fault (at Hirabayashi, Iwaya and Kabutoyama), in central Japan, which had been obtained by the National Research Institute for Earth Science and Disaster Prevention (NIED) by the stress measurement with the hydraulic fracturing method. The measured tensile strength data reveals that the in situ re- opening pressure, which is one of the parameters needed for the determination of the maximum in situ horizontal stress, was obviously biased. We re-estimated the re-opening pressure using the measured tensile strength and the in situ breakdown pressure, and re-calculated the in situ stress around the Atera fault. Although the past dislocation of the Atera fault has been considered to be left lateral from the geographical features around the fault, the re-estimated stress suggests that the present dislocation of the Atera fault is right lateral. And the shear stress decreases from the fault. The right lateral dislocation is also supported by the present-day horizontal crustal deformation observed by the triangular and GPS surveys by Geographical Survey Institute in Japan. Therefore, the dislocation direction of the Atera fault seems to change from left lateral to right lateral some time ago. The amount of accumulated right lateral dislocation estimated from the stress data with the dislocation model by Okada (1992) is 2.2-2.6 m. Because the current slip rate from the GPS survey is 2.1-2.3 mm/yr, the accumulation period of the dislocation becomes 960-1240 years if the slip rate is stable. This estimation suggests that during the last 1586 Tensho earthquake the Atera fault dislocated right laterally.
T13A-1928
Slip Sequences in Laboratory Experiments as Analogues to Earthquakes Associated with a Fault Edge
Faults in the earth's crust are intrinsically heterogeneous at multiple scales with steps, jogs, asperities, and edges. This complexity governs some of the dominant properties of earthquakes, e. g., frequency-magnitude relations and repeating earthquakes. Here we experimentally explore how edges may affect earthquake and slip dynamics along a fault. We consider two flat blocks in frictional contact, where shear is applied at one edge of the system. We show that slip takes place via a sequence of rapid events which transfer the shear load across the entire block. Each rapid slip event arrests after a finite distance and both extends the slip size and causes a progressively larger change of the contact area along the contact surface. These slip events significantly modify the initially uniform contact area along the block surface, forming an asperity near the edge and largely reducing the contact area beyond. This rapid slip sequence culminates by a slow slip event that leads to major, unarrested slip along the entire contact surface. These results show that a simple deviation from uniform shear loading configuration can significantly and qualitatively affect both earthquake nucleation processes and the resulting evolution of fault complexity.
T13A-1929
Measuring the evolution of slip surface roughness with LiDAR
Roughness of faults plays an important role in controlling the resistance of faults to slip, yet all faults do not
have the same roughness. A previous study used ground-based LIDAR to analyze 12 fault surfaces from 8
localities in the Western United States, and inferred that small slip faults (slip <1 m) are rougher in the slip
parallel direction than large slip faults (slip 10 m - 300 m) [Sagy et al., 2006]. Here we more than double the
dataset and expand the types of faults studied in order to test this hypothesis. We have applied the same
methods to 13 Italian faults. The new faults are primarily normal faults with intermediate slip distances (slip 1
m - 150 m). The data includes strike-slip faults exhumed from depth, a fault type not included in the previous
study. The new data set also controls for lithology by focusing on carbonate rocks.
We measure roughness using mean power spectra of the topography in both the slip perpendicular and slip
parallel directions. The roughness in the slip perpendicular direction is identical to the Western US faults. Slip
perpendicular roughness is insensitive to slip and rock type. The slip parallel direction is smoother for all
faults than the slip perpendicular direction. Preliminary results suggest that roughness is intermediate
between the small slip (<1 m) and large slip (>10 m) faults studied previously. Roughness is also
sensitive to time of exposure as revealed by a fault surface that has been progressively exposed during
roadwork.
The large-scale data analysis is being enabled by Slugview, a program written by the UC Santa Cruz
Seismology Group. Slugview is a 3D point cloud visualizer used to manipulate raw data. The beta version of
SlugView 2.0 is available to anyone interested, and can be downloaded at
www.pmc.ucsc.edu/~msteffec/SlugView/.
http://www.pmc.ucsc.edu/~msteffec/SlugView/
T13A-1930
Differential Geometry of Fault Surfaces
The recent advent in ground-based light detection and ranging (Lidar) instruments permit a new and more detailed look at the three-dimensional geometry of fault surfaces at the mm to m scale. We utilize Lidar data from exposed fault surfaces that were recorded at a resolution of 3mm yielding approximately 11 million individual data points on the fault. We study the geometry of fault surfaces using the tools of differential geometry, which admit quantification to second order. These methods maintain the spatial coherence of the data and may thus augment existing measures of surface geometry (e.g. roughness along profiles) and yield new insights into the mechanics of faulting. The analyses reveal repetitive patterns of the fundamental parabolic, elliptical, and hyperbolic shapes with significant principal curvatures across many scales. This indicates that faulting is more complex than sliding along planar surfaces or sub-parallel corrugations. To test the implications for fault mechanics we conduct simple tests using boundary element methods to solve the boundary value problem of frictional sliding under idealized conditions. These tests illustrate the importance of the three-dimensionality of slip surfaces and the locally non-zero principal curvatures. The three- dimensionality alters the energy budget of faulting by accentuating the heterogeneity in work done against friction and the distribution of strain energy density near the fault surface.
T13A-1931
Influence of Inhomogeneous Stress Field on the Displacements and Stresses Induced by Normal Fault With Thickness
The fault dislocations from geological surveys are the accumulations of many earthquakes or geological events. Therefore, to investigate of the deformation features of the ground surface and fault dislocations can be attributed to studying the displacement and stress fields induced by the earthquakes. The displacement field is not only related to rock properties and tectonic stress field, but also to fault geometry and the way of the fault failure. In seismology, the dislocation theory based on homogeneous, isotropic elastic semi-infinite space is used to study displacement and stress fields caused by the fault dislocation which is specified on the fault surfaces in advance. It can provide analytical solutions about the displacement and stress fields. These results are very helpful for understandings about the deformation figures of the ground surface. However, it cannot be suitable to predict the displacement and stress fields caused by the dislocation of the fault with thickness if the tectonic stress field is inhomogeneous and the material is heterogeneous. In this research, a new finite element method (FEM) in consideration of inhomogeneous stress field is introduced to study the dislocation of a normal fault with thickness and its displacement and stress fields. New results reveal: 1) The maximum dislocation of the fault is at its lower part instead of its top. 2) The maximum horizontal and vertical displacements on the ground surface are not at the fault, but at a distance away from it. 3) The dislocation may cause two failure regions near the ground surface, one is close to the fault top, another is on the hanging wall and away from the fault, the successive normal faults observed in geological surveys are usually found there. 4) The results of normal faults with the same dip are apparently different from those of listric normal fault.
T13A-1932
On the genetic connection between misorientation and weakness: slip-tendency analysis of exhumed fault zones in the Alps
Crustal-scale fault zones which show a dip-slip component (either normal or reverse) and have been active for relevant times (e.g. some million years) are very often characterised by an asymmetric distribution of fault rocks, with rocks in the footwall or hangingwall (for normal or reverse faults resp.) showing a transition from relatively higher temperature crystal-plastic deformation mechanisms to low temperature brittle-cataclastic mechanisms. This is the result of progressive exhumation during a deformation continuum and may be predicted with the classic Sibson-Scholz fault zone model. This asymmetric distribution of fault rocks has been verified in exhumed fault zones from the metamorphic core of the Alps (Austroalpine and Penninic domains), such as the extensional Simplon and Brenner detachments, and studied in detail in the Sprechenstein-Mules fault zone (part of the eastern segment of the 700-km-long Periadriatic Fault System). Greenschist facies phyllonites, from a wide shear zone which constitutes the ductile precursor to the Sprechenstein-Mules brittle fault, are exposed at the hangingwall and are characterised by a pervasive SCC' composite foliation, marked by alternating phyllosilicate- and quartz-feldspar-rich layers. Centimetre- to micrometre-scale cataclastic shear zones develop along S, C and C' inherited surfaces. Hence, the hanging wall of the Sprechenstein-Mules fault zone is characterised by a strong mechanical anisotropy, which controls the mode of deformation under brittle conditions. However, given its origin in the plastic-metamorphic environment, this anisotropy is strongly misoriented for reactivation under brittle conditions. To investigate to control exerted by pre-existing ductile foliations on brittle faulting, we applied a development of slip tendency analysis that includes the effect of anisotropy. It shows that, given the mechanical anisotropy and under a realistic palaeo-state of stress, continuing activity along a misoriented and weak fault zone, with brittle re-activation of inherited metamorphic fabrics, must be considered not only possible, but even more probable than the development of Andersonian faults. This is in agreement with field and microstructural observations integrated in a 3D fault zone architecture model.
T13A-1933
Geometrical, Spatial and Mineralogical Properties of Pressure Solution Seams in Clastic Rocks, South-West Ireland
Pressure solution seams are the product of grain dissolution driven by compressive stress concentrations between grain contacts. Dissolved material is then transported to areas of lower stress via a thin fluid film trapped between the grains. Material with lower solubility may accumulate within this region to form a seam that acts as a low-porosity/permeability feature. These structures are, in effect, closing mode and may cause a bulk volume loss if dissolved material is taken out of the system. The deep-water clastic and carbonate rock units of south-west Ireland provide field data that allows for the characterization of pressure solution seams in clastic material within a fold and thrust belt. The pressure solution seams in the sandstone has been documented from grain to outcrop scales. At outcrop scale, pressure solution seams have a sinuous trace and often have a dark residue preserved within its core. At thin section scale, minor pressure solution seams are pervasive in the rock adjacent to the major, macroscopic seams. The seams often coalesce to form longer structures. Microprobe analysis shows that within a microscopic seam, there are a number of inter-granular dissolution surfaces. Quartz is subject to dissolution, with clays remaining within the seam as residual material. The permeability of a sandstone can be significantly reduced in an orientation orthogonal to the pressure solution seams, but may not be markedly affected in an orientation parallel to them. The orientation of a pressure solution array can be predicted as they form sub-parallel to regional fold axes and thrust fault traces. However, some pressure solution seams have been reported to assist fluid and hydrocarbon flow.
T13A-1934
Orientation Patterns of Chalcedony Veins and Clastic Dikes in Tertiary strata of NW Nebraska and SW South Dakota
Stratiform bound arrays of subvertical chalcedony veins occur in distinct patches of varying size in White River Group strata in the study area. At Toadstool Geologic Park in Nebraska they are associated with normal faults, and within a patch display orthogonal patterns with one direction dominant. Junctures and tip interactions indicate the orthogonal veins formed together. Two different patches at Toadstool show different orientations. At three separate study sites in Badlands National Park patches of chalcedony vein patches are not associated with faults, but are associated with clastic dike systems. Field relations indicate clastic dike formation preceded vein formation, but common orientation patterns, composite dike-vein features and a spatial association suggest similar timing and a genetic linkage. While vein strikes for two sites appear polygonal and are statistically random, within each site subareas are organized, and do not display a simple polygonal triple junction geometry. Whether strike distributions are random or organized is partly scale dependent. The third site shows a strong preferred direction. At all localities chalcedony veins exhibit vertical shortening, explained most simply by compaction. The vein arrays are similar to polygonal faults, and are interpreted to be the result of stratigraphically controlled diagenetic driven deformation. Clastic dikes can traverse most of the Tertiary section, but are concentrated in the Sharps Formation, and taper downwards into the Chadron Formation, where some merge with veins. Clastic dike orientations have been studied at three localities, two of which show different preferred directions, and one of which appears random. A provisional model includes basal high fluid pressures in silt/mudstones rich in volcanic ash that aids overlying clastic dike formation, producing fracture drainage paths that trigger diagenetically driven deformation, and associated compaction, forming chalcedony veins. An initial local vein-fracture may organize the area around it, but the larger array of initial features was random. Whether a regional stress field contributed to the orientation patterns is yet unclear. A long list of undergraduate students participated in this research and are thanked for their contributions.