T14A-01 INVITED
The Geology of Impact Fragmentation and Crater Controlled Fracture Networks
Overall, the study of planetary impact crater physics has made significant advances in the past two decades, thanks to modeling and observation. However, our understanding of the fragmentation network beyond a crater wall remains poorly understood. At the fundamental level, we do not yet understand where the transition takes place, and over what scale, between the so-called "strength regime" (where strength and other rheological properties dominate) and the "gravity regime" of impact crater scaling. The issue is muddled by the fact that numerical shock-hydrodynamical models (hydrocodes) have difficulty accurately evolving a stress tensor in a solid undergoing shock and hydrodynamic motion. Furthermore, the evolution of the stress tensor during the production of gravity-scaled craters is a subject of much debate, for if strength were truly negligible, there would be no final craters. Further still, and of subject here, only a few published efforts have made inroads into the topic of what damage occurs beyond the crater walls in a planetary impact. The latter subject is one of some importance, for it is speculated that large, deep-seated craters on Mars may dominate the subsequent regional hydrologic evolution. If so, then the same may be true, to a more limited extent given its other activity, of Earth's lithosphere. Radial fissures in crater systems, resulting primarily from hoop stresses and listric faults towards the crater interior, may set the stage for post-impact volatile evolution. On Europa, there is evidence that small secondary impact craters create fracture networks in an ice lithosphere a few km thick, and that these fractures accommodate the far-field stresses of impact rebound by connecting into long faults. One Enceladus, Dione, Ganymede and other icy bodies, the effect of impact cratering beyond the rim is observed in fault patterns. Here I present evidence for crater controlled fracture networks in planetary lithospheres, and provide an fundamental overview of what is known about the process of impact fragmentation, at geologic scales, in both high and low strain rate regimes, including the transition from spall spatial scales and high strain rates, to large spatial scales and low strain rates. The principal challenges of implementing realistic fracture models in numerical hydrocodes will be addressed.
T14A-02
Fragmentation on a grand scale: A slab fragment dislodged from the descending Pacific plate slab wedged Beneath Tokyo
The Eurasian-Philippine Sea-Pacific plate triple junction located 250 km southeast of Tokyo not only produces megathrust subduction earthquakes along the Sagami trough (including the 1703 M=8.3 Genroku and 1923 M=7.9 Kanto shocks) but is also associated with enigmatic M~6-7 earthquakes directly beneath Tokyo (such as the 1855 M~7.1 Ansei-Edo earthquake and 2006 M=6.0 Chiba shocks). In the past, the Tokyo quakes have been interpreted as resulting from a highly warped portion of the Philippine Sea slab scraping against the Pacific slab below. However, on the basis of microearthquake distributions, seismic tomography, bathymetric analysis, paleomagnetic data on past plate motions, magnetic seafloor anomalies, and seismic stress inversion, we argue instead that a 100-km-wide by 25-km-thick fragment of the Pacific plate lithosphere detached from the descending slab several million years ago and is now jammed between the Pacific, Philippine Sea and Eurasian plates beneath Tokyo. We explore and visualize the complex three-dimensional nature of this problem through a series of seismicity and plate animations. We identify a pronounced ‘snow-plow' shape to the double seismic zone of the descending Pacific plate slab, which closely parallels the sharp curvature of the overlying volcanic front. We argue that the fragment was probably dislodged when two buoyant seamount chains on the Pacific plate collided with the Japan-Izu trench east of Tokyo. Since breakup, the fragment has been pushed 100 km to the northwest by the leading edge of the Philippine Sea plate slab. Whether the plate-scale fragmentation that we find is common or rare is difficult to assess, because our analysis is possible only because of the extraordinarily rich data of southwest Japan. Nevertheless, subducting seamount chains appear to deform trenches and bend slabs in many circum-Pacific settings, modifying seismic behavior and perhaps causing slab fracture and fragmentation.
T14A-03 INVITED
Damage Characterization in Sandstones Along the Mojave Section of the San Andreas Fault With a new Method: Initial Results and Implications for the Depth and Mechanism of Dynamic Rock Fragmentation
Following theoretical expectations that pervasive rock damage during seismic faulting is limited to the top few km of the crust (Ben-Zion and Shi, 2005; Rice et al., 2005), and uncertainties regarding the depth of rock pulverization within fault zones (Dor et al. EPSL 2006), we evaluate the fragmentation intensity in sedimentary rocks that have never been buried deeply and were displaced by the San Andreas Fault (SAF). For the analysis of damage in the sandstones we use a new method that compares the original perimeter length of a grain to the total perimeter length of its fragments, applied to a statistically representative population of grains from each sample. We employ this method on samples from the Juniper Hills formation, a tectono-stratigraphic unit that has been deposited adjacent to active strands of the SAF system. This unit, like many of the other examined sandstones in the vicinity of the Mojave section of the SAF, displays minimal or complete absence of significant SAF-parallel shear, although in places it is not as cohesive as older sandstone units along the fault. Results of a transect on the southwest side of the SAF delineate a damage zone of about a 100 m wide that is likely associated with SAF faulting events. Together with other considerations, the damage content within those sandstones suggest that dynamic fragmentation on the microscale occurs very close to the Earth surface. When we apply the method on three mutually perpendicular sections of a sample collected 10 m from the fault we observe an anisotropic damage pattern. In addition we observe preferred orientation of microfractures and many microscale damage elements likely associated with grain contact pressure. Those observations are compatible with failure in an overall compressional field. The orientation of microfractures in a sample near the fault is normal to the SAF and leaning to verticality, in agreement with modeling of the possible orientation of maximum compressive stress during cyclic loading associated with slip events on rough frictional fault surface (Chester and Chester, 2000). A change in the preferred orientation of microfractures between this sample and a more distant sample can possibly reflect a variability of the stresses throughout the damage zone that may be associated with strong dynamic reduction of normal stress or fault opening.
T14A-04
Fractal Fragmentation in Fault Zones
There is mounting field evidence that the size distribution of particles in fault gouge and breccia is power law with a slope (fractal capacity dimension D) that increases with increasing strain from values near D=2.6 at low strain to values approaching D=3.0 at large strains. This trend is also observed in 3-dimensional computer simulations that allow multiple fragmentation of composite particles. Both the power law distribution and the increase of fractal dimension with strain can be understood using a simple nearest neighbor fragmentation model for "constrained comminution" that predicts a fractal dimension of D=2.6 at low-strain which increases to a stable value of D=3.0 at high strain. The evolution from a gouge with D=2.6 to a cataclasite with D=3.0 is illustrated using an automaton that implements the evolving fracture probabilities calculated using this constrained comminution model.
T14A-05
Stress generation and hierarchical fracturing in reactive systems
Hierarchical fracture patterns are the result of a slowly driven fracturing process that successively divides the rocks into smaller domains. In quasi-2D systems, such fracture patterns are characterized by four sided domains, and T-junctions where new fractures stop at right angles to pre-existing fractures. We describe fracturing of mm to dm thick enstatite layers in a dunite matrix from the Leka ophiolite complex in Norway. The fracturing process is driven by expansion of the dunite matrix during serpentinization. The cumulative distributions of fracture lengths show a scaling behavior that lies between a log – normal and power law (fractal) distribution. This is consistent with a simple fragmentation model in which domains are divided according to a ‘top hat' distribution of new fracture positions within unfractured domains. Reaction-assisted hierarchical fracturing is also likely to be responsible for other (3-D) structures commonly observed in serpentinized ultramafic rocks, including the mesh-textures observed in individual olivine grains, and the high abundance of rectangular domains at a wide range of scales. Spectacular examples of 3-D hierarchical fracture patterns also form during the weathering of basaltic intrusions (dolerites). Incipient chemical weathering of dolerites in the Karoo Basin in South Africa occurs around water- filled fractures, originally produced by thermal contraction or by externally imposed stresses. This chemical weathering causes local expansion of the rock matrix and generates elastic stresses. On a mm to cm scale, these stresses lead to mechanical layer-by-layer spalling, producing the characteristic spheroidal weathering patterns. However, our field observations and computer simulations demonstrate that in confined environments, the spalling process alone is unable to relieve the elastic stresses. In such cases, chemical weathering drives a much larger scale hierarchical fracturing process in which fresh dolerite undergoes a continuous domain division that effectively regenerates fresh surfaces. This process produces the characteristic weathering patterns seen both in Karoo and a wide-range of other geological environments. In summary, hierarchical fracturing leads to a continuous production of fresh reactive surface area during hydration processes such as serpentinization and weathering, and provides first-order rate control during both serptinization and weathering. It thus has wide ranging implications for global geochemical budgets, landscape evolution, and a number of other important geological features.
T14A-06 INVITED
Hierarchical Cracking Seen as a Successive Space Division
Contraction of a thin material layer can lead to hierarchical cracking. One fracture succeed the other, and while the older fractures do have an important impact on the younger ones, a younger fracture can effect the preceding ones. This hierarchical interaction is reflected in the geometry on the resulting crack pattern. We present a description of this pattern formation based on the fact that the fractures divide the two-dimensional surface into distinct domains. We show that this description leads directly to a mathematical constraint: the average number of sides of these domains must be four. On the other hand, we use this description to design a simple experiment that, together with a theoretical analysis, let us understand the crack-crack interaction. References: S.Bohn, S; Douady, Y. Couder (2005) Phys. Rev. Lett. 94, 054503 S. Bohn, L. Pauchard, Y. Couder (2005) Phys. Rev. E 71, 046214 S. Bohn, J. Platkiewicz, B. Andreotti, M. Adda-Bedia, and Y. Couder (2005) Phys. Rev. E 71, 046215
T14A-07
Fragmenting magma: experimental investigations of explosive eruptions
Observation and numerical modeling have long been the dominant means of understanding eruption dynamics of volcanoes. Lab based techniques, developed by the Experimental Volcanology Goup at LMU, were designed to gain insight into the physics of magma fragmentation and explosive eruption dynamics. Various experimental approaches have been aimed at filling the gap between field observations and modeling since the region of interest is beyond direct access. Analysis of the fragmentation threshold led to a systematic investigation of parameters governing the eruption behaviour (e.g., density, permeability, texture; etc.). Further investigations such as measurement of the fragmentation front velocity have yielded insight into the parameters of eruption dynamics, which can be correlated and incorporated into numerical models. As an accompanying measure, the physical characterisation of pyroclastic materials ejected during a single eruption have been the subject of intensive field studies with the first reliable density distributions resulting. By combining the experiments and field observations we have strived to better ascertain the physical constraints on eruption dynamics. Measurements of the permeability of porous magma under turbulent flow regime have also been performed to accompany the fragmentation studies. The experiments reviewed here have significantly influenced the current thinking on explosive eruptions from magma fragmentation to the pyroclastic deposits. Here we will explain why and how.
T14A-08 INVITED
The dynamic similarity between columnar joints in basalt and starch
We compared the scales of columnar joints due to thermal contraction in basalt to the scales of columnar joints due to desiccation in corn starch slurries. We made extensive field measurements of basalt columns at several sites including the island of Staffa, the Giant's Causeway and the Columbia Plateau. At each site, we examined in detail the scaling relationship between the column width and the height of the striae. Striae are the surface traces left on the sides of the columns by individual fracture advances. These data can be analyzed using thermal and fracture models in such a way that the average fracture advance speed and cooling rate can be determined. We found that the column radius and striae size are proportional to each other, and inversely proportional to the cooling rate. Controlled laboratory experiments in desiccating corn starch produce uniform fracture advance rates and uniform column widths. Starch columns are 100 times smaller than their basalt counterparts. We show that both basalt and starch joints form at similar values of the dimensionless Peclet number --- the ratio of the fracture advance rate times the column width to the diffusion constant of heat or moisture. This parameter arises because the two systems can be described by essentially the same continuum advection-diffusion equations. The steady advance speed of the cracks, due to ground water boiloff in basalt joints or the controlled evaporation rate in the starch experiments, corresponds to a non-equilibrium steady state of these equations and hence naturally scales with the Peclet number. We show that this scaling holds for a wide range of column widths in both cases.