Laboratory measurements of rock strength provide limiting values of lithospheric stress, provided that one effective principal stress is known. Fracture strengths are too variable to be useful; however, rocks at shallow depth are probably fractured so that frictional strength may apply. A single linear friction law, termed Byerlee's law, holds for all materials except clays, to pressures of more than 1 GPa, to temperatures of 500°C, and over a wide range of strain rates. Byerlee's law, converted to maximum or minimum stress, is a good upper or lower bound to observed in situ stresses to 5 km, for pore pressure hydrostatic or subhydrostatic. Byerlee's law combined with the quartz or olivine flow law provides a maximum stress profile to about 25 or 50 km, respectively. For a temperature gradient of 15°K/km, stress will be close to zero at the surface and at 25 km (quartz) or 50 km (olivine) and reaches a maximum of 600 MPa (quartz) or 1100 MPa (olivine) for hydrostatic pore pressure. Some new permeability studies of crystalline rocks suggest that pore pressure will be low in the absence of a thick argillaceous cover.

The permeability of Westerly granite was measured as a function of effective pressure to 4 kb. A transient method was used, in which the decay of a small incremental change of pressure was observed; decay characteristics, when combined with dimensions of the sample and compressibility and viscosity of the fluid (water or argon) yielded permeability, k. k of the granite ranged from 350 nd (nanodarcy = 10−17 cm2) at 100‐bar pressure to 4 nd at 4000 bars. Based on linear decay characteristics, Darcy's law apparently held even at this lowest value. Both k and electrical resistivity, ρs, of Westerly granite vary markedly with pressure, and the two are closely related by k = Cρs−1.5±0.1, where C is a constant. With this relationship, an extrapolated value of k at 10‐kb pressure would be about 0.5 nd. This value is roughly equivalent to flow rates involved in solute diffusion but is still a great deal more rapid than volume diffusion. Measured permeability and porosity enable hydraulic radius and, hence, the shape of pore spaces in the granite to be estimated. The shapes (flat slits at low pressure, equidimensional pores at high pressure) are consistent with those deduced from elastic characteristics of the rock. From the strong dependence of k on effective pressure, rocks subject to high pore pressure will probably be relatively permeable.

Volume changes of a granite, a marble, and an aplite were measured during deformation in triaxial compression at confining pressure of as much as 8 kb. Stress‐volumetric strain behavior is qualitatively the same for these rocks and a wide variety of other rocks and concrete studied elsewhere. Volume changes are purely elastic at low stress. As the maximum stress becomes one‐third to two‐thirds the fracture stress at a given pressure, the rocks become dilatant; that is, volume increases relative to elastic changes. The magnitude of the dilatancy, with a few exceptions, ranges from 0.2 to 2.0 times the elastic volume changes that would have occurred were the rock simply elastic. The magnitude of the dilatancy is not markedly affected by pressure, for the range of conditions studied here. For granite, the stress at which dilatancy was first detected was strongly time dependent; the higher the loading rate the higher the stress. Dilatancy, which represents an increase in porosity, was traced in the granite to open cracks which form parallel with the direction of maximum compression.

Stick-slip often accompanies frictional sliding in laboratory experi ments with geologic materials. Shallow focus earthquakes may represent stick slip during sliding along old or newly formed faults in the earth In such a situation, observed stress drops repre sent release of a small fraction of the stress supported by the rock surround ing the earthquake focus.

No abstract

We reanalyze the flow model proposed by Wyllie and Rose (1950) in which the complicated flow network through the pore phase of rock is replaced by a single representative conduit. Although the model is a very simple representation of the complicated pore phase in rock, we find that it provides an adequate simulation of how the transport properties vary with external pressure. Expressions derived for fluid permeability k and formation factor F are combined to give an expression for the mean hydraulic radius of the pore phase. Using this expression, we show that the exponent r in the empirical relationship k oc F-' must fall in the range 1 < r < 3. Also, we use the expression for hydraulic radius to estimate the crack area per unit volume and the standard deviation of the height of the asperities on the microcrack surfaces for two granites. The values are in reasonable agreement with other estimates.

No abstract

The growth of cracks in photoelastic material and glass under compression is being studied as part of an investigation of brittle fracture of rock. In compression the most severely stressed crack is inclined at about 30° to the axis of compression. Such cracks, when either isolated or placed in an array, grow along a curved path which becomes parallel with the direction of compression. When this direction is attained, growth stops, unless applied compression is increased considerably. Cracks in certain en échelon arrays start to grow at much smaller applied stress than that required to enlarge an isolated crack.