S. A. F. Murrell

The effect of decomposition of hydrous minerals on the mechanical properties of rocks at high pressures and temperatures

Abstract Extensive experiments have been carried out in which specimens of gypsum, a partially serpentinized peridotite, a serpentinite and a chloritite have been subjected to pressures up to 0.662 GPa together with temperatures up to 780°C and have been deformed at a fixed strain rate of ∼10 −5 /s. The commencement of decomposition of the hydrous minerals is accompanied in sealed specimens by loss of strength, a reduction in sliding friction, and embrittlement of the rocks. Dilatancy-hardening effects are observed. Specimens which are drained to the atmosphere remain strong. In gypsum there is a ten-fold reduction in strength between temperatures of ∼50°C and ∼150°C. The partially serpentinized peridotite (∼40% forsterite, ∼60% antigorite) which contains ∼1% of brucite shows a reduction in strength ∼50% at ∼300°C, followed by a further ten-fold reduction between ∼300°C and ∼700°C. The serpentinite (∼90% lizardite and chrysotile) shows a ten-fold reduction in strength between ∼400°C and 600°C. The chloritite (∼85% ripidolite) shows a reduction of strength by about a half at 300°C; the strength remains approximately constant between 300°C and 600°C, and there is a further five-fold reduction in strength between 600°C and 700°C. The phase changes in the hydrous minerals have been studied by means of differential thermal analysis, X-ray diffraction and optical microscopy, and will be more fully described elsewhere. A detailed discussion is given of the deformation characteristics and mechanisms, with particular emphasis on the role of pore pressure and dilatancy. There is a range of temperature for each of these rocks in which the deformation of sealed samples is most stable, in the sense that brittle faulting accompanied by a stress-drop does not occur. At higher temperatures the rocks become unstable and very weak. Under conditions corresponding to geothermal gradients between 5 K km −1 and 100 K km −1 these rocks would be brittle and weak at shallow depths, and would again become brittle and weak at depths below some level which depends on the rock. Possible implications are discussed in connection with faulting and earthquakes, with syntectonic metamorphism, and with the emplacement of Alpine-type peridotites.

Experimental and theoretical fracture mechanics applied to Antarctic ice fracture and surface crevassing

Recent disintegration of ice shelves on the Antarctic Peninsula has highlighted the need for a better understanding of ice shelf fracture processes generally. In this paper we present a fracture criterion, incorporating new experimental fracture data, coupled with an ice shelf flow model to predict the spatial distribution of surface crevassing on the Filchner-Ronne Ice Shelf. We have developed experiments that have enabled us to quantify, for the first time, quasi-stable crack growth in Antarctic ice core specimens using a fracture initiation toughness, Kinit, for which crack growth commences. The tests cover a full range of near-surface densities, ρ = 560–871 kg m−3 (10.9–75.7 m depth). Results indicate an apparently linear dependence of fracture toughness on porosity such that Kinit = 0.257 ρ-80.7, predicting a zero-porosity toughness of Ko = 155 kPa m1/2. We have used this data to test the applicability to crevassing of a two-dimensional fracture mechanics criterion for the propagation of a small sharp crack in a biaxial stress field. The growth of an initial flaw into a larger crevasse, which involves a purely tensile crack opening, depends on the size of the flaw, the magnitude of Kinit and the nature of the applied stress field. By incorporating the criterion into a stress map of the Filchner-Ronne Ice Shelf derived from a depth-integrated finite element model of the strain-rate field, we have been able to predict regions of potential crevassing. These agree well with satellite imagery provided an initial flaw size is assumed in the range 5–50 cm.

Microcracking during triaxial deformation of porous rocks monitored by changes in rock physical properties, I. Elastic-wave propagation measurements on dry rocks

We present results from two series of triaxial deformation experiments performed on “dry” samples of two sandstones (Darley Dale and Gosford) carried out at confining pressures from 25 MPa to 200 MPa. Over this pressure range the mode of failure in both these sandstones passes from localized brittle failure with a clear through-going fault to distributed cataclastic flow. During these experiments, stress, strain, compressional-wave velocity (VP) and shear-wave velocity (VS) measurements were made simultaneously in the direction of the maximum principal compressive loading axis. Initial application of the hydrostatic confining pressure causes both VP and VS to increase, and upon raising the axial stress above the confining pressure both velocities increase further at first (generally by only a few percent), but then decrease as dilatant crack growth commences. During dilatancy, VS decreases proportionately more than VP, and this decrease is generally of the order of 10–15%. These velocity measurements allow changes in rock physical properties to be calculated along with the axial and transverse crack volume density parameters, ϵX and ϵZ. The results from two selected tests are analysed in detail. These tests were chosen because they exhibit; (a) typical brittle shear failure, and (b) typical ductile cataclastic flow, respectively. The full interrogating elastic waveforms were also recorded during testing, and these have been used to calculate the seismic quality factors QP and QS. To our knowledge, this is the first time this has been reported for rock samples undergoing triaxial deformation. The changes in Q values generally exhibit similar trends to those observed in the velocity measurements, but the percentage changes in Q are an order of magnitude greater, suggesting that this parameter is a more sensitive measure of dilatant crack damage. The measurements on dry rock samples reported here provide the basis for comparison with measurments of changes in complementary physical properties made on water-saturated rock samples under the same experimental conditions, and reported in a companion paper in this issue (Read et al., 1995).

SCANNING ELECTRON MICROSCOPE AND ACOUSTIC EMISSION STUDIES OF CRACK DEVELOPMENT IN ROCKS

Abstract The microcracks and voids contained in rocks affect their fracture, elastic, thermal and electrical properties. This investigation assesses methods for revealing cracks in unstressed rock and cracks developed during uniaxial and triaxial compression up to failure, using three different rocks: microgranodiorite, dolerite and marble. The techniques used are scanning electron microscope (SEM) analysis of ion-beam etched surfaces, and acoustic emission (AE) measurements. The results are compared with other published data on cracks in rocks and from this a more general picture can be drawn of void/crack development and of damage growth during deformation from zero load up to failure in compression.

Rheology of the lithosphere — Experimental indications

Abstract Models are presented of the structure, composition, temperature, and elastic properties of three types of lithosphere (steady-state oceanic, continental shield, and ‘Basin and Range’ continental). Laboratory studies of creep in granitic, basaltic and ultrabasic rocks are then used to construct models of lithosphere rheology. Good agreement is found with geophysical observations if the stress increases from ~1 MPa (10 bar) at the base of the lithosphere to values which range from ~1 MPa to >100 MPa (10 bar to >1 kbar) in the crust.

Ionic surface electrical conductivity in sandstone

Recent analyses of complex conductivity measurements have indicated that high-frequency dispersions encountered in rocks saturated with low-salinity fluids are due to ionic surface conduction and that the form of these dispersions may be dependent upon the nature of the pore and crack surfaces within the rock (Ruffet et al., 1991). Unfortunately, the mechanisms of surface conduction are not well understood, and no model based on rigorous physical principles exists. This paper is split into two parts: an experimental section followed by the development of a theoretical description of adsorption of ions onto mineral surfaces. We have made complex conductivity measurements upon samples of sandstone saturated with a range of different types and concentrations of aqueous solution with a frequency range of 20 Hz to 1 MHz. The frequency dependence of complex conductivity was analyzed using the empirical model of Cole and Cole (1941). The “fractal” surface models of Le Mehaute and Crepy (1983), Po Zen Wong (1987), and Ruffet et al. (1991) were used to calculate apparent fractal pore surface dimensions for samples saturated with different solution types and concentrations. These showed a pronounced decrease of apparent fractal surface dimension with decreasing electrolyte concentration and a decrease of apparent fractal dimension with increasing relative ionic radius of the dominant cation in solution. A model for ionic surface concentration (ISCOM I) has been developed as the first step in producing a rigorous physicochemical model of surface conduction in quartz-dominated rocks. The results from ISCOM I show that quartz surfaces are overwhelmingly dominated by adsorbed Na+ when saturated with NaCl solutions of salinities and pH found in actual geological situations. ISCOM I also shows that the concentration threshold for dominance of surface conduction over bulk conduction is aided by depletion of ions from the bulk fluid as a result of their adsorption onto the mineral surfaces as well as by changes in the ionic mobility in the surface conduction double-layer as the wetting solution becomes more dilute.

Microcracking during triaxial deformation of porous rocks monitored by changes in rock physical properties, II. Pore volumometry and acoustic emission measurements on water-saturated rocks

Abstract We report results from a series of laboratory triaxial-deformation experiments performed on samples of Darley Dale sandstone at servo-controlled constant strain rate and constant pore-fluid pressure. During deformation, the volume of water either expelled from the samples during compaction or injected into the samples during dilatancy in order to maintain a constant pore-fluid pressure was continuously monitored throughout each experiment. In addition, the parameters of differential axial stress, axial strain, and acoustic emission (AE) characteristics were also recorded. Complete AE waveforms were also captured and stored by means of a transient recorder, and Fourier analysis of these waveforms was subsequently carried out. The reported experiments were all performed in a conventional, high-pressure, gas-medium triaxial cell, but with the pore-fluid pressure maintained constant using a new servo-controlled fluid pressure intensifier and pore volumometer. The AE and pore volumometry measurements on water-saturated rocks reported in this paper are complementary to the measurements of elastic-wave propagation parameters made on dry rocks and reported in a companion paper in this issue. The two suites of data have been integrated to infer relations between crack density parameters calculated from elastic-wave velocity measurements and the directly measured pore volume during deformation. Our results show a distinct, positive correlation between changes in the axial crack density parameter ( ϵ X ) and changes in the pore volume during deformation once fluid expulsion due to elastic pore collapse has been corrected for. These two parameters have then been combined to obtain an estimate of changes in the mean aspect ratio of propagating dilatant cracks. The results suggest that, under moderate confining pressures, axially aligned dilatant cracks are likely to bow open elastically prior to significant crack extension and growth.

Fracture of multiyear sea ice

The fracture and flow of multiyear sea ice was investigated under triaxial compression and uniaxial tension in the temperature range −40° to −3.5°C, for strain rates from 10−7 to 10−2s−1 and for confining pressures up to 30 MPa using 40 mm diameter specimens. Specimens both in the horizontal plane of the multiyear floe and perpendicular to this plane were tested. The results of short-rod fracture toughness tests on multiyear and first-year sea ice at temperatures −20°C are also reported. The multiyear sea ice came from an unridged portion of a single floe about 7 m thick, which was found to be massive and not blocky with large voids. The ice had low salinity and high porosity. The inelastic deformation of multiyear sea ice was found to be strongly dependent upon strain rate, temperature, and confining pressure. In compression, four main types of deformation were observed. (1) Under uniaxial compression, completely brittle fracture at high strain rates (of 10−3 to 10−2 s−1) was characterized by multiple axial splitting. (2) Application of even a small confining pressure inhibited splitting, and fracture took place by the formation of a narrow shear fault inclined at 45±3°. (3) At higher confining pressures, plastic deformation accompanied substantial cracking activity. (4) However, at still higher confining pressures, cracking was completely inhibited and deformation was entirely plastic. At −20°C, shear fracture occurred according to a maximum shear stress criterion and hence was pressure independent, with crack nucleation dominating the fracture behavior. At −40°C, however, the shear fracture stress was found to be strongly pressure dependent up to 14 MPa and could be described in terms of a Coulombic failure envelope. The unusual 45° orientation of ice shear fractures, together with the unusual pressure dependencies of ice peak strengths, may be explained by the fact that low-stress slip and cleavage occurs in the basal planes of ice crystals.

Acoustic emission and fluid permeability measurements on thermally cracked rocks

Abstract Thermal cracking experiments have been conducted to investigate the role of anisotropic thermal expansion of the constituent minerals of a basalt, a microgabbro and a microgranite in changing the physical properties of the rocks. Samples of an Icelandic basalt, a microgabbro from Sweden, and a fine-grained microgranite from Ailsa Craig, Scotland, have been selected for these experiments. Samples were heated to a range of peak temperatures up to 800°C at ambient pressure. Thermal cracking was monitored through acoustic emission and elastic wave velocity measurements. Fluid permeability of the basalt was measured under hydrostatic conditions at elevated confining pressure using the steady state flow technique.

Modelling the stress-strain behaviour of saturated rocks undergoing triaxial deformation using complex electrical conductivity measurements

Measurement of complex electrical conductivity as a function of frequency is an extremely sensitive probe for changes in pore and crack volume, crack connectivity, and crack surface topography. Such measurements have been made as a function of pore fluid chemistry, hydrostatic confining pressure, as well as uniaxial and triaxial deformation. This paper will; (1) describe the effects of triaxial deformation on the complex electrical conductivity of saturated porous rocks, (2) use the electrical data to model the mechanical stress-strain behaviour, and (3) compare the modelled behaviour with the stress-strain behaviour measured during the deformation. Experimental conductivity data tracks how the rock undergoes compaction with progressive loss of crack volume, followed by dilatation due to new crack formation, growth of existing cracks, crack interlinkage, and finally failure, as axial strain is increased. We have used the complex electrical data to produce a direction-sensitive (anisotropic) crack damage parameter, and used it to calculate the effective Youngs modulus by employing the models of Walsh and Bruner. Comparison of the synthetic stress-strain curves so produced, with the experimentally derived stress-strain curves shows good agreement, particularly for undrained tests. This modelling is an improvement on similar curves produced using isotropic crack damage parameters derived from acoustic emission data. The improvement is likely to be due to the directional sensitivity of the electrical conductivity measurement, and its ability to discriminate between the formation of isolated cracks, and those cracks that contribute to the inter-connected crack space i.e. those cracks upon which transport properties of the rock such as electrical conductivity, and mechanical properties depend most critically during triaxial deformation.