No abstract

SUMMARY It is usually suggested that thermal cracking in a quartz‐bearing rock results from the anomalously high volumetric expansion coefficients of quartz (e.g. Simmons & Cooper 1978). It has also been recognized that thermal expansion mismatch and mineral anisotropy contribute to thermal cracking in materials that consist of a polycrystalline aggregate composed of several anisotropic minerals even in the absence of a temperature gradient (Friedrich & Wong 1986). Experiments investigating thermal cracking in rocks commonly involve imaging and quantitative stereology of crack populations induced in rocks treated to various peak temperatures (e.g. Friedrich & Johnson 1978; Homand‐Etienne & Troalan 1984; Atkinson, McDonald & Meredith 1984; Meredith & Atkinson 1985). Here we report on acoustic‐emission experiments that monitor the process of thermal cracking as it occurs during heating, supported by measurements of crack surface area, pore‐fluid permeability, porosity and surface conductivity carried out on rock samples treated to various peak temperatures. The acoustic‐emission measurements show a strong peak of microcracking at the phase transition temperature for quartz (˜573°C) superimposed upon a background of microcracking due to thermal expansion. There is also a clear peak of microcracking at higher temperatures (˜800°C) that can be attributed to oxidation‐dehydroxylation reactions of hornblende and chlorite. Measurements of fluid permeability, pore surface area, porosity and electrical conductivity, made on samples that have been heat treated to various maximum temperatures, show increases associated with a major episode of cracking in the 500‐600°C temperature range, indicating that the new cracks form a well‐interconnected network. This has been confirmed by SEM and optical microscopy These results have implications for the electrical conductivity of the continental crust, providing a mechanism enabling the high pore‐fluid connectivity needed to explain zones of high electrical conductivity at depth providing that cracks opened in this way remain open at the high pressures existing at depth. It should be recognized, however, that th***se measurements are limited in their direct application since they were obtained under initially dry conditions at laboratory pressures.

Creep of forsterite single crystals has been studied with respect to the orientation of the differential stress. Three orientations have been investigated: [110]c, [101]c, and [011]c. Specimens were deformed at high temperature (T ≥ 1400°C) and moderate stresses (15 < σ < 110 MPa) in a dead load creep apparatus at room pressure and under controlled atmosphere. Assuming that the creep law has the general form = Aσn exp −(Q/RT), two different tests were performed. Stress step and temperature step experiments gave the stress exponent n and the activation energy Q, respectively. A thorough analysis of the dislocation microstructures was carried out in the potential glide planes using the optical microscope (decorated thin section) and transmission electron microscope. Comparison of both mechanical and microstructural data allowed the determination of the different flow laws for the three orientations studied. Forsterite and olivine data are consistent; both flow laws and dislocation microstructures are similar. In light of these results, extrapolation of flow laws to the stress conditions of the mantle is not straightforward. Very high activation energies at low stresses (σ < 15 MPa) may suggest a change of flow mechanisms. Accordingly, further investigations in the low‐stress domain are required.

Permeability of granite specimens heated up to 650°C were measured in the laboratory with respect to both confining and pore pressure. An unexpected behavior (permeability decrease) was found in the low temperature range (20–125°C). Permeability variations are interpreted with the help of a theoretical model of porosity. Complementary measurements of porosity and absolute surface area have also been obtained. A discussion in terms of microstructural controlling parameters concludes to a decrease of crack aperture at low temperatures (20–125°C), followed by an increase at higher temperatures (T >125°C).