Task 6.2: Experimental results and discussion

Starting material

The measurements reported here were made on samples of a macroscopically isotropic basalt collected from a roadstone quarry located southeast of Reykjavík, Iceland. Microscopically it has an aphyric texture, comprising euhedral laths of unaltered plagioclase averaging 0.2 mm in length, and anhedral augite microphenocrysts averaging 0.1 mm in diameter, with accessory anhedral oxides up to 0.1 mm in diameter. No free quartz was visible under either optical or SEM microscopy.

Summary of previously reported results

We previously reported results of measurements made at room temperature on samples that have previously been heat-treated to temperatures up to a maximum of 900^oC in order to induce thermal crack damage. In a shallow crustal environment where the geothermal gradient is anomalously high, such as in Iceland, thermal stresses may well be large enough to induce such fracturing. Furthermore, where enough fractures propagate and link up to provide an interconnected network, they can provide permeable pathways for fluid flow which can in turn lead to embrittlement and weakening of the rock. The thermal cracking was monitored by measuring the compressional (P) and shear (S) wave velocities through the samples both prior to and following heat-treatment. Both P-wave and S-wave velocities remained essentially constant up to 400^oC, with values of about 5.3 km/s and 3.0 km/s respectively. For higher temperatures the velocities decrease rapidly, so that by $800^\circ$ they decreased to about 3.4 km/s and 2.2 km/s repectively.

We presented results from a series of fracture toughness measurements on heat-treated specimens of Icelandic basalt using the ISRM recommended methodology. Fracture toughness at ambient temperature was 2.71 MPa/m², and this remained essentially constant up to 400^oC. There was then a very rapid decrease in fracture resistance between 400^oC and 600^oC, with relatively little change between 600^oC and the highest heat-treatment temperature of 900^oC. This pattern of behaviour was considered to be entirely consistent with the wave velocity data.

Finally, we presented the results of permeability measurements on heat-treated cores of basalt. The mean permeability of the basalt prior to heat-treatment was 9.4 nanodarcy (9.4 x 10^-21m²). Similar to the previous results, the permeability remained essentially constant after heat treatment to temperatures up to 300^oC, and showed only a slight increase after treatment to 400^oC. At higher temperatures, however, the normalized permeability changed dramatically, increasing by an order of magnitude at 700^oC and by a factor of 40 by 800^oC. We concluded that such a large increase in permeability was unlikely to result merely from the increase in the number or size of thermally-induced cracks, but from crack linkage processes above some percolation threshold to form extensive sample-spanning permeable pathways for fluid flow.

New experimental results

In our most recent programme, we have performed suites of experiments on both pre-heat treated (PHT) and non-heat-treated (NHT) specimens of basalt. The reason for this is that when a measurement is made at elevated temperature, there are two potential effects on the measurement, which act simultaneously. First, the high temperature can lead to thermal cracking and hence in a change to the microstructure of the material being tested. Second, the temperature can influence the actual deformation mechanism. By performing these two suites of experiments, we hope to be able to discriminate betwen these two effects.

 

Figure 43: Results of measurements of Young's modulus of elasticity for both PHT and NHT specimens up to 650^oC.

Figure 43 shows the results of measurements of Young's modulus of elasticity for both PHT and NHT specimens up to 650^oC (all PHT specimens were pre-heated to 750^oC). The two datasets show very similar trends and values, suggesting that the pre-heat treatment has relatively little effect on this parameter. As the temperature is raised from room temperature to 100^oC, there is a significant increase in the modulus. This may appear counterintuitive, but is consistent with previously reported data on deformation properties of brittle rocks over this temperature interval (e.g. Meredith and Atkinson 1985). It is considered to be due to thermal expansion leading to microcrack closure and hence higher stiffness and crack resistance. At all higher temperatures, the modulus decreases monotonically with increasing temperature.

 

Figure: Results of measurements of tensile strength up to $650^\circ$C for both PHT and NHT specimens.

Figure 44 shows the results of measurements of tensile strength up to $650^\circ$C, again for both PHT and NHT specimens. Here, the strength values for both datasets at first decreases for temperatures up to $250^\circ$C, and then actually increases for temperatures up to $650^\circ$C. Furthermore, tensile strength values for the PHT specimens are consistently higher than for the NHT specimens. The reasons for these variations are not fully understood at present, but we consider that it is likely to be due to local microplasticity at crack tips. Microplasticity is generally enhanced by high stresses, and hence is more likely to occur at crack tips than elsewhere in the rock because of the high stress concentrations at these sites. Microplasticity acts to blunt crack tips and therefore make crack propagation more difficult. If this is the case, then we could expect specimens to be less susceptible to tensile crack growth at higher temperatures. Furthermore, we might also expect the PHT specimens to exhibit higher strength because they have previously been subjected to a higher temperature ($750^\circ$C) and hence potentially to more microplasticity. This explanation is consistent with the observation of little difference in modulus between PHT and NHT specimens. The modulus is controlled by the mere presence of cracks and not by their growth, and hence crack tip blunting is not likely to affect modulus values.

These are very new results, and we will be investigating this possible explanation microstructurally with a combination of scanning electron and optical microscopy.