The crushing and permeability tests described in sections 4.7 and 4.9 have established definitions of three major coefficients which accurately charac- terize rock material: coefficient o\M which represents the dispersion of material compression test results, coefficient R10lR60, which characterizes the scale effect for uniaxial crushing strength tests on samples 10 mm and
0.5- 0.4- 0.3- 0.2- 0.1- 0
ojM
•2 D
4 •5
•
T 10 50 100 •fe
i
2—
1
6 0
"11
•2
•3 4•
•5
10 50 100
Fig. 4.26 Relative dispersion ajM versus percolation factor, S: (1) St Vaast lime- stone; (2) biotite gneiss; (3) fissured Jurassic limestone; (4) Malpasset gneiss (right abutment); (5) Malpasset gneiss (left abutment) Habib & Bernaix.
Fig. 4.27 Scale factor Rxo/Reo versus percolation factor, S: (1) St Vaast limestone;
(2) biotite gneiss; (3) fissured Jurassic limestone; (4) Malpasset gneiss (right abut- ment); (5) Malpasset gneiss (left abutment) Habib & Bernaix.
60 mm in diameter, and finally coefficient S discussed in the previous section.
Only a limited number of rocks have been sufficiently tested to allow the calculation of all three coefficients. However, Habib & Bernaix (1966) have been able to publish two diagrams of major interest (figs. 4.26 and 4.27).
They reveal a very narrow correlation between different rock characteristics from which it is apparent that the highly fissured Malpasset gneiss is in a class of its own.
Further detailed research on different rock types has shown that most of these properties can be explained by interpreting the theory of rupture of fissured rock by Griffith (section 4.11).
Rock fissuring starts with microfissures at the scale of the crystals. Minute dislocations caused by internal strains may exist in the crystalline mesh.
Extension of such microfissures can cause larger microfractures from which
Permeability and mechanical properties 69 rock fracture may develop. There are also some correlations between etch- pits, the minute dislocations and plastic deformations (d'Albissin, 1968;
Keith & Gilman, 1960).
The following testing technique has been described by d'Albissin (1968).
Small cylindrical samples 36 mm diameter and 72 mm high (or d = 10 mm, h = 20 mm for single crystals) are tested in a triaxial test rig capable of developing a lateral pressure of 1000 kg/cm and an axial thrust of 8500 kg/cm2, causing strain and deformation of the samples.
165 200
temperature (°C)
Fig. 4.28 Tests on marble of Mosset. Variations of the natural thermoluminescence with lateral pressure: (1) intact marble; (2-5) later pressures in kg/cm2: (2) 245, (3) 390, (4) 980, (5) 5000; (6) crushed under pressure of 5000 kg/cm2 (after d'Albissin, 1968).
These internal dislocations can be detected by thermoluminescence whereby electrons falling from one energy level to a lower one produce light. Crystal dislocations can trap electrons and modify the light emitted. Incident gamma radiation is reflected at another wavelength.
The strained rock samples are reduced to powder, the dimensions of the grains being about 250 to 315 jum (200-315 /urn for single crystals) and are then irradiated. The emitted light measured from highly strained and less strained samples at different temperatures (figs. 4.28 and 4.29) gives a measure of the internal dislocations within the rock material. Further correlations were established from other tests involving acid corrosion of polished rock surfaces.
The curves in fig. 4.28 show that a moderate strain causes an increase of the thermoluminescence which then decreases with higher strains (Handin
5 £
100- 75- 50- 25-
2000 4000 6000 -1 kg/cm2
Fig. 4.29 Tests on marble of Mosset. Area under the thermoluminescence curves of Fig. 4.28 plotted against average pressure in kg/cm2. Grains of 250-315 pm (after d'Albissin, 1968).
et al., 1957; d'Albissin, 1968). Plastic deformations occurred at lateral pressures of 980 kg/cm2 and 5000 kg/cm2 and temperatures of about 280 °C.
Additional permeability tests with air have been devised to detect any incidence of chemical change in the rock material (Perami & Thenoz, 1968).
Air can penetrate through minute fissures which would not normally accept any water. As the air does not react chemically with the mineral rock materials it does not cause alterations to the rpcks. Perami & Thenoz (1968) introduced a void coefficient:
k
where V = dry volume of the rock sample reduced to powder and vp the volume of the pores. This formula yields:
The values V and V + vp are obtained by measuring the apparent specific weight of the rock sample and the true specific weight when reduced to powder. (The usual technique of Serafim measures the sample wet and after drying for 24 or 48 hours at 105 °C.)
Tests on several granites have shown that when the permeability to air is below a certain limit there is no chemical change in the rock. With higher permeability, granites can be chemically altered by percolating water. The same tests have shown (fig. 4.30) that there are two phases in the permeability curve. During a first phase of increasing uniaxial pressure, the sample is consolidated and the ratio KjKQ < 1 (Ko = permeability factor for no load). When the pressure increases beyond a iimit causing microfissuration' the K/KQ ratio increases rapidly, showing that intensive microfissuration occurs.
Rock fracture 71
1600- 1200- 800- 400-
kg/cm2
microfissuration icro
rupture limit where microfissuration
consolidation of rock I sample
i P • I • • • ' * M
1 2 3 4 5 6 7 8 9 10 11
Fig. 4.30 Increasing uniaxial loading causes the granite sample to consolidate at first. Beyond a certain load, microfissuration starts and the relative permeability ratio, KJKQ, increases (after Perami & Thenoz, 1968).