Electromagnetic Radiation from Rock During Uniaxial Compression Testing: the Effects of Rock Characteristics and Test Conditions
5. Discussion

5.1 Classification of rocks type according to EMR characteristics during failure

The seven types of rock examined in this study can be classified as follows based on the characteristics of EMR events during failure.

1) Rocks that generate strong EMR events during failure (Group 1)
The Inada granite is in this category, as strong EMR events were detected for all samples around the failure point.

2) Rocks that generate weaker EMR events during failure (Group 2)
The Honkomatsu andesite and Kuzuu dolomite are in this category, as EMR events were detected in only 2 of 3 tests for each of these rocks, and the electric field intensities of the events were lower observed for Group 1.

3) Rocks that do not produce detectable EMR during failure (Group 3)
This category includes the remaining rock types; the Akiyoshi marble, Kimachi sandstone, Sanjome andesite and mortar. No EMR events above the noise level (65 dBƒÊV/m) were detected for any of these specimens. However, it should be noted that it is possible that EMR events were generated but at intensities below the level of noise.

The characteristics common to the members of each group are listed in Table 1 and are discussed here in relation to the physical characteristics of the specimens.

a) Uniaxial compressive strength
There is a clear difference between Groups 1 and 2, which exhibited uniaxial compressive strengths of over 200 MPa, and Group 3, which displayed strengths of below 110 MPa. In the present tests, higher uniaxial compressive strength tended to result in more intense EMR events.

b) Indirect tensile strength
The indirect tensile strength of Groups 1 and 2 was over 10 MPa, while that of Group 3 was much lower. Similar to the case for uniaxial compressive strength, it appears that the higher the indirect tensile strength, the more intense the EMR events tend to be.

c) Young's modulus
Groups 1 and 2 had Young's moduli of over 30 GPa, whereas the rocks of Group 3 (except for the Akiyoshi marble) had Young's moduli below 25 GPa. This result is consistent with those of Khatiashvili (1984), who observed that the intensity of EMR events increased with shear modulus for alkaline-haloid crystals (LiF, NaCl, KCl, CaCO3).

d) Poisson's ratio
Separate tests were necessary to evaluate the Poisson ratio, and there is much missing data. The obtained values varied from 0.21 to 0.27, and no patterns were seen within groups.

e) Schmidt hammer rebound hardness
Tests must be carried out on rock blocks before boring the specimens for use in the Schmidt hammer test. Since some of the blocks had not been pre-tested, there is much missing data. The available data indicate values of 68.6 and 67.0 for Groups 1 and 2, differing considerably from those for Group 3, 60.2 and 62.0. It appears that the Schmidt hammer rebound hardness, which is easy to perform in situ, has the potential to be useful as an in-situ indicator for classifying the tendency for bedrock to emit EMR.

f) Brittleness index (uniaxial compressive strength/indirect tensile strength)
Groups 1 and 2 exhibited brittleness indices of over 16, while Group 3 displayed values below 14.

g) Shape of stress-strain curve in post-failure region
Group 3 exhibited quite stable processes of failure in the post-failure region under constant strain rate conditions, whereas Groups 1 and 2 exhibited abrupt failure processes. Class II characteristics were observed for Groups 1 and 2 under linear combination control of stress and strain. Thus, it appears from the present tests that rocks exhibiting Class II characteristics or brittle deformation characteristics tend to emit EMR during failure.

5.2 Summary of Effect of Test Conditions on EMR Emission

The influence of test conditions, such as strain rate, control method, and sample end condition, on the emission of EMR was examined using the Inada granite. EMR events were generated under all conditions. Four conditions were examined; stress feedback gain ƒ¿ of 0 or 0.6, and loading rate C of 10-4 s-1 or 10-5 s-1. Figure 6 shows the relationship between the maximum field intensity of EMR events and the decrease in stress. Despite considerably scatter, larger stress drops produced larger electric field intensities under otherwise identical conditions. Mathematically, the relationship can be expressed as

electric field intensity = (stress drop)0.5 @@(2)

The magnitude of the stress drop is affected to a certain extent by the loading rate, although the difference between the stress drops at ƒ¿ = 0 and ƒ¿ = 0.6 was minor. The emitted EMR tended to be more intense at the higher loading rate of 10-4 s-1 compared to 10-5 s-1.

Changing a from 0 to 0.6 resulted in a lower stress drop for both loading rates, and the electric field intensity tended to be lower for the same levels of stress drop. This indicates that the EMR events were less intense during stable failure.