抄録
The concept that the compressive strength of rocks usually falls within the range of two or three times the strength of its minimum value seems to be accepted in engineering fields in Japan. The authors suggest that such a wide distribution of the strength was from the experimental conditions. Carefully arranged experiments (experiments: No. 1) and normal experiments (No. 2) were carried out. Although the same granite specimens were used, the scattering of the data in No. 1 was very small but the data in No. 2 was scattered widely. The applied bending moment was calculated from the strain deviation of four axial strains at the periphery of each cylindrical specimen. The observed strength decreased with increasing bending moment. The strike of the fault was nearly parallel to the rift plane for the carefully arranged experiments. However, in experiments No. 2, the strike varied widely, showing that the intrinsic nature of the rock can not be easily discovered because of the disturbances of the experimental techniques.
