The stability of a rock structure depends primarily upon the extent to which fracture develops within the structure. The prediction of the extent of potential fracture, the first step in the study of fracture development, involves a knowledge of both the stress distribution in the structure and the criterion which governs rock fracture under various stress conditions.
While a great deal of research effort has been devoted to the development of techniques for analyzing the stresses in mining and civil engineering structures 1-3 and to the investigation of the fracture mechanism of rock 4-6 relatively little attention has been paid to the application of available knowledge to the solution of practical engineering problems.
This paper describes a photoelastic technique for the determination of potential fracture zones in certain types of rock structures. A practical example, based upon a problem which might be encountered in designing the layout of deep-level mine excavations, is given as an illustration of the application of this technique.
In deciding upon the technique to be used for the solution of a problem involving the prediction of rock fracture it is necessary to consider: 1) What type of stress information is required to satisfy the fracture criterion for rock; and 2) How this information can most conveniently be obtained.
It has been established, both theoretically 4,7 and experimentally 5, 6, 8 that the fracture of hard rock depends primarily upon the relationship between the algebraically greatest and the algebraically smallest of the three principal stresses. Consequently, any stress analysis technique used in the prediction of rock fracture must provide information, either directly or indirectly, on the three principal stresses acting at each point in the structure.
In the case of structures of complex geometry, where theoretical stress analysis becomes difficult or even impossible, photoelasticity offers one of the most convenient means of obtaining the stress distribution in the structure. Unfortunately, the photoelastic isochromatic pattern, giving principal stress difference contours, does not contain sufficient information for the prediction of rock fracture. Hence, it becomes necessary to obtain further data for the separation of the principal stresses.
One of the most effective experimental stress separation techniques is to combine the photoelastic isochromatics with isopachics (contours of the sum of the two principal stresses in a plane) which can be obtained by means of the electrical analogy described in Appendix I.
Since the experimental data obtained from photoelastic and conducting paper models is in the form of the difference and the sum of the principal stresses, the most convenient representation of the fracture criterion for rock is that in which the principal stress difference is plotted against the principal stress sum at fracture.
The triaxial fracture data for a typical Witwatersrand quartzite, listed in Table I, is plotted in the form of a conventional Mohr diagram in Fig. 1.
