ABSTRACT
The strength of rock is a very important parameter for a number of engineering applications. In practice, isotropic failure criteria are widely used to characterize rock strength. For example, the Mohr-Coulomb criterion and the Hoek-Brown criterion are among the most widely used ones. These criteria assume that the rock is isotropic, and that the intermediate principal stress does not affect strength. This is not surprising since a vast amount of experimental data on rock is obtained from conventional triaxial tests, in which two principal stresses are kept the same. In reality, however, the loading/unloading stress path in engineering practice is complicated, and two identical principal stresses are usually not guaranteed. It is therefore necessary to perform true triaxial tests to examine the behavior of rock under such complicated stress paths. Indeed, many true triaxial tests were performed, as reported by Hojem and Cook (1968), Mogi (1969, 1970), Michelis (1985), Hunsche and Albrecht (1990), Smart (1995), Wawersik et al. (1997), Haimson and Chang (2000), Chang and Haimson (2000), Oku et al. (2007), etc. As a result, the effect of the intermediate stress on rock strength is better understood. Failure criteria that can describe rock strength under true triaxial stress state were also reported, for instance, Mogi’s criterion (Mogi, 1971), Lade’s criterion (Lade, 1993), the Mogi-Coulomb criterion (Al-Ajmi & Zimmerman, 2005), and the modified Lade criterion (Ewy, 1999). These criteria are still isotropic, however, as principal stresses are the only variables.
