The significant effect of the intermediate principal stress on rock strength has been recognised for many years, leading to the development of a number of polyaxial peak strength criteria for rock. Recently, the Mogi-Coulomb and Zhang-Zhu polyaxial peak strength criteria, which can be defined by the conventional Mohr-Coulomb and Hoek-Brown strength parameters, respectively, have been developed. In this paper, we examine how Mohr-Coulomb and Hoek-Brown rock strength parameters determined from standard triaxial tests () can be directly applied to these polyaxial peak strength criteria. 32σ=σ Through a re-evaluation of published polyaxial peak strength data, for fifteen dif-ferent rock types, we examine the error between the strength measured under poly-axial stress conditions and the strength predicted by polyaxial strength criteria using strength parameters determined from triaxial tests. As a result, we conclude that the Mogi-Coulomb and Zhang-Zhu polyaxial criteria provide estimates of peak strength with mean errors generally less than ±5%. We also show that as the stress state devi-ates from triaxial, i.e. becomes progressively more polyaxial, these criteria provide better estimates of intact rock peak strength than do the corresponding triaxial criteria.
The effect of the intermediate principal stress on the peak strength of rock is known to be significant. That is, if is held constant and increased from triaxial compres-sion (i.e.) to triaxial extension (i.e.), the peak strength increases to a maximum at some intermediate value of before decreasing to a value higher than that obtained in triaxial compression (see Figure 1). This variation in strength is often a substantial proportion of the uniaxial compressive strength, suggesting that, in those engineering applications where the induced state of stress is not triaxial, the applica-tion of polyaxial peak strength criteria is warranted (Bedi & Harrison, 2012).
