Influence of end effect on rock strength in true triaxial compression test

The influence of the end effect on rock strength in true triaxial compression testing was studied using a numerical approach. The influence of the intermediate principal stress (sigma(2)) on rock strength was isolated by using the two-dimensional Mohr-Coulomb failure criterion that depends only on the major principal stress (sigma(1)) and minor principal stress (sigma(3)). Thus, any enhancement to the rock strength with the increase of sigma(2) can be attributed to the end effect. It was shown that the end effect can result in an apparent sigma(2) effect, as long as the coefficient of friction (mu) at the rock specimen-steel platen contacts is not zero and the specimen in the sigma(2) loading direction is squat. When the strengthening due to the increase of sigma(2) predicted by a theoretical failure criterion was added to the strengthening due to the end effect, the results were in good agreement with the observed sigma(2) effect from some previous laboratory tests, indicating that the observed sigma(2) effect in true triaxial compression testing could be partially influenced by the end effect, particularly when sigma(3) was low. It is suggested to decrease the end effect to a level where the apparent sigma(2) effect is very small so that the obtained test results are more meaningful to characterize the actual sigma(2) effect.