2013 Volume 129 Issue 10_11 Pages 635-641
In the previous paper, based on the summarized experimental results, long-term strength (failure criterion at extremely low loading-rate) was proposed. The simplest form of the long-term strength was σ1 = aσ3 where σ1 and σ3 were major and minor principal stresses, and constant 'a' varied from 2 to 6 depending on rock. For validation of the proposed failure criterion, rock stresses were plotted on σ1 and σ3 diagram. It was found that 86 data except one were plotted under the line (σ1=4σ3). After the previous paper was published, Okubo et al. conducted push-in test in which a failed rock sample was compacted in a steel cylinder. The envelope of Mohr's stress circles through the test were approximated byτ= Aσ0.7. Under this curve, it was considered that strength of the failed sample was recovered up to a certain extent. More than 150 data of rock stresses were newly added to examine the long-term strength. It was found that almost all Mohr's circles were plotted under the curve τ = 2σ0.7. This was the better result than the previous one, σ1=4σ3. As stated in the previous paper, it is reasonable to consider that, in relatively stable rock mass, rock stress is maintained at a certain value lower than a long-term strength. If so, this result supports the validity of the newly proposed long-term strength, τ = 2σ0.7.