Static strength has traditionally been considered as rock strength in the foundation bedrock for seismic design in Japan. The appropriate evaluation of dynamic strength instead of static strength became necessary for dynamic response analysis following the Great East Japan Earthquake. Therefore, we previously proposed a mathematical model to evaluate the dynamic strength for intact rocks. In this study, we performed laboratory tests and validated the model for discontinuity by using an artificial rock. As a result, the dynamic strength exceeded the static strength in the experimental and calculated results. However, the influence of fatigue and loading rate were less, compared to the cases of intact rocks. Therefore, the dynamic strength of the discontinuity does not greatly differ from static strength.
In seismic design for the foundation bedrock of nuclear power plants in Japan, static strength has traditionally been used as rock strength, based on the fact that "dynamic strength ≧ static strength". This relationship has been validated for various rock types (Nishi & Esashi 1982, Yoshinaka et al.1987, Sugiyama et al. 2001). However, those test conditions are limited. Dynamic strength does not have a clear definition. Besides, it is difficult to formulate a single definition of dynamic strength because seismic waveforms are diverse, and the stress waveform inside the ground also varies depending on the location. To resolve these issues, we proposed a mathematical model to evaluate dynamic strength (Okada & Naya 2018). Using the mathematical model, the dynamic strength can be obtained by monotonic loading tests performed at several loading rates in different orders and cyclic loading tests at several different shear stress amplitudes. In the previous research, we performed laboratory tests to validate the model, using intact rocks, but we have never conducted the test using rock joints.
In this study, monotonic loading and cyclic loading direct shear tests were conducted on the artificial rock joint we made from plaster in order to obtain the parameters of the mathematical model. Afterwards, multistep direct shear tests under cyclic and seismic-wave loadings were performed, followed by simulations using the mathematical model. The test results showed that the dynamic strength exceeds the static strength, corroborating previous research. Furthermore, the dynamic strength calculated from the mathematical model of the artificial rock joint was generally consistent with the dynamic strength obtained from the experimental data.