The geomechanical models for slope design, which was mainly based on shear strength parameters estimated by using RMR classification system, are compared with the real post Ð construction conditions of cuts. The results of this study show that RMR classification system generally leads in a rock mass quality underestimation, which mainly due to the rating adjustment for discontinuity orientations as especially proposed for rock slopes. The case records evaluation indicate that the rock mass rating without taking into account the discontinuities adjustment give a more realistic rate of rock mass shear strength characteristics for preliminary slope stability analysis. Also, the shear strength parameters, as determined by the use of GSI, seems to lead in an overestimation of slope stability of these limestone cuts.
The determination of representative shear strength parameters of a rock mass is often a time consuming, costly and difficult process. Large-scale in situ shear tests are desirable but very expensive. It is known that there are the following lines of approach that can be used to obtain information on in situ behaviour of the rock mass for slope design:
direct measurements by means of field and laboratory tests,
evaluation by means of mathematical simulations,
evaluation of stable and unstable natural and excavated slopes and,
empirical methods using rock mass classification systems.
The last approach is often applied in many cases for the preliminary design of high rock slopes in Greece and mainly for heavily jointed rock masses assuming isotropic rock mass behaviour. The Geomechanics Classification Rating (RMR) system developed by Beniawski (1976, 1979, 1989) is the common way to obtain the range of shear strength parameters and then to estimate the safety factors by using slope design charts for circular failure surface or other calculation methods (Hoek and Bray, 1974).