Granular soils (sand and gravel, and maybe some silt) have very little interparticle interactive forces due to rather larger particle sizes. Therefore, most of their shear resistance comes from a frictional component. Thus, c can be assumed to be zero for granular soils. In addition, these soils have a rather high permeability, so the pore water pressure will normally dissipate very quickly in most applications. Accordingly, for both wet and dry granular soils, Equation is used without the “c” component:
The failure envelope starts at the origin of the σ–τ diagraph with φ angle inclination, and the angle of internal friction φ is the sole parameter to determine the shear strength. The φ values are influenced by soil’s various properties, such as density or void ratio, gradation (uniform or well graded), angularity (rounded or angular), grain surface roughness, etc. Among these, density (or void ratio) would most significantly influence the φ values of granular materials. The typical values of φ are given in Table 1.
A straight line failure envelope or a constant φ angle for a given soil with a given density is generally assumed for granular soils as discussed. However, in reality, it was empirically observed that the failure envelope for granular soils is slightly curved, as seen in Figure 1. This implies that a slightly higher φ angle is at a lower confining stress and a lower φ angle is at a higher stress. The importance of this fact is addressed by a non conservative application of small-scale model test data in the evaluation of in-situ earth structure behavior. In a small-scale model test, the stress level, which mostly comes from the gravitational force of soils, is small, and thus the failure phenomenon is controlled by a rather high φ value, while in an in-situ earth structure, the stress level is high, which provides a lower φ value.
The geotechnical centrifuge model test is meant to overcome the previously mentioned shortcomings of the small-scale model test.
For example, if a 0.5 m high model earth dam is subjected to 20 g of the centrifugal gravity on a rotating platform, its stress level in the model increases to the level of a 10 m (0.5 m × 20) high earth dam, and thus similar φ values would be utilized in both the 0.5 m high model under 20 g centrifugal force and the 10 m high prototype under 1 g conditions.
Readers are referred to other references (Taylor 1995) for details of geotechnical centrifuge testing.