One thing that takes a while for the budding geotechs to digest is the difference between **undrained** and **drained** parameters, and when to use what. Actually, it is simple and is common sense. When a saturated clay is loaded, it will not let the water out immediately (i.e. remains undrained) and that is when most of the failures occur. In the short-term, the clay can be treated as an undrained homogeneous material where we will not separate the grains and water. Here, we carry out the undrained analysis in terms of total stresses, using undrained shear strength cu (phi = 0).

In the long term (after some months or years), the clay will drain out some water until the excess pore water pressure is fully dissipated and the pore water pressure is in equilibrium with the in situ conditions. Now, it is prudent to carry out an effective stress analysis using c’ and phi’, where we separate the stresses acting on the pore water (pore water pressure) and the grains (effective stresses).

Undrained analysis is often much easier to carry out, inexpensive to get the design parameters, and is necessary to assess the short-term stability which can be more critical than the long term stability. For undrained loading, the failure envelope in terms of total stresses is horizontal and hence we only need one parameter, undrained cohesion cu (phi = 0). Undrained cohesion can be derived from an unconfined compression test, UU triaxial, vane shear test (lab or field) or simply using a pocket penetrometer. Drained analysis needs c’ and phi’ which are derived from more expensive consolidated drained or undrained triaxial tests or in situ tests (and estimated using correlations). They are necessary when working with effective stresses. We generally charge A$1500-A$2000 for a CD or CU triaxial test on three specimens at different confining pressures, while an unconfined compression test costs less than A$100.

Granular soils drain very quickly, and hence they are always treated as drained and analysed in terms of effective stresses using phi’ (c’=0). For normally consolidated clays c’ = 0. Even for other clays (compacted or overconsolidated), c’ is not very large and is in the order of 0-25 kPa. Danish code suggests that c’ can be taken as 0.1 cu.

I*n summary, short-term analysis is carried out in terms of total stresses, using undrained shear strength parameters cu and phi = 0. Long-term analysis is carried out in terms of effective stresses, using drained shear strength parameters c’ and phi’.*