Modeling Vacuum Cconsolidation with Finite Elements

In conclusion we can summarize the procedure and arrangements needed to model vacuum consolidation in Plaxis. After the basic model has been created following the standard procedure to implement any model in Plaxis, vacuum loading can be modeled through a groundwater boundary conditions, i.e. imposing an head. The groundwater head can be imposed on a line or else on a drain. In our examples we always used an head h= -10 m, but the value that has to be imposed actually depends on the efficiency of the apparatus.

Some remarks to keep in mind when dealing with vacuum consolidation in Plaxis are:

* The soil subjected to the vacuum loading has to be set as always fully saturated, because the use of a water retention curve lead to an high reduction in permeability when suction is applied. This will create a kind of impermeable layer that will stop the consolidation process.

* At the end of the calculation the phreatic level displayed by the program is not the real phreatic, i.e. the surface where the air and the water pressures are equal. In fact the phreatic level computed by Plaxis is just based on the steady pore-water pressure (or on the water pore pressure for fully coupled analysis), but doesn’t take into account the drop of air pressure.

* Furthermore, the consideration of a not physical phreatic level will lead the program to not take correctly into account the soil weight. For a real fully saturated soil this problem can be easily solved changing manually the material weight for the dry soil and put it equal to the fully saturated weight, but for partially saturated soils this problem cannot be solved so easily.

* Even if suction appears in Plaxis output this is just a matter of how suction is defined in the Plaxis kernel. This has not a physical meaning and there is not tensile stress into the pores when the soil is under the real phreatic level.

* When drains are used, the users must set their behavior as vacuum. In fact, if the behavior is set as normal, the drainage above the displayed phreatic level is not allowed. This lead to inconsistent results for our purpose of study.