17-06-2021 | Posted by Joaquín Martí
Out of the many materials involved in our constructions, the ground is, by far, the more troublesome one. And it is the only material featured in absolutely all our structures; at the very least, their gravity loads must be transferred somewhere.
Note that most other materials, like concrete and steel, are man-made according to stringent specifications and, with considerable approximation, we know their behaviour and what they can and cannot provide. But the ground is already in place, we simply inherit it from nature and, apart from some improvement measures, all we can do is to try to evaluate its properties and take them into account when designing our structures.
To arrive at those properties, we typically drill boreholes and dig trenches, take some discrete samples to the laboratory, test them, and try to construct a 3-D picture based on those local observations. Such procedures would clearly suffice if we were dealing with a reasonably uniform and well-behaved material, but reality is very different.
The ground is made of several phases, at least solids and liquids and, in many cases, also gases. For simplicity, we would hope for linearity of the stress-strain behaviour, but normal soils depart from it after ridiculously small strains, on the order of 10-5. That complicates matters enormously and, as if this were not enough, the behaviour frequently incorporates other features such as plasticity (both deviatoric and volumetric), anisotropy, strain softening, dependence on past history, rheology, etc., let alone the idiosyncrasies of its cyclic and dynamic behaviour.
To circumvent the variability of the natural ground, the complexity of its behaviour and the vagaries of our evaluation of its characteristics, our structural design calculations normally incorporate larger safety factors for the ground than for the rest of the materials. This strategy, however, is useless when we are trying to interpret a failure that already took place, a task that cannot be performed using conservative properties, but requires taking the actual behaviour into account.
In spite of these complexities, or perhaps because of them, Principia has traditionally maintained very strong simulation capabilities in geotechnical engineering. Indeed, about 15% of our projects involve major activities in geotechnical modelling.
Sometimes this arises because of pile foundations for bridges, tanks, wind turbines, even a tunnel, under static and dynamic conditions. And shallow foundations must also be studied in relation with their settlements and stability.
Other times it is mines that require the studies, whether open pit or underground, or their tailings impoundments. Or it is the stability of earth or rockfill dams, particularly under earthquakes, or that of underground repositories for nuclear waste, or that of tunnels using various excavation procedures. At times, the question relates to the potential failure of rock or soil natural slopes. The investigation of geotechnical accidents, like the Esterhazy mine in Canada or the failure of the Aznalcóllar tailings pond, poses particular difficulties because of the modelling precision that it requires.
Earthquake loading motivates many of our geotechnical activities: evaluation of seismic hazard and risk, study of the site effect, evaluation of the liquefaction potential, dynamic behaviour of shallow and deep foundations, dynamic soil-structure and soil-water-structure interaction, etc.
In summary, the ground is indeed a very difficult material that all projects inevitably incorporate. And the most robust structure will collapse if its ground support fails.
