Getting high-quality samples in ‘sensitive’ soils for advanced laboratory tests

Laboratory tests are well recognized as highly appropriate for defining the engineering properties of geomaterials, in terms of constitutive law parameters for modeling geotechnical engineering problems. The strong development of advanced techniques, both in equipment and in data interpretation, has increased the confidence in laboratory testing, while on the other hand the limitations due to the quality of soil sampling with depth and the spatial representativeness of the samples are less consensual. Still, the development of new methods for assuring high-quality samples is increasing, together with sampling quality assessment by non-destructive methods using vibration wave velocities. Interpretation methods of in situ tests for ground characterization have also evolved significantly, increasing the reliability of these methods. Their versatility to cover large areas on site and the fact that these tests are, in principle, performed at the actual state (physical and stress) conditions, as well as the improvements in the correlations between field tests and hydraulic and geomechanical parameters, allow joining the quality of data and theoretical approaches, namely through critical state soil mechanics. This keynote paper discusses some of the aspects that can and should enable the association of ground characterization from laboratory testing over undisturbed samples used in more or less advanced tests, enhancing the determinant conditioning factor, that is, the sampling technique to get representative specimens and the way this is assessed. The confidence that we expect to have on the geomechanical parameters that we need for our geotechnical activities will mostly depend on this in view of the high uncertainties of the parametrical correlations with in situ test data, therefore, important in ground characterization. This is especially relevant in sensitive soils, such as soft fine soils, loose sandy soils, or young residuals soils. These have or can have “weak” equilibria of the interparticle micro- and macrostructures (or their arrangement, fabric) that will change substantially their properties if samples are collected and conditioned with processes that do not preserve that intrinsic “ADN”. The change in these natural conditions can be evaluated by techniques of quality assessment, which will be discussed in what follows. © 2017, Springer International Publishing AG.