Dynamic Probing as an easy and practical tool in geotechnical characterization of shallow landslide source areas: results, problems and perspectives in Northern Tuscany (Italy)

In Northern Tuscany intense and long time rainfalls produce many shallow landslides, mostly soil slip-debris flows, which often cause victims and severe damage. Aiming at contributing to the characterization of the source areas of shallow landslides, this research deals with the geotechnical characterization, by means of low cost dynamic penetration tests (Dynamic Probing, DP), of the debris slope cover of the Macigno Fm. (Tuscan Nappe unit of the Tuscan Domain). This formation, composed of quartz and feldspar sandstones altering with layers of siltstone, constitutes the main Apennines ridge and underlies hundreds of square kilometres of Northern Tuscany. Most of shallow landslides triggered by critical rainfall events really involve the slope cover of the Macigno Fm., which mainly consist of sand and gravel with a minor portion of pelite. The knowledge of physic-mechanical properties of this material is fundamental to understand the triggered mechanisms of the shallow landslides. Therefore, in order to determine the typical values of relative density (Dr) and shear strength angle (φ’), the results of 220 tests were analysed. These tests were subdivided into 91 Medium Dynamic Probing (DPM), 41 Super Heavy Dynamic Probing (DPSH), 74 Standard Penetration Tests (SPT) and 14 samples with direct shear tests. The penetration resistance obtained by DPSH and DPM tests was related to the strength parameters of the debris cover by means of empirical relations. These relations consider the most typical soil categories (silty sand, gravelly sand and sandy gravel) and take into account for the Energy Ratio of the instruments used and the stress conditions of sites. Distribution and variability of shear strength angle and relative density were analysed and related to soil type, test type and probing depth. In general, the Dr and φ’ values show an increase in scattering around the mean values with the increase in grain size and sorting; DPM tests show this trend more clearly then DPSH tests. The presence of pebbles in a silty sand material, even in low percentage, probably can cause a significant increase of N (number of blows, coming from DP), and consequently in the Dr and φ’ values. Furthermore, at the low depth commonly investigated by Dynamic Probing (usually less then 5-6 m in order to investigate the cover materials involved in shallow landslides), Dr and φ’ of each soil type do not show significant increase with depth. In order to validate the results, the values of Dr and φ’ obtained by DP tests were also compared to the results obtained in similar deposits by SPT. This comparison showed small differences: in detail, the SPT provided slightly higher friction angles than DP tests and therefore they seem to be less conservative. Moreover, the φ’ values obtained by DP tests were compared with those resulting from direct shear tests (when available), performed on undisturbed samples taken in the same soil type, close to the DP test sites, at the same depth and stress conditions. The comparison evidenced a good linear correlation between the values resulting from these different tests, so confirming the satisfactory and repeatable results of DP tests. Therefore the DP could be an effective, simple, practical and inexpensive tool for the geotechnical characterization of potentially unstable slope covers. In this way actual and reliable parameters may be achieved, on which shallow landslide susceptibility and hazard maps can be based.