GNGTS 2023 - Atti del 41° Convegno Nazionale

Session 3.2 ___ GNGTS 2023 As it can be seen in Tab. 1, even if SPT-based methods gave lower LPI values compared with DPT-and – based methods, they are all included in the range of “high” LPI, except for Cao et al. (2011) which gives a value in the “very high” range. When the fines content (FC) and the plasticity index (PI) are considered (case “B”), all methods showed LPI values close to zero or in the “low” range. Table 1 – Santa Rufina test site: comparison of LPI results obtained using only in situ tests and FC measurements (case “A”), and using in situ tests, FC and PI measurements, excluding interbedded layers (case “B”) (modified after Salvatore et al. 2022). In situ test Method LPI (case “A”) LPI (case “B”) SPT Youd et al. (2001) 5.9 0.0 Idriss and Boulanger (2008) 7.0 0.0 Boulanger and Idriss (2014) 8.9 0.0 DPT Cao et al. (2013) 13.5 0.3 Rollins et al. (2021) 12.8 0.2 VS Cao et al. (2011) 24.5 2.5 Rollins et al. (2022) 13.3 0.7 Conclusions In this work we proposed an innovative and multidisciplinary approach to liquefaction assessment in gravelly soils, using a combination of in situ and laboratory geophysical and geotechnical methods. The site of Santa Rufina (Sulmona, Central Italy) is characterized by an alluvial Upper Pleistocene – Holocene profile that consists both of sandy-silty and gravelly-sand alternating layers, with high vertical and lateral variabilities that introduce relevant uncertainties in the regional extensions (e.g., microzonation studies) of the liquefaction risk assessment at discrete vertical profiles with simplified methods. The liquefaction assessment based only on the measurements gives a “very high” LPI value using Cao et al. (2011), while the LPI value derived from the triggering curve proposed by Rollins et al. (2022) results “high”, such as DPT and SPT methods tested by Salvatore et al. (2022). Similarly, when liquefiable layers between interbedded cohesive layers (e.g., layers characterised by a “clay-like” behaviour according to well-known screening criteria) are excluded, the LPI value from Cao et al. (2011) remains greater than the LPI value from Rollins et al. (2022), which is