Laboratory measurement of matric suction in pyroclastic soil using vacuum and high-suction tensiometers.

The hills in the urban area of the city of Napoli (Italy) are frequently affected by slope failures occurring in unsaturated pyroclastic soils. Previous studies have shown that these failures are mainly caused by loss of suction brought on by infiltrating rainwater (Nicotera 1998, 2000; Scotto di Santolo, 2000; Evangelista et al. 2002). To design landslide countermeasures and early warning systems, it is essential to monitor suction in the field. Vacuum tensiometers (range 0-80 kPa) were therefore installed at four slopes, the tensiometers being part of a broader set of instruments (psychrometers, tiltmeters, rain gauge, and radiometer). To assess the performance of vacuum tensiometers, a testing programme was carried out at the Geotechnical Laboratory of the Università di Napoli Federico II. The pyroclastic soil was placed in a caisson using wet tamping procedure. A vacuum tensiometer was then permanently buried in the soil and its measurement was compared with that of a portable tensiometer, inserted into the soil at regular intervals. To further investigate the performance of the vacuum tensiometers, suction measurements were performed on undisturbed pyroclastic samples at the Università di Trento using the TNT high-range tensiometer (0-1800 kPa). The paper examines the problems associated with matric suction measurement in coarser materials. As these are poor in clay fraction, it is not easy to make contact between the ceramic filter and the soil. Another problem addressed by the paper is the long-term response of the vacuum tensiometers, as air tends to come out of solution thus invalidating the measurement. Finally, measurements by the vacuum tensiometer are compared with those of the high range tensiometer and differences are then discussed.