Author:
Monforte Lluís,Arroyo Marcos,Gens Antonio
Abstract
Sampling disturbance is a perennial problem of geotechnical site investigation. The structure of soft soils is particularly fragile and prone to disturbance during sampling. The extent by which different sampling procedures modify soil structure has been generally assessed on an empirical basis, observing the outcomes of tests performed on specimens obtained with different sampling technologies. That kind of experiment is slow, costly and laborious. In this communication we present a procedure to perform similar experiments “in silico”, by simulating the sampling procedure, examining the effect of tube sampling on the initial soil structure, obtaining sub specimens from the simulated core and testing them afterwards in simulated triaxial undrained compression and in the oedometer. The simulations are based on G-PFEM, a code developed for soil-structure interaction at large strains. The numerical experiments clarify the origin of sampling related property variability.