Laboratory Tests on Gravel-Rubber Mixtures (GRM): FEM Modelling Versus Experimental Observations

In the field of earthquake engineering, many studies have been conducted to mitigate dynamic effects on structures, adopting alternative solutions including Geotechnical Seismic Isolation (GSI) systems. The main idea is of improving the mechanical properties of the foundation soil and dissipating the seismic motion before reaching the structure. Among several materials proposed for use in GSI systems, Gravel-Rubber Mixtures (GRMs) represent an effective and eco-friendly solution. Laboratory and in-situ tests performed on GRMs have highlighted their excellent static and dynamic mechanical properties. In addition, the rubber particles adopted for GRMs are recycled from End-of-Life Tires, whose disposal is a critical environmental and socio-economic issue, worldwide; therefore, their use could help to partially solve this problem. This paper presents a non-linear finite element numerical study to simulate the behavior of GRMs undergoing monotonic drained triaxial tests. The non-linear GRMs response is simulated through a hardening elasto-plastic constitutive model, namely the Hardening Soil model with Small-Strain Stiffness. In addition, the elastic-perfectly plastic Mohr-Coulomb model, which is commonly used in many numerical codes, is also used for comparison. The calibration of the model parameters is a key point of the performed analyses; it is carried out comparing the numerical results with available experimental data, performed on GRMs with different volumetric rubber content and confining pressures. The calibrated numerical models represent a useful tool to investigate geotechnical problems involving GRMs (e.g., soil-GRMs-structure interaction), highlighting their adequate constitutive modelling.