Numerical simulation of a free fall penetrometer deployment using the material point method

Free Fall Penetrometer (FFP) testing consist of a torpedo-shaped body freefalling into a soil target. The use of this type of device is becoming popular for the characterization of shallow sediments in near-shore and off-shore environments because it is a fast, versatile, and non-expensive test capable of recording acceleration and pore pressures. In recent years, the data analysis advanced considerably, but the soil behavior during fast penetration is still uncertain. Hence, there is a need to develop numerical models capable of simulating this process to improve its understanding. This paper proposes a numerical framework to simulate the deployment of an FFP device in dry sands using the Material Point Method (MPM). A moving mesh technique is used to ensure the accurate geometry of the FFP device throughout the calculation, and the soil-FFP interaction is modelled with a frictional contact algorithm. Moreover, a rigid body algorithm is proposed to model the FFP device, which enhances the performance of the computation and reduces its computational cost. The sand is simulated by using two constitutive models, a non-associate Mohr-Coulomb (MC) and a Strain-Softening Mohr-Coulomb (SSMC) that reduces, exponentially, the strength parameters with the accumulated plastic deviatoric deformation (Yerro et al., 2016) Variable dilatancy, which reduces as a function of the plastic strain, is also taken into account, and the strain-rate effects have been evaluated in terms of peak friction angle. In general, the behavior predicted by the MPM simulations is consistent with the experimental test. The results indicate that the soil stiffness has a big impact on the deceleration time-history and the development of a failure mechanism, but less influence on the magnitude of the peak deceleration and the penetration depth; the soil dilatancy reduces the FFP rebound, and the FFP-soil contact friction angle and the peak friction angle are highly linked to the peak deceleration. (C) 2020 Production and hosting by Elsevier B.V. on behalf of The Japanese Geotechnical Society.