Determination of in-situ engineering properties of soil using an inverse solution technique and limited field tests

Abstract The goal of this research is to develop a response surface based inverse solution technique to determine in-situ engineering properties of soil for use in mobility and traction prediction models from the force displacement data generated by a cone penetrometer device. The nonlinear elasto-plastic behavior of soil was characterized by a six-parameter constitutive model – two elastic parameters (i.e., bulk modulus, K and the Poisson’s ratio, υ ), three plastic parameters (i.e., angle of internal friction φ , cohesion c , and soil hardening parameter λ ), and one soil physical condition parameter (i.e. initial void ratio, e ∗ ). Soil failure was represented by the Drucker-Prager yield criterion and associated flow rule. LS-DYNA FEM software package was used to model the soil-cone interaction problem. FEM simulations were conducted for a set of soil properties properly selected within the defined parameter space. The analysis of FEM simulations indicated that the cone penetration force-displacement curves could be represented by two piecewise smooth functions – a parabola followed by a straight line. The coefficients of these two curves were used to create fourth order response surfaces using a stepwise multiple linear regression technique. The results showed both cohesion and soil hardening parameter could be predicted using this methodology.