Comparison of Lagrangian, SPH and MM-ALE Approaches for Modeling Large Deformations in Soil

Ongoing research at the USDOT funded Transportation Research and Analysis Computing Center (TRACC) at Argonne National Laboratory on bridge stability for bridges with piers in scour holes relies greatly on LS- DYNA® capabilities for modeling large deformations in soil and fluid structure interaction. When it comes to soil modeling, material model MAT_005 Soil and Crushable Foam is often used as a first approximation. It is especially useful when little material characterization is performed--which is usually the case for riverbed soil. Although an abundance of reports can be found where MAT_005 was used with the Lagrangian approach, its use in Smoothed Particle Hydrodynamic (SPH) and Multi Material Arbitrary Lagrangian - Eulerian (MM-ALE) approaches is considerably less documented. This paper presents a comparative study of the performance of MAT_005 with Lagrangian, SPH and MM-ALE approaches for predicting large deformation soil response. For the purpose of validation, simulations were performed for the in-situ experiment of a steel loading pad penetrating into silty clay sand. Using LS-OPT® metamodel based sensitivity analysis was conducted to identify the most relevant material and loading parameters. The results show that the three formulations can produce reasonable predictions at large penetrations. Although very suitable for soil penetration problems, MM-ALE requires iterative adjustments of contact parameters to eliminate spurious leakage. The LS-OPT sensitivity analysis on material parameters indicates the yield function parameters have the greatest influence on the results. Further important parameters are the soil density and appropriate modeling of the soil island boundaries. It was also noted that the choice of the type of domain decomposition greatly affects the compute time.