Assessment of Finite Element Softwares for Geotechnical Calculations

Abstract: When building new constructions under ground level in cities it is important that the foundations of neighboring buildings are undisturbed. The regulations regarding soil deformations are hence very strict and conventional methods of predicting soil behavior is often too crude. Finite element calculations are therefore becoming an increasingly important tool used to predict soil behavior in and around a construction site. When modeling soil behavior today, specialized finite element softwares for analysis of soil and rock behavior are generally used. But as the need for numerical computations within geotechnical applications increase developers of more generalized finite element softwares enter the market. It is therefore of interest to compare how a generalized finite element software handles geotechnical problems compared to a specialized geotechnical finite element software. In this study an evaluation of how a general purpose finite element software, Comsol Multiphysics, performed in geotechnical simulations compared to two finite element softwares, Plaxis 2D and Z-soil, that where developed specially for modeling of soil and geotechnical calculations. As one important aspect to finite element calculations is the constitutive models used to capture the material behavior of the problem at hand. The two most widely used material models for simulating soil behavior are the Mohr-Coulomb criterion and the Drucker-Prager criterion, these material models have therefore been reviewed in this report. A simulation of a triaxial test has also been conducted to compare how these material models perform in the three softwares. In the work with this report a simulation has also been carried out to compare the functionality of the three softwares for a common geotechnical application. The simulation consisted of an idealized excavation problem with a tie back retaining wall. Modeling of a simple excavation project like the one in this report where the geometry and boundary conditions change over time is a bit more complicated and time consuming using Comsol Multiphysics compared to the two special purpose softwares. On the other hand, Comsol Multiphysics offers the user possibilities to take additional physical phenomena into account. Plaxis 2D was the software where a model could be established in the shortest time. However Plaxis 2D offers fewer options for the user than Comsol Multiphysics and Z-soil. Using Z-soil the user is given more complex alternatives to model constructions that in Plaxis 2D, but the software is on the other hand somewhat more difficult to learn. It also takes longer time to establish a model in Z-soil than in Plaxis 2D. From the calculations no major differences in soil deformation or forces in the ground anchoring could be identified from the three softwares, although some variations were observed.