FINITE ELEMENT MODELLING OF THE DIRECT SHEAR TEST

Resumo

The direct shear test was developed to obtain the failure envelope through a simple and
direct analysis. It is one of the oldest and simplest tests for this purpose applied to soil analysis.
However, during the direct shear test, the stress distribution in the sample is complex and not uniform.
In order to improve the interpretation of the direct shear test results, the present work aims to study the
stresses and strains in the sample during the execution of the direct shear test using a model based on
finite elements (FEM). The Finite Element model was implemented in the software ABAQUS®
,
student version. The model is a two-dimensional representation of a section in the direction of the
shear force. The shear is considered by applying displacements in a steady state analysis. For the
material properties a Mohr-Coulomb elastoplastic model was considered. An evaluation of the chosen
model was realised by a direct comparison of model results and laboratory tests, showing in general a
good agreement. To evaluate the adherence between upper and lower plates to the soil specimens two
geometric models were conceived: one considering toothed plates and another with smooth plates. It
was found that the models were able to describe the typical stress and strain curves of the laboratory
direct shear test when the results of shear are obtained from the fixed box reaction, although, with
some discrepancies between the toothed plates and smooth plates models. The differences observed
can be explained observing the failure zone which is distinct in the models: for the toothed plates
model, the failure zone as expected, occurs in the central part of the model; as for the smooth plates
model the failure zone falls out the central region. Differences can also be highlighted when the failure
envelope is reinterpreted from the fixed box reaction, smooth plate models feature slightly smaller
parameters, reinforcing the need for good plate crimping.