Most soil models in PLAXIS use the Mohr-Coulomb failure criterion in combination with tension cut-off to describe ultimate stress states in the soil. This is because the Mohr-Coulomb criterion is a realistic and well-accepted shear failure criterion for soils. Other criteria exist, which may be slightly better for some soils, like Matsuoka-Nakai or Lade, but these criteria may as well lead to non-conservative results compared to Mohr-Coulomb, and they are not compatible with conventional methods for soil bearing capacity or safety factor analysis. Hence, by using the generalized Mohr-Coulomb failure criterion, PLAXIS uses a reliable and consistent formulation of soil failure.

When using the Mohr-Coulomb failure criterion in combination with undrained shear strength (c = s_u and φ = 0, i.e. undrained B or undrained C), the criterion turns into the well-known Tresca criterion, so PLAXIS implicitly includes the Tresca model as well.

In addition, criteria exist that are NOT good for soils in general, such as Drucker-Prager and Von Mises. These criteria may only give acceptable results for particular homogeneous stress states, but not in arbitrary stress paths that generally occur in 2D or 3D applications. Therefore, we do not recommend using these criteria or models, and that is why such models are not available as standard soil models in PLAXIS.

Different failure criteria