Guidelines for Plasticity, Damage and Failure Models
The convergence for nonlinear materials can be bad if the simulation settings are not set up properly. Therefore, this section contains some best practices for these simulation types.
Note! If possible, do not use the Force Controlled load type. |
With the Force Controlled load type, the applied load might be too high for the materials in the structure: For plastic materials and for the damage or failure models, the maximal stress is limited by the parameters of the material model. This can leads to unsolvable equations.
If it is important to run the simulation with a force controlled load (as e.g., in a cyclic load case, where the stresses should decrease to zero again), we recommend one of the following workarounds
Nevertheless, even if the maximum allowed load is not reached, the convergence might be slow. For example, when using the exponential hardening model, a load slightly lower than the maximal allowed load might lead to strains near to infinity (which is also not realistic).
The main difference between the damage model and the failure model is that the failure model decreases the stiffness suddenly, which introduces a discontinuity into the stress-strain relationship. The damage models lead to a gradual reduction of the stiffness. Thus, the damage models introduce no discontinuity, this means that the simulation is more stable and usually faster. Nevertheless, this gradual reduction of stiffness can be arbitrarily fast – this means that nearly the same results as with the failure model can be achieved. Additionally, the damage models can be combined with other models like e.g., plasticity.
The main advantage of the failure model is that it is easier to set up, since it contains less parameters. Additionally, the damage models are not available for the transverse isotropic case. Altogether, it is strongly recommended to use the damage models instead of the failure model, whenever possible.