The majority of contact implementation within OptiStruct
involves the penalty-based process.
This can be activated using the CONTACT (PCONT, if
necessary) and TIE Bulk Data Entries. The TIE
entry supports both Penalty-based and Lagrange
Multipliers-based formulations. Both Node-to-Surface (N2S) and
Surface-to-Surface (S2S) discretization methods utilize this process to implement
contact between parts. The discretized contact elements are assigned variable stiffness
values (penalties) depending on the proximity of the contact surfaces to each other and
the type of contact forces on the element. The primary purpose of the variable stiffness
values is to minimize penetration of the contact surfaces.
The stiffness of each contact element typically depends on the type of analysis, linear or
nonlinear, that is performed. For Linear Analysis, the contact stiffness is constant
throughout the solution. The contact stiffness depends on the initial state of the
contact. If the contact is initially open, it remains open for the duration of the run.
The same goes for closed contacts. The contact status or the contact stiffness values
are not updated during the course of the solution for linear runs. The initial contact
stiffness of each contact element depends on the initial contact gap opening distance
(U0), which is calculated based on the positions of the Secondary
and Main (including contributions from the GPAD field, if any). For
open contact elements, a very small stiffness value of KB=10-14 * STIFF is used to avoid
numerical singularities.
Figure 1. Penalty-based Contact Element Force-Displacement Curves for Linear
Analysis
