Numerical Controls for General Contact in Abaqus/Standard

Numerical controls associated with the general contact algorithm in Abaqus/Standard:

  • should not be modified from their default settings for the majority of problems;

  • can be used for problems where the default settings do not provide cost-effective solutions;

  • can be used to control the main-secondary roles and the sliding formulation; and

  • in some cases can be applied selectively to particular regions within a general contact domain.

This page discusses:

Contact Formulation

By default, the general contact algorithm uses the finite-sliding, surface-to-surface contact formulation, which is discussed in Contact Formulations in Abaqus/Standard. In addition, it is supplemented by the edge-to-surface, edge-to-edge and vertex-to-surface formulations, which are also based on the finite-sliding tracking approach. Optionally, you can specify the small-sliding tracking approach for portions of the general contact domain and, by extension, the entire general contact domain.

Numerical Controls for Friction

Numerical controls associated with friction are discussed in Frictional Behavior.

Main and Secondary Surface Roles of a Contact Formulation

The surface-to-surface contact formulation used by general contact generates individual contact constraints using a main-secondary approach, as discussed in Contact Formulations in Abaqus/Standard. Abaqus/Standard assigns default pure main-secondary surface roles of a contact formulation for contact involving disconnected bodies within the general contact domain. Bodies consisting of connected beam and truss elements are considered disconnected bodies even though these bodies may share nodes with other faceted bodies. Internal surfaces are generated automatically using the naming convention General_Contact_Faces_k, where k corresponds to an automatically assigned component number. By default, the lower-numbered component surfaces act as main surfaces to the higher-numbered component surfaces. An exception is when component surfaces consisting of beam and truss elements interact with faceted component surfaces in the edge-to-surface contact formulation. A component surface consisting of beam and truss elements acts as the main surface in the edge-to-surface formulation if half of the average element radius is larger than the average smallest facet length of the faceted component surface. Self-contact within a body is treated with balanced main-secondary contact by default, with each surface node acting as a main node in some constraints and as a secondary node in other constraints.

For example, if the general contact domain spans three disconnected bodies, the following three internal “component-surfaces” for general contact are created automatically:

  • General_Contact_Faces_1

  • General_Contact_Faces_2

  • General_Contact_Faces_3

By default, the first surface listed acts as a main surface to the other two, and General_Contact_Faces_2 acts as a main surface to General_Contact_Faces_3. If any of these surfaces contain beam or truss elements interacting with other faceted surfaces in the edge-to-surface contact formulation, the decision to use these as the main surfaces will depend on the average element radius and the average smallest facet length of the faceted surfaces. By default, self-contact within each of these three surfaces is modeled with balanced main-secondary contact.

All XFEM-based crack surfaces in the general contact domain are assigned to a separate component and assigned the highest component number. Therefore, crack surfaces act as secondary surfaces s to other components by default. Contact between portions of crack surfaces are handled with balanced main-secondary contact since they all belong to a single component.

Specifying Nondefault Main-Secondary Roles

You can override the default main-secondary roles by specifying pure main-secondary roles or by specifying that balanced main-secondary contact should be used. The default main-secondary treatment works well in most cases. Keep the following points in mind when modifying the main-secondary assignments, in addition to other factors discussed in this section:

  • Do not use the internally generated component surfaces when assigning alternative main-secondary roles (instead, use surface names that you define). Named XFEM-based crack surfaces are not supported for specifying main-secondary roles.

  • The main-secondary role assignments are part of the model definition and cannot be modified from step to step.

  • The guidelines for assigning pure main-secondary roles for contact pairs discussed in Defining Contact between Two Separate Surfaces are also applicable for situations in which you reassign pure main-secondary roles for general contact.

  • The limitations of balanced (symmetric) main-secondary contact pairs discussed in Using Symmetric Main-Secondary Contact Pairs to Improve Contact Modeling are also applicable for situations in which you reassign balanced main-secondary contact for general contact. Balanced main-secondary contact can result in reduced robustness due to the increased number of constraints and the possibility of overconstraints.

Automatically Generated Contact Exclusions

Abaqus/Standard automatically generates contact exclusions for the main-secondary roles opposite to specified pure main-secondary roles; therefore, self-contact is excluded for any regions of the two surfaces that overlap. For example, specifying that the general contact interaction between surf_A and surf_B should use pure main-secondary contact with surf_A considered to be the secondary surface would result in exclusions being generated internally for main faces of surf_A contacting secondary faces of surf_B; self-contact would be excluded for the region of overlap between surf_A and surf_B. An error message is issued if the second surface name is omitted or is the same as the first surface name since this input would result in the exclusion of self-contact for the surface.

Specifying Small Sliding within General Contact

You can specify the small-sliding tracking approach for interactions involving portions of the entire general contact domain and, by extension, the entire contact domain. The small-sliding approach avoids repeated contact tracking and the need to re-establish the nodes involved in the constraint connectivity, which makes it more efficient than the default finite-sliding approach for general contact. However, you should be aware of the approximations involved in small sliding to decide whether they are appropriate (The Small-Sliding Tracking Approach). The small-sliding approach involves only the surface-to-surface contact over opposing surfaces; edges and vertices are deactivated. The surface pairs you specify to utilize small sliding are excluded automatically for the default finite-sliding approach. Subsequently, none of the types of contact based on finite-sliding assumptions (such as surface-to-surface contact or edge and vertex contact) are active over the pairings with the small-sliding specification.

You should not use the internally generated component surfaces when assigning the small-sliding approach; instead, you should define the surface names. Named XFEM-based crack surfaces are not allowed for specifying the small-sliding approach.

Smoothness of Contact Force Redistribution upon Sliding

You can control the smoothness of nodal contact force redistribution upon sliding. The default setting, which is generally appropriate, results in the smoothness of the nodal force redistribution being of the same order as the elements underlying the secondary surface; that is, linear redistribution smoothness for linear elements, and quadratic redistribution smoothness for second-order elements. Quadratic redistribution smoothness usually tends to improve convergence behavior and improve resolution of contact stresses within regions of rapidly varying contact stresses. However, quadratic redistribution smoothness tends to increase the number of nodes involved in each constraint, which can increase the computational cost of the equation solver. Linear redistribution smoothness tends to provide better resolution of contact stresses near edges of active contact regions and, therefore, occasionally results in better convergence behavior.

Additional Global Numerical Controls for General Contact

Some additional numerical contact controls can be modified globally from step-to-step for general contact; you cannot specify contact controls for individual surface pairings within the general contact domain. You can apply contact stabilization to address rigid body modes that occur prior to the establishment of contact in the model, and you can adjust the tolerances used by Abaqus/Standard to determine contact penetrations and separations; both techniques are discussed in Adjusting Contact Controls in Abaqus/Standard.