Contact Interface Parameters (Contact Control)

Orientation of Contact Pushout Force (MORIENT)

The orientation of the contact push out force from the main surface can be defined on the MORIENT field on the CONTACT entry. It applies only to main that consist of shell elements or patches of grids. Main defined on solid elements always push outwards irrespective of this flag. MORIENT is the direction of contact force that main surface exerts on secondary nodes.

It is important to note that, in most practical applications, leaving this field blank will provide correct resolution of contact, irrespective of the orientation of surface normals. Only in cases of main surfaces defined as shells or patches of grids, and combined with initial pre-penetration, is MORIENT needed.

By default, MORIENT is ignored for solid elements - it applies only to main surfaces that consist of shell elements or patches of grids. Main surfaces defined as faces of solid elements always push outwards, irrespective of the surface normals, or whether the contact gap is initially open or closed.

In OPENGAP (default when TRACK is not set to FINITE/CONSLI), the push out direction is defined using the assumption that the gap between secondary and main is initially open, and the contact condition should prevent their contact (gap padding GPAD from the PCONT card is ignored in defining the push out direction - this direction is based strictly on the positions of main and secondary nodes).

Figure 1. OPENGAP Example
OVERLAP assumes the reverse, namely that the secondary and main bodies are already overlapping and the contact condition should push them apart (this is useful in pre-penetrating models when the entire secondary set is pre-penetrating into the main object).

Figure 2. OVERLAP Example
With the NORM option (default for TRACK=FINITE/CONSLI), the push out force is oriented along the normal vector to the main surface. (The surface normal may be reversed relative to the default normal to a shell element, if a FLIP flag is present on the main SURF definition. This behavior corresponds to that of the reverse normals check box on the contactsurfs panel in HyperMesh). In cases when the secondary node does not have a direct normal projection onto the main surface, and the "shortest distance" projection is used (GAPGPRJ set to SHORT on the GAPPRM card), the push out force is oriented along the shortest distance line, yet with the orientation aligned with the normal vector.

Figure 3. NORM Example
REVNORM creates push out force reversed relative to the NORM option.

By default, MORIENT does not apply to mains that are defined on solid elements - such mains always push outwards. This can be changed by choosing CONTPRM,CORIENT,ONALL which extends the meaning of MORIENT to all contact surfaces. In which case, it should be noted that the default normal is pointing inwards unless a FLIP flag appears on the main SURF definition for surfaces on solid elements, making the surface normal point outwards. (When creating contact surfaces in HyperMesh, this behavior corresponds to that of the reverse normals check box on the contactsurfs panel).

Search Distance (SRCHDIS)

The SRCHDIS field on the CONTACT and TIE entries is the search distance criterion for creating the contact interface. When specified, only secondary nodes that are within the SRCHDIS distance from the main surface will have the contact condition checked.

The default value is equal to twice the average edge length of the main surface. For FREEZE contact, it is equal to half the average edge length.

For shell elements, the contact and tie search considers the shell thicknesses. Which implies that the defined search distance is expected to be the true distance between the shell surfaces facing each other. For example, in the case of shells without offset, if the geometric distance between the two shell surfaces facing each other is 5.0, and the shell thicknesses are 2.0 each, then the actual distance between the shell surfaces facing each other is 3.0. If the Search distance field is now set to 3.0, then the contact is generated as expected (in this scenario, if SRCHDIS is set lower than 3.0, then there is no contact generated).

Contact Adjustment (ADJUST)

Contact Adjustment can be used to adjust the secondary nodes onto the main surface at the beginning of the run. The ADJUST field on the CONTACT and TIE entries can be used, with the available options:

NO: No adjustment (default).
AUTO: A real value equal to 5% of the average edge length on the main surface is internally assigned as the depth criterion.
Real > 0.0: Value of the depth criterion which defines the zone in which a search is conducted for secondary nodes (for which contact elements have been created). These secondary nodes (with created contact elements) are then adjusted onto the main surface. The assigned depth criterion is used to define the searching zone in the push out direction.
Integer > 0: Identification number of a SET entry with TYPE=GRID. Only the nodes on the secondary entity which also belong to this SET will be selected for adjustment.

The adjustment of secondary nodes does not create any strain in the model. If DISCRET=N2S is selected, it is treated as a change in the initial model geometry. If DISCRET=S2S is selected, it is treated as a change in the initial contact opening/penetration.

If a node on the secondary entity lies outside the projection zone of the main surface, it will always be skipped during adjustment since no contact element has been constructed for it.

Contact interface padding, GPAD (on the PCONT entry) is used to determine the direction and distance of secondary node adjustment. GPAD is also used to augment the searching zone for adjustment. If the MORIENT field is OPENGAP or OVERLAP while the GPAD field in the referred PCONT entry is NONE or zero, the nodal adjustment will be skipped, since for OPENGAP or OVERLAP there is no way to decide the main push out direction if secondary nodes are adjusted to be exactly on the main face.

If different contact interfaces involve the same nodes, nodal adjustment definitions are processed sequentially in the order of identification numbers of the contact interfaces. Care must be taken to avoid conflicts between the nodal adjustments; otherwise, contact element errors or lack of compliance may occur.

The ADJUST field must be set to NO for self-contact.
If a real value (the searching depth criterion for adjustment) is input for the ADJUST field, a searching zone for adjustment is defined. If GPAD is also defined, then the value of GPAD augments the searching zone for ADJUST (see Figure with GPAD below). The secondary nodes in this searching zone, for which contact elements have been created, will be adjusted. If ADJUST is larger than or equal to SRCHDIS, all the secondary nodes for which contact elements have been created, will be adjusted.

Figure 4. No Contact Interface Padding (GPAD)

Figure 5. With Contact Interface Padding (GPAD)
Note: The depth criterion (A non-negative real value for ADJUST) is used to define the searching zone for adjustment, as shown above. If GPAD is defined, then the value of GPAD is also used to augment the searching zone for ADJUST, as shown in the image above. This searching zone is created in the push out direction up to a distance equal to the value of the ADJUST field. The secondary nodes within the searching zone (with defined contact elements) are then considered for adjustment based on the rules specified within this section.
If the ADJUST field is set to an integer value (the identification number of a grid SET entry), the nodes shared by the secondary entity and the grid SET will be checked for contact creation, that is, SRCHDIS will be ignored for these nodes, and then adjusted if a projection is found. The nodes belonging to the grid SET but not to the secondary entity will be simply ignored.

Contact Clearance (`CLEARANCE`)

Clearance can be defined on the CONTACT or PCONT entries on the CLEARANCE field. It can also be defined via U0 on the PGAP entry. Clearance does not physically move the nodes. The contact is considered to be closed, if the clearance between the two surfaces is equal to (or less than) the specified clearance, regardless of the physical location of the nodes. Make sure that the surface nodes are not highly irregular or the actual physical distance between some (or all) parts of the surfaces is much greater than the specified clearance. The resulting contact parameters applied to the actual physical nodes after each iteration or over the entire solution may be inaccurate in such cases. A blank clearance field is the same as U0=AUTO.

Using CLEARANCE overrides the default contact behavior of calculating initial gap opening from the actual distance between Secondary and Main. CLEARANCE is now equal to the distance that Secondary and Main have to move towards each other in order to close the contact. Negative value of CLEARANCE indicates that the bodies have initial pre-penetration.

CLEARANCE cannot be set in conjunction with (non-zero) Contact Interface Padding (GPAD). Blank GPAD field in presence of CLEARANCE is interpreted as NONE.

Warning: If CLEARANCE is used, it is important to correctly restrict the contact zones and pick search distance SRCHDIS, so that only desired Secondary-Main pairs are involved. Using CLEARANCE, all contact elements created on a given interface, even those where Secondaries are geometrically distant from the respective Main surface, will be considered to be at given initial gap and participate in resolving the contact condition.

Note:

CLEARANCE on the CONTACT entry cannot be used in conjunction with PID of the PCONT entry. In this case, clearance must be specified on the PCONT entry.
The consideration of SRCHDIS in presence of CLEARANCE can be controlled using the SRCHDCLR entry in system settings (SYSSETTING) or in the Solver Configuration File (.cfg) file.

Contact Smoothing (`SMOOTH`)

Optional SMOOTH continuation line(s) is used to define contact smoothing for region(s) of main and/or secondary surfaces.

If SMSIDE is MAIN and SMREG is ALL (or same as MSID), there must be only one SMOOTH continuation line to define smoothing of the Main side.
If SMSIDE is SECOND and SMREG is ALL (or same as SSID), there must be only one SMOOTH continuation line to define smoothing of the Secondary side.
If SMSIDE is BOTH, SMREG must be ALL and there must be only one SMOOTH continuation line.
Smoothing is now supported for both N2S and S2S contact.

Finite Sliding (`TRACK`)

Finite Sliding (TRACK=FINITE or CONSLI) option is currently supported only if TYPE=SLIDE or if friction (via MU1/CONTPRM/PCONT) is defined.

For TRACK=FINITE, the contact search is conducted for every load increment, while with TRACK=CONSLI, termed continuous sliding, the search is done for every iteration. The CONSLI option is expected to produce more accurate results and, in some cases, better convergence robustness, especially when very large sliding and/or distortion is present.

For non-solid elements, the MORIENT field should not be set to OPENGAP or OVERLAP for Finite Sliding (FINITE/CONSLI). If CONTPRM,CORIENT,ONALL is active, MORIENT is applied to solid elements. In such cases, MORIENT should not be set to OPENGAP or OVERLAP for solid elements also.

To activate Finite Sliding, it is important to review the following:

Finite Sliding (FINITE/CONSLI) is effective only for Large Displacement Nonlinear Analysis.
Finite Sliding (FINITE/CONSLI) works only with Hash Assembly and the MUMPS solver. Hash Assembly (PARAM,HASHASSM,YES) is automatically turned on for Large Displacement Nonlinear Analysis.
Finite Sliding (only CONSLI) is supported for Self-Contact.

Contact Interface Padding (GPAD)

"Padding" of the interface to account for additional layers, such as shell thickness. This value is subtracted from the contact gap opening as calculated from the location of nodes. It is set on the GPAD field of the CONTACT and PCONT Bulk Data Entries. CLEARANCE cannot be set in conjunction with (non-zero) GPAD. Blank GPAD field in presence of CLEARANCE is interpreted as NONE.

The initial contact gap opening is calculated automatically based on the relative location of secondary and main nodes (in the original, undeformed mesh). To account for additional material layers covering main and secondary objects, the GPAD entry can be used. GPAD option THICK automatically accounts for shell thickness on both sides of the contact interface (this also includes the effects of shell element offset ZOFFS or composite offset Z0), when the main and/or secondary are shell surfaces (SET of shell element types or SURF of shell element faces). The THICK option will only apply automatic padding if shells are selected for main/secondary in the contact interface (for example, padding is not applied for "skinned" solid elements that are selected as a main/secondary in the contact interface).

Contact Stabilization (CNTSTB entry and PARAM,`EXPERTNL`,CNTSTB)

Defines parameters for stabilization control of surface-to-surface contact and large displacement node-to-surface contact. A CNTSTB Bulk Data Entry should be referenced by a CNTSTB Subcase Information Entry to be applied in a particular subcase. Also, PARAM,EXPERTNL,CNTSTB can be used to apply contact stabilization.

Resolution of Pre-penetration (CONTPRM,`SFPRPEN`)

The CONTACT capability in NLSTAT solution is designed to correctly resolve initial pre-penetration, such as happens in press fit, and so on. This usually works reliably with correct identification of main and secondary surfaces. However, in some cases contact surfaces by property are created for convenience, which results in contact surfaces enveloping the entire solid bodies. Also, the secondary and main could receive the same ID, which is known as self-contact (not recommended, in spite of the convenience factor). In such cases, it is possible to encounter false self-penetrations (Figure 6).

In the case above, the secondary node will be identified as if pre-penetrating the main face, while in reality it is on the other side of the same solid body. The result from nonlinear CONTACT solution will be such that this portion of the body will be "squeezed" to have practically zero thickness, with very high stresses obviously resulting.

Apart from correctly identifying potential secondary and main sets, a possible remedy to avoid such situations is to make sure that SRCHDIS is smaller than minimum thickness of the solid bodies which are enveloped by self-contacting surfaces.

An alternative, viable when there are no actual pre-penetrations in the problem, is to choose SFPRPEN=NO, which will ignore initial pre-penetrations on self-contacting surfaces (some minor pre-penetrations, due to variations of nodal positions will still be correctly resolved - up to the minimum element size on the respective contact surfaces).

Note: SFPRPEN affects only surfaces that actually have self-penetration, as in a case where the secondary node and main face belong to the same contact set or surface. On properly defined, disjoint Secondary and Main surfaces, the initial pre-penetrations will be resolved irrespective of this parameter.

Contact Friendly Elements (CONTFEL)

Contact-friendly elements can be activated using PARAM,CONTFEL,YES. Contact-friendly TETRA10, HEXA20, and PENTA15 elements are available and can be combined in a single model. Modified shape functions are used for possible improved runs with contact friendly elements.

No Separation Contact (SEPARATION)

Flag indicating whether the main and the secondary can separate once the contact has been closed. This is set on the PCONT Bulk Data Entry, via the SEPARATION field. If SEPARATION is set to NO, then the main and secondary do not separate after contact is closed. Applied only to frictional, SLIDE and STICK contacts with S2S or large-displacement N2S.

Self-Contact

There are two ways to activate self-contact:

Main and secondary set identification numbers can point to the same set.
Main set identification number can be blank.

Self-contact is available as an option to define contact in the following situations:

Surfaces which after large deformation can come into contact with themselves (for example, large buckling deformations). In such situations, some material can collapse onto itself, and it is usually difficult to decide which part will be in contact.
When a lot of potential contact pairs (main-secondary) can exist and it is not feasible to define all of them manually.
Self-contact can be tried, if manual contact setup fails for some reason.