-
-
Notifications
You must be signed in to change notification settings - Fork 34
/
CONTACT_PAIR.html
6 lines (5 loc) · 5.24 KB
/
CONTACT_PAIR.html
1
2
3
4
5
6
<html><head><link rel="stylesheet" type="text/css" href="style.css"/></head><body> <H2> <BR> *CONTACT PAIR </H2> <P> Keyword type: model definition <P> This option is used to express that two surfaces can make contact. There are two required parameters: INTERACTION and TYPE, and two optional parameters: SMALL SLIDING and ADJUST. The dependent surface is called the slave surface, the independent surface is the master surface. Surfaces are defined using the *SURFACE keyword card. The dependent surface can be defined as a nodal surface (option TYPE=NODE on the *SURFACE keyword) or as an element face surface (default for the *SURFACE card), whereas the independent surface has to be defined as an element face surface. If you are using quadratic elements, or if you select face-to-face contact, however, the slave (= dependent) surface has to be defined based on element faces too and not on nodes. <P> If the master surface is made up of edges of axisymmetric elements make sure that none of the edges contains nodes on the axis of symmetry. Indeed, such edges are expanded into collapsed quadrilaterals the normals on which cannot be determined in the usual way. <P> The INTERACTION parameter takes the name of the surface interaction (keyword *SURFACE INTERACTION) which applies to the contact pair. The surface interaction defines the nature of the contact (hard versus soft contact..) <P> The TYPE parameter can take the value NODE TO SURFACE, SURFACE TO SURFACE, MORTAR, LINMORTAR, PGLINMORTAR or MASSLESS. NODE TO SURFACE triggers node-to-face penalty contact, SURFACE TO SURFACE face-to-face penalty contact. MORTAR triggers the mortar method with standard dual shape functions for the Lagrange multipliers, LINMORTAR the mortar method with linear dual shape functions and PGLINMORTAR the Petrov-Galerkin method in which the usual shape functions are used to describe the variation of the Lagrange multiplier. For details the reader is referred to Section 6.7.7 and [86]-[89]. If the reader wants to apply mortar contact, it is suggested to start with MORTAR contact and to use LINMORTAR or PGLINMORTAR only if MORTAR fails. Finally, MASSLESS triggers the massless contact explicit dynamics procedure. Notice that although several *CONTACT PAIR cards can be used within one and the same input deck, all must be of the same type. It is not allowed to mix NODE TO SURFACE, SURFACE TO SURFACE MORTAR, LINMORTAR, PGLINMORTAR and MASSLESS contact within one and the same input deck. <P> The SMALL SLIDING parameter only applies to node-to-face penalty contact. If it is not active, the contact is large sliding. This means that the pairing between the nodes belonging to the dependent surface and faces of the independent surface is performed anew in every iteration. If the SMALL SLIDING parameter is active, the pairing is done once at the start of every increment and kept during the complete increment. SMALL SLIDING usually converges better than LARGE SLIDING, since changes in the pairing can deteriorate the convergence rate. For face-to-face contact (SURFACE TO SURFACE, MORTAR, LINMORTAR or PGLINMORTAR) small sliding is active by default. <P> The ADJUST parameter allows the user to move selected slave nodes at the start of the calculation (i.e. at the start of the first step) such that they make contact with the master surface. This is a change of coordinates, i.e. the geometry of the structure at the start of the calculation is changed. This can be helpful if due to inaccuracies in the modeling a slave node which should lie on the master surface at the start of the calculation actually does not. Especially in static calculations this can lead to a failure to detect contact in the first increment and large displacements (i.e. acceleration due to a failure to establish equilibrium). These large displacements may jeopardize convergence in any subsequent iteration. The ADJUST parameter can be used with a node set as argument or with a nonnegative real number. If a node set is selected, all nodes in the set are adjusted at the start of the calculation. If a real number is specified, all nodes for which the clearance is smaller or equal to this number are adjusted. Penetration is interpreted as a negative clearance and consequently all penetrating nodes are always adjusted, no matter how small the adjustment size (which must be nonnegative). Notice that large adjustments can lead to deteriorated element quality. The adjustments are done along a vector through the slave node and locally orthogonal to the master surface. <P><P> <BR> <P> First line: <UL> <LI>*CONTACT PAIR </LI> <LI>enter the required parameter INTERACTION and any optional parameters. </LI> </UL> <P> Following line: <UL> <LI>Name of the slave surface (can be nodal or element face based). </LI> <LI>Name of the master surface (must be based on element faces). </LI> </UL> Repeat this line if needed. <P> <PRE>
Example:
*CONTACT PAIR,INTERACTION=IN1,ADJUST=0.01
dep,ind
</PRE> <P> defines a contact pair consisting of the surface dep as dependent surface and the element face surface ind as independent surface. The name of the surface interaction is IN1. All slave nodes for which the clearance is smaller than or equal to 0.01 will be moved onto the master surface. <P> <P><P> <BR> Example files: contact1, contact2. <P> </body></html>