Linear Imperfect Interface

From OSUPDOCS
Jump to navigation Jump to search

Description

This imperfect interface contact law assumes the normal and tangential tractions are linear and depend only on the normal and tangentatial displacement discontinuities, respectively:

      [math]\displaystyle{ T_n = D_n[u_n] }[/math]

      [math]\displaystyle{ T_t = D_t[u_t] }[/math]

where Dn and Dt are two interface parameters, which are infinite for a perfect interface (zero displacement discontinuity) and 0 for a debonded interface (zero traction). But this linear law would allow the two materials to interpenetrate, especially if Dn was low. To correct this issue, the traction law in the normal direction is allowed to be bilinear:

      [math]\displaystyle{ T_n = \left\{ \begin{array}{ll} D_{nt}[u_n] & {\rm \ if\ }[u_n]\gt 0 \\ D_{nc}[u_n] & {\rm \ otherwise} \end{array} \right. }[/math]

where Dnt and Dnc are separate interface parameters for the interface being in tension or compression.

Properties

The properties for this law are:

Property Description Units Default
Dnt (or Dn) Imperfect interface parameter for normal direction when in tension pressure/length units -1
Dnc Imperfect interface parameter for normal direction when in compression. If Dnc is not specified, it set equal to Dnt. pressure/length units none
Dt Imperfect interface parameter for tangential direction pressure/length units -1

In theory, very large interface parameters will result in a perfect interface, but they also make the numerical equations very "stiff" and therefore potentially unstable. To solve this issue, you can set any interface parameter to -1 to indicate an infinite interface parameter or a perfect interface. When a parameter is -1, the equations are handled differently to avoid stability issues.

Examples

 Material "interfaceID","My Imperfect Interface","LinearInterface"
   Dn 500
   Dt 0
 Done