Restricting displacement to a slide surface

Element surfaces can be defined as slide surfaces in order to force the design nodes onto this contour (CHECK_SLIDE). This option offers more flexibility than restricting the direction of displacement. The limiting surfaces are formed by shell structures. Those surfaces can be generated automatically with a link shape command.

A master node group is required describing sufficiently the contour of the surface. The following command defines a surface of revolution with rotation axis Z in the global coordinate system:

      LINK_SHAPE
        ID_NAME  = surface
        MASTER = NDGR, slide_master_ndgr
        CLIENT = SURF_TURN
        CLIENT_DIR = 0, 0, 1
        TOL = 0.1, 0.1, 0.1
        CS = CS_0
      END_
    

This slide surface is connected to a node group via a DVCON_SHAPE command:

      DVCON_SHAPE
        ID_NAME  = slide_restriction
        CHECK_SLIDE = surface
        ND_GROUP = restricted_nodes
      END_
    

Alternatively the limiting surfaces are generated in the FE preprocessor and loaded via the interface (FEM_INPUT command, ADD_FILE parameter) in the optimization preprocessor. With a LINK_SHAPE command

      LINK_SHAPE
        ID_NAME  = surface
        MASTER = NDGR, slide_ndgr
        CLIENT = FREE_FORM
        TOL = 0.1, 0.1, 0.1
        CS = CS_0
      END_
    

Here, the node group slide_ndgr is used to generate the surface on which the design nodes will be restricted. The CLIENT = FREE_FORM allows for freeform surfaces.

Important: CHECK_SLIDE restrictions can be performed for surface nodes as well as for inner nodes. To limit the exterior form of a component, it only makes sense to restrict surface nodes. The restriction of inner nodes can be undertaken. However, the node displacements in the MESH_SMOOTH area should be limited. To simplify the definition of the contact check it is useful to divide the limiting surfaces by assigning various element property numbers (or materials). This greatly simplifies the selection and assembly of groups in the optimization preprocessor. If the elements being used in the check are loaded with the ADD_FILE parameter of FEM_INPUT command, pay attention that node or element IDs are not used twice since SIMULIA Tosca Structure cannot process duplicated IDs. This design variable constraint is not supported in sensitivity based optimizations.