LS Dyna

LS Dyna

Topics relate to LS-DYNA, Autodyn, Explicit STR and more

Nodes translation restriction (only moving forward)

    • Yuqian TU
      Subscriber

      Hello! 

      Is there a keyword or a way to constrain nodes (either in a shell/solid element or rigid node set) at X and Y directions that nodes are only allowed to move forward. When the nodes want to move backward, the displacement is set to 0.

      I was thinking about local coordinate system with rigidwall, but didn’t seem to work.

      Thanks a lot!  

    • Jim Day
      Ansys Employee

      A couple of possibilities:

      1. Attach a stiff *ELEMENT_SEATBELT and an accompanying *ELEMENT_SEATBELT_RETRACTOR such that the seatbelt resists backsliding and is reeled in as the part moves forward.
      2. To resist backsliding, a feedback control system of sorts could be employed to apply a resisting force.

      For example, here’s an example in which a force is applied to part #1 in accordance with the control law shown below so that it translates to a specified target point.

      Force = -gain1 * (displacement – target) – gain2 * velocity

      The feedback control system is governed by the following keywords which are used to:

      1. define the gains and other pertinent parameters
      2. define the feedback control law
      3. apply the load

       

      *PARAMETER
      R gain1 0.060000
      R gain2 0.010000
      R target 2.000000
      $
      *DEFINE_CURVE_FUNCTION
      $# lcid sidr sfa sfo offa offo dattyp
      100 0 0.000 0.000 0.000 0.000 0
      $# function
      -gain1*(dx(28)-target) – gain2*vx(28)
      $
      *LOAD_RIGID_BODY
      $# pid dof lcid sf cid m1 m2 m3
      2 1 100 0.000 0 0 0 0

      In *DEFINE_CURVE_FUNCTION dx(28) is used to monitor the x-displacement of node #28 while vx(28) returns the node’s velocity. These responses are updated continuously with the current values as the simulation progresses. The function itself could be a control law that involves any type of model response, such as, nodal forces, accelerations, stresses, etc.

Viewing 1 reply thread
  • You must be logged in to reply to this topic.