General Mechanical

General Mechanical

Norms for FEM

    • momidor
      Subscriber

      Dears, 


      In the past I heard that there is some norms which may help in process of interpretations of the results. My main concern is following doubts:


      1. If we hava a maximum stress in the certain point ( see snapshot ) do we have to take it as it is or, should we rather omit it and take an average stress from the surroudings ?


      2. If from surroudings, how to designate the radius ?  


      Thanks in advanced 

    • peteroznewman
      Subscriber

      Dear momidor,


      If a mesh refinement study has been done that shows convergence to the exact value, then you should take that value as is. A mesh refinement study example is shown in this discussion.


      When a singularity exists in the geometry, such as a sharp 90 degree interior corner, and if the recommended corrective action of adding a blend radius to that sharp corner is not added to the model, then a rough estimate of the maximum stress is to use the value of stress one element away from the stress singularity.


      Regards,


      Peter

    • momidor
      Subscriber

      Dear Peter,


      You are reliable as always... but the key word is " then a rough estimate of the maximum stress is to use the value of stress one element away from the stress singularity."


      I'm sure that someone said that there is kind of guidelines which says that where is this "away from" area.


      Regards  

    • peteroznewman
      Subscriber

      Dear momidor,


      The snapshot you attached is not a singularity because the mesh refinement study converged, so you take it as is.


      An example of a stress singularity is an E-ring groove in a shaft that has a lateral load. The geometry is modeled as a sharp interior corner, so a mesh refinement study would not show convergence. This is the case when the one element guide may be used. The direction is clearly understood as along the axis away from the corner where the maximum stress is.



      Regards

    • peteroznewman
      Subscriber

      @vineesh karthikeyan, I moved your question to a new discussion.

Viewing 4 reply threads
  • You must be logged in to reply to this topic.