General Mechanical

General Mechanical

Element shape checking with Pre-integrated shell section

    • isoo


      Recently i have been experiencing a strange problem with Mechanical APDL, regarding to radius/thickness ratio with pre-integrated sections;

      • I provide several pre-integrated user defined shell section via SSPA,SSPB,SSPD,SSPM commands and assign those to individual surfaces using selection,
      • Yet, solver does a thickness/radius check before solution and raises errors and aborts solution
      • Element Shape checking was turned off via SHPP,OFF command, yet solver keeps checking element shape and raising errors then aborting solution
      • As the last option, i have used the exact section properties which MAPDL gives as output of the SLIST,FULL command (for a section which solution processes and produces results without any shape errors). Solution is aborted with GENS, while it process with casual section definition command which sets a thickness value.

      A few questions;

      How does Mechanical APDL conduct element shape check routine (thickness/radius) in case of user-defined shell sections which an actual thickness value is absent (all thichkness value seems as "1" , which is in meters, in error lines) ?

      My shell section integration is consistent and fits to the theory and gives accurate results via external solvers but why Ansys Mechanical APDL does not accept it to be valid?

      Why Ansys does not pass Element shape checking (except of nodal and property consistency and availability), despite of its being turned off by user?

      Further attempts to diagnose, I have also used ESCHK command to view problematic elements yet those were not in abrupt shape, passes casual shape check (jacobian, aspect ratio etc) but fails curvature radius/thickness criteria (eventhough shape checking is turned off) and solution aborts.

      Best Regards.

    • isoo
      I have further analyzed the process and i figured out the solver uses "unit" length in place of pseudo-thickness to check element shape validity,if user-generated section is used. So if user is working in milimeter, the assumed thickness is 1 milimeter for a any shell element, while it is 1 meter if user is working on meter scale.
      And in my opinion this is a possible bug.
    • 1shan
      Ansys Employee
      Could you try issuing the command - /NERR,,99999999 and see if this solves the issue?
      Regards Ishan.
    • isoo
      I did also try that command and it does not resolve the issue.
Viewing 3 reply threads
  • You must be logged in to reply to this topic.