General Mechanical

General Mechanical

Verification of repeated eigenfrequencies

    • julius.langenhorst
      Subscriber

      Hello,

      I have a complex truss structure that has the following natural frequencies. Some of them are multiples and some are very close to the multiples. How can I verify mathematically that I really have repeated eigenfrequencies and not some frequencies that are just close to each other?

      13252.43697076
      13253.72670625
      13253.72670625
      13319.90610080
      13319.90610080
      13319.90610080
      15125.04970060
      15125.04970060
      15125.05037702
      17822.02177338
      17822.65139503
      17822.65139503

      Is it possible to recognize repeated eigenfrequencies from the values alone? If yes, at which decimal place do numerical errors start to falsify the results?

      Thank you for your help!

    • peteroznewman
      Subscriber

      If the structure has two planes of symmetry (both geometric and boundary conditions), then you expect repeated eigenfrequencies because the structure will move in the XY plane for one and move in the YZ plane for the repeated frequency due to those being the planes of symmetry. If the structure, including the mesh, is not perfectly symmetric, then a small difference can be introduced between the two planes, but I would still call that a repeated eigenfrequency.

      It’s possible to have two very close frequencies that have totally different mode shapes. You can see that when you animate the two modes. You can’t tell from the values alone, you need to look at the mode shape.

    • julius.langenhorst
      Subscriber

      Thank you for your answer!

      I'll have a look at animating the modes. But it's save to say that frequencies which are equal up to the 6th decimal place or so are repeated eigenfrequencies?

    • peteroznewman
      Subscriber

      Yes, that is a safe assumption.

    • julius.langenhorst
      Subscriber

      Thank you!

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