What causes a warning about unprocessed component while post-processing a cyclic symmetry solution in Mechanical APDL?

    • FAQFAQ

      After /CYCEXPAND,,OFF (or if no /CYCEXPAND has been issued), the following warning message is issued (may be PRNSOL, PRESOL or other):
      *** WARNING ***
      PRNSOL is listing an unprocessed component of a cyclic symmetry
      solution. It can combine all appropriate components to produce the
      actual nodal solution for sectors 1 through 6 if cyclic expansion
      (/CYCEXPAND) and PowerGraphics (/GRAPH) are both active.

      Various commands (*GET, *VGET, …) gives you the unprocessed results but not as real and imaginary part but as real values for original and duplicate sector, so if you have n nodes in the sector, you need to get results for both nodes ii and ii+n

      NODE UZ
      24 83.275
      120 190.35

      To compute expanded results, User need eq.3-2 (Mechanical APDL > Cyclic Symmetry Analysis Guide >Cyclic Symmetry Overview > Postprocessing a Cyclic Symmetry > Using the /CYCEXPAND Command)

      The amplitude can be retrieve from the unprocessed results AMP = SQRT(UA^2+UB^2). A phase sweep can also be performed (as with CYCPHASE command) by adding a Cyclic Phase angle to both cos and sin terms in eq.3-2.
      Extra precaution has to be observed when dealing with non-component results (USUM, SEQV, SINT) or linearized stress, component results need to be obtained with eq.3-2 before computing non-component results: combining non-component results with eq.3-2 for basic and duplicated sector is not correct.

      In addition, and for modal results only, expanded results are normalized with respect to the full mass matrix (N * [M]), while unprocessed results are normalized to the cyclic mass matrix (2 * [M], except if k = 0 or N/2, where 1 * [M]) (N = number of sectors, [M] = mass matrix of the original sector), so that we have to multiply the above results by SQRT(2/N) (or SQRT(1/N) if k = 0 or N/2).

      With /CYCEPAND,,ON or within Workbench Mechanical, expanded results can be plotted (Maximum Over Cyclic Phase, Cyclic Phase of Maximum, Cyclic Phase)