I have a complex part that contains features of the form shown in the figure below.

hollow shell

Because of the cost of 3D finite element simulation of the part, I want to try an analysis with 2D thick shells.

However, I am running into a problem when trying to extract the mid-surface of the part. Automated tools provide me with a mid-surface that is shown by the red dashed line in the figure. That leads to interpenetration of the top and bottom surfaces in bending.

I would like a mid-surface that is shown by the green dashed line.

Can anyone point me to automated mid-surface extraction algorithms that can do what I seek?

  • $\begingroup$ Did you try computing the alpha shape of the section? $\endgroup$
    – nicoguaro
    Aug 13, 2021 at 22:45
  • $\begingroup$ No, I've only use commercial tools so far. But I'll look up what alpha shape means. Also, I'm not dealing with something that can be termed a section. $\endgroup$ Aug 14, 2021 at 2:07
  • $\begingroup$ I don't think anyone can help you if you don't tell us in what format you have the geometry. $\endgroup$ Aug 14, 2021 at 3:25
  • $\begingroup$ I guess, think that my suggestion might not be useful. For some reason, I mixed alpha shapes with polygon skeletonization. $\endgroup$
    – nicoguaro
    Aug 14, 2021 at 15:59
  • 1
    $\begingroup$ @WolfgangBangerth: That's intriguing. Could you explain why? I would have though that the format doesn't matter. I can store the geometry either as topological entities (points, curve, curve loops, surfaces, surface loops, volumes) or as triangulated surfaces. And save the geometry is a number of file formats, STEP (topological) is the most convenient. $\endgroup$ Aug 14, 2021 at 20:19


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.