Mirrored simple surface shows naked edge on one only

I am trying to get this simple shape to even behave itself, avoid bugs bad geometry etc.

I delete all surfaces except the side of a wing, I mirror it, I sweep it, I join it and one edge is naked and the other is not yet they are mirrored so how can that be ?

What do I have to do to get one wing complete and ready for mirroring to make second wing.

when I then join it to the cone it says bad geometry, but I want to at least get each stage of this simple thing healthy.

I start over with one wing side, mirror, sweep, mirror to other half, add planar top, then bottom and get failure to join bottom, I cap planar hole and success, but then its naked edged.

can anyone manage to take this simple wing side, and make a wingnut from this ?
Naked edge yet A and B are mirrored.3dm (403.9 KB)


dear steve

some issues already discussed here:

the “_what” command on the initial wing shows an edge tolerance up to 0.002
your document tolerance is 0.001
→ _trim again
→ _join again
depends a bit on what you re after - but please notice:
if you use a real cone or an revolve the Edge (A) might not be a line.
if you do a sweep2 the surface will not be 100% rotational./ or not a perfect cone.

it s realy hard to guess where the inprecision comes from - but above workflow will fix your naked edges:
(i used your _sweep2 approach)

Naked edge yet A and B are mirrored_02.3dm (3.9 MB)

Naked edge yet A and B are mirrored_02_rh5.3dm (375.7 KB)

kind regards -tom

Hi Tom_P
I am unable to open the file, says a newer version of windows was used.
I am V5



update my message above with a rh5 version.

Hi Tom_P
no naked edges :slight_smile:

Now then I need to obviously manage this myself.
Before opening the Rh5 from you, I went back to basics.

  1. delete top and bottom circles for cone, redraw them 0.55 and 0.7inch dia and position them as before at where the wing becomes horiz at the upper and base if wing at bottom.
  2. draw line at quads, check dir for circles and sweep2rail.
  3. use surface planar curve command and create ends of cone, join and check for edges, all good.
  4. delete rhs wing, and the swept faces of the wing.
  5. Redo the sweep2rail of all but the end areas, checking all dir or ok.
  6. add an extra straight profile ‘A’ at the top rather than use the blue curve ‘B’ to sweep to the straight line, sweep2rail (F G rails A E as profiles) and get a better surface, then use clone of sides of wing and trimmed to get sides C D as profiles for sweep2rail the last little bit with the new straight line A and curve B as rails.
  7. at base use sweep2rails , rails J K and the blue segment H sweep to the straight line I.
  8. Join all the wing surfaces, run show edges and it shows 7 though I expected 4, and cannot see the other three.
  9. Mirror the wing and run Intersect, get two ‘rectangles’ (red layer) on my cone.
  10. Trim cone surface with these.
  11. Select the two wings and cone and join, and get an error message.

wingnut v2 gives error.3dm (341.4 KB)

Why ?

I have been so systematic and careful. Its so simple and defeating me.

Your model perfick !


sorry if i did not followed all your steps …

Did you consider, that cutting a cone with a plane will not result in a straight line (only in raw / spacial cases)

the red curve in this example is not a (straight) line:

I bet a lot of folk didnt know that. not if cone is straight sided as opposed to an aircraft spinner on a spitfire etc.
However I simply ran intersect of wings to cone, and used result to trim cone to accept the wings in join command.
Should that not work ?
I cant in my mind see why it fails, and how one would establish the cut lines of one object with another for a join.