Bug Report? Twisted bodies not straigft

[RobbieKnobbie]

Playing with twisted extrusions for fidget twisters. First core was made of two SKETCH ellipses crossed at 90 degrees. Worked out fine (orange)

Then I tried to extrude and twist a STAR shape. 120mm high and 360 degrees of twist. As you can see in the attached picture. (light blue) a star extrusion from a SKETCH did the same thing.

This compared to an extruded straight square (red)

I tried to attribute this to the ellipses being 2 axis symmetrical, while the star shape is only symmetric on a single axis, so I tried a six pointed star...

Look into the X axis, the twist is bound by the red and green blocks... you can see that the lobes 2,3,5, and 6 nearly touch the blocks, while lobes 1,4, and 7 fall short.

Look into the Y axis, the twist is bound by the blue and grey blocks. Now lobes 1,3,4, and 6 are proud while lobes 2 and 5 now appear undersized...

Anybody have any ideas what's going on here? Is layerling's kernal just not that mathematically precise (being more graphically oriented) so the axis of rotation is allowed to gyrate? I'm not a programmer, obviously.

[henmedia]

@RobbieKnobbie, thank you for the pictures and for measuring so carefully, you found a real fault, and you are not doing anything wrong. When a body is twisted, layerling turns it around the middle of its bounding box, not around the middle of the star. A star is not centred in its box (a five-pointed one is about 2 mm off), so the middle of the star circles round the axis instead of staying on it. I measured it here: with a twist of 360 degrees the star reaches 30 mm from its middle at the bottom and 35 mm in the middle of the height, which is the lobes you saw falling short. I will change it so that a star turns around its own middle.

[henmedia]

@RobbieKnobbie, the twisted star is fixed in 1.86.0: a star now turns around its own middle, in the editor, in the exact body and in the exports, so the tips keep the same distance at every height. Taper works about the same point. Other shapes still turn around the middle of their box. Please try your 120 mm star with 360 degrees and tell me if anything is still off.