The polysurface is composed of ruled surfaces so they should all unroll as one new flat polysurface. All of the surface normals are pointing in the same direction. But for some reason, the revolve surfaces sometimes flip direction and produce the folded mess on the left. I usually fix this by just unrolling using the explode option and then manually flipping the incorrectly oriented revolve surface upright and stitching the polysurface back together.
To be clear, this doesn’t happen to all revolve surfaces. Notice that in this polysurface, there are two revolve surfaces, and one unrolls correctly while the other does not. It’s the inconsistency that is curious.
Since this has happened over a few years, I figured I’d post to see if there was something I’m missing. This is part of a larger tutorial for students on how to unroll surfaces, so I actually like when problems come up like this as I can teach them why the problem might be happening and how to solve it. So feel free to suggest alternatives or explanations so I can share the community’s knowledge with the students.
Not all ruled surfaces will unroll without distortion. A surface which will unroll without distortion is a developable surface. All developable surfaces are ruled surfaces BUT not all ruled surfaces are developable surfaces.
The geometry is undistorted, just flipped. The specific problem that I’m asking about is a surface made from the revolution command. The surfaces were made revolving a rectangular profile along an axis parallel with one of the rectangle’s edges. This should result in two types of surfaces:
For the sides of the rectangle that are parallel to the axis of revolution, it should result in portions of a cylindrical surface. As such, it should be both ruled and developable.
For the sides of the rectangle that are perpendicular to the axes of revolution, it should result in flat surfaces that are both ruled and developable.
So, my question is why does the unrollsrf command fail by flipping revolution surfaces opposite their surface normals, producing a folded/overlapping result? And why does it do this inconsistently?
The other surfaces were made using the Blend command, which can result in non-developable surfaces in some instances. However, they seem to be unrolling either correctly or with distortions that are within the material’s tolerance. The problem seems to be related only to the surfaces of revolution.
What is interesting is if you replace that one surface (the larger rev surface) with an EdgeSrf made from its edges and join it all back up, it unrolls perfectly.
These are old bugs which have plagued unroll surface for 10 years at least. I run into this all the time with revolve surfaces, sometimes it even happens on other objects not just revolved.
One would think that over the years with the myriad of bug posts unroll would get a little more love perhaps this is still on the bug list and will be fixed for V9?
RM
