I’m trying to layout four cross section to four rails
so that all four cross section curves are tangent
to each other and on the same plane.
I was using orient, but I need to adjust points later and with need to make many copy of these I’m trying to find more efficient way to do it.
Main purpose is making a sweep like surface with cross sections for better control. As sweep on it’s own would bump up or dent surface along the way without the cross sections.
Also use edgesrf for surface depending on what looks better.
Quick drawing of what I would like to achieve below.
Not sure to understand, but why don’t use networkSfr ? you could use your 4 rails and only 2 cross sections at start and end ( and tweak that with history if you need to adjust)
I think i’ll do that with network, keep in mind you can add multiple sections in network and keep ability to tweak all rails.
if you need more sections through 4 rails without use orient each time , you can create initial network, then project straight curves on it (close to what Richard mentionned with isocurves ) , but you could tweak orientation and placement of those curves on a fly, just rebuild once you satisfied, and run network again with all sections and rails.
edit : this is possible if you create only one continuous surface, for polysurface i follow you on multisweep
Made some sample surface.
When I need to make it, I’d need more cross sections for controlling shape better.
White surface would be keep out surface, say representing a finger.
The cross section may be more complex than just rounded shape.
In anycase, to control the cross sections, I felt preferable that all the points on the cross section curves sit on a same plane, so movin them
around wouldn’t add unneccesary twist to the surface.
->so I’d like to simplify in anyway to make these cross sections.
Also, for things that I can design freely it’s not as strict, but for things I need to copy from original model into 3D I need lots of control…
@Cyver Thank you for your advice.
I’ll try the project->rebuild method, see if that can control the surface.
I used extract iso curves, but project would be much better to get cross sections.
It would be nice, if the rail curves would be the absolute rail that the created surface follows. When I tried networksrf, if the csec curves are not exactly on the rails it sort of averages the surface and sometimes the outcome is not as imagined…
