This is one of those little things that slows me down daily. And I wish that all of these ‘splitattangents’ options were actually ‘split like source curves split when exploded’, because I need those splits so that I can edit polysurfaces efficiently.
Hi Peter,
do you have some example geometry ?
c.
Sure - see attached. I’ll create an example of why I need this, too, in a bit.example.3dm (244.1 KB)
Hi Peter,
you may run _DivideAlongCreases on the resulting Sweep1 surface. Set the option to _SplitAtTangents=Yes
c.
I tried that in an example and it worked very well. Thank you - that will help me a lot!
peter
Well, today I have my first real-world opportunity to use that tool, and it does not work in this case. I guess I really need things to split at the same places that the cross section would split if I exploded it.
Attached is an example where dividealongcreases does not help me. The section curve is copied from 3 surface edges. The surfaces are G1, but even if they were G2 or higher, I’d want to end up with 3 surfaces without manually splitting them.splitatcreases.3dm (217.7 KB)
just to be sure…
are you using the isocurve option in _Split when manually splitting the surfaces afterwards?
if not, that should speed things up for you… if that’s how you’re already doing it, i don’t have any advices 
Hi Peter,
i see. one way around it at this time would be to explode the cross section and use sweep1 three times using the same rail. Alternatively, you could enable the knot snap and split the surface along isocurve as jeff suggested.
c.
Hi Peter - I see - I guess in this specific case Explode and Revolve will get you the surfaces you want as well I guess. But I see what you’re after. I think the current splitting behavior is designed to be a compromise so that not all splits happen at strictly speaking G1 locations - this was, if I remember right, to avoid large numbers of surfaces from imperfect but not horrible curves. I’ll see if @dalelear has an idea… if I remember right the splitting criteria were typed by Dale.
-Pascal
Jeff, yes I use that, but it’s tedious to do it over & over. Clement, yes, I could sweep each curve individually, but that’s also tedious.