Regular Topology for NetworkSurface

Hi there,

I’m trying to build a dummy surface for flowing a regularly(ish) spaced diamond texture. I’m having some trouble with Network Surface creating irregular topology though, caused I expect by my input curves. The tip I’ve received in the past is to then rebuild the NetworkSrf to a dense point count which creates a pretty much square topology for even, gradual flowing. You can see in the below screenshots there are very different iso densities in between the construction curves. I set interior curve to 0.01 (the whole idea is to follow very closely the underlying surface which I eventually Boolean with, but to guide the direction of the texture (so it can look straight in side view.

networksrf-curves.3dm (819.6 KB)

Design in progress: not far away, but just not optimum.

Also considered: I know the green curves are point dense, that’s why I chose NetworkSrf.

Hi Jonathan - probably the best way to make a target surface is to simply Rebuild the surface with a dense point-count. The UV distribution on a rebuilt surface is nice an even.

-Pascal

Hi Pascal,

This is the best that I got to. The problem is that density-differences.PNG was the rebuilt surface. So because the original was so far apart, rebuilding just didn’t seem to be able to pull it back enough. So I faked it by inserting some knots at midpoints then rebuilding - so this was luck more than anything. Is there a way to imply or get an ‘expected’ result for NetworkSrf in terms of the topology? E.g. spacing of the section curves or point density?

networksrf-goodish-topo.3dm (1.1 MB)

Also I should add I’m using CreateUVCrv and a plane for my flowing.

Hi Jonathan - no - NetworkSrf does its own thing, independent of input curve structure. Even spacing may help but generally I would not expect an evenly distributed surface on a complex shape. I’m not clear on why a rebuild on this thing is not the way out - like 150 by 36 points

-Pascal

Hi Pascal, okay thanks for the clarification. There’s also a surface parallel to the ground which connects to the red surface, so it creeps under and then the flowing has to go across the seam - which was my reason for creating the separate Network dummy surface (shown in Green).

The result which is not so bad really - whilst it’s certainly a manual process, I guess once you know what you’ll encounter it’s straightforward in a way.

Oh also I need to recreate the underlying surface to get the straightness in certain areas - so the green horizontal curves help guide the NetworkSrf topology so the texture is then straight when viewed from the side.