Network Surface and Curve Direction Troubles

Hello All,

I would like to create a Network Srf, by feeding Grasshopper with some Rhino geometries.
Here we have two surfaces (the closest boundaries between them are the U components) and 3 Blend Curves as the V components.
Network doesn’t work.

So I’ve found that the vector directions of the blends acting weird when approching the Position (the continuity between the Net and the lower surface might be G0).

  • On the first the vector disappears.
  • On the second it flips (as shown in the image below)
  • On the third it rotates strangely.

Furthermore - when the Network works - how can I keep G2 between it and the upper surface?

Thank you in advance. (44.7 KB)

This could be one way of what you want? I’m not sure… (48.5 KB)

1 Like

Hello and thank you for the reply.

That is one cool way to have something like a fillet and it’s really useful, but…
In this case I would like to have something more adjustable (that is why I’ve placed 3 blend curves with different continuities). Everytime I create a blend curve and set G0 something goes wrong with the vector directions :frowning:

1 Like

Yes, I see.

But what I want to achieve is the total control of the surface, by a set of sections (curve blends).
In your way I can control the tension, but that tension is constant all along the rails.


The Blend component with the bulge=0 and continuity=2 options seem to create a curve with three duplicate control points at the bottom. I think this is what is making NetworkSrf fail.

Sweep2 seems a good alternative, but I think @HS_Kim solution is better.

You can Merge 3 sliders to control the values for each section.

Thank you for the reply.

Merge the blend curves fix the problem, I’m sure the duplicate points were the reason for the network failure.
Anyway with the Sweep2 in this case I cannot control the continuity all along the Rail 1.

Going in another direction, I tried to used the RhinoCommon method for BlendSrf, which seems to be the appropriate tool for this (neither Sweep2 not NetworkSrf maintain G2 along rail 1, especially with only 3 sections).

But, unfortunately there are a lot less parameters available in the method, compared to the Rhino command… as reported here.

So I tried to move the first two points by equal factors, which I thought would maintain G2 continuity. After that I used Surface From Points and encountered a strange issue…

The surfaces won’t join if I use the result from Surface From Points. They do if I use what C# BlendSrf spits out.
And this even happens if I don’t move the points at all… See attached.

The surface rebuilt from Surface From Points appears to have some waves… and won’t join even in Rhino.

Which is not the case for the blend surface from RhinoCommon.

So, there must be something with the surface, and after checking points, degree and order, the last thing I thought of was the knots. It happens the vectors differ !..
So I changed them in another C# and hooray, everything joins.

The Zebra analysis seems to give similar results between a manual BlendSrf in Rhino, with bulge factor = 0.25, and the reconstruction in GH. (47.7 KB)

Thank you,

Your solution works perfectly and it’s a cool way to adjust that surface; even if sometimes I guess some easier strategies should be find, for example a GH version of the Rhino Blend Surface or a Network with continuity options.

Sorry to bother you again, but…
What if I want to add more adjustments here? For example a point between the center and sides? How can it be feasible?

Something like this.
You can add as many intermediate values as you want, as long as both lists have the same amount of items. (50.1 KB)

That is AWESOME.

Thank you.