Hello.
I’m struggling with one (easy for expert?) issue.
As in the picture I placed surface on network curves. The curves are given.
By now only profiles are tanget to the red edges, enything between looses tangency.
How to make “smart” surface which tanget betwen profiles to the edge?
Thank you,
Codi.
surface_not_tangent.3dm (96.8 KB)
Hi Codi - if you extrude the long curves and use the edges of this extrusion as the NetworkSrf inputs there rather than the curves, you will be able to set tangency in the NetworkSrf controls.
Also, using just the curves, SetSurfaceTangent might help but I don’t know if the result is the same, exactly, as making the surface tangent to an extruded surface from the NetowrkSrf controls.
-Pascal
Hi Pascal.
I’ve tried do my best and implemented recommended ways.
The surface now is tangent to curve (actually to extruded curve) but on a short distance there is a large distortion.
It looks like “tangency” is modifing the surface in a tight neighborhood of the tangent direction to the edge.
Below trials which I did:
Surface with “pure tangency issue”
Surface with tangent to extrusion
Surface with “SetSurfaceTangent”
Really appreciate ideas and ways “how to”.
Thank you,
Codi.
Hi Codi - I am not sure what you’re looking at but here’s what I get:
surface_not_tangent_PG.3dm (212.6 KB)
Does that look right or?
-Pascal
Hi Pascal.
There is sudden tangency change between profiles in a short distance on the surface. The effect is seen in a surface analyze.
Both guiding profiles are “smooth” and the isocurve between them are “tangency bend” just before extruded edge. It is clear visible when there is larger difference between shape of profiles and their size (begining of the shape).
Most visible issue is at beggining of the shape.
As a result offseted surface is deformed.
Thank you, Codi.
Hi Codi - there’s a change in curvature, yes, but as far as I can see the tangency is correct, right?
Here’s what I get, offsetting .5:
surface_not_tangent_PG.3dm (772.2 KB)
Does that look reasonable? I am not sure yet what you’re seeing that is bugging you…
-Pascal
Do you mean the rapid change in curvature as shown by the curvature graphs?
@Dear davidcockey
Yes, I’m trying to figure out how to make region, next to the red edges, smooth like on the profiles 1 & 2. In my imagination sections between profiles 1 & 2 should be similar to profiles generated by “tween” function
@Dear pascal
You are right, tangency is correct -> there is one issue still existing (my headache now) -> it changes rapidly just before red edge line. Profiles 1 & 2 are “smooth” (actualy profile 1 is half circle, so it has constant tangency).
On the isocurves I mark it in red circle (previous post).
I did offset not by -0.5 but -5 and it becomes like in the picture below.
It is even more clear at the end of shape…
What could be a right a way to build a primary surface in a such a way, that after offesting it by -5, the new one would be without red circled “bumps or deflections”?
Thank you,
Codi.
Hi Codi - Well, I still say the best way to do this is with matching curves and a series of simple surfaces. NetworkSrf makes a dense surface, so the second row of points, which set tangency, is very close to the edge, which is what is driving that curvature up near the edge. But, my surface posted above offset cleanly, I think.
-Pascal
Hi Pascal.
Above screens are with your surface 
after offset -5 mm.
The best results which I tried is with 2 rails (profile 1 & 2 works as a rails) and cross sections are 3 lines (below picture).
The result is … perfect … again there is one “but” … it works only on separated segments 
If I split the shape for segments and treat all of them with “2 rails” - the tangency between segments is different and so far I didn’t find a way how to connect them smoothly.
Any ideas?
Thank you,
Codi.
Hi Pascal.
First of all thank you for your contribution and support!
I apreciate your suggestions which helped me to work out the best solution to me. Hint with “SetEdgeTangentDirection” was like dot over i 
I belive I found the way which gives to me a perfect result.
If there is a easier way I would appreciate any feedback.
Maybe it will be helpfull to others, so I will describe my way:
surfaces from Sweep2Rails and splited lines can be deleted
what stays is 3 lines (outline) and a lot of sections - use SurfaceFromNetworkOfCurves
Final result:
Previous:
Best regards,
Codi.
