Hey there,
We had the discussion with a colleague of mine about lofts options.
I do not understand the difference between normal and tight. So I made a quick try : 3 blue lines (flat on xy plane) Two lofts (one pink, one orange). I get different control point for the edges of the loft (see the duplicated line selected with different control point turned on, on the right). Can someone explain what Rhino tries to achieves here ?
Thank you in advance 
It becomes more apparent when the distribution between the lofted curves is not even.
Normal and Tight use a different curve parameterization.
It is similar to InterpCrv when choosing Knots=Chord vs Knots=SqrtChord
Thanks for your answer Gijs - if understood correctly (after checking your answer and the help page of the interpcrv command) : the loft option influence the mathematical approach behind the loft.
I am not sure which one is best for me.
Let me expand a bit about why I asked about my question in the first place :
I am trying to model a ramp in 3D. Knowing my ramp width, and my path I used different methods (it acutally happens in grasshopper) But here is a decomposition in Rhino I jsut made by hand:
1 - I can use the center-axis of my path and after placing my profile perpendicular to the start use a Sweep 1
2 - i can offset my center-axis and use a sweep 2 instead (a method I usually enjoy, especialy when provideing the section at start and end - but I had torsional twist with a center-axis curve that has elevation variation recently - therefore i am experienceing with lofts)
3 - My colleagues made a quick script that position the profile along my center-axis and then loft (using normal)
4 - Alternatively we could use tigh-option
Measureing from one of the quad from the generated intrados of the ramp to the perpedicular opposite edge on the extrados, I realised the ramp do not have the same width with lofts !
Worth noting : seing the isocurves, it seems that sweep 2 is a bit crazy - but probably a valide options still.
Here is a zoom of the inside edge of my loop. We can see that blue and green edges perfectly overlap (or it seems to be) but I get different edges for the lofts - I think following the type of parametrization: I ASsume by adding more and more section to the loft methode will make the deviation smaller and smaller (within tolerance range)
It seems sweeps are more consistent with the geometry - so from my perspective a better choice. but I hade cases of internal torsional rotation with sweep 2 - thats why we prefer the loft method.
Would you have a suggestion on how to approach that ?
Please let me know if my explanation are confused, I can add more details
I attached my test file here for your convinience.
sweep loft.3dm (345.2 KB)
What’s wrong with the Sweep1 solution?
In Sweep2 you can use “Add Slash” to control the isocurve direction.
Hello,
Thanks again for you answers
The reason i am not so pleased with sweep is because the output get some twist as soon as I start using 3d. (see below)
I would have liked that any section is tangent to the 3D-axis but not twisting (i.e. a horizontal line stays horizontal all along the axis).
I tried to nullify that by addind an end section to the sweep but it seems to be twisting between the sections and I do not know how to avoid that !
My initial idea to correct that was to work with sweep2 but then : i do not know how to generate the edges rails without twist (see the attached my attempt with a 10% slope)
Please find attached rhino and GH script of my attempt !
loft sweep.gh (81.3 KB)
sweep loft.3dm (379.5 KB)
Hopefully this is just me misunderstanding the tools and there is an elegant way to do that. Let me know your thoughts.
As a last minute idea while finishing this message : I thought about sweeping the section along “unrolled profile” (see the script attached) to get a “pure” straight geometry, and then flow along the curve in plan.
I think it solved my challenge - however I have no idea how to check if the geometry corresponds to what I was looking for (i can only say it looks good - but so did the loft until I started zooming and measuring)
Kind regards
Sweep1 in gh has not enough control, but you can use c# or python to call the more elaborate Sweep function in RhinoCommon:
Hey !
Wow ! Thank you so much for your time and answer.
Your function is perfectly what I had in mind but did not know how to approach.
I really need to invest in C# and Rhino /GH API…
Thank you again to both of you !
