Help required for creating a chair shell from curves

Hi all, trying to dust off my previous rhino skills with a chair design, im modelling half the chair shell so i can mirror it and the extrude / offset surface and cap for thickness of material.

I want the edges of the chair to wrap around the red curved tubing ( its 2mm bigger tha nthe intended tube for tolerance, so preferably as tight a fit as possible. I have tried loft command and surface control point grid.

Loft doesn’t seem to work for me. Perhaps something is missing in my approach to using the curves, i’m not really sure can anyone get a decent loft out of the curves,

The Surface control point grid command kind of works but i really want a surface to fit tightly against the curves that i’ve created on the U and V directions. I’ve attached the file for comment.

Any help would be greatly appreciated.


NickCHAIR SKIN.3dm (509.3 KB)

Hi Nicholas - I’d think about this as three surfaces, possibly two - one that wraps, basically an offset of the pipe, the chair surface and a transition surface between.

CHAIR SKIN_Maybe.3dm (686.7 KB)

Where you trim back the surfaces will determine the transition - I’m not sure I really got it looking right here, but just to indicate a layout…


Thanks Pascal, how did you create the transition section, a loft from extracted iso curves?

I’m going to rework the pipe to a j shape, so that it can be fitted / draped against the frame rather than snap fit.

It’s been 4 years since I used rhino as I’m primarily solid works nowadays so appreciate the shortcut tips!



Hi Nicholas - yes, it’s a loft and then MatchSrf to the two input surfaces - the trims on those would be important to get in the right place - they are probably isocurves on the pipe for at least some of the way but not all, I’d think. Once the loft is in place, you can move the edges and with History on MatchSrf it will stay stuck to the pipe or chair surfaces.


Amazing thank you, i’ll post up with result once i’m finished for comment. Always good to get a second opinion on technique!




So I have made a loft to join my 2 surfaces together. However when I join all three surfaces and then offset them to create my desired 9mm thickness i find patches of missing surface on the new offset surface. Why is this so? Do i need to re trim the edges joining them together or is there a command that i can use to tidy the surface up?

I’CHAIR SKIN 21-1-19.3dm (1.6 MB)
ve attached the model for reference.

Hi Nicholas - a couple of thingd - one is, if you are going to offset surfaces, make the design surface the primary - the one you build, and the offset the ‘back’ as there is always some loss in surface quality on offsetting. So here, I think I’d build to offset back.

Another is, there is a step missing which is making sure your transition surface is at least tangent to both the inputs all along the edges, preferably curvature continuous.


But for that to look right, you need to carefully adjust how wide the surface is and where it meets the inputs, especially if you match (MatchSrf)for curvature. With tangency, you can get away with what you have I think.

CHAIR SKIN 21-1-19_PG.3dm (323.1 KB)


Hi Pascal, are there commands to assist me in 'curvature continuous or is it a case of rebuilding that transition section (I used ‘‘sweep2’’ with about 5 cross sections) until I can offset it correctly?

I don’t understand your last comments on adjusting how ‘wide the surface’ is and ‘where it meets the inputs’

I feel so close to getting this shell finished!


Ok so ive added some more cross sections and the result is an offset surface with no visible holes or missing sections. I do notice there is some unwanted detail along the centre cross section of the chair on both the front down stand and the back rest. I can trim this off using a line 0.1mm offset, but just wondered why is it this appears?


CHAIR SKIN 22-1-19.3dm (1.2 MB)

Hi Nicholas - maybe it is fine - but if you look at a shaded view in dark blue you’ll never see how discontinuous the surfaces are for tangency -

If that is good enough, you’re done, but I’d say probably not…? You can do better by making cross section curves that are actually tangent to the surface edges they are bridging

Basically, I think, in order to get this type of ‘smooth’ surfacing down, you need to get the concepts of curvature and tangent continuity ‘internalized’ so to speak- the level 2 training material here has some info that will be useful -


Many thanks Pascal!