Joining/ Merging/ Combining two overlapping/ intersecting polysurfaces

Hi there!

I am having trouble joining two polysurfaces.
Although there are two nice tutorials on youtube that address this topic (Intersecting multiple pipes + Y branch modeling)
they do not offer a solution to my problem.

My setup is as follows:

I applied a sweep1 command

And those are the areas that need to be joined with a smooth transition in between.

I am dabbeling in creating a pendant and plan to send it to shapeways for 3d printing.
Hence, the file should be .stl
Does anyone have advice on how to go about this?
(Will a “group” of polysurfaces suffice or must all parts of the model be joined perfectly for proper 3d printing?)


Hi Octavia- I would assume the latter- some systems can handle inter-penetrating solids, but I do not think that is the rule - certainly Shapeways can tell you.

If you can post your object so far, we may be able to suggest ways of making the transition surfaces.


Hi Pascal!

As you suggested I made an inquiry and shapeways (and also a smaller company in my hometown) do prefer nicely booleaned, airtight 3D models.

That is my 3d model so far:
transition.3dm (278.9 KB)

I watched another batch of tutorials and was thinking that maybe I should go this route with my model:

Sweep1 command --> Extract wireframe (Curve From Object) --> split the wireframe at those critical points --> explode my section curves --> Sweep2 and build the transitions manually.

What do you think? Is there perhaps a more elegant approach?


Hi Octavia- probably there is a better way- but I am not completely clear on what transitions you’d like to smooth- something like this?


Hi Pascal,

yes, something like the transition in your picture (the upper model) - The Y-Branch :smile:

and for the X-like intersecting pieces a some kind of trim-the-inferior-and-blendsrf-with-superior-polysrf-operation…
Add-boolean does not work because the base of both sweep-geometries are located at the same height in z-axis, I guess.

Ok, so, I think it is a matter of planning a bit- you’ll need to decide the limits of the transitions at each location and trim back the surfaces to make the required gaps to fill with the transitions. It will be a bit of work- but you’ll get the idea. The attached is by no means clean enough or complete, but just by way of showing you how you might approach things.
transition_Pg.3dm (452.2 KB)


I would blend the ends like this.

Hi Pascal,

thank you for taking the time and handling the file!! :smile:

So did you basically just trim the polysurface with a geometry and the blendcrv the curves?

And how about the transition in the lower left model of your previous picture?
Did you perhaps sweep1 with only one half (trimmed) of a section curve on “each side”/the trunk and joined or blendcrv-ed it?

What do you suggest is the most efficient way to join the x-intersecting part of the 2 polysurfaces?
Using simpler section curves (with no fillet, just corner) -> sweep1 -> -> and then manually rebuild the intersecting piece --> filletedge chained edges?
Or are there more efficient solutions?


Hello Stratosfear,

well that looks neat!
Do you mind explaining how you achieved that transition?


Hi Octavia- I’ll attach a little cleaner file that will give you an idea how to construct this sort of thing. One thing that I noticed is that the small rounded edges on your sweeps are no at all tangent to the upper and side surfaces everywhere- that will make life difficult in adding details. I would make this as a hard-edged shape and kill the edges with Fillets or Blends after the fact.

Stratosfear has reworked your design much more extensively- pay attention to his approach, he’s good at this.

One thing I think will happen as you go is you’ll think more clearly about exactly what you want the design to be like - as posted, your idea is incomplete, you need to have a good idea what those transitions really look like, then it becomes easier to come up with strategies for surfacing them.

transition_pg.3dm (557.4 KB)



Hello Pascal,

hmm, what do you mean by “small rounded edges on your sweeps are no at all tangent to the upper and side surfaces everywhere” ?
How do I recognize if my fillets/round edges are not tangent?

Yes, you are most probably right. I was also thinking about making those sweeps with hard-edged section curves, blend them and then apply a filletedge.
Basically, I had a transition like that in mind, although in this picture it is a ring:

Thank you for that “cleaner” file, but to be honest I do not get any wiser as to what methods or commands you have used just by looking at your model.
And I can not follow your steps just as easily as I could if this were 3ds max + modifier stack.
I suppose you exploded the polysurface, trimmed the y-branch and did a couple of blendcrv commands and the patched or joined it? yes or no? Am I close?

Could you also please break down or at least give some hints on how you modelled this 2nd transition:

It looks nice as well and I would be also interested in that technique.

As to the second transition, the x-intersection, this (your file) is exactly what I had in mind, but again I am just vaguely guessing how you might have done it.
Did you explode the thicker-arm-polysurface, trimmed it with the thinner arm or a copy thereof and then blended the edges?

I am also somewhat confused about “Stratosfear has reworked your design much more extensively- pay attention to his approach”
Is a Rhino-Pro able to recognize the modelling-technique by merely looking at the isocurves of the transition? If so, then I am really sitting in the back of the class…

I am a beginner, so sorry for all those questions and thank you for your patience :confused:


Hi Octavia- I’ll make a blow by blow file for the first transition - the second one (‘X’) is done with FilletSrf (surface to surface fillet tool) then trimmed after all these are made. I had to replace your edge ‘fillet’ surface there thoug because they were not tangent - see the image below- the red is a section curve through your object with a shaded closeup on the right:

Probably not what you expected, right?

I pointed out stratosfear’s input because, yes, as you suggest, 9/10 of the battle is figuring out a clean and logical arrangement for the surfaces, and he shows that nicely in the image.

Here’s an updated file- transition_pg_steps.3dm (1.2 MB)



At least you should post a file to look into…


Should I make a Class-A modeling tutorial of this? If that’s necessary, I would find a time to do this.

Well, just looked Pascal’s file. He did it…

Hello Pascal,

thank you very much for your effort and the file!
Sadly, I have been ill since last week and did not feel well enough to boot up my computer and have a look.
I will though in the weekend.