Hello, had a client who needed me to put a ~2mm fillet on these connecting planes.
Since theyre different angles I had a tricky time to make them meet, but with more hours than I’m willing to admit I put in to solve this “simple” task I made it.
Got me thinking that there must be a simple solution?
What would your solution be to solving this request?
Hi, I exploded the polysrf, then used surface fillet. Split the fillets near the junctions. Then used blend curve to trim the main surface. Then used Xnurbs to fill in the five sided hole.—Mark
Rhino fillet.3dm (252.0 KB)
Nice!
Seems like Xnurbs speeds things up ALOT, have you had any problems with the plugin or is it a must have for surface modelers?
The part I did took 5 minutes.
xnurbs is helpful and no problems on my part. I bet if other people looked at your model, they could have done it without xnurbs. —-Mark
If the image below is an acceptable solution, then Google “atomic bomb fillet” to learn this technique. The surface at the intersection was made with _Patch.
As long as your input geometry is clean, XNurbs will give you geometrically good looking and matching (g0, g1, g2 or a mix you choose) results that are often hard to match by just plain Rhino tools in a fraction of the time it would require doing things manually.
Using the given surfaces, a “ball-like-fillet only solution” might look like this.
all green surfaces are done with
_filletSrf
the violet part is a
_sphere
but as a designer - i appreciate solutions that handle the filleting as part of the design.
if you not draw the surfaces first, and then look for a fillet-solution - but already think about filleting, when the surfaces are constructed
If 4 surfaces meet in one corner some kind of hierarchy is pleasant for the eye.
or
a solution without a hierarchy, but with “ongoing edges” … for this solution i had to move the surfacest:
kind regards - tom
Here is an all rollingball constant radius way-
Rhino fillet_PG.3dm (223.4 KB)
It is a bit of a puzzle sometimes to see what needs to be filleted with what. The little triangular piece on the right is a trimmed sphere.
-Pascal
any reasons why you don t use the typology i posted above ?
i am not a big fan of those small triangles that form a sharp corner
what about this ?
(giving more priority to another sphere that is rolling along…less to the sphere that is doing the inward fillet )
Rhino fillet_v2.3dm (3.2 MB)
by the way, i like the emap you re using - is there any ressource for downloading the corresponding emap-pictures ?
That looks like the standard fluorescent.tube.bmp file which is/was included with Rhino.
Hi Tom - I am sure there are many perfectly good solutions - I was intrested in one which was all ‘true’ constant radius fillets.
I do not think this surface is a fillet, at least I do not see it:
The fillet beween that plane and the wrapping fillet at least, is not that surface.
Hm - all of this is somewhat far from the origin, I see- far enough to mess with the display, here.
-Pascal
it s all _filletSrf
OK - no objections from me then… =)
This also looked possible:
but in practice it does not line up quite -
The fillet surfaces come out much simpler if everything is moved to the origin.
-Pascal
of course there is some mystery about the fillet algorithm in rhino:
i think there is a small _filletSrf missing. - then it lines up nicely.
… adjusting the initial surfaces will allow nicer results.
that s why i wrote the “design” section in my initial post.
Making the edges like your design sample makes it a lot more interesting, thanks for sharing!
Is it a normal “_Patch” you did to connect everything? all I get is naked edges if I try “_patch”
Here’s another possible option.
Make the first fillet along the top edge a little bigger, this will give you the space to blend the smaller fillets through the bigger fillet. ends up looking pretty decent.
Just use the command fillet edge and change Railtype ‘Rollingball’ to ‘DisbetweenRail’ then you will get the basic layout.
