that looks close to perfect 
2 Likes
Now I feel like I should be using deg 5âs more often 
Just a slightly tuned version of yours to reach 0 degrees tangency.
2 Likes
In many cases itâs more important to have a smoother transition (nearly zero degrees of tangency) between two adjacent surfaces than a perfect curvature.
1 Like
Would you consider 0.02 degrees to be ânearly zero degrees of tangency?â
2 Likes
0.02 deg⊠ATN(0.02*PI/180) ~ about 3 tenths of a millimeter over a meter of surface length and we know the surface has further curvature so you donât have even a theoretically easy to measure 0.02 deg angle between two large flat plates.
It obviously depends on the application, but assuming that whatâs in the drawing is a part for an object sized somewhere between jewelry and a car and that someone is going to mold/cast/glue/weld/etc. it, 0.02 deg is probably tighter than the engineerâs intended tolerances and indistinguishable from zero to QA and the customer unless there are some unusual circumstances in which tolerances would have been specified.
It might even be a joint thatâs going to get a surface treatment which would make the joined surfaces effectively continuous in production. Sanding, polish, even paint or a thin clearcoat.
Edit: just checked the drawing settings- degree tolerance is 1 deg, so yes 0.02 deg is ânearly zeroâ
2 Likes
In my opinion 0 degrees is closer to ânearly zero degreesâ than 0,02 degrees.
1 Like
And how much further, in your opinion, is a 0.3 curvature deviation from a .08 deviation?
1 Like
I actually had this problem recently trying to make various panels on the car Iâve been making.
For that particular application, i have always chosen tiny tangency breaks over the curvature, as car paint materials and the zebra tended to be completely unforgiving with curvature breaks over tangency. Anything over 0.2 curvature looked awful. On the instances where there were tangency breaks, I often really struggle to see them, other than at really odd angles.
But again, this is a very specific application.
1 Like
In this particular case, the surface we talk about is highly curved and bent at about 90 degrees, which means that the tangency break is easier to spot than a curvature break. Curvature is more important on nearly flat and large car body panels, but even then there are exceptions, because the general control point flow is even more important.
1 Like
Could you explain how you âglueâ the surface edge onto the flat front surface? Is there a âmatch-and-projectâ method in Rhino to do that? You mention that for that last matching, you used âReplaceEdgeâ and âOnSurface=Yesâ, but the exact workflow order, step-by-step, is unclear to me. Achieving G1 or G2 on the other three edges works well. But there is little how-to information about âReplaceEdgeâ online. Thanks!
1 Like
@Nathan_Bossett That corner is part of a forestry work vehicle, and 380 mm wide (large flat front surface), and moulded from glass-fiber reinforced plastic. That particular OEM has very high surface quality standards. A somewhat comparable product, the CLAAS LEXION 8000, has highly controlled surfaces (it won design awards).
1 Like
Hello- you can use MatchSrf > OnSurface=Yes and then skip choosing a target curve - the edge will be pulled to the target surface by closest points. (or you can put a curve on the surfface there and use that)
The nice thing is, this ignores trims and matches to the underlying surface.
In this case you probably want âMatch by closest pointâ and âPreserve isocurve directionâ
-Pascal
3 Likes
Itâs worth noting that in this case - since the surfacing being matched to is planar, ANY position of the verts along that edge, so long as they sit on the same plane of the surface, will form a valid, watertight edge. Youâll notice that the second row of verts that hold tangency are planar as well, and the same goes for them. When matching to a planar surface, you have an incredible amount of freedom to sculpt the edge and surface.
2 Likes
Agreed
Thanks for the details about intended use @Lagom !