SetPt is a command that can make what you did with the corner curves a lot easier (matching the points of the top horizontal curve with the mid-horizontal_ish curve…
what this will do is project all of the points of the curve onto a single plane then you can then move this plane & flattened curve in a direction constrained to the X,Y, or Z option you’ve chosen…
so, these two curve’s points now have the same exact coordinates if you were viewing them from the top:
Okay, and thank you for the help, but that’s the reverse of what’s needed. I don’t want to match the curve generated in the middle with a blend but rather match the curve setup in the original guide lines. I can project a curve to a cplane and flatten it -I’ve progressed that far but again that doesn’t suit the task. That’s why I’ve resorted to using the pull function earlier - to conform the middle curve to the upper. The upper curve is the parameter that must be met. Your results are very clean but as far as I can see they don’t coincide with the radiuses of the top and bottom horizontals.
Hi Jeff,
The top curve is a copy of the guide line created for the perimeter of the camera. That’s the ‘master curve’ that I want to conform to on the vertical axis.
The vertical surface is made with the edge curves tool and the horizontal is planar curves. Blend surface is then run along the edges with the tangent section for both sides. Additional shapes are added at the arrows as well as the two arcs on the corner. Adding two additional shapes at the quadrants on the vertical surface seem to help the overall smoothness.
but unlike OSTexo’s beautiful output, I still get less than stellar zebra results.
but he’s outside of the constraints you’re placing.
where are you getting these dimensions from?
upon further inspection of your model, your ‘musts’ are making for an impossible blend… or, impossible to get the type of blend i think you’re imagining… personally, i think the surfaces leading up to the corner need to change where they stop a little… as well as allow for a non-arc or two through blend…
going with the outer curves of your corner, the surface will be something like this:
I tried to explain above that the combination of straight lines and arcs, and associated surfaces is fundamentally incompatible with G2 blends without areas of tighter curvature. What @Jkayten wants is not possible without changing from arcs to smooth blend curves.
[quote=“JKayten, post:27, topic:28619”]
davidcockey:
What @Jkayten wants is not possible without changing from arcs to smooth blend curves.
David-So you mean the arcs that are defining the horizontal curves? Those should be upped to a g rating?
[/quote]I mean that there is a fundamental geometry challenge with what you are trying to do, which is independent of whether you working with Rhino, other CAD software, paper and pencil, or a physical object. Starting with arcs and straight lines; or flat, cylindrical and spherical surfaces which have tangent (G1) continuity and trying to smooth the transitions to curvature (G2) or higher continuity while maintaining a portion of the arcs will lead to areas of higher curvature in the transitions. These areas will be apparent in curvature combs, zebra plots and reflections (and by feel if a physical object). Increasing the B-Spline degree and/or continuity level will not eliminate the problem.
If you want G2 or higher continuity between straight lines or flat surfaces do not start with arcs and cylindrical surfaces. In Rhino use the blend commands to fill in the gaps in the straight lines and flat surfaces. But that does not how your camera was made. It appears to have G1 (tangency) continuity.
you’re trying to fill in the green hole above… that you put an arc (red) at the bottom and move the sides in only deceives and makes it look like the hole is looser/more bendable… but your requirements of the mid-horizontal line matching the top horizontal line still creates this funky situation:
use Scale on each arc in the array… 1st click on the point where the arrayed arc and the red arc meet… 2nd click where the vertical line and the ground line intersect… 3rd click at the intersection points of the ground line and the blue surface:
…or, at least, that’s how you’d handle it if you were building a skateboard ramp
it’s not going to be “perfect” since the arrayed arcs are only perpendicular to the red arc and not the blue wall but… definitely close enough to prevent any bumps and valleys and whatnot.
that said, i finally saw that there’s a camera picture hidden in your file so, i don’t think this is an approach you should follow… the camera designers almost certainly did not go about the design like this… something between what they did and what you’re trying to do are different.
Terminology: “G ratings” such as G0, G1, G2, etc are used for the level on continuity between adjacent surfaces and/or curves. They are not generally used to describe B-splines. B-splines are generally described by degree, such as degree 3.
I think there must be something amiss with the mesh settings or some other aspect in my program. I’ve simplified everything and I get a reasonable looking zebra with the surfaces in the attached model however upon joining the surfaces that were used to create the blend a ridiculous result appears.
