top and side view egg shaped closed curves and widest part of the egg as elipse.
is this not enough to surface it ?
asks for more , but tracing the photos this is all I can get.
The egg hoops should tell the green elipse what size to take on at any one point along their length.
Whats the method ?
Its a pig of a shape, as it has gullies at N S E and W.
so tracing plan view makes you think you have the max shape but one has to allow for the material missing with the gulley extraction !
so the end view with elipse riding over the gulley, then scaling the egg hoops to fit that, sees me here, after much head scratching !
here is the file surface from three curves.3dm (41.8 KB)
Its asymetric so make it from two ellipsoids of different lengths but identical mid-section, trimmed at the green line and joined. Use the radii of your three curves to define the dimensions of the ellipsoids.
Hi,
Revolve, is a non varying radius.
which curves did you pick for the command as the radius varies.
Green is not a circle.
Ellipsoids,
It wasnt drawn as such, I traced the porcelain item which was hand crafted.
I need to honour the actual trace drawn with curve tool.
I try for Ellipsoid but get a mismatch.
Hi, I think using Sweep1 will match the curve tool made curve, I had used CurvatureGraph after a Helvetosaur macro I placed into a custom button.
! _CurveThroughPt _Pause _Degree=3 _CurveType=_Interpolated _Knots=_Uniform _Closed=_No _Enter _SelLast _Rebuild
I had placed points along the edge, then run this. I had to adjust the end controls to get start and end tangent.
Perhaps an option in this to pick a point and tangent direction for start and end would be good
so the tips of my lozenge are dead flat perfect.
I split the curves as seen, and then do 4 sweeps, green purple magenta purple pick order each time.
after these four Sweep1 I join surfaces then zebra it and see a slight step at the sharper end, the other end is perfect.
Why is that ?
is this Sweep1 an ok method, or prone to small problems ?
For me it uses the exact curve, so is max accurate.
Should I have purple and mag as rails and green as profile ?
A rather odd way of doing things. Just redraw your two main curves as proper NURBS curves (see how many CPs they have, and their weird curvature comb), Revolve, then scale a tad sideways.
Hi, Ok I will try that,
I use Rebuild and 10 points degree 5.
The end points are not tangent so move them then nudge and the next one and so on to get a visual best match again.
Has Rebuild an option to stop end points moving inwards so spoiling the tangent of the original curve ?
do this on both magenta curve and purple curve.
decide revolve entire magenta, then non uniform scale wider to get to purple.
Nothing to ‘latch onto’ so use an eye match.
My sweep1 used actual rails, . surface from three curves split and sweep1rail.3dm (973.9 KB)
I fail to see why a simple Revolve with minimal curves, then scaling the egg according to the underlay, is not the way to go. You obtain the exact same form, without any noodling, and for the irregularities you want to introduce, you can just noodle some CPs. Revolve and Scale.3dm (2.2 MB)
we only need 1/8 of the “egg”.
dark green: a helper surface to mirror the surface (the surfaces matches tangent - the final continuity will be curvature)
light green: a helper surface to rotate (!) the surface - the edge is symmetrical so the rotation gives a nice transition , the current cplane in the screenshot and the cyan arrow show the rotation-plane - the final continuity will only be tangent … but close to curvature - and curvature in the center… ok-ish. or match G2 to the helper surface resulting
spheres, ellipsoids, revolve, railrevolve… all those surfaces end up with a pole.
as soon as you want to design / CV-Edit this kind of surfaces the pole - which is a collapsed, 0 length edge - will always be unsatisfying. So if the final object does not need the section of the pole - skip it already for the primary surface. (?)
@steve - does your initial part has this symmetry ? or does it have different geometry for each cut-out ? - i still would follow this approach - just do 2 different surfaces, that are used 4 times each.
Just a bit of trivia: these objects were used as insulators in broadcast and shortwave antenna systems to connect tuned length antenna elements to supporting wires . Also used in antenna tower guy wires to break them up into shorter sections to “detune” them from resonating at the broadcast frequencies. Incidentally a great illustration of the strength of ceramics if you thought about what they were actually doing. I imagine they found other uses as well.
