Egg Design... With a surface fill or loft?

O.K.

If one we going to model one half of an egg (top or bottom) with an surface fill or loft how would you go about it. I know I could probably do a surface revolve, but then suppose I wanted my egg to be elipitical about it's center rather than round.

Many thanks in advance,

Dave

Reply to
VFR
Loading thread data ...

Reply to
Sporkman

There are more than a few ways to do this and don't know many round eggs or why you are not wanting to use a revolve, the easiest way, imho.

Or, you can do a combination of sketches which use either a spline or a ellipse or arc to achieve this as well using a revolve.

Anyhow, don't know why you want to use a loft or fill but.. here some very simple, start with a circle extrude a disk, do a dome X-distance (ellipse), cut away the disk, do another dome (ellipse) X distance, then you have your elliptical shape on each end.

Anyway, to me, this is a egg.

formatting link
BTW, there is a elliptical egg on the SW parts library doing what you want.

.. 8^)

VFR wrote:

Reply to
Paul Salvador

I understand the question differently: I think Dave is wanting (in effect) a quarter circle or quarter ellipse, swept about a quarter-elliptical path (rather than a circular path, as in a revolve) and then mirrored in x and y. In other words, a hardboiled naked egg, cut in half through its LONG axis, lying on its cut face in a serious gravity field. Or the carapace of a ladybird (ladybug).

Then again, I may be wrong!

If I am not, the above would be one way to do it, but tangency at the ends (so the finished egg does not have visible "seams" at the quarter points) would require a bit of extra construction. Perhaps a better way is to loft just one quarter of the half-egg, with a quarter ellipitical guide curve, and specify tangency at start and finish as "normal to profile". (Make sure you turn on "advanced smoothing", but don't bother with "maintain tangency")

Perhaps I should explain that you can loft between two profiles which are joined along one edge- it works best if they are at 90 degrees, as in this case (or less). This is not widely known.

This method gives good parametric control over the shape: essentially the whole model is driven by three mutually perpendicular sketches, each consisting of two straight, mutually perpendicular lines meeting at the origin (each converted from one of the other sketches), and a quarter ellipse joining their tips. The loft profile sketches will all be solid lines; in the case of the guide curve sketch, change the straight lines to construction geometry. The three sketches will only need three dimensions in total (two in the first, one in the second, none in the third) This produces a result which, when mirrored about x and y, passes the zebra stripe test. (No detectable seams, but ever so slightly dodgy around the peak - this is almost unavoidable, as far as I know -- in SldWks -- in the region of singularities, especially on three-sided faces)

The simplest way would be to do a revolve or use Paul's dome method, and then scale it down in one axis only.

Another, and in fact I think the best as well as easiest method, is to dome an ellipse with an elliptical dome.

[[-- I have done this to model a stiffening emboss in the floorpan of a domestic oven, and the draw tooling to produce it. Actually, that's not strictly true, I finished the job about a week before elliptical domes were implemented ! I used the above loft method, but back then SolidWorks could not loft right to the singularity at the junction of the profiles, I had to stop the profiles just short of touching, which left a tiny hole down this junction I had to patch over.--]]

This method gives an ABSOLUTELY perfectly seamless result, even when mirrored yet again to yield a complete oblate ellipsoid. Good enough to be used as the "squashed" keel bulb on an IACC yacht, to which it looks quite similar if I adjust the parameters suitably.

To get the result I think Dave wants, you dome, rather than mirror, in the opposite direction. This does give a seam, although is only C2 tangency which is missing, and this is inevitable wherever two half-ellipses of differing aspect ratios meet. (I should explain it only looks like a seam with zebra stripes on, with high-quality shading, even shiny + specular, it looks fine)

The most off-the-wall way approach to this type of problem is to extrude an ellipse to the desired height, then use "fillet with hold line" (the hold line being the original ellipse) to trim it back- there used to be an example in training of a conduit head (looked a bit like a conquistador's helmet) done this way - very classy. In this case, however, it is not actually usable, because fillets are essentially circular in cross section, whereas this job requires an elliptical "fillet".

I hope I've been sufficiently clear, because I'm just heading upcountry on a trip (holiday, yee ha !) so won't be able to clarify further.

Reply to
Andrew Troup

What about first constructing an axis aligned half egg using a surface revolve, and then stretching/squashing it by inserting a non-uniform scale feature to make it elliptical about its center?

Reply to
Paul Delhanty

Actually, I'd say there's a good reason to use a loft. This is because VFR is looking to make an egg that is elliptical (or otherwise non-round) in at least two axes. The lofted shape can escape the limitation of the circular path of a revolve.

With a series of sketch profiles (plus one or more guide curves, if desired) the double-obblong egg shape is quite easy to develop. This method, however, depends upon using a sketch consisting of nothing more than a single point at each end of the ellipsoid shape. (The profiles are sketched on parallel planes and smoother transitions in shape are created when the number of profiles is more numerous closer to the tip(s) and the distance between the sketch planes should diminish as the tip is neared.)

It's rather amazing that the loft can end (or begin and end) at just a point...

Reply to
Per O. Hoel

The weird trick is to make the loft end (or begin) at the point 'normal to profile'. It doesn't make a lick of sense, but give it a try - you should be able to get rid of all guide curves and xtra profiles, and just loft between your initial ellipse and the point, and use the end-tangency magnitude to dictate how bulbous the loft is. I repeat... give it a try. It will take ten seconds of your time, and will get you closer, quicker, than any guide curve approach. If you use 'normal to profile' as your start tangency condition on the elliptical profile, you will really be in control.

-Ed

Reply to
Edward T Eaton

Ah, ok, after reading the others respond, Per and Ed's last response, i understand more what this want was about (I did not read the original right at all!)..

so,.. for anyone wanting a egg model using a ellipse mid profile and lofted to start and end points...

formatting link
..

VFR wrote:

Reply to
Paul Salvador

I've done many "nose cone" types of features. When it is not a plain revolve, I find it sometimes best to divide the cone into quarters (once down the symmetry plane, once perpendicular to that). Then loft one quarter, then the second quarter on the same side of the symmetry plane. Make sure you maintain curvature continuity between edges of the first two quarters, and maintain edges normal to symmetry plane.

Curvature continuity is implicit when morroring across a plane if the edges are normal to the symmetry plane.

Reply to
TheTick

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.