Ogive Curve Help Needed

[Vermin_King]

One thing I figured out on the minions that I did for my cousin is that I would use the truncated cones like Dave did as formers. Glue that inside the body and lay the petals over to create a nice shape and nice looking top. To give credit, I used the head pieces from the S&P minions inside the heads of the Paper Replika ones

[airdave]

What I don't like about petal assemblies is you cannot control the overall shape of the dome itself.
Or at least, it is much more difficult, than using ring/conical sections.

If the object is a half sphere dome, then you know what you are trying to achieve with the petals.
But egg shapes (ovoids) become a lot more difficult without some sort of form.

If you use the cone assembly as your inner form (reduced in scale of course)
it might be possible to achieve the correct ovoid shape with the petals.

I noticed, on the actual locomotive, that there is at least one ring/seam
in the bullet nose. So, one of the paper seams would be visually accurate.

[Diderick A. den Bakker]

Airdave: of course it's all very theoretical. But what one could do is make a negative template based on my suggestions above; then constuct / glue the dome to make quite sure you get the shape you want. Earlier, I have described a way to fill the dome with a mixture of glue and tissue paper. I followed this method with the Tintin moon ship (attached). If you are really ambitious about this, that will make the dome strong and hard enough to sandpaper the outside - then apply a thin layer of paint.
Let me know if you (or anyone else) would appreciate another look at this tissue/glue technique, and I will try to find it!

[sgoti]

Thank you one and all for your suggestions, contributions, ideas, perspectives, and most importantly, willingness to help!

I'll try to report back in with results when I have some to report (work changed, again, and free time got reshuffled).

Thanks again!

[aansorge]

a good first cut at petal shapes can be had from a spreadsheet. using pi times diameter equals circumference, and the graphing tools, one can see the shape and how the petals change (select points to graph along the length). the graph can be copied right into the model, and the number of petals doesn't matter, at that point all is relative.

In this case, with the beautiful non symmetrical shape, make two shapes, one for the top and one for the bottom. They will blend together, hopefully satisfactorily by the third attempt!

I haven't figured out the base shape, try an existing nosecone for that. Or make use of the higher math than I managed to achieve.

Goodness! I hope this works.

[zubie]

Wow, I really wouldn't want to try it without CAD tools. Ages ago I was working on an Aries 1b (it got shelved waiting for better refs, and learning how to use Blender, cause I was really getting tired of all the spreadsheets and trig!). At any rate the thing I wanted was a good, admittedly spherical (as opposed to ogival) shape.

I'd say, if you are working this thing by hand, don't necessarily give up on the petals. I'll tell you why. The argument of the structure, well consider making up two petal sets that end at the tip in a cone section (the headlamp). The first set of petals will be fractionally smaller and act as the underpinning of the second set which will be the skin of the outside. The inner petals should be cut into a slightly feathered shape to allow the structure to better follow the compound curve you are going after. The outer "skin" petals should be aligned with the midline of the underlying "tab" petal. Also, while the math will give you a strict set of points to join, draw them with a french curve (or whatever tool you use if are using software to draw these). The smooth curve will promote a better shape along the length from tip to base - which is where the petals may beat the rings.

In my specific case the petals didn't go quite to the nose as that was done as a cone section (for a hemisphere that's nearly flat). Below are the urls of the images of the preliminary prototype where I tested the assembly process (yes it looks kind of crappy, but I was just testing this idea for assembly).


http://i721.photobucket.com/albums/w...FirstLayer.jpg The long line was for aligning the skin overlay, the dotted lines where the cut points for the feathering. The whole thing was set on formers at this stage. The arrows were a reminder for assembly order.


http://i721.photobucket.com/albums/w...inLowAngle.jpg Here with half the skin placed on top.


http://i721.photobucket.com/albums/w...b/PB224013.jpg The second prototype, note this was done without using the formers; instead IIRC I just went slowly around the circle, align under tab petal, set skin on top, attach next tab petal, then skin petal, until I went all the way around. It was then overlayed with some seam hiding layers of plain paper as well as an overlay of thicker stock parts to match the original detailing.

[zubie]

Got a bit obsessed on this problem. Also thought it would be good practice to try out boolean operations in Blender. Anyway, I came up with this alternate part pattern that is a combination of petal and ring given the patterns illustrated in the photographs of the loco.
Basically there appears to be 2(?) ring patterns, but also there are the 2 horizontal bands. So this pattern follows the two horizontal bands and are rounded by 3 ring sections (circular hole is for the headlamp). Sorry, but I don't quite know what the exact bullet arc should be so this is a very rough guess. The attached pdf is the larger printer size. Of course more rings will improve the curvature of the upper and lower parts at the price of more seams.

I still have the blender file if this will help.