Hi Kevin (and other addicts).
Ha! The issue of conic curvature is the bit I was referring to when I said I would get back to "natural" curvature. The short answer is no. That is, no, I don't have the answer to how the perfect geometrical development will actually come out in a finished boat. I wrote a post in "boat design forums" at
that was aimed at stitch and tape plywood design, but the same principles apply. In my experience, with somewhere in the order of a dozen boats built in ply on frame and stitch and tape construction, plywood is closer to being a "perfect" sheet material than alloy, as it has a much higher Youngs modulus than alloy or steel ( effectively it stretches less). Among other things, this means that ply is less likely than alloy to conform to compound curvature.
By the way, please forgive me if I tend to over explain things that are probably obvious to you, as I am aware that others who are just embarking on the addictive path to metal boat building may be completely bamboozled by some of these concepts.
I always use a conic development process to predict how conically developable a hull shape may be but I have come to expect that the forward third of the hull will have more curve that I have predicted. The form of the curve is generally parabolic with the most curve toward the keel. Is this your experience?
I start with a set of lines that I think will be developable (the unused part of the brain gets very good at this sort of thing - I understand that the mathematics that the brain performs when you walk forward and reach up to catch a thrown ball is as complex as computing a ballistic trajectory for a moon mission) then I use the process described in the boatdesignforums post to predict where the keel profile line needs to be. If I can’t get the profile I want by moving the cone apexes I might have to alter the chine line in either profile or plan view but not very often.
My theory, and I have not got an engineering degree, is that when you make a valid conic projection, it is only one possible solution to the geometrical problem. It is not necessarily the only solution or the best one. If you use your conically developed hull shape to build a hull “plate on frame”, you can force the plates to conform to your solution, and because alloy especially is a more ductile material than mild steel ( the E value for mild steel is normally around three times that given for marine aluminium alloys), alloy will conform to a surprising degree of forced compound curvature.
As an aside, Dennis Ganley was a NZ designer of steel boats who seemed to really understand what could be done with steel plates. His designs were late on frame but he used very strong male jigs to restrain plates as they were forced to accept compound shape that got locked in as the hull went together. Sadly, he died as a relatively young man a few years back and we cannot consult him on this.
Anyway, for this reason I normally allow between 10 and 40 mm trim allowance on the bottom edge of bulkheads and frames forward of amidships. I also make an allowance for induced curvature in the side plating, as I am more fussy about pulling in hollows where they are obvious. The amount of curvature induced is obviously a function of the change of angle in the sections, the “twist” if you like, in the hull. If there is no change of angle (this would equate to a constant deadrise in bottom plates) then the hull is cylindrical rather than conical in form, and there should be no transverse curvature in the hull plates, as viewed in section.
As a matter of principle, if I am working with an apparently flat panel, I normally pre-stress the plates by artificially forcing some curvature into the panel by creating curved frames or bulkheads. I use a 1:40 rise/chord ratio for the span involved and this seems to be easily achievable. I find that if I don’t do this, the small devious gods of alloy boat plating create edge shrinkage and I get sections of plate between frames that are uncommitted or slack and prone to “oil canning”. Prestressing minimise this. I have had cracks form in fuel tanks as a result of having areas of slack plating and I am more careful now.
I am a reluctant convert to alloy as a boat building material. I love working with it. It is clean, light, easy to grind, amazingly fast to weld, has a high scrap value, it does not really need painting above the waterline and it allows you to build things that you can’t seal inside of. This last feature makes designs for steel and alloy potentially quite different. Alloy boats have a higher resale value and have the potential to be stronger or perform better due to the higher weight/strength characteristics of alloy.
As of a few months ago (global financial disasters aside) it was actually cheaper to build in alloy than steel here in Australia. 6mm steel plate was $AUD2950/tonne, vs $AUD6500/tonne for grade 5083 plate in 4mm and 5mm. As alloy weighs 2660 kg/m3 against steels 7850 kg/m3 and as alloy thickness is typically 30 to 50% more than equivalent steel thickness, the actual material costs start just about level, even before you start waving a sandblasting hose or epoxy paint around.
On the down side, alloy just does not like holding paint, it has become environmentally or politically incorrect to use effective antifouling paints, welding is less forgiving of incompetence, alloy loses a fair bit of strength in the heat affected zones around welds, it is not normally practical to build outdoors or even under an open roof as shielding gases get blown away, alloy is very susceptible to several forms of electrical corrosion and alloy boats can be noisy if you are not very careful about isolating mechanical noises.
The clincher for me is that it is easy in Western Australia to order plates up to 10 mtrs x 2.2 mtrs from the supplier, who will plasma or router cut to an NC cutting plan. The router cuts plates are marvellous, apart from the razor sharp edges. I tacked up my 22 foot centre console runabout in a few hours, including breaking out the pieces from the plates. The pieces are pre-marked with an ink marker with frames, waterlines or anything else I want marked. I just email the supplier, confirm the availability of the plate sizes I want, email through a cutting plan in dxf format and post a cheque.
A lazy mans boat building at its best.
Fremantle Western Australia