Playwood shell hull design

Try to find some method or software for design hull surface for playwood shell. like i understand such hull must consist of conical surfaces only. Long time ago i saw book with such method, but can't find it animore. Can someone help me?

 

Actually you don't need to use conical surfaces only.
Look thru the Gougeon book on Boat Construction. Its a free download - 5th edition.
http://www.westsystem.com/ss/assets/HowTo-Publications/GougeonBook 061205.pdf
You can specifically look at plywood (conical) structure or you can look at tortured ply.
In the past there have been kayaks and A and B class catamaran sailboats built in the same method.
There was a recent thread talking about tortured plywood ( or compound curvature).

Are you actually trying to make a rowing shell (racing type boat)?

 

Have a look on this

 

Friends of mine going to make boat quite similar as on TANSL example. He ask me to make production lines fairing. Idea is to avoid any unecessary deformation. In general shell plate should be bended by own weight and fit with frames sections without gaps and extra bending.

 

For metal plates and plywood to bend easily the surface shape should be "developable". Conical surfaces are one type of developable surface. Successful design of developable surfaces requires an understanding of what makes a surface developable, not just software. For example frames for developable surfaces are usually not straight. An article on developable surfaces by Chris Barry, who occasionally contributes here, is available at http://www.boatdesign.net/articles/developable-surfaces/

Rhino3D has a free plug-in, DevSrf, for designing developable surfaces http://wiki.mcneel.com/labs/devsrf Rhino3D has a command Loft which has a developable surface option but the results are not as reliable as those from DevSrf.

 

Without wishing to disturb anyone I would like to comment as follows, because this is an issue that I think is very interesting:

What are the other types?

In my opinion the shape of the frames has nothing to do with whether or not the shell is developable. A parabolic shaped frame may be in a developable surface and a straight frame may be in a no developable surface. There are not developable ruled surfaces.

When a surface is developable, its development must be accurate. It is unacceptable that Rhino or other application development so imprecise a developable surface. I do not think this is so. You may be talking about non developable surfaces, for which they use certain algorithms that produce logically less accurate development (sufficient accuracy though), since they are not developable.

 

Besides a conical surface, a cylindrical surface is developed as well. There are combinations of flat plane and these too.

Frames are shaped to the surface, not the other way around, so yes, frames employed in a developed surface hull form, are typically not straight, except at the intersections of some edges.

The fames have nothing to do with the surface development, but simply follow what the surface model requires.

Lots of software packages have difficulty with developed surfaces, some are better than others. Often it's simply the edge pressure settings, but everyone I know that's designed a CNC file for a developed set of panels, has had to go back and refine those edges, to get accurate fits, after the panels were arranged on a jig (also CNC cut). With these "refinements", the cut file can be trusted, but clearly some software is better than others in this regard.

 

"Conical surfaces" are developable surfaces as previously noted, and a general conical surface can be created by sweeping a line along an arbitrary curve while holding one end of the line at a fixed point, the vertex of the surface.

"Cylindrical surfaces" are also developable surfaces as noted by PAR. A general cylindrical surface can be created by sweeping a line along an arbitrary curve while keeping the line pointing in the same direction. That means the line stays parallel to its initial orientation as it is swept along the curve. A cylindrical surface can also be considered to be a conical surface with the vertex an infinite distance away from the sweeping curve.

Another type of surface which is developable are "tangent surfaces" which can be created by sweeping a line along an arbitrary curve while keeping the line tangent to the curve.

Planar surfaces are trivial examples of a developable surface, and can be considered as a special case of conical, cylindrical surfaces where the sweeping curve is a straight line, and of tangent surfaces where the sweeping curve is planar.

There are also "rolling surfaces" such as Oloids https://en.wikipedia.org/wiki/Oloid and Sphericons https://en.wikipedia.org/wiki/Sphericon which are developable.

All developable surfaces are ruled surfaces. A ruled surface is a surface which can be described as set of straight lines. However, not all ruled surfaces are developable surfaces. A helix is an example of a ruled surface which is not developable. See Chris Barry's article (link in previous post) for another simple example of a ruled surface which is not developable.

A developable surface can be created by combining portions of simple developable surfaces. A junction between two simple surfaces will be a straight line and a ruling line for each of the two simple surfaces. The two simple surfaces can be tangent to each other along the entire length of the junction, but except for a few special cases curvature continuity is only possible at one location along the junction.

 

TANSL, I'm not sure what you are trying to say. My comment about frames for developable surface not usually being straight was made because frequently individuals who are not familiar with the properties of developable surfaces believe that a design with straight frames will be developable. Usually a decision has to be made of whether to have straight frames or a developable surface.

 

Can anyone point me to a chart that shows plywood minimum radius of curvature vs thickness? I am referring to normal dry bending no tortured ply.

 

http://www.ewp.asn.au/library/downloads/ewpaa_plywood_bending.pdf

 

CDokey, maybe I do not understand you well, but I think the decision need not to be "whether to Have a straight frames or developable surface", but it is possible to have straight frames and developable surface.

milo12, attached a file that can help you, hope.

 

Straight frames are only possible with a true developable surface if the frame edges are parallel with each other; ie constant deadrise/flare. Otherwise the frames for a true developable surface will not be straight. (This is for a conventional arrangement of frames in parallel planes, and no kinks in the surface.) That's from the basic properties of developable surfaces.

Straight frames may be possible with a surface which is "close enough" but not quite developable depending on the characteristics of the shape and construction methods and tolerances.

 

It is rather a nuisance that a straight-framed planing boat makes a better running hull than a developable boat, generally speaking. Or at least in the most commonly used configurations.

 

Dear TANSL in general frames can't be straight. Well you can make straight frames, but you will not be able to put on it shell correctly. Frames in this case is like intersection of cone by general plane. Result of intersection can be straight line only if this plane goes across cone's axis.

I think this more realistic example for developed surfaces and frames are not straight.