New to cat design - so many questions...

Hi!

I have a concept, some layout work and a little theory but I need to know an awful lot more and want to identify some APPROPRIATE resources (i.e good advice over bad) so would appreciate any advice on books to read besides "The cruising multihull" that I have already read.

My design is an electric cruiser with solar, wind and diesel generator back up / range extender.

The intention is to get the basic concept and layout drawings done with a fairly robust specification and design statement which I can then pass on to a professional designer, so any recommendations for designers who specialise in cruising cats (specifically electric if possible) would be gratefully received.

I have no experience in boat building and most of my composite experience has been in glider design and repair. I would like to build using foam sandwich construction, but again would very much appreciate any recommendations for study resources.

I also have a question (more curiosity really)...
As I have been looking at multihull hulls, I realised that many of the profile stations can be represented as quadratic equations and wondered if this is how hull designs are designed / represented / analysed by the professionals?
If so, how are the transitions from station to station managed / handled (assuming a non linear transition) and if this is a process that professionals use, can anyone recommend a book or other resources on the math of hull design?
I don't intend designing the wet portion of the hulls myself (would prefer to leave that to the experts) but I am curious and would like to understand a bit more about this subject. I understand the principles of computational fluid dynamics and can perform some CFD analysis on solid objects exported from 3ds max and processed in matlab....but there seems to be a gap between my amateur iterative process of create, analyse, modify, rinse and repeat and how the pros must do it (do they know that a given quadratic equation will produce a hull shape suited to a given task? do they already know that a certain transition profile from station to station will suit a particular type of cat or yield a certain type of behaviour?)
I need to understand the math...

Thank you for any / all help in advance

Yours

EK

 

EK,
I don't know if the professionals use anything like this, but ever since I came across this paper by John Letcher "A New Approach to Numerical Fairing and Lofting" (http://www.sname.org/HigherLogic/Sy...tFileKey=99e5b2af-ce60-4961-970a-882aa88705d1) I thought it was an interesting idea. He uses just 6 curves to describe the entire hull - sheer, profile, freeboard, two station curves and a sigmoidal function for a smooth transition between the stations as you move along the length of the boat. He published it back in 1972 and the computing power wasn't like it is now which may be why it didn't fully catch on. I haven't found any later papers that referenced it, but the mathematics are relatively straight-forward.

I've just started working on a Matlab implementation just to see if I can generate a set of lines drawings that might be importable into FreeShip. It sounds like we've been thinking very much along the same lines - generate a basic concept of a solar/electric catamaran and then pass it on to a professional for a real design.

 

Most boats nowadays are drawn using software. I suspect that is why the curves are representations of relatively simple formulas. Drafting with battens, the curve is whatever the designer decides, which could be a hugely complicated equation. Software approximates equations to those curves.

 

Today a curve is not represented by a single equation, it would be an outrage. A curve is approximated by many polynomials between first order and fourth order, which are very simple equations. The curves generated by software can be exactly the same as those generated by hand with battens.

 

I'm working in 3ds max and experimenting with the lofting function. This is the first time I've used it for a boat though I can see right away that scripting or even a custom addon could be produced to take some criteria and produce an entire hull, which could then be exported to matlab or similar for processing. However, it is also obvious that this approach would suffer quite significantly from the "garbage in, world war 3 out" problem!
It's sort of moot though anyway as I only need an approximation for the purposes of layout; I wouldn't dream of trying to build something based on my own hull design...I'll leave it to the expert lol

Glad to touch bases with someone else who likes the idea of solar electric though! I really like the idea of having a power cat without using diesel

 

Wind power is quite different from solar or diesel power because sails create large heeling and pitching moments, while the other sources of power have thrust lines that pass near the center of gravity. As a result, you will need to let the wind power dictate the form of your hull. So the best approach would be to treat this like a sailboat with solar & diesel auxiliary power, albeit with an unusually large auxiliary power plant.

I think a sailing catamaran would be the best starting point for your design.

 

Quadratic relationships are used for simplicity in theoretical analyses because they have a limited number of parameters to optimize. Practical designs usually have more complex hull forms. Low order curves also have the advantage that they are inherently fair - free of bumps and hollows - and their modest changes in curvature make it easier to build the hull.

It is possible to develop a fairly sophisticated hull form using simple curves to control the variation in the shape parameters. Here's an example of how to do it.

 

Hull Form Math

Today most hulls are designed by using some variant of a B-Spline for the stations, and then the "transitions from station to station" are handled by more B-splines.

The details are of course more subtle than the above one-sentence description.

Letcher describes the math quite thoroughly in "The Principles of Naval Architecture - Volume 1: Geometry" published by SNAME.

 

Yes, that is the current procedure. Splines are third-degree polynomials that allow inflexion points. But splines have the problem that the representation of straight lines and circles with them is difficult. Therefore, the curves used in shipbuilding to represent, for example frames, are a mixture of polynomials of varying degrees (straight lines, arcs and splines). The frame line finally becomes what is called a "polyline". In addition, numerical control machines does not understand about splines so you have to turn everything into straight line segments and circular arcs.

 

Tspeer,
by "wind power" I meant wind turbine. As in generating electricity from a pair of wind turbines. Not sails.

 

Oh wow! thank you! Much appreciated. Sorry to impose, but could you recommend any boat design books? preferably specific to multihull but any authoritative text would again be very much appreciated. I'm itching to get a better understanding of the math behind hull design

 

Outstanding! I'll source that book and begin digesting...such a fascinating subject!
Is the b-spline equivalent to a nurbs spline? I'm trying to cross the naval architecture to 3d modelling divide at the moment. A lot of the terms used in things I'm finding in internet searches are quite different to what I'm used to..

 

Tansl,
presumably CNC machines don't have a problem with geometry though (i.e. splines that join together to form meshes)?

 

If you have access to Matlab there's B-spline GUI on the FEX: http://www.mathworks.com/matlabcentral/fileexchange/27374-b-splines that shows how they work. It's interactive, so you can put down the basic curve and then fine tune it by moving the control points around.

If you have a more recent version, it won't work as is but I found that you can fix it by commenting out the first line in gui_bind_event.m. With something like this and using John Letcher's paper, it should be possible to quickly generate a basic outline of the hull.

 

It doesn't matter if they are sails or wind turbines, they will still generate a heeling moment as a non-linear function of the power they create, so unless you propose to have very small turbines just for a few hundred watts, you'll need to consider the overturning moment they can generate.