The Design Process and Tools

From http://www.boatdesign.net/forums/boat-design/cad-vs-paper-31267-6.html#post345956:

I think it would pass a lot of interesting ideas around. It was also suggested that it would make a good stand-alone thread so here goes:

As I mentioned before, a pencil, scratchpad and hand-calculator are always within reach and are the first step in any design process of mine. How soon I get it into the computer depends on how novel the concept is, but as a minimum I like to record any interesting idea no matter how unlikely in case I figure out how to solve the problem later on. In my computer the boat folder is second only in size to the pictures folder and has more than 70 sub-folders.

My hull designs feature very simple construction, suitable for a first build; it is a particular passion for me. For example, as an exercise a year back I wanted to get an attractive 5-plank boat out of a single sheet of ply using straight cuts for the sheer plank edges and a minimum of forms.

The sheers would curve around a single form with constant flare angle and this shape would define both the sheerline and the shape of the flat bottom. The bent sheer shape could be used as a pattern for the bottom to eliminate lofting. This resulted in the bilge planks having a constant flare angle, allowing the bevels on the sheers and bottom edges where the bilge planks were glued on to be precut with a router.

It sounded like an interesting concept. First I scribbled some sketches until I had the principle worked out. It turned out I needed to work out some math, for which I created a spreadsheet to calculate the shape of a batten bent around one or two forms. I had the spreadsheet create a graph which I copied to a graphics application, where I created the orthogonal views. It was looking good so I input the major parameters into Free!Ship to check out the perspective appearance.

The hull design looked good in perspective so I had Free!Ship develop the planks. It needed a minor adjustment to get the sheer plank developments exactly straight, I then used Free!Ship to pack the planks as efficiently as I could and copied them back into the graphics application to check if they would still fit into a single 4' x 8' ply sheet. This could be done with a minute adjustment.

The last step in the design process was to cut out a card profile, including paddler, and balance it about the boat's projected CoG. This revealed that the boat would to head into the wind like a weathercock, so I had to raise the bow to counteract this tendency: this required a curve to be introduced into the forward end of the sheer development in order to achieve the other objectives, but I was able to keep 90% of the sheer cutting straight.

I built the boat and it went together quite well, although I found there were still things to learn as there always are during a build. The hull still needs finishing as a sailboat build got in the way and I had some novel sailing kayak concepts I needed to try out, but I will get back to it soon. It has the look of a nice handling boat, fast, light (15 lb unfinished) and agile.

The concept was successful enough that I adapted the sailboat design, which I got from PAR, to the same principle and it worked out very well.

 

Interesting Terry
i think a lot of top designers use batten and weights to draw the most important of lines, the sheer , once they have it right, by eyeballing along the line, they then stick it onto the Puter One sees so many sheers on so many CAD boats that are just not right. usually I like to prove that line on the loft full scale

 

Working out a sheer line in a 2D drawing is tricky. What looks good in 2D drawings might have a bit of reverse curve from some angle.

There's a theorem which says that unless a curve is planar (the curve lies entirely in one plane) then in some projection the curve will have reverse curvature.

 

That is one of the reasons I use Free!Ship to check out the perspective appearance, it allows me to look at the hull from different angles.

That's an interesting theory and new to me, although easy to understand now I am aware of it. When a sheerplank has a constant flare angle and the development has a straight top edge the hull's sheerline will be planar, per the theory.

for the design example I used I had to curve the forward part of the sheerline up in order to achieve balance in a crosswind. This gave it a reverse curve but only when viewed from close-up, below the sheerline. This is easy enough to visualize. Most small vessels are not seen from this viewpoint when afloat, which explains why I did not notice it before. I can see it when the boat's hanging on the garage wall.

For a hull with varying flare angle it is harder to visualize how the sheerline would appear from any angle; here Free!Ship and other hull design programs are a great help.

 

A boat I have planned for several years should start fairly soon. The idea began when I read about the Wee Lassie single-handed canoe built in some quantity around 1896 by J. Henry Rushton. Much of the technical detail is documented in other threads so I will try to just describe my thought process and the evolution of the project to date rather than the nuts and bolts of how.

I approached it along 2 separate paths, the design for the lines, and the construction methodology. Although there were 2 separate types of activity they were pursued more or less in parallel as an encounter in one would often lead to an idea on the other activity.

I liked the history and few pictures of the original boat which is why I bought the plans from the Algonquin Museum. The lines were drawn for a lapstrake boat and I used a computer spreadsheet to adapt them step by step, which lead to new lines for a non-lapped or butted seam boat. I put the lines into FreeShip, checked out the center of buoyancy, cross-curves and displacement vs draft, then reluctantly had to stretch it to accommodate my weight. I also made a few improvements (IMHO) to the sheer - higher at the bow to reduce the weathercocking predicted by my card cutout - and the rocker - increased for more agility in the narrow waters I usually haunt.

The original construction was rich in almost vanished woodcrafting skills which I do not possess, and it did not lend itself to my usual methods developed for 3- and 5-plank boats. With 12 planks (6 per side) it would have been a beast to attempt in stitch and glue, so I conceived a method which I called the Zipper Seam or ZS. I now understand the name has been used for another system entirely. The method itself is a blend of a technique I used in model aircraft back in the 1950's for joining small sheets together, with a system called the metallic batten wide-board canoe which I read about in a book (The Canoe by John Jennings). This puts the roots of ZS 60 to 110 years ago: nothing like recycling history, I says.

ZS allows the hull planks to be butted along the seams without using either stitches or glass-tape reinforced resin fillets, aiming at a wood-only inside finish, my nod towards tradition if you like. I am making some mockups and small scale test subjects to verify the method both as a process and to perform destructive strength testing before undertaking an actual boat, as there is no precedent to draw upon for experience.

The biggest and most time-consuming task has been to find a way around the Rushton method of keel building, hand-carved from solid oak, rolling bevels for the garboards, presumably made using factory jigs and gauges not available to me. Oak is not necessary since I will not use canoe tacks to hold the boat together, as Rushton's men did so long ago. I now have a system worked out for building up the keel from 4 softwood pieces to avoid the work of carving and to ensure the close fitting joints which glued assembly requires. Despite all the curves and rolling bevels found on such a keel all the cuts are straight ones: neat trick, but I have tested about 75% of it. During this I had to reverse an earlier modification to change the construction from 6 planks/side to 5, by merging the garboards and bilge planks, as the resulting plank generated excess stress and distortion when formed into the boat's shape.

More recently I have addressed the task of ensuring the boat goes together correctly, getting the inner stems installed, and finally how to hold the boat together using friction while the seams are glued.

Given a successful conclusion to the strength testing, I will then proceed to a 2/3 scale, 1/4 boat model using cheap door skin plywood to ensure the entire process can be completed, before I undertake a full scale build in marine ply, cedar, pine, mahogany and cherry (if I can get it).

The oldest computer records for this endeavour are more than 3 years old, and it has been going on for longer than that. This is an unusually long gestation time for one of my projects, which are normally conducted in a tearing rush! I guess I am slowing down in my old age.

One aspect I should not neglect is the amount of help I received from the members of this forum: much of it was invaluable, some of the advice I did not take - being me - but thoughts were generated, problems were identified and - I hope - resolved. We will see ...