Boat Design Geekery II (specialized)

[catsittingstill]

I'm going to talk a bit about my experiences with BearBoat here.

BearBoat:

 

Okay, for starters, this program was put together by a kayaker, for designing kayaks. This means its default idea for what a boat should look like is pretty kayaky, which looks a little odd when you're working on a canoe. You can change things around to make your boat canoe-like (and there's an option for leaving the deck off so you can see your gunwales) but it starts out with an eye-popping amount of rocker, and more than usually asymmetric, and that kind of thing.

 

You start to design a boat by putting in a few sketchy details about what you have in mind (length, beam, displacement (weight of boat plus contents), and so on.) None of this is set in stone; you can always go back and change it later. The program pops you up a boat. You change the design by cycling between four views.

 

The curve of areas. This view is very abstract—it's just a plot of the cross sectional area of the underwater part of the boat, versus its position on the long axis. It looks like a bell curve. There are four numbers you can change that change the shape of the bell curve somewhat. You can think of it kind of like floatation—you can squeeze the flotation into the middle of the boat, or you can squeeze it more out to the ends. This affects things like how blunt and how deep the ends will be, which affects speed, maneuverability, tracking, and how well the ends will tend to rise to the waves or slice through them.

 

The birdview. This is the view looking down on the boat, both of the shape of the gunwales (called the “seam” because in a kayak it would be) and the shape of the waterline. There are four more numbers you can change that change the shape of the canoe—forcing the ends to be fatter or thinner. Since the cross sectional area of the ends has been set by the curve of areas, when you force the ends thinner you force more of the flotation farther down (perhaps by bowing out the sides of the kayak) and when you force the ends wider you force the flotation (just volume inside the boat, in this case) higher up—the cross sections of the ends can get positively weird if you don't keep an eye on what you are doing.

 

The side view. This looks at your canoe from the side so you can see the shape of the keel (how much rocker you have) and the shape of the sheer (whether the gunwales curve up at the ends and how much). Again, each is shaped by four numbers that control the curve. Forcing the keel to have almost no rocker can be done, but keep an eye on the cross section or you'll get a knife like keel that would be very impractical in real life.

 

The cross section. This is actually a bit of a misnomer; you get a bunch of cross sections all the way through the boat. This is where you look for weirdnesses you introduced while changing parameters in the other three views. This is where you reduce the number of cross sections to something reasonable (one every 30 centimeters is about right) and check to see if this boat resembles anything you ever built or anything you've seen on the internet because this end-on view is the one the sellers-of-plans show you.

 

Here is an example:

This shows contour lines every five cm. The portion from the bow to the middle is shown on the left; the portion from the middle to the stern is shown on the right.  The yellow line is where the cursor happens to be positioned on the side view.  The bow slopes down more gradually than the stern, the keel angle is constrained to be about 81 degrees or less (so it can't go all the way flat on the bottom) but there's kind of a lump where the hull emerges from the water that I'm not all the way happy with yet.  But this is very much a work in progress.

Comments

[randwolf.livejournal.com]

Hmmm. I count 12 shape parameters, plus the handful that you use to get started. Is that enough, I wonder?

If you export a DXF file and send it to me, I could run it through Rhino, & perhaps produce some perspective views of the boat, at least if Rhino can read Bearboat's idea of DXF.

[catsittingstill.livejournal.com]

Oh, hmm. I hadn't thought of that. Ha! I know someone who knows Rhino already! Oh, cool!

This was a bit of an overview--you can also set the center of buoyancy, the center of greatest cross sectional area (why those two even *can* be separate parameters I don't understand), the enter (think of it as the overhang of the bow) the leave (ditto for stern), max beam at gunwale, max beam at waterline, max angle at keel, max angle at waterline, there are provisions for changing the sheer and the rocker beyond what the four parameters for each allow...

Actually, if you want to play with it, it's here: BearBoatSP (http://web.mac.com/rlivingston/BearboatSP/BearboatSP.html)

I will see about exporting a DXF file and, if you have the time and interest, you can let me know 1) if Rhino can read it and 2) if Rhino can generate a table of offsets from it.

[boywizard.livejournal.com]

I've been planning to experiment with BearBoat also, but the holiday has interfered with my playtime, and I still need to finish my description of physical modeling. Looks like a potentially useful tool, though.

[catsittingstill.livejournal.com]

I'd love to hear more about the physical modeling, but there's no particular deadline on that if you want to try out BearBoat first.

[boywizard.livejournal.com]

This is an approach to physical modeling that I hadn't considered, but I bet it would work pretty well, and would eliminate the need to slice up the completed model, since it would be essentially pre-sliced. We might call it the 'wedding cake approach', since the model is made from layers. If I were making a half-model, I would make each section a couple of inches wider than the actual half-section. Let's say that the max beam station would fit within a 4x4 rectangle (representing a 1/4 scale 32 inch beam, 16 inch deep section), and that the stations on the building form are a foot apart. I would cut a block measuring 6 by 4 by 3 inches, with the grain running in the 3 inch direction (i.e., it'll be fore and aft on the model). Each subsequent block can decrease in width as the model tapers toward the bow and stern, but retain the extra two inches of width. I would then drill a pair of half-inch holes in each block in what we will call the waste portion (that extra two inches). A simple drilling jig will guarantee that all the holes will line up. We then slide a piece of 3/8 all-thread through each hole, with washers and nuts on each end. Tighten very firmly, and we have a block we can then shape to our preference using spokeshaves, rasps, drawknives, sanders, or whatever technique we want. When the model is finished, we simply unbolt all the blocks, giving us the shapes of all the stations which can readily be transferred to graph paper for upscaling.<p>

If we already know the shape of the hull at each station, but we want a model  just to see if it 'looks right', we would lay out the shapes from offsets on graph paper and then glue it to the blocks. We could then cut to the lines with a band saw or jigsaw, assemble into our master block, and finish as before, but with a lot less waste to cut away. Check the link for a sketch of what this technique would look like.<p>

And, happy birthday, cat!

<img src="http://i1089.photobucket.com/albums/i351/glasslightdesign/boatcake2.jpg >

[catsittingstill.livejournal.com]

Thanks for the birthday wishes! I have been having a wonderful birthday :-)

The nice thing about the pre-sliced approach is after transferring the outlines to graph paper, you can put the model back together if you like and save it.

What I'm wondering is, if we know the shapes at the stations, could we set up a 1/4 scale strongback and glue 1/16 inch strips together over it? But we wouldn't be able to sand enough to get that smooth, really. Hmm.

Actually I may end up using your plan or something like it if I can't figure out another way to get offsets for the stems. Bearboat will print out offsets for the stations, but apparently the programmer normally sets up forms 4 inches apart and doesn't bother with stem forms. I can't afford that much plywood (I normally set forms 1 foot apart) and I can't picture polyisocyano-uranate foam being strong enough to hold the strips.

I got the bearboat design moved into Rhino but Rhino does so many things I'm having a hard time figuring out if one of them will turn out to be the thing I want.