Building Allie Kat
A Somes Sound 12 1/2
First Impulse
I resisted the first impulse to build the jig. It seems like a logical place to start, but Brooks recommends constructing as many of the boat parts as is possible before building the jig. This makes a lot of sense. It takes room to edge glue the four boards that will make up the transom, and ripping then scarfing the plywood planks will require even more real estate. Laminating strips of wood to make the inner and outer stem is another task that I would not want to do in tight quarters.
Making Templates
The plans include full size paper patterns for the stem and transom, and for the web frames and molds. A builder can definitely use the patterns to get accurate shapes cut and installed on the jig. I elected to go with the extra step of making thin-plywood patterns. This gave me a chance to look carefully at the plans and the shapes and start working out in my head how it will all go together. Additionally, an error cutting out a pattern would be considerably less expensive than an error cutting a fine piece of mahogany.
Shapes from the full size patterns were transferred to the sheet plywood by laying the plan page over the plywood and punching holes through with a sharp scratch awl. I marked points every inch or so on small radius curves and spread the marks further apart on larger curves. I also marked points for transferring waterlines and floors, and where the planks will land.
After connecting the dots I rough cut the shape out of the plywood sheet with an oscillating saw with a really sharp Japanese saw style blade. Next I cut to about a 16th of an inch on the band saw and finally sanded a fair curve on a stationary bench sander. It's possible to fair the curve with a block plane (I tried) but it's hard work. The sander is faster and gives a nice 90 degree edge as a bonus.
I had a few nicks where I got the band saw too close to the line. Errors like that are easily repaired with thickened 5-minute epoxy and then resanding to a fair line.
Constructing Molds
Molds are made of 1 x 10 #2 white pine. Each side has an upper and a lower connected at a miter joint with a gusset glued and screwed on top. The sides are mirror images and are connected at the center with a third gusset.
It's embarrassing the amount of lumber I ruined teaching myself how to make this simple shape.
First, I tried using the template and a flush trim router bit. The curve of the mold arcs across the grain of the board and back again. The router bit split the board numerous times near the side and at the ends of the board. After the second failure in four tries I gave up.
Second, I tried the process Brooks describes in his book. Each mold half (left and right) has an upper and a lower that are joined together with a gusset screwed and glued over the miter joint. Once the upper and lower parts are joined, the joined halves are laid one on the other and screwed together. Then the paired-up sets are run through the band saw. This was too much wood for my small band saw, plus the gusset on the bottom half ran into the saw table forcing me to lift everything up and free hand the wood for a short distance which seemed like a dangerous move.
Third, I tried cutting out each side individually on the band saw and that worked fine but the narrow blade on my saw made it difficult for me to cut with enough control to confidently cut right to the line. So I cut to about an eighth of an inch and finished smoothing to the line on the bench sander.
Finally, I used a jig saw. I have a decent jig saw; it's one of the best portable power tools I own. Since the best I could do on the band saw was leave an eighth of an inch I decided to give the jig saw a try. I installed an aggressive blade and it cut twice as fast as the band saw. The blade will tuck under a little, especially on tight turns, but the cut is only little off from plumb. The blade never wandered further than the eighth inch that I cut away from the line. After the cut I finished on the stationary belt sander as before. This final procedure was ultimately faster and most important it was less stressful.
I finished each mold by marking the DWL, the shear and the plank landing points closest to the shear.
Procedure:
1) Bisect the arc with a line on the template heading toward the center point of the curve. Then position the boards so that the template fits.
2) Transfer the bisecting line to the boards.
3) Screw the boards together
4) Use the table saw to make the cut
5 & 6) Dry fit the boards together and transfer the mold shape to the boards
7) Use the jig saw to cut close to the line.
8) After both upper and lower parts are rough cut, permanently join them with a gusset block fastened with glue and screws.
9) Use the bench sander to smooth the curve to the line. 80 grit cuts through the soft pine very quickly and the machine leaves a nice 90 degree edge.
10) Lay out both halves together and check positioning with the story stick
Screw and glue a gusset block at the center.
Story Stick
"Measure twice, cut once, but don't measure at all if you can avoid it."
- Norm Abrams
Offsets included in the plans are given to the outside of planking which is not helpful for checking the spread of the legs of the mold. From the full size plans I picked up the distance at each station from centerline to the shear and recorded it on a story stick. Then I used the stick to double check the width of the mold before fastening the two halves together.
Constructing Web Frames
There are three "web frames" in the boat. They are constructed of plywood and stay in the boat. I went back to the template and router method to cut the shapes, and failed again. On my first attempt I nailed the template to two rough cut web frames and cut both sides with one pass of the router. All went well with the cut – in fact the shape was as perfect as I can imagine is possible. Then I tried to pull the nails out. I'm using Sapele mahogany and it is really dense and hard stuff. I almost couldn't get the nails out. I won't be putting anymore nails in my boat. Once the nails were out I had station 6 complete. On the next web frame I used clamps to hold the template and two rough cut boards together. That went well until I got to one end and had to move a clamp. Even though everything was firmly re-clamped together, the boards and the template somehow slipped while the router was cutting – the pieces were ruined. On the next attempt I used screws with pilot holes drilled through the plywood. This method worked but I tipped the router in a couple of places causing some divots. This completed web frame was usable, but not perfect. Finally, for the last web frame I took the rough cut plywood to the stationary belt sander and smoothed the curves to the line, which was the same procedure that I eventually got to when constructing the molds.
Completed web frames and molds:
Scarfing Plywood
It was a lot easier than I thought.
The plans specify planks made of 3/8 inch marine grade plywood, and Brooks recommends using Mahogany. The boat is fifteen and a half feet long, so only one scarf is required per plank. In the plans there are instructions on how to efficiently cut the 9 pairs (port and starboard) of planks out of 6 sheets of plywood. The plywood is ripped into 24 inch wide panels and scarfed end to end.
I had never done anything like this and worried plenty about messing up my expensive plywood.
There are scores of YouTube videos available showing all kinds of scarfing methods/schemes – most involving hand power tools mounted in jigs. It goes fast (after the jig is built), but I didn't think the results were particularly impressive. And I worried about cutting under a jig where I wouldn't have a good view of what was happening to my wood. I decided to hand-plane the scarf.
One Practice Scarf: On the full size plans there's a scale drawing of the Coaming which wraps around the cockpit at the deck edge – it's about 13 feet long. The instructions call for making a pattern from the scale drawing so I ripped two 14 inch wide pieces out of a sheet of 1/4 inch spruce plywood and scarfed them end to end to have a piece of pattern stock long enough. What I learned from the practice is that you need a very strong/steady work bench to work on. I attached my bench to the wall so that it will not move no matter what direction I'm cutting with the plane.
I scarfed six sheets of planking stock at a time. I began by cutting with the hand plane from side to side keeping the gradually widening cuts even by keeping the exposed layers of the ply parallel. The plane is held at an angle to the cut so that the heel of the sole rides on the layer up one from the cut. This sets a constant angle. There's a scrap of of plywood at the top to set the angle for the top piece. When I was nearly finished I had some tearing of the bottom layers (the surface layer on the bottoms of the sheets) so I changed my direction of attack to down the face of the scarf to finish.
Gluing the joint is straightforward:
I did the gluing on a sixteen foot long table I originally built for the Kaholo SUP project last year. I only glue two pieces to make one long plank at a time - no stacking multiple layers.
I arranged the panels so that both factory edges lined up with a straight line I marked on the table. During the "dry-fit" I clamped one panel to the table and put a mark on the edge of the other panel, well away from the gluing area, and extended it onto the table. The mark is used to set the proper amount of overlap between the scarf surfaces. When the thickened epoxy is in the joint, the scarf is hidden and it is hard to see the line up. Before applying glue I put masking tape next to the joints.
First I spread un-thickened epoxy onto both mating surfaces using a foam brush. Second I spread thickened epoxy on the lower surface. Finally I flipped the upper surface board over and placed it into position. I clamped the joint using a 6-inch wide board and weights. The table and the board were protected with wax paper.
After the epoxy cured, the tape came up with some careful peeling and pulling. Most of the "squeeze-out" comes up with the tape. I tried using a paint scraper to get the remaining epoxy off, but it gouged the plywood. I had good results cutting/slicing the epoxy with a sharp wood chisel. Then I sanded the joint with 180 and 220 grit in a random orbital sander.
The Transom
Making the blank. The transom is made up of four 7/8" boards edge glued together. I used splines to strengthen the joint and help line up the boards for gluing. The edges were smoothed using the router and straight bit method described by Brooks in his book. Dados for the splines were done with a router too. I made splines the length of the joint and that was a mistake. It was hard to get the long wobbly stick covered in glue into the the slot without making a big mess. Next time I'll make short sticks and put them in side by side.
Cutting the shape. I used my templates to layout the inboard and outboard (Forward and Aft) shapes and then cut to near the inboard line, which is the larger side of the transom. The transom is beveled to receive the planks and the bevel makes the aft side smaller than the forward one. Cutting the bevels took a leap of faith for me. A lot of wood is removed while beveling. It's hard to visualize the shape to begin with and the bevels look too sharp. I did the cuts anyway, having faith in the plans. I started the bevel with a Japanese Pull Saw slicing cuts perpendicular to the edge. Then used a wood chisel to first knock the chips out and then smooth the cut. I finished with a Saw Rasp.
Laminating the Inner and Outer Stem
It was harder than I thought
The inner stem is made up of twelve 3/16 by 2-1/2 inch wide mahogany veneers. The veneers were cut out of a single 2-1/2 inch wide board.
Tension: Smaller pieces cut out of a stable timber can warp out of shape because of a property called "tension" that comes from naturally inconsistent growth and irregular grain in the wood. In my case about half of the veneers developed a crook – lying flat they warped side-to-side. In some cases the movement was more than a quarter inch over the length of the thin boards. This was a problem. The finished stem is 2-1/4 inches wide. The stock was only 2-1/2 inches, that gave me only 1/8th inch from each side to plane the laminated pieces to get two smooth (and parallel!) edges. The answer was to try to force the crook out while the pieces were in the clamping blocks. Next time I will start with wider veneers and straighten them on the table saw.
Gluing and getting the veneers into the clamping jig is easy with the help of a friend. I arranged the veneers with warps alternating up and down – that was a mistake. The four or five that didn't have any warp went on the inner side of the stack, which is the widest side after the stem is finally cut to the finished shape. I started with the stack of veneers held in place in the clamping blocks but not clamped tight. Then I clamped down hard on the top of the stack in two places near the middle of the bend and that got most of the warp out. Next I started tightening clamps working from one end to the other. At the ends I tried knocking the warped pieces into line with a board and hammer but holding them in the straight alignment was a problem because I didn't want to squeeze too tight with the clamps and then squeeze too much glue out from between the veneers. In the end I was able to straighten the crooks to less than 1/8th inch deviation which gave me just enough material to plane the uneven edges to two smooth and parallel sides 2-1/4 inches wide.
Spring back: I over-bent the veneers in the clamping blocks by 1/8th inch on each end – it was a lucky guess.
I don't have a surface planer so I used a powered hand planer to smooth the sides. To start I marked straight lines at the edges and the center with a marker pen. I used two clamping blocks to hold the veneer stack steady and set it on the flattest smooth surface I have - the kitchen counter. The pen was glued onto a sliding holder that sat on the counter and produced a level line. To cut the shape (the stem is widest in the middle) I used the router with a template bit and a plywood template. Then I went back and redrew the straight lines that got shaved off. Finally, there's a taper on the forward edge where the planks will land and I got it started with the router with a 45 degree bit. The taper is not 45 degrees so I left about half the taper to finish with the spokeshave. I've never used a spokeshave before and after about 30 minutes my stem looked like I had been hacking at it with a dull axe. I finally got the hang of it just as I was finishing. I left the taper about 1/8th inch from the line – to be finished when laying on the planks.
While I was at it, I glued up the outer stem too. This time I put the warps all in the same direction – bowing up in the middle. The downward clamp pressure straightened all the veneers and the shape came out nearly perfectly. When released from the clamping blocks, spring back was about double what I saw on the inner stem which makes sense since there are half as many veneers. The outer stem will have to wait weeks or maybe months before getting installed on the boat. To keep it from losing shape I bent it back to the desired curve and held it there with small battens tacked to the ends. I won't smooth or cut it to shape until it is on the boat, which won't be until after all of the planking is in place.
