For a cosmetic refit of the 32’ yawl ARROW’s cockpit, new 3⁄8”-thick teak strips were laid out and edge-glued with tenacious caulking to make panels that could easily be vacuum-bagged into position to emulate a traditional laid deck.
Original teak veneers had nearly worn through since the author built the boat in 1998.
ARROW, which I designed based on L. Francis Herreshoff’s ARAMINTA and built in 1998, has been in continuous use since then. Now in the hands of a new owner, she came back to my shop last fall for what you might call cosmetic or aesthetic work. Most of this was straightforward—repainting the deck and renewing its nonskid surface, stripping old varnish off the cabinsides and cockpit coamings to renew its luster with six new coats of varnish, and repainting the topsides. But it was also clear that the time had come to replace the teak cockpit seats and cockpit sole.
These were originally made of 1⁄4″-thick teak planks 4″ wide, edge-glued to each other and then glued down to a plywood substrate. Seventeen years of continuous use had worn the teak paper-thin in places, and the edges of the teak had chipped around the hatches. In other spots, the teak was wearing down close to the glueline.
Sanding would have rejuvenated the appearance, and the owner could have gotten maybe another season out of it, but sanding would have taken off even more thickness. Sooner or later, there would be a risk of wearing through to the plywood below. Replacement was the best course.
Originally, the teak veneers in ARROW’s cockpit didn’t have caulking seams. The veneers were simply epoxied together to appear as a single, wide piece. They were laid up over a meranti plywood substrate made up of two 3⁄8″ layers for the seats and one 1⁄2″ layer for the cockpit sole. These substrates, coated as they were with epoxy, were still in very good shape, so the new owner and I saw a chance to not only make repairs but to improve the aesthetics by using 3⁄8″-thick and 13⁄4″-wide teak strips with 1⁄4″-wide caulking seams filled with black caulking compound.
The new look would be first-rate, much like a traditional laid deck, and would complement all the effort put into renewing the brightwork. This resheathing was the most challenging part of the project. But we greatly simplified the work by making up sections of decking ahead of time, off the boat. We made accurate patterns of the areas to be sheathed, then made up slightly oversized deck sections of individual planks separated by a uniform 1/4″, then filled these seams with adhesive caulk made by Teak Deck Systems.
After a couple of days of curing, the individual deck sections could be handled very much like plywood panels, which allowed us to focus on accurate planning and cutting. Vacuum-bagging is a great way to install this kind of sheathing, and making up caulked panels ahead of time simplifies the work.
If we were vacuum-bagging individual planks to be caulked later, the epoxy would partly fill the seams, and then somebody would have to come back with a tool to clean out the cured glue before caulking. By making the whole panel up on the bench, all we have to do after vacuum-bagging is to clean up the perimeter. For me, that’s the main reason for using this method.
Mike Moros (who manages WoodenBoat School’s shop during summers) started off by carefully fitting three 1⁄4” plywood patterns. One spanned the width of the cockpit aft and butted against the aft bulkhead, and the other two followed the curve of the port and starboard seats and butted against the forward bulkhead as well as the previous pattern. Inset—Notations on the patterns showed the exact locations of things such as hinges, hatches, and the mizzenmast. The two diagonal lines are used to line up the pieces on the workbench later. The original teak veneers under the side decks were not worn, so we retained them.
Using the patterns, we determined the layout of the panels and laid out the strips following the pattern—in this case, for a panel that runs the length of the cockpit and the curvature and width of the side seat, minus trim. Spacers of 1⁄4” plywood guaranteed a uniform seam width. The planks were wedged together from the sides to keep the setup tight. Lead bricks held everything down. When ready, Mike hot-glued crossbars down to hold the spacing.
Using a power caulking gun, Mike filled the seams with caulk, taking out spacers as he went along and making sure to fill the space under the crossbars. The best way to fill a seam is to push the caulking gun, rather than drag it. In my experience, one advantage of caulking the seams before gluing down the strips is that the caulk adheres to the seam for its full depth. After each seam was filled, Mike used a putty knife to force the caulking deeply into place (as has been done on the first two seams shown here) to make sure there would be no voids. A sheet of plastic under the panel keeps it from being glued to the workbench.
It takes two days for the caulking to cure fully. During that time, Mike worked on fairing mating surfaces on the boat, then came back to work on the panels again. He used a heat gun to loosen the hot glue on the crossbars and a putty knife to pry them off. An 8” sander-grinder with 60-grit discs made short work of the excess caulking, taking care to sand as little of the wood as possible.
On the workbench, Mike lined up the new panels on the patterns. In the middle part of the aft seat, we planned one broad panel with an even number of planks, which would then be cut down the middle. Each side piece was glued to its corresponding side panel, leaving a centerline gap that would be filled by a kingplank. Careful work both in planning how pieces fit together and how margin boards, bullnose pieces, and trim would fit with the run of the strips paid off later. The curved piece visible at the lower left is a margin board laminated up of four half-width teak planking pieces, with the grain matched as well as possible to look like a full-width plank. Later, pieces were added to the margin boards to widen them because we came up with a change in the layout and decided that slightly wider margin boards would look better.
We planned to nib the planks into the margin boards to avoid feather edges and to give the planking a traditional look. It’s the right thing to do. Here, Mike laid out the nibs to blend in well with neighboring pieces, for example the seat hatches and a wide athwartships strip that fits under the mainsheet traveler. Especially at the hatches, we didn’t want to leave a small piece of a nib extending past the hatch seam where it would look out of place. The nibs also matched side-to-side, in mirror image.
After cutting and finishing the outboard side of the first panel, including its nibs, Mike cut to the line and finished off the edges with a block plane, chisel plane, and sandpaper. Then he transferred the shapes to the opposite side and made the same cuts there. Next, he traced the outboard profile of the seat panels directly onto the margin board, as shown here.
After tracing the profile of the panel, Mike marked another line on the margin board exactly 1⁄4” outboard of the traced line to allow for the caulking seam. Then he used a sabersaw to cut to the outer line and cleaned up the edges.
We used the completed port-side panel to make a mirror-image for the starboard side. Fortunately, things on the boat were pretty close to symmetrical, and any minor errors could be taken up during the margin board’s final fitting. The point of doing all this fitting so carefully on the workbench, with lead bricks holding everything in position, was to fit the margin boards so that we could glue them to the panels using the adhesive caulking. That way, we would be final-fitting only one full-length panel per side, from the cabin bulkhead to the aft bulkhead, with everything lined up, nice even caulking seams, and matching grain.
Once all the cutting and finishing were done, the margin board (port side in this case) was aligned with the planking panel using spacers. Lead bricks hold things in place, ready for crossbars to be hot- glued to hold the spacing while the seam was filled with caulking, making it part of the panel.
During test-fitting, more decisions had to be made, almost all of them based on aesthetics. Here, Mike and I were puzzling out the fit of the plank located under the mainsheet traveler. We also talked about the type of trim to use in way of the mizzenmast, and exactly where to cut the hatches. I knew I wanted margin boards at the forward and aft ends, but I hadn’t decided on their widths. It’s easier to visualize these details when the panels are in place. I wanted the forward and aft margin boards to butt tight to the side margins. By this stage, the two small panels that form part of the aft seat had been edge-glued with caulking to the long side panels, making the kingplank’s layout clear, and we decided the width of the aft margin board and the mizzen trim. There are several ways you could treat it. You could border the hatches out, but I’m not a big fan of bordering everything. I like a more traditional look, based on what would have been done if it were a real laid deck. It’s all aesthetics, and therein lie some decisions and changes.
While caulking seams were curing, Mike turned to surface preparation and fairing. He used a router to cut a uniform depth in a checkerboard pattern, then followed up with a broad chisel to break out the pieces. Even though there was a little variation in the thickness of the old veneer, we could mill off an even amount. After that came sanding. That’s boatbuilding—90 percent of it is sanding. And if I’m wrong about 90 percent, then it might be 95.
After fairing the old surface as much as possible, Mike applied thickened epoxy to fill low spots. He made a plywood tool to screed the epoxy to reestablish the camber of the seats, which had relaxed over time. After two applications and a final sanding, shown here, the surface was fair, the camber was right, and we were ready to install the new panels.
Having made up his beautiful, long deck panels, Mike then had to cut them up. The clean cut he’s making here is along one of the seams of a lift hatch in the seat. We planned to protect these edges with thin bronze trim pieces set flush on each side. Cutting out the hatch openings left us with four separate pieces of panel to vacuum-bag aboard the boat in a single operation. The hatches themselves were easier to vacuum-bag on the workbench.
Preparation for vacuum-bagging begins with cleaning up the whole area, then putting down paper-backed sealant tape around each area to be bagged. The idea is to have a continuous air seal, with no breaks. Bagging an awkward area can be tricky, and no place was trickier than the hole where the mizzenmast passes through the aft seat. Mike put sealant tape around the hole on the underside of the seat, then put a small plastic patch over the hole itself. Next, as shown here, he applied tape to the top of the plastic sheet. The tape also went up onto the bulkheads, then down to the side decks. Sealant tape, when being removed, can tear the surface off bare wood, so we first coated the wood under the side decks with epoxy, to be sanded off later.
Mike started applying the plastic “bag,” which is big enough to cover this particular panel, to the sealant tape. The outboard side, which is hardest to get at, came first. The visible yellow of the sealant tape shows the extent of this first attachment. Then we flipped the plastic up over the coaming to be out of the way while epoxy was troweled on and the panel laid down over it. The blue tape visible here made double-sure that no epoxy would get on the sealant tape. If that happens, the bag won’t seal, even if you wipe it off. It’s like grease.
Using a 1⁄16”-notched trowel, Mike spread thickened epoxy on the surfaces, making a very even layer, not too thick. Meanwhile, Richard Will had rolled unthickened epoxy on the underside of the four panel sections we were going to vacuum-bag.
Next, one of the glued- up panels was set in place, with the margin board butting against the remaining planking under the side deck and its forward end butting against the cabin bulkhead. The forward end of the planking on this panel has been cut back to allow room for the forward margin board and a 1⁄4” caulking seam.
After this first panel was in place—in this case, the port aft corner—a cut-to-size layer of peel-ply was laid down, followed by a cut-to-size breather cloth. The peel-ply prevents the cloth from bonding to the panel, and the sole purpose of the breather cloth is to allow air to pass though, for a uniform pressure all around the vacuum bag. I like to put a little extra breather cloth under the vacuum port, the bottom part of which is visible here. Then, the vacuum bag was flipped back into place, spread over the port fitting, and its edges sealed by pressing them into against the tape.
While peeling the backing off the sealant tape and simultaneously removing the blue masking tape a few inches at a time, the plastic vacuum-bag’s edges were pressed into the sealant tape, taking care to get a continuous seal. Good preparation and careful attention prevents leaks.
After the bag was sealed to our satisfaction, we poked a hole in the plastic and joined the bottom part of the vacuum port with the top part, sandwiching the bag in between. This section was now ready for the vacuum pump to be turned on.
We planned four separate vacuum bags to be done simultaneously, two forward and two aft. The aft bags ended at sealant tape that was placed in the void where the kingplank was going to go in later. We worked as fast as we could to get the plastic taped in place and everything ready to go, checking at the same time to see that the panels were lined up just right.
Once the vacuum pump was turned on, we could see right away if we had any leaks even without looking at the gauge. If you grab a wrinkle and you can’t pick it up, you’ve got a good seal. If you can pick it up, you have a leak. On the port side aft, we had a few leaks, and chased them down. Richard, working from the aft deck, found out that the sealant tape had taken the paint off the bulkhead, which caused one of the main leaks. I worked under the seat to check the mizzen patch. If you have air leakage, the trick is to not give up, but to be tenacious about it and find the leaks. Then, all a sudden, whump, you’ve got it.
After all the vacuum bags were checked and rechecked for leaks, we left the vacuum pump on overnight, pressing down with 14.7 lbs per square inch on the panels while the epoxy cured. The seat hatches and cockpit sole pieces were also vacuum-bagged, but they were done on the workbench, since they were much simpler.
After the epoxy cured, only minor cleanup was needed, after which the remaining pieces were fitted and glued in place individually. Here, the kingplank between the aft seat panels takes shape, soon to be followed by fitted pieces fitting around the box- sectioned mizzenmast, rails on the sides, a piece under the mainsheet traveler, and bullnose pieces along the edge of the cockpit seats. Caulking these final seams finished the job.
Brion Rieff is the proprietor of Brion Rieff Boat Building, 76 Flye Point Rd., Brooklin, ME 04616; 207–359–4455; http://www.rieffboats.net.