When I bought All Ways (then Northstar), my 1964 Pearson Triton, I knew I would have a major project on my hands. The surveyor told me to walk away, as the cost of repairing the decks was greater than the boat's value even if it were in perfect condition. And that was just the start of his list. But I have three (soon to be four) kids in college and had more time for a project like this than money to spend buying a boat. The owner already had another boat and was anxious to sell, so I bought it for something like 15 percent of the initial asking price.

 From the start, I was determined to make All Ways look the classic she was. To me, that meant, among other things, a laid-teak deck. Hey, I had to rebuild the deck anyway, so why not make it yare? Thus began my research into ways to rebuild a fiberglass deck with a rotten core and how to construct a laid-teak deck.

First came the unglamorous task of repairing the rotten core.

After reading everything I could on the subject, I decided to combine several approaches. For the deck, I elected to use the doughnut approach. The first step was to tap out the entire deck with a plastic hammer, marking the areas that would need repair. That was both side decks and a goodly section of the foredeck. Then I began cutting circles out of the marked areas. In all, I cut more than 200 circles, each 3 inches in diameter and each carefully numbered to allow replacing in the original location. The holes were placed on an approximate 6-inch grid so nearly 50 percent of the deck surface was removed.

The smaller holes were cut with a spade bit used to get close to the toe rail. I quickly gave up on it and used the 3-inch hole saw. By adjusting the center drill bit, I was able to cut only through the top skin and the core, leaving the bottom skin intact (except for a few misses). Next I took a spare Allen wrench, ground the short end to a point, chucked it into an electric drill, and removed the core from the holes, undercutting them by nearly an inch around the full circumference. Any additional loose core was vacuumed out at this time as well.

While the remaining exposed core dried out, I cut 200-plus circles of new core from sheets of Airex foam using the same hole saw. Note that although the old core was 1/2-inch thick, I used 3/8-inch Airex to allow for the thickness of the epoxy used to glue them in place on top and bottom. After flooding the remaining core with acetone and allowing it to evaporate overnight, I was ready to reassemble the deck. I used West epoxy mixed to a peanut-butter consistency with 404 filler for strength and microballoons for volume and weight reduction. Using a 1-inch-wide spatula, I forced this goop into the recess cut around the perimeter of each hole. I then buttered each foam circle, pushed it in place, buttered the previously removed circle of top skin (from which all core had been removed), and set it in place weighing down the whole mess with plastic grocery bags filled with sand.

This process creates a doughnut of epoxy around each foam circle, effectively isolating it from any remaining core and creating hundreds of columns gluing the top and bottom skins together and supporting the deck. I used epoxy rather than polyester for this process because it is a much better and more flexible glue, allowing it to remain glued to both surfaces even if there is some flex to the finished deck. (See Good Old Boat, Nov., 1999, for a discussion of virtues and risks of epoxy vs. polyester).

Now I was ready for my laid deck! The Gougeon Brothers have a great description of using epoxy to create a laid deck without all the traditional screws and bungs that are prone to leakage. It also uses very thin teak (1/8 inch), which saves cost and weight. My only hesitation was all the articles I was reading about folks tearing off teak decks due to the deterioration from smog and pollution and the resultant constant cleaning and upkeep of what was supposed to be a care-free deck. I decided to research other options. I considered and rejected Treadmaster pads and painting the deck with non-skid like I used for the cabintop. I like the sweeping curves of a laid deck! I looked at some commercial composite products (one is cork) that come already assembled as a laid deck in sections, but the prices were outrageous. At about this time I discovered a product called Trex.

This is a manufactured wood substitute that is 50 percent wood and 50 percent polyethylene, both recycled (plastic bags and sawdust). It's produced by Mobile Corp. and used primarily in 2 x 6s for outdoor decks around pools and such. It has a 25-year guarantee against deterioration and meets the Americans with Disabilities Act rating for non-skid with no treatment - even when wet. (This act specifies everything from how steep an access ramp can be to how non-skid is measured for public areas like swimming pool decks. -Ed.)

It weathers to a natural silver-gray that is very similar to untreated teak. There is also a brown version, but the color is not very stable from batch to batch and fades unevenly, so I would not recommend it. Trex can be cut, drilled, and sanded with regular woodworking tools, and although most things won't stick to it, epoxy glues it well. In the last few years, since I chose Trex for my deck, several similar products have come on the market, one of which uses cedar for the wood. I do not have any experience with these and so cannot recommend them, but you might want to research the expanded possibilities if you choose to follow this path.

Testing Trex

Could Trex be what I was looking for? I wanted to see before making a commitment, so I talked my local lumberyard out of an odd piece of 2 x 4 Trex and cut it into 1/8 x 1 1/2-inch slabs. I cut a piece of 1/2-inch plywood to match the curve of my sidedeck and made up a test bed using all the techniques I would use on the final project. I left the piece outdoors through our New Hampshire winter, exposing it to sun, rain, ice, snow, and even driving over it with the car. In spring, it still looked as good as new. I tried prying layers apart but split the plywood before I could break the glue bond. I put one end on a block and drove the car over it, severely bending the piece. I soaked it under water and poured boiling water on it. Nothing seemed to harm it. It was time to put it on the boat.

Several elements enter into the design of a laid deck. (See Chapelle's, Boatbuilding, for a discussion of traditional laid-deck planking.) Features may include a wide cover board at the sheer, a king plank at the centerline, and nibbing pieces wherever planks will run into another structure at an acute angle. The planks themselves may be laid straight (parallel to the centerline) or sprung to the sheer. Rectangular structures, like cockpit seat hatches, can just be laid with straight planks or have edge boards that meet at a 45-degree angle, or are butted, and these edge boards may be the same width or slightly wider than the planks. Likewise, the edges of cabinsides and cockpit coamings can have a nibbed margin, or a wide plank can be cut to the curve. Each boat will require different options, depending on how it looks to the owner.

Wide cover board
	I wanted maximum yare and chose a wide cover board at the sheer, nibbed king plank on the centerline fore and aft, and margin boards for cabinside butt joints for the front of the cabin and the coamings. All rectangular elements in the cockpit were treated to edge boards meeting at 45 degrees, and, using the fact that I was cutting these elements from 2-inch stock, I chose to cut the edge pieces in an L shape where there was an exposed outside corner to cover  The material for the deck was cut from 2-inch stock (2 x 6, 2 x 8, and so on).
While the exact thickness is not important (sanding will remove any variation), the width must be uniform for it to look right. This is accomplished by using a table saw or a bandsaw to rip a 2-inch board into 1/8-inch strips, all of which will be a very uniform 1 1/2 inches wide. Wider strips for cover boards, the king plank, and so forth, can be cut on a bandsaw with the 2-inch board on edge. After the design is complete and strips cut, the next step is to dry-fit the entire deck. I wanted wide cover boards at the sheer, but my testbed experience convinced me that also making the outer board thicker, as is traditionally done, was not practical as there was no way to sand the joint between the cover board and the first plank.
Thus, I cut all pieces to the same thickness. The cover boards were cut from a 2 x 4 on edge, resulting in 3 1/2-inch strips. These proved too wide and were narrowed to 3 inches.
More supports
Trex is rather floppy. When used in construction, it must be supported much more closely than same-sized fir or pine. Despite this, bending 3 x 1/8-inch planks to the sweep of the sheer was a daunting task.  It never would have happened with teak. (The alternative is to cut the outer plank from wider pieces, which are then scribed, cut to the curve, and joined with scarf joints, thus eliminating the bending.) As I bent each piece into place, it buckled and refused to lie flat. Several 20-pound weights and several days' time encouraged them to take the curve more gracefully. The inner edge was screwed down with sheet-metal screws, and washers placed in the joint line and spaced every 6 inches or closer, if necessary. This is a tedious process, as you must force the board into the curve, drill a pilot hole, and then hold it in place while screwing in the screw. Two drills are essential - one to drill and one to screw. Extra hands are helpful, but not necessary - I did most of it alone.

The screws are initially left loose so the outer edge of the next board can be slipped under them. Once the inner edge of that board is secured loosely, the screws along its outer edge are tightened down. This is repeated, ad nauseam, for each plank. The midline end of each plank must initially be cut square and carefully positioned along a line parallel to the centerline and spaced out several inches to create the appropriately shaped king plank. A screw placed at the end will aid in repositioning the plank later.

For reasons that will be obvious after the nibs are cut, this screw should be placed about 1/4 inch in from the outer edge. Once all of the planks are laid for one side of the deck, the ends must be nibbed for the king plank and margin boards, which are fitted later. Alternatively, the planks can be butted port and starboard, creating a herringbone pattern, but I don't think it is as attractive, and it requires working on both sides of the deck at once.

A word about screws: you will need several thousand. In my opinion, a square drive is a must as are correctly sized pilot holes. Fiberglass does not like taking screws. So-called sheet-metal screws have a straight shank, whereas wood screws are tapered. Once you cut the initial thread in the hole, a sheet-metal screw screws down smoothly, but a wood screw is a struggle all the way. Also, since you will be removing and replacing all of these screws after applying the glue, a wood screw would not grip as securely, since it could not be driven in to the same depth as it had been originally due to the thickness of the glue.

This is not a problem with sheet-metal screws. The screws should penetrate the outer skin but stop within the core. I used #8 5/8-inch screws, which create the correct 1/8-inch seam. If the screws are too long, double washers may be used when dry fitting. It is important to set the torque on your screwdriver so it can start the threads but will not crush the Trex or squeeze out all of the glue. Once half the deck is dry-fit, and screwed down, it's time to take everything back up.

First number each board so you can put it back in the same place. Also, mark the location of each screw on both sides of each board. Sand the deck surface with 30-or 40-grit to create a tooth for the epoxy to grip. Decide how many courses you can lay at a time and mix up enough epoxy to wet out both the fiberglass and the Trex.

Three or four courses
	Be sure to wet out the bottom side of the boards, both to preserve your markings and because of the nibbed ends and the set curve they will have taken. I found three or four courses to be about right, depending on the length of the run. Use the slow hardener or you will go crazy. I used 9 ounces of unthickened epoxy for four strips and the same area of deck. Once the epoxy reaches the green stage, mix another batch, this time adding carbon and 404 to create a mayonnaise-consistency mixture. My formula for four strips was: 12 pumps each of resin and hardener (5:1 from the way the pumps are set up, for a total of about 16 ounces), 2 cups of 404, and 1 1/2 tablespoons of carbon. Mix very well. Spread this on the deck with a 1/8-inch notched trowel. The screw holes are an excellent guide as to where to spread the goop.
As you position each board, use an awl to locate the screw holes and re-insert the screws. If you use the awl to locate and mark one hole and then locate the next hole, you can insert the awl in the second hole to force and hold the plank into place while you insert the screw in the first hole.
Once again, you insert the screws loosely until the next plank is slipped under and secured and then go back and tighten them down. This time, however, you must spray the screws and washers with non-stick kitchen spray (Pam, or something similar) to prevent their being glued in by the epoxy. I put them in a shallow container and sprayed away while shaking them about, to be sure they were well coated.

Messy process
	If you use the right amount of goop, and apply the right amount of torque to the screws, the seam between the planks should just be filled with squeeze-out.This is a messy process and while epoxy vapors are of little concern, skin contact is, so long sleeves and gloves are mandatory, as is frequent cleaning of your tools with vinegar or acetone. The screws should create an even spacing of the seam, but if one spot wants to close up (near the end of a plank usually), either use more screws or insert a short piece of 1/8-inch Plexiglas sprayed with non-stick.
After the last plank in this batch is laid, it is important to scrape up all the oozed-out epoxy on the deck surface so it does not interfere with laying the next course. You can then proceed with wetout of the next course, and so on. At the end of the day, you will probably not be finished. Wipe up any oozed-out epoxy on the remaining uncovered surface with acetone. Even so, I chose to re-sand that area the next day to remove all traces of black, and ensure a good bond.

It is important to remove all the screws the next morning as the epoxy will not have reached maximum strength yet, and you will be more successful with any screws that were not adequately coated with non-stick. This is where the square drive is a real must. It allows you to walk the screws right out. If one gives you a hard time, try tightening slightly first. If that doesn't work, try hitting it with a hammer while twisting. Heat will release the epoxy, but it will also burn or even melt the Trex, so I do not recommend it except as a last resort. If you do use heat, be careful, and be prepared for repairs.

When all the deck has been laid, and all screws removed, it is time to fill in the screw holes and any voids in the squeeze-out. I tried several ways of doing this. If you mix the epoxy a little thinner (say half as much 404), you can use a syringe to fill the holes. This works reasonably well, but be sure to mound up some excess or you will have to refill them as the goop flows into the hole. If you have lots of the seam to fill, in addition to the holes, just trowel epoxy goop over the whole deck, scraping as much off the Trex as possible. In fact, in areas where the epoxy oozed out almost, but not quite, up to the level of the Trex, I was concerned that the next, very thin layer would not bond well, so I routed out the seam with a 1/8-inch bit in a small trim router. I modified the base with a vertical fin to serve as a guide and make following the flowing curve of the seams easier. This will also smooth out any seams that were not quite uniform.

After all seams are filled, don your mask and begin sanding. Epoxy dust is extremely toxic and a particulate face mask is mandatory whenever sanding epoxy. Use 80-grit in a belt sander, and sand everything flush. If the epoxy seams are quite proud of the Trex, or if the Trex is uneven, you can sand at right angles to the planks for the initial leveling, but final sanding should be parallel to the planks. You will undoubtedly find a few areas in the seam that remain shiny, indicating that the epoxy is below the level of the surface you sanded. As long as there are no holes through to the core, you can either re-rout, refill, and re-sand, or tolerate the shininess; it will fade with time. Another option for small areas is to hand sand, with a folded piece of sandpaper to dull the surface.

At this point you are either done or almost done. The scratches left from the 80-grit leave a grain pattern in the Trex that can please the eye and increase the non-skid value. Alternatively, the deck can now be sanded with a random-orbit sander and 80-grit to create a smoother surface, which still has good non-skid properties. The rougher surface will catch more dust and dirt and require more frequent hosing off.

Decide on the width of the king plank. This will be dictated by the width of available stock and your eye. My king plank is 5 inches wide, which allows a little error when cutting from 5 1/2-inch (2 x 6) stock. (If in doubt about what looks right, mark out various options on the deck with a marker.) Establish your centerline and mark parallel lines spaced half of the king plank width to each side (fig a). These lines need to be thin for accurate placement and cutting. One option to achieve this is to make a thick line with a marker and then score a thin line through the marker with an awl or knife. The top photo provides an overview of the process.

Cut all planks square, and lay them so that the foremost outboard corner lands exactly on the scribed line (fig b). Be sure spacing is exactly 1/8 inch. The planks will run parallel to the sheer. A straight edge across the corresponding plank corners should be perpendicular to the centerline of the king plank.

Make a nib-marking tool from scrap wood (illustration in use in photos c-e). It should be about 4 inches long (exact length not critical). Assuming you are using 1 1/2-inch planks, make the piece 5/8-inch wide, then cut away 1/8 x 1/2 inch along one edge, leaving a 1/8 x 1/8-inch edge. The critical thing is that you have a 1/8-inch ridge for the seam, and that the shelf that remains is 1/3 of the width of the plank. The thickness is not critical, but 1/2 inch works well. I made mine from a scrap of teak because I like nice tools, but any close-grained wood will do.

Starting at the aft end, place the marking tool in the seam between the innermost plank and the next, flush with the end of the first plank. The 1/8-inch ridge fits in the seam. Make a mark along the edge of the tool that will be 1/2 inch from the outer edge of the plank (fig c). Fine lines are again important. Use a very sharp pencil or marking knife.

Before moving the tool, also mark the inner edge of the next plank (fig d). This line will be a projection of the outer end of the first plank. Move forward to the seam between the second and third plank and repeat.

Use the edge of your tool, or a straightedge, to connect the second mark (at 1/2 inch) with the first mark you made on the second plank (fig e). (On the first plank only, the inner mark will be where the inner aft edge of the plank intersects the guideline.). For all others, the point will be the projection of the previous plank's edge, which is actually slightly outboard of the line. See illustration above.) This is the line you cut for the nib. Repeat for each plank as you move forward. Cut all planks and glue down.

Lay a piece of paper over the entire length of the space where the king plank will be (fig f). I used a roll of butcher's paper, which is excellent for making patterns. Be sure the shiny side is down and secure in place with tape to prevent movement. Rub all the edges of the planks through the paper with your thumb. This is known as the dirty finger technique. A little of the carbon powder works well also. You don't want to press down into the void under the paper, just pick up the sharp edge of the planks.

Remove the paper and carefully cut along the marks (fig g). Test fit the pattern in place and trim or shim as necessary. Transfer the pattern to your king plank. An easy way to do this is to tape the pattern in place and then use a fat-tip marker to straddle the edge of the paper, marking on both the pattern and the board (fig h). When the pattern is removed, you will have a nice sharp line. (Photo shows paper slightly moved aside.)

Cut the king plank to fit. If you are confident in your markings and the width of your blade, you can cut inside the lines, as you eventually want the king plank to have a 1/8-inch seam all around it. On the other hand, it is far easier to cut it twice than it is to stretch it, so if the first cut is without seam allowances, you can use your marking tool to mark the 1/8-inch seam allowance during the test fit. Remember to leave this line when you make the final cut. Finally, glue the king plank down, holding it in place with the screws at the end of each plank and weights along the center (fig i).

fig a	fig b	fig c
fig d	fig e	fig f
fig g	fig h	fig i

One of the advantages of Trex is that it can be repaired almost invisibly, since it has no grain to match. To affect a repair, first rout out the damaged area down to the level of the deck. Next, use your sander and some scrap Trex to create sawdust. Mix this with epoxy to create a thick, peanut-butter paste, and trowel it into the defect. Sand smooth for a barely detectable repair.

This article is about an innovative technique for applying an unusual deck veneer as a substitute for a laid-teak deck. The author did extensive evaluation and sample testing before he committed his time, resources, and boat to this process. We asked that he evaluate his finished deck for one season before we printed his article. He did and is satisfied with his work.

A process such as this will be controversial. It closely follows a process published by Gougeon for laying teak. System Three does not recommend using a thickened graphite/epoxy mixture to fill the expansion joints between the teak strips. They recommend using their epoxy for gluing teak down and polysulfide compound for filling the gaps between the planks. Their concern is with the expansion properties of teak, which is not sealed with epoxy on top. The issue is that epoxy between the planks will not be flexible enough to allow for expanding and contracting. The partners at MAS Epoxies, like Gougeon and System Three, have experience gluing teak, but have concerns about using Trex because they are unfamiliar with it and
because they know epoxy will not bond to the polyethylene in the Trex without very specialized pretreatment. They recognize that epoxy will bond to the wood in the Trex. The folks at Westcoast Teak Products Ltd., who lay teak decks do not use epoxy. They prefer two polysulfide compounds, one for gluing the strips down and one for filling the gaps between them. All of these people are being cautious, as they rightly should be. They have experience with teak but not with Trex.

The deck described here is not teak, and as careful as Mark Parker has been in developing this process, it is a new process with only one satisfactory season of proven use. Mark is betting on a satisfactory outcome. Sailors familiar with traditional teak over fiberglass decks with screws and plugs realize that method of construction can have serious problems when the wood wears, the plugs fall out, and leaks through the screw holes start.

We are offering this article to you the way Mark offered it to us. It looks like an interesting alternative way to lay a "wood" deck. If you consider doing this, you will certainly want to make a thorough and independent evaluation of the method and perhaps follow up with Mark to see how his deck is doing. Either way, our hats are off to Mark for taking the process this far. We needed a better way. Maybe this is it. - (Ed. Good Old Boat)

Mark Parker, M.D., is director of the Emergency Care Center at The Cheshire Medical Center in Keene, N.H. He's been sailing since college - Sunfishes, Lasers, Hobie Cats. His work on a 16-foot trimaran, a "work in progress," was temporarily sidelined when the Pearson Triton, All Ways, received a higher priority rating. Mark sails with his family on Narragansett Bay.