Please visit our home site at

Anke and I live aboard WAYWARD, and wrote about it's design and construction at

Access to the net comes and goes, so I'll be writing in fits and spurts.Please feel free to browse the archives, leave comments where you will and write... I'll respond as I can.

Fair winds!

Dave and Anke
triloboats swirly gmail daughter com

Wednesday, February 8, 2012

Barge/Scow Bottom Planking - Making It So

Setting up the Jig
It laughs at the sea, that bow does; it challenges the sea; it snorts defiance at the sea. And withal it is a beautiful bow; the lines of it are dreamlike; I doubt if ever a boat was blessed with a more beautiful and at the same time a more capable bow. It was made to punch storms. To touch that bow is to rest one's hand on the cosmic nose of things.
 From The Voyage of the Snark by Jack London

NOTE: We've developed an improved variation on this theme, posted here, though this post introduces helpful terminology.

Barge/Scow Bottom Planking - Making It So

Let's start by noting that the bottom of a TriloBoat hull (box barge/scow) is not rocket science. It has a couple of curves, yes, but these are easily dealt with. So lets roll up our sleeves and walk through it.

There are two ways to go about building any boat... right-side up or upside down. For TriloBoats, I'm tending these days to right-side up. A couple of steps are harder, this way, but we don't have to build any temporary structures, and we don't have to turn the boat... a task that still freaks me out despite many successful roll-overs (no failures, so far). For this post, I'll stick with right-side up.

For Andy Stoner's MARY ELIZABETH (T32x12), we spent a week indoors building components. Deadflat sections (the flat portion of the bottom between end curves), side sections, bulkheads and transoms.

At the end of that week, we set up the building jig. This consists of two substantial beams that run about one half the overall boat length. You should set up as low as possible, while still being able to swing a hammer underneath the boat-to-be.

Prop 'em up on solid blocking, and parallel. To check for parallelity, measure between them and adjust until equidistant (parallel in plan view). Then eyeball across opposite edges and adjust one end of one beam until edges align (parallel in profile). Finally , lay a straight edge across their upper faces and adjust until it lies flat across both at all points (coordinate their rotation). Repeat until all satisfy, then block and shim securely into position.

Note that this accounts for square and true, but not plumb. TriloBoats are self-rectifying, so plumb is elective. If you're type-A, go ahead and adjust for plumb with a level.

A few terms for our discussion

8ft sheets on 12ft beam...
Lots of extra work.

Note jig is not level.
Once the jig is ready, lay out the deadflat components in order. Depending on how you're going about things, these could range from plain, full sheets of ply to pre-fabbed and finished stacks of ply (layers offset to provide overlap at the butts... minimum 1ft). Glue 'em up and join them, one component at a time.

I prefer at least one of these layers to be 3/4" ply. This is thick enough to accept ringshank nails, which provide a good lock and clamping pressure until glue dries. Butts may be backed up, later, by fiberglass tape set in resin, straddling the butt seams. Waterproof, and adds a LOT of strength.

4 hrs from flat platform!
Once the deadflat is together, you've got a big, flat platform. Join and frame the side panels... let set up overnight. Next day, pull aside, erect bulkheads on your marks, position sides and join to bulkheads and bottom. This day is exciting. Lots of friends help... reward with pizza, beverages and music.

Blah, blah, blah... boatbuilding deets... you may build the rest of the boat before you get to the bottom end curves.

The important thing to know, at this point, is that the side panels are framed along their bottom edge with 2x lumber. Installed, they form the jig for shaping the bottom. Their framing's lower edge faces provide area for gluing and beef for nailing. All thwartships, bottom planks (sheets of ply) overlap side panel framing and planking, ending up flush with the outboard faces. All sheets full size!

Both ends of the deadflat have offset courses of ply, resulting in a flap at each end. Make sure this is clean and clear of glue (may have to cut some goosh out). End curve courses join this flap, and continue the offsets in the steps below.


Thick or stiff ply doesn't like to bend. It want's to stay flat or pretty close. So we coax it along.

One thing that helps, while laying out the curves on the side panels (you will have done this a while back) is to let the outboard ends of the curves - 12 to 18 inches  - run free (let your batten spring to its natural lack of curve). Your ply courses will be much easier to bring into line, this way, than if you had made the curve continuous (I may have picked this trick up from Lofting by Alan Vaitses?).

SLACKTIDE's kerfed 1st course
The next thing we'll do is kerf our ply where it's curved (make thwartships, partial-depth cuts). This lets it bend similar to a roll-top desk. Experiment, and leave as many plies intact as you can while still being able to bend the sheet. If your wood is stiffer, you will need deeper kerfs and space them more closely. For 3/4 AC Fir, we find that leaving 2-3 plies out of 7, and spacing every 3-4 inches works well. Half inch AC Fir may not need kerfing at all.

Lastly, plan ahead to let the bottom run out past the ends. Once done, you'll trim flush with the transoms, at your leisure.

The first course will be kerfed, if necessary, on it's outboard face. Don't need to kerf where the curve runs flat toward its end. When joining, fill kerfs with glue. Plan ahead so that nails are embedded in solid wood (avoid kerfs).

Installing this first course is easy. A cake-walk (that reminds me... don't start hungry!).

We like polyurethane glue-from-a-caulking-tube for these external hull joins (e.g., 3M5200). It's moisture activated, slow curing (won't kick today), gap filling, water-proof, elastomeric, and wicked strong.

Spread glue along flap and framing, a bit further than one sheet will cover. Be generous (want no voids), but don't over do it (expensive). Look for a bit of glue oozing from between sheet and framing, inside and out, to ensure good contact.

Position sheet, and fasten with nails (we like ring-shanks), working from inboard to outboard. Remember it's the glue doing the work, the nails are there to clamp and provide moral support. See if one, every other 'plank' (space between kerfs), is sufficient; one every plank, if not.

Repeat until first course is laid.

1st layer on ME was 1/2" (kerfs in 3/4", 2nd layer, only)
The piecing, shown here, is unnecessary with 12ft sheets (or 8ft beam)...
This has been cobbled from 8fters

The second course is trickier due to time sensitive glue, large area gluing and the need to close gaps between sheets. It's going to be a much easier job for two than one. Make sure your help knows every step of what's going to happen before you start. Talk it over and try to spot problem areas in advance.

If the second layer requires kerfing, do so on it's inboard face (when joined, the kerfs should be hidden between courses). Stagger the kerfs so that they do not match up. Mark guides for nail pattern on the outboard face, avoiding kerfs and first course nail pattern, so you don't have to guess under pressure.

We like LPU (Liquid PolyUrethane, e.g. Gorilla Glue) for laminating plywood sheets. It's moisture activated, temperature tolerant, relatively easy clean-up, foams to be gap-filling, and doesn't take much. Looks and acts like honey. Sets up in about 45 minutes in cold weather, faster when warm. Time a test batch to get the feel. Set up materials near to hand before starting. We get about one sheet per 20oz bottle. Have back-ups... if you run short, you're in trouble.
Ready? Breathe. Note the time and start.

Squeeze glue from the bottle into all kerfs (both layers... use large syringe for overhead kerfs). Squirt a pattern on the second course's face and spread with long, flat spreaders (paint stir sticks work great). Do this before installing, while it's lying flat. Move efficiently and without panic. LPU can be spread thin, but you should leave a skim of glue (no areas scraped down to the wood). More won't hurt, and it's false economy to use too little.

Position the sheet and start nailing into place, working from inboard toward the ends, forcing gap slack ahead of where you're fastening. Have persuaders ready to go (props, jacks, screws* with drivers) and use 'em at the first sign of trouble (gaposis). Get it in position and locked down with the effective minimum of fasteners.

Repeat until this section is planked. On this course, try to hit each butt seam between sheets with glue. Once all sheets are in place, go back and fill in your fastener pattern. If in doubt, add more 2x clamps*.

Sit back. Breathe. Laugh, dance and hug. Admire your work. Kiss and make up for any harsh and hasty words loosed in the heat of battle. Wipe up excess glue. Assess, address and debrief for the next round.

Job well done!

Closing up the bow... note blocking, bottle jack and 2x4 clamp,
old clothes, gloves and knee-pads...
and yes, she's got glue in her hair!

*Screws are very useful for closing up gaps. If there's no framing to screw into, set up a short section of 2x4 and screw into it, drawing courses together (see photo above). Remove and plug holes after glue sets up. We like high thread count, square-drive screws and a brace or cordless drill. Size screw length for bottom thickness plus 1 1/2+ inches.


  1. Dave,
    Gorilla Glue does make a gap-filling epoxy, but their foaming polyurethane glue, while it indeed fills up gaps, has little strength when the two surfaces are not closely aligned. The ring-shank nails you are using may be holding your boats together more than you realize :)

    From the Gorilla Glue website:

    Gorilla Glue as a filler --

    "Since Gorilla Glue, a polyurethane glue, will foam and expand 3 to 4 times the amount applied, it will fill any gaps or voids. However, it will only fill the void cosmetically, not structurally. It cannot be used as a filler or a sealer. For the tightest bond possible, Gorilla Glue can only be used on tight-fitting surfaces.

    You will need to use an epoxy. An epoxy has gap filling ability that a polyurethane glue does not have.

    I hope this helps.


    Best, John

    1. Hi John,

      Very good points, though I'll partially disagree. Thanks for raising them!

      I agree completely re the superior strength of epoxy to EXPANDED LPU (voids). The only problems with epoxy are cost, and occasionally environmental sensitivity. But no doubt it's superior to LPU (both, at full strength, exceed the fiber strength of softwoods and most hardwoods).

      The LPU is a shoestring choice. Our tests (sawing apart various pieces to examine the joins, then beating them to smithereens) indicate that excellent contact is the rule along framing, and at every 'clamping' nail driven into 3/4" ply. MOST everywhere else has been tight, as well, with only very rare, paper-thin voids of a few square inches.

      We used to use Weldwood (plastic resin glue)for laminations (large surface area), but voids were not filled. Didn't seem to make a strength difference, and voids were plasticoated, but it worried us. The LPU expands to fill these, excluding water.

      The structural value of expanded LPU is somewhat better than Mary Ellen indicates. There are 4608 square inches per sheet of ply. If there's even 5lbs of adhesion per, we're over a ton of adhesion per sheet. I'm not sure what the expanded foam weighs in at, but it's probably in the 20+lb/ range? Can be wedged apart, but certainly doesn't want to just fall apart... this in an area that has no forces TRYing to pull it apart, and surrounded by solid, full-strength bonds. Not excellent, but, as a rare occurrence, nothing that worries me.

      I didn't mention it, but our general nailing pattern is on 6" centers in mid- sheet. Each nail produces an 'island' of solid contact... these overlap to virtually eliminate voids.

      The only problems we've seen are where the ply has had an extreme, localized warp from exposure to moisture. We avoid these for lamination.


      It's a different situation with 3M5200 (also PolyUrethane) . It's adhesion is on a par with epoxy, is cheaper (though not by much) and easier to use, and elastomeric. It's solids make it gap filling (though, like thickened epoxy, won't expand into a void). High viscosity, though, has made it difficult to use for lamination. We use it for bonding hull edge joins (including the first bottom course).

      Here, I like it better than epoxy. In Port Townsend, we often bought 70ft masts, professionally laminated from old growth lumber for $25 or so. They were ruined. A little join failure at one end would walk the length of the mast. We'd finish the process with wedges, remill 'em and use 'em for whatever.

      What I think happened is that the epoxy (and resorcinol) glues being used are plastic, but not elastomeric. A bit of separation at one end opens a slim trianglular wedge, with the tip at the line at the leading edge of intact glue. Weathering concentrates stresses at that line. Since it can't 'give' stresses rise toward infinite (divide by near-zero). The epoxy doesn't fail, but the wood next to it does. We'd see wood fibers emerging from the entire length of the glue stratum).

      In contrast, an elastomeric adhesive distributes those tip forces along a REGION of glue (3" deep into the glue, say, rather than a line), absorbing them relatively safely (divide by considerably MORE than zero). I've never seen one of our 3M5200 joins fail or even weather.


      So, in conclusion, I won't argue anyone out of epoxy. We use it ourselves for many jobs. But, at this point, the LPU has proven amply strong for ply sheet lamination. It's other advantages keep us interested for this and other, select jobs.

    2. Hi Dave,
      What a treasure trove, all the details here! Those of us reading are lucky campers indeed. Another glue question: when you and Anke were splitting delaminated masts in Port Townsend, did you find that the ones that were glued with resorcinol were coming apart just like the ones done with epoxy? We had the epoxy fail on AUKLET's masts because of heat from storage problems, but I recall the Pardey's favoring resorcinol for masts, specifically to avoid the heat delamination problem in the tropics. But it sounds like this splitting that you're talking about is a problem for either glue? Guess that's an argument for those nice solid trees for masts!
      -- Shemaya

    3. Hi Shemaya,

      Those masts were sometimes epoxy and sometimes resorcinol... both had the same symptoms, but the epoxy (which looked as though it had been thickened)left more of a footprint. Some of the stretches looked totally solid, but most could be popped apart with slim wedges. Again, it seemed to be wood fiber failure, not the glue itself... I think it's from high stress load at a vanishingly small fiber line due to non-elastic bond.

      Hadn't remembered that about the Pardeys, but they were writing earlier in epoxy's career.

      Aside from rigidity, there are important steps that that even professionals can miss, which can degrade the bond.

      Epoxy, for example, benefits from an initial, saturating coat and relatively loose contact between sides being joined (epoxy starved joins are weak). It bonds molecule for molecule, so precise proportions and mixing are important. Temperature and humidity play subtle roles, as do handling contaminants, 'tooth' and amine blush.

      All in all, it's robust enough that some error usually gets taken in stride, but each step away from optimal is a step downhill in performance.

      One trick we use with any laminated spar is to seize the ends with tarred, synthetic marlin (seine twine). This mechanically limits any weather induced glue failure, which tends to start at the ends (moisture intrusion at the endgrain, relatively low thermal mass and lack of reenforcing, neighboring structure). Cheap and effective insurance.

      I came on the scene after resorcinol was popular. I understand that it requires excellent contact and clamping (we're excellent at neither!). Again, every step away from optimal is downhill.

      One reason we like the 3M5200 is that it's near idiot proof (LPU not so much). As long as it's clean and dry and you use enough, you'll have to chisel it apart and grind the residue (I hear heat works, but generates nasty, toxic fumes).

    4. Hi Dave,
      Thanks for expanding on that! So 5200 for spar laminations? Not to be a detail pest here... fascinating, if it's been successful for spars. Makes sense about the way it would allow for stretch, rather than cracking apart.

      Many thanks,

    5. Hi Shemaya,

      Our 'Quick and Dirty' Master (mentor) turned us onto the 3M5200. He did use it for a mast, which, so far as I know is still sailing 20 years later (haven't heard otherwise, but not sure we would have).

      We've never used it on a spar... only spar we've laminated up was a short mizzen mast and sprit, and we used Weldwood.

      So, while it makes sense to me, no personal knowledge, and not too many examples.

  2. Technical writing techniques on this post. Accompanying video would do wonders in spelling this out for the beginner.

    Risking heretical blatherings here but all the kerfing could be eliminated by spreading the curve out the whole length of the boat, advanced sharpie style. Also a chance to incorporate a increased chance at self righting by locating the heavy bottom lower. And a chance for the classic 45 degree side-bottom join rumored by Bolger to lower resistance by no little to no chine eddying. Getting into more work here though and just straight barge sides with curved bottom would work well too. Perhaps on drying out the couple could sit on opposite ends of their big barge and see-saw... just kidding...well... maybe not! Built a 37 foot sharpie myself this way with two layers of 3/4" on the bottom with no kerfing. BTW.... is this a precursor chapter of the upcoming triloboat build book? Lots of great info in this post and, if continued, may well allow a crafty proletariat kinda dude or dudette to build a paid off wonderful floating sail powered home from these posts alone (a book would be better though but a seriously motivated person, perhaps one who just had all his worldly posessions confiscated by the hungry state, could do it via library printer). Thanks for the effort here and hopefully the right folks are beginning to see just how do-able these simplistic barge homes truly are.

    1. Hi Gomez,

      You're right about continuous rocker easing the curves and avoiding kerfs. The barge/scow deadflat pushes the curves into a small region toward the ends (I use 1/4 of the LOA at each end), forcing them into a more abrupt curve. Rather than try to bend chine logs in this section, I specify longitudinal, sawn frames. Or tape 'n' glue.

      But WAIT! There are trade-offs.

      If you took a barge shaped bar of soap and carved continuous rocker into it, you'd end up with a pile of shavings. Shavings represent lost displacement and interior volume. If you want to regain displacement, you'll have to go deeper (oops... add to the 4ft side panel... there's some work!). Headroom inside the hull would diminish both directions from the deepest point of rocker.

      Building-wise, you now have to lay out and cut the side panels over the dead-flat to the new curve. If you're using a chine log, it can no longer be any taller than you can bend to the curve (LUNA, for example, bent 2x2s... the Trilo version would use unbent 2x4s, and sawn 2x planks on edge fore and aft).

      Your mid-ships interior is no longer rectilinear. All bottom edges of all furnishings will have to be fit (spiled or patterned, and cut to the curve). If you want to install lubber furnishings, they'll need shimming or trimming. If you don't want a curved sole, you'll have to add a flat, false floor. All thwartships frames require bevels.

      One thing we observed with LUNA was, while trying to move it with rollers, it was uphill to the top of one, then down and bottom out before the next one. It wasn't as easy as it looked!

      The deadflat works great with rollers. On trailers, a barge is super easy to secure... rockered hulls need blocking. That easy jig we used? Doesn't work with rockered hulls.

      So there are trade-offs! Small things all, but they add up.


      RE self-righting, on a GIVEN draft - The barge has a lower CG (both hull and ballast) and CB. You can lower them by increasing the draft on a rockered hull, but that's cheating. 8) Next thing you know, we won't be shoal draft.

      In general, both forms approach self-righting by having sides that are high for the beam, and higher, water-tight deck structures that add 'deep' floatation if and when inverted.


      RE matched side and bottom curves (a la Bolger ADVANCED SHARPIES) - The 45deg line you mention isn't an actual line ON the boat... It's what you see in section view (looking from one end) at the foreshortened chine between matched sides and bottom... if they're perfectly matched (seldom the case, even in AS hulls), you will see an APPARENT 45deg line (it shows up that way on a 2D, drawn plan).

      You can do the same thing with an ADVANCED BARGE (to follow Phil's nomenclature)... at the end curves, the sides would curve in an equal amount. Transoms would be typically wider, in ABs, than ASs, as the curves are shorter, longitudinally.

      In practice, I think the aft curves aren't as critical... any turbulent cross-flow generated aft (which the matching works to avoid), doesn't have much hull length left to cause problems. Phil, also, seemed to relax the matched curves aft to widen the aft transom (increasing cockpit volume).

      We've toyed with thoughts of building an AB. But we keep concluding that costs outweigh benefits (they're MUCH harder to build). If we want the match, we'd go AS, again.


      RE My long-awaited DIY book... yes. And yes. Anyone can draw one of these up and build it. That's why my plans are so cheap! 8)

    2. Any word on when the book will be available? Definitely interested in getting a copy. :)

    3. See DIY Book evasion below, in reply to GOMEZ.

  3. Re the gorilla glue: small boat designer John Wellsford did actual strength/failure tests with epoxy, gorilla glue, and a few others and found epoxy a lot stronger. Findings on his site I think. Epoxy is difficult to work and tough to target the right setting temperature window. I always suit up and use the maximum filter mask. Do I remember you once writing about using only 3M 5200 backed up by metal angle on the chines? I often wonder how epoxy would behave in really cold conditions if the chine hit a rock or something: brittle and thus prone to cracking failure? Seems 5200 would be more elastic in such situations.

    1. Hi Gomez,

      I'll have to check John W's site. As I mentioned above, our tests have shown well bonded LPG to be at least as strong as the wood. After that, it's kind of moot.

      Epoxy is great for its versatility, and there are methods (plasticizing wood) for example, that are way expensive, but miraculous in strength.

      It's interesting to note that the first thing anyone does with epoxy's potential is to reduce scantlings (lighter boat = more speed). The RESULTING strength is not any higher than that of a heavier, slower boat built with less expensive methods and materials.

      We should recall, though, that epoxy is a relative new-comer. Many other glues are plenty strong for the jobs they're called on to perform, and have long track records. For most situations, epoxy exceeds the need, and does so at considerable cost.


      We do use the 3M5200 (or SikaFlex)... see reply to John (first comment, this post).

      We like the bronze angle, which provides mechanical protection to the chine and covers the gap where copper plates meet along side and bottom. But it shouldn't really be structural. If it WERE to come into play, the critical bond along the chines will have already failed.


      I've never seen epoxy get brittle. Even when it's pretty hard, it's still plastic.

      Weldwood, on the other hand, does get brittle... almost like glass (can cut you bad, too, if you brush along a slice). It will craze if flexed in the open... not sure what it's doing in a tight contact join. Good track record, though... Dynamite Payson built hundreds of boats with it

      These days, though, in all the places we used it (flat side of bulkhead framing) we're tending to Titebond III... it's already mixed (no toxic, powdery, breathable GLUE to suck into the lungs), NON-toxic and costs the same per resulting volume, even here in AK (must be cheaper south where water weight doesn't cost so much).


    1. Foxy Babe is Anke. No CLUE as to how this guy got so lucky! 8D

      Seriously, Anke likes building more than I do (if I had a magic wand, I'd USE it! I wanna go sailing!!). She would build for fun. Not me.

      Women have been up and coming in sailing/building circles for a good long while, now. Not a "man's world", anymore... and good riddance to that!

  5. Thanks for the attention to detail in replying to these posts. I learned a lot. What a archive this is going to be for a first time boatbuilder or bargebuilder. Once all your findings are compiled in a logical order for a book it's going to be about all a person needs to get the job done. Probably better than George Buehlers "Backyard Boatbuilding" although George reveled in outright counterculture IRREVERANCE.

    1. Hi Gomez,

      Discussion helps me, too. Challenges the little ruts in my head.

      One of the purposes of the blog is to make progress toward the book. It's mostly written, with only a few areas to be simplified or fleshed out. Then illustrations and go. These blog posts (and discussions) are helping me cut to the chase.

      At THIS point, I'm hoping the DIY book will be ready sometime this summer. But, as you know, I've been saying "check's in the mail" for quite a while!

      RE BEUHLER'S BACKYARD BOATBUILDING - my book will definitely NOT address carvel construction, engine installation, or even (thoroughly) sail plan design. I'll be referring to other resources, such as his, where it's already been adequately and entertainingly explained.

      There's some overlap, but I'm focusing on TriloBoat (box barge/scow), ply construction and the design of hulls, decks and interior. I'll discuss mast placement and tabernacles, but not rig details.

      Hmm... I feel a post coming on!

    2. Dave,
      If you're writing a book, based in part on what you're writing here, I might mention my confusion regarding the post you wrote on Traveling Off-Center Boards. In particular the diagram labeled "Schematic for components above and aft of Window" I found confusing. I'm not sure how that diagram fits into the previous diagram. I also wish you could draw more lines from caption to item. For instance, I'm puzzled what item "Eye Bolt" and "Horn Cleat" and "CLAM?? cleat" etc,refer to.

    3. Hi John,

      If you don't mind, I'm going to quote and answer you on that post (link via BLOG ARCHIVE in the bar to the right of these posts).

    4. Thanks for the excellent feedback!

  6. Happened to find this after much research into Gorilla glue for lamination. My last boat, a 31' Bolger topaz had a double bottom which led to ongoing problems to this day.

    I have built many boats including the topaz and an earlier F9A trimaran of epoxy, but laminating with epoxy, left voids in the bottom. What I find is, I had or have a small leak which filled the void between lamination, even though the bottom is still strong. Had I had gorilla glue, this would have been avoided.

    I will be using this in my building of the 32' this summer. Thanks for posting this experience, you have answered my question.