Just out of the box... puzzle joints and nubs

This is my first stab at a stitch and glue kayak.  I had two surprises as I unboxed Mill Creek 16.5 today:

- puzzle joints instead of scarfs:  there are no references to these in the manual!  Where do I look for technical guidance on gluing them up?

- nubs on panel edges.  No reference to cleaning these up in the manual.  Sanding?  Planing?  What's recommended?

18 replies:

« Previous Post       List of Posts       Next Post »

RE: Just out of the box... puzzle joints and nubs

I am building a Shearwater right now.  Regarding the nubs, I took a small piece of 2x4 and wrapped a piece of 60 grit sandpaper around it and found I could remove each nub in less than 30 seconds.   

RE: Just out of the box... puzzle joints and nubs

Nubs - yes, sand or plane, whichever you prefer. Personally, I like a Shinto rasp, but whatever you find easiest and most comfortable. They're just nubs.

Puzzle joints - paint the edges with unthickened epoxy, let it soak in for a few minutes, then brush thickened epoxy on the edges, and put the fingers in the holes. Sandwich between 2 pieces of polyethylene and put a weight on top to hold it all together and flat until the epoxy cures.

Use the same mix that the manual tells you to use for gluing wood together. If you have never done a puzzle joint before, dry assemble one first for practice.

Have fun,



RE: Just out of the box... puzzle joints and nubs

Thanks, folks.  That was my intuition, but thought I'd check in!


RE: Just out of the box... puzzle joints and nubs

   Follow-up question:

Lot's of emphasis on pressure required to clamp these puzzle joints in this forum.  But these are really just fancy, curly butt joints, aren't they?  Pressure sufficient to flatten everything makes sense to me, but you're not actually pushing the edges together (the puzzle handles that), so why so much concern about weight, screws, etc.?  

RE: Just out of the box... puzzle joints and nubs

To hold them in place. Fresh epoxy acts as a lubricant. Unless the joint is immobilized the panels could slip out of alignment. On an 18-ft panel it doesn't take much to give you an unfair edge.

Realisically speaking, though, the CNC joint tolerances are so tight that any misalignment and resulting unfairness is tiny and easily fixed with a plane or sandpaper or a light touch with a router, etc. As long as the same fix is made to the corresponding panel on the other side. whether it needs it or not, it will be invisible. Some new builders don't realize that fairness and symmetry are more important than accuracy so they focus on accuracy and you see all the resulting immobolization concerns.

Another reason is holdover habits from experienced scarfers where the joint is free to slide everywhere unless measures are taken.

But that's the joy of building your own. You get to do what is easiest and makes you most confident of success.


RE: Just out of the box... puzzle joints and nubs


    You are correct regarding pressure on the puzzle joints.  Lots of weight/pressure on a perfectly flat surface will ensure that the joint is flush.  If the joint is not flush, you end up with a cosmetic issue when you san it flush.  Personally, I use a technique that shows up in the newer kayak manuals.  Assemble the puzzle joints dry then tack in place with drops of CA glue.  If you go slow and do a couple drops at a time, you will end up with a perfectly flush joint.  The CA sets instantly when you hit it with a spritz of accelerant.  Once the joint is glued together, back it with a strip of fiberglass.  The front (out) side will get coated when you glass the hull.  Shearwater Double panels pictured immediately after glassing the back side.

RE: Just out of the box... puzzle joints and nubs

   1 gallon jug of epoxy as a weight worked well for me.

Plastic sheet and a board on top to evenly distribute the weight. I have stacked 4 strakes with no probllem.


RE: Just out of the box... puzzle joints and nubs

Careful removing the nubs!

Some are there to help position parts during assembly. Those're generally a bit longer than the vestigal bits left from the CNC process where the workpiece being machined needs little tabs to hold parts already cut out in place.

I used a pull saw on my nubs, cut close to the design edge then planed / sanded to the line. Shinto rasp'd work well too, even a sharp chisel used with caution, or box cutter.

One tip on puzzle joints: I found knocking the edge fuzz off with a chainsaw file made assembly much easier. Those errant edge fibers can get in between things, prevent a good close fit.

RE: Just out of the box... puzzle joints and nubs

   @spclark - Thanks.   Clearly there are "tabs" in this Mill Creek 16.5 design that position the bulkheads in the bilge (I probably could have measured...) but the rest appear to be bona fide "nubs".  

RE: Just out of the box... puzzle joints and nubs

   And in the unlikely event you epoxy a puzzle joint to the wrong part or wrong alignment................. a heat gun will soften up the epoxy so you can get it loose and make the corrections.  Just in case. 

RE: Just out of the box... puzzle joints and nubs

OP, you asked the questions I was afraid to ask (with my Wood Duckling)!  Thanks.  I was generating excuses for delaying my puzzle joint epoxying yet again (settled on "Not sure I have a good brush for the epoxy").  Glad to see some detailed advice up above.  I'm afraid I'm out of excuses.  ;-)

I was surprised to discover that some of the pieces in the wood duckling kit were still connected (post-CNC).  I tried using a Japanese pull saw but quickly realized the teeth on the back of the saw are hard to keep away from the adjacent piece of plywood.  I ordered a bonsai saw from CLC to cut out the last few hatch pieces (concentrically connected), though I'm now wondering (given comment above) whether a box cutter might be sufficient.

RE: Just out of the box... puzzle joints and nubs

We found a finger-sized block plane handy for shaving down the nubs and other similar little jobs (building our PMD) where we were afraid the saw teeth might tear up the plywood.  Very clean cutting if the blade is kept sharp.


RE: Just out of the box... puzzle joints and nubs

   @Andrew. One of my first questions/problems in the build also. See these


RE: Just out of the box... puzzle joints and nubs


Kinda haunted me early in the build and did affect the fairness somewhat

RE: Just out of the box... puzzle joints and nubs

   For the nubs I find that a "File Sander" with a 100 grit belt makes short work of them. Less than $40. The sander also comes in handy elsewhere - 

RE: Just out of the box... puzzle joints and nubs

Original Poster here.

Thanks all!  I nipped off the most of the nub waste with diagonals and finished with a block plane.  Quick and easy.

Put the puzzle joints together with CA glue spots and accelerant, per Nick Schade's tip on this site.  Likewise, quick and easy!  Debating whether to back with glass, as suggested above, or just carry on - they seem pretty solid, and if it is good enough for Schade it's probably good enough for me!


RE: Just out of the box... puzzle joints and nubs


Some advice I got on the subject to consider 

To address the one issue/idea presented about using CA glue on the puzzle joint.  I wouldn't recommend it.  This isn't a "tack weld" situation.  Even though covered later with glass (on only one side?), you want a good strong bond between the panel ends.  The CA glue will wick into the wood and leave a much "drier" and weaker joint, and one that might not hold up well when doing all the bending and twisting of the panel that will happend during the stitching process that shapes the boat.  I also suspect that there would be a not completely filled line left that might cause the puzzle joint to show, making it harder to sand and fill as a part of the epoxy coating or glassing-over process.  I'm not sure about your specific panel on the MC bilge area, but often only one side of a panel is glassed, the other is only coated with epoxy prior to painting or varnishing.  It is very hard to "hide" imperfections that are not able to be sanded perfectly smooth and flat during the process.  Even a haiir-line sized imperfection (stray chip brush bristle or glass fiber) will tend to show.  So for reasons of both strengh and the ability to sand the panels absolutely flat and fair as a part of the finishing process, I say no to CA on puzzle joints.

RE: Just out of the box... puzzle joints and nubs

This is a great thread!  Thank you everyone for the great advise.  The puzzle joints I put together on my NED over the past weeks were a lot easier than the scarf joints that were on my Ches. 17., however some of the first joints I did not perfectly align.  The floor board, this was the first one I did and I did not get enough clamping pressure down on it and there was a noticable step between the two boards.  I decided on the floor board, to sand it smooth, since I am painting the outside of the hull and possibly the floor (undecided yet).  I did not break through the plywood layers, so that was good.  Learning from this, I pretty much was able to get the other boards aligned, except for the left stern #4 joint.  At first this joint looked good, except for where the lap is milled in.  Apparently, for whatever reason, the thin lap section was off by more than 1/2 of its thickness.  It looks like the fwd section, poped out.  I contemplated what to do and read this thread over and over.  Since this was dealing with the thin lap section, I was concerned about breaking the wood if I tried to separate the two with a heat gun, so I went with the sand - fill - sand method described above.  Sanded the rought epoxy (from wrinkles in my plastic sheets), then I applied unthickened epoxy to the low area on the lap on the inside.  Gave it a night, and repeated the epoxy application to the outside area.  Lots of epoxy running, but overall I was able to fill the low spots and it is fairly smooth.  On the outside, it looks like I will sand through about 50-75% of the wood thickness to get smooth, but this has now been replaced with epoxy on the inside.  I don't plan on sanding into the wood on the inside, just sanding the excess epoxy off to get the joint smooth. 

« Previous Post     List of Posts     Next Post »

Please login or register to post a reply.