Puzzle joint epoxy non-idealities

I'm obsessing a bit on the puzzle joints.  I know that the fiberglass tape backing is probably not very critical in terms of structural integrity, but I have concerns about bubbles (up to dime-sized) that somehow appeared despite my best efforts to (1) apply epoxy on the wood beneath, (2) saturate the tape, and (3) apply smooth plastic (I used relatively thick file folders) to make the surface smooth.

It seems like the epoxy ended up extra thick, which might explain how I ended up with large bubbles beneath it.  I also ended up with parts of the tape that were starved of epoxy.  The picture illustrates my problems well.

After reviewing dozens of posts on the subject, my guess is that the best path forward is to use a cabinet scraper and/or sander to open up these bubbles so they can be filled with epoxy later (when sealing cockpit) to ensure the wood isn't exposed to an air bubble / potential water intrusion later.  Is that right?


16 replies:

« Previous Post       List of Posts       Next Post »

RE: Puzzle joint epoxy non-idealities

To clarify, I don't intend to focus on those bubbles when sanding but to sand back uniformly so that epoxy is smooth and they are exposed (and to feather into wood on both sides of joint).  

RE: Puzzle joint epoxy non-idealities

Thinner is better when it comes to using plastic - of any kind - for covering puzzle joints during this stage. The heavier the plastic the greater the force you have to apply to squeeze epoxy uniformly across the joints you're assembling.

If your plywood pieces aren't too warped, you're really careful in assembly and work on a nice, flat surface you can squeegee the plastic smooth with a plastic "Bondo-type" blade thing quite nicely, leaving the weave filled and no bubbles.

I'd fill the voids in your fill coat ASAP before takling a cabinet scraper to the cured epoxy that's there now. The idea being to fill the surface uniformly over the weave rather than trying to pare away the hard epoxy now without cutting into the cloth that's barely covered.

Where you had bubbles it looks like the cloth is saturated well, just not filled. So the sooner you get more epoxy on those areas the better the bond will be.

And off-topic I have to commend your skill in getting an image file to appear! What hosting service do you use? Was the process fairly easy to get us the image we see in your post? Others have been less successful, to varying degrees, myself included.

RE: Puzzle joint epoxy non-idealities


Thanks for the suggestion.  Before sanding it back I will plan to drill two holes in each bubble - one for a syringe and one for air to escape - as I've seen other suggest.  It's probably been about 4 days since I epoxied it - hope that's not too long.

Maybe my gravest mistake was not using a spreader/scraper to thin out the epoxy (I only used a brush).  I certainly didn't expect the epoxy to end up as thick as it did - I think this is the thickest epoxy of the 10 I've done (and the worst joint overall).  I thought the weight I put on top (probably 20-30 lbs) would ensure that the epoxy spread out and ended up thin, but maybe that's not enough weight or the epoxy had already started curing.  Either way, sounds like I would have been better off thinning it with a scraper/squeegee (which I have) rather than relying on gravity and compression to squeeze it out.

RE: Puzzle joint epoxy non-idealities

Oh - and I'm using photobucket.  Amazingly, I had to pay something like $5-$10 for a month of picture hosting.  Surely there is a service out there that allows you to share a small rather of pictures (rather than hundreds) for a lower fee, but I don't know what it is and didn't have the patience to look for it.  It seems like we have somehow arrived in the dark ages of photo hosting.  No way to monetize it, I guess, unlike YouTube and the like.

RE: Puzzle joint epoxy non-idealities

Er...seems like there should be a service that allows you to share/link to a small number* of pictures for a more modest fee, that is.

By the way, it sounds like you're recommending adding epoxy to the thin areas on top of the fiberglass too before starting to sand, so I'll do that while filling the voids/bubbles (and just smooth it with a scraper/squeegee).  This obviously adds a lot of tedium to the process, but I'm happy to do it to get good results.

RE: Puzzle joint epoxy non-idealities

   mmmm......you must be reading the instructions. I've never put reinforcing tape on a puzzle joint that I can remember. Of course........... I'm old.  

That plastic and weight thing. Naw, 'don't do that either. Just squeege the epoxy onto the cloth on the hull and work it in so it wets the wood when you do the whole thing. 

You have some sanding to do or you will have lines in the finish. I'd sand it down to the "fabric" level,  all over. You could also load it up with more epoxy to fill in the weave but that would make it a speed bump when you add the cloth later. (and heavier)  It also would not soften the premiter line of the epoxy puddle (foreground of picture).   Unless this is going to be burried inside and not covered with cloth. 


That dime size white dot happens when you don't work the epoxy through the cloth brush or squeege. ....OR when the wood got contaminated like with .....oily fingerprint, sweat drop, etc.  See Nick Schade videos. 


Note that many when squeeging on the fill coat. Use the squeege to wipe off the excess and scrape the extra into a cup to throw away. Keeps it from getting too thick. 

RE: Puzzle joint epoxy non-idealities

Thanks, Grumpy.

I had already filled the bubbles with epoxy (after drilling entrance hole for epoxy from syringe and exit hole for air) and tried to fill in the weave by squeegeeing more epoxy into the low spots by the time I read your comment.  It was surprisingly easy to fill the large bubbles with epoxy (took maybe 30 minutes to patch up 5-6 joints), but I still have a bunch of small bubbles (maybe as large as 1/8").  Not sure those are worth trying to fill.

I plan to sand down to the fabric once the epoxy has cured (as you suggested).  I heard that the epoxy is toxic until it fully cures, so I might wait a week or so to knock it down with a cabinet scraper and sand.

I'll check out Nick's video on contamination (I'd forgotten that nugget of wisdom from the many threads I reviewed).  Sounds like I can improve my odds by cleaning the surface of the wood or (maybe easier) sanding it a bit just before doing the puzzle joints and by keeping the epoxy thin by squeegeeing.  I have not been doing either.

I applied the weight during curing because I saw horror stories from others about the two planks becoming misaligned during curing, requiring heavy sanding of one side (and maybe sanding through the veneer on the outside, prohibiting a bright finish).  At least my planks are level.  I'd be strongly contemplating a viking funeral otherwise!

RE: Puzzle joint epoxy non-idealities

Manual instructions aside, it's easy to disassemble misaligned puzzle-jointed parts using heat to soften the epoxy.

If 'glass cloth has been used at the same time as this basic assembly step this kind of operation becomes much more difficult.

For those who are new to puzzle joints, simply assembling them with epoxy alone is simpler, easier to work out a means by which the two ends can be held flat. Once cured, any excess epoxy can carefully be scraped / sanded away. It's at this stage a layer of glass cloth might be applied if warranted, once the joint itself's been found to be satisfactory.

More typical practice, where joints are in areas that will receive high stress or below waterline, is to cover them  with 'glass cloth bonded in epoxy at the same time as the rest of a hull's inner and outer surfaces.


RE: Puzzle joint epoxy non-idealities


Glass tape over puzzle joints depends on which boat you're building. My WD12 specified it so I used it.


One way I've avoided this kind of problem for years now is to roll up the tape and put it into a container of epoxy to let it soak up what it wants. After a minute or so, I take the soggy roll out of the container and knead it in my gloved hand to work the epoxy from the outside into the center. When it's all soaked through, I squeeze from one side of the roll to the other to remove as much of the excess as I can. Then I finally unroll it onto the puzzle joint and squeegee off any excess with my gloved finger onto the surrounding wood. By pre-impregnating the glass before applying it to the wood I get a perfect wet-out every time. If it isn't perfect, I just don't put it onto/into the boat until it's fixed. Contaminated glass shows its treachery right away and is discarded and replaced with a fresh piece before even touching the wood.


As far as plastic and weights go, if the wood is thin enough it needs the weights to maintain alignment. If it needs the weights, it also needs some way to keep the weights from sticking to the joint. Again, this step depends on which boat you're building. I used weights and plastic and glass tape on my thin-hulled WD12 but not on my schooner that had 1/4" and 1/2" thick wood that was thick enough to self-align.



RE: Puzzle joint epoxy non-idealities

I'm builgind a wood duckling - forgot to mention that.  The manual does call for applying fiberglass tape to the puzzle joints, yes.  It does not state that the fiberglass tape should be applied immediately after gluing the puzzle joint and applying the seal coat of epoxy to the wood, but I'd somehow come to the conclusion that it was necessary to do so in order to get a good bond between the tape and the seal coat.  I really like the idea of decoupling these steps to (1) afford myself the option of easy undoing via heatgun as spclark suggested and (2) to allow me to carefully inspect the wetting of the fiberglass after application as Laszlo suggested (rather than covering it with plastic and weights and thus being blind to it during curing).  Laszlo, your suggestion to pre-wet the fiberglass is interesting.  I think I saw that later on in the manual in the section about applying tape to the interior of the cockpit.

I think my new process for each fiberglassed WD8 puzzle joint will be the following (please correct if you're bored enough to read and spot any obvious errors):

0. (maybe) Briefly sand or use a solvent like denatured alcohol (and a ~"sanitary" rag or pape towel) to clean the wood surface of contaminants on the inside of each puzzle joint.

1. As Laszlo has suggested in other threads, brush UNthickened epoxy onto the mating surfaces to ensure they don't end up starved of epoxy.

2. Brush thickened epoxy (epoxy + cel-o-fill) onto the mating surfaces and push puzzle joint pieces together, removing excess "glue" to the extent possible (at least on the inside) with a squeegee.

*** Allow puzzle joint epoxy to cure ***

3. Brush thin seal coat of (unthickened) epoxy onto both sides of puzzle joint and allow to cure (with plastic on both sides + weights to ensure joint stays flush).

*** Might also allow this to cure before applying tape.  Seems this might reduce odds of surprise bubbles in tape later ***

4.  Apply epoxy-saturated fiberglass tape to the puzzle joint.  Use a squeegee (either finger or tool) to ensure that epoxy is appropriately thin.

5. Using cabinet scraper and/or sandpaper, smooth epoxy and wood/epoxy interfaces.

RE: Puzzle joint epoxy non-idealities

   Herbie, you said " I heard that the epoxy is toxic until it fully cures, so I might wait a week or so to knock it down with a cabinet scraper and sand."

Someone correct me if I'm wrong but my understanding is a cabinet scraper or other methods of cutting green epoxy is fine, dust from the green epoxy is the problem. In fact green epoxy is when cabinet scrapers shine.

RE: Puzzle joint epoxy non-idealities

   Since I have almost the same question I am curious how important it is to fix this since it is inside the hull?  I am more curious how to avoid it in the future as I glass and epoxy the outside of the hull. 

Does the type of brush matter?  I have been using some foam brushes and it seems they don't like the epoxy much.

Problem shown in linked video here looks the same as above.


RE: Puzzle joint epoxy non-idealities

   So things I am seeing here.

  • Keep the Epoxy thin and squeegy it out. 
  • Thinner plastic will help between epoxy and weights.
  • A thin layer of epoxy on the Joint which is allowed to cure before taping might help smooth it out. (Is the outerfull epoxied before the glass is applied?)

How much is it important to sand the taped joint on the inside.  I know it is weight but really how much?

RE: Puzzle joint epoxy non-idealities

Neither dust nor scraper shavings from 'green' epoxy is good for you with regard to body contact.

(Scrapers may be better for some tasks than sanding. Get used to doing both so you can choose what's best for the task to be undertaken.)

The big issue is chemical reactivity; epoxy cures by chemical reaction. Hardeners are typically formulated to be strongly alkaline (like lye). Liquid hardener can burn your skin, and in combination with resin can stay reactive until the cure is well along, sometimes days under decent climate conditions. Time & temperature are the ruling factors here - a day in one case might be enough for 'green' epoxy to cure to a level permitting sandability, where it may be as little as a five or six hours if it's warmer, or several days if it's too cool.

I was sanding MAS LV epoxy this morning that I applied yesterday afternoon. With fast hardener the batch was hard to the touch by 7 PM. The operation was hand-sanding so I was wearing chemical cartridge-equipped respirator all the while, and periodically vaccuming up the dust being produced.

There's a lot of debate over chemical vs. physical bond with epoxy. Going back to the reaction that makes it useful to us, when hardener and resin molecules link up the reaction is going the way we want it to. If the mix is wrong, excess hardener may be left unlinked, the material won't be optimum. Having excess resin does the same thing, why we strive to measure accurately with each batch.

With re-coating, if the reaction is still going on in a previous application the bonds are still being formed if the material is at all malleable, 'fingernail-impressionable' is a good description. Fresh epoxy will readily bond to a previously applied batch. Once the reaction has produced a hard, sandable surface there are too few open molecule links for a fresh batch to bond to. Which is why we sand, or at minimum abrade the surface with something like ScotchBrite. You're not only breaking up the hard glossy finish but you're breaking a significant number of molecular bonds that will remain open until a frersh batch is applied.

So waiting long enough to sand epoxy properly isn't a bad thing when your plans are to add more. The resulting bond is chemical as well as physical in nature.

What's to be avoided at all costs is contamination that might prevent the bonding you desire.


RE: Puzzle joint epoxy non-idealities


I'm a novice (which is obvious if you realize that I'm the one who started this thread), but I do not recall seeing others use foam brushes to apply epoxy.  It seems like tape is typically saturated with either a chip brush or by pre-soaking (as Laszlo mentioned above).  It seems like squeegees (a.k.a. spreaders) are the tool of choice when epoxying big pieces of fiberglass fabric, but I've also seen people use foam rollers there.

The only thing I remember seeing foam brushes used for was "tipping" of the varnish (after brushing it on) to get rid of bubbles.

RE: Puzzle joint epoxy non-idealities


I have switched to chip brushes.  I think it made a difference.  The other thing that made a difference was the thinner plastic. 

I point out both in the third installment of my vids   

Good luck with your build. Mine will be taking shape soon.

« Previous Post     List of Posts     Next Post »

Please login or register to post a reply.