Lost Art Press

I am embarking on a project to (slowly) repopulate my deck with a better grade of furniture. First up is a pair of small folding tables. The design is adapted from one published a few years ago in Popular Woodworking. The table shown here is 20″ square by 38″…er, no…24.5″ high.

Designing furniture that can survive being left outdoors in the sun and rain without eventually degrading into a pile of sticks (or worse) is a whole ’nother enchilada compared to ordinary furniture making. There is significant overlap with campaign furniture, with the added bonus that wood that gets rained on moves. A lot. In fact, outdoor furniture has much in common with boatbuilding, and so one looks to suppliers like Jamestown Distributors for fasteners and finishes, SailRite for upholstery fabrics, etc.

The woods used must be durable (resistant to decay), of course. That limits one’s choices to the usual suspects: teak, mahogany and its relatives, white oak, etc. Some “cedars” are also suitable, though their relatively low strength means that the various components usually need to be beefier than in this table. Other North American woods that would be suitable are black locust and honey mesquite, both of which can be hard to find but are probably worth looking for. I haven’t worked with mesquite, but it’s on my to-do list, as it is supposed to have exceptional dimensional stability with changing humidity.

These tables are in sapele, an African relative of mahogany that’s rated as “moderately durable.” To give them a leg up (pun intended) in terms of surviving the elements, I’ve fitted them out with “shoes” made of UHMW polyethylene. The shoes are held in by a 1/8″-diameter oak pin, which can easily be drilled out to allow for replacement, in case they wear out or the whole experiment turns out to have been a bad idea.

There are four legs and four top supports, and no two are exactly the same. They come in each of the four combinations of inner/outer and left/right mirror-image pairs. It got to be so confusing that I made up some custom labels before I drilled all of the holes and counterbores.

For want of a good locknut…

Chris isn’t the only one with hardware woes. Because the joints of this table need to allow folding for storage, the fasteners have to stay put without being fully snugged up to the wood (which in any case is only going to be a temporary condition as the wood shrinks and swells). So some kind of locking fastener is called for. The standard solution is a nylon-insert locknut (leftmost in the photo below). These work well, but I didn’t want to use them, for two reasons. The main reason is that they’re thick, quite a bit thicker than an ordinary hex nut, which would mean having to reduce the thickness of wood left at the bottom of the counterbores more than I was comfortable with. The secondary reason is that nylon is not UV stable, meaning that they would degrade over time (although, to be honest, so will the wood).

Back when I was in the cyclotron business, we used some aircraft-grade locknuts that were all stainless steel, and worked by having a thinned-down collar that looked like it had been slightly squished in a vise. I wasn’t able to find that kind of locknut, but I did find some at McMaster-Carr that were superficially similar. I ordered a pack of the center-lock style (second from the left). You can see a small indent on the flat; there’s a matching one on the other side, and together they deform the thread just enough to create a locking action. Or at least that’s how they’re supposed to work. I found the nuts to be wildly inconsistent from one to the next, and most barely locked at all.

So I ordered a pack of the top-lock style (third from the left). These have three small deformations on the top of the nut, which you can just barely make out in the photo. These turned out to be a lot more consistent than the center-lock variety, although there are a still a few bad apples in the pack. These are the ones I’m using in these tables, but to hedge my bets, I also ordered a pack of low-profile nylon-insert locknuts (rightmost), in case the top-lock nuts turned out as bad as the center-lock ones.

Finishing touches

I haven’t yet applied any finish to the table. One option would be to leave it unfinished and let it go gray (like Megan the teak chair in the photo). I have a piece of sapele that’s been sitting outdoors for about eight years, and while there’s a fair bit of surface checking, it still looks pretty good, and remains structurally sound. I will most likely go with Osmo “One Coat Only.” I’m currently field-testing another piece of sapele with that on it, and it’s holding up well after a couple of months, but I’m going to see how it survives the winter before making a final decision. I decided early on against a traditional exterior varnish finish (e.g., Sikkens Cetol); I just can’t stand the look of varnish on unfilled open-grain woods, and I wasn’t about to try to use a pore filler on an outdoor piece.

In order to facilitate the inevitability of refinishing, the only parts of the table that are glued are the components of the two halves of the top. Everything else is bolted or screwed together. The top panels and the two leg braces are attached with #6-32 screws and brass inserts. The screws (also from McMaster-Carr) have a patch of locking goop that you can see in the photo. The makeshift insert installation tool on the right, along with a drill press to keep everything square, works better than any commercial tool that I’ve tried.

Lessons learned

If I were to build these tables again, I’d increase the thickness of the top from 1/2″ to 9/16″ and reduce the thickness of the legs and supports from 3/4″ to 11/16″. I also realized after it was way too late that bronze saw nuts like these might be just the thing for the leg/support joints.

I think the most important lesson I learned, though, is that writing a blog post using an iPad (from a hotel room) is never a good idea…

–Steve Schafer