Friday, August 14, 2009

Restoration--Just Another Day

I think the biggest crime against furniture has to be nails. Most places that a builder would use nails can be eliminated during the design process. That leaves nails to people doing really bad repairs.
The chest of drawers that I'm currently restoring is full of nails; in the moulding, holding the feet on, fastening the back to the frame. There's just bloody nails everywhere.

Don't get me wrong--nails have their place, it's just not in furniture. And not the way most people just bang them in at random. Most of the antique chairs that I've restored have had nails in them--it seems to be the answer to a glue failure in a mortise and tenon joint; bang a nail into the joint to hold them together. It doesn't work. Nails almost never work.

The problem is that wood moves and nails don't. Not that crap about wood having to "breathe" or needing to be "fed." Wood naturally moves over the course of a year, swelling when the average humidity goes up, shrinking when it goes down. It moves in very specific ways; less in thickness and length, more in width. If you imagine a fence, with a 150mm wide board nailed to a 50mm x 100mm (2"x4") cross member, if you nail the face board to the side of the cross member with 2 nails in the face, you are ultimately condemning the board to split.

The board will, over the course of each year, expand and contract over the same plane that the nails are holding it in a fixed position. By the end of a hot summer, the board may be 15% or more narrower than it was in the spring. The two nails, on the other hand, haven't moved at all. And then, heading down from the nail holes, two cracks. A board hung from a single nail top and bottom, on the other hand, will last so much longer, as the natural movement of the wood is not being resisted by the nails.

Movement differs between thinner and thicker wood, wider and narrower, but all wood is going to move. Movement is slowed and moderated by finishes like paint and varnish, but again, only slowed and moderated, not stopped. Wood always moves, just like gravity always wins.

After pulling nails out of the chest of drawers, I tossed a handful on the scanner. I've found three different kinds of nails: the earliest are the square-shaft nails you can see above. These were originally used to attach the back boards to the carcase. Since then small finishing nails have been added to the mix, and what look to be wire nails from an air-nailer. The latter have clipped rather than pointed ends--although they may have been finishing nails hand-clipped, I kind of doubt it from looking at them.

The front feet are proving to be the biggest problem in this restoration. At "A" in the above shot, you can see where, when the feet were being built, the 45° joint was "strengthened" by a nail. But because it was nailed into the end grain, it actually weakened the wood by forcing the fibres apart. Coupled with really crappy construction, it pretty much guaranteed problems down the line--and, as both front feet failed at the same point for the same reasons, I can feel pretty confident in saying that.

The feet were not made at the same time as the rest of the chest, pretty clearly being added later. Why is less clear--perhaps the originals were lost or destroyed, or possibly the piece is a made-up, being re-purposed later from a larger piece. But the quality of the work in the chest is streets ahead of the workmanship of the feet.

The ogee shape of the feet have been bandsawn out of some 25mm oak, reducing the thickness of the board to ~15mm at its thinnest. This thinness is why feet are usually made out of 50mm or thicker stock. The finish is much worse on the feet than on the rest--the feet are badly sanded (if at all) and covered with a stain so thick as to be paint rather than the beautiful colour and figure of the rest of the chest.

You can also see in the picture above that the pad which backs up the feet on the bottom of the chest has 2 screws holding it on--and three big nails at "B". There are also nails in random places inside the carcase, which I've removed, re-gluing pieces where necessary.

The bottom shows some of the problems furniture faces over the course of surviving a century; the holes above the crack are from wood-borer beetles, who dug into the unprotected wood and built galleries. These bore holes weakened the wood, and contributed to the crack that you can see. Inside the crack are the exposed holes, where the wood has been turned into Swiss cheese (except less tasty and more brittle).

A long life also loosened the dovetails on the carcase--but here is where good design helps. The dovetails don't allow the sides to move out and because the lifted piece is actually the bottom (the carcase was resting on its top when I took the picture), it doesn't generally move either. So even though the glue had failed at this point, the piece of furniture didn't fail, and, in fact, wasn't even damaged by the failed glue joint. So some hide glue back on the pins and tails and once again the carcase is stiff rather than loose, but other than that, not much difference will be noticed.

This shot actually shows the joint between the side and the top--again, the carcase is resting on its top rather than its feet. This shows a lot of the construction techniques used by whomever built this piece. The carcase itself is built out of softwood boards, edge-joined and then veneered on the outside. The inside is left unfinished (who's going to look?), and the corners are joined with dovetails which were then covered with moulding to produce the look of a solid hardwood chest.
It also shows where the anti-drop board was. This is just a filler piece that prevents the drawer from dropping its leading edge as it is pulled out of the chest. The top back corners try to go up and run into this filler piece. It makes for a nicer experience using the drawer, and is usually unnoticed--until its not there and the drawer drops.
But here there are four pieces of wood glued edge to edge to make up the top of the chest, and the anti-drop board is glued in across them. As the wood cycles through the seasons, swelling and contracting, it eventually breaks this glue bond, as the cross-grain board doesn't shrink and swell in the same direction. So every few years or decades, you have to glue it back in place. Not a mistake or a problem, but what I consider normal maintenance, like renewing the wax over the finish every so often.

The same construction techniques were used on the drawers, but instead of the double-bead moulding being applied, cock-beading was used around the outside of each drawer. Imagine an upper-case "D" rotated 90° counter-clockwise and the left edge aligned with the left edge of a lower case "L", so it looks like a hat hanging off one side. When you put this on the side of a drawer, the "l" covers the side of the drawer-front with the same type of wood, and the "D" covers both the top of the "l" and the edge of the veneer on the drawer-front, making a little bump all the way around the perimeter of the drawer. Looks really classy and takes quite a beating over time. And, as in the photo above, it often gets nailed back on. I removed the nails and re-glued the cock-beading with hide glue.

Hide glue is exactly that; a glue made from (usually cow) hides. The pearls, the dried glue, is re-hydrated and melted (I use an electric wok that heats a metal mixing bowl full of glue in a water bath). Hide glue, being almost pure protein, is wood-friendly, environmentally friendly, stainable, and, most importantly,reversible. If a joint breaks, it can be reheated and moistened and will reset the glue. Joints don't have to be perfectly cleaned, as hide glue happily bonds to old hide glue.
The downside? It stinks. A lot. Whoo. Now, I've been using it outdoors, so not so much a problem. But the smell of hot hide glue can be pretty nauseating. Also, being an organic product, it spoils. And when it does, it often makes hot hide glue smell appetising by comparison.
Hide glue also doesn't have long open time. As it cools, it very quickly becomes less tacky and more of a jelly. In wood-heated shops, the wood to be glued was often kept near the fire so that it would be warm when glued, lengthening the open time of the glue and improving the bond.
But in antique restoration, reversibility is a good thing, something to be strived towards. Everything I do to this chest I want to be reversible, non-damaging, and as close to the original methods as possible. If everyone who touched antiques took the same approach, well, there'd be a lot more antiques around. But people whack nails into them and use unsuitable glues that mean if you make a mistake, there's no way for someone else to fix it without massive damage to the piece of furniture.


  1. Bernie thank you so much for writing about the restoration of the chest of drawers. It looks wonderful!! I like the blog, hope you do more linocuts and restorations...interesting reading.

  2. Cool nails pic. Like all your art it has a vaugely fetishistic quality, unlike mine of course.

  3. *laughs uproariously at Ben's comment*

    No fetish in my work either.