Wednesday, October 31, 2012

Mizuya (4)

Now we come to a tricky part of the design: drawers. Now, drawers, per se, are not generally an inherently tricky thing, though getting a good fit can certainly be a challenge, and there are different ideas out there among the drawer making fanatics as to what constitutes a good fit. There are aesthetic factors to consider of course, whether to mold or bead the drawer front, what sort of pulls to fit, etc. Those are matters to work out to be sure in this build, but the tricky bit for me at this juncture concerns drawer construction.

I've set myself up to struggle in this cabinet build by trying to use glue and/or metal fasteners absolutely minimally, and as one commenter, given those constraints,  noted a few posts back that he'd,  "...like to see how you make drawers". Hah! I have been wondering much the same thing myself. This is not the first time I've pondered drawer construction in depth, though it has been a long time since I made a cabinet so drawer construction issues have been well on the back burner for some time now. Not any longer, so let's proceed....

At the outset, there is a fork in the road, design-wise: metal sliders or not. Metal sliders are strong and smooth, easy to install, and replaceable. They allow the drawer to be opened all the way for full access to the contents, and there are even fancy soft-close mechanisms for them. On the flip side, those metal mechanisms take up a certain amount of room, and that room has to be subtracted from the drawer itself. Drawers with metal slides have a drawer face which promises capaciousness, however it is not something that is delivered when the drawer is opened. Also, the construction of many drawers these days involves a box to which a face - a facade - is affixed, and this doubling of the front elements seems wasteful and the facade part does not appeal much, if simply on philosophical terms. But really, more than anything else, because this is a cabinet modeled on traditional lines - my interpretation of them at least - I really can't 'go there' with the metal drawer slides. So, my drawers will be all wood.

Going 'all wood' still leaves one with plenty of choices, and in fact there are possibilities to use wooden substitutes for the metal slides, like dovetailed side-mounted battens, and so forth. I gave those sorts of mechanisms some considerable thought, but I saw an issue in there, literally, with wooden drawer slide mechanisms, where the combination of dust and dirt, wearing surfaces, and shrinking/swelling materials tended to suggest to me that performance may not be great for much of the year. In other climates perhaps, but not around these parts where humidity cycles are vicious.

So, that would appear to, er, draw things down a bit, wouldn't it? Now all I need to worry about is how to put a drawer box together. Ignoring the quick and cheap options of biscuits and dowels, the typical way of joining the drawer sides, front, and back, together involves carcase joints - either finger joints, or dovetails. In western culture I observe a veritable fetish has developed around dovetailed drawers. Indeed, when I check out a cabinet, one of the first things I look at is the side of the drawer to see how it is attached. Dovetails, if well executed, are a good sign, but let's face it, there are only a sign and not the whole story. The larger furniture makers seem to know this piece well, as machined, gang-cut dovetails are fairly common on middle-grade furniture and up - at least on the joins between the sides and front of the drawer. The back of the drawer is often attached in an after-thought sort of manner, nail-gunned in there. A guess a lot of folks can't tell such a dovetail from a finely made one, or could care less anyhow. This is very similar to the bicycles one sees for sale in stores, where the manufacturers will often throw on a flashy handlebar and pimp rear derailleur, while skimping elsewhere in the important places where people just don't notice, like the spokes, bearings, hub spindles, etc. Just another aspect of our make-believe, stage-set material culture.

The typical well-constructed drawer will often be dovetailed at all corners. Vast battalions of woodchucks out there, obsessed with dovetailed drawer work, will hand cut the drawers so the pins are as slim as practicable, showing clearly, proudly and loudly, to a world that cares not, or is stubbornly unaware, that they cut their dovetails by hand. They may have machined the rest of the joints, but the dovetails are sacrosanct for some reason, and you're really letting the side down if you don't hand cut them. This is not something I believe personally. So long as the joint is well cut and fitted, and mechanically sound, that is that main thing to me, and if that result is obtained by many hours of contemplative sawing and chiseling, or blasted out with a Leigh jig, all good. Either way, the execution can range from 'flawless' to 'hack job'. I realize that for some out there these dovetail decisions are life-and-death, "I'll never speak to you again", sort of matters, but not here, where I have other obsessions, other fish to fry.

Dovetail variations include through-dovetails, blind dovetails, concealed mitered dovetails, and a few others. Often the sides of the drawer are made in a secondary wood like maple, oak, etc, which is light in color and given the darker wood for the drawer front provides maximal contrast to the dovetailed joint. This is to make extra certain that the world can see those dovetails - if they are paying any attention at all, which, most of the time, I would suspect they are not.

All that said - and I am poking a little bit of fun in my commentary above- a carcase dovetail is a strong and attractive joint. Well, it is if it is cut well. It is also a joint that, while having good mechanical interlock, is a glued connection. Indeed, given some of the popular practices of undercutting the iside faces of the tails and pins, the glue is often a large portion of the mechanical connection. 

I've been thinking about whether there might be another way to put a drawer together, but before I get into that I wanted to look at another aspect of drawer construction which is fairly significant - the floors, and how they connect to the rest of the drawer. The first choice is between solid wood floors and plywood floors. I'm going for solid floors. Plywood floors are very practical, easy to make, etc.,  a perfect compliment for the metal drawer slides - and I have similar reservations about how appropriate they might be to a 'traditional' piece. In order to deal with the solid wood movement issue, the drawer bottoms will be quartersawn material, and they will be attached to the rest of the drawer so that there is build-in accommodation for movement.

And how does one attach drawer bottoms to the rest of the drawer? Well, there are three methods I'm aware of:

  1. Nail and/or glue the drawer to the bottom of the drawer carcase.
  2. Cut a dado in the drawer side and front, cut a tongue on the panel, and slide it into position.
  3. Use a dadoed drawer slip.
Method 1 is commonly seen on Japanese tansu, affixed crosswise to the run of the drawer, and I just don't think it is a good way to do things. Usually the 'nails' are actually wooden pegs, other times, on cheaper pieces, they are metal. One advantage to this simple attachment method is that it maximizes available interior drawer space. Against that, however, even if one employs floor panels which do not move too much seasonally and cause a problem, eventually the nails corrode a bit, the glue bonds start to fail, and then the connection between floor and carcase weakens. One day you open the drawer and the bottom is starting to spearate away, with bits and piece getting jammed in the crevice. The owner then attempts to repair the floor separation with more glue and metal nails, often splitting the boards as they are driven in, and if these bits of metal stick out they can trash the runners inside the cabinet. Finally, the entire drawer runs in and out of the cabinet on its floor, which is a lot of surface area and friction. The drawers are typically quite light in weight so this friction is not a huge problem, but in a heavy wood it would be. I'll give that construction method a pass.

Method 2 is quite a common one, and a method I have used in the past. It can provide a strong connection between the floor and the rest of the drawer, and allows for wood movement, however it is not without its defects. The two principal defects are this: the dado in the drawer side weakens the lower portion of the drawer side significantly; and the raising of the floor panel up to fit in a dado means that a certain amount of potential interior space is surrendered.  One solution to the weakening of the drawer side from dadoing is to simply make the drawer side thicker, but in most cases one could anticipate a drawer side having to be in the range of 0.625"~0.75" to remain adequately strong, and that incurs a drawback of making the drawer heavier, both in actual weight and in visual terms, and the thicker sides mean potential interior space gets gobbled up. Many drawers built with dadoed floor panels also feature triangular glue blocks on the underside to reinforce a vulnerable area. You'll find these many years later sitting in the drawer below, having fallen off from one moisture cycle too many.

Another drawback to the commonly seen method of dadoing in the floor panel into the sides and front of the drawer is that the drawer then runs in and out of the cabinet on the surfaces of the drawer side lower edges alone. Taking a drawer in and out thousands of times, particularly if it gets loaded up heavily, entails all the wear and tear being concentrated on these relatively narrow strips, and something has to give in either the drawer's lower edge or the cabinets drawer runners, whichever is softer yielding soonest. Given the dado already weakening the lower portion of the drawer side as an issue, add in the wearing of that lower edge making the remaining mean left below the dado thinner and thinner, which means weaker and weaker.

A solution seen, as it turns out, in English high class cabinetry, is our number 3, drawer slips. Drawer slips are a piece of wood, dadoed to accept the floor panel, which are glued to the lower inside face of the drawer side. Here's an example a couple of different drawer slips, in cross-section view:


Some slips have an upper edge flush to the drawer floor, others are done so that the slip is proud of the floor. In either case, it is possible to profile the slip if so desired. 

A view of a slip attached to the inside of a drawer, all fitted together:


I trolled the internet for pictures of drawer slips, and found a short article on them on Popular Woodworking's website, which is where the above image originates.

Slips solve a lot of problems just by gluing a little strip into place. The eliminate the problem of weakening the drawer side from dadoing, and, because the dado is no longer in the drawer side, related problems with dovetail placement on the lower portion of the side are solved. The applied slips widen the effective lower edge of the drawer, and that means that the load is spread out over a greater area and thus wear and tear issues are greatly ameliorated. It's a clever solution.

The drawer slips are typically attached to the drawer sides, though some also affix a slip to the drawer front, which means that the slips could meet one another at a miter if so desired, perfect for those nosy house guests who not only want to inspect your drawer dovetails but want to pull the drawers out and examine the bottom as well.

I really like the drawer slip method, and if I were not so concerned with the glue minimization issue - if I wasn't so otaku about that - I imagine I would be perfectly happy making dovetailed drawers with mitered slips. I've been tossing and turning and scheming however to see if I could put a drawer together without using glue, with solid wood, and with joinery, and I think I may have hit upon a solution. I may also be barking mad - jury is out still. I'll share the delirium with you in the next post in this thread.

Thanks for coming by the Carpentry Way. I hope to see you next time 'round. On to post 5

12 comments:

  1. Chris,
    As always, I am enjoying your posts.

    I just had what I consider a brainwave, out of no-where. I could be delusional but I'll mention it none-the-less.
    As you mentioned, the bottom of the drawer is the issue here, especially when you dado for the drawer bottom. Since the drawer side effectively acts as a beam, the top in compression and the bottom in tension, the section of the drawer side does not need to be "traditional". The internal stresses at the top of the drawer side will be significantly less than those at the bottom. I was thinking, what if you made the section of the drawer side a trapezoid?
    I've never seen this before but it just makes sense that you would want the bottom to be substantially thicker than the top since the forces there are different. You could have a dado in a 3/4" (19mm) drawer side at the bottom, where you require strength, and that drawer side then tapers to anything you tike at the top of the drawer - 1/8" (3mm for the Aussie's here). Then I was thinking you could cut one single piece of rectilinear stock on any angle you deem appropriate (on a table saw/bandsaw) and get two drawer sides from the same piece!

    You would remove the need for gluing a slip, you would increase the strength of the dado, reduce weight with a tapered drawer side and I think it would look mighty sexy to boot. Sure that's a personal judgement.
    I am just throwing mud at the wall and seeing if I end up with a cob house...
    I think it's worth consideration anyway.
    Tom
    p.s. Additionally, if you wanted to go the dado route, you could implement some gussets into the dado so it didn't run as far unsupported (sorry, I have an engineering background, not sure what it might be called with timber). All you'd have to do would be to run a router along to create your dado, but stop halfway or however regular you want your gussets. I have a sample in the shop of this if you want a photo. I could email it though if you like. Then you would just notch out a little step at the gussets so the drawer base sits around each gusset. This would help support the dado even more. I'm getting pretty nitpicky now.

    ReplyDelete
    Replies
    1. Tom,

      yes, trapezoidal drawer side panels...Shakers came up with that one, so I guess great minds think alike!

      Check out: http://tonykonovaloff.com/?page_id=61

      Thanks as always for your comments and pleased to see my post got your brainwaves going big time!

      ~C

      Delete
  2. Hi, being an autodidact, drawer construction, like every other aspect of cabinetmaking, is a source of endless experimentation for me. I am always torn between the smooth, stout action of modern technology (drawer slides) and the more traditional methods you describe. One method I have not tried and don't see you mention it either is the NK method. Do you have any experience with this? If you don't, there is a story about it in FWW #150 if you're interested. Seems like a modified drawer slip/slide method. Anyway, keep up the good blogs,

    Dan

    ReplyDelete
    Replies
    1. Dan,

      thanks for your comment, and mention of the NK drawers. I used to have that issue of FWW, but I must have sold it or tossed it out some time ago, however I do dimly remember the NK drawer system.

      I just re-read that article, accessing it online, and my drawer construction idea does share some points with NK. Who's to say how long the germ of the thought from that old FWWW article may have been oscillating around in the dim reaches of my consciousness, only to come to flower in recent days?

      ~C

      Delete
  3. Chris,

    I wonder if you might be over emphasizing the weakening of the drawer sides with a dado, if done prudently? Your point of adding some thickness and consequently, weight, could be a concern on larger drawers. A lighter wood like Katsura or an equivalent, makes a good choice in terms of weight, for the sides and back. There is some wear factor. I can see facing the bottom of the drawer with a thin strip of harder wood that slides easily, if it doesn't conflict with your glue concerns.

    It seldom gets mentioned, but even a dovetailed drawer can often have a bit of play in it to go slightly out of square to the case. The only solution seems to be with the arrangement with the drawer bottom to solve that. A lot of Japanese chests have the bottom extending beneath and beyond the drawer back and acting as the stop. I don't recall ever seeing that method done elsewhere. It may not be the most attractive way, say as compared to the bottom completely enclosed, but it does make for easy adjustment when fitting.

    ReplyDelete
    Replies
    1. Dennis,

      thanks for chiming in. I agree that sizing drawer sides for the overall size of the drawer makes sense. Little drawers with 3/4" sides look goofy, and cavernous drawers with 3/8" sides look flimsy. The typical English trad. cabinet drawer with slips I'm talking about a fairly standard 12"~15" wide drawer - does have thinner sides than the typical American piece, which allows for a tad more internal volume and a bit more graceful look.

      I've seen western drawers with the bottom extended beyond the rest of the carcase, however not quite in the same vein as the Japanese drawers you mention. Drawer stops are another kettle of fish of course.... I have an idea percolating for a stop mechanism as well. I'm sure it's all been done before elsewhere but it is fun to at least feel one is in new territory, even if it is likely illusory.

      ~C

      Delete
  4. I like using dowel pins to fasten slips and, even occasionally, the dovetails top and bottom....

    ReplyDelete
    Replies
    1. Pchast,

      I appreciate your comment. I can see the sense in using dowels to help locate and reinforce the glued slip connections, but am less convinced there is any point in adding dowels to dovetails, as the primary glue surfaces are on the inside faces of pins and tails. Here I'm presuming that the dovetail joints are tightly fitted. A dowel would have less than full strength being glued through and into the joint, as close to 50% of the glue are in the dowel hole would be end grain on either board. I guess it adds a little bit of strength, just not sure it is worth the added effort for what little is gained. If you ask me, it makes more sense to slim down and increase the number of dovetails towards the outside corners of any dovetail joint, as this increases surface area, improving mechanical interlock and glue-able surfaces.

      ~C

      Delete
  5. Hi Chris, I've been reading your blog for a few months, and have really been digging your recent posts, particularly the extreme attention to detail in this mizuya design. I also really liked your discussion of the commercialization of live edge slabs a few posts back. So, just wanted to say keep up the good work, and Thanks!

    ReplyDelete
    Replies
    1. Lars,

      great to hear from you and thanks for your kind words. I hope you'll keep coming back for more!

      ~C

      Delete
  6. Hi Chris Another option you might consider for the drawer joinery, since you don't seem to care if you can see the dovetails,is a sliding dovetail to join the sides to the drawer front and back. You could pin it with a "wooden nail" from the side to lock the joint. I have enjoyed following along with the design process of this project, looking forward to see to solutions you pick.
    -Erik

    ReplyDelete
    Replies
    1. Erik,

      your comment is appreciated, and I think you'll find my solution to drawer construction interesting. Glad you're enjoying the account of the design phase of this build!

      ~C

      Delete

All comments are moderated. Spamming and comments containing links unrelated to blog content will be deleted.