Finishing Steps, and the Completed Table

For me, finishing is like sharpening. There’s a lot of ways to do it, and I often think it’s best to pick one or two and stick with them.  I think you’re likely to produce better results that way than trying to tailor your finish to each individual piece, rather than practicing one or two different ones.

In my case, I use three different finishes:

Oil only: 6-7 coats of hand-rubbed danish oil (which is actually an oil/varnish blend. I use Watco Danish Oil.

Buffed: one coat of oil, followed by a pass through the Beall Wood Buffer system. This isn’t a furniture finish, so it doesn’t really count. Only usable for small, fairly sturdy items.

Lacquer: This is my standard furniture finish. I believe it was described by the Danish master Tage Frid. It’s almost foolproof, durable, and easy to apply. The order is:

  • Sand to 220 grit
  • Danish oil: 1 heavy coat. Dry 24 hours. Scuff-sand with 220 grit.
  • Sanding sealer: 1 coat. You must use Lacquer-based sanding sealer. Dry 2 hours. Sand with 220 grit (remove the shine). Wipe very clean.
  • Lacquer: 1st coat. Dry 3 hours. Sand 320 grit (remove the shine). Wipe very clean.
  • Lacquer: 2nd coat. Dry 3 hours. Sand as above.
  • Lacquer: 3rd coat. Dry 3 hours. Sand as above.
  • Lacquer: 4th coat (optional). Dry 24 hours.
  • Lightly rub with 0000 steel wool, and buff with paste wax.

There’s not too much that can go wrong with it.  I don’t own a HVLP sprayer; I don’t do enough work to make it a reasonable investment.  Good ones aren’t cheap.  I use Deft Spray Lacquer, by the can. It works beautifully; it’s just not as cost-effective.  Even coverage with spray takes some practice. It’s important to keep the nozzle facing the work at the same angle as much as possible, moving back and forth at a consistent distance from the work.  It’s easy to catch yourself swinging your arm in an arc, which will cause both drips and runs in the center of the arc, and spotty coverage at the edges.

This finish takes three days to complete.  I have yet to have a major problem with it, and it’s about all I use on furniture at this point. It makes an even, clear, medium-gloss finish that’s pretty durable, and suitable for about any indoor furniture.

Of course, nothing ever goes quite as predicted.

Everything was fine until It was done – and I slid the drawer into place. Or rather, tried to.  I used figure-8 fasteners to attach the table top, to allow for expansion room for humidity changes.  If you don’t the table can literally tear itself apart. Whatever method you use to attach the top has to be able to shift slightly.

The figure-8 fasteners attach in small recesses in the sides and back of the top of the aprons in one hole, and to the top through their other hole.  The problem (completely unexpected, of course) was that the fasteners are about 1/8″ thick, and there wasn’t clearance below them to allow the drawer to close.  That one, I didn’t see coming.

The answer was to remove the top and the fasteners, flatten one end of them in a vise (they were stamped metal, not solid), and drill a slight recess in the top with a forstner bit.  Then, when attached to the top, there was enough clearance for the drawer.

I tapped the pegs into place on the stretcher, and it was finished. It looks very much like I had visualized in its new home.


This was an amazing project. In many ways, it was the most difficult thing I’ve ever attempted – the difficulty of the hand-cut joinery, my first real dovetails, the angles of the casework, the copper fills, and the hardness and unpredictable grain of the wood all combined to make it a real challenge. But almost four months after it started, I’m a happy camper.

Thanks to those of you that have followed along on this little odyssey.  Originally, I started this blog for this particular project, but it seems to have grown into more than that.  So I’m going to continue, with further projects as they come along. Next will probably be a couple of small band-sawn jewelry boxes for a friend of mine, and then I’m planning to start a Roubo workbench.   But right now, I have a lot of resharpening to do.

Stay tuned. Same Bat-time, same Bat-channel.

Ross Henton
Frisco, Texas
April 2010

Advertisements

Nobody’s perfect, and neither is their work

Almost every project requires some minor touchup when it’s assembled. This is handwork, after all – imperfections are part of it, and what sets it apart from machine work.  The goal is to minimize the amount of corrections and fixes that have to be done.  In that respect this was pretty successful. But now that it’s assembled, I’ve found a couple of places that need minor fixes.

Most of this sort of work is pretty simple.  For example, one of the lower rails had a small nick where it joined the leg, less than 2mm across. The solution was to take a chisel and cut a sliver about the size of the end of a toothpick from a piece of scrap mesquite. I dipped the end in some cyanoacrylate glue (super glue, if you get it at Wally-World), and wedged it in the nick.  Once it set, I carved it flush with a chisel, and sanded over it. Viola, invisible repair.

A note about the glue: I really, really like the Titebond CA glues.  I don’t think there’s much difference between them and the competitive brands – except the bottle. Which is much less prone to clogging and mess. The gel grade is really thick, the medium grade is what I use most of, and the thin grade is wonderful for small repairs.

Mesquite is prone to a lot of surface variation… cracks, knots, holes, and such. I took a couple of small cracks, added a little of the medium-grade CA glue with a toothpick, and sanded over it while still wet with a piece of 22o-grit sandpaper. The dust bonds with the glue, and fills the crack almost invisibly.  I’ve used this particular technique a lot; occasionally using a random orbital sander instead of sanding by hand.

The one serious problem I found was a gap along the edge of the rear apron.  I think the tenon was probably miscut at a slight angle, and although it looked fine when doing the test fitting, squaring the assembly inside the clamps probably pulled it open just a hair. There are a couple of ways of dealing with this (tearing the table apart is not one of them).

Option 1: cut a thin, tapering wedge the length of the gap, brush it with glue, tap it in place, and shave it off.  For gaps this size, a narrow slice of veneer works even better. It usually looks good, but not in this case – because the gap runs along the cross-grain of the apron, and the wedge would almost have to be made lengthwise to the grain, or it would be too fragile to handle.

Option 2: Fill it with wood filler. And no, I’m not talking about any of the grades of commercial filler. Minwax, Plastic Wood, Elmer’s, Famowood, any of them. They’re all terrible. Even the ones that insist that they’re “stainable” really aren’t. The colored ones won’t match your wood. They don’t take stain well. Don’t even try.Can you see it?

Make your own. All you have to do is save some of the 220-grit sanding dust from the wood you’re working on, and mix it with a few drops of the first finish you plan to apply. In this case, that’s danish oil.  It took about three drops and a quarter-teaspoon of dust to make the paste, and I patched into the crack with a toothpick. It sands flat (but requires a light touch), will always stain or finish to match the wood, and is nearly invisible. Note that this is not for structural repairs – only small cosmetic repairs.  The repair took about 10 minutes, and I doubt anyone but me will ever know it’s there.

The Copper Filler Odyssey

My first attempts at the copper fill I intended to do for the voids in the mesquite were a resounding failure.  I really didn’t think it would be difficult; just melt some copper rod with a blowtorch, and let it drip into the void.  I couldn’t have been more wrong.

The problem is that copper is a terrific conductor – both of electricity and of heat.  I hit the end of a 1/8″ copper rod with a MAPP gas torch for almost 15 minutes, and it didn’t even soften appreciably.   It turns out that copper is usually melted in a crucible, at about 1800 degrees. And I didn’t happen to have an industrial ceramic crucible lying around the garage.

So, plan B: I got some 1500 mesh copper powder and mixed it into some System Three epoxy.  My first results were lousy; I didn’t get enough powder mixed into it, and the results looked like grayish plastic.  The second attempt was more powder than epoxy, and made a very stiff mixture.  When dried, it still looked awful – but when I scraped it flat, it looked better.  Sanding didn’t improve it much, until I realized that it had to be sanded to a much finer grit than I generally use in woodworking.  Instead of stopping at 220, I sanded it up through 1000 grit – and all of a sudden, it looked right.

I mixed up a larger batch (much larger than I actually needed, as it turned out), and packed it into all the open knots and voids on the table.  After about 24 hours setting time – enough to be hard, but not completely cured – I flattened it with a plane (yes, it dulled the blade) and a card scraper, and polished it up to 1000 grit. It looks exactly like I had visualized.

The Drawer Pull Shaping the pull

The last piece to make was the drawer pull.  I glued up a small block of mesquite, cut the rough shape on the band saw, and shaped it with a rasp and sandpaper.  It was a simple process, but when I wasn’t at all sure how it would look when I started.  It turned out to be pretty easy, but kind of a fiddly, picky process to get it shaped the way I wanted.  I cut a small hole in a piece of soft pine, and set it in a vise with the crown of the knob in the hole, so I could brace it securely and not deform the wood of the knob.  It was mounted to the front of the drawer with a brass screw, with the hole countersunk on the inside of the drawer to accommodate the screw head. The finished pull

Next time: Final assembly and finishing.

The Rubber Meets the Road: The Drawer Joinery

And here we are.  There’s no putting it off any longer.  After weeks of occasional practice at cutting dovetails, it’s finally time to do this for real.  A lot of my satisfaction with this project hinges on how the drawer construction turns out.  These are the first hand-cut dovetails I’ve done as anything other than practice. I don’t intend to belabor all the details of the construction; there’s probably more written about the subject of cutting dovetails than any other topic in woodworking.

Pins-first or tails-first? That’s one of the most argued topics.  I learned pins-first, and I’m sticking with it. First step was to take a marking gauge, set it to the thickness of the drawer side,  and mark layout lines on the inside face (only) of the drawer front.  Remember: the front dovetail is half-blind. It doesn’t go all the way through the board, so it’s not visible on the front.  Then I cut the ends of the drawer sides to the required 3 1/2 degree angle (the last cut like this I’ll be making for this project), and inscribed the layout line on all four sides of both ends of the sides and the drawer back. Pin layouts (pencil)

Then I clamped the mesquite face of the drawer vertically in the vise, and drew the layout of the four pins for the half-blind dovetails with a pencil.   Pin layouts (knife)I started with a pencil and not a marking knife, so I could see how the layout looked to my eye before committing myself to the irreversible step of cutting in the marking lines.  This was a good thing – because I originally intended to do the spacing slightly differently, and I didn’t like the way it looked. So I cleaned the pencil lines off with some denatured alcohol (works better than an eraser), and started over. The second pass looked much better.  I went back with my little 8-degree gauge (the angle of the dovetails), inscribed the pencil layout lines, and carried them down to the layout lines in front.

Then I walked away for about half an hour, started some laundry, sent a couple of emails, and played with Angus while I worked up my nerve.

There comes a time when you just sigh and start cutting the wood.  This was it. The angles of the dovetails were cut out with my little Veritas joinery saw (which I dearly love for detail work).  Came out pretty well, with one slight overcut about 2mm long in one place.  I couldn’t use the dovetail jig I described in an earlier post, because these cuts weren’t square to the face of the wood, but at a downward angle.Half-Blind Dovetails

I clamped the front of the drawer down to the bench, and chopped out the first set.  Because the dovetails are wider at the bottom, I had to be extremely careful to not damage the upper edge of each one as I pried chips out, or to ding the edges with the chisel.  I switched to a narrower chisel, so I could angle it slightly into the cuts. The first set actually looked pretty good. I switched to the other side, repeated the layout, and discovered that my chisels were too dull to cut butter, much less mesquite.  The heavy downward chopping had trashed the edges of all three chisels I used, practically folding them over in a couple of cases. So I had to stop and resharpen.Through Dovetails

The second set went about like the first, with comparable results.  Then I clamped the first side down, placed a 1-2-3 block against the face of the drawer front, and set both across the end of the first drawer side.  I transferred the angles of the pins to the drawer side with a sharp pencil, unclamped it, and carried the lines around the other side of the tails.  Cutting the tails is always where I get into trouble.  This is one instance where practice paid off.  They weren’t pristinely perfect, but looked pretty good. They fit into the first pins with only minimal paring.  I also rubbed some pencil lead on the inside of the pins, and started tapping them together until they started to jam.  The transfer of pencil lead to the tail board showed me where the high areas were that needed cutting.  Switching to the lyptus for the inside was an incredible relief from the hard mesquite – in fact, chopping one set of the pins for the drawer front took longer than all the rest of the drawer joinery combined.

Once the tails for the front were finished, I switched to the back.  The back corners are simpler, 3-pin through dovetails. One set came out perfectly; it tapped into place with no gaps or irregularities, and no paring or trimming. That was a really, really good moment.  The second set wasn’t as good – I had missed a cutline, and had a gap that would have to be patched later.  Fortunately, that’s not terribly difficult, and will look fine when it’s finished.

Groove for drawer bottomA 1/4″ recess for the drawer bottom had to be cut in each piece.  The best way to align it was between the lowest two pins in the drawer front – that way, the groove isn’t visible from the front of the drawer.  This was a simple operation – because I did not attempt to do it by hand. I did it in two 1/8″ passes on the router. There are hand tools specifically for cutting grooves like this, but it happens I don’t own any of them, and the thought of attempting to do it with a dovetail saw and a chisel just didn’t do it for me. It probably would have been a disaster.

Once completed, I dry-assembled the drawer. The bottom is two loose panels, with a small half-lap groove halfway through the thickness of each panel where they join, so you won’t ever be able to see light between them if it contracts.  The fit of the bottom was deliberately slightly loose, because I don’t want the drawer to tear itself apart if the lyptus panels expand in the humidity. Problem is, although they have expansion room, they rattled. To stop the rattle, I cut a couple of small strips from a rubber bracelet and set them in one of the drawer grooves. It stopped the rattling, and the drawer bottom can expand into the groove against the rubber if needed.  You can’t glue the bottom in place – expansion of the wood in the bottom will literally tear the drawer apart. The same goes for the top of the table; when it’s mounted, it has to be attached with allowances for changing humidity.Clamped

I knocked the assembly apart again, and brushed glue on the tails and into the pin sockets.  The drawer was clamped up in parallel-jaw clamps, with two angled blocks faced with sandpaper (saved from the base assembly) against the drawer front.  I measured the diagonal angles of the drawer, and almost fell over in shock – they were the same to within about 1/32 of an inch. Not bad.

I let the drawer dry overnight, and turned my attention to some touch-up work on the legs. I had missed some machine marks when I sanded the legs.  The easiest way to remove them is with the best-of-all finish tools – the humble card scraper.  Really just a rectangular piece of metal, sharpened with a burr along one edge. It takes minimal time to learn to sharpen and handle, and takes wonderful thin shavings.  I’ve found that placing a refrigerator magnet on one side helps stop the transfer of heat to your thumbs, and makes working with it more comfortable. In about 5 minutes, I had the areas I had missed on the legs completely cleaned up.Scraper shavings

Next morning, I took the drawer out of clamps.  There was some touchup and cleanup to do in the dovetails, but not much (more on that next time).  I clamped down the drawer, and planed the top and bottom of the drawer front down to the required 3 1/2 degree angle followed by the legs.  I clamped the frame for the drawer rails into the base, placed the drawer in position, and adjusted the frame up and down slightly until the space above and below the drawer front was where I wanted it, squared to rails to the top of the aprons with my Odd Job (another great tool) and shifted the clamps holding the rails in place. I was originally going to glue the frame in place, but I didn’t want to risk getting glue smeared around on the lower part of the top aprons… and this gave me a chance to use a new toy.  My family got me a Grex headless pinner for Christmas this year, and I’ve been itching to use it. It’s wonderful; I had the rails firmly attached in about 30 seconds.  I added two strips along the sides of the rails, against the aprons to prevent the drawer shifting laterally as it moves. Then I attached a stop block to the rear of the frame so it doesn’t try and slide in too far.

SuccessAll in all, I’m happy with how this part went. Next is touch up and minor repairs, finishing the top, the copper fills, making the knob, and final assembly and finishing.  Stay tuned… this is going to be fun.

Graduation Day – and new lessons

“I love it when a plan comes together.”

–Hannibal Smith, The A-Team

I started out in woodworking with power tools.  Since then, I’ve been working my way backwards… slowly incorporating more hand tools into the process.  I find it relaxing, often more precise, and producing (if not better) somewhat different results.  A year or so ago, I took Howard Hale’s excellent Hand-Cut Dovetails class at our local Woodcraft.  One of the things he said that stuck in my mind was, yes: machine-cut dovetails are easier, and more precise. They usually take as long (or longer) to do, because of the machine setup required. Machine-cut dovetails will look exactly like they’re from any piece of nice furniture from Haverty’s. Yes, hand-cut dovetails will be slightly irregular by comparison. It will clearly be hand work. And that’s why it costs five times as much.

I realized when I started today (the first of two full back-to-back days in the shop, which is an incredible luxury for me – thanks, Cathrine!) that I had miscut the back apron.  I made one of the first mistakes in furniture making, and one that I’ve written about in the past.  I measured something.

Today’s Lesson #1: Not a new one, but worth reiterating. Don’t measure stuff. Just don’t. Mark the pieces from each other after the first one is cut.

The distance between the tenons for the back apron was off from the front and back stretchers by about 1/8″.  There was nothing to do but remake the piece, or lengthen the tenons on everything else. Much easier to recut just the one.  The hitch was that when I thicknessed all the parts, I didn’t leave any extra – and that meant changing machine setups in my small shop (again).

Today’s Lesson #2: When you’re thicknessing lumber from raw stock, do extra.

I had a piece of stock about the right size – an inch or two big in all dimensions – and flat on one side. I could move the dust collector hose, reset the planer to the right thickness, plane it flat, move to the miter saw, cut it to length, change the dust collector gates to the table saw, cut it to width, switch the dust collector gates to the jointer, and joint the edge.

Instead – without thinking, this time – I cut it to width with a small joinery saw and a bench hook, ripsawed it to width with a bowsaw, planed it flat, and squared the edges with my Veritas edge-trimming plane. Elapsed time: 8 minutes. No noise,  no hoses to move. It didn’t run Angus out of the shop because of the racket. No stress.  No mess. Just some curls and shavings. Then I realized that I had just done this process without thinking about it… I had moved from power-tools first to hand-tools first for this (admittedly simple) operation.  For me, that was Graduation Day.

To avoid repeating this same error, I marked the tenons on the front drawer rail directly from the front lower rail.  I aligned the two pieces in a bench ho0k to do the marking, and then marked and cut the tenons for the new back apron in the same way. This led to…

Today’s Lesson #3: Bench hooks aren’t just for sawing. They’re perfect for bracing your work for almost any reason – like making clean, straight layout lines.  Much better than either clamping work in a vise, which tends to put things at uncomfortable visual angles, or (worse yet) juggling them around in the air or braced against your body.

Once the back apron was finished, it was time to cut the curves in the front and back rails.  On this one, I decided to cheat, and go back to power tools.  I was originally going to try to cut the curves by hand with a coping saw.  The blade on my bowsaw is too wide to cut these curves, and I didn’t really have another hand tool that would work well for this. Lots of stuff for cutting joinery, less for cutting curves.  But due to the hardness of the wood, and the need for some accuracy here, I cut the curves on the band saw instead.  I marked the line with the curved edge of a serving platter we had in the house, made the cuts, and cleaned them up with a small spokeshave and a piece of 120-grit adhesive sandpaper attached to one of the cutoff pieces.

Then the plan started to come together: I assembled the front and back sections. I scraped and sanded everything first, and knocked down the sharp edges with a block plane and a piece of sandpaper. The simple card scraper is still the best finishing tool ever invented.  I don’t remember where I read this… but if you put a refrigerator magnet on the scraper, it’ll stop a lot of the heat transfer to your thumbs, and make its use a lot more comfortable.

It was a pretty simple glueup, and came out straight the first time.  The tenons mated well, except in one place in the back I found (too late) to have a misaligned shoulder that’s going to take some repair when I’m finished.

This was working well… and was time to move to a more difficult piece: the lower stretcher.

The through tenons are fairly simple pieces; the challenge is that the shoulders of the tenons have to match the 3 1/2 degree angle of the front and back legs where they meet the rails.  The problem is that there’s no way to calculate the dimensions of the distance between the tenons, due to the angle of the front and back legs.  The distance between the lower ends of the stretcher and between the upper ends is different… and this is a pretty critical measurement.  And there’s no way to test it except to assemble it – and you can’t until both tenons are cut. At first, I thought I’d have to cut the shoulders too long, and trim to fit – and the though of that in this hard mesquite gave me the cold shudders.

The answer was to violate lesson #1: go ahead and measure it. I dry-fitted the leg sections and the side rails together without the stretcher, and then used a steel rule to measure the distance between the upper edge of the mortises in the lower rails. I started the measurement at the 1″ mark in the back (to make it easier to align the ruler), measured it at 19 9/16″, and subtracted the 1″ offset from the other end (don’t forget that! It’s easy to forget.), so the distance between the upper edge of the tenons is 18 9/16″.  Once the angle is transferred to the sides, the lower edges will line up correctly. I knocked the assembly apart again.  And yes, you’re correct: in the photo, you can see a small tearout at one end of the mortise – but it’s going to be inside the shoulder of the tenon, and completely hidden by it.

I marked the tenons the correct distance apart for the top side, and then marked the sides of the tenons outward with the same angle jig I made when I started the project. Then I carried the line across the bottom at the base of the new angle.  I’ve found the easiest way to do that is to put the edge of a marking knife in the cut, and slide a square up against it to position it for the marking. I chiseled out the edge of the tenons first (see my previous post), and used the jig to align the saw prior to starting the cuts.

Today’s Lesson #4: Don’t Panic. Lots of mishaps are repairable, and will be absolutely fine with a little extra work.

If you don’t have the book Fixing and Avoiding Woodworking Mistakes, by Szandor Nagyszalanczy, I highly recommend it.  it’s where I learned the basis of what saved me today.

I had cut the stretcher tenons slightly oversized, so I could trim them to fit tightly into the mortises on the rails. As I was cleaning up the edges of the tenons on the with a chisel, I knocked a large chip out of one of the tenons. Of course, I could have just glued it back in place – but it flew off into a big pile of dust and shavings, and I never did find it. It could also have shattered, but was unrecoverable, in any case.  There were two causes of this mishap:

1) the tenon is an area of wild, turning grain.

2) I got careless.

This wasn’t just a little ding that can be filled, or a tearout in a place that’s hidden behind a tenon shoulder: it was the outside part of the front through-tenon on the stretcher. The answer was to cut it out slightly further until the sides were square, glue in a small diagonal piece of square scrap with cyanoacrylate glue, and trim it even with a block plane. Once the finish is applied, I think even I will be very hard pressed to find it.

The project instructions state that the author thought the best way to cut the holes in the tenons is with a “single stroke” of a hollow-chisel mortiser. I agree; it’s the right tool, and worked beautifully. But “single stroke” means he was working with oak, not mesquite. It took somewhat more than that, but did an excellent job. I reassembled the pieces, marked the edges where the holes should start, and then shifted both holes inward slightly, so that the pegs will pull the rails tight against the stretcher. The pins were cut out of a piece of scrap, and shaved to the correct angle with a block plane.

I knocked it all down again, assembled it with the pegs and checked the fit, and then knocked it apart for the last time. The clamps have to pull the angled frame tight, so I took a minute and cut some scrap to the correct angle and attached adhesive sandpaper along the angled face, to use as clamping blocks. I brushed glue on the tenons on the upper aprons, and assembled the frame. Five pieces this time: the upper rails, the stretcher, and the front and back leg assemblies. The blocks keep the clamps pulling straight across the angled frame.

I checked the diagonal distances to make sure it was square (it was) and set it aside. While the assembly was drying, I turned my attention to a couple of other things.

My first attempt at making the copper/resin filler for the voids was unsuccessful.  It came out kind of a dirty brown.  This time, I tried the same thing with a much higher proportion of powdered copper. It seems to have worked much better; the surface of the fill is almost iridescent. The epoxy is terribly slow to dry – it takes about 6 hours to lose its tack, and at least two full days to set completely.  I’m going to try a couple of more mixtures with some 5-minute epoxy, in hopes of not having to wait two days to see if it’s going to work or not.

Next was the frame that supports the drawer. It’s poplar, which is like handling balsa wood after the hardness of the mesquite.  The original plans called for the frame to be joined together with mortise and tenons. But honestly, I think that’s overkill.  It’s not supporting much weight, and was an added degree of complexity that doesn’t add anything much to the project. So my inner frame is half-lap joinery – essentially, two facing one-sided tenons, glued into an overlap at the corners. This photo shows it set into place, but it hasn’t been glued in yet – I’m going to hold off until the drawer construction is finished.

Before taking the framework out of the clamps, I tapped the pegs into place, marked them 3/4″ above and below the tenons, and trimmed them to fit. Then I tapped them back in, and took the assembly out of the clamps.

It’s starting to look like it might be a table when it grows up.

From here, the next major step is to build the drawer.  That means actually cutting the dovetails I’ve been practicing for real (no, I’m not at all nervous about it, why do you ask?). Then mount the drawer frame and rails, make the knob, and finish the top.

Meanwhile, Angus says the shop is a huge mess, and I need to stop and clean it up. Who am I to argue?

What a great weekend.

Next time: The Drawer Joinery.

The World’s Oldest Joint

I decided to re-examine the problem of cutting the mortises on the lower aprons on the mortising machine.  As I said in my previous post, the mesquite is so tough that the 1/2″ chisel just wouldn’t cut through the piece without burning and jamming.  That bothered me: mesquite is tough, but the steel of the mortising chisels is tougher. It should have cut without so much of a fight.

I examined the 1/4″ chisel that I used to cut the leg mortises, and all four corners were blunted.  Sharpening the chisels requires a very specialized tool, available from Woodcraft, Rockler, and similar places). I had to go ahead and shell out for one.  Fortunately, it’s fast to use, and works extremely well.

I decided to touch up the 1/2″ chisel at the same time, and I discovered that it had become clogged badly – no wonder it wouldn’t cut properly. After removing the clogs and touching it up, I tried the apron mortises again, and it cut well enough. Still slow, but likely better results than I would have gotten with a coping saw.

The other half of the joint is the tenons.  Mortise and tenon joinery is very strong, and one of the world’s oldest construction techniques.  It appears in finds thousands of years old, including the Khufu Ship, which dates to about 2500 BC.  Don’t let that surprise you; the Egyptians knew a lot about woodworking that we still use,  including this type of joinery, the bow saw, intarsia, barrel hinges, and plywood. Kari Hultman’s wonderful blog, The Village Carpenter, discusses ancient Egyptian woodworking at some length.

The tenon is a stub that protrudes from the end of the (usually) lateral pieces, and inserts in the mortise. The purpose is to present surface grain to be glued into the mortise.  End grain will not take glue at all; it just absorbs it, and you can snap it apart with your fingers.  A properly glued mortise and tenon joint is stronger than the original wood, and the stress that would cause it to fail will usually rip the wood completely apart. Note:

The tricky part of the tenon is the shoulder, where the surface of the apron meets the leg.  It has to be smooth, square to the tenon itself (called the “cheek”), even on both sides, and free of chip-out.  I’ve cut this kind of joint before, but only on the table saw and band saw. This is my first excursion cutting them by hand.

The cut line has to follow the 3 1/2″ angle of the legs, but the angle gauge I discussed in a previous post isn’t big enough <sigh>.  I made it about 3″ deep, but these are 5″ aprons. Yes, that’s really annoying.  The solution was to use the angle gauge to set my bevel gauge, and mark the cut lines with a marking knife.

The cut line follows the apron all the way around.  The lateral sides of the tenon are marked the same way, and run down the apron to the crosscut line of the shoulder.  Then the crosscut is made with a small joinery saw, to 1/3 of the thickness of the apron.

I’ve learned a lot of my woodworking by reading, and this was no exception.  I read several things on the subject before attempting these cuts, and I’m glad I did. I found one particular technique that improved the results greatly.  I started the cut line across the tenon with a chisel first – much like starting the initial chisel cuts in a dovetail joint.  That creates a sharp shoulder edge to start the saw against, and keeps the initial crosscut of the wood fibers clean. Then, because the saw starts slightly below the face of the wood, the sharp chisel line is what shows against the leg. Once that’s done, the cheeks of the tenon are cut, and the ends trimmed to the correct width. The flat surfaces of the tenon are trimmed square with a shoulder plane.

I had four of these to cut, two for each of the side aprons.  Honestly, my first attempt wasn’t all that great. The fit in the mortise was too loose. The second was better, but one of the shoulder cuts had wandered slightly, and the two shoulders weren’t quite even. The third was better yet, and the fourth was close to perfect.

Fixing the loose tenon was easy – all you have to do is glue a small piece of veneer to one side of the tenon (the side that was cut too deep) and plane it down gently until it’s tight. The fix for the second one was more difficult, because both sides have to be exactly the same length between the tenons – so after squaring the edge on one, the other had to be brought down slightly with a shoulder plane to match. But when that fourth tenon tapped perfectly into place with a mallet, that made my day.  These appear to be square, and will glue up tight.

Each adjoining face is marked, so that I know which tenon was fitted to which mortise.

Cutting the tenons was labor intensive, but felt really good. Slow and clean was better than fast and full of dust. When it was done, I had a small pile of sawdust and a handful of shavings – not a face full of choking dust, and I spent the day cutting wood instead of setting up the table saw. Ask me which one I like better.

Next time: Tenons for the back apron and drawer stile, and starting the curved lower aprons. And (maybe) back to the top, to test out the copper filler.

Chopping Holes in the Table

One of the best things in life is a wife that says, “Why don’t you take the day in the shop today,” on Valentine’s Day. There’s not much else I can say about that. I’m a lucky guy.

——————————

After consideration, I decided to stick with lyptus for the secondary wood. I haven’t tried to cut dovetails in it yet, but it can’t be much worse than the hard maple I started with. All the pieces are cut and dimensioned, including two 1/4″ panels for the drawer bottom. The lyptus machined extremely well, and I like the way it looks with the mesquite. A few pieces still have to be cut to final length, and they will actually be cut to fit the openings to ensure that they’re tight.

The first step in cutting the joinery is cutting the legs to length, and establishing the top and bottom angles. At this point, the legs are identical blanks – same size and dimensions. But that’s about to come to an end, and this is where you have to start being really careful. I can’t overemphasize this: label your parts clearly. Each leg is different. The two back legs have one mortise in the front, and two on the inside, facing each other. The front legs have one mortise in the back, one on the inside, and a second 1/4″ square mortise towards the top for the front drawer support. Because the legs are angled outward front and back, the placement of each mortise is crucial. If I get two of my legs switched, I’ll ruin a leg by cutting a mortise in the wrong place. That means squaring more wood, and cutting and gluing up another leg blank.

The legs and other mesquite parts are labeled in white crayon. It’s much easier to read than darker pencil lines against the dark wood. Each leg is labeled on its outside front- or back-oriented face, with it’s left/right position, and an arrow pointing out the top. It’s extremely easy to cut something on the wrong end of a leg, which is just as disastrous as the wrong face.

The first angled cuts (on the top and bottoms of the legs) are made by hand, with the guide block I discussed previously. I want these cuts to be fine and as free of splinters as possible, so the cuts are made with a Japanese dozuki saw – an extremely fine crosscut saw. Its teeth are somewhat fragile, and can’t be resharpened. I was a little concerned about damaging them in the hard mesquite, but they handled it nicely.

The leg and apron joinery is mortise-and-tenon. If you’re not familiar, it’s basically a tab-and-slot system. The holes are called mortises, and can be cut several different ways. My intention for this project was to do as much of the work by hand as possible… but that resolution just failed.

Hand chopping mortises is difficult, difficult, slow work. We get used to the speed and ease of work with power tools, and forget sometimes that centuries of beautiful furniture was cut strictly by hand. Mortises are cut by hand with a mallet and a heavy-bladed mortising chisel. Mine works fine, and I’ve done it before – but never in wood as hard as mesquite. I started cutting a mortise on a test piece to see how it went, and I quickly discovered that I’d be working all week to cut even these small 1/4″ mortises by hand. The wood is just so tough that it becomes an incredibly labor-intensive job. In this case, the ease of work provided by power tools wins out.

The mortises are laid out with pencil marks, with the area to be removed marked with X’s. This is pretty crucial markup, because although the lower aprons are centered on the legs, the upper aprons are offset. The positioning of the offset of the aprons is not specified anywhere in the plans… which means that the lower aprons have to be cut to fit once the upper dimensions are established. These plans have turned out to be rather sketchy, in my opinion. There are several places I’ve found where dimensions aren’t specified, offsets aren’t indicated, and so on. In this case, I’ve worked backwards from the thickness of the aprons, the thickness of the tenons, and made an educated guess about the best offset.

There are several different ways to cut mortises by machine. I use a mortising attachment for my (fairly cheapo) Delta drill press. The cutters are best described as four-sided chisels, with a center auger bit that cleans out the bulk of the waste. Once it’s attached and aligned, it makes short work of cutting the mortise. The drill press cuts downward to the depth set, and the waste comes out the side of the hollow chisel. It takes about eight or ten strokes to cut the mortise, and the sides are pretty smooth. A little cleanup with a file and chisel is all that’s necessary.

The mortises for the side aprons are cut to the same angle as the legs. In this case, there’s not much way to use the guide block I made to align the mortising tool, so the legs are supported by a small piece of scrap wood which determines the angle of the leg on the machine. The scrap is clamped in piece, and the leg slides across it when it moves to maintain the cutting angle.

Side note: Handling the legs and realigning them for the mortises started to get painful. The side edges of the legs are extremely sharp, and the hard mesquite was starting to tear my hands up. Two or three minutes with an apron plane softened the very sharp corners so they were easier to handle – which is a step I was going to have to do sooner or later anyway, and doesn’t affect the layout of the joints in any way.

Once the side mortises are cut, the mortises for the back apron are cut into the back legs, on an adjacent side; then the lower mortises are laid out and cut. The 1/4″ square hole for the drawer support is cut using the same technique. Since the lower apron and back apron are parallel to the legs, the scrap support to establish the angle isn’t necessary. Once the mortises were finished, I examined the chisel used to cut them. Surprise, surprise… it’s completely trashed. All four chisel points are dull. That means it’s going to have to be resharpened before it can be used again. I don’t have a sharpener for these – it requires a cone-shaped sharpening tool to get the inside of the chisel. Looks like I’m going to have to order one. This is tough wood.

The lower aprons have a 1/2″ x 2″ through tenon, secured with pins in the front and back. It also has to be cut at the angle of the legs, so I decided to use the 1/2″ mortising chisel on the machine to cut those as well. That’s where things started to go wrong.

Problem #1: It’s too flippin’ tough. The 1/2″ chisel won’t cut through the wood like the 1/4″ chisel does. It jammed… it tried to smoke and burn the wood… it took forever to get just the first of the four cuts made for one of the two aprons.

Problem #2: Because this mortise goes all the way through the wood, it needed to have some support for the back when the chisel came through. I should have backed up the wood with another board, to prevent tearout. When the chisel finally came through, it tore out a piece about the size of my thumbnail adjacent to the chisel. Fortunately, it won’t show – the tearout was all in the area that will be cut away for the through tenon. But I’m going to have to watch that; it’s a mistake I’ve made before.

Problem #3: When the wood started burning, I saw a tiny hot coal go into the dust collector. That’s bad. The odds of it starting a fire in the dust collector bag are very, very small, but I won’t take chances. I don’t think I want to explain to the wife-unit how I managed to start a huge, explosive dust fire in the attic on Valentine’s Day. But that’s okay; I needed to empty the bag anyway.

The second and third problems can be solved – by slowing down on the cuts and not trying to rush the machine, and by backing up the wood to prevent tearout. But the best solution, I think, is going to be to cut these through tenons with a coping saw. I’ll have to be careful, and cut a little shy of the line and clean it up with a chisel to ensure a tight fit for the stretcher.

In one of my classes at Woodcraft, Howard Hale told a story. He had been to a homestead museum in the Northeast, and it had a preservation of an 18th-century woodworker’s shop. In one of the displays was the craftsman’s diary, and the page it was opened to had this entry:

“Cut three mortises today.”

Circled above that, was this notation: “Good day!”

It’s easy to forget how much hard work was involved on operations we think of today as pretty easy. I cut twelve mortises today, by machine, and I still feel like that was a good day.

Next: Finishing the mortises, starting the tenons, and back to the top (again).

Labor

To me, the joinery and detail work is the most fun part of any project.  The dimensioning of lumber? Well… not so much.  It is, however, necessary – and kind of gratifying when you have the pile of dimensioned lumber in front of you.  But it’s just straightforward, repetitive work jointing, planing, cutting, trimming. It’s also extremely loud.  The alternative is to dimension by hand and plane to thickness, but there’s a rational limit to the time I have if I’m to finish this project in my available time. Angus doesn’t like the planer or jointer noise, and as you can see, he’s glad this part is over.

As you have probably seen, cutting into the wood sometimes reveals its beauty.  But it can also uncover problems.  For example, this crack – known usually as a “check” when it appears in the end grain, as it does here – was internal to the board. It didn’t appear until the board was cut.  It has to be cut away to prevent the board from splitting later.  It could  be glued together with superglue, and would probably be strong enough.  But I’d rather remove it and not take the chance.  Fortunately, I have stock to spare for this project.

At this point, all the mesquite lumber is cut, sized, and squared.   That leaves some poplar for the drawer frame, and the secondary wood for the drawer itself. You’ll probably notice that most furniture uses different wood inside the drawers and for the drawer sides than the rest of the piece is made of.  Traditionally, that’s called “secondary wood”, and refers to a a less-expensive wood used for parts of the piece that don’t readily show. It also saves on the consumption of less-sustainable woods (such as mahogany, which is typically harvested from old-growth forests).

To me, it’s also about how the primary and secondary woods look together.  The original Stickley table from the plans was white oak and maple.  This one is mesquite, and it makes the maple look bright yellow next to it.  My first choice for secondary wood was sassafras, but I’ve decided that it’s too light and soft for this project. I’ve settled (for now) on a sustainable hardwood called Lyptus.  It’s harder and denser, and I like the way the two woods look together.

This also creates another challenge.  There’s no plywood faced in Lyptus to use for the drawer bottom.  Which means resawing the lyptus into 1/4″ panels, and making a slatted drawer bottom.  I may yet decide to back off and go with maple, since I can get maple-faced plywood… this is going to be a lot of additional work for the inside of a drawer. Stay tuned.