Casualties of Peace

Let hammer on anvil ring,
And the forge fire brightly shine;
Let wars rage still,
While I work with a will
At this peaceful trade of mine.

— Harry Bache Smith

Rule of thumb: If you take a mallet and beat on a heavy object long enough, eventually either the object or the mallet will break.


The Roubo du Garage had one casualty – my mallet. I’m sentimental about it. It was the first hand tool I ever made. It was a streak of luck for me: I didn’t really know what I wanted when I made it. But it’s the perfect shape, balance, weight, face angle, length, and fit for my hand. It’s my go-to for bench chisels, mortise chisels, light to medium joinery, and every other similar task. It doesn’t drive nails, but it works for anything else that requires beating on something with a hammer-like object. It’s made out of some scrap rock maple, with a couple of small walnut inlays.

When it failed, it really, really failed. I was almost through chopping out the mortises in the Roubo when it practically exploded. I had to hunt for pieces. One of the sides split in two, and the center lamination broke through on both ends. In retrospect, I’m amazed it didn’t happen sooner. It was made from three laminations of maple, with the handle fitted through the center section. I added a couple of thin strips of walnut inlay, just for fun. It took me about an hour to make. And it was only held together with glue. As much as I’ve used it, and as hard as I’ve whacked on stuff with it, that’s a testament to the strength of Titebond.


But it was fixable, because all the pieces fit together without any gaps or missing chunks. And if it fails again, it’ll be because it completely disintegrated. I set the pieces together and drilled two 11/32″ holes all the way through the handle. Then I glued everything together, drove two 3/8″ rived oak pegs through the holes, and clamped it up for the afternoon.


When dry, I cleaned out the old laminations with a router plane, slightly deeper than the originals. I cut two blanks of scrap mesquite to fit the inlay slots, glued them in place, and clamped it up again. I trimmed the excess off on the band saw, and planed them flat with a block plane. The two mesquite inlays hide the pegged joinery that hold it all together. Before the handle comes lose, the pegs would have to be broken, and that ain’t gonna happen. The glue joints holding the laminations together at the ends might split, but that’s very unlikely. More likely, the wood itself would fail. And when it came apart this time, the glue joints stayed together. It failed along the wood, not the glue (which is normal). One coat of oil, a pass through the Beall wood buffer, and viola, back in business.
It’s not going to fail easily; I expect to beat on things with it for a lot more years. The end grain of the maple will take a tremendous beating without getting significantly dinged up, but sooner or later it will have to be refaced – or replaced.

It’s funny how much satisfaction I get out of picking up this simple tool. I smile every time I use it, and probably always will. That’s true of most of the tools I’ve made or restored. They have history and character like no other tools do. Making simple tools isn’t difficult, and the planes, shaves, mallets, and handles I’ve made all contain part of my own history as a woodworker.


There are a number of excellent resources on the web for toolmaking. My favorites include In the Woodshop, Toolmaking Art, and Hock Tools (where I get my blades). As a woodworker, a lot of your work often goes to other people. Toolmaking is something you do for yourself, and your own pleasure and education. Go ahead. Indulge yourself.

Ross Henton

Got fish?

Time to quit babbling about the bench and actually use it. I needed to make something for a friend’s birthday, and I recently found a wonderful book at my favorite used bookstore.  Now out of print (but sometimes still available), The Art of Elegant Wood Kitchenware by Tony Lydgate has a number of beautiful designs, and looks like many months of small projects. Can’t have too many of those, can we?

The instructions in the book for this sushi tray were a little sketchy – but this isn’t a terribly difficult project, and there’s a lot of wiggle-room for using the techniques you’re comfortable with.

The slats are curly maple, and the chopsticks, feet, and condiment holder are mesquite.  I found it a little easier to do cut the dados in a slightly different order. First, starting with a solid block of maple, cut the dados for the feet, then the chopstick holder.  Then – and only then – switch back to a regular blade and rip the slats to width. Then switch back to the dado stack, and cut the dado for the condiment holder.

Take the block that will become the feet, and trim it to thickness to fit in the angled dados. Rather than make two cuts for the feet, since their finished height isn’t really critical, take the block the feet are made from and rip it down the middle with one cut at the correct angle.  That ensures that the angle will be exactly the same on both feet. Then, clamp them together, and trim them to the correct height with a hand plane – you shouldn’t have to touch the angled edges except to sand slightly.

The edges of the slats are eased with a block plane instead of a router… I think it makes the finished tray look slightly more organic. Once sanded, the tray was glued up with 1/4″ spacers between the slats.  The slots for the chopsticks and the condiment tray were cleaned up with sanding blocks.

The chopsticks were made – no, I’m not kidding – by shanking 5/16″ blanks in a drill, and holding them against a belt sander to turn the tapers. It worked quite well, but I got more even results by marking across them with a pencil at where I wanted the tapers to stop, so I kept them turned evenly.

The hardest part was the condiment tray. Partly, because I insisted on making it in mesquite… again.  (Well, it’s a Texas-made tray, for Texas sushi. Probably catfish).  After cutting the block to length, I hogged out a rough (very rough) version of the cups with a hand-held router, and grabbed a leather glove and a carving gouge.  It was sharp enough that once the cups were cut, I had fairly minimal sanding to do.  Not a difficult process – this was the first hand carving I’ve ever done.  Just labor-intensive. My shoulder was singing pretty hard when I was finished.

The finish needed to be both food-safe and easily renewable.  All finishes are food-safe, once they’re cured.  But I only had three days in which to get this made, so I just wiped it down with six or seven heavy coats of mineral oil.  That way, it will be easy for its owner to renew it after washing.

This was the first real AB (after-bench) project.  There’s no doubt, the Roubo is the perfect planing bench.  I am, however, going to have to build a riser for detail work in the very short future.  At my Advanced Age, I need to bring small parts a little closer.  But the best part was that I never even had to think about workholding. Every clamping operation was easy and instinctive. Gotta love that.

Music was from Yo-Yo Ma’s Silk Road Ensemble. Seemed appropriate.

More little stuff ahead. Stay tuned.

Ross Henton

Finishing the Roubo du Garage

Ten months and one shoulder surgery later, it’s finished. I’ve used the bench several times over the past several days), it it only took a few minutes to know that I love working on it – the height, the workholding, the stability. I already know a couple of things I’ll be changing, but I’m extremely happy with the results.20110912-064350.jpg

The Planing Stop

The planing stop is a sliding, friction-fit square post above the face vise. I really, really wasn’t convinced about its utility, but Christoper Schwarz insisted that it’s fundamental to the design of the bench, and it didn’t take long to figure out that he was right.

It allows you to but a board against it and start planing, without any other (more complicated) workholding required. It adjusts up and down with hand pressure, no tools required. It’s position is ideal for planing long boards, and can be used as a brace for other workholding (more on that in the future).

Its simplicity was only exceeded by the almost ridiculous difficulty in getting it made. I mean, really… it’s just a square hole in the bench. It has to be fairly accurate, but it’s just a hole.


The problem I faced was with the wood. Southern yellow pine apparently has tremendous variability in the density of the rings. That means that the downwards pressure of the chisel is concentrated in a tiny area, which causes the edge to crumble. I’ve worked with several extremely dense woods, from ebony to mesquite, and I’ve never seen anything more damaging to share edges than cutting cross-grain in this resinous pine.

From Adam Cherubini’s “Arts and Mysteries” blog:
“Interestingly, the toughest wood on my chisels is pine. If you’ve followed my logic so far, you instantly know why. The hard/soft rings of pine, especially quartered pine (the use of which I advocate), poses a unique problem for an edge tool. Pressure is concentrated discreetly, instead of spread evenly across the cutting edge. And I believe that the hard parts in pine are pretty darned hard compared to other woods (like those previously mentioned), but I don’t know that for sure.”

I think this is a good assessment. Chopping the hole for the planing stop was difficult, and I had to stop and resharpen my larger chisels twice. But it’s such a joy to use that it was worth ten times the effort. Just butt a board against it and start planing.

The Sliding Deadman

This was the easy part.
The deadman was made from a glue up of some leftover pieces of mesquite, and cut to shape on the band saw. The instructions in Christopher Schwarz’s book had recommended cutting the lower notch for the deadman on the table saw, with the blade angled to produce the notch. However, I had the bandsaw pulled out and ready, and its resaw capacity is huge – so I just cut the notch on the bandsaw right before cutting it to shape. 20110912-070211.jpg

Same for the tenon at the top that holds it in the track. A few minutes with a shoulder plane to trim the tenon, and to clean up the edges with a spokeshave. The holes were fast to mark and drill, and viola – it popped into place.

Dog Holes

Drilling the dog holes was easy – and tedious. The best way I found was to start the holes with a forester bit, and then switch to a spade bit at about a third of the way through. That way the top of the holes was nice and even, and the longer (and more aggressive) spade bit made fairly short work of the rest. I only drilled out one line across the front, four in the back, and one above the plane stop. It matches the Roubo drawings I’ve seen, and I can always add more later if necessary.

To Crochet or Not

This was a touch decision, just because it looks so cool. Ultimately, the answer was not. The clearance between the end of the bench and the face vise is shorter than standard, because the overall length of the bench was reduced. The reviews of the traditional crochet I’ve read have been mixed as to whether or not it’s really much of an advantage. I could have added a (slightly shorter) one, but I have a plan to retrofit one later if I decide to (stay tuned).


Now, these were a wonderful idea… and proof that sometimes, the older ways of doing things have advantages. A holdfast is just a J-shaped piece of metal; tough enough to handle being banged on without breaking. All you do is put the long end through a dog hole, set the J-end on a workpiece, and bang the top with a mallet a couple of times. 20110912-072936.jpgIt jams in the hole (theoretically), and holds the workpiece down to the table. You need to put a thin piece of wood between the head of the holdfast and the workpiece, to avoid marring the work. To remove it, bang the back side of the holdfast (by the post) and it pops out.

Holdfasts come in several varieties – ranging from beautiful hand-forged work to cheap wrought-iron versions. This is one area where I did quite a bit of reading. Reviews for included very expensive (and lovely) hand-forged ones, to some that apparently broke the first time they were set. I settled on a middle ground = utilitarian, but well-made holdfasts from Gramercy Tools.

One thing you may read is that sometimes they don’t want to grip in wood that’s this thick (about 4 1/4″).  The manufacturer’s recommendation is that you scratch them up across the stems with some coarse (80-100 grit) sandpaper. That really does help. Aside from that, I had no trouble getting them to hold in this thickness.

The Lower Shelf

Adding the lower shelf was easy, partly because I deliberately didn’t do anything fancy. It’s just sections of 2×4 with chamfered edges, butted against each other and cut to fit. They sit on ledgers inside the lower framework, and aren’t attached to the bench itself – they’re attached to two lateral runners underneath them. The entire shelf lifts out, and is only held in place by its own weight.

There’s method to my madness. I couldn’t decide whether or not to do solid shiplapped joints, or to do open slats to let dust fall through. This way, I can easily remove it and replace it later if I change my mind. As it is, it holds the tool carriers I made last year for sanders and air tools, and its weight adds even more to the stability of the bench.


And now… Proof again that a lot of the most famous woodworkers didn’t live in Texas. Including Tage Frid.

Professor Frid, author of the tremendous Tage Frid Teaches Woodworking, had an amazing amount of useful commentary about different joint types and their construction, and a wonderful no-nonsense Git-R-Done attitude towards woodworking. He also recommended a particular formula for finishing benches, which advances from raw linseed oil, boiled linseed oil, BLO and japan drier, and paste wax. It takes nine days, start to finish.

This all proves that Prof. Frid never set foot in Texas, or at least he wasn’t working with southern yellow pine in Texas in the summer of 2011.

All started out well, but the result at the end of day 5 was goo. It became a sticky, uncured mess that couldn’t be scrubbed off, much less sanded. I had to resort to a scraper. Fortunately, Howard Hale (a student of Prof. Frid who teaches at the local Dallas Woodcraft) set me straight: “If the oil is no longest being soaked up by the resinous wood, the you’re done. Wait two days for it to ‘bleed’ oil up from the surface, give it one coat of straight japan drier, and wax and polish it. And that’s all.”

As usual, Howard was right. I was done in three more days. The result? I can dump super glue on the surface, spray it with activator, and pop it off with an fingernail. Gotta love that. This was for the top only – the mesquite used for the vises and other parts got several coats of oil, but that’s all.

Final Notes, and Hindsight

Yes, I love working on this bench. It only took a few minutes of watching it not vibrate to figure that out. Also, I haven’t had to stop to think about the best way to hold something, or figure out where to put clamps. Workholding is fast, easy, and instinctive. But there are some things I know I’d do differently.

Choice of wood: Wow… this is a big one. But I wouldn’t use southern yellow pine again, in spite of its availability and cost. I found it difficult to work due to its variation in density across grain – it was really hell on chisels. Also, it’s prone to warping badly. This was after months of seasoning time, but it still warped after cut and sections were glued up. I’d probably go ahead and spring for beech or maple, even knowing how hard they can be.

The Veritas vise: I really like this vise, and I have nothing bad to say about its manufacture or mechanics. But even setting aside the problems I had with the instructions, and the (comparative) difficulty of installing a wagon vise, I’d probably go ahead and get the Benchcrafted tail vise. The Veritas vise does require additional room at the end of the bench to open and close it; that’s not a problem with wagon vises.

Positioning of the tail vise: I’d move it a couple of inches inward on the bench, so the dog holes don’t get into the track of the sliding deadman (my bad).

Rouboverload. Rouboverkill. Rouboverdone. This was a much, much longer project than I had expected, and I already have ideas about changes, fixtures, and add-ons… making dogs, more planing stops, and a riser/Moxon Vise for doing close-up work. The bench is the largest hand tool I’ve ever made. Building it will always be something I look on as a turning point in my woodworking. Maybe projects will be BB or AB (before-bench or after-bench).




Much, much after.

We’ll see. Stay tuned.