Roubo Is Coming… Part 5

After several months of work, this is the moment when the various components of the workbench start to come together and look like something. There’s a great deal of distance left to go, but I now have the makings of a workbench undercarriage.


Marking the tenon shoulder with the Vesper 4″ square and Hamilton marking knife

Cutting the mortises for the short stretchers was enjoyable but slow work, so I was looking forward to swapping my chisels for a tenon saw and cutting the second half of the joint. This is much faster work, even accounting for fine tuning the joint. When I laid out the mortises I had locked my marking gauge in position, and it will stay at that setting until the undercarriage is completed. I laid out the bare faced tenons with the marking gauge, and then scribed the shoulders with a square and marking knife, before shading in all of the waste with a pencil.


The Roubo Beastmaster is the definition of controlled aggression – it ploughed through this 3″ oak tenon like the Apprentice ploughs through chocolate brownies

To cut the tenons I reached for my Bad Axe Roubo Beastmaster tenon saw. This build was one of the reasons I originally ordered this saw, and at 18″ long with 5″ depth saw plate it is a substantial tool. The key to cutting any joint is to establish a saw kerf to guide the saw through the entire joint, and cutting the tenon cheeks takes three cuts. First I set the stretcher at an angle so I can work to the layout lines on the end of the stretcher as well as one side. This keeps me cutting straight on both vertical and horizontal planes. I start at the corner nearest me, and keep sawing until I hit the far corner of the tenon end and the baseline on the side facing me. I then flip the tenon over and work down to the baseline on the newly exposed face, using the saw kerf along the end of the tenon to guide me. Once I hit that baseline, I place the stretcher vertical in the vise and cut out the triangle of waste that remains above the baseline.


After making the first cut, turn the stretcher over and cut the second side, using the existing saw kerf to guide you.

Once the cheeks are cut, I cut the shoulders (this time with a Bad Axe Bayonet saw, filed hybrid – this is my “do everything” utility saw, as I’ve written before). The tenon shoulders are a “first class cut” and so I keepen the knife line with a wide chisel and pare out a small trench to guide the saw. Placing a finger gently on the toe of the saw prevents it from jumping out of the cut, and I saw down to the layout line until the waste pops out.


Cutting the shoulders with the Bad Axe Bayonet

Placing the tenon in the corresponding mortise helps to identify if there is any tuning or further material to be removed. A broad chisel helps to remove any small humps from the tenon cheek, but to be honest if you have cut the tenon carefully you shouldn’t need too much tuning. Gently chamfering the corners of the tenon helps it to slide in smoothly without any junk in the corner of the mortise (often the hardest parts to clean out) fouling the fit.


Chris Vesper says this is square, which is good enough for me

Once I had a pair of tenons cut I did a test fit with a pair of legs, checking for square with my Vesper 10″ square (the final arbiter of square in my workshop). I now have two subassemblies of legs connected by short stretchers, and next will be to cut the mortises and tenons for the long stretchers.


One of the two completed sub-assemblies. Now to cut the joinery on the long stretchers and connect them all up.

Roubo Is Coming… Part 4


Laying out the mortises with the Vesper 10″ square and Hamilton marking knife

After processing the legs and stretchers for the Roubo bench I was looking forward to cutting some joinery. Roubo explains that the stretchers are flush to the front of the legs (which also means that the long stretchers will be in the same vertical plane as the edge of the bench top) and situated between 4 and 5 pouce above the floor. He is silent on how the stretchers should be fixed to the legs, although mortise and tenon joinery is the obvious choice for this element of the build. I decided to follow Chris’ approach to this joint and use bare faced tenons, the advantages of which are that you only have one tenon shoulder to worry about, and you can include a good solid 1 pouce thick tenon (and 3 pouce long) in the joint. As with everything on this build, the joinery is big, which is why this bench should last forever.


Laying out tenon width also established the dimension of the mortise

To break up this element of the build into manageable chunks, I decided to cut the joinery for the short stretchers first to create two sub-assemblies of legs, and then cut the joinery for the long stretchers as a second round which will leave me with a completed undercarriage. Whether you cut mortises or tenons first is, I think, purely a matter of personal preference. I tend to cut mortises first. What is important is careful layout of both halves of the joint, which I much prefer to do in one sitting so as to avoid any small (but significant) changes in measurement.


Boring out the waste with a 1920-era North Bros. brace and 7/8″ auger bit. The Vesper 4″ square and my great uncle’s Starrett combination square keep me plumb to the workpiece

The Veritas mortise gauge is my go-to tool for laying out the joint, and I typically start by marking out the tenons. This establishes the setting for both halves of the joint and the gauge is then locked into that position for the remainder of this project. With the tenons marked out, I then laid out the mortises, measuring from the bottom end of each leg. The ends of the mortises were scribed across the full width of each leg, as this will enable me to wrap round the positions to the sides of the leg for the long stretcher mortises.


Ready for paring

As I’ve said before, this oak is tough stuff. There are no medals for cutting these mortises with only a chisel and mallet, only blisters. Also, the widest mortise chisel I could find on the internet was 1/2″ wide, which would effectively require cutting two mortises for each joint, and removing the thin webbing between them. That sounds a lot like work to me, and so I took a different tack while staying firmly hand tool only. So, with my big Jet drillpress standing tantalisingly close, I fitted a 7/8″ Jennings pattern auger bit (from Tools for Working Wood) to my 1920s era North Bros. brace, and bored out much of the waste by hand. Yes the Jet would have eaten much of this oak for breakfast, but I really enjoy using the brace and bit, and taking it at a steady pace this did not feel like a particularly tough job. The key is to be bold and use the largest bit you can, to remove as much waste as possible. A knife line down the middle of the mortise also helps to give a clear centre line in which to place the lead screw of the auger bit.


Squaring the ends of the mortise

With much of the waste removed to depth, the mortise could then be cleaned up with chisels. Effectively paring the mortise walks square and true relies on taking small bites with a chisel, and nibbling your way back to the baseline, very much as you would when chopping dovetails. First I removed any webbing left between the auger holes, and then square up the ends of the mortise using a 3/4″ chisel. The end grain is brutal hard stuff, even with a freshly honed chisel, and a sharp tap with a mallet helped ease things along.


Paring the sides with a wide chisel

The sides of the mortises were easier going, and I used a 1 1/2″ wide Blue Spruce bench chisel for these – that width gave me a very consistent straight wall but without being so wide that it needed too much force to push through the cut. Again, paring the excess away in small bites and working back to the knife line, is essential. I also use a two handed grip for this work – my right hand is on the chisel handle with my thumb looped over the end for maximum downward pressure, while two fingers of my left hand are wrapped round the front of the blade, pulling it flat against the wall of the mortise. Each hand only has one job to do – the right hand drives the chisel and the left keeps it running in the right direction. This grip keeps the chisel from drifting out of the cut, and also from under cutting the wall of the mortise.  A small combination square can help test the walls of the mortise – if the blade of the square does not make contact with the top of the mortise then there is a bump which needs paring flat.


Checking for square with the vintage Starrett

Next up will be cutting the tenons for the short stretchers, and tuning the fit of the joints. And then repeating the whole process for the long stretchers.


Two of the four mortises cleaned up and ready for tenons

Roubo Is Coming… Part 3


And with that, the undercarriage is all processed. Four legs and the same number of stretchers. My reward for all of this stock prepartation will be cutting 8 mortise and tenon joints, and fitting the vise hardware to the front left leg. Not that I’m done with planing stock for this project – far from it. There is still the vise chop and lower shelf to process, not to mention the small matter of flattening the slab top. But after six weeks or so of nothing but planing oak until it is square and true, I’m looking forward to some joinery. At the outset I did wonder if processing this quantity of material by hand without any distractions might be repetative, but while it has been heavy work with occasionally truculent timber, it has all been very enjoyable and offered many learning opportunities.

Now time to put the hand planes to one side and sharpen my chisels in readiness for cutting those joints. It’s going to be good to see the base come together.

Roubo Is Coming… Part 2


Timber selection – picking the final two legs and vise chop out of these beams. In the rough this timber looks a little gnarly, but it cleans up very nicely.

Although I’ve not written about the Roubo bench build for a few weeks, I have been hard at work processing stock for the undercarriage. All four stretchers are now square and cut to length, and as of today three of the legs are also processed. Only one leg remains and then I will be ready to start cutting joinery (eight mortise and tenon joints to connect the stretchers to the legs). Processing stock of this size, in oak, by hand is a significant undertaking, and cutting some joinery will be a real treat. That being said, I am really enjoying the change in scale with this work, and instead of a sprint to get all of the stock processing done, I’ve been taking my time and enjoying the journey.


Transforming grey timber into something beautiful and useful. The scrub plane is perfect for raidly removing high spots

Processing all of the stock by hand has partly been an ideological choice – for this build I want to, insofar as is possible, get as close to the bench described by Roubo, and that includes the process of building it. It is also a practical decision – my 12″ Dewalt lunchbox thicknesser is a fine workhorse for furniture sized material, but this bench is more akin to timber framing and I have serious doubts that it would cope with some of the monstrous timber I need to clean up for this build. So, meat-powered tools are pretty much the only game in town for this work.


Smoothing surfaces with the No3

To be honest, although the work is much heavier than processing furniture sized material, it hasn’t been too arduous, and there have been opportunities for technical insight. I am used to processing all of my stock by hand – it is how I’ve always worked. But working with oak beam stock has offered some valuable lessons in honing those skills because the scale of the work ,and the toughness of the oak, means that you want to work as efficiently as possible.


Mutton tallow – lubricating the soles of planes makes for much easier work when taking heavy cuts in difficult timber.

Firstly, I’ve been remembering to lubricate the sole of my planes more frequently than I often do. I always have a tin of mutton tallow in my apron pocket for lubricating planes and saws, but when working more forgiving timber such as pine I am not always fastidious about using it. This oak is tough stuff, and it is remarkable how a smear of tallow can improve the ease of hogging off material. Later in the build I will be building a grease pot to hold my tallow, and right now I can easily say that I am a convert to lubricating planes. Ease of work is also directly influenced by sharpening. Now, we all know that sharp fixes everything, but how many people sharpen as often as they need to? No one, that’s who. We can (and should) all sharpen more. This oak lets me know immediately if my planes are not sharp enough, and as a result I’ve been getting in the habit of sharpening much more frequently as I work (on occasion twice per leg). Hopefully that will be a habit that sticks, because in reality breaking for three minutes to freshen an edge is far quicker than battling on with a dull tool. So thank you, truculent oak for the helpful reminder.


Trimming the legs to length. The Skelton Panel Saw cuts so accurately I can get a square end straight off the saw without any need to clean up with a block plane

Early on when cleaning up the legs, I asked myself whether I wanted to treat each and every surface with the smoothing plane, and if not, what sort of surface would be acceptable. I’m not advocating sloppy work, but these large beams are going to be turned into a workbench, which will see a lot of abuse over the decades I work at it (and hopefully long after I’ve returned to the soil). So really it was a question of proportionality. But also, I know that I’m going to spend a lot of time standing at the bench, and a big patch of tearout will frustrate the dickens out of me, especially as it’ll be a total bear to try and pretty up some of the surfaces once the bench is assembled. In reality, a smoothing plane is a great tool for making localised adjustments, so I’ve been hitting every surface with the smoother, although also not sweating minor surface imperfections if they are on inward facing surfaces. I think that’s the right balance.


It’s not square until Chris Vesper says its square.

Next weekend I will tackle that final leg, and then put my planes down for a while to cut the mortise and tenon joinery, and fit the crisscross for the leg vise (this is much easier to do before the legs are fitted to the bench). After which it will be time to process the slab top – a task which will require some perseverence, grit, and maybe an assistant.


Current progress – three legs and four stretchers all processed and ready for joinery.

Roubo Is Coming… Part 1


My new Hamilton Woodworks traditional marking gauge in use

It is hard to believe, but I first started thinking about a Roubo bench build four years ago. And finally, I have started the build in earnest. This promises to be a fascinating project with plenty of learning opportunities, both in terms of building the bench and then using it. The scale of the work involved is quite different to anything I’ve done before – in many ways it feels more akin to timber framing than to furniture making. And that is one of the two reasons why I decided to start by making the under carriage, rather than tackling the (frankly monolithic) slap top – working my way up from smallest component to largest means that I will have my process figured out, and adjusted to the scale of work, by the time I address that imposing slab of oak. The other reason is that while the top continues to acclimate to the workshop (a process that could take some time) it is liable to move, which means that starting with the top would result in several flattening sessions being needed by the time I assemble the bench. In contrast, the components for the undercarriage have all acclimatised, which means that they will be much more dimensionally stable.


All measurements for this build will be taken from the Pied du Roi ruler made by Brendan Gaffney

Following the principle of working smallest to largest component, I started by processing material for the four stretchers. These are 4″ wide, 2″ thick, and will have 3″ barefaced tenons on each end. Regular readers may remember that my brief for this project was working from the description in With All the Precision Possible, and using the 18th century French Pouce as my base unit of measurement. All measurements I use on the blog for this project will, unless otherwise stated, refer to the Pouce rather than a modern inch (to be honest this is as much for my benefit as anyone else – if I have to swap between two different measurement systems I’m going to get very confused).


Processing the long stretchers on the staked saw benches

Processing the stretchers was the same process as for any other furniture project, with a few significant exceptions.The first was that the long stretchers were quite a bit longer than my Sjoberg benchtop, and too wide to fit in the face vise. To plane the wide sides I would be unable to work at the bench, unless I was prepared to clamp the stretcher to the benchtop and work it in sections. That sounded like a slow way to work, so instead I took the opportunity to try some of the techniques explained in Ingenious Mechanicks. I don’t have a low workbench (yet) but I figured my saw benches would work at a pinch. Setting up the long stretchers on the saw benches, I was able to sit on the workpiece to keep it in place, and plane the section in front of me, before shuffling back a few feet and working the next section. Not having to adjust any clamps made this a very quick and efficient way to work, and I was surprised at just how comfortable working while sitting down was. The saw benches are incredibly versatile, and have really earned their keep since I built them in 2016.


Sitting down on the job (literally) is a comfortable and efficient way of processing stock

If I were working this way regularly I think a set of wooden bench planes would be beneficial, including a shorter jointer plane (the No8 is quite a weight, and length, to be extending out in front of you from a sitting position), but on the whole this was an enjoyable and efficient way to work. The one area where the saw benches did not work quite as well as a dedicated low bench was the last section of the stretcher, which you can’t sit on as you work it – a low bench would have a planing stop or series of pegs to secure the workpiece while working at the far end. At this point I did reach for a couple of clamps, but as that was the final short section it did not slow me down too much.


Ripping the stretchers to width with my trusty 1900 era Disston D8

The second difference to processing furniture sized stock was that the thickness of the stretchers exceeded what my shooting board plane could comfortably handle. I want all of the stock for this project to be as accurately processed as possible, so as to avoid any unpleasant surprises come glue-up, and that includes accurately squared-off ends. With my shooting board out of the question, I turned to my mitre box and Bad Axe mitre saw. Having knifed in clear layout lines on all four sides of the stretcher ends, I cut a deep kerf along each knife line using the mitre box. The four kerfs then guided the saw to leave a clean and square end from which I can layout the tenons.


Squaring off the stretchers with a mitre box. Kerfing in the layout line on all four sides offers a “path of least resistance” which guides the saw

With the stretchers processed and ready for tenoning, the next task will be to process the legs. These are 6″ wide and 4″ thick, so will represent a step-up in dimensions and scale of work. It will be interesting seeing what lateral thinking they require.

Roubo is Coming

Roubo is coming, and he’s a heavy dude. I know this, because on Monday night I unloaded an oak slab that is destined to become the top of my new workbench. It’s hard to believe that over two years have passed since I last wrote about the ideas and options for a new workbench. While I may not have written about benches since then, I’ve spent a lot of the intervening period thinking about what I wanted from a workbench. Often when I was straining the capabilities of my current Sjoberg bench. In that time, I’ve refined my plans, settled on the design, and given some serious thought to how I’m going to approach the build. This is a big project to undertake, and so I wanted a narrative arc to carry me through. I think I’ve found one.


Stretchers, planing stop, and slab all stickered. The legs and vise chop are stacked up separately.

It was Ethan who convinced me at the start of this year to stop putting this project off for another day, and to get to building. Since then I’ve been trying to source the timber – no small endeavour for a project like this, but which came to fruition on Monday, when a very heavily laden truck arrived at the workshop. But what exactly am I going to build with all of this oak?


My instruction manual for this buld – Roubo’s desciption of the work bench, and the Pied du Roi ruler

Well, as you’ve probably guessed by now the new workbench will be Roubo-orientated. In my last workbench posts I was considering a dovetailed end cap, sliding deadman, wagon vise. Lots of accoutrements that Roubo would not have recognised. Instead, I’ve decided against doing a modern interpretation of a Roubo bench, and will follow Roubo’s description of the workbench (and in the excellent Lost Art Press translation of With All the Precision Possible) together with the illustration in Plate 11. So no sliding deadman, tail vise, or other modern additions. This is going to be as close to the illustration in Plate 11, and Roubo’s description, as I can make it – including the grease pot and drawer.

And this extends to the basic unit of measurement. In an effort to follow Roubo’s description faithfully I will be using the period appropriate Pied du Roi as my unit of measurement, in preference to the modern inch, with the aid of Brendan Gaffney’s excellent rulers. One of the practical research questions that I’m interested in exploring with this build is whether using a different unit of measurement has any practical implications once the bench is in use.


Benchcrafted Glide C vise – the texture of the casting, and beech knobs, work really nicely with the oak. And who doesn’t want a steampunk ship’s wheel on theit workbench?

So a slab top oak bench, with the iconic sliding dovetail leg joint. That sounds pretty good. For work holding I have a traditional planing stop made by Peter Ross and my one concession to modernity – the Benchcrafted Glide C vise. To supplement the planing stop I’ll be using 1″ diameter holdfasts and Does’ feet. At this stage I’m not convinced by the need for a crotchet, so will be omitting that from the build. But it is always something I can add at a later date if it becomes useful.


Peter Ross planing stop

How about the dimensions of the bench? As it currently sits in my ‘shop, the slab top is 6″ x 22″ x 110’ in the rough. Ultimately. I’m shooting for a top that is 8 Pied du Roi long (in the region of 102 modern inches). This should be long enough that I’ll never outgrow it, but short enough that it is easily housed. My hope is that this will be the only bench I ever build. Sourcing oak of this size has not been an easy task, but Matthew Platt of Workshop Heaven kindly put me in touch with Acremans Timber, who were very helpful in supplying oak in what is an unusual set of dimensions.


The slab top – 110″ of green oak ready to be transformed into a workbench

The oak is now stacked up and stickered in the ‘shop while I wrap up a few projects. Towards the end of the year I’ll start the build in earnest – that will have given the legs and stretchers time to acclimatise. As for the slab top, well, given that it is both air dried and huge, it will take decades to dry out. My moisture meter tells me the end of the slab is at 16% moisture content, but the middle of the slab is likely to be much, much wetter. One of the other learning experiences of this build will be working with a monolithic slab which is, to all intents and purposes, still green. It promises to be a very interesting (and exciting) build, and I hope you all join me for the ride.

Ingenious Mechanicks – on review


Aside from my round-up of favourite woodworking books earlier this year, book reviews are not really something I’ve done on the blog. But as soon as I finished reading Chris’ new book, Ingenious Mechanicks on holiday earlier this month, I knew that I was going to have to write about it here.

If somehow this book has passed you by, then it can briefly be summed up as an investigation into low workbenches and workholding solutions spanning a 2,000 year period from the Romans to 20th century Estonia, by way of China and South Ameria. In the book, Chris draws upon a vast body of history of art, written primary sources, not to mention the Saalburg bench – the oldest surviving workbench in the Western world (dating from 187 AD).  The research is then rigorously tested by building four low bench designs and working at them to unlock techniques and working methods. So far so niche, you might think, but actually the information and lessons in this book have a wide application.

The book contains a useful guide to building (any type of) workbench with a thick slab top, which is a very welcome and helpful supplement to the Workbenches book. Next follows a detailed survey of the source material, and explanation as to how the research led Chris to build the benches as well as the simple appliances which increase their functionality. The production values of this book are high, as you’d expect from a Lost Art Press release, and the benefit is that the pictoral sources are reproduced with great clarity so that the reader is able to interrogate the images and reach their own conclusions about how artists have represented workbenches and woodworking. As I’ve written previously, history is one of my first loves, so this book was always likely to be like catnip for me. I wasn’t disappointed.

But putting aside my particular interests for a moment, as I was reading Ingenious Mechanicks I was struck by the recurring thought – this book is quite possibly the most important work Chris has published since the Anarchist’s Tool Chest. That’s a pretty bold statement, I know. But there are two reasons why I think this.

First is that Chris is presenting a way of working which is (for most people) entirely new – one which presents solutions for age old workholding problems, as well as new challenges. The vast majority of the way we work is from handed down knowledge – we learn from books, Youtube videos, or (heaven forbid) other woodworking humans in the same room as us. These are all healthy and invaluable ways of gathering and transmitting knowledge. But what I find exciting about Ingenious Mechanicks is that it offers something truly unique – the opportunity to experience a vibrant way of working for which not all the answers are known, and for which there is not a pre-existing corpus of knowledge on which to draw. Instead, the reader is presented with a chance to explore the benefits and limits of low workbenches and to report back from the frontline of woodwork research as they happen upon working methods and techniques. The only price of admission is to suspend pre-conceptions, and spend some time using unfamiliar workholding solutions.

Secondly, Chris has laid bare his workings and research methods. In doing so he provides the tools for others to research historic woodworking practices, and demonstrates how a rich seam of art history, archaeology, written primary sources, and an inquisitive disposition, can provide valuable insight into woodworking practices that are both historic and relevant today. Ingenious Mechanicks therefore becomes part woodworking book on pre-16th century workbenches, and part woodwork research methodology.

As I was reading the book I found myself asking questions about how familiar operations would translate to a low bench. Would wooden bodied planes respond differenty, or feel more comfortable, than metal bodied planes when processing stock? What sort of body position and posture would be necessary for sawing accurate dovetails? Would ow benches encourage the use of specific joinery techniques? There are few answers yet, and I know that there is really only one way to understand how these benches work. The oak for my slab-top Roubo bench is on order and should be here in the next couple of weeks. And now I am measuring up the workshop and trying to decide whether I should build one of these low benches too. And when all is said and done, persuading a reader to build one of these curious old benches goes to show just how effective Ingenious Mechanicks is.