I was flattening some panels by hand the other day (too wide for my machines), and that got me thinking about plane blade camber. If you search online for discussions of blade camber, you’ll find that a great many electrons have been spilled on the topic. One common thread in these discussions is frequent confusion over the fact that a bevel-up blade requires more camber (i.e., the center of the blade needs to protrude further beyond its corners) than a bevel-down blade to have the same effect.
On the one hand, everyone seems comfortable with the notion that as the blade’s bedding angle decreases, the effective radius of curvature of its edge increases. This is easy to see. First, find yourself a thin disk (e.g., a CD or DVD) and hold it up at arm’s length:
When the disc is perpendicular to your line of sight, the apparent radius of its lower edge is equal to its actual radius (2-3/8″ in the case of a CD/DVD). But start tilting it from perpendicular, and the curve flattens; its apparent radius increases:
Tilt even more, and it keeps increasing:
From the point of view of the wood fiber that’s about to have its head chopped off by an oncoming blade, the greater the tilt from vertical, the greater the apparent radius of curvature, and consequently the less the depth of cut at the center of the blade. And since the blade in a bevel-up plane is tilted further from perpendicular, its apparent radius of curvature is larger than that of the bevel-down blade unless we make its actual radius of curvature smaller (i.e., increase its camber). Easy.
On the other hand, we’ve also all seen diagrams of bevel-down vs. bevel-up planes seated on their respective frogs:
The resulting cutting geometries in the two cases are identical. The blade’s cutting edge comprises two intersecting planes, one formed by the back surface, and the other by the bevel. The only difference between the two configurations is that these two planar surfaces switch roles.
This is where I think some people get confused. If the two setups are equivalent, why can’t we measure the blade camber in the same way with both? In truth, we sort of can, but there’s a difference between the bevel in a cambered blade vs. a straight blade. When the camber is small, that difference is also small (and negligible), but with a strongly cambered blade, such as one we might use in a fore or scrub plane, it’s not. With a cambered blade, the bevel is not planar. In fact, the bevel is a section of the surface of a cone:
That’s where the equivalence breaks down, as it’s no longer possible to directly superimpose the cutting geometry of a bevel-up blade onto that of a bevel-down blade. And so we go back to always measuring the camber with respect to the back of the blade.
Anyway, is any of this important? Only to the extent that you get a feel for how the different parameters interact, so that you’ll know how much to camber your blade to achieve a given depth of cut.
I’m avoiding the math here, because it’s been covered before (such as here and here), but I did put together a little online app that lets you plug in some numbers to see how this all works. Here’s a screenshot:
You can find the app here. To use it, enter your bed angle and blade width, and one of the other three values. The app will compute the other two corresponding values for you, dynamically updating the display as you modify the values. The bed angle is in degrees; the other values can be in whatever length units you choose, as long as you’re consistent (inches, millimetres, furlongs, it makes no difference).
Now, I know that someone is going to read this and then get out their micrometer and measure their blade camber to three decimal places, to which I say,
STOP!! PLEASE STEP AWAY FROM THE PLANE!!
The point of the app is intuition, not prescription. The precise value of camber that you end up with is largely irrelevant, as long as you’re in the ballpark.
If you have been interested in the low Roman workbenches I’ve been writing about, here’s your chance to follow along while two woodworkers build them. You can even join in and build your own with home-center materials and firewood (more on that in a minute).
Joshua Klein and Mike Updegraff of Mortise & Tenon magazine are each going to build Roman workbenches and blog about the experience starting on Feb. 20. You can read more details about their plans here.
When starting with rough materials, these workbenches take me about 10 hours to build (that includes the time to document the process with photos and notes). But I have an electric lathe. I think that balances out the equation – I think anyone can build this bench in about 10 hours.
If you don’t have a slab on hand, here’s what I would do: Buy a 12’-long 2×12. Crosscut it in half. Glue the two halves face to face. That’s the benchtop. For the legs, go buy some firewood from the grocery store if you don’t have a big firewood pile already. Split the legs out of firewood billets.
The low Roman workbench is a lot of fun to build. But it’s even more fun to use at the end. You get to sit while you work – nice!
It’s also funny how the low bench has become the community center for my workshop. When I have visitors, they naturally gravitate to the low bench and sit there (I’m the only one who sits on my Roubo bench).
If you’d like more details on why the Roman bench is a marvel of early technology and workholding, check out the article I wrote on it for Issue 2 of Mortise & Tenon magazine.
I’ll definitely be following Joshua and Mike’s progress that week. But I won’t be building along. I’ve already got one….
Yesterday Chris and I had a Q & A session about stick chairs, the Hall’s Croft chair, chair design and more. It was only towards the end of the evening portion of our chat, and after he had consumed two beers, that it was necessary to redact a line or two.
Suzanne: The last time we talked about chairs was in January 2015, your pre-condition was we had to be naked (although we were about 550 miles apart) and it was titled ‘Naked Necessity.’ What pre-condition do you have this time?
Chris: Let’s say hirsutus maximus.
Suzanne: Sorry, I’m rejecting your hairy pre-condition and going with a jolt of Tia Maria in my afternoon coffee. Let’s get started.
When you see a stick chair what do you find pleasing to your eye?
Chris: Well first it’s the angles. Peter Galbert, the bard of chairmakers, nailed it when he wrote this: “The angles of the legs, along with their design, help give the chair a ‘gesture.’ Whether the desired result is a visual lightness and sense of action or stability and weight, angles are important.”
I’m looking for a gesture that is somewhere in the neighborhood of “f-you world.” I like chairs that have an animalistic stance – like they would jump up and lick your face or tear you to shreds.
Most Windsor chairs have a stateliness that leaves me cold. In contrast, Welsh stick chairs are more like a crazy uncle.
Suzanne: For the woodworker in you what do you like about these chairs?
Chris: These chairs were not manufactured. And in many cases they were built by the same people who used them. So every chair is different and is connected to a person.
Plus, the makers didn’t follow the same rulebook as the Windsor makers of High Wycombe. They used angles that were more rakish and severe (and got away with it). They used construction methods that were simpler (and many of these chairs survived 200 or more years). And they used found materials. The armbow of many of these chairs is a curved branch they nicked from a coppice or from their own land.
You don’t have to be a professional chairmaker to make nice Welsh stick chairs. You just have to have some sticks, a plank for the seat and a few tools.
The Hall’s Croft chair.
Suzanne: When you and Roy Underhill stumbled upon the Hall’s Croft chair what were your first impressions?
Chris: Roy and I had spent the entire day crawling around the floors of the dwellings of Stratford-on-Avon, photographing all the stuff that was fascinating (I filled a 32gb SD card). There was a short bed, for example. Why is it so short? Was it because people were shorter back then? Or was it because beliefs at the time were that you should not sleep flat – you should sleep upright – so evil spirits didn’t get in through your mouth.
When we saw the Halls Croft chair we both just stopped for a minute. Unlike a lot of the stuff we’d seen that day, this chair was out of the norm (by the way, I really doubt it was contemporary to the house; many of these chairs are much younger than dealers suspect or advertise).
The first thing we did was set up a perimeter. I poked my head into the dining room to make sure the docent was facing the cafe. Then Roy started putting objects on the chair that were an identifiable dimension – such as a touristy pamphlet – so we could scale the chair’s parts when we got back to the States. We took dozens of photos each whilst I kept a lookout for the very helpful employees of the house museum.
We did it without upsetting anyone and without anyone (me) having to say: I’ll create a diversion!
I think Roy liked the odd crest rail. I really liked the birdcage-like structure of the spindles.
Suzanne: You encourage woodworkers to explore many furniture forms to develop their knowledge of joinery and their own designs and suggest carrying a sketchbook, camera, etc. What else do you do to get a good record of a piece of furniture?
Chris: I always carry a camera with me. It’s a habit I picked up as a newspaper reporter and has served me well as a furniture designer. I also carry a credit card – not to pay anyone off but to put it in photos so I can scale the object in Photoshop. And I try to take photos that resemble construction drawings: a straight-on elevation, a profile and (if possible) a plan. Then I take a “beauty shot” to remind me of how all these pieces add up together, and I take photos of the important details.
I rarely make replicas. But knowing what a maker did – exactly – with a beautiful piece is solid gold information.
Suzanne: As for measuring the chair I’m surprised you didn’t use body parts as measuring devises. And no, not that body part (this isn’t ancient Rome after all). I mean the width of your palm, elbow to wrist, etc.
Chris: Using body parts works in a pinch. I usually have a 6” rule in my man-purse when I travel – that’s the easiest gnomon to deal with because you can pick out 1/16”s easily. I know all this sounds a bit wacko, but a good image inventory of pieces you’ve encountered is a huge help when designing. It’s like a sketchbook of other people’s work that you love.
Suzanne: Would you say your experience as a chairmaker plus the image library you have built provides you with a “muscle memory” of seat proportions, back splay, etc.?
Chris: That’s a good way to put it. Once you see thousands of designs you quickly see any design as a collection of angles, segments of circles, boxes and other assorted shapes. It’s a bit like seeing the code in “The Matrix” or the magic point where you think in a foreign language.
Suzanne: To use Peter Galbert’s term “gesture” of the chair the features that caught my eye, besides the crest rail, are the roundness of the arms and the gap in the back. The arms curve around to embrace the sitter plus the surface of the arms are rounded. The gap in the back adds a lightness overall. You posted a photo of a similar chair. In your study of these chairs have you seen this feature very often?
Chris: Sitting in the chair is very much like receiving a hug. There is an amazing compactness to it. It’s so close to you that it feels like an exoskeleton or a carapace.
While the compactness of the chair isn’t common, having the arms threaded by the back spindles is fairly common. As I have been told by our John Brown team, Welsh chairmakers didn’t do much steam bending, so this technique allows them to cut the arms from solid material (no bending) and yet create a pleasing horseshoe shape.
To be honest I was skeptical of this style of armbow until I sat in one. They are amazingly rigid thanks to the spindles below.
Another “birdcage” example.
Suzanne: The original chair was made from elm. You chose sycamore. Why and what do you like and not like about sycamore for this chair?
Chris: Vernacular chairs were generally made from whatever materials were on hand. So that’s the philosophy I use when building chairs. Elm is difficult to get here – you have to find it and cut it yourself. And Dutch elm disease made finding elm a tricky business.
When you look at the materials available around the Midwest, sycamore is a logical choice. It’s a junk tree of no real commercial value. Its grain is interlocked (like that of elm), which makes it impossible to split. (That’s a good thing with seat material.) And it can be had if you ask around.
Like elm, sycamore is an enormous challenge to work. If your tools are not razor sharp, it will tear out horribly. Its density varies greatly depending on the color of the wood. But if you take the time to conquer it, the rewards are spectacular. The quartersawn figure is like a field of stars.
Suzanne: The Hall’s Croft chair has a unique crest rail which I have dubbed the Trinocular. You indicated the form might be an exercise in geometry. Explain, or do we need to bring in Jim Tolpin?
Chris: Well one of the themes underlying the geometry of woodworking tools is that if you set your dividers to the radius of any circle, then that distance can be stepped off exactly six times around that circle’s circumference. Hollows and rounds are one example of how this plays out in our tools. If you want to know what radii a certain plane cuts, you measure the cutter’s width. That width equals its radius. That makes layout predictable.
So the crest rail is three half circles. That means the length of the crest rail is exactly six times the radius of each circle. The radius also equaled the width of the area below the half-circles. So the maker laid out the entire crest rail with one setting of his or her dividers. I don’t know if they were lazy, in a hurry or winking at the person who stumbled on it 200 years later.
Suzanne: I just had a flashback to 8th grade Geometry class.
You made several different crest rails and finally put aside the Trinocular. You also made other design changes to the chair. Describe what you did and why. Did your changes include resizing the chair for the modern body?
Chris: I don’t make replicas unless a customer requests it specifically. I made replicas for many years to get inside the heads of early makers, but I’m at the point now where I sit in a chair and know exactly what needs to be changed to make it suit me and the modern frame.
For my first version of the chair I kept the seat dimensions and leg angles true to the original. I wanted to see how the chair sat because I didn’t get to sit in the original (promise!). But when it came to the crest rail, I had to make changes. The trinoc crest rail was too quirky, low and flat. I made a couple trinoc crests and just couldn’t fall in love. So I increased the length of the four back spindles and carved a curved crest out of solid beech.
I also made some minor changes to the seat profile and arms, but nothing major.
For the third version of the chair, which I’m building now, I’ve changed a whole host of things. The seat is slightly wider and deeper but retains the same overall feeling of getting a chair hug. The rake and splay of the leg angles are all new. I wanted to give it a slightly more aggressive stance and make it more stable in back.
I saddled the seat to add comfort (the original had a flat seat). And I’m working on a slightly different crest rail that will tuck under the sitter’s shoulder blades. Most people will see it as the same chair. But the third one is a different animal.
Suzanne: What did you learn from making this chair? Did making the Hall’s Croft chair help you with your design of the staked armchair that you didn’t get to include in the “The Anarchist’ Design Book”?
Chris: I really love the birdcage effect of this chair’s spindles and will use that a lot in my future work. This chair gave me some clues about how to deal with a staked armchair a la “The Anarchist’s Design Book,” but that chair is on hold right now. I pushed things a little too far with its design and ended up pinching the sitter’s side meat – not good. So I’m finishing up this other chair and am putzing around with the staked armchair. As of now, I’m detaching the arms from the back spindles and trying to see if the chair still feels durable.
Or it will go in the burn pile.
Suzanne: You are also planning a staked settle. Where are you in designing that piece? Have you made a settle before?
Chris: I have made a number of settles over the years. They’re kind of a weird form with their own sets of rules that aren’t exactly like chairs.
I’ve designed this staked settle a couple times, and I think I have it nailed. But I won’t know until I build it.
Suzanne: I am ever hopeful you will build one of those Welsh pub settles with the bacon compartment.
Chris: Anything with a bacon compartment is a good thing. One might call it the “meat pocket.”
Suzanne: What kind of finish did you use for the chair?
Chris: Organic linseed oil and beeswax.
Suzanne: I am slightly obsessed with the Trinocular crest rail. Do you see any other use for it? Door stop? Bookend? Trivet?
Chris: [Redacted – Heavily Redacted] OK, I’ve had two beers. Please excuse that.
I don’t know. They look like a pair of “wooden knuckles” to me. Maybe they could be used in a massage situation. The shape is utterly odd – like the face of a three-eyed frog. I like it, truth be told. But I can’t see it as a component in my furniture – yet.
Suzanne: Do you have any questions for me? I take that back. Chris, thank you! We will have to do this again in another two years.
Chris: Thanks for doing this little chat. It’s actually an interesting exercise to put some of this stuff into words that has been swimming around in my head.
Off to find beer No. 3.
Suzanne: While you enjoy your beer I’m going to update my woodworking dictionary with some meat-based terms: meat clamp, sitter’s side meat, meat bushing and meat pocket.
You can read our first chair Q & A, ‘Naked Necessity’ here.
You can read more about the Hall’s Croft chair here.
Chris did five posts on building a stick chair. Click on the titles listed below to go directly to the article.
We are down to 30-something copies of the letterpress version of “Roman Workbenches.” This book is currently in production and will ship sometime in April 2017.
Brian Stuparyk at Steam Whistle Letterpress and I are trying to create a book that is as perfect as the technology will allow, but no more. Using a sheet-fed proofing press, there are limits to how precisely you can get 16 pages to line up on both sides of a 19” x 25” sheet.
Brian is a maestro with his Vandercook press, so I know that the pages will be in near-perfect registration. But we’ve been negotiating with the bindery, which is accustomed to laser-line precision. That’s not what we’re after with this job.
As with anything handmade, there are small (very small) imperfections that accumulate to produce an object that is not technically perfect, but is aesthetically so. So a page might be 1/32” out of register. Another page might have its image tilted a fraction of a degree. These things are not visible to the eye. They can only be measured with precision tools. But they can be sensed.
Will we succeed? I have every faith in Brian. We’ve worked on a couple of very tricky jobs together (and many non-tricky ones). “Roman Workbenches” is going to be something to see.
The elongation of wood should also be put among the number of assemblages, its application being very useful, given the impossibility of always having wood of the necessary length, or supposing that it is, the defect being that they sometimes are
not of a perfect quality along the entire length, but being corrected by this method.
There are two ways to elongate wood: the first, by notching half of each piece with tongue and grooves at the ends of each piece of wood, which you hold together by means of glue and pegs, Figs. 1 & 4.
The second way to elongate the wood is with Jupiter’s thunderbolts (apparently named thus because the shape of the cuts is a bit similar to that which you give to the gap which you wish to represent). [This is a notched and pegged scarf joint, most likely named because the configuration of the joint looks somewhat like a lightning bolt.]
There are two types of Jupiter’s thunderbolts, one which you make by notching half of each piece and by forming a second notch to receive the [inserted tapered] key. One must note to make this second notch off-set toward the end of the piece, so that the key forced against it finds no resistance in the opposite side of the other notch, and consequently it better draws the joints together [so that it acts like a draw pin], Figs. 2 & 5.
The second way is to trace in the middle of the piece two parallel lines a–b, c–d, which give you the thickness of the notch. After having determined the length of the notch, and having traced the position of the key in the middle, you cut out all the wood from the front of the wood (assuming you are looking at the front of the notch) up to the first parallel line. From the position of the key up to distance e, you make a second notch a–e, such that in each piece, what there is more of takes the place of what there is less of in the depth of the notch, and makes space for the key. For the ends of these notches, they make tongue and grooves, or only an angle, but the little tongues are better, Figs. 3,6 & 7.
This second way is very strong, and is much better than the first because the key bears all the thickness, instead of the other way, which has only half as much. What’s more, a key bearing only half [the thickness] is subject to rolling, and consequently to open the joint. Even if the joint does not open up, the key can be eaten up [word down] and forced, bearing on the opposite side of the groove, which loses its desired effect, see the figures above.
This assembly is very useful and very strong, and is in use not only by Joiners, but also by carpenters, as much for buildings as for ships.
When the entire length of the wood which you wish to elongate is taken up by mouldings, and you cannot or do not wish to make Jupiter’s thunderbolts, for fear that the key and the grooves will not meet up in the mouldings, you use an assembly called a flute, or a scarf joint, which is made in this way.
After having divided the width of your piece into two equal parts, as indicated by line f–f–g, you make the length that you wish to give to your grooves by h–i–l–m. From this line to the end of your piece, you draw diagonals r–o–p–i, and f–q–m–n, some from one side of the line and the others from the other, such that these notches are made in two pieces with much precision, are at the same time a solid and very tight assembly. You must take care that these grooves be made going from right to left, so that when you wish to elaborate with mouldings, they will not be subject to splitting, Fig. 8.
Although I said that you must separate the piece into two pieces to make these types of notches, this rule is not however general.When you have many pieces of mouldings in the piece, you put the joint in the loosening of one from the other, if it is found in the middle, or in the middle of the groove, as you can see in Fig. 9.
When you elongate pieces ornamented with mouldings using Jupiter’s thunderbolts, you should take care to make notches according to the depth of the moulding, if there is not a groove, so that the key is not uncovered, Fig. 10.
You can also lengthen curved pieces, both on their face and on their edge, using Jupiter’s thunderbolts, as indicated in Figs. 11 & 12. For as many pieces as are curved on the face, and for as little as they are curved, you should never make any tenons, because they will become too sliced up, and consequently less solid. You should fit them together by making at the end of the piece a forking of little depth and of the thickness of the tenon. In this forking you make three or four holes for placing pegs or dowels from the tenon that you fit together. These types of tenons are called tenons a peignes [toothed tenons, doweled tenons], Fig. 12.
There you have it, all the different assemblies that are used for the construction of joinery. I have detailed them the best that was possible for me; this matter, lifeless by itself, not being able to be rendered with as much clarity as I would have wished. You will have recourse to the plates where I have illustrated all the different assemblies, either joined or separated, so that you can see their effect better. I have also indicated all those that are hidden by punctuated lines. I hope that for as little as you may wish to pay attention, the demonstration that I have made will supplement that which one could find obscure in this discussion.
