Shaved Windsor chair; take 2

Thursday I resumed work on my 2nd version of Curtis Buchanan’s democratic chair. The first piece of furniture I assembled this year was my initial attempt at this chair back in January. https://pfollansbee.wordpress.com/2020/01/06/finished-my-curtis-buchanan-chair/

For my first Windsor-style chair in decades, I was happy enough with that one. Which is different from “I was happy with it.” One problem I had was the legs splitting at the joints.

splits in rear leg & side stretcher

I suspected my tenons were too large. I talked to Curtis about it, and if I remember right, he said because of the steep angle (say where the side stretcher meets the leg) you can get splitting. Suggested yes, make the tenons a bit smaller than I might in a ladderback chair. I was using white ash. I think another factor was the auger bit I used. It has a thick lead screw that might have contributed to the splitting.

For me, one of the most glaring problems was boring the mortises directly in the tangential plane of the legs. And the orientation of the leg is dictated by the orientation of the seat. In this case:

  • the long fibers (the “grain”) of the seat run front-to-back.
  • The growth rings in the legs’ tenons (the whole leg actually) are therefore oriented so they run perpendicular to the fibers in the seat. 
  • The leg is split so the wedge is also perpendicular to the fibers of the seat. 

This makes the front view of the front legs (and back legs) the tangential plane. And it means when boring the mortises for the side stretchers you’re boring directly into the growth ring plane. Where ring porous woods split very easily. 

BUT – I hate looking at the tangential plane of ring porous hardwoods like oak, ash, etc. And on my first democratic chair, the front of the front legs (and back of the back legs) is this plane. 

front leg shaved chair, Jan 2020

I thought about switching the leg orientation 90 degrees to the “usual” format. Then you wedge it just the same. This puts the leg’s radial plane, which changes less than the tangential plane, running in the direction in which the seat moves the most from one season to the next. So in a worst-case scenario the seat could split I guess, if it shrank a lot versus the leg tenon which wouldn’t shrink much. I think if you use a softwood seat like white pine, and hardwood legs, the seat will compress before it will split from the legs. I bet any problem would be at assembly, not afterwards. I could, of course, be wrong. It’s been a long time since I was really a chairmaker of this sort.

In the usual orientation, the leg’s radial plane lines up with the long fibers of the seat. The seat does not shrink in this direction at all; maybe the tiniest fraction of an inch. If you were to make the change I was thinking about, the radial face would now be the front view of the legs. Better visually for me, but now you’d still be boring into the radial face./growth ring plane, where ring porous hardwoods also split very easily. 

In the ladderbacks I learned from John Alexander and Drew Langsner, we positioned the posts’ growth rings at an angle to the rungs…they coined the phrase “post and rung compromise.” They didn’t make up the concept, it came from studying old chairs. The reason for it is to reduce the chance of splitting the legs when driving the rungs in – the mortises are bored between the growth ring plane and the radial plane. Below is one of Chester Cornett’s chairs, showing the front post oriented with this post & rung compromise. The radial crack bisects the angle between the front & side rungs. But you see how neither of those mortises are in the radial or growth ring plane. Some of Chester’s chairs that I saw used sassafrass posts, this one was white oak, very slowly grown.

I decided to try something this on my democratic chair #2. I was moderately successful. The legs & stretchers on this chair are also white ash, and had grown a bit too slow for this application. Too many growth rings make them a bit weak. I oriented the stretchers in the usual way – their hardwood-to-hardwood mortise & tenon joints seemed more critical to me than the hardwood/softwood leg-to-seat joints. So the growth rings on the side stretchers are parallel to the floor – those on the center stretcher are perpendicular to the floor. This means the mortises for the center stretcher are bored into the radial face, directly in the growth ring plane.

boring the mortise for the center stretcher

One side stretcher cracked slightly like on the first chair. I keep learning.  

But it was the legs that I moved around. I re-positioned them just a bit, turning them in their tapered mortises so the growth rings were angled to the fibers in the seat. The wedge, and the wedging action, are still perpendicular to the long fibers in the seat.

rear leg in the seat

This way I was boring the mortises for the side stretchers one facet off from the growth ring or radial plane.

It worked pretty well, one leg has a small split, but that might be more due to the slow growth rings than anything else. Next time, I’d choose a faster-growing log, and I might turn the tenons rather than spoke-shaving them. It’s better than first time out, almost to where I’m happy with it. 

8 thoughts on “Shaved Windsor chair; take 2

  1. Don’t most period Windsors have the long fibers oftheseat side-to-side? Except maybe some seats with tail braces, I’ve seen diagrams with the long grain running on a diagonal.

    • yes, Trent – every which way goes the seat fibers. I haven’t got out Nancy Evans’ books in a while; I think most side-chairs have the fibers running front to back. But I’m an amateur in windsor chair work.

    • Thanks, Gerry – I forgot about the bit. On this 2nd chair, I switched to an old Stanley Power-bore bit that we used in making Jennie Alexander’s ladderback chairs. No longer made, a comparable one is the Stern bit made in Germany. (but a web search right now shows they’ve “redesigned” this bit and I’m not sure how the new one would work in a brace. And I’m not about to go testing it. I have enough to do.) Center bits are a good idea too.

  2. I spoke with Drew Langsner last night. First off, he’s doing quite well recovering from his surgery. Walking a treadmill 3 times a week, getting stronger, etc.

    And had some thoughts about my Windsor chair post. He’d left a comment, but it got lost in the ether. Never had a leg split after his very first JA chair; suspects the tenons were too large. Reached the conclusion many years ago to always turn tenons – even on shaved ladderback chairs. A shaved tenon, no matter how carefully made, is a polygon and all those small facets affect the bond between mortise & tenon. Turned tenons mate with round mortises the best. Barely oversize, said maybe 1/100th of an inch. Stopped flattening the sides of the tenons.

    Drew’s main points were to always orient the growth rings perpendicular to the seat fibers. What happens below doesn’t matter, says he. The seat is not going to shrink that direction, and the tenon will shrink the least in that direction. Make sure the tenon is bone-dry.

    There must have been more, we talked for a long time, but that’s what I remember.

  3. As you know, High Wycombe Windsor chairs, for the most part, used local Beech for the undercarriage and superstructure of the chair set into wide Elm seats. That’s when there was Elm to be had in this country – alas, no more.

    There’s a method I came across in a book some years ago…..(the title has vanished from the memory, I’m afraid) where a ‘wet and dry’ method of glue-less assembly of the leg/stretcher arrangement was made.

    The leg/stretcher joint involved turning the tenon, as you suggest, on a lathe and turning a small groove, a key-way if you prefer with your chisel, about 1/8 inch or so deep positioned about halfway up the inserted tenon section.

    This stretcher/tenon stick was well seasoned and dry when this was done. It was inserted into a freshly riven and formed leg/mortise that was wet. Over time the leg section shrank onto the stretcher as it dried and a small section inside the joint deformed to fill up the key-way, thus locking the glue-less joint in place.

    ……..If I could find that book or article, it showed one of these joints dissected longitudinally showing the inherent strength in the union.

  4. What if you orient the seat grain side-to-side? I may be wrong, but that would allow the leg and spindle grain to be in the optimal direction for not splitting. I would mean that the rays are visible to the front, but it would be stronger. Thanks for sharing.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s