Chair stretcher joinery question

[Sean Hughto]

The dowel thread got me thinking about chair stretcher joints, in part beacuse that was the actual focus of Hoadley’s article. Also, I’m in the midst of making a sort of welsh stick chair a la John Brown, though I’m planning to add stretchers to the undercarriage, whereas many of his chairs lack them.

Chair stretcher joints are notorious for coming loose. I know Dunbar beleives in having them be oversized so as to functionally "push the legs apart" rather than hold them together. I have heard of using green wood and spoon bits to essentially make a tenon that gets locked when the leg shrinks. But I’m wondering why simple mechanical things have not been used as a mater of course. Like:

- a simple dowel that pierces the leg through the tenon at 90 degrees - think draw bore

- making the tenon go through the leg (perhaps with the last half of its length narrowed to half the entry circumference) and then wedging it

- also too - further to the critique of round M&Ts - why not square tenons on the stretcher in squared mortises on the leg?

So my question is, are there obvious reasons why these are bad or unworkable ideas?

I can imagine worries about weakening the leg, but I would think that making turnings where the stretcher holes fall at maximum leg diameter spots (as more or less usual) would leave plenty of meat for strength.

Thanks for any thoughts or insights.

[john brenton]

I wouldn't give my own opinion, but I think I had read somewhere that in the minds of classic chairmakers or the past and today, if the delicate chair is well made history has shown it will hold just as good as anything else, but if it's going to break down it's going to break down. I wish I could remember where I read it, but it was a well known chair maker. He was addressing a "why not beef it up?" question similar to yours.

Those welsh stick chair legs are pretty scrawny, I don't know that anything (other than heeding Dunbar's advice) can keep it together. You can't control who sits on it or how they sit on it.



The dowel thread got me thinking about chair stretcher joints, in part beacuse that was the actual focus of Hoadley’s article. Also, I’m in the midst of making a sort of welsh stick chair a la John Brown, though I’m planning to add stretchers to the undercarriage, whereas many of his chairs lack them.

Chair stretcher joints are notorious for coming loose. I know Dunbar beleives in having them be oversized so as to functionally "push the legs apart" rather than hold them together. I have heard of using green wood and spoon bits to essentially make a tenon that gets locked when the leg shrinks. But I’m wondering why simple mechanical things have not been used as a mater of course. Like:

- a simple dowel that pierces the leg through the tenon at 90 degrees - think draw bore

- making the tenon go through the leg (perhaps with the last half of its length narrowed to half the entry circumference) and then wedging it

- also too - further to the critique of round M&Ts - why not square tenons on the stretcher in squared mortises on the leg?

So my question is, are there obvious reasons why these are bad or unworkable ideas?

I can imagine worries about weakening the leg, but I would think that making turnings where the stretcher holes fall at maximum leg diameter spots (as more or less usual) would leave plenty of meat for strength.

Thanks for any thoughts or insights.

[David Weaver]

I'm not a chairmaker, but they must've come to the conclusion that for the amount of work involved and for the results, that a round M&T joint and a (spoon?) bored hole was the best compromise. They may have felt that a chair that needed some attention to a loose spindle was still a chair with an indefinite life, and the repair probably quick and routine.

(that didn't keep people from rapping kids who leaned back on chairs on the back of the head when I was little "you'll loosen the legs, don't lean back!!". )

From a theoretical standpoint, the wedged ends seem the most satisfying solution to me, still a round mortise just so it looks nice. That way, you wouldn't disturb any of the long grain that you would if you pinned the tenon. I think I would wedge the wood with the same wood as the tenon so that it didn't show up, but the solution from a mechanical standpoint seems viable - of course, only in the situation where a through tenon is OK.

[Jim Koepke]

My information on this is limited to what my father told me.

He sold furniture and a lot of Windsor style maple chairs. I still have some that I grew up with around our dinner table. The rest of them are serving duty with one or our children's family.

What my father told me is the stretchers with a bulge in the middle where the cross piece is inserted are drilled with holes that are tapered to be bigger at the bottom than at the entry. When the parts are assembled a wedge is in the mortise to spread the tenon as it is driven home.

He also told me the chairs without the bulge most likely are either just glued or have a pin driven into the tenon.

Maybe I should take some pictures later. Nothing shows this like pictures.

jtk

[Sean Hughto]

Fox wedging, right?

[Jim Koepke]

Fox wedging, right?

Yes, I had to Google > fox wedge < to understand the term.

The images show it clearly.

The term seems to apply to both through and blind mortise and tenon joints.

jtk

[Brian Millspaw]

Hi Sean,

In John Brown's book he stated that he only omits stretchers if he is using a one piece elm seat at least 1 3/4" thick.

I've made around 20 chairs in the last year and a half, so I'm just beginning, but here are my observations :

1. I made a stick chair without stretchers, with a glued up walnut seat. I think that this is OK as long as the rake and splay isn't too extreme.

2. I pinned the outer spindle on a fan back type chair with an 1/8" dowel and later the spindle broke right where the dowel was

3. I've wedged all 3 stretchers instead on just the medial stretcher. I think that this works well, although it doesn't look as good. There is a picture in the beginning of Drew Langner's book of a stick chair done this way.

4. A square mortise and tenon would be a lot more work, and because the tenons are always intersecting the leg at an angle, they would show their "squareness" on the obtuse side of the joint.

That's all
-Brian

[Sean Hughto]

Hi Sean,

In John Brown's book he stated that he only omits stretchers if he is using a one piece elm seat at least 1 3/4" thick.

Yeah, I saw that. I am actually using elm, and it started out at just under 2" rough sawn, but by the time I flattened it , and then scooped out the seat (not terrible deep, just moderately saddled), I think the mortises will not be more than 1 1/2". I have seen some place their legs in the full thickness outer rim - nonsaddled - part of the seat, but intend to have the legs more in the saddle area. A long way of saying that he wasn't very specific about where that 1 3/4" had to be. I assumed the mortise/hole.


I've made around 20 chairs in the last year and a half, so I'm just beginning, but here are my observations :

1. I made a stick chair without stretchers, with a glued up walnut seat. I think that this is OK as long as the rake and splay isn't too extreme.

I tend to favor a decent amount of rake and splay. It gives the chair a nicer look to my eye.


2. I pinned the outer spindle on a fan back type chair with an 1/8" dowel and later the spindle broke right where the dowel was

Interesting. I think a spindle might be more likely to have this issue as there is just not enough meat there.


3. I've wedged all 3 stretchers instead on just the medial stretcher. I think that this works well, although it doesn't look as good. There is a picture in the beginning of Drew Langner's book of a stick chair done this way.

I've seen that picture. I think it looks okay. Did you step the tenon down in diameter on the wedged side? Did you taper the exiting mortise (use a reamer to make it trumpet belled a bit?).


4. A square mortise and tenon would be a lot more work, and because the tenons are always intersecting the leg at an angle, they would show their "squareness" on the obtuse side of the joint.

Definitely more work and more demanding in that the parts can't spin at all to help assembly at glue up.

Thanks for the thoughts. I'd love to see some pics of your chairs. By 20, you must be getting it down.

[Brian Millspaw]

This is the chair that I wedged all the stretchers through tenons. Because of the extreme slouch and pretty good rake, I thought that it might help.

197576197577

I didn't ream the outside or anything. The exit hole always seems a bit bigger than the entry hole, so I thought that that was good enough.

Here's some others

197585197584197583197582

-Brian

[Sean Hughto]

Really nice! Very inspiring. I like the black one especially, but they are all terrific!

Are they as comfortable as they are beautiful?

I think my first one is bound to be a far bit rougher/more rustic then these - sort of like as made by a one-eyed Welsh farmer ... with only one good arm .. and only an ax ... in the dark.

[Kevin Adams]

Brian, very nice work. You must have worked with Joe Graham at one point...I can definitely see the organic influence. I like bold and different...good work. What kind of glue are you using?

Take care,
Kevin

[David Keller NC]

Sean - One aspect of the traditional Welsh stick-chair is worth mentioning here; I suspect that such a chair made without riven legs and without stretchers would run into difficulties, particularly if the legs aren't oak or hickory (which strongly resist shearing across the grain, much more so than maple or cherry). One other thought is that it would make some sense that one would need a relatively strong wood for the seat to avoid breaking the glue and enlarging the mortise in the seat, Drew Langsner notes that the plank seat of a stick windsor should be "no softer than tulip poplar", and they're most often of oak, ash or elm.

[Tom McMahon]

Brian really great work I really like the continuous arm rocker.
Sean many if not most 17th and early 18th century chairs were pegged. Pegging went away with the industrial revolution. I recently restored a set of Louis XV chairs that had all square tenons.

[Sean Hughto]

I keep coming back to look some more. Brian, your chairs are really excellent. Thanks again for posting the pics.

Maybe you could make a chair-beginner happy and tell me that your first few weren't this excellent? It would give me hope. ;-)

[Tom McMahon]

I agree every time I look I see something new. The bent spindles on the side chair are great.

[Jim Koepke]

An inspiration to anyone who wants to make some of their own originals.

Fantastic work Brian.

jtk

[David Keller NC]

Indeed, very impressive work, Brian. The dark one (walnut?) has sort of a science-fiction arachnid look to it. It's not easy applying new/different aesthetics to a form that has such ancient roots.

[Bob Glenn]

The usual reason stretchers loosen is because the leg/seat joint fails. This leads to the stretcher joint failure. Manufactured chairs usually do not have a through mortise and tenon joint, only stopped mortise in the seat with the leg glued in the hole. Sooner or later, this joint will fail with use. Hand made Windsor chairs use a tapered mortise and tenon joint for the legs. I make my stretchers a quarter inch longer than the inside measurement of the opposing mortises and do not use any glue in the stretcher/leg joints and have never had a problem.

[Eric Brown]

Don't know if it will contribute much to this thread, but I want to share a picture of an old rocker my grandparents got from a hotel in Dayton Ohio that was remodeling. I'm estimating they got it used over sixty years ago. What I'm trying to show is a chair designed to withstand hard long usage. And it did! I don't see anybody today making anything like it. Notice the curved round stretchers and the carved areas on the back. Pine cone tops. Has a round leather seat. Enjoy.

Eric

197915

[Brian Millspaw]

Hey Sean,

Rest easy, my first chairs were not so good. The first chair I made was at Joe Graham's chair camp. That one came out great. He makes everything easy for his students. Joe offered all of his patterns to copy, but I thought that I should figure it out for myself. The first chair that I tried alone was the high back stick chair from the Langsder book. A spindle broke and I had to cut it down to a low back. Then I tried the bowback side chair. I was never happy with the leg angles, so I started to experiment. Next I made the "slouch" chair that I pictured first. That one looked good and I was on my way. All of the chairs are comfortable, I consider that the goal.
Joe graham and I use titebond II for glue

-Brian