Pick Yer Brains on Chairs... Please.

[Argus]

I'm in the process of thinking through an upright armchair. A wainscot-type design in principle, but with the back raked slightly instead of being upright and the seat narrowing to the back.

I've no problems with the frame and construction, but I'm not decided on the type of joint that unites the arm with the raked back.
What I don't want to do is to screw or pin the arm from the back, which is the obvious solution, but I do want to have a strong mortised construction where the flat arm joins the back on the front and side faces. The whole thing is in effect a set of compound angles because the front seat narrows toward the back.

I'm sure that it's been done before, so can I ask if there are any pointers on the type of joint?

All best and thanks in advance.

PS..... I'm using some very hard Oak.

 

[marcros]

have you seen Derek Cohen's chair build. Whilst not the same, there were some complex angles and joints- well worth a look

 

[rdesign]

a loose mortice can help solve awkward connections.

Regards Richard

 

[yetloh]

I agree with Richard. I'm making four of these chairs -

all slightly different. There are te joints in the back alone and all have loose tenons. Dominos except for the rear seat rail where the mortices are routed. There are no straight lines anywhereandd the joints are at various angles and bringing them in so that all are tight is a considerable challenge - doable with loose tenons, it would have been impossible, at least for me, with traditional mortices and tenons.

Jim

 

[custard]

I'm in the process of thinking through an upright armchair. A wainscot-type design in principle, but with the back raked slightly instead of being upright and the seat narrowing to the back.

I've no problems with the frame and construction, but I'm not decided on the type of joint that unites the arm with the raked back.
What I don't want to do is to screw or pin the arm from the back, which is the obvious solution, but I do want to have a strong mortised construction where the flat arm joins the back on the front and side faces. The whole thing is in effect a set of compound angles because the front seat narrows toward the back.

I'm sure that it's been done before, so can I ask if there are any pointers on the type of joint?

The simple answer to your question is that chairs are almost always morticed and tenoned. The most critical joint is where the side rail joins the back leg, this is where chairs almost always give way first, and once this joint loosens then all the other joints will progressively give way too. Consequently when you design a chair this is the joint that gets priority, and every other joint is secondary to the needs of this one. As Richard pointed out there are some advantages in using loose tenons, chiefly it frees you from the "angled tenon versus angled mortice" dilemma.

Peter Follansbee has written extensively about constructing this style of chair, and his work should help guide you through the process. Here's a photo of one of his chairs that shows the jointing detail relevant to your question,

I've never made a wainscot chair, but I've made plenty of jointed chairs that broadly follow the designs introduced from the mid 18th century onwards. This is where still more angles were introduced to add strength and comfort whilst reducing weight. The rule of thumb for these later chairs is that building in arms generally increases the production time by at least 50%, and from my own experience that's about right!

It's a cracking project that you're undertaking, so please keep us posted and the very best of luck!

 

[Argus]

Thanks to all for the responses….. all of it much food for thought.

I should have been clearer on two things – first, it’s all done by hand, secondly the chair is already made, compound mortise/tenon joints and all, to the extent that it is at the unglued fit-up stage, minus the arms of course, which prompted the question regarding the arm mortice with the back. It all seemed a bit too complicated for what I needed. This was the point where I stood back for some thought and inspiration.

The chair is, to put it into context a hybrid between an Arts-and-Crafts style of chair, one of Stickley’s designs and my own interpretation of the two.
I’m indebted to Custard for reminding me of Peter Follansbee who has dropped off my radar since he wandered off his very interesting Jacobean-style furniture making activities some while ago, but looking at his chair I think that I see the solution.

In general shape and form my chair is very similar, without of course, the solid back (mine’s slatted) or the extensive carving. The raked back is similar, as is the rise of the main front legs which I intend to terminate with a through tenon and a raised, domed head above the arm.

Follansbee’s chair is a replica, I believe and has an arm that is substantial and inserted into the back with a longitudinal tenon. I had envisioned a flat, wide arm on mine about one inch in thickness and about four in width – hence the concern about the strength and shape of the joint section where it met the back. The solution, I believe, is to provide a square arm section that is mortised into the front leg section and the rear leg then to mount the wider flat arm on top, relying on the main under section for strength. Most of the under-arm would be hidden by the wider top.

I’m going to do a bit of experimenting next to see how it will work, how it looks and feels before I commit myself with chisels and saw, but thanks to all for the help.

All best

 

[custard]

Interesting project. I don't know if this is your first chair or your hundredth, so if I'm teaching granny to suck eggs then ignore, but a couple of thoughts that may be relevant.

-Back slats or back bars are rarely easy, but there are some techniques that help. The key problem is getting really tight, gap free, glue lines where they joint into both the back seat rail and the crest rail, this is tricky because by this stage you're often off from the rod or plan by a mill or two, also there are often angles or compound angles that complicate the situation. One option is to construct them as a blind tenons and make the mortices a bit deeper than necessary, that way they'll "float" and take up any slack. The other option, which personally I prefer as it's neater, is to make the crest rail "float" by making the mortices in the back legs about 3 or 4 mill wider than the tenons in the crest rail. This allows you one or even two rounds of scribing the shoulders of the tenons to achieve a perfect fit. If it's a Stickley style chair then I'll assume the back bars are fairly straight. If so this massively simplifies the glue up as you can run some light sash cramps from the crest rail to the back seat rail to bring the joints up tight. If the back bars have significant curves you may need to make a jig/fixture to bring pressure to bear where it's needed. That can get pretty involved as in these arrangements,

In any event the best way of proceeding with the glue up (especially if you're using quick setting PVA) is to glue up the back bar/crest rail/back seat rail assembly, only loosely attaching the back legs to keep everything true (i.e. no glue), then glue up the back legs in a separate operation. If I'm honest I often chicken out and use a slow setting UF glue even though I was taught that UF glue is to be avoided like the plague on chairs as it's too rigid and won't soak up the flexing that chairs are subject to.

-Attaching arms is normally done in one of two ways. With the front legs extended and the arm anchored directly onto them. Or with a separate component that joints into the side rail. There are advantages with both methods. By extending the front leg as you propose the job is slightly simplified and it's slightly stronger, but (there's always a "but" isn't there!) the seat needs to be about 50-75mm wider at the front or else it makes for a very restrictive experience for the user. Sometimes the aesthetic of the chair doesn't allow for this, or you want the chair narrower to squeeze more chairs around a table. Then you have no choice but to anchor the arm onto the side rail via another component, and jointing that component into the side rail can be a bit of a pig as you'll normally use a big, thick sliding dovetail that needs to be bang on accurate if it's to last and fit correctly.

-The wide, flat arms that you're proposing will make the chair wider still (they're generally flush on the inner surfaces so all the extra width is outboard), and the issue as you've correctly identified is that then the outer edges of the arm cantilever out in space. Given that people generally press down on the arms as they raise themselves out of the chair this needs to be pretty sturdy if it's going to last. Personally I wouldn't use the method you've suggested (a supporting component beneath the arm) as it won't address the cantilevering problem. Just my opinion, but I'd use a corbel, certainly one fastened to the front leg, maybe even a second fastened to the back leg unless you taper the arm thinner at the back. If you google "Morris Chair" you'll see how this arrangement works.

Good luck!

 

[Argus]

Thanks, Custard. I'm not entirely new to chairs, I make a few from time to time not quite as classy as yours, by the looks of it, that's a nice comfortable dining chair!

I had identified the stresses on the outboard extremities of the arm and also factored in a corbel at the front; I was considering a slightly smaller one as well (from an aesthetic point of view) at the rear where the overhang would be less, but still subject to stress. My main concern was with the strength of the meeting joint of the arm at the back, as I did not want to fix it through the back rail with screws. It was all hinging, if that's the correct term, on a combination of what I thought was a rather insubstantial tenon/miniature trench joint at the back rail. The more I thought about it the less I liked it.

I think that I will draw up a full sized sketch and see how it looks and how it stacks up with the expected loads. To add a fourth dimension, I'm now considering mounting the whole thing on rockers.........

Thanks again to all who have contributed to my line of thought.

All best