Technical question on mortise fit

[Carl Beckett]

I am in process of a batch of chairs, and will be using a through mortise on the rails as they extend through the back chair leg. Its hard maple (which I observe doesnt always suck up glue the way other woods do)

Right now the tenons are cut to have a small gap with the height of the mortise (I have just always done it this way - especially any time the tenon is close to the end of a board I didnt want it to expand and pop out the end grain). These are then drawbored, and also glued in place.

But given this joint takes an incredible amount of stress - should I be making the tenon height tight to the mortise to prevent rotation? (the glue is titebond II, and is in shear - and two 1/4" pins on the drawbore). It would be mechanical strength, but not much glue strength because its gluing into end grain.

An alternative would be to taper the back side and use a wedge (or more likely two). I might have to put a shim on the tenon height to gain back some material (a little tricky because they are compound angles... but do-able). But this would mechanically lock the joint into place and rely less on glue or pins.

Im curious what you experts advise?

[jamie shard]

It will be interesting to hear what others say.

One thing I've noticed is 90+% of the time (totally made up number), whenever we ask a question like these, it means that our gut is telling us the design really should be adjusted to build in more strength/stability! :D

[Jessica Pierce-LaRose]

If you're drawboring, won't the pegs help resist any rotation if you use more than one? Particularly if the drawbore is pulling the joint tight against some shoulders on the tenoned piece?

I always aim for well-fitting joints, but I'm surprised at how much slop you can get away with when drawboring.

[Carl Beckett]

If you're drawboring, won't the pegs help resist any rotation if you use more than one? Particularly if the drawbore is pulling the joint tight against some shoulders on the tenoned piece?

I always aim for well-fitting joints, but I'm surprised at how much slop you can get away with when drawboring.

Indeed this is the case.

The drawboring pulls it tight. That went great. The sides are very nice fitting - so not a sloppy joint per se - only play on the ends which I did intentionally and then started thinking about it. And then the cheeks are glued, all the way through, to the leg inside the mortise (so a pretty large glue surface area times 2 - but will the glue ever work loose over the years?).

If the chair is rocked back while being sit on, then the rotation is carried by the lower contact on the shoulder - no problem - but the top shoulder will be pulled away. Thus the load goes through the pegs, or the glue joint (or both). The pegs would be 1/4" walnut, in shear at 4 locations total - 'pretty' strong all by themselves. (I could use larger dia pegs...)

But chairs take a lot of stress. (and I can always imagine some crazy worse case scenario and result in something terribly over designed). Another option is to just try them and see if they loosen up and then worry about it if that ever happens. But would be better to tap into the wisdom and experience here on the 'right' way to do it. :)

237560

[Jim Koepke]

I am not an expert on this.

My father sold maple furniture. He explained to me that the cross pieces were cut so they had a cone shape and with a kerf across the length of the tenon. The tenon was wider at the bottom or end. The mortise was also cone shaped wider at the bottom. A wedge was placed so that it would expand the tenon as it was driven into the mortise.

He also showed me how chairs made like this had wider turnings at the joints. Chairs put together with staples or draw pins didn't require more wood to survive the pressure of assembling the joint.

jtk

[Adam Cherubini]

Written a couple articles about this you may find helpful. The joint you are using is typical of 18th century "Philadelphia Chippendale" chairs. These were typically thru joints, hot hide glued, wedged, and often pegged. The hot hide glue is very strong and can be reversed for repair. Titebond can not be easily reversed and doesn't generally adhere to old tite bond surfaces. Hide glue does.

The chief stress on the joint is the horizontal (lean back) load on the chair back. For this load, the height of the mortise is most critical. The wedges help. Wedges were rarely mroe than 3/16" thick. Typically 2 are inserted 3/16" or so from the ends of the mortise (vertically). Good pictures of all this in my articles.

Corner blocks typically complete the joint. These were typically made up. Many look like SYP.

Sounds like you are doing something very similar, which in my mind at least, is good.

[Carl Beckett]

Thanks Adam,

Very helpful. (many of your articles are locked on copyrighted sites - but I found a lot of info on your chair reference).

Im inclined to keep the pegs, and at the same time add wedges (and taper the mortise out a little to help the wedges pull it tight). It doesnt sound like there is too much worry about overconstraining the joint by using both pegs and wedges. It does mean pulling the joint tight in the vertical fit, but thats fine, and do-able.

Was going to skip corner blocks.

[Carl Beckett]

Ok - Im sold on the concept of adding the rear wedges. So it will be through tenons, drawbored, and then wedged from the back side. And everything glued.

The tenon material is hard maple. The wedge will be walnut (just the theme of the piece)

I did a test sample. The angle is 10 degrees inclusive. The test board thickness was 3/4" - whereas the actual tenon will be 1/2" thick.

Note that the base material split at the root of the wedge notch. Should I care?? It will be hidden inside the joint, and it seems it has to expand somehow so cracking is inevitable

If I hold a straightedge along the edge of the board, indeed the end is slightly wider. Perhaps .005" (meaning not a lot).

Also I wonder if I should slightly taper the ends of the mortise to have a small angle - to allow the tenon to expand outward. But given the very small amount of movement on this test piece - .005" isnt going to do much.

Thoughts?

239966239967

[Pat Barry]

That wedge looks to be 1/8" thickness, maybe a bit more. If you can only measure 0.005" of width increase, then I suspect the wood grains are compressing somewhere because, where did the 1/8" go? I'd lean toward two wedges, 1 near the top and 1 near the bottom. Also, tapering the end of the mortis is very good plan. I just saw this on Rough Cuts with Tommy Mac episode of a trestle table he made.

[Carl Beckett]

Yes, the wedge is approx 3/16 (havent measured exactly). But then I cut a notch in the mortise at the same angle. By no means am I moving 3/16" - again, more like .005" to .010" (max) of actual expansion across the width. I could drive the wedge deeper and it would expand more, but would split past where I cut the notch.

Maybe my question is this: How much interference in the wedge should I design for? Force it out 1/16" (a long ways to go to get to 1/16)

And yes I will use two wedges spaced at top and bottom - was just doing a test here.

[Matthew N. Masail]

We just moved apartments so I don't have internet and Im using my phone to replay, I just wanted to say - if you drill a hole in your tenon, at the bottom of the kerf you make for the wedge, you'll spread the tension and could avoid splliting

[Carl Beckett]

Thanks Matthew, a good tip.

The other option I was considering (and will do a test on), is to just saw it deeper before using the wedge. This will in effect allow it to spring open without splitting (maybe....).

Im sure I am overthinking it all, but its a critical joint and am wanting to learn from the phenomenal knowledge here.

[Terry Beadle]

If you go with this spring open wedging, be sure to drill a slightly bigger hole in just the end of the split so that the split won't continue splitting over time.

[Jessica Pierce-LaRose]

I have a hard time guesstimating from the pictures the actual size/taper of the wedges, but I feel like the couple of times I've wedged joints, I've gone with a more tapered wedge - I just kind of eyeballed it, but I feel like I went with something close to what those pine shims you buy pre-made and packaged at the hardware store come like. I have no idea if this is enough, I've only done this once (and that piece ended up in the burn pile for other, non-mortise related problems) but it seemed to work well enough.