Another zigzag dining chair

[David Dundas]

Unlike the zigzag dining chair described in the May 2007 issue of Woodcraft Magazine, whose joints were reinforced by three locking tenons, this chair's miter joints (Photo 1) are reinforced by a single locking tenon. The mortises (Photo 2) for the tenon were routed perpendicular to the miter face using a 3/8" spiral bit on my mortising jig (Photo 3). The jig is described in the June and August issues of American Router.

The wood used is rose gum (Eucalyptus grandis), and the locking tenons are jarrah (Eucalyptus marginata).

David Dundas

[Rick Christopherson]

David,
As you know, I am intimately familiar with your previous chair design and the mechanical joinery, but I have some serious concerns about this version, especially if you are going to be sell them. First and foremost, a single tenon cannot be a locking tenon, as it requires opposing angles for a tenon to be considered locking. With this new design of your tenon, the entire mechanical strength of the joint is relying exclusively on the tensile strength of the tenon and the adhesion of the joint. Unlike your previous design, this joint can fail catastrophically without warning, or without tearing the surrounding structure apart.

The mitered joint has a great deal of stress on it to begin with, and the single tenon design places a great mechanical advantage on tension of the tenon. When this joint is new, it will support your weight, as I am sure you have tested, but with a little age and repeated usage, it will fail, and it will do so without much warning to the occupant.

If you are not sure about what I am saying, build a joint without glue or fasteners. As a comparison, do the same with your previous design, and compare how they both stand.

[David Dundas]

Rick,

I don't share your pessism about the durability of this chair's joints. The joints have been tested with three people standing on the chair, so there is no question that it is strong enough when new.

If, for the sake of argument, the epoxy for some reason completely failed, I still believe that the brass pinning screws would keep the joint intact. If the joint is to fail, then the screws would have to shear through 1/2" of the 3/8" thick jarrah tenons. Jarrah has similar properties to American maple; and I very much doubt that a load of, say, 200 lbs shared between the two sides of the chair - i.e. 100 lbs on each side - would be sufficient to cause the tenon to shear.

I don't see any reason to believe that the epoxy would fail under normal conditions. The pins should protect it from any excessive stress. I read somewhere that the shear strength of West Systems epoxy is about 7000 psi. The surface area of each tenon on my chair is about three square inches, so there should be a huge margin of safety.

I am not sure what data are available on the long-term durability of epoxy. But I would be very surprised if it would deteriorate to any significant extent in under 20 years. It is after all used in very demanding situations on boats, where the potential for disaster is much greater than in a dining room.

Another possibility would be to glue up the miter joint with an alignment floating tenon, or domino, and then to rout the mortise for the locking tenon perpendicular to the top surface of the seat side-rail, so that the tenon would be no longer be perpendicular to the plane of the miter joint. The problem with this is that the required depth of cut is greater than the 2" maximum cutting length of spiral bits, unless the mortise were located close to the apex of the miter.

David Dundas

[David Dundas]

Rick's post prompted me to try an experiment to see the effect of total glue failure on a frame made in the same way as this chair. Unfortunately, I don't have a photo, since I got too ambitious.

The Z-frame was built without any glue at all, with the tenons held in position solely by the pinning screws. The frame supported my weight (170lbs) without damage. However, when I got my long-suffering neighbor to try standing on the frame with me, the frame did collapse, with the screws shearing through the tenons, and the wood beside the tenons splitting along the grain. However, I believe that this test was unrealistically severe, since total failure of the epoxy is an unlikely scenario.

David Dundas

[Dan Stuewe]

Not to put words in Rick's mouth (or keyboard, as the case may be), but I think his concern is the long term durability of the joints. It isn't a couple of hundred pounds loading the chair for a few minutes, it is 100-200 lbs plopping down onto the chair hundreds of times, along with sliding the chair up to and away from the table numerous times as well (and the sliding away from the table will probably have a bigger load than before the meal ;) ). Anyway, I am not an expert on the strength of joints, but chairs do get quite a bit of, for lack of a better word, dynamic use and that is what leads to joint failure, in my experience.

Be that as it may, the chairs are as striking as usual, and I thank you for sharing them with us.

[David Dundas]

Dan,

I agree that the joints of traditional chairs built using traditional glues (hide glue or PVA) are notoriously prone to failure after prolonged use. But I wonder if there are cases where chairs glued with epoxy have suffered similar damage. I have not heard of any, but perhaps this is because epoxy has only been generally available relatively recently. Alternatively, it may be because epoxy can take the punishment that chairs receive in normal use.

David Dundas

[Rick Christopherson]

Yes Dan, that's exactly what I was alluding to with the comment about the life of the adhesive. This joint will be under high-stress and high-flex every time someone sits down, so over time the adhesive will degrade. I pointed out what I saw from an engineering perspective, so I don't want to belabor the issue.

By the way David, your description of the failure mode of the brass pins was exactly what I described they would do six months ago when I was putting your design into SolidWorks for you. Granted, in this new version of the design, those pins are required, but in the previous design, those pins actually weakened the joint.

[David Dundas]

Rick,

I am inclined to think that you are correct in saying that the pins weaken the joint in my original chair described in Woodcraft Magazine. But I also believe that that chair, like this one, is entirely dependent on the epoxy for its strength. The fact that its locking tenons are oblique to the direction of the force that is tending to open the joint does little to add strength, in the absence of effective glue. Some time ago, I carried out a similar experiment, making a frame in that configuration, without glue or pins, and the frame collapsed under my weight, since the wood beside the locking tenons split.

The question really boils down to whether epoxy is subject to fatigue in the same way as metals are, and, if so, after how long can fatigue be expected to set in. I have zigzag chairs which have been in regular use for two years without any sign of deterioration. But it would be interesting to find out whether experimental data exists on the long-term strength of epoxy in joints subject to high stress.

David Dundas