Box Joints vs. Dovetails for strength

That was interesting. Thanks for sharing.

I’ve always felt a good box joint was as strong as a dovetail joint, but a dovetail joint can be a little showier.

This confirmed that.

-- Rich, Cedar Rapids, IA - I'm a woodworker. I don't create beauty, I reveal it.

Thanks Lee, you are right, an impressive simple test.

If I hadn’t read Rich’s comment, I would not have thought the dovetail would have been weaker.

-- Randy "You are judged as much by the questions you ask as the answers you give..."

Not a very fair test when he’s got nearly twice as many finger joints as dovetail joints which means nearly double the glue surface.

-- Don - I wood work if I could. Redmond WA.

But if you tried to put that many dovetails on the same piece, would it be any stronger? I wonder what the optimal number of dovetails would be in that example. Also, even with more fingers, the box joint was quicker to make.

The way I see it is that I would expect the optimal number of fingers in a box joint would be stronger than the optimal dovetail joint.

-- Scroll saw patterns @ http://www.sheilalandrydesigns.com

I remember this test he did a while ago….

-- Childress Woodworks

Box joint is obviously stronger than dovetail (given sufficient glue area in both cases). No need for testing here.
In finger joint cross sectional area of all fingers at their base, where failure occurs, on one piece (A1) = cross sectional area of all fingers on the other piece (A2), and both =0.5 of the total cross section (At).
A1+A2=At
In dovetail joint regardless of the number of pins and tails or their arrangement either A1 < 0.5At or A2 < 0.5At.
A1+A2 < At
This happens because tails have sloping sides, i.e. portion of the grain is undercut.

In fact, finger joint has a threshold after which failure occurs in fingers rather than along glued surface. After this threshold increase in number of fingers does not compromise strength.
In dovetails, after optimal number and proportion of pins and tails is reached adding more pin/tails will weaken the joint.

i believe i saw this on his site long time ago.
for me dovetails are something of the past when all was done by hand, and when modern glue wasn’t available.
but today it’s purely for visual effect, since a glue joint today is more solid than wood, it’s the wood that will fail first. so with dovetails you put all the stress in the narrowest part of the dovetails.

but then this is kind of pointless, both methods almost overengeneered anyway, theres 100+ year old drawer chests here that have huge drawers joined by simple nailed rabbets, and they hold up just fine!

Viktor – I find your graft interesting and insightful. THank you.

Can you comment on the optimal number of pin/tails?

-- Rich, Cedar Rapids, IA - I'm a woodworker. I don't create beauty, I reveal it.

Matthias Wandel is impressive. Fine WoodWorking did a more exten$ive joint test. -Jack

Great points Skarp. I’m also in the “Does it look nice, and still remain strong?” camp. They’re both strong enough and making anything stronger than strong enough is wasted effort. So deciding which to use should be based primarily on how nice you want something to look and how much work it takes. I don’t think that means dovetails will always win though. Green and Green style has some pretty nice looking finger joints.

-- Don - I wood work if I could. Redmond WA.

Rich, optimal number will depend on glue strength and wood properties (compression vs tensile strength, etc.). In other words I don’t know :-) May be someone cares to conduct an experiment?