Saturday 22 February 2014

Sometimes I'm Just Not In The Mood For A Little Tail...

I was just getting ready to layout the tails for the dovetail joints in each corner of the frame for the portable vise, but as I started to set up the dividers to mark-out some really cool, thin-tailed doves, a thought hit me that maybe what I had in mind was not the best way to go for something like this.

I am so enthralled with dovetail joints, I didn't even think about it. I just started to go at them with my only thought being how they would look, not how they would perform. I decided to hold off for a bit and take a serious look at what this type of joint is all about.

The first item on my new agenda was to create a couple of dovetail illustrations in different styles so I had some visuals to ponder.

The first illustration I created was the dovetail joint I was planning to use, one with thin pins...

I then created a dovetail joint with equal sized tails and pins so I could see the difference...

With my illustrations in front of me, I sat back and tried to remember the different points about them that I had read in the past.

The most common praise for dovetail joints that I remembered was about their ability to deal with wood movement. I came to my first conclusion that this, to me, isn't an issue in this application because the pieces being joined together are all sawn in the same manner and will share the same environment. For me, movement would only be relevant if the dovetail joint is used to joint two boards of different species together, or one board would spend its life in a damp environment while the other in a dry one (how that could happen, I have no idea).

So if wood movement isn't a factor in this application for using dovetail joints, what is?

Well for one thing - stresses!

Somewhere I read that dovetail joints, due to their mechanical design, deal with stresses quite well. Thinking about it in layman's terms, I came to the conclusion that the stresses on the joint come in six directions; forwards and backwards, up and down and out and in. If you have a look at the illustration below, you can see how a dovetail joint deals with five of these six directions, and why it excels at each. The angles of the tails and how they fit to the angle of the pins handle forward stress while the cheeks between the tails deal with the backward stress. The cheeks of both the pins and tails resist the upward and the downward stresses while the cheeks between the pins alone deal with the inward stress. The thing is, I also realized that a dovetail joint can't deal with outward stress.

Thinking about it, I realized that the way a dovetail joint goes together is its achilles heel. The straight pins allow the tails to slide into them during assembly, but they also allow stress to reverse the process as well. If you cut both the tails and the pins on the angle to help the joint deal with the outward stresses, you wouldn't be able to assemble the joint. 

My second conclusion was that, when it comes to keeping the joint together, the dovetail joint excels in five out of the six stress directions.

Is there a way to overcome a dovetail joint's achilles heel?

Yes - glueing!

You wouldn't glue two boards together using a butt-joint because glueing end-grain surfaces, even to a long-grain surface, is a waste of time. A glue joint only works if the surfaces being glued together are both long-grain.

If you look at either of the two dovetail illustrations again, you will note that both the side-cheeks of the tails and the side-cheeks of the pins, where they come in contact with each other, have long-grain surfaces. It is these surfaces that hold the joint together, and while we all like to spread a little glue on the cheeks between the tails and pins, we are really wasting our time as they are end-grain, and have no glueing integrity.

My third conclusion was that these long-grain glueing surfaces are another beauty found in the dovetail's design. It creates long-grain glueing surfaces where none existed before and it is these glueing surfaces that not only add to the joints mechanical strength, but compensate for its achilles heel as well, the one stress direction the joint can't deal with.

So would more tails result in a stronger joint?

It depends!

In the case of the first two illustrations, from a glueing perspective, the narrow pins have the distinct advantage. The traditional dovetail layout has two tails and three pins, but by narrowing the pins considerably, an additional tail and pin can be added. This addition adds 50% more long-grain to long-grain glueing area, which is considerable under any definition.

This was the style of dovetails I had planned to use on the portable vise frame, but looking at the illustrations, I came to my forth conclusion that the joint, with thin pins, won't be strong enough. True, there is lots of that wonderful long-grain to long-grain glueing area, but I think what the thin-pinned joint gains in glueing strength, it looses in pin integrity.

If you compare the two illustrations of dovetail styles again, it is pretty obvious that the 50% addition in glueing area came at a cost of almost 40% in pin material. For the application at hand, I think the thin-pin style isn't up to the task.

In traditional drawer construction, the tails are always added to the side pieces. This is because the mechanical ability of the dovetail joint helps it withstand the stress of the drawer being pulled out thousands of times over the course of the piece's life (foreword stress). It can do this because the tails are wedged-shaped, so the drawer face can't be pulled away, unless, of course, the stress is so great the puller breaks either the tails, the pins or both.

In normal use, a drawer has to endure very little outward stress. The only time this stress direction would effect a drawer is if it is owned by someone like my wife, a lovely lady who loves to stuff 10 cubic feet of clothes in a drawer designed to hold 6, or thinks that they put two drawer pulls on wider drawer faces for looks as any drawer, no matter what its width, can be opened with one pull. Under normal use, however, the long-grain to long-grain glueing surfaces can easily withstand the outward stresses. If a dovetail joint holding a drawer together does give-way, it is either because my wife is using it, or it is a result of glue failure.

The portable vise, however, is a horse of a completely different colour. There will be constant stresses forward on the front joints and backwards on the rear joints, the results of clamping a piece of wood between the two. What also has to be considered, though, is that not all the stock placed in this vise will be square-ended, meaning there will be a considerable amount of stress in the outward direction as well. As the vise is tightened on oddly shaped stock, the stresses will try to rack the vise, putting far more outward stress on the dovetail joint, the one stress direction that the joint is incapable of coping with. These outward stresses can only be coped with by not only the integrity of the glued surfaces, but the integrity of the tails and pins as well.

As this frame is made from 1½" by 3½" stock, two beefy tails will give me enough long-grain to long-grain surface to create a more than adequate glue joint to handle the stresses. If I used the thin pin design, however, I very much doubt that they would be strong enough to handle the outward stresses, resulting in failed joints over time. They wouldn't fail because of problems with the glue, they would fail because the pins would break off. A problem, after all the work I am putting into this vise, that I would rather avoid.

After all this deep thought while contemplating dovetail joints, it is time I headed off to layout some; all with equal sized tails and pins.



1 comment:

  1. You might like reading this :

    see text below fig. 14

    Obviously it only applies to drawers with half blind dovetails.