Questions Answered – Spring Edge Edge-jointing Boards

Question:

Paul (and Team)

I was curious if you could give us your thoughts on spring joints.  I’ve seen them mentioned in various places when discussing panel glue ups.  I think I understand them in principle.  By creating the slight concavity in the middle of the two boards to be joined, the ends are automatically pulled tight during glue up.  My questions have to do with the purpose of such a joint and if they are practical for everyday use.  Does this make for a tighter joint?  Does this help to reduce the number of clamps needed for a glue up?  Is there any benefit in strength of the joint compared to joining two parallel edges or even match planed edges?

Thanks for all you and your team do!

Adam

Answer:

This was first presented as an article in one of the famed US woodworking mags back in the early 90s I think, but the concept originates from the days and centuries before we had screw-threaded cramps or clamps. The idea is to create as you say an even convex along two long board edges and then glue along the edges to form wider panels.

DSC_0010

The panels can then be clamped using only two clamps or you can use two timber dogs or nail dogs as they are also called to pull the two ends tight, which automatically pulls the joint gap-lessly together along the centre section and indeed along the whole length of the joint line. Of course the convex only has to be small, the article showed a lot, but any slight, slight belly works and the more belly you have the greater the applied or necessary pressure to either end to close up the gaps. The problem that can occur with too large a gap is the boards can result in a dishing hollow, so get the camber as slight as possible if you choose this method.

The practicality of it was that you didn’t need too many clamps; two only per panel as I said.

Dogs were indeed commonly used for such glue ups; I have several pairs and they can be useful to have because some times clamps don’t work in some situations.

The dogs had the angles inside and this effectively applied pressure on the extreme end of the board and pulled the two together as shown above.

Also, it’s good to remember that screw-threaded clamps were not at all always as common to woodworkers as they are today and so were not commonly used in earlier centuries. Wedges and dogs took care of many glue ups; and also remember that that was the age of no alternative glue to animal glues such as hide and fish glues either. Glues such as PVA and epoxies are new kids on the block. These old glues were different in that they ‘snatched’, which meant that during the cure the glue ‘pulled’ surfaces together.

The most common of all methods for jointing for edge joining was not clamping, wedging or dogging but simply rub-jointing. This is where we rub the two glued edges to be joined along one another moving one back and forth against the other, which is held in the vise, until the glue is evenly and thinly dispersed. During this process the glue gets as thin as possible between the surfaces and then, at a certain point, the glue ‘grabs’ or ‘snatches’ and the parts no longer move.

Leaning two or three sticks inclined and sighted in against one another (to ensure no twist in them) against a wall was a resting place for the glued boards to stand leaning in toward the wall on edge but flat against the sticks. The boards were left unclamped until the next day when full cure was achieved. Of course this latter method, being the more common of, all meant that the board edges were first trued by plane exactly and with no gaps or convexed edges. The practice of convexed edges may well have been common in earlier centuries purely because of the lack of screw-threaded clamps or even nail-dogs. Dogs work remarkably well, but of course they leave their telltale square holes in the endgrain of the boards.

For some work this is of no consequence.

Once clamps or dogs are applied the boards at the other end open dramatically because of the compression of the cells. It’s a good idea to have the clamps or dogs in place and started so the boards don’t part too much as shown.

Below you can see that the method is effective as the nearest clamps are 36″ apart and yet the glue can be seen in the mid section.

I suppose the point here is is convexing the board or boards better than clamping? My answer would be probably no. I have always found it best to go the extra mile and remove all contention between parts so long as it relies on me. Yes the boards will hold, but it may well have been more a lazy way with good and plausible excuses. We generally work to move the materials we work to their extreme limits as say we do in the fitting of dovetail joints and mortise and tenons and also in making musical instruments such as cellos and violins, guitars and so on. That is, we want extreme fits of perfection for tightness without splits or too much pressure and so on. It’s always a thin line, pun intended. After we have created the harmony between the parts they are united in the common cause of serving the owner and it’s at this point when what we make is subjected from here on to the highest levels of stress imaginable. A chair is scooted, cocked, leaned back on with an excess of 150lbs every day and all day in some cases. And a violin with strings taut and played is stretched to its most extreme limits in the hands of a maestro. Creating harmony relies on the crafting artisan to ensure he has done his utmost to remove any and all contention so that what he or she makes can indeed withstand the pressures of life.

20 Comments

  1. Wow, this “blew my mind”. I had always thought “spring joints” are the exact opposite to what you describe i.e., a concavity in the edge. This way the ends meet and it’s the centre that you draw together with a clamp. I could have sworn I’d read this in Robert Wearing’s The Essential Woodworker, but I can’t find it (he simply advises a dead straight edge). Next I tried The Joiner and Cabinetmaker, but no luck there either. Thus, I think I got the idea from Fine Woodworking magazine and/or Chris Schwarz. I suppose it’s the same principle, just turned on its head, but your thesis that screw clamps were relatively rare makes sense. Perhaps this will become another woodworking contest on the primacy of sources.

    1. You’re not wrong, Alex, Concave was equally common but harder to achieve even clamping on hollows that often needed more than one or two clamps but three to close off the gap because the physics in the pressures are applied quite differently.
      Also, remember that the timber dogs were extremely common too and they only worked on convex or straight edges.

  2. The same here. I’ve never seen yet a sprung joint with convex edges, always concave. On a short panel that would even allow just one clamp, instead of two, and it should prevent the ends of the joint opening overtime. No idea how much an old wives tale that is.

    For example: http://books.google.nl/books?id=jPsDAAAAMBAJ&pg=PA28&lpg=PA28&dq=sprung+joint&source=bl&ots=EMXzOr53ww&sig=ByP5vErqKUZlFN1OeMqpkMm7T6s&hl=en&sa=X&ei=Qc9QVKHAHIXYPJSxgYgH&ved=0CEsQ6AEwBQ#v=onepage&q=sprung%20joint&f=false

    1. It sounds as though this is a more common practice in your sphere of woodworking. I’ve not seen this done since I was a boy apprentice in my teens except in a couple of for-your-interest articles. Most of the woodworkers I know have a dozen or more clamps floating around these days so it has become quite uncommon and unnecessary in our woodworking culture. It’s handy to know about for those occasions when clamps are scarce though, and there are still hundreds of countries and cultures with millions of people that don’t have the wealth our culture offers but do have access to the internet so I am glad to put these articles out there to help the unknown situations people struggle under. Most people can make a few dogs to dog their joints and as I just pointed out to Alex, timber- or nail-dogs only work with convexed edges and they were commonly used.

    1. This is definitely the first time I’ve ever heard this! Most people I’ve talked to (including a well known teacher and past recipient of the SAPFM Cartouche award) would describe the joint as both edges having a slight concave, not a convex! One reason I’ve heard stated is that it helps compensate for seasonal movement which occurs closer to the end grain then the center of a board.

  3. I read somewhere that a spring joint was conCAVE so that as the wood dried the two ends wouldn’t separate and leave the panel joined only in the middle. Have you heard of this?

    1. There is no benefit that convinces me one way is better than the other. Both work fine.

  4. yes! this caught me by surprise also i was to believe a slight concave along the centre would be the best method,however from my own experience and humble opinion i believe the wider the boards to be jointed the more need for a perfect meeting (no gaps) .unlike for example the workbench top when i was gluing 4″x 2″ face to face a slight gap either way was not much of a problem because of the woods natural tendency to move more readily that way with little clamp pressure . regards David . p.s love these blog post thanks

    1. I don’t think anyone is saying you shouldn’t have perfect meeting lines at all, that’s always been my personal choice and recommendation and has always been so. I was passing on what’s always been one common practice to my knowledge and experience. There are and will always be more than one way and more logical answers than one.

  5. Uh, I think what you have illustrated is completely wrong. You’ve shown a bowed joint, not a sprung joint. That is, a joint even more likely to open or split at the board ends than if the boards were prepared with straight meeting edges, or if they were sprung.

    As Dennis Ryan pointed out above, the purpose of a sprung joint has less to do with the clamping up equipment than dealing with movement in the wood as a result of moisture loss after the pieces are put together. Wood loses/gains moisture more rapidly at end grain locations than elsewhere, and if the boards were put together with dead straight, perfectly meeting edges, then after several years of moisture cycling the tendency would be for the glue joints to open up at the ends, or split if the glue happens to be stronger than the resistance of the wood to tension perpendicular to grain. Moisture cycling, when there is gain, means grain compression, and when the moisture is lost the tendency, eventually, is for the joint to open slightly. At first the wood is elastic enough to recover from the compression, but over time and repeated cycles of gain/loss, it takes more of a compression set and recovers less.

    By springing the joint, the grain at ends of the board are pre-compressed when clamped up, and thus are better able to tolerate later moisture loss. The amount of springing depends upon the stiffness and elasticity of the wood – generally we are talking about 1/64″ or so in most cases. No dogs needed.

  6. One of the sources often cited for the concave version is Tage Frid’s books. It never quite made sense to me – the convex version seems more logical.

    1. There are indeed always shrinkage considerations but my recommendation is to stock your wood ahead for as long as possible and let it season longer if your shop is dry enough. In 50 years I have never had a glue line separate and this is because indeed wood has elasticity that allows for some considerable movement. I have measured tables that expanded 10mm in the humid season and shrink back that much when dry. It’s a bit like the fact that I have seen hundreds of wide tables with the tops screwed to the apron frames and no cracks caused by that. Yet today we say that if we don’t used turn buttons the tops will crack. That’s possible but highly unlikely.

  7. HI PAUL, AND WHIT DRAWER’S BOTTOMS? THEY WILL DISTORT AND MAKE THE DRAWER STUCK IN THE CARCASE IF I NAIL OR PEG THEM FROM THE UNDERSIDE, INSTEAD OF FITTING THEM IN GROOVES? I’M TRYING TO EXPLAIN WHY THE DRAWER’S BOTTOMS WERE DEVELOPED LIKE THE ONES THAT WE MAKE IN OUR DAYS, IF SIMPLE NAILING WOULD BE SUFFICIENT (UNLESS BIG DRAWERS WHERE THERE IS THE MECHANICAL SUPPORT OF THE GROOVES DUE TO THE WEIGHT OF THE OBJECT THAT CAN BE PUT IN THE DRAWER). I SEE OLD TABLE TOP NAILED OR GLUED OR PEGGED, WITH NO CRACKS OR DISTORTIONS, LIKE OLD BACKS OF CARCASS CASE FURNITURE FOR EXAMPLE, WHERE THE BIG TOP/PANEL WERE FIXED AND HAD HOLD WELL, SO I DON’T UNDERSTUND WHY NOT THE SAME FOR DRAWER’S BOTTOMS.
    THANKS VERY MUCH.

  8. The use of spring joints was described in the book, “Classic Guitar Making”, by Arthur Overholtzer, published in 1974. He spoke of using a concave joint for joining sides and back.

    1. It’s a little superfluous really as it was from a bygone pre-clamp era. It’s just handy from time to time to keep it alive.

Comments are closed.

Privacy Notice

You must enter certain information to submit the form on this page. We take the handling of personal information seriously and appreciate your trust in us. Our Privacy Policy sets out important information about us and how we use and protect your personal data and it also explains your legal rights in respect of it. Please click here to read it before you provide any information on this form.