New WoodChat!
Home Articles Forums Gallery Marketplace Videos

WoodCentral Forums

Knowledge You Can Build On — Est. 1998

Fast and affordable laser cutting – so you can make the parts you can’t find anywhere else

Sprung joints

Posts

Re: Sprung joints

#26

Re: Starved joints

Barry

Toothing is not necessary. All you need is straight, flat stock and a well-fitting butt joint. A starved joint is where you tighten the clamps so tight, you squeeze all the glue out. Tighten the clamps until the joint just comes together, then stop. This is really easy. Let's not make it too difficult!

Re: Sprung joints

#27

Re: Sprung joints

Pam Niedermayer - Austin, TX

Spend less time on joints and more time on other aspects of woodworking techniques.

I think this is dubious advice, although I think I understand what you're trying to say (hey, starved joints is a triviality); but I think achieving mastery in joinery is just as important, perhaps more so, than flattening stock and preparing it for glue.

Pam

Re: Sprung joints

#28

Re: Sprung joints - Clarification

Barry

I didn't say that very well. Starved joints are not trivial, especially if you are getting them. The fact is, there is really no need for sprung joints and certainly no need to get a starved joint. Just because you bought the newest, highest-tech clamps capable of 1500 pounds per square inch, doesn't mean you have to crank those clamps down to maximum possible force on every joint. Sometimes, less is more. Tighten the clamps until the joint pulls together and a uniform line of glue squeezes out, then stop. Well-fitting joints should be the goal of every woodworker, whether or not we use sprung joints.

Re: Sprung joints

#29

Moses Yoder

Re: Sprung joints

Moses Yoder

The ends of the boards will dry out faster than the center, and I have a table that shows this very symptom. Whoever built the table was not a master of any part of woodworking. I usually take the edge of my boards over a power jointer then smooth them with a jointer plane then take 3 or 4 passes out of the center with a smoothing plane to make a sort of very slight concave edge. This does not take much effort and makes a very slight sprung joint that I think is worth the insurance.

Re: Sprung joints

#30

The fact is...

Bob Hackett

Even the best joint for the job executed in the finest possble manner means nothing if you don`t understand the nature of the wood you are using.

Any woodworking project requires a system of moving through a progression of steps.From drying the wood and stock preparation all the way to finishing and even the planned environment it will live in and be expected to perform under.

Make a poor decision in any of these many steps and you set yourself and the project up for failure.

To take a small point derived from one of those steps and examine it in detail(glue starved joints for instance) has the potential to illuminate and improve that step but if taken out of context and not applied to the big picture contained by that system then it loses it`s ability to relate to the whole and add to the ultimate successful execution of the work.

Adhesives are meant to hold things together and must be appropriate to the project,material and the surrounding environment.Forget one of those factors and I don`t care how carefully you prepare the joint,failure will result.

Splitting of panels at or near the end grain can be as much a fault of poor material choice,grain orientation, the finish not moderating the surrounding environment properly or a host of other things as the glue used or joint preparation.

Without an in depth look at and analysis of the big picture much of the causes of our failures are just unproven theory and conjecture.

Re: Sprung joints

#31

Ease of making sprung joints vs. butt joints

wilbur

I would argue that it's actually easier to make a sprung joint than a perfectly fitting butt joint. With a butt joint, you have to have two perfectly mating surfaces. With a sprung joint, you have to have two surfaces that mate at the ends, and have a small hollow in the middle. The actual depth of that hollow is not critical, nor does the consistency of the hollow critical, and is easy to make with a couple of passes with a plane. Just start in an inch or so from the end, and pull the plane up an inch short of the other end.

When I've done panel glue ups, I tend to make a sprung joint because it takes less time.

Re: Sprung joints

#32

Re: Ease of making sprung joints vs. butt joints

ThomD

Wilbur, I agree with that, sprung joints is what a pane makes naturally, unless one uses something like a shooting board, planes don't naturally make straight edges. That said. If I had to glue up say 6 6" pieces to make a 36" wide table. I would get into a pretty good groove, set the plane to take a good pass, and make for the sprung joint. If all the factors led to too much spring, I would then just work back from the ends, and do a final pass, which allows on to get rid of the spring, or some of it, with pretty much equal certainty. So while sprung joints are easy, straightening them out is pretty easy also.

Re: Sprung joints

#33

Re: The fact is...  My early steps

ThomD

You certainly do need to understand how to use the wood. When I started woodworking I was most influenced by the Gougeon Brothers, and Krenov. Both in different ways emphasized moisture content, and working with the wood. As I proceeded and read a few hundred books, I developed even greater paranoia about the delicacies of working wood. When it came time to make a large dinning room table, I was all worked up about the wood movement issues. This was not entirely unreasonable, because I was modeling the table after ones in use at a local woodworking store. There the display tables were beautiful massive cherry tables, that had been made with breadoard joints. They had all snapped down the middle etc... with large cracks. The guy who owned the place (the excaliber saw and fence guy) said when the first 0ne split while he was there at night, it had sounded like a rifle shot.

I was a little worried... What's more, the texts give contradictory advise on stuff like whether to put all the rings in the same direction, or mix them up. So I decided I should use one of my lifelines and call local living treasure, Michael fortune. I asked him what I should do, and how best to deal with the momentous issues. His response was not to worry about it. All he did was make the table top up using the most pleasing to the eye orientation of the wood! That was a long time ago, so things may have changed. It can also be an issue if you decide to move your furniture from Chicago to Texas, or sell stuff to far flung clients.

As far as wood moisture content is concerned I have found I acquire wood faster than I use it. And despite early info I got, it seems this climate isn't all that harsh as far as materials survival is concerned. Sure I do stuff like bring boards into the living area before I work them, etc... But overall i have found these issues are all pretty easy to take care of with basic knowledge and practices.

Re: Sprung joints

#34

What Franklin tech staff says about starved joints

Bill Tindall

1. It is impossible to created a starved joint with PVA glue. They went to 1000psi on a joint and it was not starved. 150-300 psi is recommended for PVA glue. I have tried it with all the pressure I could muster from a full sized machinest vice on a 2 sq in joint and I got 90+% wood failure.

2. It is possible to starve (that is excessively squeeze out the glue) a liquid hide glue joint with excessive pressure. This result surprised me as it would seem impossible to squeeze a viscous liquid from a joint that only requires 2 molecular layers to form a bond (on a perfectly smooth surface).

Re: Sprung joints

#35

Re: What Franklin tech staff says about starved jo

ThomD

That makes a certain amount of sense. Though whether a few guys experience with a few joints is able to dismiss long run experience is another mater. There are two weasel aspects to that report, one is the failure rate. How evenly distributed is the 90%. The other more serious concern is the perfect joint thing.

One dynamic that is rarely refered to is the way glue squeezes out of a joint. One guy I read mentioned taking up pressure on your clamps to about a medium level at first, then waiting till the glue was displaced before final tightening (probably not a good idea with sprung joints or joints that aren't clamped like dovetails. This is a great technique and for one thing is far more accurate, since the materials don't skid all over the place. What it shows is that if you tighten all the way down, then you don't actually press the glue out of the joint, but it seeps out for a while, causing the joint skid at times. The skid indicates what was happening in that the joint wasn't bottomed out. Once the glue leaves the joint isn't any longer fully compressed. So one may be getting not so much a starved joint as one that has the wrong gap. This may explain why "perfect" joints with no closing pressure are so effective. Frankly I don't know what is actually going on to a scientific certainty. But I think it's more complex than a few lab samples would indicate. My experience with some of these things is based on titebond, since it is the go to glue for guitarbuilding.

Re: Sprung joints

#36

Re: What Franklin tech staff says about starved jo

Barry

In 50+ years of woodworking, I have never had a "starved joint". I suspect what other people may call a "starved joint" is actually a poorly executed joint that failed to pull completely together upon clamping and failed because of this. If this joint fails, it was because it was poorly fitted, not because it was a "starved joint". If you must apply maximum pressure to clamps to get a joint to close, you need to refine your joint-making technique. You certainly don't need bigger clamps. Drawing from my own personal experience, I think "starved joints" belong in the realm of "cow tipping" and "snipe hunts".

Re: Sprung joints

#37

What does "90+% wood failure" mean?

Pam Niedermayer - Austin, TX

Re: Sprung joints

#38

the other reason

Bill Tindall

If the surface of the applied glue dries to much before the joint is closed the glue molecules at the interface with not co-mingle and the joint will fail at this interface. A "rule of thumb" time for Titebond regular is 5 minutes (according to Frankline tech staff) open time for joint preparation. Of course this time depends on temperature and humidity.

Re: Sprung joints

#39

joint testing

Bill Tindall

The goal is a joint stronger than the wood. One way to test a glue, or glue joint, is to prepare a lap joint, cure it and then bust it apart. Either the wood or the glue will fail and it is easy to tell which broke. The joint surface is examined and the amount of wood and glue failure is estimated. The goal for a good joint/good glue is something in the neighborhood of 80-90% wood failure. It is an easy test to do and handy for assessing the success of some experimental joint preparation-for example gluing an oily wood. When we first started gluing IPE we were getting 10% wood failure, clearly unacceptable joint strength.

Re: Sprung joints

#40

Re: What Franklin tech staff says about starved jo

Adam Cherubini

For what we're discussing, I don't think 1000psi is anywhere near enough. Remember that psi is pounds per square inch. As the contact patch decreases, the pressure increases for a given load (think high heeled shoes on an aircraft floor).

For typical machine work, starvation is probably no issue. Like Barry says, you'd have to be pretty sloppy to have your mating surfaces off so much that you wouldn't have a very large contact patch.

But hand tool work is really a different issue. Think about the hollow made by a plane with a curved iron for example. When 2 boards come together like that, you could see a couple thousands of an inch of gap between them. My guess is, we crush the contacting edges with our clamps until the centers touch.

Most woods have about 3000psi of compression strength across the grain and we all have crushed wood with clamps. So we shouldn't discuss that sort of pressure as unachievable. I don't know what PVA does under that sort of pressure. I wouldn't presume to guess. But that could be exactly the sort of pressure we're getting at the edges.

I don't think this is a big deal. But I would caution guys not to speculate on this particular issue based on even 50 years of experience. The last 50 years of joints may have been different than what folks are doing now. The OP here was asking specifically about sprung joints and how much spring to plane in. I raised this issue with the glue because I don't know what effects one might get at the ends. This isn't similar to preparing boards on machines and clamping them together. 1000psi may not be out of the realm of achievability.

Adam

Re: Sprung joints

#41

You're mixing your pressures...

John in NM

Sure you can develop 3,000 psi or more at the clamp face, but your initial statement reflected, correctly, that the important place is the joint face where there is more surface area and thus less pressure. The math here is not difficult, if we keep the image straight. Say you have a 5' joint, just two boards 3/4" thick. That is 45 square inches. Apply 3000 psi at 5 points using 5 clamps for a total of 15000 psi (I'm assuming that clamping pressure is cumulative, as if we used weights rather than clamps). Divide the total pressure by the total area of distribution, and you get 333.3 pounds. This is assuming even distribution on the joint, and obviously, sprung joints do not have an even distribution. If the Franklin people are talking about indexing their figure to the joint area this way, then I think 1000 pounds is reasonable.

Re: Sprung joints

#42

Re: You're mixing your pressures...

Adam Cherubini

John, I'm sorry I didn't follow 100%. Send me an email. I must be thick.

What i do know is that the reaction in the bond line is not constant in any way. And woodworkers should know this. We'll see higher bond line pressures under the clamps and less between them. We could make a NASTRAN model and figure out exactly what is going on (but why bother).

If Franklin's PVA requires pressure to develop strength, we'll likely only see the desired pressure in spots under the clamp heads. The reason is that wood is not infinitely stiff.

THEREFORE: When gluing up boards edge to edge, use a lot of clamps!

Adam

Re: Sprung joints

#43

Re: You're mixing your pressures...

Bill Houghton, Sebastopol, CA

Are these calculated pressures attenuated by "give" in the wood between clamp and glueline? I don't know how much elasticity wood cells have in that setting; and, if they do, I imagine it varies by species.

Re: Sprung joints

#44

Re: What Franklin tech staff says about starved jo

ThomD

Is cow tipping a myth?

I think what the lab is claiming is that you can't push out the last two molecules and that there isn't

any situation where you could do this, within the parameters they have described. I suppose for techies one could ask is there no such thing for a glue joint as excess pressure (either in absolute terms, or within what our technology allows)".

Your suggestion about poor fit seems like a good one, though possibly not as you may have intended it. You seem to be suggesting the problem area is where the joint surfaces don't meet. Whereas the area that is called starved in a clamped sprung joint would be the ends which are under the heaviest pressure. So as you say more clamping pressure isn't going to solve the "problem", or possibly just move it to another section in the joint.

Re: Sprung joints

#45

Re: joint testing

ThomD

Does go to show the crazy world they live in. The glue of theirs I use most, the original wood glue has a base failure rate of 77%, So 90 would be way up there. This is a glue respected for it's ability to hold bridges on guitars. TB III is 57 percent. Not too sure where the 90 comes from. Apparently they get these base results while applying hundreds of psi to the joint. So one would wonder if somewhere along the holds bridges on with no clamping pressure, to 77% performance, they are experiencing a glue starvation result. I mean, if that 250 psi pressure where biased towards the ends by a badly sprung joint, they would be up into thousands of PSI, and what does that do along the glue starvation continuum.

Also, none of these adhesives are professional grade in their categories. Luthiers and cabinetmakers are all just foolin around.

Re: Sprung joints

#46

Re: What Franklin tech staff says about starved jo

ThomD

"For what we're discussing, I don't think 1000psi is anywhere near enough." etc..

+1

I think you can easily start to get enormous pressures on the ends. Only difference I would have is I have virtually never used cambered jointers for edge jointing, and for decades that was not taught. I continue not to for much the reason you highlight.

Re: Sprung joints

#47

Re: You're mixing your pressures...

ThomD

Normal clamp diagrams suggest the force radiates out at a 90 degrees, so you use pressure and spacing consistent with that idea, and dependent on the width of the boards. This model assumes no spring, for what good that may do. But the spring assumption being that it evens out the inputs of few clamps, though it become extreme at the ends.

Re: Sprung joints

#48

Re: You're mixing your pressures...

Adam Cherubini

"Are these calculated pressures attenuated by "give" in the wood between clamp and glueline? I don't know how much elasticity wood cells have in that setting; and, if they do, I imagine it varies by species."

No Bill. I'm talking about the modulus of elasticity for wood itself. Yes it varies a great deal by species.

Tho we may not think of in in these terms, wide boards have a much higher moment of inertia so they are much much stiffer making the need to have just the right spring or just the right fit much greater. I glued up 18-20" wide boards for my dining table and they were pretty tough to pull a little spring up.

Just for those playing along in our home edition stiffness equals modulus times moment of inertia or EI.

PW is currently showcasing a great deal of articles delving into fine art. Several authors are dealing with this previously taboo subject. I'm all for it and have been willing to go there and beyond. Obtuse as these discussions may seem, I feel woodworkers are equally if not better served by a healthy serving of engineering and material science. I frequently see woodworkers of different sorts making engineering mistakes. One I see most often is the use of a single huge fastener where a group of small screws would have better served. I also see fasteners and supports where none are needed. Work benches are a good place to se such things.

I think there's an assumption that all woodworkers are technical minded folks that know all these sorts of things already. That may have been true once upon a time, but I've met several woodworkers who do not even OWN a pocket protector.

Adam

Re: Sprung joints

#49

Just a ballpark figure...

John in NM

That calculation was pretty rough, and fraught with potential errors in assumption.... but, yes, the clamping pressure is attenuated in the center of a sprung joint by the elasticity of the wood. It can be calculated as clamping force less the amount of force required to bend out the spring and then indexed to the area of the joint in question. Just a guess without doing an actual force problem on it, a sprung joint from my example will see the full 333 pounds of force at the ends only, whereas the center will be less by the amount of force it takes to un-spring the boards.

Adam's got the right idea with the stiffness, though technically moment of inertia and bending moment are not the same thing.

I would like to see an actual technical analysis of this question by some one who knows the physics better than I do - my method is really a bit too much "seat of the pants"

Re: Sprung joints

#50

I'm mixing terms a bit too...

John in NM

I really need to write it all out to make sure its right, but the confusion comes from mixing up pressure with force. Force is in pounds, pressure is force indexed to an area. So the 3000 psi it takes to crush wood at the face of a 2 square inch clamp is not the same thing as 1000 psi on a 45 square inch joint surface.

One mistake I made in my post above was not fixing the clamp face area to get a total force. There may well be others. But my point is that I'm not 100% certain what Franklin's figure of 1000 psi is referring to - pressure measured at the clamp or at the joint. If it's at the joint, I find it to be a quite adequate "overkill" figure for testing their glue.