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

Moisture loss and regain and sprung joints

Posts

Moisture loss and regain and sprung joints

#1

Moisture loss and regain and sprung joints

Bill Tindall, E.Tn.

The recent sprung joint discussion inspired an experiment. The experiment involved taking two pieces of wood from the same board that had been stored in an outside shed, damp winter conditions. One piece was coated with two typical coats of Kemvar M conversion varnish. (I used this stuff because it would dry at low temperatures.) The pieces were then brought inside to a small room with the heat opened to it. The register in this small room caused the temperature to rise to 78 degrees when the heat was running hard, like during the recent cold spell. RH got as low as 30%. After several days the pieces were returned to the outdoor shed when the outside temperature and humidity rose to damp winter conditions. This step is ongoing.

Each day the weight of the two pieces were measured to the nearest gram(0.01%), and the widths were measured at each end and in the middle.

Preliminary results. The uncoated piece equilibrated to the warm environment amazingly fast, with the ends initially shrinking much faster than the middle. The coated board lost moisture much slower with a much less difference in rate of moisture loss between the ends and middle.

Preliminary Conclusions: Because moisture loss out the ends of lumber is faster than out the middle, the ends of an unfinished glued up panel will be stressed if the panel is prepared from lumber that is drying in a low humidity shop environment.

The coating I used (and maybe any coating)not only greatly slowed down the rate of moisture loss it lessened the difference in rate between the ends and center.

A sprung joint could lessen the stress on an the ends of an unfinished panel under drying conditions. The benefit is much less for a finished panel where the ends as well as the faces are finished.

When finishing furniture sealing the end grain will lessen the drying stress on boards and panels.

When high moisture content lumber is brought into a dryer environment the lumber responds and equilibrates to this dryer environment in days not weeks.

Re: Moisture loss and regain and sprung joints

#2

Re: Moisture loss and regain and sprung joints

don stephan

Thanks for making the experiment and sharing the results. I'll continue using spring joints because they seem to reduce the chance of the ends opening during dry periods.

Re: Moisture loss and regain and sprung joints

#3

Re: Moisture loss and regain and sprung joints

Chuck Bender

Bill,

It is my experience that, if you continue the experiment for years/decades/centuries, eventually the middle will shrink more than the ends. Wood is in a constant state of expansion and contraction (depending on the relative humidity) but the long term trend for material that is kept indoors, where the heat and humidity are relatively controlled, is toward moisture loss/shrinking. As you point out, initially that is more rapid at the ends of the board but, eventually, those fibers reach relative equilibrium while the fibers in the center of the board are still in a greater state of change. Because the fibers at the ends of the board had greater moisture loss from the time the tree was felled, the center of the board will eventually end up shrinking more because those portions of the fibers were wetter and larger (think dry sponge versus wet sponge) they have more room to contract as moisture is lost. This eventually results in table tops that are narrower in the middle than at the ends (which is the general trend in antiques of the 18th century variety at any rate).

Glad my response to a reader's inquiry spawned such interest, thought and lively debate. I will endeavor to continue that behavior.

Re: Moisture loss and regain and sprung joints

#4

Re: Moisture loss and regain and sprung joints

Bob Lang

Or perhaps they are narrower because somebody planed them narrower to begin with?

Smart-a$$ answer aside, the factor of time is important and largely ignored, and this can make a difference when there is an annual cycle from dry to damp. Simply put, it takes a while for moisture in the middle of a hunk of wood to reach equilibrium and in many cases the environmental cycle heads in the other direction before that happens. What happens to a wide tenon buried inside a thick leg is not the same as what happens to a similar piece exposed to air. We also introduce moisture at the joint when we apply glue and it takes a while for that H2O to migrate. Eventually everything will even out, but it might take more than one lifetime or a radical change in environment.

I, for one appreciate Bill's scientific approach to this. Maybe if he submitted his data and photos to one of the woodworking magazines it would get published.

Bob Lang

Re: Moisture loss and regain and sprung joints

#5

Indeed

Bill Tindall, E.Tn.

"Simply put, it takes a while for moisture in the middle of a hunk of wood to reach equilibrium and in many cases the environmental cycle heads in the other direction before that happens."

The "dry" room did not stay at the same humidity nor temperature. When it was cold outside (a couple of record colds and hence low humidity when this air go warmed up) and the heat was on a lot the test pieces continued to dry, over the course of 10 days. But there was one day that it got above freezing, the humidity went up to 45 % , and the unfinished piece regained weight and width that day.

What was surprising to me was how quick the unfinished board responded and how effective the finish was in slowing this response. The most important result for me is that I will become more careful about getting end grain sealed in critical locations.

I would want to repeat the experiment before making claims in too wide an audience.

Re: Moisture loss and regain and sprung joints

#6

Re:A kind of dumb ?

Larry in NW Ohio

I have read this post on sprung glue joints with great interest, for one thing I have never heard of it, and another why. I could never figure why a person would want an opening in the middle of the joint on purpose. What is the purpose of having an opening in the joint on purpose. I always thought you wanted the joint to be as tight as possible. Thank You. Larry

Re: Moisture loss and regain and sprung joints

#7

Not "dumb" at all-the why of it

Bill Tindall, E.Tn.

The data I gathered shows that end grain shrinks at a greater rate than the middles of boards or panels under a drying situation(this observation is a well known fact. I don't mean to have discovered it). Such a situation could be in the fall when the heat goes on or when lumber stored outside is brought into a lower humidity shop, or when building from lumber that was not thoroughly dry to begin with.

The theory of the spring joint is to put compression on the ends of glued up panels so that if they dry, and therefore shrink, at a faster rate than the middle the ends will be less likely to shrink so much more that the end of the panel splits.

However, furniture factories strive for two perfectly straight edges when edge joining boards. (these edges are often obtained with an accurate rip saw designed for this purpose.As they say, "don't try this at home")

Now here is a quandary. Suppose you have a 21" board you wish to use for a cabinet side or table top. (there is a dandy at the Spring Hill,KY Amish town) The potential stress on the ends of this piece are going to be the same as if the 21" width was created from gluing up 3- 7" boards. If one truly believed that compressing the ends is essential for something this wide, then it would follow that the 21" board should be ripped into three pieces and glued back together with spring joints.

An alternate explanation:

I believe that the true essence of the spring joint is that for those skilled in its application it is the fastest way to ensure the ends of the joint aren't gappy during assembly of a panel. If a perfect joint can be created otherwise the spring joint is not essential.

Re: Moisture loss and regain and sprung joints

#8

In theory....

John in NM

The idea is that the center of a board is the most stable over time, it's change in EMC being buffered by the time it takes atmospheric moisture to migrate in from the ends. By springing the joint slightly, the builder is putting the stress on the joint where the wood is most stable, not at the ends where it reacts to environmental changes most quickly.

I have not read up on the history of the technique, my explanation is based on my own understanding of why it works, not why an old timer might tell you to do it. There could also be something to the alternate theory that Bill just outlined for you, that if it's difficult to make a perfect joint, so you make it imperfect in the strongest possible way. I think that this is also a likely explanation for starting the practice, though perhaps not an explanation of why it works well.

For what its worth, one of the most catastrophic examples of table joint failure I've seen is a factory piece from the 1960's. They did everything right by production standards, and the table lasted about 40 years.

It is my belief that factory methods are not "better" in terms of long term survival, but that they are perfect for balancing production cost and consumer expectations. If the table lasts only 40 years, the consumer is perhaps disappointed, but does not feel cheated.

Factories do not set out to make future antiques. They set out to make furniture as cheaply as they can that will satisfy their customer's expectations at the price they pay. If they're having expensive stuff come back after 5-10 years with failed joints, they will address the problem. After 40 they probably never see them returned for a myriad of reasons.

Re: Moisture loss and regain and sprung joints

#9

Re: Moisture loss and regain and sprung joints

Tom

Which is why I use epoxy as an end grain treatment where others use bread board joints.

Re: Moisture loss and regain and sprung joints

#10

VS. the real world

Tom

Two things that differ in the real world are that the extremes are rarely that extreme time wise. So while woods dry quickly. Usually the environment they live in, won't switch that badly.

And not all woods dry the same way. Beech (few of us use that for furniture), was a real surprise to me. It dries incredibly fast from all over the board, not just the ends. Which raise the question of the degree to which all boards are drying from the ends, in accordance with the normal model.

Re: Moisture loss and regain and sprung joints

#11

Clint Searl

Ashow of hands...

Clint Searl

by all who've had a joint fail because it wasn't sprung.

Re: Moisture loss and regain and sprung joints

#12

Re: "What was surprising to me...

Tom

how quick the unfinished board responded and how effective the finish was in slowing this response. The most important result for me is that I will become more careful about getting end grain sealed in critical locations."

This is something that one learns real fast air drying wood, or roughing wet bowls. The wood can split in an hour, usually one gets a few days, but if the ends aren't coated then bad things happen.

It appears we don't learn very well from reality. Everyone knows one doesn't leave the ends uncoated during initial drying, but then they don't do that to stop wood moisture in pieces. We are still caught in the past of using wood configuration to solve the problem. Which is does, up to the point it doesn't. There are certainly pieces of old furniture that didn't survive, and cracked along the way, or distorted. Plugging the end grain pretty much eliminates that problem.

Re: Moisture loss and regain and sprung joints

#13

Re: Dumb is as dumb does

Tom

"An alternate explanation:

I believe that the true essence of the spring joint is that for those skilled in its application it is the fastest way to ensure the ends of the joint aren't gappy during assembly of a panel. If a perfect joint can be created otherwise the spring joint is not essential. "

That is basically my argument in the earlier thread. As I pointed out at the time, the sprung concept does not work in a variety of situation: Multiple joints; pieces that are too stiff to be sprung; rubbed joints. These are not special situations with some other cancelling factor. Some of these are mega situations where if spring were required one would expect them to be all the more required in this case. If spring is only required when it is convenient to produce it; if it is never really calibrated, how valid is it?

At the time I said a departure point for Bill's exercise would be to document the actual existence of the myth. It reeks of some clever idea someone stuffed in a magazine article at some point and it became an eternal truth. But I don't think this has happened. Where is the evidece that this is an eternal truth handed down from generations. It sounds like a post rubbed joint thing for on. We are using junk science to create some junk history, or vice versa.

Re: Moisture loss and regain and sprung joints

#14

Re: Ashow of hands...

Tom

I can give you examples of where a joint won't assemble if it is sprung, or will fail if it is sprung. I also have plenty of examples of where it held together when it was sprung, but none of I can think of where a joint came apart because it wasn't sprung (unless it was otherwise done badly).

The whole issue of extreme changes in humidity is mostly a post central heating issue. Even when I lived in the UK/Ireland in the 60s and 70s, there was no central heating in those homes, which was typical.

Re: Moisture loss and regain and sprung joints

#15

Re: In theory, reguarding planes....

Tom

I have been pushing this point for 20 years now. The natural result of using a jointer plane to joint boards is the creation of spring. That is why jointer planes are long. People say it is because they take the ripple out by panning highs and lows, which is a whole other discussions. Suffice it to say that in the real world an edge has already been hit first by another plane or power jointer, so the rippling should not be an issue. So why is the sole long, because a long sole pre-sets the edge of the board for a given amount of spring. If you don't want spring, then once the plane stops cutting you back out the ends, and take one full pass. All this can be done without looking at the board, it is an auto pilot thing. However, the degree of spring varies depending on plane geometry. But that isn't much different than any other thing in cabinetmaking: our benches are only long enough for work that is traditional height in furniture; Or planes are only wide enough for average edges, and chisels wide enough for average joints. There are different crafts with different tools. So yeah, spring is a plane thing. But it doesn't work for all planed butt joints, so we are back to where the "myth" comes from, and that is not so easy to figure. And when I say "work", I don't know whether it is a splitting thing or not.

For it to be useful it would be sufficient for it to not make splitting worse. Let's say that is was an advantage for reducing splitting, so where it came off the tools, and worked with the clamps, it was a good idea. But let's say the degree of benefit was at the margin. A kind of mittens are warmer than gloves, but you still need to be dressed in something else to be warm out of doors, in winter, kind of thing. So then your process and results are fully consistent, but Bill is going to be up against it proving it one way or another.

Re: Moisture loss and regain and sprung joints

#16

Re: Ashow of hands...

Dick Coers

"I can give you examples of where a joint won't assemble if it is sprung, or will fail if it is sprung."

What are they?

Re: Moisture loss and regain and sprung joints

#17

Another thought....

John in NM

I remember thinking this in the last discussion, but I'd forgotten it until I saw your mention of rub joints....

I doubt springing is all that old a habit, as there is no point to it at all without significant clamping ability. As far as I know, the ability to apply a great deal of clamping pressure is recent (however, I do not know how recent -- I would guess early industrial).

I do not doubt that it helps though, as I said in the first paragraph before, it concentrates the stress on the glue joint in the place most able to take it.

I wonder now, if it became a practice due to the ability to clamp tightly, or it became a practice due to the inability to simply re-glue failed PVA joints? Repairing a failed hide glue rub joint is trivial, not so with PVA.

From what others (with more traditional educations than mine) have said on the subject, I tend to think its the former rather than the latter.

Re: Moisture loss and regain and sprung joints

#18

Re: Indeed

Bob Lang

And the problem is that like people, each board is a unique individual. Any results would really only apply to the specific sample. If you used a zillion samples you would have a better average, but you wouldn't have any certainty that the next board off the stack would behave. I've thought of "following" samples over the course of a year or two, and have held on to cut-offs from wide glue ups with intentions of measuring them every few weeks. Either I forget or someone cleans the shop and tosses my experiment.

If you ever want to make yourself crazy, try milling rough lumber while following all, most or even some of the published rules and recommendations.

Bob Lang

Re: Moisture loss and regain and sprung joints

#19

many years of data

Bill Tindall, E.Tn.

I have been following two wide panels of walnut and another of elm for many years. The humidity range they experience is 30 to 65%. I have been surprised how little they change, far less that the shrinkage tables would predict.

That said, I fitted some drawers too snug during the late winter and when I conditioned them at high humidity to be assured they would work seasonally I found I had to plane some more off. So this dimension stability issue is not something that can be ignored.

Re: Moisture loss and regain and sprung joints

#20

Re: many years of data

Bob Lang

Absolutely not. Although it may look like science I think it is more art and educated guesses based on experience. To get it right you need to be able to assess the character of the piece of wood you have (and be prepared to shave something off if you guess wrong).

Bob Lang

Re: Moisture loss and regain and sprung joints

#21

Re: Ashow of hands...

Sgian Dubh

Over the years I have had a few edge joints fail for one reason or another. This can be put down to a variety of factors. For instance an incorrectly formed sprung joint may be a cause, so too can incorrectly formed unsprung joint. For instance, if the joint is sprung but the spring is too great excessive force is needed to pull the centre together - I made this mistake on at least one occasion that I can recall. Where the intention was to create two perfectly matched edges, whether straight or curved, if the ends had to be forced together that can cause the joint to start opening at the end, and once the joint starts to open it will frequently continue to open. Another source of problems can be heat applied locally to a joint; for instance a laptop computer left running for long periods over a joint may cause localised drying leading to localised opening up of the joint. I've seen this happen a couple or three times, once to a piece I made, which was frustrating because it meant I was called in to undertake a repair, and I didn't charge for it.

Moving on to a general point about sprung joints, I'm forming the impression from reading comments on the technique, both here and in other discussions, that perhaps some people have developed a misconception about them. A well formed sprung joint needs very little force to close the gap at the centre, and I wonder if people may be thinking that significant clamping pressure is needed. If this is the case the spring is too great. All that should be needed in my experience is to butt the two edges together, put one hand around the outer edge of one board, and the other hand around the edge of the other so that your palms rests against the edges, and gently squeeze. I've said it before but the kind of pressure I'm talking about is I think ideally no more than you'd need to suspend a 4 or 5 lb book with the spine gripped between the fingers and thumb of one hand. This kind of pressure is piddling compared to the kind of pressure you can exert with a firmly tightened sash clamp.

With regard to when sprung joints came into general usage my understanding is that this occurred at about the same time as modern glues such as urea formaldehyde, PVA types and other synthetic types began to take over in common usage in preference to animal glue types which, as is well known, have a good deal of 'suck' and are therefore well suited for rub jointing techniques like edge joinery. This is not to say all other glue types don't have 'suck', because PVA and aliphatic types for instance can also be used with a fair degree of success for some rub joint applications. Slainte.

Re: Moisture loss and regain and sprung joints

#22

Re: Moisture loss and regain and sprung joints

Sgian Dubh

Bill, the results of your testing are much as I expected, unless I'm missing something, which I don't think is the case. I think the time you've spent on this has been worthwhile and interesting. I concur with you that applying a finish can greatly affect the rate at which water moves in and out of wood -- some finishes are more resistant to water ingress and egress than others of course. I recall doing a somewhat similar experiment some years ago to show how gaps develop in mitred joints as the wood dries or gets wetter, and I was struck by how little time it took for the bare wood joint samples to respond to changing conditions of RH and temperature. Slainte.

Re: Moisture loss and regain and sprung joints

#23

Re: Ashow of hands...

Tom

Basically cases where the width to length ratio (aspect ratio) of a board is such that it can't be deflected. Of course, maybe you could make one in the 1/10000s, but that isn't what the planes do, and any practical moisture benefit would be lost.

So for instance, my table has boards with an aspect ratio of 12-1. A guitar has soundboards with ratios of about 2-1. On an aspect ratio basis the soundboard pieces are 216 times stiffer than the pieces that went into the table. For all intents and purposes, they are rigid, and can't be deflected. There were 5 boards in the table, so what I did was glue up, sprung, the two outside ones (this was like 20 years ago). Then after they were through the planer to get the glue off. They were glued to the remaining centre piece. the newly paired sticks were at that aspect ratio 8 times stiffer than during the first glue up.

So to answer the question, if the aspect ratio is too high, they can't be bent, then that is either that, or you may get glue starvation. Another problem is potentially different flexibility, where one side does all the bending.

Re: Moisture loss and regain and sprung joints

#24

Re: Ashow of hands...

Tom

"I've said it before but the kind of pressure I'm talking about is I think ideally no more than you'd need to suspend a 4 or 5 lb book with the spine gripped between the fingers and thumb of one hand."

So I actually agree with just about everything you said, but I think you might be understating the degree.

Myself, I don't really see that. They do call it a spring, you are taking about such a minor pressure I don't even know how you could feel five pounds. That is less than the weight of a jointer plane, so that would mean the whole board is so flexy it bends under 1/3rd the weight of the plane, not to mention any leaning on it. You might be supporting the board on a surface, but a lot of planing is done with, say, one end of the board in a vise and the other end on a jack. And those boards don't deflect under a few pounds pressure.

I think the same error was propagated in the PW article which spoke of a few thou between the boards. I have been around long enough to see plenty of demoes where the amount was more like 1/16" on a 5 foot pair, and yes, more. There are a lot of variables. It should not be a struggle to get the boards together, but if it takes nothing, two things are happening; one the jointer is probably being used improperly, and the "spring" is so slight it can't possible be doing anything. I am not betting on boards that are sprung one or two though, or bananad one or two thou, doing anything pro or con (to call on the PW example). If I could be guaranteed one or two thou of reverse spring every time, I would just slap the clamps on.

Re: Moisture loss and regain and sprung joints

#25

Had or seen?

John in NM

I don't recall having one fail, but then I've been springing them more often than not for most of the time I've been building stuff.

The factory table I referred to earlier has about 1/3 of the joints failed partially or completely. There are a lot of joints, none of the boards making the table top being wider than 4", and most are less than 3". While it would be time consuming to build such a top with sprung joints, it's clear to me that everything they did to avoid problems were unsuccessful. Might as well have built a top out of 12" wide boards and glued it up with hide glue, probably would have fared better.

Another one I recall vividly was a 2" thick mesquite table top we had to replace in a shop I worked for. Top was only 6-8 years old at a guess. The shop had had a run of bad tops like that one early on, and had then changed to this kind of strange looking wrap of 4/4 material to mimic 8/4 - cut 4" off the ends of the glued up top and glued it on the bottom for a thick look to the edge (kind of book matched end grain looking). I would have tried springing the joints first, but I have to admit that the alternate method seemed to work ok if you didn't mind the look.

Then there are countless "good furniture" tables I've seen in junk shops. Nothing much is older than the 1950's out here, maybe 1920's at a stretch, and all of it mass produced. We rarely see real antiques for sale. A bad joint or two in such tables doesn't seem to preclude their sale, though I suspect the really bad ones like our handmedown table never even make it to the shop or are turned away by the vendors.

Perhaps I see more of them out here due to the dryness. You can air dry lumber to below 6% MC depending on season and where you live out here.