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

Why glue joints fail

Posts

Why glue joints fail

#1

Why glue joints fail

Bill Tindall

It was shown below that excessive clamping pressure in NOT a cause of failed glue joints. So, what is the cause? Glue works by "wetting" the surfaces to be joined. It follows that both surfaces must be wetted for the bond to form. A common cause of a wetting failure is allowing the surface of the glue to dry some before assembling the joint. This kind of failure can result when glue is applied to only one side of the joint. Any drying will hinder transfer and wetting of the mating wood surface.

A second problem resulting from letting glue dry too much before assembly is a thick glue line. As the glue dries its viscosity increases dramatically. If the viscosity is too high before clamping then the excess glue will not be squooze (E. TN word) from the joint and a thick glue line will result.

Another source of failure is the result of gaps. The strength of the glue dried in a gap is less than that of the wood so a thick glue line will result in fracture through the weaker glue rather than the stronger wood. It takes but the tiniest thickness of glue to form the strongest possible joint. Any amount of glue in the joint greater than this optimum thickness will therefore weaken the joint. Perhaps gapping joints beget excessive clamping pressure to try to close them and that is the source of the clamp pressure myth.

Re: Why glue joints fail

#2

Clint Searl

Re: Why glue joints fail

Clint Searl

Most of the time, the "laboratory" quality of a glued joint doesn't matter. A properly designed and fitted joint won't be subjected to the kinds of stress that would cause it to fail. You don't need much glue; you don't need to smear glue on both surface; and you don't need tons of clamping pressure.

Clint

Re: Why glue joints fail

#3

Re: Why glue joints fail

john lucas

Are we talking about failing right now or failing 10 years down the road. I repair a number of pieces each year and I would say that most joints fail due to wood movement or improper storage of the piece. When you get a lot moisture around an old piece of furniture it seems to cause excessive or uneven swelling. This along with the moisture added to the glue seems like it could be a culprit.

Mostly the joints that fail are miters or cross grain joints. The uneven movement of the woods in these joints are probably what cause them to break.

Of course this is just an educated guess based on seeing lots of the same thing and studying the joint. What do you all think.

Re: Why glue joints fail

#4

agreed

Bill Tindall

Adding, or removing, moisture from a wood joint can put great stress on it due to differential response of the wood in the joint to moisture content. A dowel is a classic example. It will become elliptical through loss or gain in moisture as will the hole it fits in. Only by chance will these deformations compensate; typically they won't and the resulting stress will break the joint. Which is why I avoid using dowels.

Age and moisture must also affect the glue because it seems to be common to see glue line failure in old pieces in addition to busted wood.

From the failures in some old stuff it looks like lumber less than optimally dried was sometimes used.

Re: Why glue joints fail

#5

Re: Why glue joints fail

Jesse Cloud

I dunno about that. I used to design pieces proportional to the anticipated stress. Then I saw a kid dancing on the night table that was supposed to hold an alarm clock and a book. That shelf that was supposed to hold a delicate piece of china wound up crammed with books and sagging.

Now I always try to build it way stronger than I think it needs to be.

I agree, though, that the amount of glue used and the amount of clamping pressure is hardly ever the problem...

Re: Why glue joints fail

#6

Re: Why glue joints fail

Jim Follette in Fayetteville, NY

I disagree that clamp pressure has nothing to do with it. PVA glue works by interdigitating with the cellulose skeletal structure of the wood, then polymerizes into a more solid structure. This solid mass with extended little fingers into the wood on both sides of the joint is what holds the two pieces of wood together. The more the glue is forces into the cellulose the better. This requires more clamping pressure for hardwoods that soft because the cellulose skeleton is packed closer together.

This is not true of polyurethane glue which expands on its own and forces its way into the cellulose as long as the wood pieces are held to each other (and not allowed to be spread apart by the expansion of the urethane as it hardens) by appropriate clamping force.

As someone mentioned, polymerized glue is not as strong as the polymer interdigitated with the cellulose so gaps filled only with glue are not as strong as the joint where the wood pieces are tight together.

Moisture content is important in this regard...too much moisture and the water cannot be easily displaced along the joint and the glue will not polymerize properly (the molecules are too few and far apart when mixed with the moisture in the cellulose). However, if the wood is too dry the liquid glue might be too viscous to penetrate as well as when there is some H2O to decrease viscosity.

Re: Why glue joints fail

#7

Re: Why glue joints fail

FredD

Everybody is talking about pressure and never say how much they use approximately.

My opinion of a good joint is one that fits so closely laid together that the line tends to disappear before the glue is applied.

Then I believe the clamping pressure is only enough to keep it that way.

Torqueing clamps down to crush a poorly fitted joint together is a waste of time, the time should be spent on fitting the joint before the glue stage is entered.

Works for me.

FredD

Re: Why glue joints fail

#8

Re: Why glue joints fail

Chris Friesen

"Then I believe the clamping pressure is only enough to keep it that way. "

The glue manufacturers would disagree with you. The best PVA glue line is one that fits perfectly before the glue is applied, has both surfaces wetted with a thin layer of glue, and is then clamped to the manufacturer-recommended pressure (100-250psi, depending on species).

Incidentally, it's amazing how much difference a small unevenness in width can make in the pressure on the glue line. There's an interesting walkthrough here: http://www.fpl.fs.fed.us/documnts/fplrn/fplrn259.pdf that looks at edge-gluing 3"x30" flatswawn red oak. They talk about the case where the shop air is more humid than the wood, so the last 2 inches of each end swell to 0.010" wider than the rest of the board. They then clamp 2 boards together using 6000lbs of pressure using rigid cauls. Because the ends are slightly swollen, they end up receiving the vast majority of the clamping force, at 980psi. The rest of the glue line only gets 80psi, which might end up making a glue line that is too thick. A home shop would be slightly different, because pressure is applied via individual clamps. This would warp the pieces so the pressure at the glue line would be more uniform, though still not as good as if the pieces had been glued up right after machining.

Re: Why glue joints fail

#9

Re: Why glue joints fail

FredD

"The glue manufacturers would disagree with you. The best PVA glue line is one that fits perfectly before the glue is applied, has both surfaces wetted with a thin layer of glue, and is then clamped to the manufacturer-recommended pressure (100-250psi, depending on species)".

You apparently misunderstand what I mean about "enough clamping pressure to keep it that way"

The pressure I apply with a Plano Glue Press is enough to pop a row of tiny beads of glue.

Once this has occurred any more pressure then that is counter productive.

It's kinda like shooting pool , many think a rack can be manipulated , but if all the balls touch that's it , no more room for improvement.

Once the glue line is beaded , you have a good joint , always assuming you took the time to fit the joint first. My glue lines are not visible if there is a good grain match.

Last thing about PVA , it is not the only glue that follows these guidelines.

FredD

Re: Why glue joints fail

#10

Re: Why glue joints fail

Chris Friesen

The simple fact is that the manufacturer of the glue (and the government research departments who research this stuff professionally) recommend a specific pressure (much higher than most small shops use) in addition to a well-prepared surface. They have nothing to gain in recommending a higher pressure than necessary, so I tend to believe them.

That said, it's likely that a lower pressure will result in a joint that while not "as strong as possible" is still "strong enough" for most purposes.

Re: Why glue joints fail

#11

Re: Why glue joints fail

FredD

I am not a authority on glue and it's various uses , but I know when a joint is snugged by many years of doing it .

Anybody who has glued up panels for years know what I am talking about.

What I don't know and it seems highly irrelevant is how many molecules in thickness my typical glue line is or how much the applied pressure is , or how many angstroms I have going for me, though it could very well be in the range your claiming since none of my joints fail and except for grain change in mismatched panels the joint line is not visible unless you look on the ends and after finishing that tends to disappear also.

One thing I have noticed over the years is that PVA glue manufacturers never mention how the joint can creep a long time after assembly using that type glue.

Until the day pipe clamps and similar clamping devices come with a pressure gauge and thickness gauge for the glue line I am not going to sweat it or any dubious science promoted by self appointed oracles.

Just call me skeptical

FredD

Re: Why glue joints fail

#12

Thanks

Bill Tindall

I have been a voice in the wilderness for the past years telling folks that lots of pressure is recommended and good. I don't know where these notions of wimpy pressure came from; maybe fear of "starved joints"?

Re: Why glue joints fail

#13

testing

Bill Tindall

If you occasionally test a glue line and get 90+% wood failure you can't do any better and your technique needs no improvement. However, if a panel joint hasn't failed, it may be that it hasn't been stressed by the current generation of owners. Personally I don't make guesses about my glue joints. I test them. It is easy when gluing up panels. Fracture the cut-offs and look for wood failure.

BTW, it is easy to get required pressure form professional quality clamps. An Acme thread is capable of generating tons of force with moderate torque. Some fraction of this force is lost to thread friction but still a lot of pressure can be brought to bear on a joint with a good clamp.

Re: Why glue joints fail

#14

Re: testing

FredD

I have demonstrated the wood failing before the glue does many times to people who were clients or just curious about strong a glue joint can be.

I first saw this in wood shop in high school when the teacher showed how strong Franklins Horsehide glue really was. Early 1950's.

No PVA then.

It would be interesting to me to see the results of a Independant Laboratory doing a test of glue joints and what the conclusion would be for appropriate pressure for individual wood species.

I don't believe until I see that test stated in layman's terms,that exerting "tons of pressure" on a glue joint is beneficial or necessary.

I do agree that Acme threaded clamps like I use have the capacity to exert far pressure then is desirable, but why exactly would you want to.

While testing occasionally may make you feel better, how many years do you need to do this before you think you got the hang of it?

FredD