Clamp Time For Glue

[Tom Watson]

I was gluing up these chairs today and was thinking about ideal clamp time.

I pretty much have a habit of leaving everything clamped up overnight in conditioned space that is ideally 72/50, or 72 degrees at 50% relative humidity.

So, for somewhat less than ideal conditions you adjust a little bit.

I worked in a stair shop for about six months and they would take flat panels out of clamps after an hour and run them through the Timesaver.

I don't hold with that but that is a feeling and it doesn't reference anything empirical.

When running my own cabinet shop I never took things out of clamps unless they had sat overnight. That was in the day of mortice and tenon joinery without additional mechanical help.

When I switched to using pocket screws as a hidden helper on the joints I would take the clamps off immediately. I know longer feel that is appropriate for things like doors that will enjoy hig stress during their lives. It seems to me that they need to sit in situ until the chemistry is complete enough to give them the maximum chance of survival.

Many of the glue bottles suggest a clamp time of from 30 minutes to 60 minutes.

It seems to me that a complex glue up, like a dowelled chair, where you are 'encouraging' imperfect fits to do your bidding have needs greater than this.

Once again, this is a matter of feel and not backed by data.

I've checked my Hoadley and looked at Forest Products Laboratory but I'm not finding the definitive stuff that I am looking for.

What I'm interested in is things like shear strength v. time and tension strength v. time for the typical PVA and A resin glues that we all use.

If someone can point me to data on this, it would be much appreciated.

Regards,

Tom Watson

[dpb]

Tom Watson wrote: ...

Don't have a reference per se; if there is something fairly readily available I'd think it would be at the Forest Products Laboratory site. If it isn't there, must not be easily findable. That would make me wonder if the basic answer is my thinking--"when it's dry, it's dry".

I have seen some published work (summarized in a FWW article some months back) on optimal clamping pressures for production woodworking applications but afaic(an)r(ecall) clamping time wasn't a variable in the testing. The upshot of that was that pressures far higher than normally achieved in hand clamping actually produced somewhat higher strength in testing. But, of course, in most cases the glue joint is stronger than the wood anyway. The point of this work was pressure for excessive squeeze out in production work, not mechanical strength, btw.

Of course, the answer depends on the glue specifically and the species to a lesser degree and the temperature/humidity factors you've already mentioned.

But, my take is that for PVA glue once it has set up for about an hour or so in ordinary weather conditions it's strong enough to machine and if it handles that it's strong enough. I don't think after that point clamping will have any further effect.

Urethane and other similar chemical bonding as opposed to drying glues are different and each follows its own regimen. I'd look at manufacturers' web sites for more specifics on them; most tend to have additional technical literature. Probably would get a response from their tech staffs to direct question as well; either reference to previous work or guidelines for specific applications if provided.

As for the question regarding relying on the glue to make ill-fitting stuff hang together, glue doesn't work well for that purpose anyway. Thick glue lines take longer to dry, obviously, but also testing shows they don't have the strength of nominal-thickness gluelines, either.

How's that for "I don't know"... :)

$0.01, etc., etc., etc., ...

--

[Tom Watson]

Thanks for the reply.

I guess what I'm looking for is something like the old concrete tables that would show strength over time. This was mostly compressive strength but the same thinking should apply.

The interesting thing about some of the old concrete stats is that they thought that final strength v. time was essentially an infinite program.

When I talk about a joint that I want to 'encourage' what I'm saying is that I have a 3" stretcher that is dead flat for 2 1/2" but, because the chair was used after the initial failure, the bottom half inch has been rounded. I'm not looking for any gap filling help at that point, and I'm not looking for that half inch to contribute to the strength of the joint, what I want to know is the tension strength of the joint, given what I have available and what allowing extra clamp time does for me regarding final tension strength of that joint.

Regards,

Tom Watson

[dpb]

But concrete isn't glue...or glue isn't concrete. :)

Concrete does continue to cure and change its chemical process through quite a long period of time. Wood glues don't act like that. Once it dries (PVA) or cures (urethanes, epoxies) it has reached it's ultimate strength--additional time, clamped or not, won't matter a whit.

So I don't think you'll ever find a set of data like what you were looking for.

Not much if anything at all once it has set. If you're pulling the joint back together to make up for material that has been either compressed or worn away (or even, to a lesser degree, simply pulling a separated piece back into position) there's a pretty good likelihood that same failure will happen again. PVA glue in particular tends to creep with time, while urethanes and even hide glue don't nearly as much. To attempt that kind of repair (which I've certainly done as well altho restoration work isn't a regular portion of what I ever did altho swmbo did have a period years ago of dragging home a lot of "antiques" that did work over) unless it is really of value I'd probably go w/ the urethane and leave it until completely dry. If it's more valuable piece, and not factory-built, then hide glue would be the choice of course as it can be reversed if ever became necessary.

--

[Tom Watson]

Please don't think that I am being argumentative on this but my entire thinking is based on what the physical chemistry is when it is 'dry'. It is also predicated on the defintion of 'dry'.

We know from experience that the squeeze-out is available to be cleaned off for some time after the joint is set but what of the 'dryness' and 'cure' of the joint itself?

If I glue up a joint that wants to separate because of the mechanics of the interface, what help may I expect from the adhesive?

This is where NOVA should jump in.

I think he actually took a strength of materials class back in the day.

Tom Watson

[Nova]

[Tom Watson]

You lie so bad, dog.

You be an engineer and you had to take this course.

Regards,

Tom Watson

[dpb]

So did I, but that didn't cover the subject... :)

The time to "dry" for PVA is dependent primarily on temperature and to lesser extent RH and what material. Since PVA glues dry by the loss of moisture from the glue line, it's not a fixed time but dependent on the external conditions as well as the moisture content of the material and the amount of glue in the joint. The typical one hour mechanical strength time is the time at which one can expect the joint to have dried sufficiently to have sufficient strength for further mechanical operations. I think almost if not everybody will suggest 12-24 hrs for complete drying for PVAs.

If you're really concerned about high stress joints I'd definitely recommend waiting for the overnight just as safe practice but I still don't think you'll ever find any data that's directly applicable to the question.

For such applications you might want to consider some of the specialty glues for laminating, etc., that don't have the same creep properties of "ordinary" PVA.

Selley products are excellent; it may be somewhat hard to find them retail altho I've not looked in ages.

a couple products you might want to look at.

--

[Nova]

specialist for what was, at one time, the major long distance company.

[Tom Watson]

Thank you for your response and I do appreciate the thought and time.

This looks like a good bet for my son's middle school science project for next year.

Somehow, that seems like a damned shame.

Regards,

Tom Watson

[Tom Watson]

Tom Watson

[Tom Watson]

You know, what I see in my mind's eye are micro-crystalline bonds that are destroyed by relaxing the clamp pressure prior to a certain time.

It is as though a mechanical as well as chemical process were occuring at the same time and that the mechanical bond was totally dependent on the completion of the chemical reaction.

I have been told that I have too much imagination for my own good.

Regards,

Tom Watson

[Lew Hodgett]

In the future, would it make any difference if I signed my posts using my business signature:

Lew Hodgett, PE

rather than the more informal

Lew

[Tom Watson]

Lew, you sign your posts any way you want but the fact that you are a PE is of great importance to me.

I've worked with good ones and bad ones but I have not worked with one who was not available to a logical argument.

That is one of the things that I like about engineers.

Their availability to a logical argument.

Carpenters are half-assed engineers and often produce half-assed results because they substitute rules-of-thumb for careful analysis.

I know, I'm a carpenter.

When I ask a question about shear strength of a glue line, or withdrawal strength (tension) on that glue line, I'm asking because I am outside my area of expertise.

I am hoping that an engineer, or a physical chemist will chime in and help me out.

It also helps that you are a sailor.

I do not countenance the term 'boater' as it includes far too many people with whom I would not choose to associate.

Especially the personal watercraft crowd.

It is a great disappointment to me that you are not a woody traditionalist but I would be interested in hearing your theories on concrete and fero-cement vessels (I can not bring myself to call them boats), should you have any.

Regards,

Tom Watson

[J. Clarke]

If it's a real structural joint where the glue is a stressed member and doesn't just hold the pieces in alignment I'd go with the old standbys, Weldwood plastic resin glue or Weldwood resorcinol glue. Understand--they have no gap filling capability to speak of and they cure _hard_, there is no flex in that joint, so wood movement will over time break down the bond, but they will not creep and the resorcinol glue is as waterproof as any adhesive for wood gets. They're not glamorous and "modern" like the epoxies but adhesives based on the same chemistry (but British brands, not American) put legions of Mosquitos over the Third Reich, some of which are still flying more than half a century later.

[Tom Watson]

Still a decent ticket that demands respect.

And yet we often ask them to have an effect.

No data.

Is there a trade off between strength and brittlenes?

Regards,

Tom Watson

[Father Haskell]

Which is why you use interlocking joints such as finger joints or dovetails.

Counter by increasing glue area.

[DanG]

Tom, I too am a carpenter. Arguing with engineers and inspectors is a bit like wrestling in the mud with a pig . . .. . you finally figure out they enjoy it.

But on the other hand, the difference between engineer boots and cowboy boots

/ / / / / the cowboy boots have the manure on the outside.

[Lew Hodgett]

Comes under the heading of poor design.

It becomes a function of "green" strength.

'GREEN' ADHESIVE WILL HOLD A JOINT IN PLACE; HOWEVER, MAY BE ABLE TO HANDLE DESIGN LOAD.

(Hit Cap Lock by mistake.

I don't have a clue.

A few years ago they had to sink some WWII Ferro boats that couldn't be broken up even using dynamite.

They sank them some place off the Oregon coast as I remember.

Regards,

Lew

[dpb]

...

Well, I'll modify the previous statement somewhat...I think you will find some data on tensile strength for various adhesives including PVA--a quick google found several references in various journals. The nearest I read the abstract for measured strength reduction as a function of strain rate but in your case the rate is essentially zero.

What I doubt you'll find will be anything about the clamping time vs strength--since the time is relatively short and it isn't nearly as complex a chemical bonding process as the concrete example I'd expect odds are very high that any test joints will have been cured and that will not have been a variable in the study.

What you might find in studies for commercial applications _might_ be some information on speeding up curing/drying time via microwave, etc., where production times might be a significant factor.

If I were to do any further working, it would be from the standpoint of that as probably the driving force for the research--what can folks like the furniture manufacturers get away with when they're gluing up reams of stock? In general those are fully or nearly fully automated processes and most probably use enhanced methods to shorten the times. What, if anything, might be of use to you I've no idea.

I think I'd go back to the original thought I had--send a query to Weldwood, et al., asking if they have any information on the subject and the reason therefore. I've had some success w/ them in the past on some specific applications.

--