Regluing Titebond and Liquid Hide joints

Re: Regluing Titebond and Liquid Hide joints

Barry Irby

Thanks for the test. Glad the glue succeeded.

Did you test a similar joint that you just glued together and clamped "normally"?

And, how about a test where you break the joint after a few minutes and then set the pieces aside for a much longer time, months or a year, to simulate a failed joint that has been neglected?

Not sure, but I think the glue (Titebond) continues to polymerize for a long time.

It was something of an Ah Ha! for me when I realized MT joints fail in shear. That emphasizes the important of the shoulders on the joints. They act as fulcrums of at least they could.

Thanks again. BTW, I used your tip yesterday abut the fact that Titebond responds much nicer to gentle clamp pressure than to hammer blows. Thanks for that too.

Re: Regluing Titebond and Liquid Hide joints

Bill Tindall, E.Tn.

"Not sure, but I think the glue (Titebond) continues to polymerize for a long time."

I have not studied the chemistry of Titebond II and III. I think they cure by a post application polymerization. Titebond regular doesn't. It cures by coalescing polymer particles where the polymer was formed during manufacturer of the glue.

If I were testing this,

Keith Newton

I would keep at it until it broke. If you've already applied all the torque you can by hand on the clamp, then rip it in half and try again, which would in effect double your shear. If that wasn't enough, then half again.

After it breaks, inspect with magnification to see if fibers tear out of each face, which would be best, or did the glue-line fail. Just showing an unbroken test doesn't show me much.

Followup test

Bill Tindall, E.Tn.

I spread a thin layer of Titebond on a plastic key card and let it dry two days. Then I spread a second coat of glue over the first, covering it and extending the new coating beyond the edges of the first. I let this second application dry 2 days. This second application extending past the first gave me a tap to gab to try to pull the two layers apart.

If there was no adhesion of new glue to cured glue the layers would easily separate. The didn't. Then I flexed the two layer assemble and detected no separation. Flexing is a good test of two layer adhesion. This test clearly demonstrates some bonding of new to cured glue.

Is this bonding enough to hold a repaired dovetail or M&T joint together? I don't think it takes much shear strength to hold these joints together, but be your own judge.

How does the strength of a new to cured bond compare to a new glue to wood bond? I don't know and don't have the equipment or inclination to test reliably.

Re: Regluing Titebond and Liquid Hide joints

Don Stephan

Thank you for taking the time to make the test and share the results.

For me, allowing the first surfaces to sit (cure/polymerize/dry/. . .) for several months and then apply fresh glue and try to shear. If memory serves, I have in the past seen statements that some finishes can simply be recoated with a couple days of the last application, but after a longer period the surface needed to be scarified so the next application could make a mechanical grab. Seems like yellow glues might be similar.

Unrealistic experiment

Bill Tindall, E.Tn.

First, taking a joint apart is going to replasticize the glue or the joint would not come apart. It will not make the glue good as new but it won't be the equivalent to glue cured for 10 years either. Then there is the problem of letting it set 10 years. In the joint it will be marginally exposed to oxygen and not at all exposed to light. The combination of light and oxygen has a large affect on the surface chemistry of polymers, including paint and other finishes. Glue in a joint never experiences this degradation. Good thing too or no glue would last decades.

I have clearly shown that Titebond can bond to itself under the conditions described. How relevant this fact is to your woodworking will have to be decided according to your use of glue.

Personally, I have never had the occasion to need to take a joint apart. Reversibility, what ever that means, is not a relevant factor in my choice of wood glue. I am not willing to give up the desirable properties of Titebond, or its equivalent, in anticipation of the unlikely event that some decades hence one of my joints will need to be disassembled and reassembled and that I should make this unlikely event easier on the repair person. Unless this repair person is stuck in some previous century's technology I have every confidence they will get my joint apart and back together. The occasion to repair hide joints will be so rare that no one will survive the repair business unless they can do Titebond type glues. Disclaimer: I don't make chairs. I might have a different view of the situation if I did.

Re: Regluing Titebond and Liquid Hide joints

John in NM

This confirms much of what I started to suspect during the recent discussion - that the problem of joints repeatedly failing and being reglued has more to do with the joint than the glue and whether or not it bonds to itself. Joints that fail will likely continue to fail, regardless of the glue used.

This is just supposition however, I have no data to go with it.

It shows me something...

John in NM

That the reputation of PVA glues for inevitable failure when applied to dried glue is misplaced or at least overstated. A non-failure under considerable pressure shows that.

I would like to see the test repeated with week old glue and month old glue. For that one would need to test to failure, probably with a hydraulic press and multiple samples of each to look for a trend.

Chairs

jesse cloud

A totally unscientific speculation, but I would bet that if we took a random sample of ten chairs that were ten years old, half would fall apart as soon as we unscrewed the corner blocks.

I took a class in chairmaking where one of the activities was taking apart old chairs to discover how they were made. These were commercially manufactured chairs. The failure after corner block removal was close to 100%.

In support of premise.....

Bill Tindall, E.Tn.

Consider a well made M&T joint. What in fact does the glue do? It keeps the tenon from withdrawing. Racking strength is almost entirely provided by the joint.

Shown is a demo desk made by the Headley shop. For programs and demoes they start with a pile of parts and proceed to knock it together in about an hour. The only screws holding it together are the hinge screws and something on the feet . Accurate joinery renders it a serviceable desk without glue. Adding the glue would prevent taking it apart, but do little more to make it a desk.

Now consider a sloppy M&T joint glued with Titebond. The only structure resisting racking is a thick foamy glue layer. Stressed it will fail. With the old Titebond scraped out and reglued it will still fail. Epoxy might work.

Re: In support of premise.....

Hank Knight

I have been known to glue a sloppy mortise and tenon joint with epoxy (blushes). It works great, but I think disassembly would be difficult.

Re: Regluing Titebond and Liquid Hide joints

Robin Corell in Atascadero

I see one potential issue with your test and that is the initial application of glue which was just a surface coat and not subjected to clamping. we all know that PVA glue gets its strength from penetrating the cell walls of the glue and thus in a proper joint that has failed there would me no available wood fiber for subsequent repairs to bond to. I'm not convinced that glue spread on the surface and scraped off is a prue simulation of a properly clamped joint that has failed

If you had made sections and looked at them under a microscope of the initial dried glue and then compared that with a section of the "repaired" joint to make sure that the initial application of glue prevented the repair from having any strength from glue penetration of wood fibers.

Premise may be incorrect

Bill Tindall, E.Tn.

Mortise and tenon joints and dovetail joints are not clamped.

Virginia Tech has an adhesives program and once I had the occasion to ask someone from it this wood gluing question. Does Titebond type glue bond to wood by adhesion, or by mechanical means through pore penetration. The reply was that it could bond by either or both mechanisms and allowed it would be hard to separate the two possibilities. So, I can't say that pore penetration is necessary. At least this adhesives expert didn't know.

I don't believe clamping forces glue into wood pores. The role of clamping is to squeeze out all the glue that can be squoze out from the joint, leaving as thin a glue line as possible behind. When tested, simple rub joints with Titebond show nearly as much wood fracture as a clamped joint. My application of the initial glue would be similar to a rub joint, presuming scraping ==rubbing.

In the case of my test the initial application of glue had penetrated as much as it would have in a typical joint and there was enough left on the surface to make it look like it had been sealed with varnish. Hence any bonding of new glue to cured glue did not involve pores for there were no pores remaining to be penetrated.

Love it, Bill

Joe shelton

"Squoze".

Sounds like down home to me.

For foreigners, I'll amplify that the past participle is "Done squoze".

The glue stuff is interesting, too. I think I'll worry less about ultimates, and concentrate on not starving any jints.

More information from Titebond

Bill Tindall, E.Tn.

Some time ago a chemist at Franklin (mfg. of Titebond) sent me a stack of technical literature which I have read again. They specifically say that the urethane glue bonds by pore penetration. (I believe this statement is not entirely correct. The urethane chemistry should chemically bond to the cellulose in the wood)

They say regular glue penetrates a few thou. but offer no information on how important this penetration is to bonding. They do warn that too thin an applied glue layer and all the glue will penetrate and none be left for bonding. They recommend that glue be applied to yield a small bead of squeeze out.

There was an interesting discussion of what they call assembly time, that time between glue application and clamping. If the joint is pushed together quickly after glue is applied the joint will "tack" much quicker than if the joint is left open a while. Contrary to what the hot hide enthusiast claim, Titebond Regular can tack quicker because it has a solids content twice as great. TB-R is better gap filler for the same reason. Tack property of a glue that is in solution, like TB- liquid hide, is much slower than emulsion glues like TB-R.

Sizing is when glue is applied to end grain, like a miter, the glue allowed to dry to seal the end grain and then new glue applied. Sizing is recommended when using Titebond Regular and should not be done with TBII or III, the crosslinking resins. This recommendation suggests to me that TB-R bonds to itself, consistent with what I saw in the experiment.

The chemistry of TB-R is different from TB-White. And, the properties are substantially different. TB-R has much less creep, it is stronger, etc.

Clamp time varies with many variables. The bottom line is that the clamps stay on till the joint has developed enough strength to resist the spring back when the clamps come off. This time can be as short as a few minutes, or as long as days in high moisture content lumber.

and one more....

Bill Tindall, E.Tn.

If you take a joint apart before the glue has set, apply fresh glue before reassembly. This situation could occur in panel assembly when the panel pieces are discovered to be misaligned or become misaligned as the clamps are tightened.

Re: More information from Titebond

Larry Clinton At Frankfort, (Central) Indiana

Bill it is very obvious you are not a fan of Hot Hide Glue. I would have to disagree about the "tack time" as the hide glue is often used to repair musical instruments by simply applying the glue to the surfaces of the parts to be bonded and "sliding" or just applying pressure when mating the parts for a very few seconds. Tite bond (all varieties) will not properly bond in this time frame. Long grain joints can also be joined by sliding the surfaces together without clamping and have excellent bonding properties. I have done this and when testing always find the glue line holds far beyond the wood strength. As far as solid content, Hot Hide glue varies according to the amount of water added, however I am not certain solid content is a valid measure of tack time. IE: cyanoacrylate glues composed of monomers have polymers (solids) added, and as I understand this delays the Tack time.

I understand you have a far better knowledge of chemistry than I. However I do have one fact, Hot Hide glue has the history of holding wood together without failure far longer than

any of the "modern adhesives". Hopefully the Titebond's and yellow glue's will prove to have as long a life as the hide glue - but unfortunately none of us will live long enough to prove that.

not the way I would state my state

Bill Tindall, E.Tn.

"Bill it is very obvious you are not a fan of Hot Hide Glue." Actually, I am almost entirely ignorant regarding the advantages hot hide glue. I would reword this assessment to say, "I have little experience with hot hide and looking for some practical reason to gain some".

To clear up some misunderstanding.....

"Solids" is a measure of how much non-volatile stuff is in the glue. The recipe for hot hide is about 1/3 glue (solids) and 2/3 water. Titebond Regular is about 50% glue and 50% water. A dilution of only 5% water to Titebond has a dramatic effect on how quick it bonds, which can be used to advantage. Cyanoacrylate and epoxy are 100% which is why these glues are effective gap fillers. With the others as the water leaves a void if left behind if the glue layer is thick.

Polystyrene is dissolved in cyanoacrylate to thicken it. This thickening agent dilutes the cyanoacrylate and will lessen its strength if brittle polystyrene has poorer properties than cured pure cynoacrylate. Cyanoacrylate decomposes over time and should not be used for permanently gluing wood.

As water leaves a joint and diffuses into the wood the bond strength increases. The more water there is in a glue the longer it will take to achieve full strength.

Rub joints rely on "tack", that initial grab of the joint.....for better or worse. Hot hide has the property that it gels as it cools and when it gels the viscosity shoots up and the joint tacks. This weak bond is often enough to hold a part in position as it dries. I have seen this property of hot hide used to advantage in marquetry/inlay. It could be a useful property when clamping is impractical.

In my woodworking "tack" is more often a nuisance in joint assembly. That said, when applied properly, Titebond tacks quickly in a rub joint. The only place I use a rub joint is corner block installation. I don't see any difference in how rapidly I can install corner blocks with a rub joint with Titebond vs the demonstrations I have seen with hot hide. As soon as the excess glue is rubbed out of the mating surfaces the glue grabs, forever. I occasionally use a Titebond rub joint for temporary joining but found the bond too permanent, ie lots of wood failure like for hot hide. I have not seen any data showing that hot hide is any better than the Titebond types of glue for rub joints. Given how rare I use this joint I have not searched very hard to find it.

The bottom line of it all is that experienced woodworkers learn to use some glue effectively. With experience the glue's properties are used to advantage, and the disadvantages tolerated. Once one gains the experience, there is little reason to learn about another glue so long as the glue at hand fits the need. If I take up marquetry I will learn to make hot hide, for example.

The situation is different for a new woodworker. They should be provided information that is accurate so that they can make an informed choice.

Re: not the way I would state my state

Rum

"Cyanoacrylate decomposes over time and should not be used for permanently gluing wood. "

Bill, do you happen to have a good reference I can point at for that? I ask not because I don't believe you but because I keep having that conversation with a group of not-woodworker users of a specific set of antique tools who have a rather large fondness for CA glues and their continued use pains me somewhat. A solid reference would be quite useful to have.

Thanks.

replied by email

From the Getty Museum

Bill Tindall, E.Tn.

From presentation by

Don Williams Getty symposium on antique conservation

PS-cyanoacrylate trivia

Bill Tindall, E.Tn.

Don is not a chemist. Cyanoacrylate does not react with nitrogen. Nitrogen doesn't react with anything under normal conditions. The polymerization of cyanoacrylate is initiated by a host of chemicals, including water and water vapor, and probably chemicals in the wood, any amine, acids, bases, protein.

BTW, I checked Wikipedia. Much of what is written there is just wrong. For example cyanoacrylate bonds perfectly to perfectly flat surfaces. As a demo of its strength, the director of research at Kodak had two small cylinders expertly machined flat in the shop. One was fastened to a convertible and the other to a crane. A drop of cyanoacrylate was placed on the surface of the cylinder and the cylinders joined. The crane then lifted the car and rider into the air. This demo appeared on "What's My Line".

Re: Thanks for the clarification.

Hank Knight

I thought nitrogen was relatively non-reactive, so I was puzzled by the original statement that it was the curing agent for cyanoacrylate adhesive. Your follow up post clarified that. Thanks.

Re: PS-cyanoacrylate trivia

Yonak

Bill, if nitrogen is non-reactive, how is it helpful in fertilizer ?