ARTICLES & REVIEWS

Bill Tindall wondered: It is often stated that an advantage of hide glues over aliphatic resin glues (white/yellow) is the ability to disassemble joints glued with them. Some time ago I asked how this was done, for example on a tight mortise. I was pointed to a link describing disassembly of some musical instrument part. It involved drilling the M&T and injecting steam while pulling on the joint with some mechanical device. It struck me at the time that heat or steam will likewise soften an aliphatic resin glue. Steam is used to clean glue from clamp stations in furniture factories.

Does anyone know for a fact that a sound hide-glued M&T is more amenable to disassembly than a joint made with the ubiquitous yellow glue? After reading the process for a sound hide glue joint I wondered if there was a significant difference.

Bruce, a MN Galoot: I've heard that vinegar would hasten the demise of a PVA joint, but I didn't know that steam would work. I suspect that the steam used in cleaning furniture factory clamping stations is probably higher pressure and more saturated than steam available to the home user. Steam can be dangerous stuff.

On the other hand, I have disassembled hide glue joints with little difficulty, just hot water and reverse clamping pressure. It's just a little scary how easily they came apart. But I sure appreciated the ease of disassembly. I've never been able to get a PVA joint apart without destroying something in the process.

Bill Tindall: Did you soak the joint (wood) in hot water—minutes, hours? For example, suppose I wanted to disassemble a M&T frame. How would the joints be treated? Submersion in hot water? It is not clear how you got the hot water to the glue? And did I understand correctly that a similar treatment did not soften yellow glue, thereby validating the difference?

Mike Recchione: One very nice thing about hide glue joints is that you can not only take them apart, you can put them back together without going back to bare wood. What I've done is to get the excess off with hot water on a rag, and then just apply new hot glue. It combines with the old glue to give a joint that's as good as the original. So, for things like musical instruments, where you may need to take them apart and put them together multiple times, it's still probably the best way to go. I've done this for a chair I was repairing as well as a guitar, and it worked very well both times.

William Duffield, on the Cohansey: I dump out the cold coffee, refill the coffee cup with water, nuke it to boiling, take the cup back to the bench, and apply hot water to the surface of the joint with an acid brush. It wicks right in, the hide glue soaks it up and softens. Usually, by the time I've applied it to several joints, the first joint is soft enough to start to move.

David Barnett - Venice, Florida: As Adam Cherubini points out in Chair Restoration—Hide Glue is that hide glue can be reactivated, making it easy to reglue a loose joint without having to scrape (or worse) the surfaces free of aliphatics, such as Titebond, or build the surfaces to a tight pressure fit. For some applications, hide glue's reversibility is critical. Luthiers, for example, must take apart and reassemble very old and valuable stringed instruments, and only hide glue makes this possible year after year. Knox brand gelatin's great for this, by the way, and comes in handy packets

John Jesseph: I suspect this refers to getting the necks off Martin-style guitars. I have not done this, but is is common practice to remove the fifteenth fret and drill a small hole to let in steam. There is no better access to the joint, based on the relationship of the back, sides, and fingerboard. This is another example of something not originally designed to come apart—not nearly as simple and elegant as a violin. There is a lot of online violin forum discussion about cussing people for using PVA etc. glue as Mike R. said, as much for regluing as much as dissassembly. Mortise and tenon is pretty straightforward, and the chances of ruining the object is not as high as a guitar.

Todd Stock: First, we are talking hot hide glue, versus the stuff in a bottle that is pretty much good only for non-structural applications such as glue size, etc.

My assumption is that you're asking if, in the event of a partial glue line or partial tenon failure in a M&T joint, which adhesive offers an easier repair. I'd say hide glue, with some reservations on where it is used.

Adam Cherubini: You asked for experience and mine is limited: I had great trouble removing/reversing recently glued hide glue joints where the joint was sloppy to begin with. I was using straight glue, no additives of a moderately low gram strength (I buy 192 from Joel but it degrades—or so I'm told—with repeated heatings). When I say “trouble” I mean failure. I broke the parts. In tight fitting joints on the same piece of furniture, reversal was no problem. It appearing that very little glue was in the joint.

I've taken apart rub joint glue blocks, installed cross grain. I applied hot water, waited seconds, then struck said glue block with my favorite patternmaker's Warrington hammer (I want to make sure you have all the details, Bill). Clearly the water reversed the glue since the blocks came right off cleanly. I should mention that I also sang an Olde English work song before striking the glue blocks which probably had the same effect as the hot water on reversing the glue. I suspect others have sung this same tune to achieve similar results (and similarly correctly attributed the result to the combination of the chemical and meta-psycho reaction).

Here's my guess: In a tight and craftsmanly M&T joint (I know you are all planing your tenon cheeks smooth with your new Lee-Valley shoulder planes—I wonder if LV will make a cheek plane soon?), the hide glued joint will be easier to remove than the PVA glued joint. Hide glue extrudes out of piston fit joints like these. On a loose fitting joint, the PVA will be easier to dissassemble because M&T really can't be clamped and PVA just won't develop properties over a significant gap. Conversely, that same sloppy joint may well be next to impossible to disassemble if hide glue is used.

Moxon said craftsmen added beer to their glue pots to give it strength. Alcohol would break the inter molecular bonds and weaken the glue. Almost anything you add weakens the glue. Our experience reversing old joints may be predicated on the poor quality of the old glue. Salts added to hide glue slow the kick time and weaken the glue, but they also make hide glue more hydroscopic which makes it even easier to reverse (double whammy—low bond strength and hydroscopicality—okay I made that word up).

Hide glue will reactivate in water—the addition of heat speeds the process, but the key ingredient is water. Soak the joint with warm water, and eventually, the hide glue will liquify. Add some fresh glue to amalgamate with the reactivated old stuff and clamp.

Tightbond (TB 1) will take longer to soften, and cannot be reactivated—it's got to be removed down to the wood surfaces in the joint prior to regluing. TB II/III are both going to require lots of heat to soften the glue line because they are water resistant (and TB III is actually pretty resistant to temps under 212° F), so in this case, steam is used to transfer the heat energy involved in the phase change from vapor to liquid to the wood. You could transfer the same amount of heat by wrapping the joint with the silicon filament heating blankets used by luthiers to remove bridges, fingerboards, etc.; however, chances are good that this would toast the finish around the joint as well, so not a great solution.

For what it's worth, the reason why the procedure for guitar neck removal is identical for TB and hide glue is that the dovetail is shorter than the socket, creating an air space under the 13th or 15th fret (12th fret or 14th join to body, respectively). This space traps both the steam and liquid water, creating lots of heat (to soften the TB), and trapping moisture (to liquify the hide glue and further desolve the TB).

The only problem with leaving space between the tenon and back mortise wall—say in a joined chair—as a feature to ease future repairs is that the tenon length versus mortise depth really needs to be known by the repair guy to make things work. Some cabinet makers provide wood/finish/glue type info to their customers, but it's unlikely that the as-built plans will be included as well.

To further complicate the choices, TB (all formulations) is more shock absorbant than instrument/general woodworking grade (gram strength) hot hide glue, so there are some good reasons to use aliphatic resin glues in furniture that gets slammed around. One of the reasons why I use a harder drying clear version of TB from one of the luthiery suppliers is that it is almost as hard as dry hide glue, but remains a little more shock tolerant. I also have gone back to TB 1 for furniture construction.