OK, let's talk about glues . . .

[rcc56]

The recent thread on headstock repair inspired me to bring this up. In one of the posts, one of our members mentioned a fish glue repair that failed twice. When one of my suppliers first started carrying the stuff a few years back, claiming all the advantages of hot hide glue without the problems of heating and immediate clamping, I was sorely tempted. But I held off, because I had not heard of the stuff, and didn't know if the long term results would be good. Since then, I have heard of many fish glue failures.

So I thought that this might be a good time to talk about glues.

Here's what I know about glues, from personal experience over a period of many years:

1. Hot hide glue-- great stuff if applied under the right conditions. A bit of a pain, and there's a definite learning curve. But it doesn't creep, and can hold for many decades, sometimes even for a century or more. But it's a recipe for failure if you don't dot all the i's and cross all the t's. In many cases, easy to disassemble when necessary. High bonding strength. Non-invasive to finishes.

2. Good old Titebond Original. An aliphatic resin-- Easy to use. Reasonable open time. Breaks down at lower temperatures than hide glue, but not unreasonably low. Creeps, but not too much. High bonding strength. Hard to disassemble without quite a bit of heat. Non-invasive to finishes. Will hold for decades.

3. Cyanoacrylate, also known as CA or super glue. Bonds nearly instantly. Invasive to finishes. Rather unpredictable stuff. One luthier states "cooks in five years." Not suitable for structural repairs, but good for inlays, loose fret ends, fingerboard divots, and the occasional shimming of mandolin bridges, but not much else. Some people use it for bindings, but then there's the danger of marring a finish.


4. The various binding cements, such as Duco and the specialized cements available from LMI, Stew-mac, etc.-- Good for bonding plastic to wood. Invasive to finishes. One problem is that so many different plastics are in use these days, it's hard to know whether you've got the right glue for the plastic you're using. I have also used Weldbond contact cement recently for a pickguard replacement on an old D-28.

5. Epoxies-- A big can of worms. Difficult to clean up, difficult to break a bond if necessary. Most people will probably gravitate towards quick cure varieties available in hardware stores, which are often not of high grade. If I find it necessary to use it, I'll probably select a high grade marine epoxy, formulated with wood in mind. Limited usefulness in lutherie.

6. Titebond liquid hide glue. Some of the advantages of hot hide glue, but risky for structural repairs. If you use it, it must be fresh. A bottle must be discarded after a year, and I recommend it be stored in the refrigerator. I generally use it only for non-structural repairs. I like it for fret jobs and gluing nuts, or to make paste filler around inlays. The last new bottle I saw had a coded expiration date, which I don't like at all.

7. Old Brown Glue. Another semi-liquid hide glue. I haven't used it. Must be heated, but only moderately, and has a longer open time than hot hide glue. Lacks the preservatives that some say are responsible for Titebond hide glue failures. Old Brown glue is hide glue with urea added. Considering the expense and limited shelf life, I have instead chosen to make my own in small quantities. I haven't yet settled on the best proportion of glue granules to urea. The last batch I mixed used 10% urea. To me, this is an unproven glue, and if I use it, I must be prepared for do-overs.

8. Wilsonart Lokweld melamine glue. An unproven quantity. Made for bonding non-porous materials to wood. Used by cabinet makers for bonding Formica to wood. Non invasive to finishes. Modest bonding strength. An unproven glue in lutherie. A cabinet maker friend suggested I try it for repairing loose bindings on finished instruments after I complained about the difficulties of using cements or CA without risking finish damage. I've used it maybe a dozen times. So far, so good, but I understand that I must not be surprised if I have failures with it. If it proves to be good over the long term, it might be a great alternative for binding repairs. But again, it is unproven. I hope a few others will take a chance with it so we can find out whether it is indeed viable.

9. Other glues: Elmer's wood glue and white glues are polyvinyls, as are Titebond 2 and 3. Anybody who has worked on many 60's and 70's Martins knows that polyvinyls are a poor choice for lutherie. I will not bother to discuss Gorilla glue and its relatives, except to say I used it once and will never use it again. I have heard enough stories about fish glue failure to satisfy my curiosity.

The above list comprises all the experience I have with glues. Polyvinyls, fish glue, and Gorilla glue have no place in my shop. I have not used epoxy in years, but I may use it again. The failed headstock repair in the other thread is a definite candidate for epoxy. I try to keep all the others in stock in my shop. I date the containers when I buy them, and generally discard them after a year, except for the cements.

Anybody else care to chime in on what they are positive as being reliable glues for our craft?

[multidon]

I think you have really summed up things well. But I did want to bring up one thing that puzzles me. Lots of instrument makers buy supplies from LMI (Luthier’s Merchantile for the uninitiated). Their proprietary yellow wood glue is polyvinyl acetate! On their website they go on and on extolling the superiority of this glue for luthiers. I have never tried to use polyvinyl for any repair projects, not because I have used it and been unsuccessful, but because I have heeded advice given here. I see no reason to try and improve on original Titebond; it is great and does the job splendidly. But why is LMI recommending a polyvinyl acetate glue if most luthiers won’t use it and consider it unsuitable? Or do some buy it because LMI is a trusted source and they assume it’s good? Or maybe they assume it’s comparable to original Titebond, even though it’s not? It looks like Titebond only because it’s yellow. You have to look at the MSDS to find out what it stuff really is.,

[HoGo]

To me each glue has its uses where it will outperform others. Often the criteria is not how well the glue will hold things together but rather what is traditional for given task (especially in classical violin making).
I use all of the glues (except those specialty glues for furniture manufacturers).
Ad1. HHG is the mainstay and I use it whenever possible for clean wood joints or cracks. The process is often misunderstood and many myths are out there like the ease of disassembly. The HHG is incredibly strong (folks call it stronger than wood) and disassembly of nice glue joint can be a nightmare if you don't know how to do it. The myth is fed by folks who misinterpret violin repairs where violins are often taken apart for repair. But the fact is that most good makers of violins are VERY conscious about how they prepare glue for each part of violin. They use different viscosities to get correct strength (or weakness) of given joint or thay use surface preparation to avoid too strong joint.
The best property of HHG is thatit bonds to old HHG joints with no problem.
2. The modern White or Yellow glues are all similar in uses and their strngth. Titebond is used by many guitar makers but others are as good. The creep at elevated temperatures has to be counted for and it's main disadvantage is that it doesn't bond to dried glue residue. Typically it will peel off easily from such surfaces. The waterproof variants are often a bit stronger and more heat resistant so luthiers don't like them as thay are harder to disassemble using moderate heat.
3. CA in some applications they are best performing, like gluing smal pieces of dark exotic woods like ebony or rosewood. Great for inlays and fillers (with wood dust) on those woods. Perfect for quick items like clamping cauls and small tools you make yourself and don't have to wait for glue to dry. Various brands can be very different and careful selection of thin/ viscous type is crucial. On light woods one must remember they are impossible to clean if they sink deep into wood, which they love to do, and the spots are sealed forever. They will typically stick perfectly to finish and you can use it to your advantage. If you spill it on finish it is best just quickly wipe away with finger and let the residue dry. It dries fast so it hopefully won't eat deep into finish and can be removed with fine sandpaper and polish. WEAR EYE PROTECTION!!!!! (DAMHIKT)
4. Glues for plastics are special items you need to know what plastic you have and select glue accordingly. Most plastics used in guitar/ mandolin making ase acetone sensitive so acetone based glues are typically used.
5. Epoxies are best for fillers and ugly contaminated joints where typical glues of choice won't work. There are many types of various strengths and durability (some are quite low heat resistant) so choosing the right brand is critical.
The rest of glues are not often used and sometimes their use is risky.
I would add one more though. These days almost forgotten but has many good properties of HHG and long working time. Casein glue. It also bonds well to dry residue of itself of HHG and can be quite water resistant (depending what base you use to mix it)
There is infinitely more that can be learned about glues. Experience is what will teach you...

[multidon]

Wasn’t good old Elmer’s White Glue a casein formula back in the old days, before they switched to polyvinyl? Does anyone make a ready to use bottled casein glue these days?

[sunburst]

Nothing to say about specific glues and their uses, but just a general statement:

If we study and learn glue theory; the 5 phases of a wood glue joint, how glues and adhesives work, relative adhesive vs cohesive properties of different glues and adhesives, mechanical properties of glues and adhesives, what materials are suitable for what glues, how to properly make a joint and prepare it for glue, temperature concerns, moisture concerns and so forth, it becomes easier to choose the right glue for the right job. We can then learn to properly use our chosen glues and adhesives and have confidence that we have done our best.

No one glue or adhesive is best for all joints, seldom is one specific glue the only one suitable for a given joint, what is best for one situation may be unsuitable for another, etc.. There is a lot of published misinformation and anecdote to sort through so learning and understanding all we can about our glues and glue joints is our defense against it all.

[rowka]

^
This

[Jeff Mando]

Great descriptions of glues and their uses! And, I agree -- use the right glue for the right job -- Dan Erlewine, himself, oftens uses three different glues for a single repair in his videos (of course, he is in the business of selling glues, applicators, clamps, etc.....)

I would break it down even more simply by saying Original Titebond is what you reach for 95 percent of the time, because it works for stuff made of wood.

Sure, I use a tiny drop of superglue once in a while for trim or the odd piece that won't stay down, but never for repairs requiring strength.

Hide glue is great for old-tyme restoration and is currently used by many boutique makers, BUT Martin still uses Titebond on all guitars EXCEPT the super high-end signature models and historic tribute models. Only those, do they use hide glue. So that should tell you a lot, right there.

[multidon]

In addition to what Jeff says, I will point out that Gibson Acoustic (talking about the guitars made in Montana, NOT the mandolins made in Nashville) uses Titebond on every joint of the guitar EXCEPT the dovetail, which is glued with hot hide glue. I assume this is to facilitate a neck reset should one ever be needed.

I would love to have someone answer my question about LMI glue. If polyvinyl is bad, why is it sold by a luthier supply company?

I did a little research on casein. I found one site that said it fell out of favor because it was vulnerable to being attacked by bacteria.

[Greg Mirken]

I can add a bit of information about epoxies; specifically a limitation of any of the hardware-store versions [including JB Weld]. You actually need much less hardener than resin, but it's very difficult to eyeball a 4 to 1 or 5 to 1 ratio of resin to hardener. So inert material is added to the hardener so that equal proportions of each are used. I don't know what is actually added, but I suspect it results in a compromised cured material. The beauty of West Systems marine epoxy is the pump cans that automatically give you the right ratio of hardener to resin, every time. [They also have a kit with pre-measured packets for occasional use.] Compared to other brands of epoxies I have used in the past, the cured West Systems epoxy is hard as a rock. NFI, by the way.
I don't use epoxy very often, but for some occasions like nasty, contaminated breaks, it's the right stuff.

[Jeff Hildreth]

In 30 years I have never had a failure with fish glue.

Gorilla glue should be banned.

[pops1]

I use mostly HHG and only on occasion or for certain things super glue, epoxy, or titebond. I do use gorilla glue super glue gel when I need it as it lasts much longer in the bottle, you have to shake it before use, than other super glues and I have much less failure.

[sunburst]

Not to be too argumentative, but the main point of my post above is: if we learn about glues and how to use them, we need not know nor particularly care what Martin or Gibson or whoever uses in their factory. We can choose our own glues for our own purposes based on our knowledge of how they work. Factories must choose limited numbers of general purpose glues and adhesives to streamline production. That's fine as long as the limited glues are at least adequate for all purposes that they are used for, but as individual builders we don't need to do that. We can choose what we understand to be the best glue for each individual joint and situation.

Edit:
Before someone brings this up, when we are doing repairs it is helpful to know what glue Martin, Gibson and others used. My post is directed toward building.

[rcc56]

Also, bear in mind that Martin used an inferior glue from circa 1965 to 1989 [that's nearly 25 years, folks], and Gibsons made in the late 40's and 50's are notorious for glue joint failure.

Anyone who uses white polyvinyl to re-glue a loose bridge on a '70's Martin because that's what the used at the factory is shooting craps.

Pay attention to what John has said in his posts. He knows what he is talking about.

Also, repair often requires a different approach from new construction.

Sorry, I have no experience with LMI's glues. But it might be instructive to know that LMI is now owned by Guitar Center/ Musician's Friend, that they only sell glues that they have exclusive rights to, and that they may be making a higher profit on their stuff than they would if they were selling Titebond.

And to fill in something John mentioned in passing, Martin used hide glue until '64 or '65, experimented briefly during '65, then went to polyvinyl until sometime in '89, when they switched to Titebond. Gibson used hide glue at least through the mid '50's, but for many years they either used a low grade and/or they mixed it too thin. I don't know when they dropped it, but in the early '70's they briefly used an RF cured glue that was one of the strangest things we've ever seen.

[O. Apitius]

I visited the Martin plant in the early 80's and I have a clear (to me) memory of them using HHG to glue on tops and backs. They used heat lamps to warm up the linings on the rims. It was explained to me that the reason HHG was used for this operation is that it was easier on their router bits when the binding channel was routed.
I'll admit that my memory is not 100% reliable but the image of those heat lamps is pretty clear in my head in fact, that is where I got the idea to use heat lamps in my own glue-ups.

signed: Lucid Dreamer

[O. Apitius]

Hide glue is great for old-tyme restoration and is currently used by many boutique makers, BUT Martin still uses Titebond on all guitars EXCEPT the super high-end signature models and historic tribute models. Only those, do they use hide glue. So that should tell you a lot, right there.

It tells me that Titebond is better for ease of production purposes but HHG is a superior glue for stringed instruments in most other respects.

[rcc56]

I visited the Martin plant in the early 80's and I have a clear (to me) memory of them using HHG to glue on tops and backs. They used heat lamps to warm up the linings on the rims. It was explained to me that the reason HHG was used for this operation is that it was easier on their router bits when the binding channel was routed.
I'll admit that my memory is not 100% reliable but the image of those heat lamps is pretty clear in my head in fact, that is where I got the idea to use heat lamps in my own glue-ups.

signed: Lucid Dreamer

I don't doubt you at all. I was going by Martin's own history book, which apparently has some flaws. I have heard elsewhere that they used hide glue for mounting tops and backs to the sides, at least for a period of time, but I don't know what years they were using that procedure. I'll start looking for glue squeeze out on the linings of Martins made after '65 when they come into my shop.

I do know from personal experience that in the later 60's and 70's, they were using polyvinyl for bridges, braces, fingerboards, and dovetail joints. I will guess that they continued this practice through the 80's.

I just realized-- I can't remember the last time I did any structural work on an '80's model.

I don't recall seeing any signs of hide glue or any yellow glue anywhere on the last two that came through, which were both from the early '70's. Both instruments had loose braces, and I gave one a neck reset. I used Titebond for the braces after cleaning the joints as best as I could, and hide glue for the neck set.

My early '65 12-string appears to have hide glue throughout the body, but the neck was attached with a yellow glue.

[multidon]

I have taken John’s advice and started doing lots of reading about glues and how they work. Talk about going down a rabbit hole! I came across some serious chemistry stuff. More on that in a bit.

I found an old archived thread here where the discussion was whether to use HHG or Titebond firba headstock break. John gave a great explanation of adhesion vs cohesion. Consensus seemed to be, epoxy is inferior to the others in terms of adhesion (sticking to the materials being joined) but superior in cohesion (sticking to itself). So if there is missing material and gaps need filling, then epoxy can be a good choice. If the surfaces are clean and a very close fitting joint is possible, it is not the best choice because it does not do well in a very thin layer. Both hide glue and Titebond are superior in this.

As for ingredients, good luck finding those. It is really interesting to see how manufacturerers view the MSDS requirements. There are those who disclose as little as possible, and those who go overboard and report as much about their ingredients as possible. Those companies who choose the latter option cover themselves legally much better in the case of unforeseen circumstances (such as a material not known to be hazardous now being re-designated as such in the future), but also risk copycat products being developed by competitors. Those companies who only disclose the bare minimum information required play it close to the vest and supposedly guard their proprietary formulae. IMHO, this is foolish. There are no secrets. If any company wanted to copy Titebond original, all they would have to do is buy some and send a sample to a lab for analysis.

It turns out that Franklin, maker of Titebond products, is one of those companies that plays it very close to the vest. There are no ingredients at all disclosed on Titebond Original’s MSDS. They can get away with that because nothing in it has been proved to be hazardous in the amounts present. So we are using a glue, and we do t know what’s in it. At least we know what’s in hide glue! But we are told that it is “aliphatic resin”. What the heck is that, anyway? Aliphatic just means “fatty”. And “resin”, that covers a whole bunch of natural or synthetic organic substances, liquid or solid, that are able to be converted into polymers. So the term “aliphatic resin” is a vague designation, kind of like “varnish”, another term that is so general that you really don’t know exactly what you have. Based on what I read on line (don’t know about accuracy, it’s hard to tell on the inter web), Titebond Original is chemically similar to polyvinyl acetate, but the molecules have been modified with aliphatic epoxide molecules. The resulting molecules are in long chains, because the aliphatic epoxide modification prevents cross linking. The fact that they don’t cross link improves the working characteristics in some ways. Titebond 2 and 3 are true PVA glues, not modified, and they cross link when cured.

Please be aware that I am no chemist. Maybe we have some chemists out there that will check the accuracy of what I’m saying. I just think it’s interesting that so many of us swear by something, and we don’t even really know what it is.

[Nevin]

A question and a point. My understanding is fish glue, like HHG is a gelatin based glue. How does it differ from HHG.

People point to Martin using Titebond. Martin is a factory with all the production concerns that implies. As Sunburst has pointed out they have made the determination to use that glue for those purposes based on their needs and desires. These can be very different from what a small shop needs or what a person doing repairs needs.

[multidon]

More research! Interesting stuff.

Many sites call Original Titebond a PVA glue, or refer to it as being "in the PVA family". At least 2 sites refer to it as "modified" PVA. So you guys who say you would never touch PVA, you're sticking with good old Titebond Original, guess what? Of course it's been "modified", if that makes you feel any better.

Also, Gibson's site calls the glue they use "Titebond 50". Never saw that in stores. So, it seems that this is a special formula not sold in stores, only in industrial quantities. I saw it in 5 gallon pails. Some sites say it is a lot like Original, but engineered for faster cure.

[Bernie Daniel]

....Hide glue is great for old-tyme restoration and is currently used by many boutique makers, BUT Martin still uses Titebond on all guitars EXCEPT the super high-end signature models and historic tribute models. Only those, do they use hide glue. So that should tell you a lot, right there.

Likewise Gibson using hot hide glue only on the top of the line Master Models (MM and DMM) -- all the other mandolins are glued with Titebond.

[Bernie Daniel]

....It turns out that Franklin, maker of Titebond products, is one of those companies that plays it very close to the vest. There are no ingredients at all disclosed on Titebond Original’s MSDS. They can get away with that because nothing in it has been proved to be hazardous in the amounts present. So we are using a glue, and we do t know what’s in it. At least we know what’s in hide glue! But we are told that it is “aliphatic resin”. What the heck is that, anyway? Aliphatic just means “fatty”. And “resin”, that covers a whole bunch of natural or synthetic organic substances, liquid or solid, that are able to be converted into polymers. So the term “aliphatic resin” is a vague designation, kind of like “varnish”, another term that is so general that you really don’t know exactly what you have. Based on what I read on line (don’t know about accuracy, it’s hard to tell on the inter web), Titebond Original is chemically similar to polyvinyl acetate, but the molecules have been modified with aliphatic epoxide molecules. The resulting molecules are in long chains, because the aliphatic epoxide modification prevents cross linking. The fact that they don’t cross link improves the working characteristics in some ways. Titebond 2 and 3 are true PVA glues, not modified, and they cross link when cured.

Please be aware that I am no chemist. Maybe we have some chemists out there that will check the accuracy of what I’m saying. I just think it’s interesting that so many of us swear by something, and we don’t even really know what it is.

Just to clarify that point "aliphatic" refers to the type of carbon bonds like the saturated aliphatic ( -CH2-) bonds in a hydrocarbon chain such as ethane or it could be unsaturated aliphatic-- e,g, (-CH=CH-) bonds in ethylene for example -- this specifically excludes the other kind of carbon bonds that is aromatic carbon linkages as found it hydrocarbons like benzene or napthalene. (each bond type e.g., saturated or unsaturated aliphatic and aromatic is a different mixture of s and p orbitals -- information likely to be of little use here).

So the choice of the term "aliphatic" in the glue class probably refers to what Franklin knows about the chemistry of the components of this glue -- information as you point out that they are not sharing -- perhaps because their product which existed long before MSD requirements might be "grandfathered"?

[Tavy]

Not much to add... except that a badly glued joint will come apart alarmingly easily whatever the glue. I've taken apart plenty of old bowlbacks that were all HHG but had many of the key joints fixed with so little glue that with a tiny bit of pressure the whole thing just pops apart. Maybe this is a hangover from violin making, or maybe it was just really sloppy building, or maybe both.

On the other hand... properly applied most of these glues are "stronger than the wood". Recently I decided to disassemble a body that was a failed build and had been sitting in the cupboard for a year or more.... figured I could re-use the rims for an experiment and ditch the top and back... well let me tell you I simply could not get the thing apart... in the end I snapped the rims trying to disassemble and had to bin it. That was all Titebond other than the plate joints and braces which were HHG. I'm not sure whether to be pleased at the strength of my joints or alarmed at the prospect of ever having to disassemble one

[sunburst]

Likewise Gibson using hot hide glue only on the top of the line Master Models (MM and DMM) -- all the other mandolins are glued with Titebond.

As much as I like hot hide glue, despite the fact that I use it for most joints in building and many in repair, despite the fact that I believe it is the best glue to use for the majority of joints in lutherie, the reasons Martin and Gibson use it only for top end products (and often publish the fact) are marketing-related.

Prewar Martin guitars and Loar-signed mandolins were made using hide glue, that makes it the best, right? That must be why they sound so much better than anything even the best builders can make today! (In case you can't tell, I'm saying that tongue-in-cheek to point out the reason for marketing "hide glue construction" as somehow being superior and leading to better sound. The simplistic notion among buyers that superior sound can be attributed to materials used is pervasive.)(FWIW, I've used red spruce... oh sorry, make that "Adi" for tops since the early 1990s, and I've used hide glue for at least as long. I never marketed my instruments as being superior because of it... guess I missed that boat.)

[Bernie Daniel]

As much as I like hot hide glue, despite the fact that I use it for most joints in building and many in repair, despite the fact that I believe it is the best glue to use for the majority of joints in lutherie, the reasons Martin and Gibson use it only for top end products (and often publish the fact) are marketing-related.

Prewar Martin guitars and Loar-signed mandolins were made using hide glue, that makes it the best, right? That must be why they sound so much better than anything even the best builders can make today! (In case you can't tell, I'm saying that tongue-in-cheek to point out the reason for marketing "hide glue construction" as somehow being superior and leading to better sound. The simplistic notion among buyers that superior sound can be attributed to materials used is pervasive.)(FWIW, I've used red spruce... oh sorry, make that "Adi" for tops since the early 1990s, and I've used hide glue for at least as long. I never marketed my instruments as being superior because of it... guess I missed that boat.)

I totally agree John. I would bet money that almost no one could distinguish a red spruce topped, HHG constructed and varnished F-5 from an equally well crafted Sitka topped, titebond constructed, lacquered F-5 unless they could see it or know before playing it. But we've all been over that one a million times.

As you say definitely a marketing ploy.

[pops1]

But..... it's HHG, it's older, and was used longer, it has to be better. My world is falling apart. Bad glue I guess.