Home Articles Forums Gallery Marketplace Videos

WoodCentral Forums

Knowledge You Can Build On — Est. 1998

Veritas - Innovation in Tools

The differences "white" vs "yellow" glue & TB III

Posts

The differences "white" vs "yellow" glue & TB III

#1

The differences "white" vs "yellow" glue & TB III

Bill Tindall

Two years ago at IWF I had a long conversation with the folks at Franklin glue company. I was led (mislead) to believe the differences in white vs yellow were minor. This year at IWF I had occasion to discuss this topic with someone more willing to provide details. I have more technical details if someone is interested but for now I will summarize.

The main adhesive component of all this family of wood glue is poly(vinyl acetate). They differ by things added to alter the set time or the properties of the cured film.

White: White is a simple mixture of two polymers poly(vinyl acetate) and poly (vinyl alcohol). The cured glue is softened by heat or water and therefore joints are readily "reversible". A cured film can function as a hot melt adhesive or a joint disassembled by moderate heat. This glue has a slow initial tack. Joints are prone to creep if stressed.

Yellow: The yellow is a result of a dye that has been added. However, other stuff is also added that makes yellow glue tack quicker (good for panel assembly), creep less than white and be more water and heat resistant than white.

Titebond III has a filler added that slows the drying of the joint resulting in longer working time. Good for mortise and tenons and the like.

Re: The differences "white" vs "yellow" glue & TB III

#2

Re: The differences "white" vs "yellow" glue & TB

Tony - memphis

Good info. Thanks for posting. So, what makes one more water resistant that the other?

Re: The differences "white" vs "yellow" glue & TB III

#3

Titebond original creepiness?

GolfSteve in Calgary

I'm building my first guitar now, so know very little. Luthiers are obsessed with glue, but most seem to think that Titebond original (white) has less tendency to creep than Titebond 2.

Re: The differences "white" vs "yellow" glue & TB III

#4

answer

Bill Tindall

The yellow glues have stuff added that provides a reversible crosslinking of the adhesive. It is not a permanent covalent chemical bond like what is found in cured Titebond II.

Re: The differences "white" vs "yellow" glue & TB III

#5

Re: Titebond original creepiness?

Bill Tindall

Disclaimer: I know nothing about building instruments.

Yellow glue will be much less creepy than white because the yellow is crosslinked by the chemical additives in this adhesive. Titebond II is covalently crosslinked so I would expect it to be even more creep resistant. However, this adhesive would be not easily reversed in a repair. White is the most easily reversible.

There is a local guitar builder. He gets thousands for a guitar so I presume he is good at what he does. He uses yellow glue, according to his report at our woodworking club meeting. Ditto for the local mandolin maker.

Re: The differences "white" vs "yellow" glue & TB III

#6

Bill, I have a question for you.

Keith Newton

I'm getting started on building a desk for a customer who had the state champion Loblolly Pine, before it got beetles and was sawn. The wood is really rich, and I have chosen edge-grain for the top, which would of course leave some really broad areas of late-wood in the glue-line.

I have never used pine for any of my furniture, but have lots for construction, where I used mechanical fasteners.

What do you think would be the best glue choice for this torch material? ha

Re: The differences "white" vs "yellow" glue & TB III

#7

I don't know......

Bill Tindall

suggest you contact this person at Franklin Adhesives:

Bob Behnke

senior tech specialist

BobBehnke@FranklinInternational.com

In any case I for sure would test what ever is recommended. Make a lap joint, let it cure and bash it apart and look for wood failure. Lee Valley once sold some epoxy that was good for oily woods. Vile smelling stuff but it worked.

Re: The differences "white" vs "yellow" glue & TB III

#8

David Barnett

(Message Deleted by Poster)


Re: The differences "white" vs "yellow" glue & TB III

#9

Re: Use a glue specifically made for luthiery

GolfSteve in Calgary

I have no problem using Titebond original - so many people (including the luthier whose plans I'm following) recommend it that I'm not worried. I should have ordered some LMI glue but didn't. If I put in another order I'll get some. Thanks for the advice.

What threw me off was that Bill reported that Franklin said Titebond original exhibits more creep than Titebond 2, which is the opposite of what I think I've read on the various luthier forums....who is correct? Classic internet forum conundrum. (*)

I plan to use hot hide glue for a few things, like bracing, the bridge, and the Spanish heel neck joint.

I'll use fish glue (from Lee Valley) for attaching the bindings and purflings.

I haven't decided which glue to use for attaching the fingerboard.

(*) here's a link to Franklin International response about creep on a luthier forum:

"By day I work as a Technical Specialist at Franklin International, so I field this exact question on a regular basis. I've recommended Titebond II exactly once for instrument construction. Why? It was for a redwood laminated top that will never need to be dismantled and Titebond II has an affinity for "sappier" woods. For everything else I recommend Titebond Original, Titebond Extend, or animal glue. All of our PVA wood glues have the same strength at room temperature, about 3200-3400 PSI, when tested in shear on hard maple. Since they all form a bond stronger than the wood we are literally testing the strength of maple blocks, which is the reason this strength range is the same for all three. At elevated temperatures (we test at 150°F) Titebond Original and Titebond II lose roughly half of their strength while Titebond III loses up to 2/3rds. Without giving up any trade secrets there is distinct difference in the strength of a joint made with Titebond III that has seen boiling temperatures at some point early in its' life compared to one that has only been to 150°. The best creep resistance of our wood glues is Liquid Hide Glue followed by Titebond Extend, Titebond Original, Titebond II, and Titebond III (note the relation to thermal resistance at 150°F, creep is a function of temperature). Animal glues are exceptional aside from their moisture resistance. Titebond Original, and non-catalized PVAs in general, can be dismantled using heat or steam. Titebond II and III, both catalyzed, will respond to heat but will not be the high point of your day.

In very simple terms: Titebond Original, Extend, or animal glue are best for interior applications such as musical instruments. Titebond II is excellent for outdoor use where pieces will be exposed to the elements. Titebond III is excellent where high temperature moisture resistance is needed, my favorite example being cutting boards. Despite your warning eventually a cutting board is likely to spend time in a dishwasher and Titebond III can withstand such treatment... such an example of an abused cutting board sits in my office and it displays no signs of separation."

http://www.luthiersforum.com/forum/viewtopic.php?f=10101&t=31065&p=411430&hilit=+titebond+original+creep#p411430

Re: The differences "white" vs "yellow" glue & TB III

#10

Re: Use a glue specifically made for luthiery

Bill Tindall

Since Titebond II is irreversibly crosslinked when it cures I would have predicted that it would have the best creep resistance. Since the person from Titebond says this is not the case, I would say my prediction is flawed.

I have tested Titebond regular in a creep test and with a 25 pound weight hanging off a 1 square inch joint for a month I could not detect any creep.

What people call creep in a joint is usually in fact a result of shrinkage of the wood after the joint has been made.

Re: The differences "white" vs "yellow" glue & TB III

#11

I did two sample test, and both were good.

Keith Newton

I had Tightbond original and TB II on hand, both of which were fine, so I guess I'll be using the II.

Re: The differences "white" vs "yellow" glue & TB III

#12

Re: The differences "white" vs "yellow" glue & TB

Hugh D. Evans

I'm a former Franklin employee and worked in the technical group until March of this year. Bill, your name sounds very familiar and I'm almost positive I've spoken with you at some point. If you're who I'm thinking of you requested an informational CD after IWF two years ago after talking with Dale.

Anyways, since departing Franklin I've founded a business that is focused on guitars (after consulting that industry for a bit over a decade.) I have no industry affiliations with respect to adhesives: So here's the straight dope on the differences:

White glues were the first PVA based adhesives and had little else to them. As a result they had lousy water resistance, were highly prone to creep, and generally wound up with a poor reputation compared to hot and liquid hide glues that were dominant in the woodworking industries at the time (mid 40s.) In order to improve their performance across the board (tack, creep resistance, thermal resistance, water resistance, to name a few) tackifying resins were added. In case you're wondering, this is all that "aliphatic resin" glues consist of. I often call them resin-modified PVA's because I consider it a less confusing term. This modification would not change the color on its own, so the decision was made to add yellow dye to differentiate them from white glues and is purely a marketing strategy. I have personally used dye free Titebond II and tested low quality yellow glues from manufacturers who have done nothing more than add dye to white glue. Uninformed individuals may try to convince you that yellow was added because it provided a better match to the surrounding wood. This is simply untrue. Not to mention, if you can see a glue line it's too thick. As part of some competitive testing I performed on Gorilla Wood Glue (not the polyurethane... The type II water resistant white glue) I found that not only was it nearly colorless and clear when dried but its performance was effectively identical to that of Titebond II. If you need something colorless for an exterior application I would not hesitate to recommend Gorilla Wood Glue. Titebond III is a different animal. It was actually the end result of a project to achieve type I water resistance from a PVA as a method to compete with urea-formaldehyde adhesives. It does accomplish it, but unless you're building a cutting board, countertop, or skateboard, I wouldn't recommend it. The product is prone to creep, and there's a reason its shear strength at 150°F isn't published: it's not great, usually around 1000 PSI which equates to a 2/3 loss of bond strength.

In terms of strength, at room temperature a good white glue can cause 100% wood failure just like Titebond Original, Extend, Titebond II, and Titebond III. Under ASTM D-905, they are also tested for shear strength at 150°F at which point from best to worst the list will look as follows: Titebond Extend, Titebond II Extend, Titebond Original, Titebond II, Titebond III. All PVA based adhesives are thermoplastics, which simply means they will melt (most of them reach their melting point somewhere above 250°F.) This is known as thermal plasticity, and in case you're wondering, it correlates exactly with creep resistance. It's also the reason that as a guitar manufacturer I personally use primarily Titebond Extend and Titebond 50. They're overkill, but if a guitar gets left in a hot car I can still feel confident that it wont come apart.

I've seen the term crosslinking a lot in this thread and some confusion about how it relates to creep. Picture plain PVA adhesive as wet spaghetti noodles: If you leave them to dry they will tangle around each other and shrink a bit, and you won't be able to pull the mass apart easily. Put that mess back into water and they will separate. Crosslinking PVAs would look more like spaghetti noodles that actually fused together as they dried and if you put them into water again those fused points (just like covalent crosslinks) would not separate. However, they do *not* crosslink with the wood itself. Instead, they have relatively weak electrostatic interactions (known as hydrogen bonds and van der waals forces) but in HUGE numbers, which is why they still create bonds stronger than wood.

I saw Bob's name mentioned, you're welcome to contact him if you wish. I helped train him, and unlike myself he comes from the sales/marketing world and still has an obvious corporate affiliation. In my case: I'm an entrepreneur with no specific allegiance to any adhesives company aside from my father who works for their industrial division. My Dad is the reason that despite being a younger guy I have been learning about the adhesives industry for over 25 years. I do my best to keep an eye on message boards because even though it's not technically my responsibility anymore, I care about getting good information out to fellow woodworkers. Not to mention, at over 5 months unemployed and just starting to get endorsements and contract work... It makes me feel like I'm still contributing something to the world.

I'm sure I've opened a can of worms, so feel free to reply or contact me directly. I'm always happy to help.

-Hugh

Re: The differences "white" vs "yellow" glue & TB III

#13

Re: Use a glue specifically made for luthiery

Hugh D. Evans

Actually, in that case, we're not talking about creep. If a glueline is excessively thick (>10 mils) it begins to expand and contract independently of the wood creating ridges. Creep results from a large stress, less than that required to cause immediate joint failure, over time and create permanent deformation of the glueline. There are a couple of ASTM specs to test adhesive creep. The most common looks like a ton of blocks glued together and stressed with a shock absorber spring that is tensioned and left to see if the assembly moves. No PVA has ever passed said test. Most that are advertised as wood glues are also more than sufficient for use in luthiery. My personal favorite is Titebond Extend.

Re: The differences "white" vs "yellow" glue & TB III

#14

Re: Bill, I have a question for you.

Hugh D. Evans

Keith,

It's really hard to go wrong with Pine just because it's so soft. Something along the lines of Titebond Original will be a great choice as would Titebond Extend. It will never hurt to wipe the gluing surfaces with acetone in case it's a bit on the oily side. Surface prep has never decreased bond strength. Use as much clamping pressure as the wood can take (stick a sacrificial piece between the clamps and wail on the until the fibers start to crush.) After that the adhesive bond should last longer than you or I will, which happens to be my definition of sufficient strength.

As for credentials, I'm a former Franklin Tech Guru and helped oversee Bob's training.

Re: The differences "white" vs "yellow" glue & TB III

#15

Bob

Re: The differences "white" vs "yellow" glue & TB

Bob

This is all good information thank you for posting it.

Question? Where does Hyde glue fit in the mix or epoxy for that matter? I know its used for veneers buy why not Titebond? Is there a book on Glues and where which is used and why I would really like to read one without all the chemistry, well maybe some where needed.

Re: The differences "white" vs "yellow" glue & TB III

#16

As a practical matter

Bill Tindall

I hung 25 pounds off a 1 sq in lap joint at room temperature and after a month saw no detectable movement of the wood.

Re: The differences "white" vs "yellow" glue & TB III

#17

Re: The differences "white" vs "yellow" glue & TB

Jesse Cloud

I can help on part of your question. Personally, I use hide glue on veneer when placing the veneer on the workpiece is going to be tricky, e..g. it has to fit just right or it wants to capture air pockets.

The main benefit of hide glue for me is that when I goof, as I often do, a little bit of heat from a hair dryer, iron, etc. will soften the glue so that the veneer comes right off and the glue stays soft long enough for me to put the veneer back on the workpiece. As they say in the post above, Ya gotta love that.

Not so much for me, but for a lot of people making period furniture, there is great satisfaction in using historically appropriate methods, and hide glue was the go-to glue for much of the antique furniture we love.

Having said this, I have also used many, many gallons of titebond on veneer with no problems.

Re: The differences "white" vs "yellow" glue & TB III

#18

If no PVA has passed the creep test

R Payne

is there a glue which has?

ron

Re: The differences "white" vs "yellow" glue & TB III

#19

Re: If no PVA has passed the creep test

Hugh D. Evans

Unfortunately, you've asked a question that is best met with another question. What creep specification do you need it to meet or exceed? Due to its plasticity, for instruments such as guitars or even pianos, creep is never a major design consideration for a couple of reasons: there generally isn't enough stress to induce it in the first place, and if any deformation does occur it typically undergoes elastic recovery (the technical term for this behavior is relaxation hysteresis.). The specifications designed to evaluate adhesive performance with respect to creep are all geared toward load bearing structural applications. When I say structural I'm not talking about a guitar... I'm talking about houses or wooden aircraft frames. The stresses involved are on the order of several tons. A steel stringed guitar might see 160 lbf... If we assume 5 sq inches of bond area for a bridge and no other reinforcement the gluelines will experience a stress of 32 PSI. I have personally tested PVA such as plain old Titebond original that cause 100% wood failure in shear using hard maple (Acer Saccharum) that failed at over 4500PSI. For creep to set in stresses must be near what is required for failure or the temperature must be greatly elevated. Does that make it any easier to understand why stresses of less than 1/100th of what they are known to withstand just aren't taken seriously as likely sources of failure?

For true structural applications, adhesives such as EPI, Resorcinol, Epoxy, Casein, and UF are all examples of chemistries that are known to have passed such tests historically. If you'd like to know more about why they work, let me know and I can elaborate.

-Hugh

Re: The differences "white" vs "yellow" glue & TB III

#20

Re: If no PVA has passed the creep test

R Payne

I don't have a need to meet or exceed. You mention PVA specifically so I got curious as to what has met the criteria.

It doesn't really matter to me if PVA meets or fails any test as I essentially stopped using it a couple years ago.

Thank you for answering.

ron