Titebond 1: indoor
Titebond 2: water resistant
Titebond 3: water proof
What is a reason for using titebond 2?

Its been made obsolete. I never found it to be great for exterior use. It could probably be used for edge to edge glueing under paint.

Radio frequency curing can be done with TB2. Unfortunately our shop stereo never seemed to work...

I have just been thru a 2-3 month trial/tribulation on gluing Western Red Cedar. It started after some routine laminations with TB3 failed 100%. That is, after overnight clamping, the joints fell apart. The glue was still wet, and had not penetrated the Cedar. In this case the failure was on about 75% of the parts, about 300 l/f of glue line total. We had a couple of minor failures prior to this and chalked it up to those mysterious demons that live in every shop....

I suspected bad glue and called Titebond. I ending up having a few good conversations with their tech guy, and one of the things that came up was using TB3 for exterior doors or other millwork - on any wood.

Due to the heat of the sun and prolonged exposure, I was told that the propensity to creep with TB3 might work against us, and that TB2 would be a better glue to use for that work. I have seen TB3 let bent laminations become unbent so to speak, and never did like the rubbery end product.

The old waterproof vs water resistant is also tricky. Sort of like saying a 30% chance of rain - is it going to rain or not? Very little of our work is going to need true waterproof - boats, etc. If our door joints are exposed to enough water to deteriorate TB2 bonds, then there are more serious problems already at hand.

As for the Cedar problem, we were told to wipe down the surface to be glued with acetone or similar solvent and paper towels until no brown came of onto the towel, then glue as soon as possible, with either TB1, 2, or 3 - depending upon our needs. This is to remove the oils on the surface that some (note that word - 'some') Cedar may have.

The alternate is to go to a solvent based glue - polyurethane, epoxy or resorcinol to avoid the oil removal steps.

We have tried wiping down with acetone and still had about 10% failure in another 300l/f of 2"wide glue line. So we now use epoxy for our exterior Cedar work, and the problem is gone.

A friend said that TB 1 also comes out of clothes...big plus. Thanks

David,

We have always used epoxy for all true exterior stuff. No need to de oil cedar or redwood. Epoxy is fine.

I haven't had any issues with TB3; however, we rarely use it.

Years ago I did some testing on cedar after a conversation with Titebond tech people. It overlap faced glued some 4' boards and then broke them apart.

I had the same experience of glue failure although of the three.....regular Titebond, TB2 and TB3..... it was TB3 that worked the best. However after those tests we went to Polyurethane and West System as the only adhesives we use on cedar.

BH Davis

Many years ago, before TB3, I glued up some exterior curved handrails in Locust, with TB2. Out in the rain & sun for more then 15 years, they have now started to come apart. I'm sure that if I had maintained a good coat of anything on them, they would still be good. Of course, I'm gonna re-glue with TB3.

The shop I work in has also had some problems with #3. We had some door panels that failed on multiple attempts. They were Sapele panels stained dark in an entry door that was behind a storm door (and was placed so that it got full sun). We were told that #3 will not fully cure unless it hits a certain temp. The heat from the sun hitting the storm door was heating the glue, but only enough to make it gummy and allow the joints to fail. I made some samples and left them on a flat roof under glass for an experiment, Devcon 5 minute epoxy, Titebond 2, and Titebond 3 all failed. The only glues that held were West System Epoxy and Gorilla glue.

Chip - I recently learned in conversation with a TB tech that TBIII does not do well in heat, that TBII would be a better choice above 120 degrees.

There is a tendency with a light weight wood like cedar to use too much pressure. We never can squeeze glue into the wood even with a lot of pressure. However, with a lot of pressure we can squeeze out too much adhesive, giving a starved joint which will be weak. All three adhesives are stronger than wood. So, a failure means that something is wrong with the surface prep (for example, oil, but cedar has very little oil) or glue joint processing (pressure, for example).

I will not pretend to be a wood or glue expert, but I thought the idea of starving a joint by excessive clamping pressure had been debunked. I remember reading a fine woodworking article in which they attempted to test this and their conclusion was: clamp the s#!% out of it, within reason. This made sense to me, as I understand that glue like tb has no real strength by itself, but depends on the two wood pieces cross linking under direct contact. Epoxy is obviously another animal altogether. I think we starve joints by not feeding them enough glue, maybe hoping for less squeezeout/cleanup. I use tbII and a lot of heavy clamping on almost everything and have never had a failure after curing.

You can apply too much pressure and loose strength...substantial strength. Indeed a joint can have too much pressure. A glue line thinner the 0.002" is considered too thin to develop maximum strength. Thousands of tests have shown this.

What does help, is that with many adhesives, the maximum joint strength is 1.5 times high than the wood strength, so you can loose 33% of the strength and still have wood failure. The problem is when there are other issues affecting the joint strength, such as an aged surface, an uneven surface (1/1000" type of unevenness), etc. Then you do not have the 33% excess. In a lab test done perfectly, you might have 33%, but in production probably not. Further, in some glue joint uses, you do not need 100% of the strength...the stresses are low. (Example. A table top is typically not under high stress unless someone is dancing on top of it, but a chair leg is.). So, I hope this explains why some tests indicate one thing, but in practice there are different issues and concerns that must be factored in. A starved (lack of glue), and weak joint because of this, is indeed possible.

Note that when applying pressure, oftentimes right at the pressure point you might have too much pressure and a thin glue line, but away from the pressure point, due to the elasticity of wood, more reasonable pressure and a good, strong joint. For this reason, uniform joint pressure is a good objective to achieve uniform joint strength.

High pressure does not squeeze the glue into the wood.

Pure academic curiosity, but how does one measure the thickness of a glue line?

I have my Starrett digital calipers in one hand, and am scratching my bald head with the other.

I can picture a method that would create a long oblique angle across a glue joint and somehow measuring the length of the glue line vs the angle, but it seems prone to wide variables.

You can measure thickness by looking at the joint with about 100 x magnification.