Walnut and Glue

[Tanus]

I've done a bit of glue-ups. Most have been successful. Some have been disastrous, but with each failure, I've been able to figure out what I did wrong and not repeat it too many times. All have been with pine, maple, or oak.

At this point, the glue-ups I find easiest to do are laminations or edge gluing. If the surfaces are clean and planed smooth, and the clamping pressure is enough, the joint will hold for me.

The other day I went to the kindling pile to get some wood for turning a knob. I didn't have the size I needed, so I decided to take 2 pieces and face-glue them together to give me the thickness I needed. I'm fairly sure the pieces were walnut. The glue was Titebond III

The first two pieces fell apart by hand. I was in a hurry and used weak clamps. I also didn't wait long enough for the glue to bond.

The second two pieces went better but fell apart when I was turning on the lathe. Same for the third set. Both of the second sets had much stronger clamps and I waited a full day.

In all cases, the faces were flat and smooth, planed by hand.

What am I doing wrong?

Tanus

[Leon]

Well, Walnut has nothing to do with it. The glue may not be fully cured and with a joint spinning and pulling it could fail if not properly cured. Walnut is NOT incredibly strong, did the joint fail or did the wood fail close to the joint?

[dpb]

Tanus wrote: ...

Well, that's obvious enough...

I don't really have a clue -- a glue joint of that type should be possible to make hold stronger than the wood itself simply w/ a "rub fit" even w/o clamping at all.

I have only a few guesses (and they're all simply that w/ no more data or observation)...

1. The source of the material from the woodpile makes me wonder--was this dry lumber or scrap firewood that could possibly be wet? Seems unlikely, but we're searching here...
2. The glue is old or your simply not using sufficient amount to get good glue surface. Since same symptom w/ "weak" or "stronger" clamps, I'm assuming you're not clamping to the point of glue starvation, but that's another possibility.
3. Surfaces aren't very flat or are excessively burnished so no porosity for glue. Would be very difficult w/ walnut given it's open-pored characteristics but again...

I can think of very few glue joints that didn't hold ever over 30 years and virtually all of them could be attributed to either trying to make ill-fitting pieces stick or pushing the envelope of temperature (cold) on the glue.

--

[Kenneth]

Howdy,

Is there any possibility that, at some point, the glue had frozen?

All the best,

[J. Clarke]

Too much pressure will result in a glue-starved joint. I think many people go overboard on the clamping pressure with Titebond and the like.

[Phisherman]

Allow the glue to cure 2 days before working the piece, longer if the piece is large. Make sure the mating surfaces fit well before you apply glue.

[dadiOH]

Hard to say but it is obvious that either your glue is no good or the wood surfaces being glued were "dirty". I suspect the latter.

Was the "kindling pile" inside or outside? Did you machine or sand the wood surfaces to be glued so that you had nice fresh, dry, unoxidized, unweathered wood for the glue?

BTW, clamp pressure depends much on the hardness of the wood; butternut is soft, doesn't need as much pressure as walnut which is harder and walnut doesn't need as much as hickory which is very hard. Same thing for planarity of surfaces...one can get a satisfactory glue job with two relatively rough, non-matching butternut surfaces by increasing the clamp pressure...can't do that with hickory - or at least I can't.

[sweet sawdust]

"Tanus" wrote in message news:g1fnaj$mb7$ [e-mail]...

I think I see a problem here. Try this: Go to the lumber yard and get a "pine" 1 x what ever( ask for a cull it is cheaper and will work fine). Cut the board into 2 inch wide strips about 10 inches long. Get a NEW bottle of glue, I like Titebond ll but any white or yellow glue will work, Try to stay away from Titebond lll for this. Now take two strips and put glue on about 6 inches on the end of one of them and put them together so that you have a unglued surface on each end and the glue joint in the middle ( you should have a "handle" on each end of the glued up wood). Clamp and set aside to dry. You should do this with several samples, changing the amount of glue, clamping pressure etc with each sample, one with a lot of glue so that is runs out of the joint, one with very little glue, one with glue spread evenly on one surface, one with glue spread on both surfaces etc etc.. You should vary your clamping method, strong clamp, light clamp, tape for clamp etc. Allow to dry for a least 4 hours the longer the better. When cured take one of the "handles" and place in a vice and whack the other end with a hammer until the joint breaks, Look at the broken joint and you will see which method is strongest and best for you. All wood should act pretty much the same with what ever gluing method and glue you used, some better some worse depending on the glue and wood but similar. Now you have a gluing method to use. Next on the faces of the boards you glued up, you hand planned them, are you sure that you got them perfectly flat? I find it hard to get them perfectly flat with a hand plane, it seems to work better with a planner or jointer when doing face glue ups.

[lmlitwin]

I read some research that proved this to be an old wives tale. Joints can be glue starved, BUT not at the pressure generated by hand clamps. The study was scientifically done and reported in a woodworking mag.

[dpb]

That would have been Fine Woodworking (about 3-4 issues ago?).

Pretty incredible recommendations in it--if followed, we'd all need hydraulic presses to meet the "optimum" clamping pressures... :)

Fortunately, experience has shown that far less clamping pressure than those determined by the research reported therein is adequate for highly reliable joints.

The likelihood of glue starvation by clamping pressure would be quite low for OP's case I would think except for two possibilities that seem present in his case--possibly not using sufficient to start with and a small glue joint area means clamp pressure is concentrated and if he was concerned in not clamping enough, could possibly have really torqued down. Don't know what kind of clamps were used, either, of course...

The possibility posted elsewhere of the glue having been subject to freezing over the winter is a most excellent one I hadn't thought of...that would do it.

--

[mark]

Maybe the glue was old???

[J. Clarke]

For certain values of "scientifically". And hardly "old wives tale" when the glue manufacturers agree with years of serious research by the likes of Forest Products Laboratories that for each type of glue there is an optimal glue line thickness, the achievement of which depends on the type of wood, the grain orientation, the viscosity of the glue, and the clamping pressure.

As for "the pressure generated by hand clamps", how much pressure is that? That depends on the joint geometry. Don't confuse "force" with "pressure".

One test reported in the popular press is not definitive research.

[Chris Friesen]

Yep. And for hard maple and PVA the optimal glue line thickness resulted from very high pressures...substantially more than the 250psi that is at the upper end of Franklin's recommendations.

However, as "dpb" mentioned, optimal is not normally necessary, and quite adequate joints can be made at much lower pressure.

Absolutely true. Common misconception.

Chris

[Leon]

No, it will not, that is an old wives tale. The only way you get glue starvation is to not apply enough in the first place or scrape it all off during insertion. A mortise and tennon joint that fits too tightly can scrape all the glue off during assembly.

[Leon]

[EXT]

Did you do the "TV woodworker" type of glue up -- squirt a little glue on one side and press together -- for the best bond, glue should be spread evenly and lightly on both pieces and then clamped together. For working on the final result, wait 24 hours, even though it looks glued, it may not be set strong enough to withstand pressures and stresses.

[Woodie]

two things come to mind:

1 - bad glue - i'd consider this most likely - go to the borg and get a new bottle of titebond and try again.

2 - oily wood? Don't know what your kindling pile is, so is it possible your two pieces are not walnut, but a more tropical wood that has a high oil content?

[J. Clarke]

If one needs "very high pressures" to achieve 3-6 mil bond line thickness then one is applying too much glue. Note that the table of "recommended clamping pressures" in the Roman Rablej article in Fine Woodworking appears to have been pulled out of Rablej's ass. He gives no source for it at all.

Which is beside the point.

[Chris Friesen]

It may not be the case that a 3-6 mil thickness is always ideal. If so, then higher pressures may be better.

As "dpb" mentioned in this newsgroup when the FWW article came out, I'd guess the pressures came from Rabiej's paper:

"The Effect of Clamping Pressure and Orthotropic Wood Structure on the Strength of Glued Bonds", Wood and Fiber Science Vol. 24, No. 3, July 1992.

From the abstract at "

""...Using this concept, the optimum clamping pressure for sugar maple was found to be 0.3 times compression strength using U-F glue and 0.5 times using PVAc glue. This approach to determining reliable clamping pressure data can lead to improved gluing practice and more precise testing procedures."

Using the FPL tables, this gives a bit over 700psi for PVA glue.

Chris

[J. Clarke]

Do you have evidence that it is not?

How about just applying the right thickness to begin with instead of glomming it on and trying to squeeze it out with clamps?

"

">> "...Using this concept, the optimum clamping pressure for sugar > maple

But (a) how much difference does he see (I'm not really interesting in paying 25 bucks to find out) and (b) has anybody replicated his results?

Personally I'll take FPL's recommendations based on decades of experience backed by research over one unreplicated paper.