Official Luthiers Forum!

I remember reading in Overholst book on guitar making, two things that I try to always do.
One is to fit the joint perfectly, so the wood isn't stressed when clamped.
Also, over tightening clamps can create a glue starved joint.
I always take a cut-off scrap from a glue up, and break it.
Usually the wood breaks, not the glue joint.

I believe this is a myth. Inadequately wetting the surfaces of the joint with glue creates a glue starved joint. Another kind of "glue starvation" occurs when the joint is loose and there isn't enough glue to fill it (but that's not a good joint to begin with). Neither is caused by clamping tight; the latter can be remedied by it (less than ideally).

If the wood soaks up a lot of glue, I can't see how leaving it loose helps. It will create voids in the joint when the glue gets sucked up.

_________________
Howard Klepper
http://www.klepperguitars.com

When all else fails, clean the shop.

I've wondered about overtightening causing a glue starved joint as well (assuming a well fitting joint). If a thicker glue line weakens the joint, how can squeezing excess glue out of the joint make it weaker? The only way I can see it happening is if the wood absorbs the glue leaving no glue in the joint.......but can that happen?

_________________
Formerly known as Adaboy.......

Interesting, Howard.
Do you wet the joint when using a glue like titebond or white git-makers glue?
The glue starved joint never really made sense to me,
but heard about it from some very experienced woodworkers.

I just wandered around the web a bit and found some interesting info on gluing practices for wood aircraft - a place where the consequences of failure are a lot worse than having the bridge pop off a guitar!

Interesting reading:
http://www.faa.gov/regulations_policies/advisory_circulars/index.cfm/go/document.information/documentID/99861
See Ch 1. The .pdf is too large to upload here directly.

The issue of clamping pressure with epoxy is addressed there, briefly.
Surface prep recommendations are interesting in light of some guitar building practices- e.g. no sanding except for 220 grit to very slightly roughen the surface to allow glue penetration, if necessary. Smooth planed surfaces without planer marks, only.
And the doc also addresses the old question of whether to apply glue to both surfaces, which comes up here from time to time. The FAA stipulation is that glue must be applied to both surfaces.

Cheers
John

It's always 'made sense' to me, and I can't ever remember applying glue to only one surface, so I must have been taught that somewhere. Certainly if you just do a very quick test - apply glue to one surface, press together and then pop apart, you will often find less-than-complete wetting. Google will turn up lots of recommendations to wet both surfaces with glue.....and probably lots of the opposite as well!

John

From the FAA doc:

1-5. BONDING PRECAUTIONS. Satisfactory
bond joints in aircraft will develop the
full strength of wood under all conditions of
stress. To produce this result, the bonding operation
must be carefully controlled to obtain a
continuous thin and uniform film of solid adhesive
in the joint with adequate adhesion and
penetration to both surfaces of the wood.
..........

Wetting tests are useful as a means of
detecting the presence of wax, old adhesive,
and finish. A drop of water placed on a surface
that is difficult to wet and thus difficult to
bond will not spread or wet the wood rapidly
(in seconds or minutes). The surface may be
difficult to wet due to the presence of wax, exposure
of the surface to heat and pressure as in
the manufacture of hot press bonded plywood,
the presence of synthetic resins or wood extractives,
or simply chemical or physical
changes in the wood surface with time. Good
wettability is only an indication that a surface
can be bonded satisfactorily. After performing
wetting tests, allow adequate time for wood to
dry before bonding. Preliminary bonding tests
and tests for bond strength are the only positive
means of actually determining the bonding
characteristics of the adhesive and material
combinations.


1-7. APPLYING THE ADHESIVE. To
make a satisfactory bonded joint, spread the
adhesive in a thin, even layer on both surfaces
to be joined. It is recommended that a clean
brush be used and care taken to see that all surfaces
are covered. Spreading of adhesive on
only one of the two surfaces is not recommended.
Be sure to read and follow the adhesive
manufacturer’s application instructions.

Over the years I've figured out through my own trial and error that wetting both surfaces just gives better coverage. Sometimes when wetting only one surface, then clamping, the second surface will not be thoroughly wetted. Doesn't mean I always do it.

_________________
Steve Smith
"Music is what feelings sound like"

I haven't seen any proof of 'glue starvation' yet, and I've clamped things in a hydraulic press...

The closest I've come to figuring out how an otherwise competent woodworker could experience 'glue starvation' is the case where clamping pressure is applied non-uniformly and so the joint closes and then opens back up a bit leaving insufficient glue to fill the gap. It would be hard to spot it happening in a lot of cases, especially where multiple clamps are being used, and it would seem like the glue had 'disappeared'. If anything, this is less likely to happen if all the clamps are tightened fully.

There was a big long discussion about adhesion here:
http://www.luthiersforum.com/forum/viewtopic.php?f=10101&t=27187&hilit=surface+energy

That thread's got pretty much everything you'll ever need to know about sticking two things together between the content and the links. I wouldn't place much trust in any document about gluing that doesn't mention wetting (chemical wetting, like in the FAA doc above) or surface energy.

_________________
Bob Garrish
Former Canonized Purveyor of Fine CNC Luthier Services

Well, the final test is in and the results are......inconclusive.

As you can see, there was a huge fillet on one side and that held really well. The epoxy in the joint itself didn't do so well except for one edge. Could this be starvation?

I'm going to glue this joint today. Now that I've scraped and sanded and re-sanded etc, my plan is to check the joint's tightness. If it looks pretty good, I'll make a caul and use titebond. If it's too wavy, I'm going to go with the system 3.

_________________
http://www.birkonium.com CNC Products for Luthiers
http://banduramaker.blogspot.com

Andy-
It looks like you had good glue coverage throughout. My guess would be poor penetration. Some plywood I've used had an almost 'waxy' feel to the surface-perhaps a leftover from the pressing during manufacture? If you are gong to do more test pieces, perhaps a good scrub with solvent (acetone/lacquer thinner/alcohol) and abrasive (Scotchbrite or similar) might be one (more) thing to try.
If one gluing surface presents a lot of end grain you can apply epoxy and let it soak in for 5-10 min before gluing, if your shop is cool. Definitely move the epoxy joint to a warmer area if you are going to test in 24hr. WEST epoxy (and System3?) takes several weeks to reach max strength, curing at 'room temperature', as I recall.

Cheers
John

ok ok, im an electric guy and i dont deal with alot of the same gluing issues as acoustic builders,however i am also a woodworker with 25 years expierience. you have to think microscopicly,most polys when dry are smooth like any other plastic.its almost non-pourous so nothing to "bite" into the surface bieng glued.try painting a piece of plastic it doesnt stick well.titebond works well because it does just the opposite,it flows into and around the open grain of the wood to "bite' in. you can "key" in poly and epoxy too but its difficult to do.even with titebond i always 'key" or rub it into both sides im gluing. this is something i learned from an old german cabinet maker,"You have to key in both sides or the glue will fail" he always said. Ive seen him break the grain on solid red oak before his glue joint failed using yellow titebond. also poly is not very flexable, niether is epoxy. ever see a surfboard bent,nope they just snap in half. so side pressure on your joint with poly will probably do the same.
remember to moisten the wood slightly to open the grain,and key the glue in well on both sides.

p.s. what kind of glue did the oldtimers use? just a thought.

HHG is not used generally in instrument building because of strength it is used because of it's crystalline structure when dried and because it is to a form self clamping. The very hardness of the glue in the first place would make it inherently weaker in certain uses but the transference of vibration would be greater.

As for self clamping, HHG is used in the glass industry to create certain scaled effects in glass. All one has to do is brush the glass with the higher strength HHG and as the glue drys it cracks the glass off the surface. Pretty GD strong!

_________________
Billy Dean Thomas
Covina, CA

"Multi famam, conscientiam, pauci verentur."
(Many fear their reputation, few their conscience)

Agreed. I thought I would get some spirited replies to that comment, but no takers. In any case, I woudn't have a problem using Titebond, LMI white, or HHG. However, since I haven't tried it, I'm not too sure about the bottled liquid hide glue.

In my very brief time with a violin maker, he taught me to size an end grain glue joint by patiently brushing white glue into it, keep it wet with glue for 10 minutes or so, then it stops absorbing and I clamp. I have always put glue on both sides, it seems to me that a glue that will absorb into the wood and then crosslink will need a moment to absorb.

I used to make doors, I would always start the glue process with the rails to get the endgrain soaked, then I apply glue to the stiles, by then the rails need more glue, then clamp.

I tested PU glue with door panels years ago, always had a glue line failure, always broke the wood with TBII
Rob

_________________
http://shastaguitar.com/
http://www.kalimbakit.com/
http://www.youtube.com/user/comfyfootgr ... ature=mhee
http://www.facebook.com/robert.renick.7

I've starved HHG joints before, it's not very difficult. However, in our labs at Franklin, we have never starved a PVA joint (even with low MC wood, rapid assembly, and 460 PSI clamp pressure). HHG has outstanding strength, although this largely depends on its' gram strength, in the same range as a modern PVA and has superior resistance to creep. People do not give it enough credit!

I would have preferred a more standard test method such as ASTM D905 shears. If there is sufficient interest I have two MTS Instron machines at my disposal with a variety of load cells. In other words, I can test any adhesive you want under a variety of experimental conditions.

Hugh-
Thanks for the offer to inject some facts into the discussion!
Like most 'science guys' I'm all in favour of experiments as long as they don't challenge my existing beliefs too strongly!
I am interested in the 'coating both sides' question (especially in a scenario similar to sanding a brace on a 60/80 grit radius dish and gluing it to a top (or even better, rosewood back) which has been sanded to 120.

Also, the epoxy 'pre-coat with epoxy before adding the epoxy with filler' technique - especially when bonding 'absorbent' woods like dry spruce or cedar - could stand a further look.

And, as Filippo has suggested, anything that results in better HHG use would be helpful. The question of optimum viscosity for HHG isn't much talked about. I understand fiddle builders use a lot thinner variety than guitar and furniture folks.

You did want to stop guitar building and become the OLF glue test lab, right???

Cheers
John

I'd be happy to slice of a length of the plywood and send it to you for testing. I do have some of the Maple Compwood left but it's a little bent and might not be usable for the test.

_________________
http://www.birkonium.com CNC Products for Luthiers
http://banduramaker.blogspot.com

No need to submit samples, I have more sinister plans in mind. Engineers among us shall be entertained. Very little has been documented regarding adhesive properties beyond yield strength. The tests I've been running recently are painting a fascinating picture prior to joint failure.

I'm investing in a cnc router with a friend and plan to exploit the instrumentation at my disposal and apply science to better understand this art.

AH! A glue troll!

I think just about anybody that builds uses different adhesives depending on the application. Creep resistance, as well as low moisture, ease of use and just plain strength of the adhesive itself would allow different places in use to be justified.

_________________
Billy Dean Thomas
Covina, CA

"Multi famam, conscientiam, pauci verentur."
(Many fear their reputation, few their conscience)

Update - I decided to go with the system 3 because of some mild waviness in the joint. Work has been so busy that I didn't get to it until yesterday morning. I'm traveling the next few days so they'll be in clamps until Wed. at the earliest.

_________________
http://www.birkonium.com CNC Products for Luthiers
http://banduramaker.blogspot.com