Official Luthiers Forum!

I'm not an expert in this... but from what I've read you need to factor in the glue type before discussing the surface preperation. It was explained to me that White glue/PVA's need a glass like surface and therefore planing is the optimal strategy. Think of it like wetting two pieces of glass and laying them face down on top of each other. The glue only needs to penetrate a small amount... and of course dry... to create an inseperable bond. The greater the surface tension the stronger the joint. You should not use PVA's to fill anything. If you have any sort of a "rough" surface, the surface tension is greatly reduced and weakens the overall joint. Sanding a joint and using PVA's/white glues will weaken the overall joint. However, if you plan to use epoxy or polyurethane glues, a toothed or sanded surface is appropriate. These glues actually "grab" the surface... so the more surface the better, ala sanding.

Long and short is that planing is best for PVA's with a glass like surface and a good tight fit is required to create a strong joint.

Yes epoxy likes a larger glue line but with poly a large glue line will make for a weak joint. Wood crumbs and torn fibers are not strong. Besides large glue lines are not what most folks want in a guitar top or back from not only a structural point but a aesthetic one as well.
Here is a hand planed, LMI glue, glued joint under magnification on a Adi top. The joint is just under the D on the penny. You can see that you can't see it (the joint) better with a bigger file size but I couldn't do it on the forum.
L.

_________________
Cut to size.....Beat to fit.....Paint to match.

I think starved joints from excess clamping pressure are another myth that needs to die. It's the standard strawman for faulty technique and bad joints. There might be situations where such a thing could occur, but none of them happen while building a guitar. IE: a perfectly fitted joint, with enough glue applied to wet both surfaces and squeeze out, which is clamped evenly and in one direction...which then fails

I have used a hydraulic press from time to time to clamp things up, at psi much higher than anyone's getting with manual clamps...no failures. I thought of doing a 'positive' experiment...but that just proves that I haven't successfully starved a joint. I think the real 'proof' experiment will have to be something involving a cash wager where I make the joints and the glue, then set up the clamps, and my rival needs to apply enough pressure to make my joint fail/weaken without actually splitting the material.

_________________
Bob Garrish
Former Canonized Purveyor of Fine CNC Luthier Services

Bob,
I am not following you totally here, are we agreeing or disagreeing ?
Are you saying that by a "straw man" you mean people will blame over clamping for a poor joint when really it is poor technique. I can see that especially sense in a poor joint there is likely to be a too much glue left in there.
The poorly fitted joint will need more clamping pressure to bring things in proximity but this is not a good situation, obviously.
The better the joint the less clamping pressure you need to bring things together. Of course this varies with different woods densities etc.
The important thing for begging and intermediate woodworkers is not to crush the wood with the clamps but instead make proper fitting joinery.
Heavy clamping pressure beyond what is needed can be troublesome with regards to distortion of clamps, the wood, the work and so on.
If your main point is that it is hard to starve a joint by too much clamping pressure I think it depends on the type of joint and the wood. A thin edge joint in a dense wood is much different than a large surface area joint in a softer wood. Cauls play a big part as well in distributing pressure and not causing what Hoadly calls the sponge effect. Which is push in the middle of a sponge and edges will lift.
So if you are saying that generally it is pretty hard to clamp the glue out of a joint, I would for the most part agree, especially with a less than perfect joint.
L.

_________________
Cut to size.....Beat to fit.....Paint to match.

I've starved one or two joints in my time, but never with Titebond. It's not exactly a 'straw man', but it's certainly less common than joints with too much glue in them, IMO.

I agree. I tried it. Got close, but still had to sand on a granite slab. That is why I bought that set of planes.

Mike

-

Yes on straw man. The thing you meantioned from Hoadley is very important. That's what I was alluding to in my description of clamping technique.

On agreement, I think it's a really agreeing shade of gray...I'd say we're definitely 97% agreeing (ie: I agree with everything you've said wrt clamping, joint prep, etc etc). Whatever percentage of 'mostly agree' is 'agree' will decide what the other 3% is . My opinion is that it wouldn't be possible to squeeze all the glue out of a wet, perfectly cut, perfectly clamped up joint with the amount of pressure any clamping system luthiers are using can provide. I'll get around to trying to do it in a hydraulic press, just to see what the actual pressure required is (and if it's less than the pressure the wood yields at!), but I do have some bill-paying to do before my wacky experiments! I suspect it might be possible with a couple perfectly prepared pieces of ebony.

There are a couple exceptions, and this also addresses 'sizing'. My statements are all in reference to a 'wet' joint (ie: both pieces with fresh adhesive on their surface). Sizing is done in cases where the glue might actually soak-up and dry on one piece before the pieces are completely clamped; while that WILL cause a joint failure, it isn't a result of clamping pressure.

_________________
Bob Garrish
Former Canonized Purveyor of Fine CNC Luthier Services

I'm pretty sure I do not in any way understand the concept of 'sizing' a joint...at least when one is using wood glue...

you see, wood glue does not work by being an adhesive...it works because it soaks into both pieces of wood being joined and then hardening...its the hard tendrils residing in both pieces of wood which hold it together...

I saw a test on TV were a guy used simple Elmers yellow glue to join to simple blocks of wood...he then tried to make the joint fail..in the end the wood around the joint failed...

it is possible to not use enough glue on certain woods, notably red oak..I typically join that wood, pull it apart and apply more glue, then do the final joining/clamping...

basically I am in agreement with the thought that most bad joints are just that...bad joints which wood glue is not going to 'fix'...

Yes I read it. And I pretty much disagree with everything you have said. I don't think there is any value in toothing a joint. Who uses toothing irons? Tell me how a toothing iron would be used ? I have never seen or heard of anyone using a toothing iron for a glue joint. How would planing a bunch of ridges in the edge of a board "allow deep penetration of the glue into the pores". Answer: it won't. Read Hoadly. Read the U.S. Forest Products lab stuff. Talk to some wood technologists. All you are doing is reducing the surface area of the joint.
For resinous woods a freshly planed edge properly glued is the way to go.
L.

_________________
Cut to size.....Beat to fit.....Paint to match.

-

Yup, Martin used to tooth the bottoms of their bridges. They also had to stop using the 'bar' bridge and go to the 'belly' shape to keep them from pulling off too often when they switched over to steel strings. Mario Proulx claims that 'bar' bridges work fine with steel strings, but he doesn't tooth them.

I tooth things I might want to remove later, such as the underside of a fiddle fingerboard. The thick glue areas and reduced effective surface do weaken the bond.

The likelihood of starving a joint is a matter of how much glue you can squeeze out and how much is left in the joint afterward. Most of the problems I've run into along those lines have to do with hot hide glue joints, where either the glue was too thin, or I warmed the joint too much, or rubbed it too long, or clamped it too hard. Usually it takes a couple of things: too much heat _and_ too many clamps, for example. One of the most embarrassing was a rubbed joint on a fiddle top: everything was perfect, but I warmed it too much (the shop was cold) and got carried away with rubbing, since the top was so warm the glue didn't gel quickly. It still took it several months to fail.

I have to wonder if the nature of Titebond, and other similar glues, doesn't reduce the likelihood of starving a joint. The glue itself is in the form of droplets of rather thick glue, suspended in water. You can squeeze out the water, and alot fo the glue, but at some point those little droplets are going to be jammed together enough to grab, and then you can't get the pieces too much closer.

-

Tony Francis asked:
"Why is the joint weakened by too much heat?"

Because it kept the glue too fluid for too long, allowing me to rub to much of it out of the joint.