I am starting a new thread so as to not get lost in the discussion below. What follows is what the Franklin tech rep had to say on this topic.
To get an optimum bond both surfaces must be wet by the glue and excess glue squeezed out of the joint. Either approach for applying glue can fail so there is no "right" way. And, either approach can succeed. Do what you find reliable.
If glue is applied to only one side there are two ways to fail. When the joint is mated insufficient glue is transferred to the mating surface to wet it. Or, the glue begins to dry before mating and insufficient wetting occurs on the mating surface.
If glue is applied to both surfaces and the surface is allowed to dry some before mating, insufficient glue could be squeezed out during clamping or the two partially dried surfaces may not properly wet and bond.
The bottom line is to understand the requirements and do what is needed to achieve optimum bonding. The idea of testing bonds by fracturing cut-offs is a good practice to follow.
Bill, from the information you have posted, it would appear to me that the major cause of glue failure is allowing the glue to "skin" or partially dry before clamping. If that is the case, I (although I do not always follow this method) that applying glue to one side of the joint only would hasten assembly and be the best method.
However, I am fortunate that I have a temperature controlled shop so I have more time prior to the glue drying than some who work in a high temperature dry environment. I always dry fit any assembly prior to glue-up, have my clamps set up etc., before applying glue. If I have an extremely large area to apply glue prior to assembly, I usually wait until my son can help me to prevent premature drying.
I try to toss my glue when it has been stored a year or a year and a half. I am sure 99% of my glue is discarded while it is still quite usable. However, the cost of glue compared to my time and other materials is so low I sure don't want to take chances. One of my pet peeves is the fact the manufactures do not put an expiration date or "best when used by date" on the containers.
I have assembled and broken test samples for every glue I use, but unless working with an exotic or oily wood I have not tested with the glue I am using I do not retest, do you?
When doing bent lams, I apply glue with a 4" foam roller from the paint department of any hardware store. Puts on a nice even coat of glue very quickly, so it's easy to do both sides.
Most often I use plastic resin glue, but I've used Tightbond III before.
Most of the bent lams I've done were for customers rather than myself, so there was no point to risking a callback for a failed lamination. While it may be possible to glue a single side and have no failure over time, I just don't see the point to risking it.
No question that glue skinning over is a horrible situation. A clamping dry run using a brush to simulate glue application can help suggest when skinning over might occur. Sometimes there's no other way than to find ways to break the glue-up into smaller sections or get good help.
I suppose we are all talking about laminated bends from thin strips. Sure there is a tradeoff in time, for the window of opportunity to get it all spread before getting the clamps on. It takes more time to do both sides of course, but isn't it a common practice to press and squirm the two parts together, when only spreading on one side? Do you take that extra step when only spreading glue only on one side?
After cutting off the waste to the finished size, do you ever see a dry line that your stain gets into like an open pore? If you do, this may be like the tip of an iceberg.
And another question, if someone using a glue which starts to tack. Is the glue line compromised if the glue has started to gel and set, then broken while the bending and clamping proceeds?
It's my observation that conventional wisdom advising abundant squeezeout and high clamping pressure is the correct glueing strategy parallels the myth of the sprung joint. My experience belies that. The "optimal" glued joint is the one that does not fail under intended use, and the key to achieving that is a gap free dry fit mating. If the dry fit is tight to begin with, very little glue and only moderate pressure is needed to make a permanent stable joint.
Except for situations like bent lams, which I don't do, I apply sparse glue to one side of the joint, and expect that moderate clamping pressure will both wet the dry side and spread the glue sufficiently.
IMO, a lot of squeezeout, especially in edge-to-edge joints, suggests that it takes too much clamping pressure than otherwise needed to get rid of the excess glue and close the joint.
"A clamping dry run using a brush to simulate glue application can help suggest when skinning over might occur."
I always do a clamping dry run when starting a glue up process. And I am often surprised. Oooh, 'I forgot the cauls' or "where's my wet rag?" - things that would be a big problem to discover after I started applying glue!
During a class, the prof asked "How much glue should you have in your joints?" Being naive, I replied "Enough to get a good squeeze out line."
The prof replied, "Yeah, thats what we told you in the intro class, but by now you should be applying enough to just get a hint of squeeze out all down the joint line."
Keeping it in the frig prolongs its life, maybe for years in the case of the Franklin product.
The Franklin glue is fundamentally different from the more traditional formulations that rely on urea for liquefying. Urea degrades to ammonia and ammonia attacks the polymer lowering its molecular weight and rendering it brittle and weak.
There are two tests for liquid hide glue. In one case rub some between fingers and as it dries pull fingers apart. If strands form between fingers it means the polymer has not degraded. Franklin recommends drying a film and testing to see if it is brittle or tough. And, one can always make a test edge glue joint and bash it apart and look for glue failure.
Re: Franklin tech staff agrees with this statement
Clint Searl
I'm sure there's a point at which greater pressure does nothing to improve the characteristics of the glue's performance. I must have found that point because none of my joints have failed.
The other thing is that the efficiency of the clamping is not all that great, and on even 1" boards I doubt the PSI is that high.
The other thing is that if you have too much, or somewhat glue in the joint. When you apply pressure it squeezes out. The right way in most cases is mild pressure so as not to skid, and once the wood to wood happens, take it up a notch. If done incorrectly, the initial pressure can drop off quite a bit and once it settles in be less. Some clamping is self-adjusting, other methods are not. Stuff like elastic bands, air hoses, some rub, weight, vac, etc... will self-adjust.
A typical clamp (Taylor or Hargrave) with an acme thread like is used in furniture factories and other shops can apply 3500 pounds of clamp pressure.
Franklin recommends over 100 psi on the joint. In a 28"long panel and 3/4" lumber that equals 2100 pounds of clamp pressure for the joint, minimum.
The point to make here is that so long as the panel or what ever you are gluing doesn't buckle, apply as mush clamp pressure as you can conveniently apply. More doesn't hurt and too little will.
The clamps apply that 3,000# at the pad, whereas at the joint surface the pressure is, roughly, the total clamping pressure divided by the joint surface area. The joint would see a few hundred pounds by my rough calculation.
On the other hand, I've never seen a properly made joint that a few pipe clamps would not clamp quite sufficiently, so I would say that Bill's info from Franklin is probably spot on - 100 psi will do it just fine
From what I've seen, if we get to the point of crushing the wood under the clamp faces, we're trying to clamp a poorly made joint. Even if the clamp will close it, I betcha it's going to fail one day.
Sure, a properly made, perfect joint will need little glue and little pressure. The idea behind excess glue (squeeze out) and sprung joints is that not all joints are going to be perfect, and not all craftsmen are going to be ever vigilant with inspecting dry fit.
These are simply short cuts if you will. Practices that compensate for inexperience and expediency in working. The idea is to have consistency in result regardless of inconsistency in experience and rate of work flow - both assuming we are discussing a business shop. Applied to the hobbyist, there is little need for either practice, since the hobbyist has the time to learn to make a joint as ideally as possible before gluing it up.
Even in commercial work, both practices (and other similar ones) should be taken with a grain of salt. The idea is that if the work has a chance of being one side of the ideal, you plan to put it on the correct (non-failing) side of that ideal. It is a question of which variable is most easily controlled to reach the desired outcome - the consistent result of non-failing joints.
As a beginner I used slathers of glue and as many clamps and as much force as I could muster. Now I make better joints, use just enough glue to wet both surfaces as they mate, and lighter weight less robust clamps.
You said, "A typical clamp (Taylor or Hargrave) with an acme thread like is used in furniture factories and other shops can apply 3500 pounds of clamp pressure." And I'm not doubting your accuracy, but seem to remember that Jourgensen Pony clamps were good for about 800#, but that comes from deep into the cobweb recesses of my "inquedible bwain" as Elmer Fudd would phrase it. These also have acme threads, with a 4.5" handle. If my memory is anywhere near right, and the thread pitch is the same, it seems like the one you mentioned would need a handle about 18" long to achieve that much without a cheater bar to compensate for the individual user's hand strength.
Ive searched to see if I could find a comparison chart for various clamps, but haven't succeeded. If anybody knows of one, I'd like to see it.
The limiting factor on pipe clamps is the pipe size/weight and to a lesser extent, length. Clamps using schedule 40 3/4 get to something like 2800 lbs, and using heavier schedule 80 pipe would get you past 3000 lbs. Iron bar clamps would probably go over 4000 lbs, depending on bar size.
I would bet the 800lb pressure you remember is at the joint, that would be about the maximum for a moderate length edge joint in 3/4" material with lots of pipe clamps on it - like about twice as many as one actually needs Of course, there are a huge number of assumptions made - do it just right and you can get the pressure up over that which the clamp produces (less surface area than the clamp face itself). But 800 would be a realistic maximum for an over the top clamping of a typical joint.
Those Dubuque aluminum bar clamps develop much less pressure than even a pipe clamp, but by all reports they work just fine too. Anything much over 1000 lb at the clamp face is probably overkill for a good joint.
Just use your normal practice on some scrap of spreading and clamping. Then after maybe 2 - 5 minutes in the clamp, take the clamps off. The glue should be squeezed as far as it will ever get, but not cured so far that a chisel can lever them apart.
If there are voids, you'll easily be able to tell whether your method is acceptable for your expectations or not.