The rights and wrongs of wood glue!

[Phil Pascoe]

I've just looked at an interior Resin W (albeit an old one) and that's a "one side only". Having once made something with a dry join (sometime way back in the mists of time ) I'd rather lag it on and wipe it off - unless of course there was a valid reason for keeping the glue off the surfaces. Certainly anything structural gets lagged.

 

[RogerP]

Many manufacturers actually specify only apply glue to one side only. They obviously have a reason, but I've asked numerous times before and never got a definitive answer from anyone. I cannot see any reason for this theory, personally.

Presumably to stop users applying too much glue?

 

[Jake]

Won't go back to PVA? Titebond is PVA.

Yebbut with some dye and tackifiers and marketing labelling which makes it sound like its not.

 

[Phil Pascoe]

Many manufacturers actually specify only apply glue to one side only. They obviously have a reason, but I've asked numerous times before and never got a definitive answer from anyone. I cannot see any reason for this theory, personally.

Presumably to stop users applying too much glue?

Why? Using too much is to their advantage.

 

[Beau]

Having two super smooth ... surfaces clamped together would squeeze out most of the glue, and only have a very small amount bonding the wood.

But that's actually a good thing. It's sometimes said that the goal is to have a single molecule of glue between the two surfaces, that's an exaggeration but it's certainly true that a glue line can be almost unbelievably thin and still amazingly strong. According to the technical sources (including one or two of the glue manufacturers) it's nearly the thinner the better, with many adhesives a glue line of 0.06mm is weaker than one of 0.02mm.

Whilst not exactly the most scientific of experiments, check out Mathias' test here https://youtu.be/14Mkc63EpMQ

The thinnest layer of glue joints were actually the weakest in practice

Never done any carful tests but often snap offcuts from butt jointed boards and the thin lines do fail easily. Dam shame as no one wants to see ugly glue lines running down tabletops so guess if you want nicely glued up tops plane them nice and smooth but put joints in to keep it strong.

 

[RogerP]

Many manufacturers actually specify only apply glue to one side only. They obviously have a reason, but I've asked numerous times before and never got a definitive answer from anyone. I cannot see any reason for this theory, personally.

Presumably to stop users applying too much glue?

Why? Using too much is to their advantage.

Not if it makes the product underperform.

 

[custard]

Never done any carful tests but often snap offcuts from butt jointed boards and the thin lines do fail easily. Dam shame as no one wants to see ugly glue lines running down tabletops so guess if you want nicely glued up tops plane them nice and smooth but put joints in to keep it strong.

My experience with PVA is the exact opposite.

Two carefully prepared boards (hand planed not machine finished), possibly with a small amount of "spring", thin coating of PVA on each surface to ensure complete wetting out, very heavy cramping, no biscuits or dowels or dominos; in my experience this all adds up to invisible glue lines and the strongest possible joint.

The devil's in the detail though, and hand planing the joint surface consistently, accurately and quickly really is a high skill challenge.

The workshop I trained at only takes fairly experienced woodworkers, and one of the early tasks was jointing up a wide surface. Their particular recipe was zero spring, but meticulous and extensive testing with a straight edge and a square of known accuracy. It was sobering to see furniture makers with quite a few years under their belts having to make multiple attempts before achieving this high standard, and by the way I was no better!

My take out is that woodworkers need to be honest with themselves, if you're genuinely good with hand tools then this is the route to the best possible joint. If you're not quite at this level but you've got excellent quality and well set up machinery with sharp tooling then you can 90% of the way there straight from the machine. However, if your skills and equipment aren't right out of the top drawer then the most reliable way to achieve "okay-ish" jointed up boards is with stopped plywood splines and/or a gap filling glue like Cascamite.

 

[Beau]

Never done any carful tests but often snap offcuts from butt jointed boards and the thin lines do fail easily. Dam shame as no one wants to see ugly glue lines running down tabletops so guess if you want nicely glued up tops plane them nice and smooth but put joints in to keep it strong.

My experience with PVA is the exact opposite.

Two carefully prepared boards (hand planed not machine finished), possibly with a small amount of "spring", thin coating of PVA on each surface to ensure complete wetting out, very heavy cramping, no biscuits or dowels or dominos; in my experience this all adds up to invisible glue lines and the strongest possible joint.

The devil's in the detail though, and hand planing the joint surface consistently, accurately and quickly really is a high skill challenge.

The workshop I trained at only takes fairly experienced woodworkers, and one of the early tasks was jointing up a wide surface. Their particular recipe was zero spring, but meticulous and extensive testing with a straight edge and a square of known accuracy. It was sobering to see furniture makers with quite a few years under their belts having to make multiple attempts before achieving this high standard, and by the way I was no better!

My take out is that woodworkers need to be honest with themselves, if you're genuinely good with hand tools then this is the route to the best possible joint. If you're not quite at this level but you've got excellent quality and well set up machinery with sharp tooling then you can 90% of the way there straight from the machine. However, if your skills and equipment aren't right out of the top drawer then the most reliable way to achieve "okay-ish" jointed up boards is with stopped plywood splines and/or a gap filling glue like Cascamite.

I think my skills are up to it but on consideration I do use PU glue a lot. Learnt with PVA and then fell into using PU but have not changed my approach to joining boards. Usually careful set up of the surfacer followed buy very light shaving to remove any ripple. Back to PVA these days so will do a few tests if I get a chance.

 

[Phil Pascoe]

Presumably to stop users applying too much glue?[/quote]
Why? Using too much is to their advantage. [/quote]Not if it makes the product underperform.[/quote]
How can the glue you've wasted make the glue you've actually used underperform?

 

[custard]

Personally I think the strongest PVA joint comes from a thin coating on each mating face, but there are times I'll use a thick coating on one face. With a big glue up the open time on PVA might not be long enough, a thick coating on one face stays open and useable much, much longer. The problem is that PVA is a great lubricant, so the components can slip badly out of alignment with this method, but sometimes it's still the best compromise.

 

[scrimper]

TBH I am always faced with the dilemma of how tight do I do up the clamps, on the one hand we are told that the adhesive is stronger than the wood so is squeezing most of it out not weakening the joint, on the other hand one does not want a glue line showing and anyway science says the wood is joined at the molecular level and the adhesive is there to smooth out the hills and troughs in the surfaces?

Being a scrimper I do clamp quite tightly and scrape off any 'clean' adhesive and use it on the next board I am making up. (If I have several to do).

 

[custard]

TBH I am always faced with the dilemma of how tight do I do up the clamps, on the one hand we are told that the adhesive is stronger than the wood so is squeezing most of it out not weakening the joint, on the other hand one does not want a glue line showing and anyway science says the wood is joined at the molecular level and the adhesive is there to smooth out the hills and troughs in the surfaces?

Titebond's own recommendation is a cramping pressure of 100-150 psi on softwood and 175-250 psi on hardwoods. That's fairly typical for PVA, read the "Application Guidelines" tab on this link,

http://www.titebond.com/product.aspx?id ... 3970f736af

So let's assume we're edge jointing two 1" thick hardwood boards that are four feet long. So that's 48 square inches, which at say 200lbs per square inch means we ideally need a cramping pressure of 9,600 lbs. A really good sash cramp, tightened hard, will deliver about 2,000 lbs. But, as Bruce Hadley points out, a bit of rust on the thread will reduce a sash cramp's maximum pressure by half. So in the real world hitting that optimum pressure target might need eight sash cramps, that's one every six inches!

The bottom line is that in any workshop outside of industry, with their hydraulic presses, it's practically impossible to apply too much pressure so there's no point worrying about it!

From a practical perspective we almost always end up with PVA glue joints that are below their theoretical maximum strength because of our cramping limitations. Applying veneers in a vacuum bag means we'll never get more than 15 lbs per square inch pressure, or when gluing up a face frame for example too many heavy sash cramps can distort the frame.

The way I resolve this dilemma is to recognise I'll only ever get "good enough" PVA glue strength rather than the theoretical maximum, but I try to recognise that and compensate as far as I can by ensuring every other part of the process is done to the best of my abilities. So I aim for flawless jointing surfaces, and I try to ensure that I've prepared those surfaces at most a few hours before the glue up so they haven't moved or oxidised or become soiled before the assembly.

One final point, PVA glue has virtually no gap filling capability, so we must be careful not to confuse "a joint that's stronger than the wood" with an "adhesive that's stronger than the wood". A gappy PVA joint is always a weak joint.

 

[ED65]

Whilst not exactly the most scientific of experiments, check out Mathias' test here https://youtu.be/14Mkc63EpMQ

The thinnest layer of glue joints were actually the weakest in practice

Something wrong with his process then! I have seen the video before and I don't want to revisit it now to list what I can spot that might be at fault but without doubt thin glue lines should be strongest using most adhesives.

If you're sceptical yourself you can confirm it with just one test comparison, the difference should be starkly evident. But this question has has exhaustively investigated since before we were all born* because industry needed to know. And a thin glue line is much more resistant to shear stresses and hence stronger. This is with most glues, deffo includes foaming polyurethanes and PVAs of all types so that covers most of what we use. Superglue as well. Epoxy is an exception for anyone who uses that sometimes.

When bonding wood this does assume very smooth joint surfaces (but not burnished/glazed) and heavy clamp pressure. It's the clamp pressure that leads to the thin glue line, not scant glue being applied.

*Earliest thing I recall reading where it was definitively stated that you should aim for the thinnest of glue lines for strength was published before WWI.

 

[ED65]

TBH I am always faced with the dilemma of how tight do I do up the clamps, on the one hand we are told that the adhesive is stronger than the wood so is squeezing most of it out not weakening the joint

Don't be afraid to crank up the clamp pressure to "I'm scared of damaging the wood." if the goal is the strongest joint possible.

Any fear of starved joints should not be from clamping, as custard says above it's practically impossible to overclamp. Starved joints are probably mostly caused by not applying enough adhesive in the first place. It can also happen due to taking too long before bringing the joint halves together, so the glue skins over and then can't fully wet the mating surface.

BTW it's not actually true that the adhesive is stronger than the wood, it's the adhesive-bound wood that's stronger than the surrounding wood (or at least equally strong). The glue itself IS the weak point if the glue line is too thick. Some numbers here might be useful so 0.1mm or four thousandths is a bit on the thick side for a glue line!

 

[EddyCurrent]

Won't go back to PVA? Titebond is PVA.

Sorry, I didn't explain that the glue I used previously was white PVA in the green bottles that's been about for years. Titebond is a different consistency and in my view far better.

 

[Phil Pascoe]

I suspect the white pva in the green bottles has been superceded by many brands - to me, it's always been decidedly average.

 

[Mike Jordan]

I was taught that the thinnest glue line is the strongest, the edges must be hand planed to remove any machine marks and make a perfect fit. Finally the timber should be glued together straight after planing to prevent the surfaces oxidising in the air for best results.