PDA

View Full Version : Glue Test: And Results



lesharo
12-15-2007, 03:55 PM
Fine Woodworking just had a glue test. Pretty amazing results ( at least to me). They tested plain PVA (Elmer's I think), Titebond III (apparently a waterproof PVA), System Three Epoxy, and Gorilla glue. They used a motice and tenon joint and different woods and different tightnesses of the joints.

The results? The worst: Gorilla Glue. The best? The PVA's. The epoxy? Pretty lousy; just above the Gorilla.

Pretty amazing! We bought the Gorilla glue just because they told us how great it was. Epoxy; apparently the same deal. And I bought the hype years ago that if you had to take apart wood glued with epoxy; you'd break the wood before you broke the glue joint. Not so in the Fine Woodworking test. It looks like reality has reared its ugly head.

Check the test out, and I'll go back and read it again myself.

GregH
12-15-2007, 04:03 PM
Was there any explanation why they chose to use a mortice & tenon joint? I ask because that joint has some mechancal properties of it's own, and they will vary depending on the fit and proportions of the joint relative to the size of the parent material.

Paul Girouard
12-15-2007, 04:09 PM
They used a bridle joint ,

http://www.taunton.com/finewoodworking/Materials/MaterialsPDF.aspx?id=28897

Funny that link has a Gorrella glue ad on it :D

What was really funny was there hasn't been any rebuttal from Gorrella glue , at least that I've seen.

I guess that ad space rotates when I check the link to be sure it worked it had a different ad.

Cuyahoga Chuck
12-15-2007, 04:20 PM
Fine Woodworking just had a glue test. Pretty amazing results ( at least to me). They tested plain PVA (Elmer's I think), Titebond III (apparently a waterproof PVA), System Three Epoxy, and Gorilla glue. They used a motice and tenon joint and different woods and different tightnesses of the joints.

The results? The worst: Gorilla Glue. The best? The PVA's. The epoxy? Pretty lousy; just above the Gorilla.

Pretty amazing! We bought the Gorilla glue just because they told us how great it was. Epoxy; apparently the same deal. And I bought the hype years ago that if you had to take apart wood glued with epoxy; you'd break the wood before you broke the glue joint. Not so in the Fine Woodworking test. It looks like reality has reared its ugly head.

Check the test out, and I'll go back and read it again myself.

Expoy's reputation rests on is abilty to funtion in woodwork where there is little, or even no, joinery.
It may be different in big boats but in the small stuff I build the only instance of a mortice and tenon joint showing up was in a skin-on-frame kayak. And, that was because the joint was lashed not glued.
The more I consider this the more I'm convinced that testing glue on a mortice and tenon is no real way to evaluate the all-around capabilities of it.
When I'm on the water in the little boat in my avatar there is nothing between me and a chilly swim but a bunch of plywood held together with epoxy resin. And, unless the epoxy undergoes some molecular degeneration I think I'll be endangering myself in this plywood wonder for years to come.

pipefitter
12-15-2007, 04:43 PM
Most woodworkers and boat builders know the limitations of their adhesives. I rarely see any craftsmen rely on glue alone (not talking about fastener free boats) and with well designed joints engineered properly for most builds, the glues such as those that were tested, become somewhat overkill. In a situation where the joints shown in the test would be used, it would most likely not see unlimited stress tests like that and if it was subject to motion fatigue, would surely be backed with a gusset or bracing system to compensate. In most cases and as composite constructions (with/out fasteners) are designed these days, epoxy is indeed overkill. The collective strength of epoxy throughout a build of a boat with it's many joints becomes rather moot. I know there is arguable exceptions to these rules but for the most part, most glues match or exceed their claims. It's when we start to tweak the usage parameters or limit them out is when we have to be responsible enough to engineer in some compensation allowances to not make the adhesive completely liable.

Bruce Hooke
12-15-2007, 04:44 PM
I've never been a big fan on Gorilla Glue. I just did not see any big advantages to it over what is already available.

On epoxy, what I've always understood the big selling points of epoxy are:

1. It has a proven track record of effectively being waterproof if used properly. The only other standard glue I trust to be waterproof is Resorcinol and it has a lot of other down sides.

2. If used properly epoxy does very well at filling large gaps. Note that this often requires adding some filler, which I don't think they did in the tests.

I don't see much point in using epoxy for indoor furniture in most situations, but for boat work I think its track record is pretty well proven.

What I found interesting about the article was how well hide glue did. For furniture work hide glue does have some significant advantages when it comes to finishing and repairability.

Overall, in interpreting the results of the tests, I would take the approach of using strength as a cut-off, not as an absolute measure. By that I mean, if a glue does not appear to be significantly weaker than other glues in a particular situation then there is no reason not to use it in place of some other glue that is in theory 10% stronger. When I've seen glue joints fail it has been because of something going wrong other than simply the glue not being strong enough. To put it another way, don't pick the glue that is strongest, pick the glue that has the right qualities and is strong enough.

Bob Smalser
12-15-2007, 05:38 PM
I don't know of a glue that's not stronger than the wood it glues in typical applications. Strength beyond that seems totally meaningless to me.

Other factors like open time, moisture content limits, clamping pressure required, performance after water saturation, and its compatibility with other glues in repairs are much more important and vary so much that there probably will never be single glue correct for every application.

Moreover, how long the glue survives in routine crossgrain applications like M/T or bridle joints subject to seasonal movement has nothing to do with its strength, and everything to do with its flexibility.

I haven't read this article so I'll stop, but so far I've heard nothing that convinces me I should pay to do so.

JimConlin
12-15-2007, 07:28 PM
I am also vitally interested in the susceptibility of a glue to creep. That is, if subjected to a sustained load, will it move? For bent laminations, this is critical. The white and yellow glues have not been good in this.

Scott Sawtelle
12-15-2007, 08:25 PM
Wasn't this a couple of months ago? Alot of people wrote in to complain that they didn't use any fillers in the epoxy to make it a suitable glue. FWW answered that the instructions on the epoxy didn't require it, so they didn't use it. They just used pure epoxy.

Tom Lathrop
12-15-2007, 10:27 PM
Wasn't this a couple of months ago? Alot of people wrote in to complain that they didn't use any fillers in the epoxy to make it a suitable glue. FWW answered that the instructions on the epoxy didn't require it, so they didn't use it. They just used pure epoxy.

Duh! When you buy a hammer, does the instructions describe which part to hit a nail with? Ignorance wins again:( FWW started out as a journal for experienced woodworkers. It has evolved into a magazine for beginners and a shill for products through testing articles and promotional blurbs. After 30 years I have let my subscription lapse.

The reason hardware stores sell epoxies in specialty formulations for specific purposes is that the average buyer/user can't be expected to know the proper way to use the raw stuff. Maybe that is where they should have bought their glue. System three has apparently bought into this way of marketing with their purpose formulated products. Most experienced builders can't be bothered with a shelf full of expensive specialty products and make up their own from a few basic resins and fillers.

I suspect that the editors of FWW are now business grads and that few if any grew up through working in actual shops.

john welsford
12-16-2007, 03:48 AM
Richard Jagels, Prof of Forest Biology and very long time writer for WoodenBoat mag featured a comparitive glue test in the magazine a few years ago, the results of my own testing as a techy advisor to companies making structural laminated beams was almost identical.

I dont use Gorilla glue or any other polyurethane anywhere that I dont have a large glue surface relative to the load on the joint.

JohnW



Fine Woodworking just had a glue test. Pretty amazing results ( at least to me). They tested plain PVA (Elmer's I think), Titebond III (apparently a waterproof PVA), System Three Epoxy, and Gorilla glue. They used a motice and tenon joint and different woods and different tightnesses of the joints.

The results? The worst: Gorilla Glue. The best? The PVA's. The epoxy? Pretty lousy; just above the Gorilla.

Pretty amazing! We bought the Gorilla glue just because they told us how great it was. Epoxy; apparently the same deal. And I bought the hype years ago that if you had to take apart wood glued with epoxy; you'd break the wood before you broke the glue joint. Not so in the Fine Woodworking test. It looks like reality has reared its ugly head.

Check the test out, and I'll go back and read it again myself.

ssor
12-16-2007, 01:38 PM
All of this is why we resort to large safety factors. If the ideal adhesive in the best controlled conditions gave a result of x pounds per square inch we would design a joint of at least 5x per unit load. One 16d nail driven perpendicular to the grain will carry my weight but I won't bet my life on it holding. The same is true of glue joints conditions can conspire against us and foil our best efforts.

kc8pql
12-16-2007, 01:50 PM
Wasn't this a couple of months ago? Alot of people wrote in to complain that they didn't use any fillers in the epoxy to make it a suitable glue. FWW answered that the instructions on the epoxy didn't require it, so they didn't use it. They just used pure epoxy.

Both System Three T-88 and Gel-Magic adhesives do NOT require fillers. There are ready to use. I have no idea however if either of those are the System Three epoxy FWW used.

Bob Smalser
12-16-2007, 04:41 PM
... We bought the Gorilla glue just because they told us how great it was. Epoxy; apparently the same deal. And I bought the hype years ago that if you had to take apart wood glued with epoxy; you'd break the wood before you broke the glue joint. Not so in the Fine Woodworking test.

Dunno about that FWW author, but all those glues sure break wood as well as the resorcinol standard for me. The polys used were Elmer's Probond and PL Premium Poly Construction Adhesive and the epoxy West System.

http://pic20.picturetrail.com/VOL12/1104763/7738131/102368835.jpg

http://pic20.picturetrail.com/VOL12/1104763/7738131/218588986.jpg

http://pic20.picturetrail.com/VOL12/1104763/7738131/103956073.jpg

http://pic20.picturetrail.com/VOL12/1104763/7738131/102357806.jpg

http://pic20.picturetrail.com/VOL12/1104763/7738131/102357565.jpg

Joel Herzel
12-22-2007, 11:03 PM
I'm a little confused. I read the article analyzing the various glues by using destructive testing. Pretty straightforward. Epoxy came in right behind the strongest Titebond, a point or two less if I remember.

Gorilla glue scored WAY behind. That is pretty much what my experience has been also.

That said, each glue has its appropriate use. Gorilla snot is good to use on wet softwoods in house construction. I have seen pieces of wood where it traveled by capillary and expansive action a good six inches up the stick in the capillary tubes. No other glue I am aware of will do anything like that. It IS a royal pain because it goes off so quickly upon contact with moisture and the overflow will really screw up a clear finish. Very difficult to clean up.

I have used gallons of urea formeldahyde (Weldwood) and it can be a good glue also, but with the major caveat that it is absolutely necessary that the temperature of the wood, the glue, the clamps and the air all be a minimum of 70 degrees fahrenheit. It says so right on the label and they are not kidding. The only failure I've had was when I was a little cavalier about that and indeed, it did fail. That high of a temperature is a very tough requirement for boatbuilding in many places. I have done a lot of pronounced bent laminations with it in white oak and it has held fine.

Epoxy gives you waterproofness, long assembly time if you need it, ability to work in pretty darn cold conditions, choice of thickeners and great longevity. I know some will argue that it doesn't work with some woods, but I have yet to have a bonafide failure with it. I make a lot of entry doors and windows that have extreme exposure to weather and temperature cycles and have yet to be disappointed by the stuff. I would never use it in a traditional "wet" boat situation such as carvel plank, but for dry wood, plywood, laminated bends, etc., I've got no gripes.

BTW, the joints tested were not mortise and tenon, but as was mentioned, bridle joints. That doesn't really matter to me anyway because they went to great lengths to make the conditions identical for each joint and each glue. And then to boot, the measuring of the pressure used to induce failure was measured by instrumentation.

In my work I use all the glues tested, but I don't believe that any one glue is a magic bullet that works in all circumstances.

Joel

Wild Wassa
12-23-2007, 12:39 AM
4 glues make a glue test? ... ... orrrhh right, ... 4 glues? ... that should be a definitive test with 4 bottem end glues?

Soak the timbers did they, just like how a boat gets soaked?

... and PVA won ... orrrh wow! Whiter than the average white?

Make all boats out of PVA then. (fair d'inkum! get real!)

Warren.

PS, How did Faber's Polyurethane glues go again, or BCP's epoxy ... but no, they tested Gorilla glue, PVA and Gorilla glue? When I was a kid I made glue out of potatoes ... how did the potato glues fair?

How did AV180 fair or any of the acid etch glues fair? ... what they didn't test them? 4 bottom end glues only make a bottom end glue test? The WBF is a boat thingy thing isn't it?

Boatmik
12-28-2007, 07:11 AM
I don't think I've ever used a mortice and tenon join on a boat.

Actually - the only place is for mast tenons - and generally I choose not to glue them - as I sometimes want to get my mast out of the boat!

:-)

The tenon join is a classical furniture join - it's purpose is to be stiff but not necessarily very strong at all - there is not much surface area for gluing and a large proportion of the total glue surface has at least one endgrain face.

Furniture joins are weak - weak - weak. I have enough experience pulling apart chairs, tables, cupboards to know that - I've seen enough of the same joins fail over a few years of use under the modest loads involved.

But while you can tear most furniture apart with your bare hands it is very difficult to pull any parts off a boat without significant leverage - a crowbar or something big heavy and blunt.

The empirical information that most boats don't fall apart despite considerable loads indicates that both joint design and bonding materials and methods (or fastening schedules in trad construction for that matter) are a world apart from such idealised testing as per the magazine article.

It can tell you what is bad - maybe - if the methodology is appropriate - BUT - I don't think the tenon joint (or dovetail joint) is the best methodology to generalise from for boatbuilding.

Cheers all!

MIK

Raka025
12-28-2007, 09:03 AM
Soak the timbers did they, just like how a boat gets soaked?



It was a Fine Woodworking Magazine, not WoodenBoat?