Gluing End Grain is a No No But it Isn't

[Chris Parks]

A very interesting video


https://www.youtube.com/watch?v=m7HxBa9WVis&ab_channel=PatrickSullivan

[Mountain Ash]

Great video CP! Nice to have assumptions tested now and again

[labr@]

The testing appears valid but the conclusion and message from the video is all wrong.

The main point should be that using scarf joints for load bearing beams/spars etc is better than end grain butt joints. The demonstration with the pine at the end of the video appears impressive but the test results shown earlier tell a different story. The numbers show the long grain is stronger than the end grain glue joint by between about 3 and 7 times for the species tested. Or to put it the other way around, the end grain glue joint is only 1/3 to 1/7 as strong as the long grain. A good scarf joint is close to the original long grain strength so is much better that an end grain joint.

As for side grain glue joints being stronger than the timber, well we already know that and design accordingly.
I don't see this as justifying changes to current practices.

[Chris Parks]

Can you add that to the comments of the video for further thought. The point about scarf joints is not relevant as he was not testing a scarf joint and not suggesting it either. I can think of a few ways to do it better and the Japanese about a thousand different ways.

[pal]

A little about scarf joints.
What is a Scarf Joint? (with pictures) (https://www.infobloom.com/what-is-a-scarf-joint.htm)

[AlexS]

Thanks for posting that. I'd like to see similar tests done with some of our hard-to-glue eucalypts, comparing untreated joints with those cleaned with a solvent like acetone, and also those primed with diluted glue.

[orraloon]

Can't argue with that as the experiment clearly shows the results. The fact that the endgrain joints always go at the glue line would give the impression that its not as strong but the reverse is clearly shown. Science to the rescue. We now just have to get all those flat earth people whipped into line.
Regards
John

[derekcohen]

If anyone read my report on building a plinth for a turntable ....

https://i.postimg.cc/bNr66wQZ/12.jpg

https://i.postimg.cc/v83qLV6R/11.jpg

... they may recall a reference to "sizing" the mitre joints: use glue to seal the end grain straws, allow to near-dry, and then glue again.

This creates a strong joint.

Regards from Perth

Derek

[rrich]

OK, gluing end grain to any other type of grain.

Understand what is going on. Essentially the process of gluing end grain is like gluing a bundle of tooth picks to a slab of wood. There just isn't that much surface area for the glue to work with.

[Bohdan]

OK, gluing end grain to any other type of grain.

Understand what is going on. Essentially the process of gluing end grain is like gluing a bundle of tooth picks to a slab of wood. There just isn't that much surface area for the glue to work with.

But the glue that soaks into the pores of the wood may in fact result in a larger area in contact with the "tooth picks" and so the ultimate load is limited by the glue strength and not the wood strength.

A stronger glue may in fact result in a stronger joint but it is not used because the current philosophy says that the limit is the wood strength, in face to face joints.

[AlexS]

I also size the mitres, Derek. I'd like to see the real differences between sized and unsized.

[pippin88]

Very interesting.

(https://www.woodworkforums.com/members/38385-labr-)labr@ makes a good point about not necessarily changing practice. Most joints / joint reinforcement is about incorporating the strength of the long grain fibres, rather than specifically avoiding end-end grain joints.

[woodPixel]

Outstanding.

Myth busted!

[Mobyturns]

But the glue that soaks into the pores of the wood may in fact result in a larger area in contact with the "tooth picks" and so the ultimate load is limited by the glue strength and not the wood strength.

A stronger glue may in fact result in a stronger joint but it is not used because the current philosophy says that the limit is the wood strength, in face to face joints.

Again we are confusing apples with oranges - a mitre joint and also a "machined finger joint" are "special cases." Strictly they are neither a long grain nor end grain joint but a combination of both, as some "over lapping" long grain fibers are present in the joint - short length but still present - which significantly improves the joint strength.

I have commented on this in another forum which I will copy across,

"I'm not surprised at all by the findings in the clip.

It is something that I have observed for myself, not scientifically but through practical experience and wide reading of available scientific validated research. I have always maintained that glue joint failures are mostly due to "operator error" - poor selection of joint type / placement / glue type etc; and most importantly what the operator does to the mating surfaces in the joint i.e. sanding faces etc.


Widely accepted procedures like wiping a joint with solvent for "oily wood" are likely to be debunked as well. I'm a fan of gluing freshly and accurately machined faces that have "suffered" minimal "torture."


One very common example of glue joint strength improvement/s through research is the "machined finger joint."

[Alkahestic]

The 'Law Of Glue Joints', as understood by me (a novice and a rookie, so please don't judge me too harshly):

1) End grain to end grain glue joints are weaker than any other glue joint.
2) When properly glued, side/long grain to side/long grain joints should never fail on the glue joint.
3) Gluing up without consideration for wood movement at right angles to the grain direction will eventually result in joint failure or wood failure.

To me this video has proven point #2. In the end grain to side grain (ES) and side grain to side grain (SS), the glue joint does NOT fail, it is the wood that fails. In the end grain to end grain (EE) sample, the glue joint fails. This would lend credence to point #1 being true, but we would need to actually have a failure of the ES or SS glue joints (and not the wood) to get a value we can compare against the actual glue joint failure in the EE tests. So to say the ES or SS joints are 'weaker' than the EE joint is false. The proper conclusion would be that the wood is weaker than the EE glue joint under a bending force in this test.

This video should also not be claiming that EE joints are "X% stronger" than other joints, partly because of the above and partly because it has only tested a bending force. What about tension, compression, shear, and torsion forces? As a purely academic exercise, I think this video has merit for testing bending forces on these glue joints but the conclusion it comes to is far too bold when the test is so limited. The test also has no allowance for strength of joint as it ages, so again, academically useful, realistically there is a lot more testing required to make any claims.

Actual furniture or anything that we create really, will usually undergo different types of forces in different areas, and I would challenge anyone to claim that an ES butt glued drawer box is as strong in practical use as a dovetailed drawer box. What about a butt joined chair? A jewellery box would probably be ok since they normally aren't moved much but then we come into the aesthetic part of woodworking - hiding or pretty-ing up the ugly bits. You don't even have to be a woodworker to appreciate that a mitred, dovetailed, or finger jointed box looks nicer than butt joins. Just because we can doesn't mean we should. I would venture to say that 'traditional' joinery and application of glues is traditional because it has, quite literally, stood the test of time.


Edit- TLDR version: My beef is with the testing not being comprehensive, making big conclusions, that I think are wrong.

[Mobyturns]

Drawing valid conclusions from a limited range and number of tests is risky, more so when the tests are not truly representative of "real world" applications and methodology or the the "real world" loading of joints.

A few simple questions will highlight why,

Why do wood workers

use mechanical "fastners" (dowels, pins, tongue & grove, corrugated (wriggly) nails, biscuits etc) to strengthen long grain joints?
use mechanical enhancements (mortice and tenon, dowel, slips, feathers, splines, lap joints etc) to strengthen long to side grain joints?
use mechanical enhancements (finger joint, scarf, etc) to end grain joints?
use a combination of specific mechanical enhancements and specific glues in "high strength" applications or particular "environments" (above / below waterline, high/low temp environments etc)?


The short answer is that practical experience with "real world" usage the failures associated with the loading of joints (static, kinetic, shock) and stresses due to environmental changes and associated wood movement etc have (mostly) been resolved through a combination of good joint design / selection, glue selection, enhancement / supplementation with mechanical fastners etc.

Yet we still see failures of the wood/glue/wood interface in long grain joints in chairs, tables etc in joints without some form of mechanical enhancement.

Yet another couple of questions to ponder -

why are end grain joints in the various layers of LVL staggered? A. because it makes for a stronger "overall" product.
why are there so many historic buildings constructed with "mechanical only" joints still standing?
why has hide glue been the "glue of choice" for so many fine furniture makers?


I'm not saying that the findings / conclusions drawn from the tests shown are invalid, but I am saying that the "test methodology" is not representative of "real world" scenarios, therefore exercise caution in applying the conclusions of those tests to "real world" applications.

[GraemeCook]

This is a very poorly designed experiment that "proves" only one thing.

Short grain is very weak ! We all knew that.

GIGO - garbage in, garbage out.

[woodPixel]

After reading everyone's comments, I feel that he has conclusively proven what was to be tested: That glue is 2 times stronger than the timber, even on end joints.

On some of the comments above, I believe that many of the Sayings we take to be Truths were made at a time when glues weren't SciFi - i.e. they were home made, or protein glues, or some other basic fabrication. With these glues the Truths are true, but with modern glues these Truths no longer apply.

As for making a chair with end grain joints, there is no need. As for making a table with end grain joints, its too hard. The point of the video wasn't to show that good mechanical adhesion and properties aren't important, it was to disprove the simple fact (now false) that "glue is the point of failure".

On unglued freestanding buildings, this dudes Instagram is highly illuminating! He makes Japanese houses and temples. Some beautiful work... Login • Instagram (https://www.instagram.com/dylaniwakuni/) Well worth looking at and studying.

[woodPixel]

Video two in the series of "end gain gluing and strengths"


https://youtu.be/2-cCCdwEhg4