End grain glue joints stronger?

[darwin-bass]

Wow. So could a better scarf joint be made if both glue surfaces were end grain rather than just one?

 

[Matt O]

Neat engineer test, but no. There are a number of factors he’s leaving out as to why joinery is necessary for end grain joints.

He has only proven that after a two day cure, PVA is strong. Yes, as long as the joint isn’t glue starved, it will be strong… for a while. Sometimes, only a short while.

After a couple seasons of movement, and once the glue starts to decay a little, an end grain joint that lacks joinery will just fall apart. He seems smart and should know this. As an engineer, he should also understand the direction wood swells and contracts seasonally (width) and why that will eventually cause a glue only end grain butt joint with no mechanical joinery support to fail.

He’s also leaving out that PVA raises side/face grain fibers because it’s water based. When applied to the side or face of a board, the raised fibers of the two glue surfaces enmesh and create a strong mechanical bond. This doesn’t / can’t happen with end grain. That’s why PVA works so well as a side/face wood glue but fails as an end grain glue. There’s no mechanical bond in end grain joints. It’s also why a scarf joint works and why a end/end butt joint headstock would fly off on the first humid day of summer.

I hope he makes some joinery free furniture and checks back with us when it falls apart like the first woodworkers using animal adhesive discovered a few thousand years ago. That would be a sweet video for sure.

Also, a miter joint is a type of short scarf joint, it’s not an end grain joint. It gets its strength from the exposed side grain in the angled surface. They can be strong glue only joints, but it’s always best to use a spline.

 

[Big B.]

What he has actually proven is that boards break more easily cross grain than they do long grain. Kids in karate classes have known that for decades. Notice that every time the end to end joint broke it was the glue line that broke. On the other hand every edge grain test broke along the grain but never once did the glue joint fail.

 

[boxerrider]

What he has actually proven is that boards break more easily cross grain than they do long grain. Kids in karate classes have known that for decades. Notice that every time the end to end joint broke it was the glue line that broke. On the other hand every edge grain test broke along the grain but never once did the glue joint fail.

Exactly. The edge grain glue joints never failed, the wood did. He did make the point that sound woodworking practices utilize long grain for strength. FWIW - his end grain joints held better than I would have predicted.

 

[Big B.]

Exactly. The edge grain glue joints never failed, the wood did. He did make the point that sound woodworking practices utilize long grain for strength. FWIW - his end grain joints held better than I would have predicted.

I think the 4"x4" squares affected that end grain result. The pressure points are almost on top of the glue joint. If those were 8"×4" pieces you would get way more leverage and the joint would fail under much less pressure. If you glued two 16"×4" boards together end to end you could break that joint over your knee fairly easily.

 

[Jeff Siddall]

I think the 4"x4" squares affected that end grain result. The pressure points are almost on top of the glue joint. If those were 8"×4" pieces you would get way more leverage and the joint would fail under much less pressure. If you glued two 16"×4" boards together end to end you could break that joint over your knee fairly easily.

What you are saying is true, but that's just mechanical leverage so the statement applies to all materials, including side glued joints, not just end glued joints.

 

[Rôckhewer]

Humbug.
Video dude's tests were silly.

Any finish carpenter's here?
I'm not an aficionado of structural engineering.
I just built interior wooden shutter systems for about 8 years.
Endgrain butt joints?.... are not a thing.
Lap-joint?.. finger joint?... comb joint?... sure.
...no butts.

 

[Gilmourisgod]

Yeah, that end grain idea pretty much violates Furniture Building 101. I'd be more convinced by some real-world examples of it used for joinery long-term. Granted, the glues are a hell of a lot stronger than old fashioned hide-glue, but every time I've tried anything end-grain, it just glue-starved and failed.

 

[Jeff Siddall]

I agree with all the great arguments for using complex wood joints rather than relying on a glued butt joint. However, all those arguments apply just as well to side-grain glue joints also. Both are subject to failure of the glue joint if the glue degrades. Glue starvation is also a real thing but the video demonstrates the glue-up technique, and his end grain glue joints are still stronger than the wood lignin.

Bottom line: ideally avoid glued butt joints on any project, regardless of the direction of the grain, but definitely avoid it on side grain joints.

 

[Arie X]

that video is one of those things that has you rewinding several times over to make sure the bs meter wasn't broken due to the click-bait factor. fortunately mine is straight and true. this does not pass peer review for long term application in a stressed musical instrument joint or any other joint that isn't a paper weight object.

 

[RichterScale]

My impression of the vid wasn't that he was telling people to use end grain glue joints. Or that they should/could be used in any particular application.
My impression was that this was simply informational and explorative. Designed to confirm or dispell the belief that end grain glue joints are weak.
All things being equal, I came away from this believing that they are not weak, on their own. In fact, the data showed it took 3x the force to cause a failure.
So, for the folks calling BS,..... What part is BS? Is the data fake? Did I miss something?

 

[Rôckhewer]

All things being equal, I came away from this believing that they are not weak, on their own. In fact, the data showed it took 3x the force to cause a failure

That's just the problem.
All things ARE equal... in his tests.
Little bitty square blocks, testing the strength of a glue joint vs the stability of a given wood fiber matrix.

Real world facts? ... in practice?
Try gluing two 80 inch boards together via butt joint... to create a stable 160" header that supports a couple hundred pounds of shutter panels.
Simply hang that assemblage cantilevered off the bench & give it a bounce.. It WILL snap at the butt joint.
A lap joint... (or scarf )... will not.

I'm only talking about endgrain butt joints.
Side to side is fine.

 

[RichterScale]

That's just the problem.
All things ARE equal... in his tests.
Little bitty square blocks, testing the strength of a glue joint vs the stability of a given wood fiber matrix.

Real world facts? ... in practice?
Try gluing two 80 inch boards together via butt joint... to create a stable 160" header that supports a couple hundred pounds of shutter panels.
Simply hang that assemblage cantilevered off the bench & give it a bounce.. It WILL snap at the butt joint.
A lap joint... (or scarf )... will not.

I'm only talking about endgrain butt joints.
Side to side is fine.

Yeah, totally get that. But you wouldn't do that with any grain orientation. There has to be some type of interlocking/increased surface area joint in a situation like that, no matter what.
I didn't get the impression that he was saying that you could take 2 long blanks and glue them end to end and it would hold up.
I think he was saying that whatever the situation, the bond wouldn't be weaker simply because it was end grain to end grain.
In the case of gluing 2 long planks together, end to end, it would fail because of the tiny glue surface and leverage. But not necessarily because it's end grain.
That's what I got out of it anyway.

 

[Rôckhewer]

In the case of gluing 2 long planks together, end to end, it would fail because of the tiny glue surface and leverage. But not necessarily because it's end grain.

Yes it would.
Cross cut the same scenario.
160 inches... side to side grain.
Now glue.
Where's it gonna break?

 

[RichterScale]

Where's it gonna break?

It will break in the grain, rather than the glue joint.
Same joint, end grain to end grain will break right at the glue joint.
But the surprising results from the vid was that, even thought the end grain to end grain joint broke at the glue joint, it took 3x the amount of force to make it fail.
Just like a chain, it's only as strong as it's weakest point.
Side to side, the weakest point is the wood itself. End to end, the weakest point is the glue itself.
But it takes 3x the force to break the glue than the wood.

 

[Rôckhewer]

But it takes 3x the force to break the glue than the wood.

Yeah... in a mesily 6 inches.
... but beyond that... "leverage" rears it's ugly head... and all bets are off.

 

[RichterScale]

I think what's more important is whether or not gluing end to end has better tone.

 

[Rôckhewer]

I think what's more important is whether or not gluing end to end has better tone.

What's the best joinery for metal?

 

[Jeff Siddall]

I think what's more important is whether or not gluing end to end has better tone.

Hah! Are you going to run that experiment?

After nearly 40 comments, many of which sound like disagreements, I actually think everyone is agreeing. To recap:

1. Use a more complex wood joint than a butt joint if possible
   
2. Use a bigger glued surface if possible
3. Reduce leverage on a joint if possible

However, if the requirement is a simple butt joint with a fixed glued area, the answer is easy: use end grain if possible. That's the part that most people find counter intuitive.

For example, to make a glued two piece bass body blank as strong as possible along the glued axis would most people choose to make it from:

1. Two pieces of 2x6, 20" long, glued on the side grain
2. Two pieces of 2x20, 6" long, glued on the end grain

The best answer is option 2, but I have never seen anyone do that. Possibly because 2x20 boards aren't common, but more likely because people assume the end grain joint would be weaker.

 

[daveman50]

What's the best joinery for metal?

A butt joint is the best joinery for disco.