interesting conclusions from a carefully designed test of glued joint

TEST ON SCRAP FIRST! If your question is about repair work, either regluing or refinishing, please post it in our Repair Section.
Post Reply
Bob Hammond
Posts: 638
Joined: Sun Jul 22, 2012 4:13 pm

interesting conclusions from a carefully designed test of glued joint

Post by Bob Hammond »

I'm a bit surprised by the results.

https://youtu.be/m7HxBa9WVis
User avatar
Bob Gramann
Posts: 1101
Joined: Fri Jan 06, 2012 11:08 am
Location: Fredericksburg, VA
Contact:

Re: interesting conclusions from a carefully designed test of glued joint

Post by Bob Gramann »

Wow! Very useful video. Thank you.
User avatar
Bryan Bear
Posts: 1375
Joined: Fri Jan 06, 2012 1:05 pm
Location: St. Louis, MO

Re: interesting conclusions from a carefully designed test of glued joint

Post by Bryan Bear »

I saw that video. Well most of it, I didn’t make it to the end.

I felt like the conclusions he drew were misstated a bit. He made it sound like end grain glue joints were stronger than side grain or end to side because those samples took more force to separate than for the others to break. I’m not disagreeing with the data points just that those data speak to the properties of wood rather than of the glue joints.

The side grain glue joints never failed, the wood failed. Wood is much weaker along the grain than across it. Just ask any luthier, firewood chopper or karate student. The end grain glue joints all failed but it took more force to break them than it did to break wood along the grain. That does not indicate that end grain glue joints are stronger just that wood is stronger along the grain and an endgrain glue joint would be the weakest link in that chain. In a side grain glue joint the wood is the weakest link.

It was interesting to see that end grain joints held up so well. It makes me worry less about heel blocks on cutaways. An interesting test would be to pit end grain joints against lap joints and scarf joints (all along the grain).
PMoMC

Take care of your feet and your feet will take care of you.
Bob Hammond
Posts: 638
Joined: Sun Jul 22, 2012 4:13 pm

Re: interesting conclusions from a carefully designed test of glued joint

Post by Bob Hammond »

my alter-ego suggested testing a long scarf joint.
Marshall Dixon
Posts: 169
Joined: Tue May 21, 2019 8:58 pm
Location: SW Oregon

Re: interesting conclusions from a carefully designed test of glued joint

Post by Marshall Dixon »

Well done presentation on the subject. I was taught that end grain joints are a no-no. Several years ago a maker told me that he used a butt joint to attach a neck in 1970 and it was still holding up. So I investigated further and found that makers going back to the 1930's used the butt joint exclusively for attaching neck to body on gypsy jazz guitars.

I copied this information several years ago but can't find the website now.

"Even pseudo experts on Busato (there are many including some well known brokers) are not aware that Busato did not use dovetail connection. The neck heel is glued flat on the neck block. Castelluccia did the same but there is a 15 mm approx. hole in the flat neck block. When you have to do a neck reset on a gypsy guitar which looks like Busato, if the joint is flat, investigate further, you might have a Busato."
User avatar
Barry Daniels
Posts: 3186
Joined: Thu Jan 05, 2012 10:58 am
Location: The Woodlands, Texas

Re: interesting conclusions from a carefully designed test of glued joint

Post by Barry Daniels »

I made my first three guitars using the method in David Russel Young's book which had an epoxy glued, butt jointed neck. None of them have come loose in the last 45 years. But that joint has a lot of surface area and should not be mis-interpreted.

I certainly don't use that style of neck joint today.
MIMF Staff
Alan Carruth
Posts: 1264
Joined: Sun Jan 15, 2012 1:11 pm

Re: interesting conclusions from a carefully designed test of glued joint

Post by Alan Carruth »

Bryan: watch the whole thing.

One thing I noted in the wood break tests with no glue lines was that the wood was clearly failing in compression on the underside before the tension side let go. If you look closely at most of the footage you can see cross grain compression lines running across the test pieces just at the edges of the pipe support. This is the defect called 'wind shake', which actually occurs when a tree is felled onto something hard, such as another down tree or a rock, which causes such compression failures. This is, of course, especially true of the softer samples, such as the pine, as opposed to the oak. I wondered if larger pore size (say, walnut vs cherry) would affect this, but the video went by too fast to really keep and eye on it. At any rate, I have to wonder if it's quite correct to say that the wood failed first in tension.

End grain joints do fail along the glue line, and that alone may have been enough to make people think that they're 'weak'. I'm reminded of the situation with side tapes. When I did similar (if not so sophisticated) testing of side tapes the tapes almost always broke when the wood did. This has lead people to conclude that they don't work. In fact, though, it took a lot more force to break the samples with tapes than the ones without, so tapes do add useful strength. They also tend to stop cracks running along a side, which can greatly simplify repairs.

I wonder about the surface energy of his glue samples. Were all the cuts made at the same time, and glued up within 15 minutes of cutting? If one side of a joint was an older surface it would tend to bond less strongly, and that could be one thing that's affect the spread of his tests.
User avatar
Bryan Bear
Posts: 1375
Joined: Fri Jan 06, 2012 1:05 pm
Location: St. Louis, MO

Re: interesting conclusions from a carefully designed test of glued joint

Post by Bryan Bear »

I watched the whole thing and I still disagree with the conclusion he draws. More specifically, how he words it. He says end grain glue joints are stronger than side grain joints. He has not demonstrated that at all. He has demonstrated that a beam with an end grain glue joint will take more force to break (the joint) than it will take to break a beam of similar dimensions with a side grain joint (the wood is breaking here).

This demonstrates that an end grain JOINT is stronger than side grain WOOD. We all knew that wood was stronger in the long Dimension than across the grain. What some of us (me) did not know was that the endgrain (GLUE) joint was also stronger than wood across the grain. That is good to know, but it does not demonstrate that end grain joint is stronger than a side grain JOINT.

As he points out, you don’t typically see side grain joints on long skinny cross grain beams. That orientation is a bad idea and the designer would run the grain along the length instead of across it. But, he then goes on to show how weak it is y breaking it. But. . . the joint never fails, the wood does in two places.

What I took from the video is not that end grain joints are stronger. I took:

Side grain joints are stronger than wood across the grain (already knew that)
End grain joints are weaker than wood along the grain (seemed inherently obvious but was demonstrated with the tests).
End grain joints are stronger than wood across the grain (I didn’t NOT know that).

I’m not sure how much this really matters for my work other than I will worry less about scarf joints and cutaway heel blocks. I don’t really do any work that would see end grain glue ups needed. Of course, now that I say that, I will find 10 of them this week.
PMoMC

Take care of your feet and your feet will take care of you.
Alan Carruth
Posts: 1264
Joined: Sun Jan 15, 2012 1:11 pm

Re: interesting conclusions from a carefully designed test of glued joint

Post by Alan Carruth »

Bryan Bear wrote:
"This demonstrates that an end grain JOINT is stronger than side grain WOOD."

OK, now I see your objection, but isn't it really a semantic one? We've long been told that "X makes joints stronger than the wood itself" and he's demonstrated that, at least for the short term under that sort of loading. Once you've got that the actual strength of the joint doesn't matter, since we know that's not where it will break.

As a friend of mine pointed out years ago, wood is actually mostly air. Glue joints can be stronger than the wood even if the glue itself is not as strong as either the cellulose or lignin the wood is made out of because it fills in air space, and 'something' is stronger than 'nothing'. That's evident from the end grain results.
User avatar
Bryan Bear
Posts: 1375
Joined: Fri Jan 06, 2012 1:05 pm
Location: St. Louis, MO

Re: interesting conclusions from a carefully designed test of glued joint

Post by Bryan Bear »

For sure, it is a semantic issue for me, but I think it is an important one. If you are presenting g something as a scientific test and results, your conclusion needs to say exactly what it means. Lots of people will watch that video and come away with the lesson that endgrain joints are the strongest. Or worse, some might decide they can butt two short prices together in place of a longer one and not worry about it’s streanth.


“We've long been told that "X makes joints stronger than the wood itself" and he's demonstrated that. . .” Yes and no, he demonstrated that side grain joints are stronger than wood in that direction AND that end grain joints are stronger than wood across the grain. But he also demonstrated that end grain joints are weaker than long grain wood.


Like I said, I learned something, that’s a good thing! I just wish the conclusion was represented accurately.

As predicted, I thought of another end grain glue up that I encounter. The lap joint in an X brace is all end to side grain. I used to feel like the glue inside the joint was not very effective and all the stength came from the glue between to top and brace as well as the glue between the brace and cap. Now I have a renewed respect for how strong that intersection really is. Though I still won’t make a long brace out of two short ones <G>
PMoMC

Take care of your feet and your feet will take care of you.
User avatar
Bryan Bear
Posts: 1375
Joined: Fri Jan 06, 2012 1:05 pm
Location: St. Louis, MO

Re: interesting conclusions from a carefully designed test of glued joint

Post by Bryan Bear »

Earlier in this thread, I mentioned that my biggest problem with this video is that the conclusion is worded in an inaccurate way and that some people will take the wrong message from it. It would appear that I was right (at least about people taking the wrong message). Both the Wood Whisperer and Stumpy Nubs have gotten enough comments about this video that they published their own response videos; not to dispute the material presented but to highlight the context of the findings.

I though Stumpy’s video was pretty spot on and worth a watch for anyone who watched the original video.

https://youtu.be/dwLWmoGh59g
PMoMC

Take care of your feet and your feet will take care of you.
Post Reply

Return to “Glues and Finishes”