End grain gluing. We all know how weak that is, don't we....

[MikeG.]

Of course we do. It's the first part of a joint where the glue will fail. Hardly worth putting any glue in there, other than as a bit of a gap filler..........

My wooden hold down failed today*. The end grain glue joint didn't, and the larger area long grain to long grain join did. Maybe we need to have a little re-think of those things that we all "know".


*This was a twenty minute experiment which I made about 3 or 4 years ago. I'm astonished it has lasted so long considering it is made of the wrong wood, and has taken some real abuse. I'll make a new one or two properly in something stronger than pine. Probably ash. In the meantime, I've glued this one up again..... :lol: :lol:

 

[doctor Bob]

Interesting, couple of pegs may increase it's lateral strength

 

[MikeG.]

I think I'll do wedged tenons, Bob.

 

[Sgian Dubh]

Maybe we need to have a little re-think of those things that we all "know".

I'll make a new one or two properly in something stronger than pine.

I'm not convinced at this point, unless you can add an additional convincing factor. It occurs to me, for example, that there may have been some weakness in the long grain of the morticed pine part, e.g., a hairline split. Slainte.

 

[MikeG.]

But the long grain glue joint has failed, in a tight fitting mortise. It's failed completely. It almost doesn't matter whether there was a weakness the pine, because the point of failure (ie the weakest point) was the glue join mortise-to-tenon cheeks.

 

[owen]

Is the one piece of timber accoya? What kind of glue did you use?

 

[xy mosian]

As I see the joint, from here, is it possible that the end grain to long grain joint was clamped? I wonder if the long grain to long grain joint was. I have not come across advice to clamp across a tenon during glue-up. No clamping means little significant glue penetration of course.
The image shows a white deposit on the tenon, that looks like dried PVA. I think it only dries like that if left with some thickness.
Just some thoughts from a fare distance.
xy

 

[MikeG.]

Is the one piece of timber accoya? What kind of glue did you use?

No, it's parana pine. The glue will almost certainly have been Everbuild D4 wood glue.

 

[MikeG.]

As I see the joint, from here, is it possible that the end grain to long grain joint was clamped? I wonder if the long grain to long grain joint was. I have not come across advice to clamp across a tenon during glue-up. No clamping means little significant glue penetration of course.
The image shows a white deposit on the tenon, that looks like dried PVA. I think it only dries like that if left with some thickness.
Just some thoughts from a fare distance.
xy

I've no idea whether or how I might have clamped it. The mortise would have been chopped out in the normal way (ie chopped across the grain, with no paring). In pine that leaves a pretty rough surface, so it is possible that there were pockets where the glue was deeper than it should have been. The only thing I remember about making that piece was that the joint was really tight. I obviously knew it was going to be under a lot of strain, and I didn't expect the hold down to last even a week, so I made the tightest joint I could.

 

[xy mosian]

I expected that would have been your method Mike. The 'standard' method.
I just cannot get my head around clamping the likes of rails on tables, etc., being clamped end to end. That method can only apply pressure to the long grain - end grain joint. As you say that is not supposed to work well.
In your case, the forces the joint has been under, I would call that a good length of time to joint failure.
xy

 

[samhay]

If the mortise was really tight, you may have squeezed most of the glue out of it when you put the joint together.
However, as someone contemplating an end grain-to-end grain joint this morning, your post gives me some solace. Thanks.

 

[Phil Pascoe]

Except it's not clamped to stick end grain, it's clamped to avoid a gap between the rail and the leg.

 

[Sgian Dubh]

But the long grain glue joint has failed, in a tight fitting mortise. It's failed completely. It almost doesn't matter whether there was a weakness the pine, because the point of failure (ie the weakest point) was the glue join mortise-to-tenon cheeks.

I see that Mike. But perhaps it's a partial failure because I can see nodules of adhesive on the faces of the tenon that show just beyond the slivers of pine still attached to it, and I see that on both sides. So how effective a glue bond is there in those globules of glue? I think we both know the answer to that, i.e., very little.

In your second photo down, looking at the open end of the mortice, it appears as if there is little or no tearing of the grain at the interface between the cheek of the mortice and the face of the tenon. So I wonder if the mortice cheeks and the tenon faces were as snug as they could be to provide a really good long grain glue bond. I do appreciate that a sawn tenon and a chiselled mortice are never going to create ideal mating surfaces in the same way that a well prepped panel edge joint can be achieved with a hand planed sprung joint, for example. It's remarkable, I think, just how resilient M&T joints are given how they're mostly executed, and I think that's down to the interlocking mechanical properties the form confers. And this piece, being a hold down, has probably experienced severe shear forces in use, greater than many an M&T, although the side rail to back leg joint in a chair might be an example of a rival for stress.

The bits of pine still attached to the tenon that meet the shoulder line may simply be better glued long grain to long grain at that point, or it's possible that there was simply less shear stress in that region than the shear stress further away from the shoulder line, i.e., deeper into the mortice.

I still don't see convincing evidence that the glued end grain of the tenon shoulders have made a better bond with the long grain of the matching edges (shoulder line) of the morticed pine than the long grain to long grain bonding, in that area of the joint, between the tenon face and the mortice cheek.

Anyway, I suspect none of this really matters - you just need a new hold down, and maybe you could get one of those cast iron holdfast jobbies, or similar from elsewhere, ha, ha. Slainte.

 

[MikeG.]

What, buy something, Richard? With actual money? You've got to be joking...... :lol:

 

[Blackswanwood]

I am not sure I would trust one of my mortice and tenons to be used in a holdfast so credit to your woodworking skills Mike that you got so long out of something you expected to last a couple of weeks.

A good few years ago I saw an Amish workshop where they were using wooden holdfasts fashioned from where a new branch had grown forming a y shape.

 

[MikeG.]

I've been on the lookout for the right shaped branch to do that for a while now, but I have to be in the woods pretty soon after the branch falls because they aren't my woods or trees.

 

[Sgian Dubh]

What, buy something, Richard? With actual money? You've got to be joking...... :lol:

Well, I know that might be pushing the boat out a bit ... but just thought you might be, er, tempted. Slainte.

 

[MikeG.]

What, buy something, Richard? With actual money? You've got to be joking...... :lol:

Well, I know that might be pushing the boat out a bit ... but just thought you might be, er, tempted. Slainte.

Never!! Not if I can make it myself. Besides, my dog holes are square.

 

[MikeG.]

I made a new one. I grabbed some wood off the ash pile (notice, I didn't say I grabbed some ash):

Turns out the shaft is sycamore, not ash. I pressed on anyway, having dimensioned it. Will I regret that?

That's the cleanest mortise I can chop be hand, I reckon. I couldn't find a dimension more than one twentieth of a mm out, and the faces are clean as you can see. No excuses this time!!!

The tenon cheeks, straight from the chisel, came up virtually polished. I love working with ash (although it's hard to finish nicely). Again, all of it that I could measure within 1/20th of a mm:

The joint is wider by 10+mm this time, the "leg" bells out to make it about 10mm thicker at the top too, and I've done a wedged edge to the part of the tenon that will be in tension:

I painted PVA wood glue on both the tenon and the inside of the mortise with a brush. I chamfered the ends of the tenon so it didn't scrape the glue out as it was driven home, then not only did I tap it home and wedge it, but I also clamped across the joint in a way I would never think of doing:

That was just about the best m&t I can imagine doing. There can be absolutely no excuses this time. I didn't rush, and with the test-to-destruction of the last one I've adapted the design and improved the construction. So, what's your guess.........how long will it last, and if it fails, where will it fail?

 

[woodbloke66]

Some years ago I was nattering to Rob Ingham no less who surprised me by saying that end grain gluing with modern adhesives can be quite strong. I certainly wasn't going to argue with him :lol: - Rob