Glue joint testing?

[K. L. McReynolds]

Ok, all the discussion(again/still/whatever ) about clamping pressures/success/technique aside, how do you actually test the glue joint?

I used to make a lot of picture frames. I have picture framing vises as well as band clamps. I experienced a lot(one out of 6 or 7) failures early on with dense or oily woods. Purpleheart, bubinga, bloodwood, hard maple, and cocobolo are examples.

The problem occurs on miter joint frames, half lap/overlap/etc frames are not affected.

Found a solution. Mini biscuits create a much stronger joint by giving the glue more area and a different density wood to join. In this case, the mini biscuit does make a stronger joint.

Now, the problem is my testing method. I simply drop the completed frame(after a 24 hour+ time in the frame vise/band clamp) so a corner hits the floor first from waist high.

However, that only works for small frames(12x12 or smaller and 1" wide or wider stock).

Larger size(or more narrow stock) get a torsion test by stretching/compressing the frame on opposing corners just enough to stress the joints.

I've glued white oak stock 1&1/4" thick(for stair landings), red oak stock for bench/cabinet tops, and 1by pine stock(for a door) and never even thought about testing the glue joints.

I glued spalted sycamore to walnut(TV tray legs and stretchers) and had 50% failure because the sycamore was too soft. (Was trying to make metal free joints) The joints held, but the sycamore stretchers simply came off, leaving a layer of wood/glue. Had to clean, reglue and add some brads.

Not the best choice of woods, but was what I had on hand when the need arose.

[jteneyck]

The down and dirty way to test a picture frame joint would be to hang it from one corner and hang a pail on the other corner and fill it with water until it breaks. Might take several pails, however. A step up would be to use a hydraulic cylinder that has a gage on it. Put it inside the frame and push opposing corners apart until it breaks. Or if you have a hydraulic frame press, put it inside and press from outside in.

Or just use loose tenons and there will be no need to worry about testing.

John

[Doug_H]

KL,

Good question. To answer directly, I seldom if ever actually test a glue joint. So, your question got me to think about why I don't. Short answer -- I have had one glue failure in 15 years. No brag, just fact.

For most applications such as M&T joints, I make a good dry fit and I believe the claim that PVA glue bonds are as strong as the wood itself. In other words I assume that the mechanical construction, not the glue alone, holds the joint together. FWW published a "torture test" of glue joints a while ago and as I recall the wood failed before the glue in most or all of them. BTW the miter joints were stronger than conventional wisdom would predict.

I don't make picture frames very often but if I do I make a keyed miter analogous to your mini-biscuit joints. I make larger boxes, drawers etc mechanically strong with drawer lock, lock miter, or finger joints.

So, what about butt joints? I used to think that I needed splines, etc for panel glue-ups but finally believed my betters who say that butt joins are strong enough and that splines, etc are mainly useful to hold the boards in alignment. I have done maybe 10 tabletops without a failure.

I make a lot of segmented turning blanks -- segments and staves. These all rely on long-grain butt joints. Segmented rings, especially, are subjected to a lot of pressure in the flattening process. Then the rings are glued together to form the blank (more butt joints) and after turning the glue surfaces are only 1/4" wide. SO -- about 300 butt joints per turning, about 100 turnings over the years -- 30,000 butt joints and only one failure. That failure occurred in a segmented lamp base made from a piece of live oak firewood that I had obviously not allowed to dry properly before I used it. When it dried more after being brought to Colorado it cracked along one of the glue lines. I repaired it with epoxy and it has held.
(Come to think of it I have melted a few joints on the lather when I got too agressive and overheated the wood.)

So -- sorry for the long answer. I don't worry much about testing glue joints.

Best regards

Doug

[Richard D.]

I don't do stress testing on my finished work. For picture frames I'd cut a few extra joints from the scrap and destructively test them for strength. In other words I see if I could pull them apart and how much force it took.

[Steve N]

Mac, on the oily exotics have you ever just used poly glue?










I find on a well sanded end grain it works like magic, even on miters.

[Axehandle]

This is only marginally relevant but since I got a domino joiner, I have scaled up my frames a bit to fit the smallest tenon domino into the miter joint.

I have never tested them and I never had a problem in the past with the miters failing. I never just did a straight glue miter though either. There was always some kind of extra connection.

I know it is not purist but why not use staples on the back side of the frame miters?

or am I missing the point?

Your idea with biscuits is a good one. I don't think you will have a problem with it. Testing it though...? I don't like destroying the things I just made. If I built it knowing I was going to try to destroy it in some kind of torture test, I would probably chintz out on materials and time investment which would immediately compromise everything.

[K. L. McReynolds]

tried Gorilla, but I should have used gloves as I am rather messy where glue is concerned.

[John Mihich]

I always cover the back with butcher paper. This really adds a lot of strength to the frame. This is how a lot of framers do some of their frames.

[Gary G™]

Wipe oily woods with acetone before gluing.
For miters, I put a thin layer of glue and let it skin before gluing the pieces together.
Mitered picture frames get a slip feather at each miter.

[K. L. McReynolds]

I did some experimenting with the woods I listed.

I was not making wide stock frames. More in the 1" to 2" wide range and under 10" by 8" in size.

Results were that wiping with acetone made little difference, probably due to the minimal surface area of the miters.

I was using a Freud 80 tooth Ultra fine finish blade. I found that blade(had two) sometimes glazed the surfaces of the miters and a touch of sanding helped.

The final(and still used today) solution was to use the mini biscuits regardless of wood species. An unexpected side effect of using the biscuits was the biscuits often held the freshly glued frame together well enough to apply the clamp.

Not such an asset when using the frame clamps, but quite welcome when using a band clamp.