Tom Watson wrote:
....

let me ask you a question, Chris.

In all the tests that I have seen, although it is true that the glue
line does not break, the break is always near the glue line. It does
not appear to be the case that the break is in another part of the
panel that is away from the glue line.

What's up with that?
....

Every test I've seen the two pieces are the same width so the joint is
in the middle. When the board is loaded and supported on the edges, the
point of highest stress is then in the middle; hence unless there is a
weak point farther towards one edge or another, the most likely place
for the failure is near the point of highest stress, the middle.

Since as noted, the glue joint actually is generally as strong as or
stronger than the material, typically it is slightly to one side or the
other of the joint where the break occurs.

Simply physics of the test geometry is the basic explanation...

The more interesting test that is illuminating is the one of the bridle
joint joint loaded perpendicularly to the grain on one piece -- even
there it is typically either the wood that breaks or a combination
rather than a glue-line failure for well machined joints.

--


--