I wonder about handle pop offs being blamed always on one glue type or another. It seems like there are all kinds of different woods with very different expansion properties glued together in handles with different thicknesses, grain and ring orientations, surface preps (as mentioned already), degrees of oiliness, and different species gluing characteristics.
I was once a boatbuilder with a smaall boatshop, and have spread well over 100 gallons of epoxy in my time. The Gougeon Brothers
.E.S.T. System inventors) always warned about gluing hardwoods with a thicknesss of over 3/4" because the forces of expansion were so great with variations in humidity. They tested woods and found that usually it was the wood itself that separated in failures, in a very thin layer just adjacent to the glue line -- even though the glue had been thought to have failed.
And also, if I remember correctly, that roughening the gluing surfaces did not improve joint strength. I could be wrong in all this, but I do wonder whether too much is made of particular glues. I think most any glue specified for wood probably is stronger than the wood itself, if properly applied to a suitable wood type. But glues are often asked to do more than they can do to hold the wood itself together with humidity cycles and shooting shocks, at the thickness people use.
Gougeon used to recommend breaking up thicknesses to help spread the swelling strains, and maybe a handle riser would better be made up of 2 layers of 3/8" instead of on single 3/4" piece, as an example. Also mechanical pinning is far more effective at reducing shear in something like a handle than relying on a glue line alone. In boatbuilding pinning a joint of any thickness is common, mainly to avoid splitting, but glue separation is a form of splitting, too, when you think about it. Just a couple cents worth here.
