Many of the glued joints in guitars are subjected to a combination of shear or tensile stresses. The bridge usually has the highest loading, with areas near the neck joint, top bracing, and top plate edges significantly lower. A few places have compressive loading, but those areas rarely fail.

Most of these stresses are produced by the tension of the strings, but that tension is not constant. The tension varies repeatedly as the string vibrates, producing what engineers call asymmetric cyclic loading. Cyclic loading causes fatigue failures, in which small cracks develop, and then grow larger with each cycle. If you've ever bent something back and forth until it broke, then you've participated in cyclic fatigue.

This applies to glued instuments in several ways. The most obvious source of cyclic stress in an acoustic instrument is the vibration of the strings, but they are specifically designed to accommodate that problem. If the glue and the wood (or different woods) don't expand exactly the same with differences in temperature or humidity, then the glue and wood(s) can be significantly stressed. Removing all the strings at once can be a significant stress cycle, as can bending the neck for tremolo.

Another factor involved in the selection of instrument glues is the elastic modulus, which is a measure of a material's stiffness. For example, rosewood is generally much stiffer than pine or cedar. When the joint between two different materials experiences cyclic loading, usually the softer of the two materials fails. This problem becomes more complicated for wood, because wood has different elastic properties depending on the direction of the grain. Also, when a glued joint is not fitted properly, the thick layer of glue can begin to act as a third material, which causes additional stresses on either side of the joint.

To sum up, the various physical properties of the glue should match the wood(s) as closely as practical to prevent cyclic loading from causing fatigue failures near the glue joints.

Most glues are much stiffer than most woods, but one of the major advantages of using hide glue is that its elastic properties can be put in the same ballpark as common hardwoods. This probably occurs when the percentages of cellulose and collagen are properly matched. Most of the modern glues, such as epoxy or cyanoacrylate are much stiffer, and also do not match the wood in expansion when temperature or humidity changes.

The commercial carpenter's glues, such as Wellbond, Elmer's, etc. have too many plasticizers to be useful for instruments because of the creep they allow in constantly stressed joints. The same holds true for most of the hot melt glues. Beta-cyanoacrylates (the HotStuff mentioned above) are thick enough to bond softwoods well, but are also slightly soluable in water. Many commercial epoxies are too viscous for thin joints because they are intended to fill large gaps.

There currently seem to be few replacements for properly fitted parts and a strong hide glue, but some of the marine polyesters may work well for tough environments. Their low viscosity, excellent strength, and moderate elasticity should be effective for instruments. I've reglued several guitar bridges and one top plate for friends who live on sailboats, and almost a decade later the pieces are still stuck together. Unfortunately, such resins are usually supplied in a minimum of quart-size cans with about a thimble-full of catalyst.

For lower stress areas such as inlays or the back plate, any cyanoacrylate, epoxy, or stearate glue designed for wood should work reasonably well, even in humid locales. However, there is little substitute for properly fitted joints and minimizing the environmental cyclic stresses.

~S~

PS. If you must cycle between hot/cold or dry/damp locations frequently, mixed woods in the sound box (and heaven forfend, plywood) should be avoided.