Glue Joints, some additional thoughts...
When wood is turned into timber suitable for engineering use all but 8 to 10% of the moisture in it is dried out.
Usually the wood becomes quite stable but it is still subject to the atmospheric moisture around it.
If the timber is stored in humid environments it will absorb additional moisture.
two considerations arise form this hydroscopic characteristic.
The first is that we coat all surfaces in a wooden aeroplane with a varnish. this pretty well waterproofs the timber and slows the absorption of moisture.
The hollow parts in the aeroplane like the nose of the rudder and elevators are all varnished internally.
We leave no unvarnished surfaces on an aeroplane.
when gluing on a ply surface that would seal a hollow we mark on the ply where it will be glued and varnish all the rest of the surface. we then make sure that the unvarnished glue areas are all fully coated with epoxy when we do the gluing.
The second consideration is that as the moisture content increases in the wood it swells up, as it dries out it shrinks.
What that means is that if we glue together two pieces of wood with different moisture contents, they will eventually stabilise on a common moisture content. while doing this there will be noticeable shear forces set up in the glue joint. since all the loads on a joint are additive we are preloading the glue joint and in effect making the joint weaker.
storing all the wood together and allowing it to stabilise on a common moisture content by not getting around to doing the building for a few months is actually a good thing.
Ok what does a good glue joint look like?
where the two timber members meet the surfaces should be of the same shape. usually this will be a pair of flat surfaces.
the members should be cut to exact length. if we put in the joint a member that is not at exact length we will set up the forces that cause the section of the structure to warp and twist. since the timber sections are quite small in most cases it is not an expensive problem to cut another piece if we find that a member is too small. if we find it too long we just sand it off a little more.
You will notice on the plan there are gussets on the joints. if you look at a joint area it isn't large. what the gusset does is increase the glued area of the joint many fold and substantially reduce the stresses in the glue bonds. This makes the glue joints much more reliable.
When we glue the actual joint we apply the epoxy to both surfaces. then we position the surfaces together and clamp them so that they have no relative movement for the entire time it takes for the glue to set and cure.
Once we have it clamped in place we then we scrape off all the excess glue. the glue bond is just in the area between the members being glued. glue fillets don't add any strength to a glue joint and just add needless weight to the aircraft.
Where Resorcinol Formaldehyde glue is used, dribbles and glue fillets are actually a liability since they crack over time and at the bottom of the crack we get bare timber which lets moulds and moisture get past the varnish protection.
When plywood is made the timbers are pressed in heated presses. we substantially improve the bonding to plywood if we give the ply a light sanding all over the glue area. this improves the glue bond. this is where our sanding blocks make the job easy.
I hope needless to tell is that we remove sanding dust from surfaces being glued.
A small anecdote on improving glue strengths.
a company here in Perth experienced poor performance from some of their laminated timber beams.
After lots of experimentation it was found that gluing old surfaces of wood compromised the glue joint.
What they found is that gluing a freshly machined wood surface made for a stronger bond. in Gluelam manufacture the section comes out of a thicknesser and is immediately coated with glue and placed in position.
In practical terms it seems that gluing a surface that was prepared no more than a quarter of an hour before consistently achieves sound glue joints.