The problem there, from examples we've seen of similar aluminium wing
failures, is when a rather thin extrusion is welded to a thicker part.
The weld pool contracts by up to 10% as it cools, generating yield-level
stresses in the adjacent thinner metal. Aluminium is vulnerable to
fatigue failure & the combination of this high residual stress with the
much lower cyclic loads of the rowing stroke is sufficient to initiate a
hard-to-detect crack at the most vulnerable location. That crack then
propagates rapidly, with the result you describe.
Even where 2 similar sections of extrusion are being welded together,
the greater thickness of the weldment & the resulting residual stresses
in the adjacent thinner material can again lead to premature
fatigue-induced failures.
If, in addition, you row in brackish or saline water then, even if the
rigger has been anodised, you will get the more rapid process of
stress-corrosion cracking accelerating this fatigue failure. Chloride
ions in the water will attack the stressed metallic grain boundaries at
any weak point in the coating &, thereafter, will drive on the crack
formation. This can often start from the inside of the rigger
extrusions since it is very hard to weld a tube which does not have a
small vent hole to atmosphere, so water easily gets into the unprotected
inside of the tube, where the damage it starts can't be seen.
Unless welded aluminium riggers are stress-relieved (the heating of
which process will soften & distort the rest of the rigger) or in other
ways treated to hold off weld-induced fatigue failure, they will
eventually fail through that combination of weld stress and cyclic loading.
Carbon riggers have a different problem: they won't corrode, but carbon
has poor impact resistance compared to metal. It will take a knock up
to a certain level without apparent injury, but at any higher load it
will fracture, not bend or deform. And, if the impact was enough to
start a fracture you may be unable to detect a defect which may then
continue to grow. It is for that reason that elaborate systems for
crack detection in aerospace carbon structures are required.
HTH