Joined: Jan 2010
Posts: 9,808
Likes: 1,781
From: Northern California
Bikes: Cheltenham-Pedersen racer, Boulder F/S Paris-Roubaix, Varsity racer, '52 Christophe, '62 Continental, '92 Merckx, '75 Limongi, '76 Presto, '72 Gitane SC, '71 Schwinn SS, etc.
The testing and quality-control of these parts which assures our safety can't go on for 30 years, except as our final "field testing" at our expense.
Aluminum does have a time-dependent "creep yield" failure mode tendency in tension that steel does not have, so near-infinite life under static tension load depends on an extra measure of designed-in stress reduction (more material) in addition to high levels of processing (i.e. alloying and cold-forging or heat-treatment).
The designers are aware of aluminum's vulnerability to such failure mode, but are designing parts intended to be competitively light.
Such strength/weight performance levels are only realized within a certain time period when parts such as a clamp or a hub flange are subjected to constant tensile stress.
It amazes me that these stem and similar hub-flange failures so often happen when the bike or wheel are sitting unused.
There are no 30-year-old Ksyrium SL wheels out there yet, but the infrequency of failure of their aluminum spokes shows careful design including sophisticated processing of the spoke material.