MikeF wrote: ... You don't want it tighter or looser than its neighbours or the wheel will very likely be out. The spokes of some factory built wheels don't seem to be that tight and don't give a good "ping". Also they tend to be uneven tension, at least after riding, and is probably why breakages occur.....
Well uneven tension certainly doesn't help matters but IMHO it is just a symptom of a wheel that hasn't been built very well, rather than a primary cause of spoke breakage in its own right.
To a first approximation when the wheel is loaded radially, the spokes at the bottom (normally about four of them) get a fair bit slacker and all the other spokes increase in tension by a tiny amount. They also see very small loads in torsion because of pedalling, somewhat higher loads in torsion for hub mounted brakes, and lateral loads. Apart from the lateral loads, which stress small groups of spokes to some extent (small +, big -, a bit like radial loading), the loads are shared. The fatigue damage arising from torsion loading is usually not very large, simply because most riders don't spend that much time on the brakes.
Spokes fail because of fatigue; sure a bump in the road or a sudden knock may seem to 'break a spoke' but invariably that spoke was already cracked and without question doomed to break; the bump or knock just administered the coup de grace.
In general terms the fatigue life of parts can be dominated by the time to initiate a crack, or the time to propagate a crack.
I have carried out a careful analysis of failed spokes and wheels with failed spokes and it is clear that the fatigue life is dominated by crack initiation, rather than crack propagation. Nothing you do will stop spokes that are already cracked from breaking later, and you should expect badly cracked spokes to fail in a few hundred (at most a thousand or two) miles of further use. The trick is to stop them from starting to crack in the first place, which is why stress-relieving wheels is so important.
It turns out that if you have a few spokes already break in a wheel, there are most likely also a similar number of spokes that are already cracked and are definitely going to break pretty soon.
The bottom line is that if you stress-relieve a used wheel in which the spokes are a good fit (and you do it properly) then any cracked spokes there might be will break in the next few hundred miles, but after that you should get thousands of miles of use out of the wheel with no further breakages, because you won't initiate any new cracks in spokes.
As far as I know no-one has subjected failing wheels to this kind of scrutiny or has been able to draw this kind of conclusion with any confidence previously. Remember folks, you read it here first....
cheers
.