Originally Posted by
Kontact
Elbows out spokes do not touch the hub flange and curve around it. So when the spoke tension changes from rolling, the resilt is that curved section flexes back and forth like a leaf spring.
18/8 stainless is a horrible spring. Much worse than even unheat treated chromoly. Whichnis why we don't build frames out of it. So the flexing elbows break.
When you seat the elbows you shape the spokes like a string, conforming to the flange in the straightest possible path. Now tension changes are all along the length of the spoke and there is no bending force. That's how spokes are supposed to work, and last nearly forever if they are maintained at a correct preload of tension.
A few thoughts.
Firstly AISI 304 or 18/8 or EN 1.4301 is a decent spring steel. One of the common uses for said steel is actually springs. One big reason why it isn't used for bicycle frames is it's price. Stainless steel in general is pretty expensive and even though 304 is fairly common stuff, it's still around the double the price of AISI 4130.
Secondly I went and looked at my wheels. It would seem that the need for seating the elbows is rare to the point of obsolete with today's wheel components. Even the lighter weight hubs I have still exhibit thick enough flanges that there is absolutely no need for setting the elbow or spoke head. With older steel hub shells exhibiting much thinner flanges it may have been more relevant. But even in those situations I'd rather use spoke washers.
Which brings me to my next point. Spoke tension should take care of any spoke path issues in due time. In fact a proper stress relief should also set the spoke paths pretty much immediately. I find it extremely unlikely that the spoke head would be able to spring out against 1200 newtons of DS tension and 700 newtons of NDS tension. We're speaking about 2mm steel wire that you can put permanent bends to with your fingertips. When we tension the wheel, the spokes experience elastic deformation. It just simply isn't possible that you'd have a floating piece of spoke flexing against the pull of the spoke while you're still having the spoke stretch for sometimes more than 1mm or so (for very long at least)
What I suspect is happening is along the lines cyccocommute wrote earlier. If I refine a bit, when a wheel is built and the component selection is done badly (ie. the flanges are thin and spoke elbows are long and no spoke washers are used) and no stress relief is done or it is done to too low of a degree, the elbows may remain "floating" away from the hub flange. Tension will however, in time, pull them into submission against the flange. Depending on how far outside the elbows are, the subsequent bending against the flange would effectively lengthen the spoke which can
drastically drop the spoke tensions of the wheel, which would then logically lead to spoke failures if not addressed.
So to conclude, I believe that these days spoke head or elbow setting is necessary only when there's some inherent incompatibility with the spokes and hub. However the same effect could be achieved with spoke washers. With most modern aluminum hubs and DT Swiss spokes (I've ever only built with DT Swiss) spoke head or elbow setting simply isn't necessary since the spokes adopt a perfect path against the flange when the wheel is properly tensioned.