Samuel D wrote:A swing in length of 0.2 mm is equivalent to a 72° turn of the M6x1 preload bolt. That’s huge.
it is and it isn't. In Nick's example the length change is a slight overestimate, but the gist of the argument is on the face of it fair.
However it should be noted that the bearings are sensitive to load, not free head tube length, in other words the stiffness of the assembly is what changes the preload in the bearings when the free length changes.
If the bearings are set at ~20C then the possible temperature excursion is +20C or -30C. This corresponds to free length changes in the head tube of ~ +0.072mm or ~ -0.108mm.
The stiffness of the head tube vs the steerer tube will likely vary from one frame design to another, but it is likely that the changes in free length are comparable with those experienced when the preload is applied. If so, in hot weather the preload might increase from (say) 200kg to 400kg and in cold weather you might lose the preload entirely, and experience movements in the bearing of ~0.1mm or so. Neither thing will give the headset bearings the kiss of death immediately, but if the preload is wrong (too high or too low) to start with, it might help cause damage.
FWIW I have long thought that it could be a good idea to include a wave washer in an Ahead headset build, just above the wedge; something that squashes flat @ ~400kgf load and 0.5mm travel would be about right I think. Setting the headset bearings would then become a matter of feeling the torque ramp up as the washer was squashed flat, and then backing off 1/4 turn. Preload wouldn't vary with settling/temperature excursion so much. A potential downside is that the wedge might move around more than it does anyway; it might be that the wedge could benefit from having its own sleeve against the steerer, such that the wedge always moved against the sleeve not the steerer. CF steerers commonly suffer worryingly in this respect and this approach would be of benefit anyway.