Originally Posted by
une_vitesse
....that bike's bb was cobbled together using campy cups and a shimano DA spindle. i installed it, i seem to remember that it spun smoothly without the cranks and had no play with the cranks on. still doesn't have play.
The issue is that while it feels good in the stand it's not under load in the stand. If it helps to understand this it's much like a poorly tensioned wheel. If someone that doesn't understand the dynamics of a spoked wheel is holding a low tensioned wheel in their hand. To them it feels just fine with the spokes all quite tight and "twang'y"
. Yet we know that inadequite tension will result in the spokes on the lower side going too loose and failing. But our wheel newbie would hold the wheel up and think "WTH, they are all still tight. What went wrong?". What went wrong is that they didn't take into account the wheel is composed of a lot of elastic elements and that as the load is put on that system it will react by very slightly altering the centering of the hub to the rim and that the rim will very slightly alter it's shape. The spokes don't stretch much as we know but it's enough to see the spoke tensions alter radically as the wheel comes under load and that slight movement is enough to lower the tension of the spokes on the bottom of the wheel. And as we know too low a tension in this case allowed the lower spokes to "take a break"... literally as well as figuratively.
Bearing shell flexing under load that causes the bearing ball preload to go away and allow the cup and cone to move out of concentricity will work the balls in a manner very analagous to that poorly tensioned wheel.
We see another example of this in setting the bearings on a QR wheel. You've likely already done these adjustments where you either clamp the QR skewer in the axle for the adjustment or play with getting just the right amount of play so that when the skewer is clamped it just goes away. The skewer in this case is compressing the axle that last thou or two that closes and preloads the wheel bearings. This is the most obvious example of what I mean when I say that steel is actually elastic. It also shows why we need preload in each bearing to allow for this elastic movement of the metal to occur without going so far that the balls lose full contact in any of the bearings.