Runout is caused by a bent shaft, or by eccentricity
of the coupling. Eccentricity literally means being off-center. It can occur if the coupling is not bored
in its physical center, or if the bore is bigger than the shaft, and the set screws
pull the coupling out of center. When measuring runout, the dial indicator
is mounted to a fixed point in space. If the coupling is disconnected, attach the
dial to the shaft OPPOSITE of the one you want to check. Or, if the coupling is connected, you can
use a mag bracket to mount the dial indicator to the base of the machines, or any other
fixed point in space. We recommend that you first check for runout
at the hub, or rim, of the coupling, because there you will pick up both eccentricity AND
bent shaft. Some people will also measure runout in the
axial direction, which, in addition, would tell you if the coupling bore is skewed. Before taking the readings, make sure the
dial is mounted perpendicular to the shaft. Set the plunger to half its travel range to
assure both negative and positive capacity. Once the dial is affixed, rotate the shafts. Make sure you rotate in only one direction. Find the greatest distance from zero in either
positive or negative direction, stop rotating the shafts and reset the dial to ZERO. Now, continue rotating the shafts and watch
the dial for the maximum amount of change, which equals the total amount of runout. If the runout is within acceptable limits
at the hub, 2 thou or less, then the coupling must be bored properly enough AND the shaft
is reasonably straight and within acceptable limits. In this case, the dial is indicating excessive
runout of 22 thou. When excessive runout is observed we next
need to determine if it is being caused by a bent shaft or by an eccentric coupling,
or both. We achieve this by repeating the same process
at the shaft. You'll want to take the measurement as close
to the coupling as possible. In this case there is 21 thou of runout at
the shaft, accounting for most of the 22 thou of runout we observed at the coupling. Therefore, the bent shaft is the problem. If our measurement at the shaft had been close
to zero the problem would have been eccentricity of the coupling. Of course, we could observe a situation where
both conditions occur, a bent shaft and an eccentric coupling. If excessive runout is present it should be
corrected before attempting precision shaft alignment.
