The goal of this chapter is to answer the question "How close is good enough?" Ideally, the alignment should be perfect. Ideally, we would like sunny weather and perfect governments, too. We live in a real and imperfect world. Perfect alignment is not achievable, even if we knew what it was and we could measure it when we arrived there. There will always be some remaining residual misalignment. In some cases it is even desirable to have some misalignment. For every machine system, there is a condition of alignment that is not perfect, is economically achievable in the allotted time, and will not cause premature failure. This is the condition that an efficient alignment mechanic will promptly head for and stop when he/she gets there, and will go no further.

Dial-indicator readings alone are poor criteria for acceptance because they do not account for bar sag, thermal growth, and the condition of the coupling. They also limit the aligner to a specific measurement tool and do not give him/her the freedom to use different measuring instruments, like laser, optics, or electronic indicators. An appropriate alignment specification should specify only the final physical condition and say nothing about the tools or the methods to arrive there. The appendix contains a generic alignment specification that is complete. This chapter will elaborate on the contents of that specification. It would be useful at this time to examine a poor alignment specification for comparison. Figure 9.1 is a copy from the installation...