The newly developed wavelet transform enables one to look at the evolution
in time of a signal's frequency content. This property makes it very suitab
le for the detection of vibration transients generated by developing locali
sed faults in gear trains. Recent applications of the wavelet technique hav
e focused on the use of the amplitude map only. In this study, both the amp
litude and phase maps of the wavelet transform are used in conjunction to a
ssess the condition of an instrumented gear test rig. The phase wavelet map
is found to display distinctive features in the presence of a cracked toot
h. Another key finding is that the amplitude wavelet map of the residual vi
bration signal offers a better indicator to the presence of faults than the
map of the actual signal. A new polar representation is also introduced th
at makes pinpointing the precise location of a crack an easy task. The case
s studied include a healthy gear, one with a simulated fatigue crack, and o
ne with a simulated chipped tooth. (C) 1999 Academic Press.