Diaphragm-type couplings are high misalignment torque and speed transfer components used in aircrafts. Crack
development in such couplings, or in the drive train in general, can lead to component failure that can bring down an
aircraft. Real time detection of crack formation and growth is important to prevent such catastrophic failures. However,
there is no single Nondestructive Monitoring method available that is capable of assessing the early stages of crack
growth in such components. While vibration based damage identification techniques are used, they cannot detect cracks
until they reach a considerable size, which makes detection of the onset of cracking extremely difficult. Acoustic
Emission (AE) can detect and monitor early stage crack growth, however excessive background noise can mask acoustic
emissions produced by crack initiation. Fusion of the two mentioned techniques can increase the accuracy of
measurement and minimize false alarms. However, a monitoring system combining both techniques could prove too
large and heavy for the already restricted space available in aircrafts.
In the present work, we will present a newly developed integrated Acoustic Emission/Vibration (AE/VIB) combined
sensor which can operate in the temperature range of -55°F to 257°F and in high EMI environment. This robust AE/VIB
sensor has a frequency range of 5 Hz-2 kHz for the vibration component and a range of 200-400 kHz for the acoustic
emission component. The sensor weight is comparable to accelerometers currently used in flying aircraft. Traditional
signal processing approaches are not effective due to high signal attenuation and strong background noise conditions,
commonly found in aircraft drive train systems. As an alternative, we will introduce a new Supervised Pattern
Recognition (SPR) methodology that allows for simultaneous processing of the signals detected by the AE/VIB sensor
and their classification in near-real time, even in these adverse conditions. Finally, we will discuss the architecture
developed to produce a fully autonomous monitoring tool based on the fusion of the AE and Vibration techniques.
|