Applications of time-frequency analysis for condition based maintenance of military helicopter power train

The objective of condition based maintenance (CBM) is to reduce the maintenance cost in time-based maintenance by performing timely repairs only upon evidence of need. However, the attractive feature of CBM requires an accurate assess tool of components conditions that indicates the appropriative maintenance to be performed. If this technical goal is achieved, we can significantly increase the fleet's operational capability with reduced maintenance cost. This thesis develop a new time-frequency techniques to assess health conditions of tail rotor drive shaft in military helicopter by use of vibration signals, collected under the different type of conditions in order to quantitatively compare with the baseline operation condition.;The classical time-frequency analysis represent time- and frequency-localized energy distribution. However, it is not an easy task to analyze correlated signals such as transient input and output signal pairs, that are simultaneously collected. The CBM center at The University of South Carolina has an AH-64 Helicopter tail rotor drive-shaft apparatus. The signals discussed in this thesis includes vibration measured by the accelerometers, temperature measured via thermocouples, and speed and torque measurements during the operational running at the experimental setup of CBM center.;In this thesis, a new concept of non-parametric detection and classification of the signals is proposed using the mutual information measures in the time-frequency domain. By use of the proposed method, different health status of the tail rotor drive shaft are represented in terms of in-phase and quadrature information. The experimental results exhibit negative increases of in-phase information in the unbalanced shaft case, and negative increases of quadrature information in the misaligned shaft case. In addition, the misaligned-unbalanced case which has two conditions at the same time, indicates one can properly identify its signatures in terms of in-phase and quadrature information measure.