Date of Award


Document Type

Campus Access Thesis


Mechanical Engineering

First Advisor

Abdel-Moez E Bayoumi


The CBM research at University of South Carolina is targeted at achieving some of the CBM objectives for rotorcrafts in collaboration with South Carolina Army National Guard. The current research study is motivated by the endeavor to enhance condition monitoring objectives by evaluating the condition indicators (CI) that have been implemented currently on the aircraft in operation.

For the purpose of this research three tail rotor gearbox have been tested to simulate severe lubricant starved conditions. The three test articles have failed in three different modes. The TGB article-1 had severe damage of the input gear teeth. For the second TGB article the testing was terminated when an input gear tooth broke. The third TGB article did not suffer significant damage to its gear teeth, but the testing was aborted due to an increase in temperature. For all the three articles, a significant increase in gear mesh second harmonics was observed but the condition indicators have not been successful in providing a warning well in advance. The only condition indicator that has triggered caution alarm was TGB lateral bearing energy CI. From the evaluation of condition indicators it is found that only the DA1 CI has shown a very sharp increase in its value. The FM0 and FM4 CI have also shown noticeable fluctuations but the variation was not severe to create any alarm. This suggests that if the existing CIs do not perform up to the expectations then an alternate CI that is either better in anticipating failure or produce similar results with relatively lower computational expense is beneficial.

In the current research, a new condition indicator that is relatively simpler compared to the FM4, ER, SLF, SI condition indicators and has the ability to offset some of the limitations of Discrete Fourier transform (DFT) is proposed. The proposed CI is defined as Asynchronous to Synchronous Energy ratio in Time domain (ASET). As the name suggests, the ASET for a component rotating at a particular speed is obtained by dividing the total energy in the raw time domain signal with the energy in the time synchronous averaged signal of the component with synchronous frequency equal to the component's frequency. The ASET CI has ability to identify modulation of synchronous frequencies, and change in gear mesh amplitudes. Since, the ASET CI does not involve any information about the spectrum of the signal, the need for DFT is eliminated which reduces computational burden as well as provides hope to detect the signals with both transient and periodic signal content. The proposed CI has been analyzed with the vibration data for the TGB article-2, TGB article-3 and the IGB article that was installed during the testing of TGB article-2 and TGB article-3. The plots obtained after computation of ASET for the test articles showed variation as hypothesized. Furthermore, ASET indicated existence of a functional relationship between FM0 and FM4 CIs.

This thesis presents the results of the experiment, evaluation of the existing CIs, the definition of the new CI and its behavior when fed with the data from the experiments conducted on test articles. The ASET CI can be a supplementary diagnostic algorithm that provides information about sidebands, gear mesh amplitudes from just the time-domain data. The proposed CI reacts to multiple failure modes with lower computational burden unlike some of the currently implemented CIs and hence has potential to substitute some of the existing CIs.


© 2011, Praveen Kumar Nooli