UVM Theses and Dissertations
Format:
Print
Author:
Menon, Praneet
Dept./Program:
Mechanical Engineering
Year:
2013
Degree:
MS
Abstract:
Helicopters are used widely for various civilian and military missions. These missions vary from medical and rescue services to landing and extracting soldiers from enemy territories. No matter what the mission profile, aircraft readiness and reliability are paramount. The ability to take-off when necessary and ensurethe safety of the passengers and aircraft also has huge cost associations.
To ensure a high level of flight safety and mission readiness, civilian and military organizations that utilize helicopters have followed maintenance practices put forth by OEMs and third party maintenance management organizations. These maintenance practices revolve around the various failure modes of helicopters. One of the main failure modes is the failure of the drive train to transmit power from the engine(s) to the main rotor and tail rotor. Failure to transmit power to the main rotor causes the aircraft to lose lift, while power loss to the tail rotor causes uncontrollable yaw. In both cases, loss of life and aircraft are highly probable.
The focus of this research was drive train diagnostics that can be added to a Condition-Based Monitoring (CBM) system to detect and predict failure to drive train components. Algorithms were developed that extracted features from the data for state assessment and ascertain the health ofa given shaft, gear or bearing. The effectiveness of the Condition Indicators (CIs), generated by the algorithms, was tested by performing a Pooled Sample t-test. This test showed that there were a number of CIs that had a positive response to the seeded faults. These CIs were then used to compute a Health Indicator value for each component in question. Once again, the t-test was used to determine whether there was significant separation between the nominal data and the faulted data. The results showed that the HIs were effective in diagnosing faulty components for all shafts and gears. However, for bearings, a fault on the ball was not picked up. Possible causes for this missed detection are also discussed in the thesis.
To ensure a high level of flight safety and mission readiness, civilian and military organizations that utilize helicopters have followed maintenance practices put forth by OEMs and third party maintenance management organizations. These maintenance practices revolve around the various failure modes of helicopters. One of the main failure modes is the failure of the drive train to transmit power from the engine(s) to the main rotor and tail rotor. Failure to transmit power to the main rotor causes the aircraft to lose lift, while power loss to the tail rotor causes uncontrollable yaw. In both cases, loss of life and aircraft are highly probable.
The focus of this research was drive train diagnostics that can be added to a Condition-Based Monitoring (CBM) system to detect and predict failure to drive train components. Algorithms were developed that extracted features from the data for state assessment and ascertain the health ofa given shaft, gear or bearing. The effectiveness of the Condition Indicators (CIs), generated by the algorithms, was tested by performing a Pooled Sample t-test. This test showed that there were a number of CIs that had a positive response to the seeded faults. These CIs were then used to compute a Health Indicator value for each component in question. Once again, the t-test was used to determine whether there was significant separation between the nominal data and the faulted data. The results showed that the HIs were effective in diagnosing faulty components for all shafts and gears. However, for bearings, a fault on the ball was not picked up. Possible causes for this missed detection are also discussed in the thesis.