Autonomous structural health monitoring technique for interplanetary drilling applications using laser Doppler velocimeters

The research work presented in this thesis is devoted to the formulation and field testing of a dynamics-based structural health monitoring system for an interplanetary subsurface exploration drill system. Structural health monitoring is the process of detecting damage or other types of defects in structural and mechanical systems that have the potential to adversely affect the current or future performance of these systems. Interplanetary exploration missions, specifically to Mars, involve operations to search for water and other signs of extant or past life. Such missions require advanced robotic systems that are more susceptible to structural and mechanical failures, which motivates a need for structural health monitoring techniques relevant to interplanetary exploration systems.;Strict design requirements for interplanetary exploration missions create unique research problems and challenges compared with structural health monitoring procedures and techniques developed to date. These challenges include implementing sensors and devices that will not interfere with the drilling operation, producing "real-time" diagnostics of the drilling condition, and developing an automation procedure for complete autonomous operations.;The first research area involves modal analysis experiments to understand the dynamic characteristics of interplanetary drill structural systems in operation. These experiments also validate the use of Laser Doppler Velocimeter sensors in real-time structural health monitoring and prove the drill motor system adequately excites the drill for dynamic measurements and modal analysis while the drill is in operation. The second research area involves the development of modal analysis procedures for rotating structures using a Chebyshev signal filter to remove harmonic component and other noise from the rotating drill signal. This filter is necessary to accurately analyze the condition of the rotating drill auger tube while in operation. The third research area involves the development of structural dynamic models to represent the drill system under nominal and expected drilling fault conditions. These models are compared with the modal analysis experimental results and provide theoretical means to analyze the drilling operation and predict fault conditions. The fourth research area involves the formulation of a complete autonomous system to collect and perform the dynamic analysis of the drill signal, identify fault-diagnostic results, and relay these results to the drill Executive computer. The formulated system includes the signal filter, trained Neural Networks, and an automation procedure. Trained Neural Networks are implemented to provide a rapid-response method of relating and comparing the current drill signal with the fault-based structural dynamic models developed in this thesis. Lastly, an automation procedure, and the corresponding software, is developed to interface the measurement equipment, signal filter, Neural Networks, and drill Executive computer to provide a complete hands-off operation of the structural health monitoring system. The fifth research area involves field testing and validation of the developed structural health monitoring system through three field campaigns. Two of these field campaigns were completed at a Mars-analog site in the Canadian Arctic.;The automated dynamics-based structural health monitoring technique developed in this thesis presents advanced research accomplishments leading to real-time, autonomous structural health monitoring, and it has been successfully demonstrated on an operating dynamic system. Other major contributions of this thesis work include the formulation and demonstration of real-time, autonomous structural health monitoring in rotating structures using Laser Doppler Velocimeter sensors.