ENHANCING THE RELIABILITY OF TURBOMACHINE CONDITION MONITORING EXPLOITING UNCONVENTIONAL RUGGED SENSORS

This work is focused on increasing the reliability of measurement systems for turbomachine condition monitoring. The motivation stems from the fact that the sensing systems that are widely used today to determine the behavior and health status of turbomachinery use advanced methods such as digital twins. A digital twin is a software representation of a physical device which relies on the data coming from the onboard sensors and can be used to increase the reliability of the device and to perform predictive maintenance to reduce the risk of costly failures. In this framework, the sensors and the measurement systems must offer the highest reliability. However, modern measurement systems rely on complex devices and electronics which do not get along with the harsh environment present in a turbomachine. In this work, the reliability enhancement of two of the main measurement systems for turbomachine condition monitoring will be addressed. The first analyzed measurement system is related to combustion monitoring with particular reference to the detection of combustion chamber pressure oscillation (pulsations). Typical pulsation monitoring techniques exploit piezoelectric sensors, which are intrinsically robust devices, but suffer from problems related to wiring robustness and, in general to the reliability of the electrical connections. A novel measurement system for pressure oscillations in the combustion chamber, which exploits pre-existing parts of the combustor, has been studied, realized and tested in the field. In particular, the developed system is based on the measurement of the ion density in the combustion chamber due to chemo-ionization, through the ion current gathered by the electrode of the spare spark plug present in the combustor. In this context, at first, a novel model of the ion current sensor and of the measurement circuit was developed and used to predict and analyze the measurement technique performance and to properly design the overall monitoring system. The model was validated through experiments in laboratory. Finally, the proposed system was tested in field in an industrial test rig. The discussion continues with the improvement of turbomachine rotor dynamics monitoring systems, with specific focus on torsional vibration detection. Measurement systems for torsional vibration monitoring usually exploit capacitive sensing. Sensors 2 are mounted in orifices in the stator case, which weaken the whole mechanical structure. Moreover, also in this case, the wiring is a criticality. In this work, the application to torsional vibration monitoring of preexisting variable reluctance sensors normally used to monitor the turbomachine rotation speed has been investigated. A novel system based on a special toothed wheel and on a tailored front-end circuit was designed, realized and laboratory tested. Also, in this case a model of the sensor and of the measurement circuit was developed and used to analyze the potential of the proposed technique. Summarizing, the aim of this work was to increase the reliability of two measurement systems crucial for turbomachine condition monitoring, exploiting pre-existing rugged components of the turbomachine itself. Two setups for the ion current measurement and for the torsional vibration measurement were tested in the laboratory obtaining satisfactory results: Moreover, the ion current measurement system was tested in the field giving promising results when compared with the performance of traditional pressure sensors.