Experimental validation of multifidelity digital twins of electromechanical actuators for prognostics.

Over the years, electromechanical actuators have revolutionized the aerospace industry, offering numerous benefits over traditional hydraulic ones. Their introduction allows the complete removal of the hydraulic system from the flight control lines, reducing weight and also offering greater control and precision. Nowadays they are mainly used for secondary flight commands but, with the development of More-Electric aircraft, they will become a key component. On the other hand, with the increasing use of electromechanical actuators, it's crucial to have a robust and reliable failure identification system in place. A malfunction in such a system could have serious consequences, potentially leading to flight disruptions or even accidents. Thanks to the evolution of onboard electronics, prognostics can now be performed by proper mathematical models that can accurately reproduce the behaviour of the component under consideration, allowing real-time detection of the failure. The thesis is set in this complex scenario with the aim of validating two different mathematical models built in Matlab-Simulink that can predict the behaviour of an electromechanical actuator. The algorithms, a High-Fidelity Model, and a Low-Fidelity Model differ in terms of the detail on which every element is modelled and the computational cost of the simulation. The validation process is carried out through experimental tests conducted on an actuator test bench located at the “Politecnico di Torino”. The tests are performed with different values of backlash and resistive load to simulate various operating conditions of an electromechanical actuator. The results of the tests demonstrate the accuracy of both models in reproducing the experimental values, confirming their reliability. These models represent the first step for future prognostic models of electromechanical actuators. In addition to their practical applications, these models also have the potential to contribute to the broader understanding of electromechanical actuators and their behaviour. Overall, they can provide a basis for further research and development, helping to advance the state of the art in this field.