When rotors operate in a supercritical regime, a phenomenon known as rotordynamic instability may occur. This can be brought about by internal friction occurring both in the material and in the structural elements, and it is a cause of self-excited lateral vibrations. Rotors containing an interference fit are particularly prone to such instabilities, so the interest arises, for the formulation of a predictive mathematical model that permits a dynamical study of the system. The main difficulty, however, resides in the characterisation of the internal damping coefficient. Therefore, the aim of this thesis is to develop a strategy to determine this damping coefficient. This can be then substituted into a suitable rotordynamic model to make simulations, predictions and verify the system’s stability. The first step is to select an appropriate predictive model for a built-up rotor, which takes into account the internal damping present at the interference fit. Then, the parameters should be determined in the most accurate way, in particular the internal damping coefficient. This model can then be used to run simulations and predict potential problems, what would be particularly useful in the branches of design and construction of rotors. The Jeffcott Rotor model with linear viscous internal damping is selected as the overall model of the system for two main reasons. First, the friction mechanism in the interference fit can be analysed independently and then easily transformed into an equivalent viscous coefficient. Secondly, it is a linear model, what enables the superposition of different friction mechanism working simultaneously. This characteristic is particularly valuable if the model wants to be expanded to include other factors that have been excluded in this paper. For the interference fit, two models have been proposed, for macro- and micro-slip friction. Both are appropriate descriptions and can be converted into an equivalent viscous coefficient. Emphasis is also made on parameter identification methods, especially for the determination of interface pressure, stiffness and the coefficient of friction. Working with accurate parameters is the key for a successful model, to make precise and useful predictions.

Outgoing