Rotor delevitation analysis of active magnetic bearing systems

Abstract

Active magnetic bearing (AMB) systems present an elegant solution to many problems associated with high-speed machinery design and operation. However, AMBs are deficient in that it allows only for contact-free suspension of rotors. Conventional rolling bearings, named backup bearings, are usually installed between the magnetic bearings and the rotor in order to avoid machine damage in the case of suspension failure. Given the critical function fulfilled by backup bearings with respect to system safety, adequacy assessment of these bearings is vital. However, literature on the subject reveals that no established procedures exist in this regard. This need is addressed in the present study by creating computer simulation models which are capable of predicting backup bearing loads during delevitation. This provides a basis on which stress-related failure safety of the backup bearings may be evaluated. The first simulation model which is developed assumes planar dynamics of the rotor and other components. Development of this model mainly serves to resolve computer implementation issues which are relevant to the intended full model. Following development of the first model, a more detailed model is created by major expansion and modification of the developed code. The detailed model accounts for all major effects present during rotor delevitation. These include a rigid rotor model capable of accounting for three-dimensional unconstrained motion, a model of rotor-bearing contact stiffness and a model of the bearing mount stiffness. In order to ensure accurate computer implementation of the models, both are extensively verified by testing against manually obtainable solutions. Following verification, the models are also subjected to a validation process to ascertain the extent to which the models are representative of real-world behaviour. This is done by comparison of model predictions with experimental observations of a practical AMB system. Many unforeseen problems are encountered during the validation process, hindering detailed validation of the models. Notwithstanding these problems, valuable lessons are learnt which can contribute significantly to improvement of future validation attempts—a greatly lacking aspect of research in the field of rotor delevitation analysis.