PhD Oral Exam - Amir Kermanizadeh, Electrical Engineering

When studying for a doctoral degree (PhD), candidates submit a thesis that provides a critical review of the current state of knowledge of the thesis subject as well as the student’s own contributions to the subject. The distinguishing criterion of doctoral graduate research is a significant and original contribution to knowledge.

Once accepted, the candidate presents the thesis orally. This oral exam is open to the public.

Abstract

This thesis investigates the thermal and vibration behaviour of electrical machines under healthy and faulted operating conditions, with emphasis on induction machines and permanent magnet synchronous machines. The thermal part develops lumped-parameter thermal networks and finite-element models for estimating component temperatures and representing the principal heat-transfer paths within squirrel-cage and wound-rotor induction machines. Deep neural networks are trained using physics-based thermal data, and transfer learning is applied to adapt temperature-prediction models to machines with different ratings and topologies using limited target-machine measurements. Fault-aware thermal models are formulated for stator inter-turn short circuits and air-gap eccentricity by including localized electrical losses, temperature-dependent parameters, and fault-related mechanical losses. Thermal modelling of conventional and soft-magnetic-composite PMSMs is also examined, together with the thermal response of a traction PMSM under controlled inter-turn faults. The vibration part examines how eccentricity and winding faults alter electromagnetic forces and the vibration transmitted to the machine housing. Three-axis accelerometer and current measurements are processed using frequency-domain, mechanical-order, and band-energy features. Supervised classifiers and ensemble learning are considered for fault detection and severity estimation under different speed and load conditions. The thesis also presents the electromagnetic, thermal, and mechanical design of an experimental PMSM developed for controlled fault studies. The machine incorporates tapped stator windings, adjustable mechanisms for static, dynamic, mixed, uniform, and non-uniform eccentricity, and a removable magnet arrangement for broken-magnet investigations. Together, these studies provide an integrated framework for thermal modelling, vibration analysis, data-driven monitoring, and repeatable laboratory investigation of electrical-machine faults across several controlled operating conditions.