PhD Oral Exam - Koteswara Rao Alla, Electrical and Computer 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

Induction motors (IMs) are commonly used motors because they are very rugged, require low maintenance, and are mechanically simple. The IM is the workhorse as a traction motor in rolling stock applications due to the inherent advantages in terms of cost, controllability, reliability, and maximum speed. An induction machine has three common faults: stator inter-turn faults, rotor broken bars, and bearing faults. These faults can affect the industry's productivity and cause outages in the transportation industry. For example, an induction motor failure in a freight locomotive can cause stalling and subsequent line congestion. Similarly, a wound rotor induction machine (WRIM) is used as a doubly fed induction generator (DFIG), the most commonly used topology for wind power applications. The most common electrical faults in DFIG are bearing faults and inter-turn winding faults in the stator and rotor. If these faults are not detected and rectified in time, they can proliferate quickly and become catastrophic, leading to unexpected outages and contributing to high operating and maintenance costs. Hence, it is essential to study the behavior of WRIM’s in the event of these faults and what corrective action can be taken in the event of a fault. Also, it is necessary to investigate the faulted IM’s behavior with accurate models and replace expensive test benches and equipment with safer and economical test procedures.

There are a lot of practical difficulties in creating the fault with machines. Even if the fault is implemented, full voltage cannot always be applied, and motors with higher loads cannot be tested since it may damage the machine faster. However, a virtual machine with power electronic converters can be tested for different fault conditions using emulation. Power-hardware-in-loop (PHIL) based emulation is increasingly used as an economical test bench for testing the drive system. With PHIL-based emulation, the electrical machine is emulated in a laboratory environment with real-time power flow using power electronics converters. With PHIL-based emulation, the faulted machine behavior can be studied with direct online start or machines fed from variable frequency drives (VFD) without damaging the motor or VFD. With emulation, the risk, time, and cost associated with a physical machine subjected to faults can be reduced. Also, the emulation test setup is designed to return the power flow from the inverter to the mains supply. This requires an active front-end converter connected to the DC link of the emulator converter. Hence, only power losses are drawn from the power supply. Thus, the laboratory power supply requirements are reduced.

The machine's mathematical model must be very accurate for better emulation accuracy. So, there is a need to develop detailed mathematical models which can account for the nonlinearities in the machine. The machine parameters vary when temperature changes. For an accurate emulation, it is necessary to consider the effect of temperature on the machine's behavior. For this, the thermal behavior of the machine is crucial. Lumped parameter thermal network (LPTN) models can be used for emulation purposes since finite element models (FEM) require long execution times which are not suitable for real-time emulation though it has higher accuracy.

This thesis investigates PHIL emulation of stator inter-turn faults and rotor inter-turn faults. Detailed and improved mathematical models for these faults are developed. An alternative mathematical model has been proposed for stator inter-turn faults, i.e., the Voltage-behind-reactance (VBR) model, which has some advantages compared to the conventional modeling approach of the voltage-in & current-out (VICO) model. Fault conditions are emulated with experimental hardware in real time. Emulation results are validated with a practical machine subjected to the faults. Thermal analysis of an induction machine with stator inter-turn fault is carried out. An analytical model is developed to estimate the temperature rise caused by the inter-turn fault, and the model is validated with experimental results.