Skip to content

This paper presents a comprehensive fault tolerance strategy for active magnetic bearing (AMB) systems in critical applications. AMBs offer numerous advantages, including reduced wear, minimal maintenance, and improved vibration performance; however, these systems must be carefully designed to manage fault conditions that could compromise stability or cause catastrophic machine failure. The primary objective of this study is to define system-level failure modes, establish a clear methodology for fault classification, and evaluate redundancy options for AMB components to enhance overall system reliability. The paper introduces a fault tolerance framework that classifies failure modes into four categories: operation outside allowable limits, performance reduction, unintended machine downtime, and catastrophic damage. These categories are mapped to specific AMB failure scenarios. The methodology emphasizes early fault detection, appropriate hardware redundancy, and the implementation of software control strategies that enable drop prevention and drop recovery. A systematic evaluation of hardware fault tolerance options is provided for actuators, position sensors, magnetic bearing controllers (MBCs), and touchdown bearings. Redundancy configurations such as 1+1, N+1, and active-active load-sharing topologies are assessed for their feasibility and system impact. The paper also details drop detection, fault prevention, and drop recovery strategies, supported by adaptive control algorithms, synchronous force rejection, and bias current management. Field-tested drop recovery software, previously demonstrated on a rotary-disc atomizer, is reviewed as a basis for future applications. Results show that integrating fault tolerance measures in both hardware and software significantly improves AMB system resilience. The conclusions support continued development and validation of fault-tolerant AMB architectures to meet the reliability demands of space, defense, and mission critical rotating machinery.

Author: | Published:
Booktitle: Proceedings of ISMB19