Date of Completion

12-8-2016

Embargo Period

12-7-2016

Keywords

gearbox vibration, system-level modeling, harmonic wavelet transform, feature extraction, fault detection

Major Advisor

Dr. Jiong Tang

Associate Advisor

Dr. Robert Gao

Associate Advisor

Dr. Brice Cassenti

Associate Advisor

Dr. Chengyu Cao

Associate Advisor

Dr. Xu Chen

Field of Study

Mechanical Engineering

Degree

Doctor of Philosophy

Open Access

Open Access

Abstract

This research presents a systematic approach to health monitoring using dynamic gearbox models (DGM) and the harmonic wavelet transforms (HWT) for vibration response analysis. A comprehensive DGM is developed, the model parameters are identified through correlated numerical and experimental investigations, and HWT analysis is performed to illustrate the fault detection and diagnosis procedure and capability of this approach. The model fidelity is validated first by spectrum analysis, using constant speed experimental data, and secondly by HWT analysis, using non-stationary experimental data. The comparison confirms that both the frequency content and the predicted, relative response magnitudes match with physical measurements. Model prediction and experimental data are compared for healthy gear operation and seeded gear faults including a pinion with a missing tooth, tooth root crack, tooth spall and varying tooth chip severities, demonstrating that fault type and severity are distinguishable. The research shows fault modeling in combination with HWT data analysis is able to identify fault types, evaluate fault relative severity, and greatly reduce pattern recognition library development. This approach can facilitate successful fault detection, diagnosis and prognosis for gearbox systems, providing a physically meaningful connection of fault indicators to the actual fault patterns thus paving the way to real-time condition monitoring.

Download

Share

COinS