This thesis presents novel acoustic metamaterial approaches for enhancing fault sensing in machinery condition monitoring through advanced acoustic wave manipulation. The research systematically develops and validates three distinct metamaterial configurations through comprehensive structural design, analytical modelling, finite element analysis, and experimental validation, each addressing specific challenges in industrial fault diagnosis while progressively building upon previous findings to achieve enhanced sensing capabilities. Initially, a space-coiling metamaterial (SCM) was developed, featuring labyrinthine channels with systematically increasing path lengths. This design effectively achieved subwavelength-scale enhancement (λ/14) through high-Q resonance, demonstrating up to 40% improvement in fault detection SNR. The SCM structure maintained strong amplification characteristics even with reduced symmetrical units, enabling device miniaturization while preserving acoustic enhancement capabilities. Building upon the identified limitations of SCMs, a graded acoustic metamaterial (GAM) was engineered, incorporating a series of gradually increasing diameter slabs to create a high gradient refractive index profile. This configuration demonstrated superior broadband wave manipulation, achieving pressure gains up to 42 times across frequencies from 500-2500 Hz. Experimental validation confirmed its effectiveness in detecting both blade and bearing faults, with up to 45% improvement in signal-to-noise ratios compared to conventional measurements. Subsequently, an innovative hybrid design integrating space-coiling and graded acoustic metamaterial principles (SC-GAM) was developed. This integration successfully combined the advantages of both approaches while mitigating their individual limitations. While maintaining compact dimensions comparable to the GAM, the SC-GAM achieved enhanced low-frequency operation and improved frequency selectivity. Experimental results demonstrated superior performance in extracting both bearing and gear fault signatures, with SNR improvements of up to 50% compared to baseline measurements. In a crucial final investigation, comprehensive directivity analysis of all three designs revealed that the GAM configuration achieved the highest directional sensitivity (DI reaching 15.2 dB), enabling fault localization without complex sensor arrays. The research included extensive finite element simulations and experimental validations, including challenging scenarios such as drone-based acoustic fault detection, establishing a robust framework for acoustic metamaterial optimization in broader machinery diagnostics applications. This research advances the field by introducing passive, cost-effective solutions for enhanced acoustic-based condition monitoring, demonstrating significant improvements in signal enhancement, frequency selectivity, and spatial resolution. The findings provide new pathways for implementing metamaterial-based sensing in industrial applications, particularly beneficial for early-stage fault detection in rotating machinery.
| Date of Award | 9 Jan 2026 |
|---|
| Original language | English |
|---|
| Sponsors | Beijing Institute of Technology |
|---|
| Supervisor | Fengshou Gu (Main Supervisor) & Helen Miao (Co-Supervisor) |
|---|