Texture characterisation in metals using diffuse ultrasound
File(s)
Author(s)
Png, Melody Ren Xin
Type
Thesis
Abstract
In safety-critical engineering applications, such as aerospace components, the mechanical performance of polycrystalline metals is strongly influenced by crystallographic texture, which reflects the predominant orientation of grains. In Hexagonal Close-Packed (HCP) alloys for example, texture governs how properties such as strength vary with loading direction, giving rise to anisotropic behaviour, non-uniform stress distributions, and increased susceptibility to failure. Reliable texture measurement is therefore essential, particularly for components with complex geometries that cannot be readily characterised using conventional diffraction-based techniques limited to flat, well-prepared samples. Ultrasonic texture inversion offers bulk sensitivity by correlating directional wave velocities with texture, but its wider application is constrained by the difficulty of accurately measuring velocities in non-planar structures. This thesis develops a non-contact ultrasonic methodology based on diffuse field Green’s Function (GF) reconstruction to determine travel times between two points, enabling bulk velocity measurement and crystallographic texture characterisation in complex engineering components.
The feasibility of using GF reconstruction for velocity measurement is demonstrated through experiments and simulations, providing insight into the principles governing velocity extraction from diffuse wavefields. Building on this, a robust methodology is developed that incorporates signal processing to mitigate noise. The approach is validated using different receiver types and geometries, including curved surfaces, achieving a velocity error below 10 m/s and demonstrating its potential for practical, non-destructive measurements in complex components.
Finally, the method is applied to a textured titanium specimen with curved geometry. Directional velocities reconstructed from diffuse fields are input into an established ultrasonic inversion algorithm, and the resulting texture shows strong agreement with Electron Backscatter Diffraction measurements. Quantitative comparison of pole figures yields 92% similarity, demonstrating successful texture characterisation in complex components. These findings highlight the potential of this approach for reliable in situ assessment, supporting manufacturing quality assurance of safety-critical HCP alloys and related materials.
The feasibility of using GF reconstruction for velocity measurement is demonstrated through experiments and simulations, providing insight into the principles governing velocity extraction from diffuse wavefields. Building on this, a robust methodology is developed that incorporates signal processing to mitigate noise. The approach is validated using different receiver types and geometries, including curved surfaces, achieving a velocity error below 10 m/s and demonstrating its potential for practical, non-destructive measurements in complex components.
Finally, the method is applied to a textured titanium specimen with curved geometry. Directional velocities reconstructed from diffuse fields are input into an established ultrasonic inversion algorithm, and the resulting texture shows strong agreement with Electron Backscatter Diffraction measurements. Quantitative comparison of pole figures yields 92% similarity, demonstrating successful texture characterisation in complex components. These findings highlight the potential of this approach for reliable in situ assessment, supporting manufacturing quality assurance of safety-critical HCP alloys and related materials.
Version
Open Access
Date Issued
2025-10-09
Date Awarded
2026-03-01
Copyright Statement
Attribution-NonCommercial 4.0 International Licence (CC BY-NC)
License URL
Advisor
Lowe, Michael J.S.
Lan, Bo
Sponsor
Singapore. Agency for Science, Technology and Research
Publisher Department
Department of Mechanical Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Engineering Doctorate (EngD)