Project Details

Layman's description

Have you ever seen a wind turbine and wondered about the material of its blades?  Well, most of wind turbine blades are made up of glass fibre or a combination of glass and carbon fibre. These wind turbines, as huge as commercial aeroplanes, require nearly a year in transportation and installation for energy production. Therefore, the selection of materials with high strength and low weight is crucial to ensure longevity and ease of transportation. Both glass and carbon fibres exhibit favourable attributes in terms of strength and weight. However, carbon fibres are lighter and stronger compared to glass fibre which directs us to research more about carbon fibre capabilities. The improvement of carbon-based materials requires a comprehensive characterization of their properties.

This PhD project is a part of a Marie Curie Project– RELIANCE which aims to develop software and instrumentation for real-time characterization of material properties from micro- to nano-scale. The classical pipeline to characterise the material properties consists of scanning, segmentation, modeling, and quantification as shown in the attached figure (Overview of a classical pipeline for material characterisation). However, executing this entire pipeline and undergoing labour-intensive visual inspection consumes a substantial amount of time. This PhD focuses on reducing the processing time of characterization by fitting a machine learning-based geometric model in place of the segmentation & modeling process shown in the figure. To develop the geometric model, we will propose a unified framework that utilises a fibre tracker algorithm and structure tensor method. This technique helps to do quantification and measure the strength of the material.

The algorithms and methodologies developed in this PhD project will be utilised in other work packages of the RELIANCE project, where developed method efficiency will be tested and validated.

Effective start/end date15/09/202314/09/2026


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.