Predictive AI for high value reuse of reclaimed carbon fibre

The use of CFRPs has grown exponentially over the past 20 years (from ~60 kT in 2008 to 285 kT in 2025), driven by their widespread adoption in aerospace and, more recently, in renewable energy applications such as wind turbines. A direct consequence of this rapid expansion is a substantial increase in associated waste streams. While recycling end-of-life composite components remains a major long-term challenge, in-process manufacturing waste represents an immediate and escalating problem.

The manufacturing waste includes prepreg offcuts which has the same quality as the virgin materials. Prepreg offcuts are typically generated by cutting a large piece of carbon fibre ply. Despite their excellent properties, these irregularly shaped offcuts are routinely discarded, with scrap rates in aerospace manufacturing reaching up to 50%. At present, most of this high-grade material is landfilled or incinerated, resulting in significant environmental and economic losses.

To reduce the environmental impact of CFRP, it is crucial to minimise waste but also utilise this waste stream for a second life application. Despite this potential, industrial uptake has been limited, largely due to uncertainty about the performance of products made from reclaimed materials. To improve the reliability of products made from reclaimed material, this project aims to establish a linkage between material and processing parameters to performance of products made from reclaimed material using advanced machine learning methods. The project aligns with EPSRC strategic theme of building a green future and priority areas such as i) Circular economy, ii) Advanced materials and iii) Artificial intelligence.

You will need to meet the minimum entry requirements for our Engineering Materials PhD programme.

Open to candidates who pay UK/home rate fees. See UKCISA for further information.