A newly developed artificial intelligence model promises to revolutionise the way designers and engineers develop high-performance composite materials, significantly reducing the need for time-consuming physical testing and complex simulations. The innovation, developed at the University of Gothenburg, is particularly relevant for product designers, automotive designers, and architects working with advanced materials in applications demanding lightweight strength, such as in transportation, sports equipment, construction components, or even experimental textile forms.

The Challenge of Designing Composites

Composite materials, such as carbon fibre-reinforced polymers, are widely used for their ability to combine lightweight properties with high strength and durability. These materials are particularly popular in aerospace, automotive, and sporting goods design, but also increasingly appear in architectural components and experimental product design.

When woven into textile-like structures, these composites offer even greater versatility. However, predicting how such woven composites will behave under stress remains a complex challenge. Traditional approaches rely heavily on physical prototypes and computational simulations, which are both expensive and time-intensive.

AI-Powered Modelling for Faster Development

To address this bottleneck, Ehsan Ghane, a PhD researcher at the University of Gothenburg, has developed a novel AI-based model that streamlines the process of evaluating the mechanical behaviour of woven composites. Unlike conventional neural networks that require vast training data and struggle with extrapolation, Ghane’s generalised model integrates material physics directly into the AI algorithm. This not only reduces the amount of data required but also enables predictions beyond the trained scenarios.

By incorporating existing data from both simulations and empirical testing of individual materials, the AI can reliably predict the durability and deformation behaviour of composite fabrics. This breakthrough could enable designers to iterate faster, reducing both material waste and development time—a clear step forward in sustainable material innovation.

Implications for Sustainable and Circular Design

The AI model aligns with broader goals in sustainable design by promoting efficient material use and supporting the development of longer-lasting components. For professionals working in automotive and product design, this tool could enable the creation of optimised lightweight parts with minimal environmental impact. In architecture, its predictive capabilities could support the use of novel composite panels and structural elements with confidence in their long-term performance.

Designers experimenting with biobased or recycled fibres in woven composites may also benefit from this AI-driven approach, as it allows for more agile testing of unconventional material blends without the need for costly physical trials.

As the demand for smart, circular, and environmentally conscious materials grows, innovations like this AI model are becoming essential tools in the designer’s toolkit.

Source: University of Gothenburg
Photo: MaterialDistrict (Flaxco)