Digital wind blade manufacture

TPI Composites is working with the US National Science Foundation (NSF) to design a composite wind blade manufacturing method based on ‘digital twin’ machine learning.

This involves using a computer program that takes real world data to create simulations that can predict how a product or process will perform.

According to TPI, this could reduce defects and enable more efficient production of wind blades, reaching predictive accuracy of more than 95% with 100 times faster computation when compared to physics-based manufacturing model.

The technology is being developed at WindSTAR, a NSF-funded research center, along with students and professors from the University of Texas at Dallas and engineers from plastics companies Olin Epoxy and Westlake Epoxy. Researchers are focusing on a vacuum assisted resin infusion molding (VARIM) process, TPI said.

“The primary value of utilizing a ML framework is leveraging historical results and data to inform current manufacturing at a pace that significantly reduces defects from occurring in a real-time production environment,” said Stephen Nolet, director at TPI.  “Additionally, this technology allows users to create alternative manufacturing scenarios to increase production velocity in manufacturing operations while simultaneously reducing infusion related problems.”

Plans are to develop the technology got larger components with greater manufacturing complexity using artificial intelligence (AI) tools to find patterns in historical data and predict outcomes on full-scale wind blade components, including blade shells, according to the company