Digital Twin-Driven Optimization of Pilot-Scale Polyurethane Aerogel Production Using SVR Modelling

Area of Science:

• Materials Science
• Chemical Engineering
• Computational Modeling

Background:

• Aerogels, particularly polyurethane (PU) aerogels, are highly sought after for sustainable and energy-efficient insulation due to their thermal properties and mechanical versatility.
• Traditional PU aerogel development relies on time-consuming and resource-intensive trial-and-error experimentation.
• There is a need for efficient methodologies to accelerate the design and optimization of PU aerogels.

Purpose of the Study:

• To present a Digital Twin (DT) framework for supporting the design of PU aerogels.
• To reduce the experimental workload in PU aerogel synthesis and development.
• To demonstrate a data-driven approach for advancing aerogel manufacturing.

Main Methods:

• Development of a pilot-scale Digital Twin (DT) framework using data from 21 synthesis experiments.
• Evaluation of two predictive models, selecting Support Vector Regression (SVR) for its accuracy (R² = 0.964).
• Utilizing the DT to map process parameters, analyze the synthesis, and estimate aerogel density.

Main Results:

• The Support Vector Regression (SVR) model accurately predicted PU aerogel density (R² = 0.964).
• The DT framework successfully identified synthesis conditions linked to lower aerogel density, potentially enhancing thermal insulation.
• The study validated the efficacy of DT-assisted modeling in material development.

Conclusions:

• Digital Twin frameworks offer a powerful tool for optimizing material design and improving process understanding in PU aerogel synthesis.
• This data-driven approach facilitates more efficient experimentation and guides the development of sustainable, scalable aerogel manufacturing.
• The developed DT framework shows significant potential for accelerating innovation in advanced insulation materials.