PHYSICS-INFORMED MACHINE LEARNING FOR PREDICTING OPTICAL TRANSMITTANCE AND HAZE IN TRANSPARENT WOOD

Transparent wood (TW) has emerged as one of the most promising sustainable materials for advanced optical and energy-related applications. By combining the hierarchical structure of wood with transparent polymers, TW exhibits a unique balance of optical transmittance, light diffusion, low density, mechanical robustness, and thermal insulation. These characteristics make it attractive for smart windows, solar energy harvesting systems, energy-efficient buildings, and optoelectronic devices. Despite significant advances in TW fabrication, optimization of optical properties is still largely based on empirical approaches. Optical transmittance and haze depend on a complex combination of variables including sample thickness, refractive index matching, residual lignin content, infiltration quality, and microstructural heterogeneity. As a result, establishing predictive structure–property relationships remains challenging.

Read the full article here: Nuvoli, D., Bifulco, A., Pugliese, D., Casciello, A., Terzi, M., Mariani, A., & Malucelli, G. (2026, September 11). PHYSICS-INFORMED MACHINE LEARNING FOR PREDICTING OPTICAL TRANSMITTANCE AND HAZE IN TRANSPARENT WOOD. XXVI National Congress of AIM - Italian Macromolecular Association (AIM 2026), Torino. https://doi.org/10.5281/zenodo.22705940