ML-Driven Surface Structure Optimization of Aluminum Alloy Superhydrophobic Surfaces for Enhanced Anti-Icing...

Area of Science:

• Materials Science
• Surface Engineering
• Computational Science

Background:

• Surface icing presents significant risks to aerospace, transportation, and power systems.
• Superhydrophobic surfaces offer anti-icing properties, but their complex surface-performance relationship hinders optimization.
• Current experimental methods for evaluating anti-icing performance are inefficient and costly.

Purpose of the Study:

• To develop a machine learning (ML)-driven approach for optimizing the anti-icing performance of superhydrophobic surfaces.
• To identify and quantify the key surface features that govern anti-icing capabilities.
• To establish a theoretical framework for the rational design of advanced anti-icing surfaces.

Main Methods:

• Utilized the gray level co-occurrence matrix (GLCM) for multiscale surface feature characterization.
• Employed feature importance analysis to determine critical parameters influencing anti-icing performance.
• Integrated key features with classical nucleation theory to build an optimization model.

Main Results:

• Successfully predicted the anti-icing performance of aluminum-based superhydrophobic surfaces.
• Identified crucial surface features that control anti-icing effectiveness.
• Quantified the synergistic effects of these features on freezing delay time.

Conclusions:

• The proposed ML-driven method effectively optimizes superhydrophobic anti-icing surfaces.
• The developed model provides a theoretical basis for designing surfaces with enhanced anti-icing properties.
• This approach offers a more efficient alternative to traditional experimental methods for anti-icing research.