Machine Learning Inversion of Layer-Wise Plasticity and Interfacial Cohesive Parameters in Multilayer Thin Films.

Area of Science:

• Materials Science
• Computational Materials Science
• Machine Learning Applications

Background:

• Traditional material parameter inversion for thin films is time-consuming and inefficient.
• Multilayer thin-film structures require accurate characterization of layer-wise properties and interfaces.

Purpose of the Study:

• To develop a fast and efficient machine learning-based method for evaluating material parameters in multilayer thin films.
• To overcome the limitations of traditional parameter inversion techniques.

Main Methods:

• Conducting nanoindentation experiments to gather data for multilayer thin films.
• Building a 2D elasto-plastic model to generate load-depth curves for machine learning training.
• Training and validating a machine learning model using simulated and experimental data.

Main Results:

• The machine learning model accurately identifies layer-wise plastic parameters and interfacial cohesive properties.
• Achieved high accuracy in inverting interlayer cohesion parameters with a correlation coefficient R² ≥ 0.99.
• Reduced parameter estimation time to under 1 hour for batch analysis, significantly improving efficiency.

Conclusions:

• The proposed method offers a highly accurate and efficient alternative for material parameter evaluation in multilayer thin films.
• This approach provides a valuable tool for performance evaluation and optimization design of thin-film materials.
• The method demonstrates broad applicability and potential for advancing thin-film material characterization.