Determination of the Bentonite Content in Molding Sands Using AI-Enhanced Electrical Impedance Spectroscopy

Area of Science:

• Materials Science
• Chemical Engineering
• Industrial Process Control

Background:

• Foundry molding sand mixtures require precise control for optimal casting and efficient recycling.
• Current in-line monitoring methods for these complex mixtures are insufficient.
• Bentonite content is a critical parameter affecting sand performance and recyclability.

Purpose of the Study:

• To investigate the feasibility of an AI-enhanced electrical impedance spectroscopy (EIS) system for in-line monitoring of molding sand.
• To develop a predictive model for estimating bentonite content in foundry sand mixtures.
• To assess the potential for optimizing foundry processes and sand recycling.

Main Methods:

• Characterization of quartz sand, bentonite, and water mixtures using EIS (20 Hz–1 MHz).
• Measurement of sample water content and density.
• Application of Principal Component Analysis (PCA) for feature extraction from EIS data.
• Training fully connected neural networks with PCA features, water content, and density to predict bentonite content.

Main Results:

• High prediction accuracy (R ² = 0.94) for bentonite content estimation.
• Successful feature extraction from EIS data using PCA.
• Demonstration of AI-enhanced EIS capability in laboratory settings.
• Correlation established between EIS data, physical properties, and bentonite concentration.

Conclusions:

• AI-enhanced EIS is a viable technology for in-line monitoring of bulk materials in the foundry industry.
• The developed system offers a pathway to optimized process control and enhanced sand recycling efficiency.
• This approach provides a novel solution for real-time quality assessment of molding sand mixtures.