AI for Corrosion Prediction

This project focuses on building machine learning models to predict the corrosion performance of coated steel plates using data such as weight change, capacitance (electrochemical measurements), and adhesion strength. It establishes a framework to scale from simple lab experiments to predictive corrosion modeling.

Goals

• Validate whether wet-dry saltwater cycling causes measurable degradation.
• Scale the experiment for broader corrosion performance prediction.

Methodology

• Single coated steel plate experiment for preliminary testing.
• Measure weight change, capacitance, and adhesion over multiple cycles.
• Analyze data using statistical and ML techniques to predict performance.

Expected Outcome

• Detectable corrosion (≥0.2g weight loss) validated.
• Framework established for full-scale machine learning experiments.