About this Abstract |
| Meeting |
MS&T26: Materials Science & Technology
|
| Symposium
|
Advanced Coatings for Wear and Corrosion Protection
|
| Presentation Title |
Sensing, Predicting and Validating Coating Lifetime Performance via Electrochemical Testing and Machine Learning |
| Author(s) |
Rebecca Crow, Homero Castaneda, Rishi Gupta, Andy Nowasielski, Ulises Martin Diaz, Victor Ponce, Shaik Marjuban |
| On-Site Speaker (Planned) |
Homero Castaneda |
| Abstract Scope |
This work characterizes and quantifies barrier and hybrid coatings under steady-state and accelerated corrosion and integrates the data into machine learning models to predict coating lifetime based on degradation modes. Coating–substrate interfaces undergo transport (electrolyte uptake) and interfacial (corrosion activation) processes, analyzed using Electrochemical Impedance Spectroscopy (EIS), surface characterization, and digital imaging. Long-term experiments (over two years) with regularly spaced EIS measurements defined the evolving coating condition under controlled exposure. Accelerated testing included wet–dry cycles and aggressive environments simulating weathering and roadway conditions. To replicate worst-case scenarios, coatings were mechanically damaged with scratches penetrating 50% and 100% of thickness. Experimental data supported a combined deterministic and machine-learning framework capable of predicting coating behavior, including water uptake and corrosion, over more than 10 years. Model predictions were validated against experimental results for up to three years, demonstrating reliable extrapolation from steady state and accelerated testing to long-term performance. |